Advances in Strategic Management Strategy Beyond Markets PDF

STRATEGY BEYOND MARKETS
ADVANCES IN STRATEGIC
MANAGEMENT
Series Editor: Brian S. Silverman
Recent Volumes:
Volume 23: Ecology and Strategy
Edited by: Joel A. C. Baum, Stanislav D. Dobrey and
Arien van Witteloostuijn
Volume 24: Real Options Theory
Edited by: Jeffrey J. Reuer and Tony W. Tong
Volume 25: Network Strategy
Edited by: Joel A. C. Baum and Tim J. Rowley
Volume 26: Economic Institutions of Strategy
Edited by: Jackson A. Nickerson and Brian S. Silverman
Volume 27: Globalization of Strategy Research
Edited by: Joel A.C. Baum and Joseph Lampel
Volume 28: Project-Based Organizing and Strategic Management
Edited by: Gino Cattani, Simone Ferriani, Lars Frederiksen
and Florian Taube
Volume 29: History and Strategy
Edited by: Steven J. Kahl, Brian S. Silverman and
Michael A. Cusumano
Volume 30: Collaboration and Competition in Business Ecosystems
Edited by: Ron Adner, Joanne E. Oxley and
Brian S. Silverman
Volume 31: Finance and Strategy
Edited by: Belen Villalonga
Volume 32: Cognition and Strategy
Edited by: Giovanni Gavetti and William Ocasio
Volume 33: Business Models and Modelling
Edited by: Charles Baden-Fuller and Vincent Mangematin
ADVANCES IN STRATEGIC MANAGEMENT VOLUME 34
STRATEGY
BEYOND MARKETS
EDITED BY
JOHN M. DE FIGUEIREDO
Duke University, Durham, NC, USA
MICHAEL LENOX
University of Virginia, Charlottesville, VA, USA
FELIX OBERHOLZER-GEE
Harvard Business School, Boston, MA, USA
RICHARD G. VANDEN BERGH

University of Vermont, Burlington, VT, USA
United Kingdom North America Japan

India Malaysia China
Emerald Group Publishing Limited
Howard House, Wagon Lane, Bingley BD16 1WA, UK
First edition 2016
Copyright r 2016 Emerald Group Publishing Limited
Reprints and permissions service

Contact: permissions@emeraldinsight.com
No part of this book may be reproduced, stored in a retrieval system, transmitted in

any form or by any means electronic, mechanical, photocopying, recording or
otherwise without either the prior written permission of the publisher or a licence
permitting restricted copying issued in the UK by The Copyright Licensing Agency
and in the USA by The Copyright Clearance Center. Any opinions expressed in the
chapters are those of the authors. Whilst Emerald makes every effort to ensure the
quality and accuracy of its content, Emerald makes no representation implied or
otherwise, as to the chapters suitability and application and disclaims any warranties,
express or implied, to their use.
British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library
ISBN: 978-1-78635-020-6
ISSN: 0742-3322 (Series)
ISOQAR certified
Management System,
awarded to Emerald
for adherence to
Environmental
standard
ISO 14001:2004.
Certificate Number 1985

ISO 14001
CONTENTS
LIST OF CONTRIBUTORS ix
INTRODUCTION xiii
STRATEGY BEYOND MARKETS: A STEP BACK

AND A LOOK FORWARD
David P. Baron 1
PART I
PUBLIC POLITICS
POLITICAL RISK AS A HOLD-UP PROBLEM:

IMPLICATIONS FOR INTEGRATED STRATEGY
Kenneth W. Shotts 57
INCORPORATING LEGISLATIVE EFFECTIVENESS

INTO NONMARKET STRATEGY: THE CASE OF
FINANCIAL SERVICES REFORM AND THE
GREAT RECESSION
Craig Volden and Alan E. Wiseman 87
A UNIFIED MODEL OF POLITICAL RISK

Benjamin A. T. Graham, Noel P. Johnston and 119
Allison F. Kingsley
MOTIVATIONS FOR CORPORATE

POLITICAL ACTIVITY
Adam Fremeth, Brian Kelleher Richter and 161
Brandon Schaufele
v
vi CONTENTS
THE MARKET FOR LEGISLATIVE INFLUENCE

OVER REGULATORY POLICY
Rui J. P. de Figueiredo Jr. and Geoff Edwards 193
PART II
PRIVATE POLITICS
CORPORATE REPUTATIONAL DYNAMICS, PRIVATE

REGULATION, AND ACTIVIST PRESSURE
Jose Miguel Abito, David Besanko and Daniel Diermeier 235
SELF-REGULATION AND REGULATORY

DISCRETION: WHY FIRMS MAY BE RELUCTANT
TO SIGNAL GREEN
Thomas P. Lyon and John W. Maxwell 301
PRIVATE POLITICS DAILY: WHAT MAKES FIRMS

THE TARGET OF INTERNET/MEDIA CRITICISM? AN
EMPIRICAL INVESTIGATION OF FIRM, INDUSTRY,
AND INSTITUTIONAL FACTORS
Dominik Breitinger and Jean-Philippe Bonardi 331
PART III
INTEGRATED POLITICAL STRATEGY
NAVIGATING NATURAL MONOPOLIES: MARKET

STRATEGY AND NONMARKET CHALLENGES IN
RADIO AND TELEVISION AUDIENCE
MEASUREMENT MARKETS
Hillary Greene and Dennis A. Yao 367
THE ORGANIZATION OF NONMARKET STRATEGY

Dylan Minor 413
COMPLEMENTARITY IN FIRMS MARKET AND

POLITICAL CAPABILITIES: AN INTEGRATED
THEORETICAL PERSPECTIVE
Nan Jia and Kyle Mayer 437
Contents vii
HOW PATENT STRATEGY AFFECTS THE

TIMING AND METHOD OF PATENT
LITIGATION RESOLUTION
Deepak Somaya 471
ABOUT THE AUTHORS 505

This page intentionally left blank
LIST OF CONTRIBUTORS
Jose Miguel Abito Department of Business Economics and

Public Policy, Wharton School of the
University of Pennsylvania, Philadelphia,
PA, USA
David P. Baron Stanford Graduate School of Business,
Stanford University, Stanford, CA, USA
David Besanko Kellogg School of Management,
Northwestern University, Evanston,
IL, USA
Jean-Philippe Bonardi HEC Lausanne, University of Lausanne,
Lausanne, Switzerland
Dominik Breitinger Finance & Capital Markets, World
Business Council for Sustainable
Development (WBCSD), Geneva,
Switzerland
John M. de Figueiredo The Law School and Fuqua School of
Business, Duke University, Durham,
NC, USA
Rui J. P. de Haas School of Business and Department
Figueiredo, Jr. of Political Science, University of
California at Berkeley, Berkeley, CA, USA
Daniel Diermeier Harris School of Public Policy, University
of Chicago, Chicago, IL, USA
Geoff Edwards European and Asia-Pacific Competition
Practices, Charles River Associates,
London, UK
Adam Fremeth Ivey Business School, University of
Western Ontario, London, Canada
ix
x LIST OF CONTRIBUTORS
Benjamin A. T. School of International Relations,

Graham University of Southern California,
Los Angeles, CA, USA
Hillary Greene School of Law, University of Connecticut
Law School, Hartford, CT, USA; Harvard
Law School, Harvard University,
Cambridge, MA, USA
Nan Jia Marshall School of Business, University of
Southern California, Los Angeles,
CA, USA
Noel P. Johnston Blavatnik School of Government, Oxford
University, Oxford, UK
Allison F. Kingsley Grossman School of Business, University
of Vermont, Burlington, VT, USA
Michael Lenox Darden School of Business, University of
Virginia, Charlottesville, VA, USA
Thomas P. Lyon Ross School of Business and the School of
Natural Resources and Environment,
University of Michigan, Ann Arbor,
MI, USA
John W. Maxwell Kelley School of Business, Indiana
University, Bloomington, IN, USA
Kyle Mayer Marshall School of Business, University of
Southern California, Los Angeles,
CA, USA
Dylan Minor Department of Managerial Economics and
Decision Sciences, Kellogg School of
Management, Northwestern University,
Evanston, IL, USA; Harvard Business
School, Harvard University, Boston,
MA, USA
Felix Oberholzer-Gee Harvard Business School, Harvard
University, Boston, MA, USA
List of Contributors xi
Brian Kelleher Richter McCombs School of Business, University

of Texas at Austin, Austin, TX, USA
Brandon Schaufele Ivey Business School, University of
Western Ontario, London, Canada
Kenneth W. Shotts Stanford Graduate School of Business,
Stanford University, Stanford, CA, USA
Deepak Somaya College of Business, University of Illinois
at Urbana-Champaign, Champaign,
IL, USA
Richard G. School of Business Administration,
Vanden Bergh University of Vermont, Burlington,
VT, USA
Craig Volden Frank Batten School of Leadership and
Public Policy and the Woodrow Wilson
Department of Politics, University of
Virginia, Charlottesville, VA, USA
Alan E. Wiseman Department of Political Science and Law,
Vanderbilt University, Nashville, TN, USA
Dennis A. Yao Harvard Business School, Harvard
University, Boston, MA, USA
INTRODUCTION
Traditional strategy scholarship is fundamentally about understanding the

choices firms make in markets to create and capture long run value.
Scholars in this space study the actions and choices of firms by focusing on
firm interactions with or responses to key market rivals and/or market-
based stakeholders such as upstream suppliers, downstream distributors,
and key employees. Strategy Beyond Markets research is also interested in
how the actions of firms affect long run value. However, the scholars in the
Strategy Beyond Markets field focus their attention on choices emanating
from firm interactions with or responses to entities beyond their primary
market stakeholders such as government regulators, community organiza-
tions, unions, advocacy groups, and policy makers. These Beyond Market
stakeholders are numerous and each have their own wants, demands, or
requirements of firms.
Typically Beyond Market stakeholders want firms to change the pro-
cesses employed to implement firm market strategy or they desire resources
from the firm to help meet their own objectives. For example, in the com-
munities where firms are located, community leaders might ask firms to
invest resources to improve education for local citizens or to help improve
basic infrastructure. Activist environmental groups might boycott firms
that do not redesign their supply chain to reduce toxic release into water-
ways. International nongovernmental organizations may protest firms that
do not eliminate the use of resources acquired from locations where the
NGOs observe violations of human rights. Trade associations may seek
protective quotas on important raw material inputs for their member firms.
Government regulators may attempt to require firms to improve safety for
manufacturing employees or to invest in the best available technology to
reduce air pollution emissions from production processes. Elected politi-
cians may try to pass new laws that increase minimum wages paid to firm
employees or laws that restrict entry by firms into industries. Courts may
restrict the funding amounts and methods used by firms to interact with
elected officials or rule on the protection of property rights affecting
the firms.
xiii
xiv INTRODUCTION
In response to these pressures, firms design strategies to address the

wants, demands and requirements of Beyond Market stakeholders. That is
the focus of this special issue of the Advances in Strategy Management. We
organize the papers in this special issue around three central themes: Public
Politics, Private Politics, and Integrated Political Strategy. Public Politics
research is concerned primarily with the strategies employed by firms to
address Beyond Market stakeholders whose influence reflects their position
in formal political institutions, namely public policy makers. Thus, Public
Politics is concerned with the interactions between firms and government
officials that have the authority to pass new legislation, promulgate regula-
tory rules or issue statutory interpretations (inter alia). By comparison
Private Politics focuses on firm strategies to manage Beyond Market stake-
holders that act outside of the public policy-making process. In the Private
Politics field scholars explore strategies employed by firms interacting with
special interest groups like activist environmental groups or international
NGOs. Integrated Political Strategy scholars seek to identify comprehen-
sive strategies that bridge Public Politics and/or Private Politics with
market-based competitive strategy.
It is probably accurate to say that since the advent of markets and
firms, various Beyond Market stakeholders have strived to change the
behavior of firms. In the 1960s, firms in the United States experienced
the proliferation of new public policies to improve employee health and
safety, the environment, and consumer protection by mandating
changes in firm behavior. These gave rise to the desire for more effec-
tive Public Politics as many firms realized that they could not simply be
passive recipients of government regulation if they wished to thrive in
the competitive marketplace. Many companies hired large legal staffs,
engaged in lobbying, and adopted environmental, health, safety and
compliance units.
Around the same period, many firms observed an increase in interest
group activity to pressure them directly to address social and environmen-
tal problems voluntarily rather than through traditional public policy
avenues. In their wake, Private Politics emerged. Standard setting efforts,
trade association codes of conduct, community outreach, and public rela-
tions became common place. Earlier adversarial relationships between acti-
vists and firms were complemented by collaborative efforts and other
attempts to mutually create value. More sophisticated firms began to recog-
nize the value of Integrated Political Strategies that leveraged public and
private politics and traditional market-based strategies.
Introduction xv
Whether challenged through public policy or private pressure, managers

see the effects on long-term profitability and have learned the importance
of Strategy Beyond Markets as a component of firm strategy. The extent to
which firms craft strategy to actively address Beyond Market stakeholders
is increasing in importance. In a recent survey, McKinsey reported that
over 95% of CEOs believe that society at large has increasing expectations
that business must address and that business should be responsive to the
wants, demands and/or requirements of Beyond Market stakeholders.1
Additionally, nearly 60% of executives are explicitly focused on building
and maintaining relationships with Beyond Market stakeholders.2 Firms
recognize that they must develop resources and capabilities to effectively
manage Beyond Market stakeholders.
Executive recognition of the importance of Strategy Beyond Markets is a
global phenomenon. The World Bank asked CEOs to rank the most impor-
tant constraints for their firms when deciding to invest. The executives
ranked the uncertainty of public policy as the second most significant
constraint, ahead of more traditional market-based constraints such as
identifying market opportunities, raising financial capital and attracting
human capital.3 Furthermore, each year the Conference Board asks CEOs
to discuss the biggest challenges their organizations face to build the cap-
abilities necessary to meet long-term objectives. Recent results from several
annual Conference Board surveys of nearly 1,000 CEOs from across the
globe reveal that Private Politics is growing in importance. By 2015, the
survey showed that Private Politics (e.g., engaging with local community or
engaging with special interest groups) was among the most important chal-
lenges CEOs face.4 In a report entitled, A marketplace without bound-
aries: Responding to disruption, PwC discusses the results of their 18th
annual global CEO survey. In one question, PwC asked CEOs How con-
cerned are you about the following potential threats to your organiza-
tions growth prospects? The number one threat identified in the survey
was the threat of government regulation. Nearly 80% of the CEOs sur-
veyed identified regulation, and thus the importance of Public Politics, as a
potential threat compared to about 55% of the CEOs identifying either
new market entry or technological change as major threats. In a related
question, CEOs viewed regulation as a greater potential disrupter to firm
success than market threats such as increased competition, changes in con-
sumer preferences, or supply chain changes. Interestingly, the view that
regulation is a significant threat is shared by CEOs not only in highly
regulated industries such as energy and pharmaceuticals, but also in
xvi INTRODUCTION
traditionally less regulated industries such as entertainment and retail.5 It is

apparent from these recent surveys that executives view Strategy Beyond
Markets as important as traditional competitive market strategy, and
perhaps more important in some situations.
As David P. Baron points out in the opening paper in this volume, the
academic interest in Strategy Beyond Markets has grown significantly over
the past couple decades. However, despite this growing interest, the impor-
tance of the field relative to other areas of strategic management does not
align with the views of executives. As illustrated above, executives view
Strategy Beyond Markets as fundamental, while the community of strategic
management scholars views it largely as a niche.
We base this assessment on a simple analysis of scholarly work pub-
lished in the Strategic Management Journal (SMJ) and of research papers
presented at Strategic Management Society conferences (SMS). SMJ is
widely regarded as the premier generalist journal in strategic management,
and SMS is similarly viewed as the premier generalist conference. We
reviewed all published papers in the SMJ (Volumes 33 through 35) and
paper presentations at the SMS6 over a three-year period from 2012
through 2014. For each article or research note published in the SMJ,
we read through the abstracts and identified if the article fit within the
Strategy Beyond Markets field. To be categorized as a Strategy Beyond
Markets article, the focus of the research had to be on the strategic interac-
tion between a firm and their Beyond Market stakeholders, or on the
strategies designed to manage these stakeholders. We erred on the side of
inclusion. For example, we included papers that addressed who, when,
why, or how to interact with Beyond Market stakeholders, and papers that
analyze the development of capabilities/resources to respond to these types
of stakeholders, or papers that included measures of strategic action or
firm performance on some relevant dimension (e.g., net cost of action;
protests; boycotts, campaign contributions, etc.). We used the same proce-
dure to identify Strategy Beyond Market papers presented during the SMS
conferences.
We reviewed 276 papers published in SMJ and 1,984 paper presentations
at SMS. The results of our simple analysis are compelling. Less than 10%
of the papers published in SMJ and 3% of papers presented at SMS were
categorized as addressing Strategy Beyond Markets. Within the set of SMJ
papers, we also determined the theme in which the research fit. Once again,
relative to the weight that executives place on policy uncertainty and gov-
ernment regulation, the fact that approximately 2% of the papers fit into
Public Politics is quite striking. We recognize that there are other excellent
Introduction xvii
journals that publish strategic management scholarship (e.g., Management

Science, Organization Science, Academy of Management Journal, etc.) and
that our analysis is suggestive but clearly incomplete. That said, an infor-
mal preliminary perusal of these other journals (and of papers presented in
the BPS division of the Academy of Management Conference) revealed
similar information.7
One goal of this volume is to create a launch pad for future scholars in
this field. From the analysis above, we cant determine the causes of the
fields relatively niche status, but we can self-reflect. Paraphrasing Chester
Barnard, if we wish to increase the importance of Strategy Beyond Markets
in the broader field of strategic management, then we need to identify what
the field lacks, and the set of factors limiting its advancement.8 In our
opinion, there are several important unanswered questions in the field and
several fruitful avenues for research.
First, we currently have limited understanding of the critical attributes
for firms to establish a sustainable Beyond Market strategy. While there
have been excellent studies highlighting the importance of preemption in
Private Politics, political ties in Public Politics, and forum shopping in
Integrated Political Strategy, it is unclear if these types of factors are sus-
tainable in the context of political or market dynamics. Furthermore, it is
unclear whether these factors differ significantly from those identified in the
broader competitive strategy literature (e.g., are political resources different
fundamentally from market resources; do firms organizing political
resources use a fundamentally different logic than firms utilize to organize
market resources).
Second, there has been very little (if any) significant research linking
firm strategy to both Beyond Market outcomes and firm performance.
Most research has developed theories and/or empirical analyses that
explore the determinants of a firms strategy beyond markets. Very few stu-
dies have shown that a firms strategy increases policy performance; almost
none have demonstrated a link between policy and firm profitability.9
Third, the topical areas of analysis appear to be limited. Environmental
issues and corporate social responsibility dominate Private Politics
research, heavily regulated firms dominate Public Politics studies, and
theoretical studies dominate research on Integrated Political Strategy. We
believe the papers in this volume answer some of these questions and may
help launch Strategy Beyond Markets into the mainstream of strategic
management scholarship.
The founder of the field of nonmarket and integrated strategy, David
P. Baron, has contributed an excellent overview of the field in his paper,
xviii INTRODUCTION
Strategy beyond Markets: A Step Back and a Look Forward. Baron has
a unique perspective, having been in the field since its inception and having
written the seminal textbook and numerous teaching cases, and continuing
to be a contributor to understanding the strategy of the firm beyond
markets. In this paper, Baron discusses the creation of the field, highlight-
ing the prominence of the political economy approach as compared to the
resource-based theory of the firm. He argues in this paper that this political
economy approach to self-regulation by firms can inform managers on how
to reduce challenges from public and private politics. For scholars, man-
agers, and students, Baron illustrates these points with three very interest-
ing case studies: Uber and public politics; the Rainforest Action Network
and Citigroup and private politics; and The Fast Food Campaign and inte-
grated strategy with both public and private politics. This paper provides a
very nice perspective on this volume as a whole.
PUBLIC POLITICS ARTICLES

One of the central questions in the literature on foreign direct investment is
how to mitigate the threat of government expropriation. Scholars have
developed a variety of theories to address this question. Kenneth W.
Shotts, in his paper, Political Risk as a Hold-Up Problem: Implications
for Integrated Strategy, encompasses many of these theories in a single,
simple, elegant hold-up model of political risk. The model is particularly
useful for three purposes. First, its breadth allows a researcher to examine
the comparative statics of many different theories in a single model.
Second, its simplicity makes the model a useful teaching tool in graduate
level courses. Finally, and perhaps most importantly, the model demon-
strates that effective management of political risk requires an integrated
strategy, consisting not only of public and government relations efforts,
but also financial, value chain, and human resources strategies designed to
reduce the governments incentives for expropriation. Overall, the paper
pushes the frontier on many fronts.
Traditional models of political strategy examine how interest groups tar-
get marginal politicians in a pivotal politics, spatial model. However, politi-
cians vary in the effectiveness at pushing policy through the legislature.
Craig Volden and Alan E. Wiseman, in their paper Incorporating
Legislative Effectiveness into Nonmarket Strategy: The Case of Financial
Services Reform and the Great Recession, develop a new and innovate
Introduction xix
empirical strategy for identifying exactly who are the politicians who are
most effective at advancing a legislative agenda. They develop an empirical
method for assigning each legislator a Legislative Effectiveness Score based
on the ability of the legislator to move bills through Congress. They then
demonstrate how corporations can use these scores in targeting their lobby-
ing, with a vibrant and detailed case study of financial reform legislation
before the U.S. Congress during the Great Recession.
In A Unified Model of Political Risk, Benjamin A. T. Graham, Noel
P. Johnston, and Allison F. Kingsley make an important contribution to
the literature on expropriation risks of (foreign) investors. Looking at indi-
vidual risks, the authors argue, scholars often miss important interdepen-
dencies. To provide a fuller picture, they construct a unified model that
analyzes different types of risk (violence, outright expropriation, capital
controls), borne by different groups of investors (direct investors, equity
and debt investors, commercial banks) who vary in their capabilities (access
to private information, ease of exit, resistance). The authors show how
specific risks vary with the distribution of investor types in the economy
and with the specific capabilities of these investors. Perhaps surprisingly,
for some groups of investors and model parameters, there is safety in small
numbers. Bank lending, for example, is less risky when direct investors
provide most of the capital.
In Motivations for Corporate Political Activity, Adam Fremeth,
Brian Kelleher Richter, and Brandon Schaufele ask whether a firms CEO
and/or PAC make campaign contributions that are motivated by strategic
or agency concerns. Strategic contributions are those that are expected to
help the firm meet its objectives. Agency contributions are made without
regard for the firms objective but rather reflect private benefits of contribu-
tions for the CEO (or PAC although the latter usually has a mission
aligned with firm interests). This is an important question for the Public
Politics literature because much of the literature finds inconsistent results
regarding the efficacy of campaign contributions for improving firm perfor-
mance. Scholars have conjectured that the lack of a consistent performance
link can be attributed to the fundamental principal/agent problem. That is
CEOs and/or PACs may have interests that are not aligned with the inter-
ests of firm shareholders and make contributions that reflect their private
benefits at the expense of their principals (firm shareholders) interests.
The authors develop an argument and testable hypotheses that strategic
and agency contributions occur simultaneously. To test their hypotheses,
they construct a unique dataset of nearly 7 million observations where the
unit of analysis is a pairing between a S&P 500 firm-linked actor (CEOs
xx INTRODUCTION
and/or PACs) and a political actor in the general election for office in the
U.S. Congress. They measure the campaign contributions associated with
these pairs over nine election cycles. They find evidence that agency motiva-
tions are apparent within PACs contributions, and that strategy is appar-
ent within CEOs personal contributions, and that elements of both agency
and strategy exist side-by-side within the actors contributions.
Rui J. P. de Figueiredo, Jr. and Geoff Edwards also address the open
question in the Public Politics literature emanating from inconsistent
empirical findings about the effect of campaign contributions on public
policy outcomes. In their paper The Market for Legislative Influence over
Regulatory Policy they argue that scholars may be overlooking an impor-
tant mechanism for campaign contributions to affect policy. Instead of
contributions of just buying support for legislation or access to legislators,
the firm may strategically allocate campaign funds to gain support from
legislators to bring pressure to bear on regulatory agencies. In their study,
de Figueiredo and Edwards provide empirical support for extant theoreti-
cal arguments that firms influence regulatory policy through indirect pres-
sure from elected officials. Additionally, they extend prior findings by
developing arguments and empirical support for when firms are likely to
respond strategically to interest group competition. The authors find
empirical support for their arguments by analyzing a dataset of campaign
contributions by competing interest groups to state level legislators in
the context of regulatory policy governing the local telecommunica-
tions industry.
PRIVATE POLITICS ARTICLES

In Corporate Reputational Dynamics, Private Regulation, and Activist
Pressure, Jose Miguel Abito, David Besanko, and Daniel Diermeier model
the interaction between firms and activists who may adopt either confron-
tational campaigns or more benign forms of criticism. To try to placate or
deflect activists, the firms may attempt to enhance their reputations
through self-regulation the voluntary effort to reduce a negative external-
ity. They find that, while a firm may face decreasing marginal utility from
externality reducing activity thus incentivizing a firm to coast on its reputa-
tion, an activist can place reputational pressures on a firm that keeps the
firm incentivized to continue to self-regulate and build its reputation.
Furthermore, they present conditions under which an activist is more likely
Introduction xxi
to rely on confrontation versus criticism and serves as a nice complement

to the Dominik Breitinger and Jean-Philippe Bonardi paper discussed below.
In Self-Regulation and Regulatory Discretion: Why Firms May Be
Reluctant to Signal Green, Thomas P. Lyon and John W. Maxwell model
the decision to self-regulate, not as a reaction to the potential actions of
activists, but to regulatory agencies. They observe that regulators maintain
enforcement discretion thus providing incentives to individual firms to self-
regulate to reduce the likelihood of fines and other punishments. They find
that self-regulation can be a doubled edged sword helping preempt
further regulation on one hand, while causing a ratcheting of expectations
among regulators who demand higher levels of compliance from greener
firms. Specifically, firms with lower compliance costs could step forward
and substantially self-regulate helping reduce the likelihood of future regu-
lation but increasing the risk of being targeted for stringent enforcement
should regulation pass.
In Private Politics Daily: What Makes Firms the Target of Internet/
Media Criticism? An Empirical Investigation of Firm, Industry, and
Institutional Factors, Breitinger and Bonardi explore what drives activists
to criticize a firm through social media and internet posts. Most research to
date on the private politics of activists focus on large campaigns such as
protests and boycotts. In this paper, the authors leverage a unique dataset
to look at a broad range of factors such as industry, country, and firm-
specific dimensions that drive activists to criticize on the internet. What
they find is that the motivations and targeting behavior of activists in social
media forums is very similar to what has been found for larger campaigns.
Large, visible firms from countries with strong institutions and high
standards of living are the most likely to be targeted. More importantly,
they present evidence that criticism through the media may be the first step
toward a broader public campaign against a firm.
INTEGRATED POLITICAL STRATEGY ARTICLES

In Navigating Natural Monopolies: Market Strategy and Nonmarket
Challenges in Radio and Television Audience Measurement Markets,
Hillary Greene and Dennis A. Yao chronicle the history of the audience
measurement industry in the United States and compare it to strategies and
outcomes in foreign markets. The market for radio and television audience
ratings tends toward monopoly as a result of economies of scale, the value
xxii INTRODUCTION
of stable ratings standards, uncertainty about the quality of ratings, and

buyers who mostly care about their relative performance in the ratings.
Greene and Yao argue that firms predominantly rely on market strategies
such as aggressive pricing and acquisitions during a first phase of market
development when monopoly positions are being established. When
faced with dominant firms, buyers switch from exit market-based
negotiations to voice strategies beyond the market that engage poli-
tical stakeholders, witness the frequent congressional hearings, and private
interest groups. In Greenes and Yaos analysis, the relative attractiveness
of exit and voice directly reflect the economics of the industry and the
resulting market structure.
One of the outstanding questions in nonmarket strategy is how should
the firm organize these nonmarket activities through a specialized unit
(government relations) or through integration and infusion within each
division of the firm. Dylan Minor develops a nice model to gain traction on
this question in his paper, The Organization of Nonmarket Strategy.
Minor argues that the advantage of integration of the activities over the
specialization of these nonmarket activities is U-shaped in the importance
of corporate social responsibility (CSR) to output. That is, if CSR is very
unimportant or very important to the firms output, then integration of the
CSR activity into the firms production activities is preferred; where CSR is
moderately important to the output of the firm, then the CSR activity
should be put into a corporate-wide, specialized unit. Along the way,
Minor provides us with a typology of organizational forms engaged in
CSR. He provides suggestive evidence of his theory from companies listed
in the Dow Jones Sustainability Index.
Firms resources and capabilities play a major role both in market com-
petition and in the political arena. As a rule, prior research analyzed these
assets in isolation. In Complementarity in Firms Market and Political
Capabilities: An Integrated Theoretical Perspective, Nan Jia and Kyle
Mayer argue, however, that there are important spillovers from one set of
capabilities to the other. For example, the authors contend that firms with
deeper technological expertise will find it easier to provide lawmakers with
pertinent regulatory information. Jia and Mayer focus on two core
capabilities the ability to shape political processes and outcomes through
the provision of information and the building of large constituencies and
discuss how these capabilities are related to R&D, marketing expertise,
the firms ability to absorb outside knowledge, and the institutional
Introduction xxiii
environment. Jia and Mayer stress the need to align market capabilities
and political tactics.
Deepak Somaya explores the Integrated Political Strategy of firms in the
computer sector and the research medicines sector in How Patent Strategy
Affects the Timing and Method of Patent Litigation Resolution. He argues
that firms time their Public Politics actions (decision to enter patent litiga-
tion) to align with one of three generic patent strategies firms utilize to create
competitive advantage in the product market. Some firms adopt proprietary
patent strategies, while others may adopt defensive strategies or leveraging
strategies. Since patent strategies employed by firms differ, Somaya argues
that the timing of patent litigation should also differ in a predictable manner
and he develops a set of testable hypotheses that suggests which type of
patent strategy in the product market is best integrated with the timing of
patent litigation in the political environment. He models the choice of firms
to either settle or adjudicate a litigation and when to do so. Empirically,
Somaya tests his propositions by employing a hazard rate model of patent
litigation termination on a sample of patent suits from the U.S. Federal
District Courts in the computer and research medicine industries. His find-
ings are supportive of the hypothesis that firms design an Integrated Political
Strategy by showing that proprietary, defensive, and leverage strategies are
important factors in how firms conduct their patent litigation activities.
In closing, while managers from all over the world consider Strategy
Beyond Markets a top priority for their firms, the academy has not yet
placed as great an emphasis on these important strategic questions. The
data presented in this introduction suggest that Strategy Beyond Markets
research is presented at major conferences and published in major journals
at relatively low rates. This special issue is designed to help address this
deficiency by publishing a set of papers by many preeminent scholars. We
believe the papers published herein represent innovative scholarship and
will provide researchers with new avenues for future inquiry into the
field avenues so that the academic research will catch up to the practice
of Strategy Beyond Markets.
John M. de Figueiredo
Michael Lenox
Felix Oberholzer-Gee
Richard G. Vanden Bergh
Editors
xxiv INTRODUCTION
NOTES
1. See CEOs on strategy and social issues, The McKinsey Quarterly,

October 2007.
2. For the results of the survey, see Sustainabilitys strategic worth: McKinsey
Global Survey results most recently accessed on July 2, 2015 at http://www.mckinsey.
com/insights/sustainability
3. See Multilateral Investment Guarantee Agency (2014).
4. See the Conference Board CEO Challenge Report 2015: Creating Opportunity
out of Adversity, Building Innovative, People-Driven Organizations. Report can be
downloaded from https://www.conference-board.org/ceo-challenge/ (most recently
accessed on July 2, 2015).
5. Report can be downloaded from http://www.pwc.com/gx/en/ceo-survey/2015/
index.jhtml (most recently accessed on July 2, 2015).
6. We did not include Volume 33(5) since this was a special issue dedicated to an
unrelated topics. In these three volumes, there was no special issue dedicated to
Strategy Beyond Markets. During these three years, the SMS conference was held in
Prague, Atlanta, and Madrid.
7. Strategy Beyond Markets research is also published in high-quality economics
journals (e.g., American Economic Review; Journal of Political Economy; Journal of
Law, Economics & Organization; Journal of Economics & Management Strategy)
and high-quality political science journal (e.g., American Political Science Review;
Journal of Politics). Strategic management scholars tend to focus on the firm as the
primary unit of analysis, but this is not so prevalent in economics or political
science. Still it would be interesting to examine how prevalent Strategy Beyond
Markets research is in the premier journals and at the premier conferences in these
two related and important disciplines.
8. See Barnard (1968).
9. The exception being does it pay to be green? Scholars have made many
attempts to quantify the effects of corporate environmental consciousness on firm
performance.
REFERENCES
Barnard, C. I. (1968). The functions of the executive (Vol. 11, pp. 204205). Cambridge, MA:
Harvard University Press.
Multilateral Investment Guarantee Agency. (2014). World investment and political risk 2013.
Washington, DC: World Bank Group. r World Bank. Retrieved from https://open-
knowledge.worldbank.org/handle/10986/16388. License: CC BY 3.0 IGO
STRATEGY BEYOND
MARKETS: A STEP BACK
AND A LOOK FORWARD
David P. Baron
ABSTRACT
This paper provides a perspective on the field of nonmarket strategy.

It does not attempt to survey the literature but instead focuses on the
substantive content of research in the field. The paper discusses the
origins of the field and the roles of nonmarket strategy. The political
economy framework is used and contrasted with the current form of the
resource-based theory. The paper argues that research should focus
on the firm level and argues that the strategy of self-regulation can be
useful in reducing the likelihood of challenges from private and public
politics. The political economy perspective is illustrated using three
examples: (1) public politics: Uber, (2) private politics: Rainforest
Action Network and Citigroup, and (3) integrated strategy and private
and public politics: The Fast Food Campaign. The paper concludes
with a discussion of research issues in theory, empirics, and norma-
tive assessment.
Keywords: Integrated strategy; political economy; private politics
Strategy Beyond Markets

Advances in Strategic Management, Volume 34, 154
Copyright r 2016 by Emerald Group Publishing Limited
All rights of reproduction in any form reserved
ISSN: 0742-3322/doi:10.1108/S0742-332220160000034002
1
2 DAVID P. BARON
INTRODUCTION
As an academic field strategy beyond markets, or nonmarket strategy, has
matured considerably during the past three decades, yet the field remains
somewhat disjointed and the remaining research agenda exceeds the accom-
plishments to date. This paper steps back and considers the events that
gave rise to the field of nonmarket strategy, considers the development of
research pertaining to nonmarket strategy, and discusses an agenda for
future research. The paper does not attempt to survey the extant research,
but instead references research contributions as illustrations of approaches
and frameworks. The research agenda is both a set of topics that warrant
further study and a set of challenges associated with theory, empirical
studies, and strategy choice.
This volume is appropriately organized into sections on public politics,
private politics, and integrated strategy. Integrated strategy can involve the
integration of public and private politics strategies, but more importantly it
refers to the integration of market and nonmarket strategies.1 This requires
a framework for market strategy, which is considered below, but the focus
here is nonmarket strategies in the presence of private and public politics.
This includes strategic corporate social responsibility, which is considered
in the context of self-regulation.
Maturation of the field is evidenced by this volume, the Routledge
Companion to Nonmarket Strategy (Lawton & Rajwani, 2015), a
volume on the economics of corporate social responsibility (McWilliams,
2015), the Oxford Handbook on Business and the Natural Environment
(Bansal & Hoffman, 2012), a 2012 Symposium on nonmarket strategy in
the Academy of Management Perspectives (Doh, Lawton, & Rajwani,
2012; Henisz & Zelner, 2012), and the Oxford Handbook on Corporate
Social Responsibility (Crane, McWilliams, Matten, Moon, & Siegel,
2008). Fourteen annual Strategy and the Business Environment confer-
ences have now been held, attracting a core group of researchers and
a broadening set of scholars who find the conferences a useful forum
for their research. Mainstream economics, political science, and strategy
journals publish research on nonmarket strategy, and specialized journals
provide a home for the research. Yet, the field is still young, and much of
the research is formative rather than definitive.2 Much more remains to
be accomplished.
Firms choose market strategies to guide their actions in markets, and
firms also choose nonmarket strategies to guide their actions in their social,
political, and legal environment.
Strategy beyond Markets: A Step Back and a Look Forward 3
Market strategy focuses on the sources of competitive advantage and

sustainable performance and addresses issues involving the lines of business
in which the firm operates, the markets in which it participates, its rivalry
with competitors, investments in innovation, organizational boundaries,
and so on. That is, market strategy focuses on positioning and choice in
markets and the performance of the firm in those markets.
Nonmarket strategy also focuses on choice among strategy alternatives,
not in markets but rather in the context of public and private politics are-
nas in which the firm competes against, and at times works with, its rivals,
interest groups, and social activists. Nonmarket strategy thus focuses on
choice as does market strategy, and the objective is to integrate market and
nonmarket strategies to maximize overall performance.
Nonmarket strategy addresses issues that affect firm performance and
are not fully resolved within markets and private contracting. Nonmarket
strategy takes as exogenous the same factors as in economics and political
science, but it does not take market strategy as given. Instead, market and
nonmarket strategies are chosen together to form an integrated strategy.
As elaborated below, Amazon had a market strategy that exploited its price
advantage resulting from not collecting state sales tax on Internet sales,
and its nonmarket strategy focused on defending against attempts by states
to force the collection of the taxes. As it became increasingly clear that its
nonmarket strategy could not be successful in the long run, Amazon
switched its market and nonmarket strategies. Its new market strategy
included emphasizing rapid delivery and building new fulfillment centers to
fill orders. The company replaced its aggressive nonmarket strategy with a
strategy that emphasized negotiating transition, using the location of new
fulfillment centers in bargaining with states regarding the transition to sales
tax collection.
Because nonmarket strategy is to be integrated with market strategy,
firm leaders and managers must be responsible for both strategies. This is
illustrated in more detail in the section Applying the Approach using
the cases of Uber, Citigroup and the Rainforest Action Network, and
McDonalds and the Fast Food Campaign.
The principal theme of this essay is that more attention should be given
to firm-level strategy choice and implementation. Political economy pro-
vides the micro-foundations for the essay. The firm is the unit of analysis,
and performance of a firm depends on its market and nonmarket strategy,
the strategies of others, the markets and institutions in whose arenas the
strategies are deployed, and exogenous factors such as demographics,
macroeconomic conditions affecting markets, and institutional features.
4 DAVID P. BARON
Identifying the effects of these exogenous factors on strategy choice and

performance is important, but the focus of the field of strategy beyond
markets is on the actions of firms and other interested parties such as com-
petitors, interest groups, activists, and government officeholders. The objec-
tive is to explain firm performance and guide strategy formulation and
choice to improve performance. The task of firm leaders is to position and
direct the firm efficaciously in both its market and nonmarket environ-
ments. The field of strategy beyond markets thus focuses on the nonmarket
environment, nonmarket strategy choice, and its integration with the mar-
ket strategy of the firm. Research on nonmarket strategy often takes the
market strategy of the firm as given, recognizing that nonmarket forces can
affect those strategies.
The next section offers a brief history of the demand for nonmarket
strategy, and the following section addresses the roles of nonmarket strat-
egy and the political economy approach applied at the firm level. The sec-
tion Applying the Approach presents cases in public politics (Uber),
private politics (Rainforest Action Network and Citigroup), and public and
private politics (the Fast Food Campaign and McDonalds). The section
Market and Nonmarket Strategy considers market and nonmarket strat-
egy and self-regulation, and the final section offers observations on research
topics.
A STYLIZED (U.S.) HISTORY OF THE DEMAND

FOR NONMARKET STRATEGY
Although the field of nonmarket strategy is relatively young, some firms
have been practicing it for a considerable time. Firms in regulated indus-
tries and state-owned firms have long interacted with government, as have
firms in resource-extraction industries. Direct social pressure has also been
a factor for some firms as evidenced by the boycott of Nestle over the mar-
keting of infant formula in developing countries and the anti-apartheid
campaigns against firms operating in South Africa. Early research focused
less on the strategies of these firms and more on the rationale for govern-
ment regulation and on the responsibilities of firms to stakeholders.
The environmental movement was ignited by Rachel Carsons Silent
Spring, and much of the early focus was on building awareness of envir-
onmental issues. Activists turned their attention to government and were
an important force in the wave of social regulation enacted during the
1970s that focused on the environment, occupational health and safety,

product safety, worker rights, and other social aspects of business.
Importantly, this social regulation dealt with issues that cut across indus-
try lines. Many firms that had operated largely free of government regula-
tion found themselves having to deal with complex regulations that
forced changes in their strategies and operating practices. Firms began to
recognize that they needed to participate in public policy processes rather
than have policy developed in response to opposing interest groups. The
importance of nonmarket strategy also increased as a result of the deregu-
lation and privatization movements. Firms honed their lobbying, coali-
tion building, advocacy, grassroots mobilization, and judicial strategies.
Researchers turned their attention from the evaluation of market failures
and government intervention to the political strategies of firms participat-
ing in the development of the rules that governed market competition and
shaped the nonmarket environment in which they operated.3
By the early 1990s, firms had sharpened their nonmarket strategies and
slowed the pace of new government regulation. Some critics of business
turned their attention from influencing government to directly targeting
companies through, for example, naming and shaming campaigns. The non-
market strategies of activist NGOs proved successful in many instances, and
particularly after Greenpeaces success in stopping Shell from disposing of
the obsolete Brent Spar oil platform in the North Atlantic, direct pressure
on firms both intensified and broadened. Firms facing criticism and attacks
on their market strategies began to develop nonmarket strategies to deal
with the social pressure they faced. Some firms, however, were too difficult
to challenge with campaigns, and activists innovated by targeting their sup-
ply and distribution chains. This exposed a much larger set of firms to social
pressure and broadened the importance of nonmarket strategy. In response
to this broadened threat from activism, firms began to self-regulate to lessen
the social pressure and avoid targeting. In implementing their self-regulation,
some firms partnered with NGOs, shifting the focus of the NGOs from con-
frontation to cooperation. This self-regulation was often called corporate
social responsibility, but its motivation was different from the earlier version
that focused on responsibilities to stakeholders. This corporate social respon-
sibility was strategic and intended to alleviate or forestall social pressure.
Some firms discovered that this strategic social responsibility could also be
used in marketing and reputation-building. For these firms, nonmarket strat-
egy was naturally integrated with their market strategy. For other firms, the
nonmarket environment was sufficiently important to their performance that
their nonmarket strategy had to be integrated with their market strategy.
6 DAVID P. BARON
The need for nonmarket strategy and its integration with a firms market
strategy increased considerably with the wave of outsourcing and globaliza-
tion. Firms that had operated in developed economies where the rule of
law was strong found themselves operating where the rule of law was weak,
enforcement of regulation was often lax, and corruption in some cases was
widespread. More firms found themselves embroiled in public and private
politics, necessitating the development of effective nonmarket strategies
integrated with their market strategies. Nonmarket strategy had evolved
from being a response to government intervention to being an integral
component of sustainable value creation.
NONMARKET STRATEGY
Roles of Nonmarket Strategy
Nonmarket strategies can be thought of as playing one or more of six roles.

One is rent seeking, as when wind and solar power companies seek to con-
tinue their generous subsidization, or burden lifting, as when firms work
for a lower corporate income tax rate. Another is handicapping competi-
tors by, for example, raising their costs, as in bricks-and-mortar retailers
working to require e-commerce firms to collect sales taxes. More important
are strategies directed at unlocking opportunities, as when firms seek to
open markets closed by a government or to deregulate an industry.
Examples include MCIs ultimately successful entry into the telecommuni-
cations market in the United States (Yoffie & Bergenstein, 1985), the finan-
cial services industrys successful efforts to repeal the GlassSteagal Act,
and Ubers entry into local passenger transportation markets. Fourth,
nonmarket strategies are used to defend against rivals and critics seeking to
restrict a firms opportunities or operations, to forestall social pressure led
by activists, and to forestall initiatives of governments seeking to impose
additional burdens on firms. For example, Internet privacy has largely
been in the realm of self-regulation by individual firms, and attempts by
Congress to address the issue have been furtive and rise and fall with
privacy breaches or their absence but leave the status quo government poli-
cies in place. As discussed below, Amazons nonmarket strategy supported
its market strategy, and when the company changed its market strategy, its
nonmarket strategy changed as well. Another example is restaurants resist-
ing social pressure to raise workers wages as considered in the section
Applying the Approach.4 Fifth, nonmarket strategy is used to attract

customers with social preferences for environmental protection, social jus-
tice, or the protection of rights, as in the strategic use of corporate social
responsibility. Sixth, nonmarket strategies are used to create value by posi-
tioning the firm to be able to anticipate and deal effectively with the non-
market strategy challenges that may arise in the future. This includes
strengthening reputations, building trust, and enhancing legitimacy. Self-
regulation, the acceptance of social responsibilities, forthrightness, and
relationship building are often components of these strategies.
The scope of nonmarket strategy does not include all activities with
social or political aspects. For example, responding to consumer social
preference with green products is straightforward profit maximization.
Shaping those social preferences through the design and implementation of
a corporate social responsibility policy requires a nonmarket strategy.
Similarly, complying with government regulation is not nonmarket strat-
egy, since companies like citizens are to obey laws and regulations.
Lobbying to influence the content of legislation involves a nonmarket strat-
egy. Public relations, involving communicating the firms activities to the
public, is not nonmarket strategy, but a communications strategy designed
to develop a favorable reputation or an advocacy campaign to complement
a lobbying strategy addressing legislation under consideration in Congress
is nonmarket strategy. Nonmarket strategy involves the choice of actions
to affect outcomes in the nonmarket environment and in conjunction with
a market strategy to affect outcomes in the market environment.
Levels of Nonmarket Strategy Research
Research on nonmarket and integrated strategy can be thought of at three

levels. The first focuses on the nonmarket environment and seeks to iden-
tify factors that can affect the choice of strategy and firm performance.
This research can focus on a country, industry, or market or on differences
across countries, industries, or markets. For example, Henisz, Dorobantu,
and Nartley (2014) study the nonmarket environment of gold mining, and
Holburn and Zelner (2010) study the global electric power industry to iden-
tify the nonmarket challenges firms face in operating in countries other
than where they are domiciled.
The second level focuses on explaining differences in performance
among firms. In the market strategy literature, this research must control
for industry-specific factors, such as the intensity of competition, that affect
8 DAVID P. BARON
all firms in a more or less similar manner. The objective then is to identify
firm-level factors, or better yet choices, that cause and hence explain differ-
ences in performance. The challenge is to identify those causal factors. Some
nonmarket issues are industry-specific and amenable to relative performance
analysis, but many nonmarket issues cut across markets and traditional
industry lines complicating the identification of causal factors.
The third level is that of the individual firm, and the focus is on strategy
choice and implementation, taking into account the nonmarket competition
on the issue. At this level the challenge is to understand the specifics of the
strategic situation, how those specifics relate to strategy alternatives, and
how those alternatives are likely to fare in competition with the strategies
of others that will be active on the issue.
The Political Economy Approach
A number of frameworks have been proposed for understanding the non-

market environment and generating hypotheses about firm behavior, strat-
egy choice, and performance. Some frameworks start at the first level, as in
the case of those that focus on country-level characteristics relevant to non-
market strategy, such as the likelihood of expropriation or the strength or
weakness of the rule of law. This approach is useful in understanding the
nonmarket environment, but the step from country analysis to strategy
choice by firms can be a long one. The political economy approach recog-
nizes the importance of understanding the nonmarket environment but
focuses on operationalizing that understanding in a manner that facilitates
strategy choice. That is, it focuses on interests that may act on an issue and
on the institutions in whose arenas the issue will be contested. The intersec-
tion of issues, interests, and institutions and their implications for indivi-
dual firms are at the heart of nonmarket strategy.
The political economy approach begins with micro-foundations for the
behavior of firms, citizens, and government officeholders, and from those
foundations builds theories to predict behavior and identify effective non-
market strategies. It focuses on actions, interactions with the actions of
others, and the likely outcomes of those interactions. This approach takes
politics seriously and views it broadly. In particular, the political economy
approach looks inside government institutions, which is necessary for strat-
egy choice and implementation. For example, in the case of an issue before
a legislature, it is important to understand the preferences of pivotal legisla-
tors and to know how much political action is likely to be taken by those
potentially affected by the proposed legislation. As Olson (1965) argued,

common interests may not lead to common action, since some may free-
ride. This means that for the Internet taxation issue unorganized interests
such as consumers are unlikely to play a role, whereas bricks-and-mortar
merchants are well organized and have strong incentives to act. In private
politics the political economy approach takes into account the strategies of
activists and NGOs and the structure of the campaigns they conduct
against firms (Baron & Diermeier, 2007). For some nonmarket strategy
issues, formal models can be constructed to identify equilibrium incentives,
strategies, and outcomes and to generate hypotheses that can be taken to
the data.
Performance attributable to a nonmarket strategy is not controlled by
the firm any more than is performance attributable to a market strategy. A
market strategy competes with the market strategies of other firms, and a
nonmarket strategy competes with the nonmarket strategies of other firms,
interest groups, and social activists. Nonmarket strategy thus focuses on
choice and competition, as does market strategy. The disciplines most
closely related to choice and competition are economics and political
science, and the political economy perspective is based primarily on those
two disciplines. This does not mean that insights from other social science
disciplines are not drawn upon in choosing strategies but instead that the
framework for approaching strategy choice begins with those two
disciplines.
The Firm Level
The challenge for a strategy field, either market or nonmarket, is to identify

the best strategy for a firm in a particular situation. At the teaching level
this is the realm of cases and basic models of strategy choice, competition,
and bargaining.4 Examples of basic nonmarket models include the median
voter theorem and extensions to gridlock theory, free-rider analysis, hold-
up problems, lobbying models, election models, activist campaign models,
self-regulation models, and so on. That is, given a fairly detailed descrip-
tion of a firm and its situation, what is the best course of action to maxi-
mize its long-term value given that other firms and interests are also
choosing their strategies optimally? This is not a static exercise but rather a
dynamic one, since ideally the strategy takes into account the situation the
firm might be in at future points in time, depending on uncertain factors
and the outcomes of earlier choices. This recognizes that the firm will have
10 DAVID P. BARON
the opportunity to reevaluate its strategy and adapt it to the situation in

which it finds itself, and this means that current strategy choices have
option values to the extent that they can be adapted to new or changing
market and nonmarket circumstances.
Since observing the specifics relevant for firm-level strategy choice is dif-
ficult from afar, a framework should include a methodology that in combi-
nation with the specific details of the competitive situation leads to the
choice of the best strategy. The framework and methodology for nonmar-
ket strategy choice used here is developed and presented in my book
Business and Its Environment, 7th ed. (Baron, 2013). In short, the methodol-
ogy is organized around a nonmarket issue or challenge faced by a firm or
collection of firms and focuses on the actions they and other interests take
to influence the outcome of the issue. This approach is illustrated in the
three cases in the section Applying the Approach.
The goal is an integrated strategy that guides a firms conduct in its mar-
ket and nonmarket environments. When conditions in one environment
change, strategy may need to change. Nonmarket strategy often enables a
market strategy, and when the market strategy changes, the nonmarket
strategy changes. Amazon had a well-integrated strategy with an aggressive
nonmarket component directed at avoiding having to collect sales taxes
from its online customers. Amazon claimed protection from an earlier
Supreme Court decision in a case involving catalog sales that held that
unless a company had a nexus in a state it was not required to collect sales
taxes. A nexus was interpreted as a retail store, warehouse, or office.
Amazon maintained that the decision applied to Internet as well as catalog
sales and hence collected sales taxes only in its home state of Washington
and in the few states in which it had facilities. It also maintained that ware-
houses did not constitute a nexus but instead were simply fulfillment cen-
ters. The company strategically limited the number of fulfillment centers to
limit its obligation to collect taxes.
Since sales taxes are state taxes, Amazon was forced to deal with the
issue on a state-by-state basis. Some states, most notably New York, took
Amazon to court in an attempt to require it to collect the state sales tax.
Amazon vigorously contested the lawsuit. Internet sales grew rapidly, and
a number of states began to consider legislation requiring large online retai-
lers to collect sales taxes from residents in the state.5 Amazon responded
aggressively by not only lobbying legislators and the governors office, but
also by threatening to sever its contracts with local associates that operated
Internet sites and received a payment when their site directed a user to
Amazons website. This caused direct harm to the associates and resulted
in hesitation in some states. In some states such as Maryland and Texas

where Amazon planned to open fulfillment centers, it threatened to cancel
its plans. Amazon, however, failed to prevent California from passing a
law requiring the collection of sales taxes.
Amazon knew that this was a battle it would eventually lose, in which
case it would lose a competitive advantage relative to bricks-and-mortar
retailers. Needing a new strategy and a new competitive advantage,
Amazon adopted a market strategy emphasizing rapid delivery and particu-
larly one-day delivery. To implement its new market strategy, it needed
more fulfillment centers. In California it used its bargaining power from
the promise of new jobs to negotiate a moratorium on sales tax collection
in exchange for committing to open several fulfillment centers in the state.
With its new market strategy it backed off its aggressive nonmarket strat-
egy and began to collect sales taxes in an increasing number of states.
APPLYING THE APPROACH

Public Politics Example: Uber
Some entrepreneurial firms are able to operate without a nonmarket strat-

egy or to free-ride on the efforts of others, but many others face nonmarket
issues early in their lives. Uber immediately faced regulations governing
taxi and limousine service, and those regulations varied not only by coun-
try and state but also by city. In many local markets interest groups includ-
ing taxi companies and their drivers, and in some countries public
transportation workers organized demonstrations and worked for more
stringent regulations to limit Ubers penetration of their markets. The prin-
cipal nonmarket opposition was taxi companies. Uber founder and CEO
Travis Kalanik said, Our opponent the Big Taxi Cartel has used dec-
ades of political contributions and influence to restrict competition, reduce
choice for consumers, and put a stranglehold on economic opportunities
for its drivers.6 This section illustrates the political economy approach
applied to Ubers strategic situation with an emphasis on its nonmar-
ket strategy.7
Ubers value proposition is to be Everyones Private Driver by coordi-
nating efficient and high-quality car service. Its market strategy is to allow
passengers to use mobile devices to arrange a pickup with a car arriving
usually in 2 or 3 minutes. When requesting service, customers click an app
12 DAVID P. BARON
and GPS gives the location to Ubers software. People requesting service
give their credit card information in advance and pay online automatically
at the end of the trip with the amount determined by the time and distance
of the trip and the time of day, so drivers do not have to carry cash. Ubers
fares were 40100% higher than taxi fares because of the quality of the cars
and the compensation of drivers, and are higher during high-demand times.
Drivers in Ubers network receive 80% of the fare with the other 20%
going to Uber. Drivers are not employees but instead are private contrac-
tors who join Ubers network and respond directly to the requests by pas-
sengers. Ubers business model is to be the connection between passengers
and car service and only the connection. For example, drivers are indepen-
dent, at-will contractors paid directly from their fares, and they provide
insurance to cover themselves and passengers in the event of an accident.
Drivers participating in the Uber network were required to take all cus-
tomers and to deliver them to any location in the local area. This policy
gives Uber a nonmarket asset, since it makes service available in areas
poorly served by taxi companies. Drivers initially were recruited from
among licensed limousine drivers and in some cases through limousine
companies, and the drivers carry insurance. Drivers make considerably
more driving for Uber than driving a taxi or a limousine. Drivers and
limousine companies own the cars. Uber cars are black with no distinctive
markings, do not have meters, and do not roam to pick up passengers, all
of which differentiate them from taxis. Uber cars are also clean and com-
fortable, and drivers are restrained. Uber relies on mobile devices and GPS
technology not only for connecting passengers and cars but also to position
cars efficiently for serving passengers. Technology also allows Uber passen-
gers to rate drivers, and drivers can report troublesome passengers.
Ubers service has strong network externalities and positive feedback,
since the more drivers in its network the better is the service for customers
and the more customers the easier it is to recruit drivers and limousine
companies. Ubers market strategy is also scalable and can readably be
deployed in new cities. The technology of Ubers service is easy to replicate,
so entry into the market is easy as Uber demonstrated. Markets are local,
and economies from operating in multiple markets are relatively small, so
entrants can range from companies serving a single market to ones operat-
ing globally. There, however, is a first-mover advantage to the extent that
the first-mover can achieve scale to reduce fares and wait times.
Nevertheless, competitive services can be expected, particularly if the
entrant can offer a price advantage. The first-mover also has an advantage
in the selection of drivers and limousine companies. As a second phase of
its market strategy in 2012, Uber introduced a lower fare service UberX
that uses hybrid vehicles to reduce operating costs, allowing fares closer to
those of taxis. Uber also introduced a low fare service Uberpop in Europe.
In June 2012, Uber raised $1.2 billion from private investors, which gave it
an implied valuation of $18.2 billion. By 2015, the implied valuation had
increased to $40 billion.
The initial market strategy decision was where and how to enter a
market, and this involved the nonmarket issue of whether its service was
legal, since in many cities its service could violate regulations established
for taxicabs and limousine service. For example, some states required a
time interval such as an hour between when a customer called for a pickup
and the car arrived and some prohibited limousine fares varying with
time and distance. Some cities had minimum fares to protect taxi compa-
nies, and some required dispatch offices in the city. A myriad of other
local regulations were present.
Uber could have chosen a nonmarket strategy of working within the
regulatory system to unlock opportunity in local passenger markets and
shape local regulations to accommodate its market strategy. Uber could
seek regulatory authorization from city or state agencies, but that could be
a lengthy process since taxi and limousine companies would surely attempt
to delay or block its attempts. Moreover, they could well be successful.
Another approach would be to enter directly a market and deal with non-
market issues and public politics as they arose. Since licensed limousine
drivers would be used, Uber had an arguably allowable service. Moreover,
if Uber were successful in serving passengers, it would have the beginning
of a rent chain on which a nonmarket strategy could be based. Passengers
are constituent of mayors and city council members, and their voice would
be heard. The best market to enter was one that was underserved by taxies,
where a constituency could be developed quickly. Since Uber planned to
serve the entire community, it could receive strong constituent support in
underserved areas of cities, many of which were low income and largely
minority. The direct entry strategy was clearly the better.
Ubers integrated strategy thus was to enter a market with scale
and deal with nonmarket issues as they developed. Its nonmarket strategy
then focuses on the response by interests and institutional actors to its
market penetration. The more successful that penetration the stronger is
the incentive of the opposing interests to restrict the company. The princi-
pal role of Ubers nonmarket strategy then is to protect its position from
opposing interests attempting to impose restrictions on it or shut it
down completely.
14 DAVID P. BARON
Kalanik decided to charge ahead and entered the underserved San

Francisco market in 2009 as UberCab. After being shut down by the State
of California for the use of the word Cab, Kalanik removed the word
and resumed operations. As the adage goes, Uber shot first and asked ques-
tions later. In each market, Uber studied the licensing and local regulations
to avoid overt violations, but it was comfortable operating in the gray areas
between what was clearly legal and what was clearly illegal. For example,
Uber did not serve hailing passengers, which would be a clear violation of
taxi regulations. The distinction, however, between hailing a taxi at the
curb and using a smartphone app to call for an Uber car was a fine one. By
mid-2014, Uber operated in 170 cities in 43 countries.
The nonmarket issues Uber faces center on the enforcement of existing
regulations and the adoption of new regulations intended to restrict its
market strategy. The nonmarket issues differ by country, state, and city,
and each market has somewhat different issues, interests, and institutions,
but there are some common characteristics. In the United States, limou-
sine services are typically licensed by the state, whereas taxi fares and reg-
ulations are set by municipal governments. Similarly, qualifications for
limousine drivers are set by the state, whereas qualifications for taxi dri-
vers are set by the municipality. The structure of municipal governments
varies, but typically they have an elected mayor and city council and a
taxi regulatory agency. The municipal government could impose regula-
tions on Ubers service, including limits on prices and conditions of ser-
vice. In many cities, each taxi is required to have a medallion, and the
number of medallions is tightly controlled with the market price of a New
York medallion trading over $1 million. In other cities entry into the taxi
market is unrestricted. Entry into the limousine market is typically unrest-
ricted other than for licensing.
The principal interests affected by Uber are passengers and taxi compa-
nies and their drivers.8 Passengers are typically thought to be costly to
organize, but Ubers passengers can at very low cost use social media to
express their support for Ubers service. This also provides an opportu-
nity for a local politician to represent their interests, which has been the
case in underserved areas of cities. Uber can also serve as a political entre-
preneur and represent their interests. Washington, DC, has 37 Advisory
Neighborhood Commissions that represent local neighborhoods and
represent their interests before the city council and city regulatory agen-
cies. Uber brought car service to many neighborhoods that were under-
served by taxis, and the neighborhood commissions expressed support for
UberX, which was under consideration by the city council.
Limousine services could compete with Ubers system, but the compa-
nies and their drivers have the substitute or alternative of joining Ubers
system. Taxi companies and their drivers are different. The taxi companies
and their drivers cannot easily join Uber, since they would have to obtain a
commercial license, which is costly to obtain. The costs of organizing the
taxi interests for nonmarket action are low, and in some cities a union
represents the drivers. The costs of organization are lowest where entry
into the taxi market is restricted and highest where it is freely open. The
taxi companies could enter Ubers market segment, but that would require
a costly upgrading of their fleet. Facing competition from Uber and other
car services, the taxi companies could improve the quality of their service.
The better alternative for the taxi companies and drivers is a nonmarket
strategy. The taxi companies and their drivers are important parts of the
electoral constituency of mayors and city council members. They also pro-
vide contributions for electoral campaigns. At times, the financial and elec-
toral support is in exchange for higher fares or more favorable regulations.
One advantage for Uber is the widespread public dislike of taxi service
because of the low quality, lack of cleanliness of the taxis, and aggressive
driving. Also, taxi companies are hesitant to serve parts of cities where a
return fare is unlikely and where there is risk to the drivers, who often carry
significant amounts of cash. Some city council members and local commu-
nity groups see Uber as better serving their constituents. Nevertheless, the
taxi companies and their drivers constitute a well-organized interest group
with significant resources and often long-established political connections.
In some countries such as France, taxi unions are aligned with public trans-
portation unions, which supported demonstrations against Uber.
The nonmarket strategy objectives of the taxi companies and their dri-
vers are to constrict Ubers ability to connect with potential passengers, to
raise its costs, to regulate its prices, and to restrict its service. In some juris-
dictions, regulations place a floor on Ubers fares, which poses a threat to
short trips and particularly to UberX. In Germany, the taxi companies
took Uber to court for violating a regulation on fares, and the court ruled
against Uber. Berlin had banned Uber, and the court ruling could lead to a
national ban. Uber announced that it would appeal the court ruling and
vowed to continue to operate.9
Ubers drivers are not employees but could be organized by Uber, but
they do not represent a traditional constituency for local elected officials.
Uber also has little experience with local or state regulators, so has limited
specialized political knowledge or a history of political connections. Uber
could form a coalition with similar firms in some markets, but in most cases
16 DAVID P. BARON
those companies are small with few political resources. In addition, Uber
and Lyft had been trading allegations that the other was trying to sabotage
its operations. Ubers greatest nonmarket asset may be the publics dissatis-
faction with taxi service and the obvious fact that the public is
better served by Ubers presence in the market. Nonmarket decisions,
however, are not always dictated by the public interest.
Uber deals with both state legislatures and regulatory agencies as well as
with municipal regulators. Although the companys rent chain consists
only of passengers and the drivers in its network, it could act as a political
entrepreneur mobilizing its customers, who clearly value its service.
Although the taxi companies have allies in the districts in which Uber oper-
ates, state legislators understand that Ubers service is very popular. Uber
hired lobbyists to bring this message to state legislatures and regulators. On
an insurance issue in California, it employed grassroots and mobilization
strategies in the district of a key state legislator, sending mailers to her con-
stituents and producing a television ad with a former professional basket-
ball player delivering the message. Ubers basic strategy when facing
restrictive regulations was to use bargaining, and the aggressive strategy in
California was intended to strengthen its bargaining position.
State law can override municipal regulations, and one of Ubers non-
market objectives at the state level is explicit language allowing its service.
In California, Uber was able to obtain regulations by the Public Utilities
Commission to allow the use of its on-demand car service app. By the end
of 2014, 18 cities had adopted ordinances allowing Uber service. Uber also
seeks state authorization for those cities in which its operations had
been blocked.
Ubers nonmarket strategy is as aggressive as its market strategy. Its prin-
cipal nonmarket strategy is lobbying city councils, mayors offices, and state
legislatures and participating in regulatory rule-making. This requires knowl-
edge of the regulatory institutions and the specific issues before them. To
provide knowledge and help address the specifics of regulations, Uber hired
a top official with the New York City Taxi and Limousine Commission.
Ubers rapid success resulted in entry at the low price end of the market.
Companies such as Lyft and Sidecar entered with a ride-sharing model.
Rather than using licensed limousine drivers, these companies allowed any-
one with a car to join their network and carry passengers. Ride-sharing
posed a major threat not only to taxis but also to Uber. Taxi companies
and drivers unions attacked the ride-sharing companies with a nonmarket
strategy to raise their costs and restrict their operations. Consumer advo-
cates criticized the companies because their drivers did not have special
licenses or commercial insurance as taxi companies were required to carry

and Uber chose to require for its network. Most cities did not enforce exist-
ing regulations on ride-sharing companies, however.
After a year, Uber was forced to respond and entered the ride-sharing
segment of the market. Its entry was soon followed by a new component of
its integrated strategy. Uber had to recruit drivers without limousine
licenses, and the recruited drivers might not have adequate insurance. To
reduce the cost advantage of ride-sharing, Uber announced that each ride-
sharing trip would be covered by a $1,000,000 per-incident insurance pol-
icy and that it will conduct extensive and strict background checks on
its ride-sharing drivers.10 Uber challenged regulators to impose these
requirements on all ride-sharing companies. Ubers nonmarket strategy
had come full circle from opposing regulation to seeking regulation to pro-
tect its position and increase the costs of its rivals.
Uber also became engaged in private politics fueled by the media.
Ubers aggressive integrated strategy attracted extensive media coverage
that included at times combative exchanges with the media. Travis Kalanik
countered with media interviews and commentary, but things turned worse
when an Uber executive was quoted at a private dinner party saying that
he would like to investigate the personal lives of critical journalists. This
ignited a sea of media coverage, forcing Kalanik to reject and criticize the
executives comments. Private politics had arrived at Uber.
The taxi companies and their drivers can be expected to continue their
efforts to restrict Uber, and the threat posed by UberX, Uberpop, and
ride-sharing strengthens their incentives. Since city markets are largely
independent, Uber faces the challenge of potentially working in a different
institutional setting in every city and state in which it operates. This
requires monitoring and action when its interests could be affected. Uber
deals with nonmarket issues city-by-city, state-by-state, and country-by-
country, but as the number of cities served increased, it became important
to develop a broader nonmarket strategy and coordinate its nonmarket
actions. In addition Ubers continued to raise new nonmarket challenges.
For example, its opponent argued that Uber drivers should be classified as
employees rather than as independent contractors. This would require
Uber to provide health insurance, pay Social Security and Medicare taxes,
and incur additional burdens. Moreover, Uber had begun to explore a
complementary service of using its fleet of drivers to transport items as well
as passengers, which would raise a new set of nonmarket issues.
In 2014, Uber hired as executive vice president David Plouffe, who was
an architect of President Obamas 2008 election campaign and served as a
18 DAVID P. BARON
top advisor to the president from 2011 to 2013. Kalanick said, Uber has
been in a political campaign but hasnt been running one. That is changing
now.11 Plouffe stated, Were on an inexorable path of progress here. Uber
is making transportation safer; its providing jobs; its cutting down on drunk
and distracted driving. I think the mission is really important.12 Matthew
Daus, a former commissioner with the New York Taxi and Limousine
Commission, commented hundreds of millions of dollars are being spent by
a Silicon Valley cartel in this hostile takeover of the citys taxi industry.13
Ubers success was accompanied by an explosion of its nonmarket issue
agenda, fueled by media coverage of its every move and every problem it
encountered. An Uber driver in India was charged with raping a passenger,
igniting a wave of protests around the globe. Critics complained about
inadequate screening of drivers safety records, resulting in calls to require
a commercial license and to assure that drivers did not have criminal
records. Uber collected data on passenger hailing locations and destina-
tions, raising calls by privacy advocates to disclose what information it col-
lects and how it is used. Complaints also arose about its surge pricing,
threatening the possibility of fare regulation. Uber was also proactive, part-
nering with MADD in studying the effect of Ubers car service on drunk
driving. Uber also partnered with an insurance-by-the-mile company to
close the insurance gap. The company strengthened the screening of drivers
and added a panic button to its app for use in case of an abusive driver.
Uber was learning the importance of nonmarket and integrated strategy.
Private Politics: Rainforest Action Network and Citigroup
The Rainforest Action Network (RAN), a radical environmental NGO

based in San Francisco, was concerned with harm to ecosystems in devel-
oping countries caused by large infrastructure projects such as roads,
telecommunications systems, pipelines, etc. The projects were owned or
authorized by national governments, which contracted with large interna-
tional construction companies to construct the projects. Many of the con-
struction companies were privately owned and were headquartered in the
United States, Europe, and Asia. There was little that RAN could do to
influence the countries, and the companies were also beyond its reach not
only because they were headquartered in different countries but because
most did not have a public face or interact with the public. Instead of
attempting to target a construction company or a government, RAN
decided to conduct a value chain campaign.
An essential factor of production for the infrastructure projects was pro-

ject finance, since 8090% of the project cost was financed by loans from
large banks. RAN reasoned that if it could limit the availability of project
finance or have environmental standards required for lending, it could limit
the harm to ecosystems. RAN researched the project finance lenders and
found that Citigroup was the largest supplier of project finance. It decided
to campaign against Citigroup on the theme that it was causing environ-
mental damage through its financing. The campaign was intended to harm
Citigroup and cause it to restrict lending. In contrast to Citigroup, RAN
was an organization consisting of 23 people with a budget of $2.5 million
for all its campaigns. The campaign and Citigroups actions are detailed in
the case entitled Anatomy of a Corporate Campaign: Rainforest Action
Network and Citigroup (A)(B)(C), published in Business and Its
Environment and available in the Stanford Graduate School of Business
case series, so only a brief sketch is provided here.
RAN planned to mount a naming and shaming campaign augmented
with more aggressive measures including demonstrations, banner hangs,
blockades of branches, and more. RAN began its campaign by writing to
Citigroup with its assessment of the harm caused by project finance and
demanding that the bank stop the destruction. Citigroup was perplexed by
the letter, since in its view it was not causing any harm to the environment.
For nearly two years Citigroup met with RAN, but from RANs perspec-
tive no progress was made. RAN ramped up its campaign including hang-
ing a banner from Citigroups headquarters, demonstrating in front of its
headquarters, chaining closed the doors to branches, and having grade
school children write to Citigroup CEO Sandy Weill. RAN also organized
days of protests with demonstrations in a number of U.S. cities and
countries, conducted by aligned activist groups. RAN also took out ads in
several publications showing the destruction caused by the projects, and
also took out an ad in The New York Times accusing Citigroup of dama-
ging the environment. RAN also urged Citi credit card holders to destroy
their cards to put pressure on the bank. The campaign actions by RAN
caused little direct harm to Citigroup, but it became clear to Citigroup that
RAN would not go away. RAN escalated its campaign by handing out
posters in Greenwich, Connecticut, where Sandy Weill lived. The posters
headline was Wanted with a picture of Weill, followed by text. RAN also
went door to door in Weills neighborhood, passing out posters. RAN even
went to Weills home and handed a poster to his son.
Citigroup had no nonmarket strategy to deal with the campaign and
largely remained silent. It knew that RAN was planning a demonstration
20 DAVID P. BARON
at its annual meeting, and on the evening before the meeting Citigroup
called RAN and asked for a cease-fire, agreeing to meet with RAN to
negotiate the issues. RAN agreed to a three-month cease-fire and negotia-
tions began.
Citigroup was also under some pressure from the International Finance
Corporation (IFC), a member of the World Bank Group, which had what
RAN viewed as a very weak set of environmental standards for project
finance. Other project finance banks in Europe also were under social pres-
sure from environmental activists and the IFC. Citigroup decided to initiate
self-regulation and began meeting with those banks to address the project
finance environmental concerns by designing more stringent environmental
standards for loans. In the midst of the cease-fire, Citigroup and three
European banks, accompanied by six other banks, issued the Equator
Principles for project finance. The Equator Principles are designed to
reduce the environmental and social risks from projects and have largely
been applauded by environmental NGOs. Eighty banks from 34 countries
now participate in the Equator Principles.
The Equator Principles were not enough for RAN, which continued
negotiations focusing on implementation and reporting details. Citigroup
agreed to provide advance notice of financing and to report on develop-
ments, and also pledged to implement a detailed set of Environmental
Initiatives. RAN ended its campaign against Citigroup and turned its atten-
tion to the Bank of America and JPMorgan Chase.
Citigroups absence of a nonmarket strategy to deal with RAN and its
campaign had delayed a serious focus on the environmental effects of pro-
ject finance. Ultimately, its nonmarket strategy had two components. One
was to negotiate with RAN to end the campaign. The other was to self-
regulate to address the environmental concerns. It is not clear how much
harm RAN had caused Citigroup, but at a minimum it was a nuisance and
a source of negative media coverage. Today, few companies targeted by an
activist group would allow a campaign to continue for nearly four years.
Integrated Strategy and Private and Public Politics: The Fast

Food Campaign
The Fast Food Campaign (FFC) illustrates the approach to integrated

strategy. The FFC and the broader restaurant and minimum wage campaigns
are well designed, well directed, and well funded. The principal targets are
fast food chains, and the analysis here takes the perspective of a chain such
as McDonalds. In contrast to many activist campaigns that are top down;

that is, an activist NGO designs a campaign with the hope that the public
will support it sufficiently to cause its target to change its practices, the FFC
originates from ongoing community organizing efforts with the objective of
empowering low-income community residents. A leader and innovator in
community organizing was Saul Alinsky, who organized Chicagos Back of
the Yards community, so named because it was in the shadow of the Chicago
stock yards. The goal of community organizing is community development
and empowerment and the local provision of public goods for the commu-
nity. The current version of these organizations is a worker center, and the
FFC and the related campaigns build on this base. The difference is that the
worker centers are funded in part by unions seeking to recruit new members.
The ostensible goal of the FFC is a wage of $15.00 an hour, but it has
four principal private and public politics objectives. The first is to impose
social pressure directly on employers to raise wages. The second is to sup-
port union organizing. The third is to spur grassroots support for higher
government minimum wages at the federal, state, and local levels. The
fourth is implicit: to strengthen the base of the Democratic party. The pub-
lic politics portion of the campaign for higher wages operates at three insti-
tutional levels. The first is federal, but divided government means there is
little chance that the increase in the minimum wage to $10.10 sought by the
Obama Administration will be enacted. The second is in states where pro-
labor Democrats have a unified government. The third is the city level,
which in some regards is the easiest, since most large cities have a govern-
ment controlled by Democrats, and in many of those cities the governments
are backed by public sector unions as well as unions such as the Service
Employees International Union (SEIU) and United Food and Chemical
Workers (UFCW). The Obama administration led by Secretary of Labor
Thomas Perez embraced the campaigns, with Perez justifying the cam-
paigns on social justice grounds.
Fast Food Forward, the coalition conducting the FFC, is led by
New York Communities for Change (NYCC), formed in 2010 by former
ACORN leaders and backed by unions as well as private foundations.
In 2012, the SEIU contributed $2.5 million to NYCC for support for
organizing.14 The SEIU has 2.1 million members with 150 local chapters.
In 2012, its cash receipts were $411 million and its expenditures were
$453 million. In the 2012 election cycle, the SEIU spent over $23 million,
almost all of which was in support of Democrats. The SEIU has substantial
resources to fund a campaign, but the key to success is the participation of
local residents and community groups.
22 DAVID P. BARON
The union money funded demonstrations. They began in 2012 with

demonstrations by 200 people against 20 New York restaurants and contin-
ued in 2013 with a series of one-day strikes, as the organizers called them.
The strikes, however, were in name only, and very few restaurant workers
participated. One-day strikes were also conducted against McDonalds for
the purpose of attracting media coverage. Demonstrations were organized
in a number of U.S. cities and countries.
Fast Food Forward funded a study that estimated the public assistance
paid from four major government programs to fast food workers. The study
found that 52% of frontline fast food workers received assistance from at
least one of four federal programs compared to 25% for all workers. The
cost of that assistance was estimated at $6.99 billion a year. Overall, 20% of
fast food workers had income below the federal poverty level, compared with
5% nationwide. The study also examined the mix of people employed in fast
food jobs.15 The study reported that members of historically disadvantaged
classes of workers were over represented among fast food workers. Twenty-
six percent of the workers had children, 23% were between 16 and 18 years
of age, and 20% were married. The study called for higher wages for fast
food workers and identified a higher minimum wage and collective bargain-
ing as effective means of achieving that objective.
The National Employment Law Project (NELP) used the Fast Food
Forward-funded study data to issue a report entitled Super-Sizing Public
Costs and estimated a $3.8 billion cost of the public assistance
attributable to workers in the 10 largest fast food companies. The report
compared the cost to the profits of the companies and the compensation of
their CEOs. The Fast Food Forward and the NELP reports were released
on the same day to coincide with planned demonstration by workers
and union representatives. Michael Saltzman of the union watchdog
Employment Policies Institute (EPI) commented, Its no surprise that two
organizations the SEIU funds generously would carry the unions water in
these specious new reports. The EPI paid for full-page advertisements in
major newspaper with the headline The Best Weapon in the War on
Poverty is a Job, showing a picture of a man in a hoodie and cutoff gloves
holding a cardboard sign reading I dont need a raise I need a job. The
advertisement opposed a minimum wage of $10.10 and stated Of those
who would receive a raise, just 13 percent live in poor families.16 The ad
referred the reader to minimumwage.com. The EPI was headed by a lobby-
ist for the restaurant industry.
Many of the jobs in question commanded the minimum wage or close to
it. The low wages were due to labor supply driven by a large number of
low-skilled workers, including millions of undocumented workers. An

increase in the minimum wage was the solution from the point of view of
the social activists, unions, and the Obama administration. Professor
Arindrajt Dube of the University of Massachusetts said that a $15.00
hourly wage would increase average wage costs by 60% and increase the
price of a $3.00 hamburger to $3.50 or $3.60. He said, Would I be
concerned about possible job losses if there were a $15 minimum wage in
the restaurant industry, yes, Id be concerned. It might lead to the substitution
of automation for workers.17 The Congressional Budget Office estimated
that the Obama administrations proposed increase in the federal minimum
wage to $10.10 would cost the economy 500,000 jobs. The Obama adminis-
trations proposal was nearly dead on arrival in Congress, and the nail in the
coffin was provided by the CBO.
The minimum wage campaign posed a threat to several industries
including restaurants, hotels, and retail. These industries are unconcen-
trated with many thousands of firms, and they conduct nonmarket strate-
gies primarily through industry associations. The National Restaurant
Association opposes any increase in the minimum wage, pointing out that
the median earnings for waiters are $1622 an hour, depending on experi-
ence. The opposition to the minimum wage campaign was spearheaded by
the Chamber of Commerce, since thousands of small companies would be
adversely impacted by a significant increase in the minimum wage. The
Chamber and the associations conduct informational lobbying and grass-
roots nonmarket action. The opposition also includes a number of conser-
vative organizations dedicated to countering union-funded campaigns.
For example, Americans for Limited Government provides a website www.
seiuwatch.com. The Chamber of Commerce hired Professor Jarol
Manheim to assess the impact of worker centers. He produced a detailed
and publicly available report (Manheim, 2013) reviewing their history, their
organization, and their funding. He concluded that the unions had out-
sourced some of their traditional organizing activities to the worker centers
in exchange for funding, but he observed that there remained a tension
between the objectives of the centers and those of the unions.
In addition to supporting the public objective of a higher minimum
wage, social pressure was directed at individual firms. With funding by the
SEIU, NYCC organized one-day strikes by some McDonalds workers.
In Chicago, Fight for 15 was the local union organizing campaign formed
by the local worker center Workers Organizing Committee of Chicago,
which is staffed and funded by the SEIU. On its website it called on
McDonalds, Whole Foods, and Sears to pay a living wage. The broader
24 DAVID P. BARON
campaign organized coordinated demonstrations in several countries, and

McDonalds was a convenient target for attracting media coverage. Some
of the demonstrators were McDonalds workers, who were made available
for media interviews. The workers argued for justice and higher wages.
A few cities and states increased their minimum wages, but typically not by
the amount demanded by the campaigns. Internally, some SEIU members
questioned why it should spend so much in an industry in which it had
never sought to represent workers.
McDonalds faced two primary nonmarket issues, one private politics
and the other public politics. The private politics issue was the social pres-
sure generated by the campaigns and the associated efforts to unionize its
employees. The public politics issue was the legislative efforts to increase
city and state minimum wages. An integrated strategy was needed.
McDonalds had long experience in dealing with nonmarket issues, and its
public and private politics strategies were often effective. It monitored its
environment carefully and maintained the capability to act quickly on
issues. It had, for example, responded effectively to health concerns such as
mad cow disease and other potential health threats. Its reputation for effec-
tively dealing with nonmarket issues was among the best.
Formulating a nonmarket strategy for addressing the private politics
campaign begins with an assessment of vulnerability of the firm as a func-
tion of the scale of the campaign. McDonalds brand is its principal asset,
and that brand is to some extent a function of public perceptions, but
McDonalds has extensive experience with such issues, including unioniza-
tion attempts. In contrast to an environmental issue, the wage issue involves
no externality and generates less sympathy. Moreover, unions are not held
in high regard by the public. The next step is to assess the degree of public
support for the issue, recognizing that the support is uncertain and can vary
by locale. This step involves assessing the support among employees for the
issue. Possible spillovers to public politics also must be assessed.
The activists capacity must also be assessed, beginning with their fervor
and resources. Worker centers are a relatively new phenomenon, so
McDonalds has little direct experience with them. It does have experience
with the unions, however, and understands that they have the resources to
conduct a multi-year campaign. Perhaps the best example for McDonalds is
the multi-year, union-funded campaign against Wal-Mart during the pre-
vious decade. The campaign had little effect on Wal-Marts employees but
may have tarnished the companys attractiveness to upper income customers.
McDonalds principal strategy alternatives are (i) to do nothing, rely on
the Chamber, industry associations, and the anti-union groups to oppose
the campaigns, and hope the SEIU gives up, (ii) respond in a low-profile
manner by providing information to employees and the media to counter
the claims by the campaigns, (iii) mount a higher-profile media response
with advertisements, interviews, and lobbying, (iv) bargain with the cam-
paigning NGOs, and (v) self-regulate by raising wages above market rates.
McDonalds workers are difficult to unionize in part because of turnover
and the expectation by many that they will quit as soon as they find a
better paying job. Most restaurant employees do not expect to make a
career at McDonalds. McDonalds knows that its workers who partici-
pated in the strikes are paid by the unions and are unlikely to strike
otherwise. John Jones, a manager at a Burger King restaurant who partici-
pated in the September 2014 demonstrations added, The majority of my
co-workers dont speak English very well, and the idea of walking off the
job is not financially feasible.18 McDonalds also knows that the concern
of many workers is the number of hours they can work and less so the
wage, since McDonalds paid the prevailing local labor market wage.
If McDonalds were to pay significantly higher wages, it would be at a
competitive disadvantage if it passed on the higher costs. If it did not pass
on the cost, its profits could be substantially lower. Moreover, providing sig-
nificantly higher wages would not likely bring in any more customers nor
is the campaign likely to convince many customers to go elsewhere. In the
midst of the minimum wage debate, Gap, Inc. announced that it was
increasing its wages for 65,000 of its 95,000 U.S. employees to $9.00 an
hour in 2014 and $10 an hour in 2015. GEO Glenn K. Murphy explained,
We are investing in our workforce, because we think going forward, stores,
and the service our workers provide will become even more important to
our customers.19 McDonalds did not have a service problem.20
The SEIU was spending nearly $3 million a year financing the cam-
paigns, and a central question was how long it would continue to do so in
the absence of success. The unions eventually gave up on their campaigns
against Wal-Mart and at some point would likely do so in this case as well.
The campaigns and the worker center strategy had electoral objectives,
however, and activities were designed to support the mobilization of the
base of the Democratic Party. On Labor Day 2014 President Obama gave
a speech in which he supported the campaigns and said that if he were a
worker in the restaurant industry, he would join a union. The campaign
organizers decided to escalate their protests in advance of the 2014 elections
by engaging in what they referred to as civil disobedience, and nearly 500
were arrested during the September 2014 demonstrations. McDonalds
replied, These are not strikes but are staged demonstrations in which
26 DAVID P. BARON
people are being transported to fast food restaurants.21 The company said
all its restaurants remained open. In their media strategy the demonstrators
continued to refer to strikes, and the mainstream media began to call them
strikes rather than demonstrations.
The social pressure and publicity generated by the campaigns was more
of a threat on the customer side, since some customers might respond to
the campaign message. This effect was likely to be small, however.
Nevertheless, it was worthwhile countering the claims made by the cam-
paign, but to do it in a low-profile manner. McDonalds principal strategy
was information provision, and while its customers were unlikely to see the
information, the media and politicians would see it. McDonalds conveyed
its messages on the Internet and in press releases. For example, it empha-
sized that wages were set in local labor markets and that the company
served as a first-step for many workers who then moved on to successful
careers at the company or elsewhere. McDonalds wrote on a company
blog, as with most small businesses, wages are based on local wage laws
and are competitive to similar jobs in that market. Our history is full of
examples of individuals who worked their first job with McDonalds and
went on to successful careers both within and outside of McDonalds.22
The FFC argued that wages were so low that many fast food employees
were receiving welfare payments and that the composition of the workforce
had changed as a result of the recession with more principal breadwinners
now working at fast food restaurants. Whereas both of these arguments had
some truth to them, the same could be said for retail and hospitality work-
ers. The same arguments had been made in the union-backed campaigns
against Wal-Mart and the associated unionization efforts with little effect.
McDonalds also knew that the economy was recovering and the unemploy-
ment rate was falling, so there were more job openings for breadwinners. If
the unemployment rate continued to drop, there would also be labor market
pressure for higher wages. The best strategy in such an environment is to
maintain a low profile and let the anti-union NGOs such as EPI directly
address the claims of the campaigns. The worker centers could target indivi-
dual restaurants for unionization, but doing so is costly, and the worker
centers, as Manheim concluded, have other issues on the agendas.
The major associations such as the Chamber of Commerce and the res-
taurant and retail associations were active on the issue, and their message
was the hardship a higher wage would have on small businesses and the
resulting loss of jobs. Wal-Mart, which had in the past supported an
increase in the minimum wage, chose not to participate in the debate this
time. Wal-Mart vice president David Tovar said, This time we decided
were going to stay neutral. Elected officials can have the dialogue and ulti-
mately decide what the right thing is to do about this.23 McDonalds
would be affected more by a higher wage than would Wal-Mart, but main-
taining a low profile and leaving the battle to others was the right strategy.
The anti-union NGOs would continue to mount a high-profile campaign.
One uncertainty for McDonalds was whether the unions would back off
after the November 2014 elections. If a movement for a higher minimum
wage were to strengthen after the election, McDonalds could begin to
actively participate in the public politics. Some cities and states would
surely increase their minimum wage adversely affecting McDonalds, but
its competitors would also be affected. Higher prices and lower demand
would result, but the magnitude of the effect was hard to predict.
McDonalds and its franchisees, however, would have stronger incentives
to reduce labor costs. The failure of Democrats in the 2014 congressional
elections took some wind out of the sails of the campaign, but the labor
market continued to tighten and a few cities and states increased their mini-
mum wages. Restaurant wages began to rise.
In terms of an integrated strategy, McDonalds must prepare for an
eventual higher minimum wage and higher labor costs. The principal
market alternative to address higher labor costs is to substitute technology
for labor. The most straightforward opportunity is in order taking. If peo-
ple can book Ubers cars by clicking an app on their mobile devices,
McDonalds customers can order on a touch screen. This makes service
less personal, but after all it is a fast food restaurant.
MARKET AND NONMARKET STRATEGY

Strategic Management
The field of strategic management is based on the value minus cost fra-
mework that focuses on specific assets that generate competitive advantage
because they are costly for others to duplicate and their value is not
competed away. This approach was initiated by Porter (1970) based on the
economics field of industrial organization. The currently popular version
of this approach is the resource-based theory (RBT), which is a theory of
management focusing on resources and capabilities that allow the firm
to generate value and sustain competitive advantage. The logic is the
same resources on which the operations of the firm are based are more
28 DAVID P. BARON
valuable the better is the performance that flows from them and the more
costly it is for others to replicate them. Capabilities are intangible counter-
parts of resources and include factors that are difficult to observe and mea-
sure. Capabilities include how the firm is organized, employee engagement,
the quality of management, leadership style, knowledge, and so on. This
resource-based approach focuses on management, and the step from
resources and capabilities to strategy choice and implementation is not
always immediate. Resource-based theory is informal, making it relatively
easy to incorporate insights from other fields and to generate hypotheses.24
As Foss (2011, p. 1420) notes, Part of the initial success of RBT argu-
ably was that it represented an ingenious way of reconciling the unique and
idiosyncratic with the basic tools of economics. RBT seemingly moved
away from the foundations of industrial organization that underpin
Porters work to more conceptual and behavioral versions. The political
economy approach has remained closer to the micro-foundations of eco-
nomics and political science, so the distance between the two approaches
may have widened.25 Economics and political science, however, have broa-
dened to incorporate behavioral factors such as moral motivation, network
relationships, and leadership. But, those fields remain less able to accom-
modate the unique and idiosyncratic, which can be important in firm-level
applications. By structuring analysis within its framework and focusing on
the firm level, the political economy approach attempts to take into
account the unique characteristics of firms and their market and nonmarket
environments in strategy formulation and implementation.
One approach used in the RBT literature for studying nonmarket strat-
egy is to consider the demand and supply of public policy based on the
Chicago school approach (Becker, 1983; Stigler, 1971) to public politics.
This approach focuses on competition, but it is both reduced-form and
institution-less. More importantly, demand and supply are large num-
ber concepts, which seems inconsistent with a world in which most of
the large number free-ride and only a small number act. This approach
explains some variation in performance, but it is less able to identify the
effectiveness of the strategies used by those on the demand side or the
institutions that supply the policy. Moreover, bringing this approach to
the firm level is a challenge. The competition also depends on the nonmar-
ket issue, and the institutions governing policy. Legislatures, regulatory
agencies, and courts are very different in how decisions are made and
hence in the strategies firms use. Bonardi, Hillman, and Keim (2005),
Bonardi, Holburn, and Vanden Bergh (2006), and Kingsley, Vanden
Bergh, and Bonardi (2012) expand the supply side by incorporating
measures such as the competitiveness of elections, partisan control, and

political constraints at the country level.26
A central research focus in RBT is explaining inter-firm heterogeneity in
performance. This typically is studied empirically through reduced-form
specifications that relate performance to resources and capabilities.
Strategies, however, are based on bundles of resources and capabilities, and
one challenge is to identify how resources and capabilities are combined
and deployed. In addition, competition within an industry has to be taken
into account, and when nonmarket considerations affect performance, the
relevant institutions in whose arenas nonmarket strategies are deployed
must be included in the empirical specification. Structural methods may be
necessary to obtain the level of specificity needed to attribute performance
to individual resources and capabilities and to identify causation.
Nonmarket Strategy
For nonmarket strategy the variation across issue, industry, and country
may be greater than it is for market strategy because of variation in political
systems, the rule of law, and culture. This means that striving for universal
characterizations of the nonmarket environment or strategy prescriptions
may be less productive than in market strategy, so theories are constructed
as much for developing understanding as for strategy choice. For example,
in deciding whether to enter a seemingly attractive country market, political
risk must be taken into account. In countries where the rule of law is weak
or regime change is possible, firms, and particularly foreign firms, may be
subject to hold-up problems. In Venezuela, this has meant expropriation,
but in most countries the practice is more subtle and may involve restrictions
on repatriating profits, currency convertibility, or favoritism of local firms.
Conceptual frameworks are useful for understanding the nonmarket
environment, but a framework should have an accompanying method for
structuring analysis and bringing it to the firm level. Some strategies are
global and are applied uniformly in all countries, but most are tailored to a
particular region, country, market, or in the case of Uber a city. Strategies
are also tailored to specific issues. For example, a firm may use informa-
tional lobbying for a legislative issue and use communication and bargain-
ing when addressing a private politics challenge.
Nonmarket actions in public politics include, for example, lobbying,
participation in regulatory hearings, the filing of a lawsuit, self-regulation
to forestall further regulation, and formation of a coalition to collectively
30 DAVID P. BARON
address an issue. In private politics the actions include, for example, devel-
oping a capability for community interaction, partnering with an NGO, self-
regulation to forestall social activism, and corporate social responsibility
(CSR). The actions may themselves determine outcomes, which is often the
case in private politics, but in public politics the actions are directed at the
institutions governing the outcome. In the case of Uber, the relevant institu-
tions are city councils, the office of the mayor, regulatory agencies, courts,
state governments, and in the case of Germany the national government.
Strategies that are effective in countries where the rules of the game are
clear are often ineffective where those rules are unclear or at times arbitrary.
Effective nonmarket strategies, for example, in emerging markets, have to
address new forms of political risk and an often country-specific set of chal-
lenges. Corruption and lax enforcement of regulations complicate the market
and nonmarket environments that firms face. The wave of outsourcing and
globalization exposed more firms to the challenges of operating in emerging
markets and increased the importance of nonmarket strategy.
In public politics, nonmarket actions are directed at formal institutions
and their officeholders, and the characteristics of those institutions can be
crucial to strategy formation. Decisions are typically made according to
majority rule, so there can be winners and losers. More often than not, no
legislation is enacted on an issue. Legislation in the United States requires
approval in both houses of Congress and the signature of the president.
The possibility of a filibuster in the Senate, a veto by the president, and
divided government means that a supermajority typically is needed to
advance legislation. This means that there is a gridlock interval, and a firm
seeking to preserve the status quo located in the gridlock interval has a con-
siderable advantage. Officeholders in regulatory agencies are arguably
more difficult to influence than are legislators, and courts are designed to
preclude influence activities.
Nonmarket knowledge is needed by all firms, and that knowledge can
reside in management. The logic that governs the nonmarket environment,
however, is different from that governing the market environment because,
for example, in public politics policies are chosen in institutions that use
majority rule instead of the unanimity rule used in markets. Experience can
be an important teacher in public and private politics, and it may take time
for managers to develop the needed knowledge. To some extent that
knowledge is available in the marketplace. Firms can hire political advisors,
specialists in nonmarket reputation-building, and lobbyists for access to
institutional officeholders. If a firm must address nonmarket issues on a
regular basis, it can bring the knowledge inside the firm by hiring people
with expertise or by using advisors. Uber, for example, hired expertise to

address municipal regulations and to help formulate and implement a
broader nonmarket strategy pertaining to states and countries. It also
engaged a privacy expert as a legal advisor. Amazon hired President
Obamas former press secretary, Apple hired the former head of the
Environmental Protection Agency in the Obama administration, and
Google hired a former congresswoman.
Nonmarket strategy ultimately involves integration with the market
strategy of a firm, but the logic is somewhat different from that in RBT.
The principal nonmarket asset for public politics, and particularly for dis-
tributive politics, is a firms rent chain, and the principal capability is
knowledge. In contrast to RBT that views resources as available in markets
and intangible capabilities as firm-specific and internally developed, the
rent chain is costly to develop and replicate, whereas knowledge and exper-
tise, while intangible, are learnable and available in the marketplace.
The rent chain is composed of all parties that earn rents or benefit from
the operations of a firm. The rent holders are constituents of legislators,
and regulators are often charged with taking their interests into account in
their rulings. Employees, suppliers, and local communities in which facil-
ities are located are in the rent chain, as are customers if they have no close
substitutes for the good or services provided by the firm. Pharmaceutical
companies often use patient advocacy groups to pressure the FDA to
approve new drugs.
A firm with a limited rent chain is at a disadvantage in distributive poli-
tics, but it may have the opportunity to form coalitions with other firms or
interest groups. If a firm does not have a significant rent chain and coali-
tions are costly to form and maintain, it may have to rely on a strategy
emphasizing the risk from change (if it seeks to preserve the status quo) or
the risk that the industry will fall behind international competitors that are
less restricted. The major money center banks have relatively small rent
chains, but the risk of disrupting credit markets by intervening in complex
and interconnected financial markets makes reform difficult and favors the
status quo. Even after the financial crisis and the reform legislation that
eventually passed, the Volker rule took three years to promulgate, albeit in
a weakened form.
The focus on resources and capabilities in RBT means the acquisition of
resources and the development of capabilities define the boundaries of the
firm. An alternative view is that boundaries are determined by the comple-
mentarities among activities (Roberts, 2004). Activities with complementa-
rities should be within the firm, and otherwise the activities should be
32 DAVID P. BARON
outside the firm and hired as needed. The rent chain includes both those
within the boundaries of the firm and rent holders outside the firm.
Although a firm without a significant rent chain does not have a constitu-
ency base that can be used for access, it may be able to hire access in the
marketplace for political capabilities. If access for the purpose of lobbying
is crucial for success on a nonmarket issue, a firm can prefer a broader to a
narrower rent chain and may select suppliers for their contribution to
access in addition to commercial reasons. If a firm is vulnerable to social
pressure through private politics, it may choose to reduce its footprint and
shift pressure to suppliers or retailers.
The principal nonmarket capabilities in private politics stem from reputa-
tion, and the principal liability is the relatively low trust the public has in
business and particularly big business. Reputation can be strengthened by
innovative products, for example. Apple has a very favorable reputation,
and allegations of poor working conditions at factories of one of its principal
suppliers, Foxconn barely dented its reputation. Self-regulation and corpo-
rate social responsibility can also strengthen a firms nonmarket reputation.
The field of nonmarket strategy sits somewhat uncomfortably with RBT
and the field of strategic management. Nonmarket strategy is chosen by
firm leaders, but it is not general management. As discussed above, strategic
management, and particularly the resource-based view of it, focuses on
management, which encompasses all the activities and components of a
firm. As such it necessarily operates at an aggregate level and focuses on
broad issues, leaving particular aspects, such as finance, operations manage-
ment, and marketing, to more specialized fields. The market strategy com-
ponent of strategic management focuses on the sources of competitive
advantage and sustainable performance and addresses issues involving the
lines of business in which the firm operates, the markets in which it partici-
pates, its rivalry with competitors, investments in innovation, organizational
boundaries, and so on. That is, market strategy focuses on positioning and
choice in markets and on the performance of the firm in those markets.
Nonmarket strategy also focuses on choice among strategy alternatives, not
in markets but rather in the context of public and private politics arenas in
which the firm competes against, and at times works with, its rivals, interest
groups, and social activists. Firms and other actors in the nonmarket envir-
onment choose nonmarket strategies, but the outcomes stemming from
those choices are typically determined in government institutions and
amongst the public. Nonmarket strategy thus focuses on choice, but like
market strategy it is not general management itself, but instead is an activity
in which managers address challenges and opportunities.
Henisz and Zelner (2012, p. 40) lament that nonmarket research has
failed to coalesce. They write, the diversity of research approaches
employed, together with the infrequency of cross-citations by scholars
working in different traditions, has hindered the fields coalescence.
Coalescence, however, is not necessary for a field to advance. An alterna-
tive perspective is that there is a long-distance horse race among alternative
approaches with each innovating, progressing, and bringing with it a
knowledgeable base for strategy formulation.
For nonmarket strategy to assume a more important position in the
broader field of strategic management, however, it must speak to the scho-
lars in that field. That remains a challenge, particularly for the political
economy approach. For example, private politics is important because of
the direct impact it can have on firm performance, but more importantly
because its threat can induce self-regulation. Self-regulation is also impor-
tant in public politics to forestall legislation and more restrictive regulation.
Self-regulation arguably is pervasive, yet social pressure and self-regulation
receive relatively little attention in the field of strategic management.
Self-Regulation
Self-regulation refers to voluntary actions by firms to mitigate the likelihood

that private or public politics will impose burdens, such as more stringent
regulations, higher taxes, or expanded duties. The burdens may come from
the actions of government or from social pressure outside the institutions of
government. Social pressure is the instrument of private politics and is direc-
ted at firms to cause them to change their policies. Self-regulation is under-
taken by individual firms and by industries. Self-regulation can forestall
public politics or private politics, or lessen their intensity. In public politics,
a firm or an industry can self-regulate to the boundary of the gridlock inter-
val, and the legislature is then unable to change it further.27 In private poli-
tics, a firm can self-regulate to the point at which an activist is indifferent
between incurring the cost of conducting a campaign and accepting the self-
regulation and not campaigning (Baron, 2014, 2016).
The most noticeable aspects of private politics are the campaigns con-
ducted against firms, but the greater effect is from the threat of private
politics. That is, many more firms are threatened by confrontational acti-
vists than can be targeted, and many of these firms self-regulate to forestall
a campaign even though they recognize that the probability they will actu-
ally be targeted is not high. Firms must self-regulate ex ante, which gives
34 DAVID P. BARON
activists leverage, since more firms are threatened than can be targeted with
a campaign. Those firms that prefer to forestall a campaign must do so in
advance of the choice of targets by activists. So to forestall a campaign, a
threatened firm must self-regulate to the point at which, if targeted, the
activist will accept its self-regulation rather than campaign. Self-regulation
forestalls a campaign because it allows the activist to avoid the cost and
possible failure of a campaign. The scope of activism, that is, the set of
firms threatened, is increasing in the saliency of the social issue, since the
activist then has a stronger incentive to campaign. A campaign against a
more vulnerable firm is more likely to succeed and hence more costly to
forestall, and some soft firms can find forestalling a campaign too costly
and choose to incur a campaign. In that case, they may self-regulate to
reduce the probability that a campaign succeeds if they are targeted.
Radical activists pose a greater threat than do moderate activists, so the
scope of activism is greater the more radical are the activists (Baron, 2016).
Some firms, however, are too hard for activism and are not threatened.
The reach of private politics thus has limits.
Gupta and Innes (2011) study the effect of boycotts on the adoption of
environmental management systems (EMS) by firms. They find that soft
firms as measured by KLD strengths for social responsibility are signifi-
cantly more likely to be the target of a boycott and proxy resolutions than
are hard firms as measured by KLD concerns. Confrontational activists
prefer to target more vulnerable or softer firms, since less vulnerable or
harder firms are either too hard to target or self-regulate sufficiently to
avoid a campaign. Soft firms must self-regulate more to avoid a campaign
than must hard firms because activists expect to obtain more from a soft
than a hard firm if a campaign is launched.
An alternative to this self-regulation is partnering with a cooperative
activist. A cooperative activist, however, has bargaining power to the
extent of the threat by confrontational activists. That is, if the cooperative
activist can provide a shield against confrontational activists, it can extract
concessions from the firm with which it partners (Baron, 2012).
Cooperative activists thus may be able to accomplish as much as can con-
frontational activists. Cooperative activists can prefer to partner with soft
rather than hard firms.
The impact of activism thus is at two levels. One is the impact of the
campaigns that are actually launched, some of which succeed and some of
which fail. The larger impact may be from the leverage on firms that are
threatened and choose to self-regulate to forestall the activists, reduce the
probability that a campaign succeeds, or partner with a cooperative
activist. This self-regulation results in smaller changes in practices than

those obtained from a successful campaign, but there are many more firms
that make the changes. This self-regulation is often given the label of cor-
porate social responsibility.
Corporate social responsibility can have three types of motivation. The
first and most fundamental is moral, where a firm acts when it is best placed
to address a moral wrong. For example, a manufacturer or a firm dealing
with a supplier may be best placed to address hidden hazards in the work-
place. Under social pressure Nike set a minimum age for employees in its
suppliers factories and chose a two-year higher minimum age for shoe
factories than for apparel factories. The reason is that the manufacture of
shoes requires the use of adhesives and other chemicals whose potential
toxicity may be unknown to workers and difficult to observe in the work-
place. Those chemicals are not used in the production of apparel. The
second motivation is strategic and intended to maximize the value of the
firm. The extent of this strategic corporate responsibility is determined by
equating the marginal return from the action to its marginal cost.28 The dis-
tinction between the first and second motivations does not mean that there
are no potential returns for Nike from its higher minimum age. Instead, it
means that Nike is to take the action regardless of whether there are any
returns. The third is to label self-regulation taken to forestall or mitigate pri-
vate or public politics as social responsibility. Rather than equating mar-
ginal returns and marginal costs, forestalling CSR involves acting to the
point at which the opposition is just willing not to act. The second and third
forms of CSR involve choices, whereas the first involves fulfilling a duty.
RESEARCH
Theory
A wide variety of private politics theory questions remain to be investi-

gated, and two are addressed here. The first is extending the models of acti-
vist and firm encounters to include agency issues on the part of firms. This
could focus on managers allocation of effort between private politics stra-
tegies and activities that more immediately yield a financial return. The
allocation of effort to private politics could be disaggregated to allow for
effort allocated to general reputation-building to reduce the harm that
could be generated by a campaign and the effort to forestall a campaign
36 DAVID P. BARON
through self-regulation. Self-regulation and reputation-building that gener-

ate accolades from NGOs and the public could not only benefit the firm in
private politics but could also represent a perquisite to managers that value
the recognition.
On the other side of private politics an important research issue is how
activists choose their targets and whether they prefer to target firms individu-
ally, sequentially, or as an industry. Baron (2016) distinguishes among firms
based on their vulnerability to an activist campaign and shows that low vul-
nerability firms forestall campaigns whereas high vulnerability firm may
choose to incur a campaign because forestalling the activist is too costly.
The second is the study of firm versus industry strategies for private poli-
tics.29 Some firms such as McDonalds act separately on private politics
issues, whereas others act through their industry or allow an association to
represent their and other firms interests. This requires recognition that
firms are heterogeneous. Acting as an industry requires addressing collec-
tive action issues as well as how a collective strategy would be chosen by
heterogeneous firms. For example, do the firms choose strategy by majority
rule, unanimity, or weighted majority? A unanimity rule would attract
more participants than would simple majority rule, but the standard chosen
would be lower and would be less effective in deterring private politics.
More generally, the choice of a governance structure affects which firms
choose to participate in the collective action and how effective the collective
action is.
Public politics theory is a flourishing research field in economics and
political science and provides a good understanding of institutional beha-
vior and outcomes. What is needed is to better characterize the strategies of
firms and other interest groups and to relate those strategies to outcomes.
Much of the research on nonmarket strategy has been in the context of
U.S. institutions, with less attention given to parliamentary systems. One
reason is that in parliamentary systems much of the nonmarket strategy
implementation takes place behind closed doors and out of the view of
the public, since policy choice in parliamentary democracies is party- or
coalition-controlled. In Westminster systems with first-past-the-post elec-
toral systems two-party competition tends to result with one party having a
majority in parliament. Since the parliament chooses the government and
the government is composed of a single party, it can choose any policy it
wants provided that it can maintain party discipline, which is often the
case because candidates for elections are chosen with strong party influ-
ence. The same is true in parliamentary systems with proportional represen-
tation electoral systems, although the mechanism is somewhat different
because more parties are present and typically no party wins a majority in
an election. Coalition governments then result, and policy can be thought
of as resulting from bargaining to form a government and, once formed,
among the coalition members subject to the threat that a party could quit
the government causing it to fall. Proportional representation electoral
systems often use list systems which give parties the power to choose the
candidates they put on the election list and the order in which they are
placed, resulting in strong party discipline. Nonmarket strategy in parlia-
mentary systems thus focuses on the majority party or government
coalition, and as such it takes place outside of formal institutions.
Research is needed on influence strategies in parliamentary systems,
and in particular how firms use lobbying and information provision to
influence policy choice.
Congressional elections in the United States have resulted in the van-
ishing center, meaning that Congress is polarized with few members with
centrist policy preferences. Theories of policy-making based on the median
voter theorem thus seem inappropriate for a world with two ideologically
separated blocks, neither of which has the votes to meet the supermajority
hurdles. This suggests that bilateral bargaining with an eye of the next elec-
tion may be a better framework than the median voter theorem for predict-
ing policy outcomes. This in turn means that theories of influence, such as
those that focus on pivotal voters, may need to be revised. It also means
that the status quo is more secure, and addressing emerging issues may be
more difficult.
Theoretical research is particularly needed for environments in which
the rule of law is weak or the government is authoritarian or single party.
Many of the market and nonmarket environments in emerging markets
have these characteristics. Assessing political risks and providing contin-
gencies for adverse government actions is important for decisions about
whether to enter these markets and how to manage once there. Both theory
and empirical work are needed, and detailed case studies can identify fac-
tors that need to be taken into account. The wave of globalization may
have peaked, but the choice of nonmarket strategy for firms operating in
multiple national market and nonmarket environments with varying
strengths of the rule of law remains an understudied research topic.
A central nonmarket strategy question is when a firm should delegate
nonmarket action to a general business association such as the Chamber
of Commerce or the National Association of Manufacturers, when it
should rely on an industry association, when it should form an ad hoc
coalition, and when it should act individually. These alternatives are
38 DAVID P. BARON
not mutually exclusive, and on some issues such as international trade

agreements firms may engage in all of them. One answer is simply to
apply the methodology for nonmarket strategy choice, but guidance from
theory or empirical evidence about which nonmarket issues and non-
market environments call for particular types of strategies would be helpful
to focus the analysis.
A natural extension of extant theory is to dynamics and the durability of
policies and strategies. As de Figueiredo (2009) discusses, policy durability
can differ between parliamentary and presidential systems. In parliamen-
tary systems policy is durable during an inter-election period but can
change as a result of an election. In a presidential system policy is deter-
mined by bargaining between the president and Congress and within
Congress under the constraint that a supermajority may be required and
that the bargaining has to take place with both chambers of Congress,
which may have different partisan control. This can give rise to a gridlock
interval, and once in the interval a policy is durable and elections that leave
the gridlock interval intact then have little effect on the policy.30
A policy in the gridlock interval is stable if the political environment is
stable, but elections and changing circumstances can result in changes in
that interval and in policy. Incorporating uncertainty is not as straight-
forward as it might appear, particularly if there are exogenous shocks and
changing events. For example, policy choice in a parliamentary system with
proportional representation elections is stable if the supporting government
coalition is durable, but exogenous events and crises can lead to changes in
the government and in policy.31 If the firm is affected by a change in
government policy, its nonmarket strategy may have to change. Dynamic
nonmarket strategy theory has yet to be developed, and dynamic political
economy theory is in its early years. Understanding these dynamics and
choosing strategies that are robust to change or can adapt to change is an
important research topic.
Empirical
Data and Methods

Empirical research that characterizes the nonmarket environment in a
country or a region is useful in identifying factors to take into account in
strategy choice. Cross-sectional studies, however, can be difficult to
interpret because of unobserved heterogeneity among the countries or
regions. Unobserved heterogeneity in a cross-sectional study can be along a

number of dimensions. The nonmarket issues can differ, institutions can
differ, the set of active participants in the politics can differ, and their char-
acteristics can differ and be difficult to observe and measure.
One approach to reducing unobserved heterogeneity is to restrict atten-
tion to issues characterized by client politics, so that only firms on one side
of the issues are active. Alternatively, an empirical study could focus only
on issues characterized by interest group politics, where interests on each
side of the issue are active. Outcomes in interest group politics are jointly
determined by the strategies of all players, and outcomes reflect equilibria
of the competition on each issue. Since outcomes are jointly determined,
identifying the effect of individual components of an integrated strategy is
a challenge. This is a classic problem of joint production, and structural
methods may be the best hope.
Another approach is to study the strategy choices of a single firm. For
example, Ingram, Yue, and Rao (2010) examine Wal-Marts attempted
store openings to identify location-specific public and private politics
factors that affect success or failure.32 Similarly, Locke, Qin, and Brause
(2007) study Nikes supplier factory inspection data to assess compliance
with the companys workplace standards. A broader approach is to develop
data on a panel of firms and use firm-fixed effects to control for unobserved
time-invariant characteristics of the firms. Then, within variation can be
used to examine firms responses to changes in the nonmarket environment.
Finding data remains a challenge, however.
A particular challenge is observing the strategy choices of firms. In mar-
ket strategy studies, accounting data are available, and data on prices and
quantities are also available for some industries and firms. Accounting data
are not particularly useful for drawing inferences about nonmarket strate-
gies, although some firms now report on their expenditures on environmen-
tal programs. Surveys and rankings can serve as a data source, but some,
such as the Fortune 100 most-admired firms, require firms to apply to be
considered, leaving an endogeneity problem. If a firms rent chain is an
important nonmarket asset, it is necessary to obtain data on where the
firms employment is located and where its suppliers are located. Few firms
make such data public.
Consistency in the data is also a major challenge. Some firms hire out-
side lobbying firms to implement an informational strategy in public
politics, and their expenditures are publicly reported and available on open-
secrets.org. Other firms use their executives for lobbying, and unless the
40 DAVID P. BARON
executives spend a substantial portion of their time on lobbying, their

efforts are not reported. This omitted variable confounds the assessment of
the effectiveness of lobbying strategies. This is a particular problem in
cross-sectional studies, and controlling for variables such as assets, size,
advertising expenditures, and other reported accounting measures is helpful
but does not remove the omitted variable problem. Attributing outcomes
to a measurable component of a nonmarket strategy, such as lobbying or
campaign contributions, thus is difficult.33
The government does not require reporting on private politics as it
does for public politics, so data sets must be developed directly. Data on
CSR is collected by KLD, now a part of MSCI, which has spurred a large
research program, but data on social pressure, the responses to it, and
self-regulation in anticipation of it are more difficult to obtain. One source
of data is media coverage of events such as demonstrations, environmental
accidents, lawsuits, and the like. Lenox and Eesley (2009), Eesley and
Lenox (2006), Gupta and Innes (2011), Minor (2013), and Kruger (2015)
have developed databases using media searches.
Causation is the most elusive issue in most empirical studies. To illus-
trate the issue, consider regressing the market value of firms on their KLD
strengths minus concerns, controlling for other factors using data from
COMPUSTAT. Suppose the coefficient is positive. One interpretation is
that a higher KLD score causes the market value of the firm to be higher,
and another interpretation is that a higher market value gives management
the resources to spend on CSR. A third possibility is that there is there is
no causal relation between CSR and financial performance and that the
positive correlation is due to unobserved firm characteristics. For
example, some managers may view CSR, and the accolades that can
accompany it, as a perquisite, and those managers, for whatever reason,
might be better than others at extracting value from the assets of a firm.
They might be more aggressive in risk-taking, which might cause the
higher market value. Selecting a closely matched control group or using a
difference-in-differences or synthetic control approach on time series data
can address some but not all concerns. Harrison and Scorse (2010) use the
differences-in-differences approach in studying the effect of the anti-
sweatshop campaign on wages in suppliers factories in Indonesia.
Aggregation is also an issue. Studies that use a count of KLD strengths
minus concerns as a measure of the net goodness of the firm or the net
contribution to firm reputation are problematic because strengths and
concerns represent different forces. Strengths represent CSR policies that
strengthen reputation, brand equity, employee engagement, etc., whereas
most concerns are factors that give rise to social pressure in either private
or public politics. For example, protests are a direct indication of private
politics social pressure, and environmental concerns are an indicator of
possible government enforcement measures, protests by environmental
NGOs, or liability exposure. These can impair reputation and employee
engagement, and they also can directly reduce market value independently
of reputation and engagement by, for example, alienating customers, busi-
ness partners, or investors. Social pressure thus should decrease market
value independently of reputation, whereas strengths should increase mar-
ket value if reputation and engagement matter. Moreover, if CSR is a per-
quisite of management, it should be increasing in market value or other
measures such as free cash. In addition, KLD strengths could represent a
resolution of social pressure or self-regulation in anticipation of social pres-
sure. Strengths then would be increasing in concerns and concerns decreas-
ing in strengths if the strengths rectify a source of social pressure. A
simultaneous equation empirical model is required to investigate these
effects. Baron, Harjoto, and Jo (2011) estimate a simultaneous equations
model that takes social pressure into account by using KLD strengths as a
measure of CSR and KLD concerns as a measure of social pressure. They
find that financial performance is decreasing in social pressure, whereas
CSR has no significant effect on financial performance. CSR is increasing
in social pressure, but social pressure is increasing in CSR. The latter may
be due to activists targeting soft firms, as Argenti (2004) suggests and
Gupta and Innes and others find.
Public Politics
Researchers naturally look for readily available data, which in the case of
public politics includes campaign contributions by corporate PACs and
expenditures for registered lobbyists. But, there is little consensus about the
purpose of corporate PAC contributions or about the relation between
contributions and lobbying expenditures. Empirical studies (Ansolabehere,
de Figueiredo, & Snyder, 2003) find little evidence that contributions affect
voting by members of Congress. The most common explanation for the
contributions is to gain access for the purpose of (informational) lobby-
ing.34 An important reason for hiring a registered lobbyist is for connec-
tions that yield access, so a natural hypothesis is that the greater are
corporate PAC contributions to members of Congress the less likely the
firms are to hire a registered lobbyist for access purposes. However, if cor-
porate PAC contributions are so important for access, why do a third of
the Fortune 500 firms not have a PAC? Are these firms more likely to hire
42 DAVID P. BARON
registered lobbyists than firms that make contributions? Also, do these

firms not make contributions because they are little affected by legislation,
have chosen to free-ride, rely on associations for their influence strategies,
have learned that contributions have little effect on outcomes, or something
else? The access value of PAC contributions to candidates campaigns may
be less, or perhaps more, important after the Michigan United ruling that
allows unlimited spending in election campaigns by unaffiliated groups.
de Figueiredo and Edwards (2007) take a different approach and esti-
mate the effect of contributions on policy outcomes rather than on indivi-
dual votes of regulatory commissioners. The policy issue studied is the
price for access to local telecommunications networks, and the contributors
are from local telecommunications companies and potential entrants and
are made to state legislators. They find a robust negative effect of contribu-
tions on access prices for local networks. Studying outcomes rather than
votes is a useful approach and may allow identification of both an agenda
setting effect and a voting effect of political contributions.
Studies of hard money contributions to candidates election campaigns
focus on the induced behavior of officeholders, but those contributions
are used by candidates to improve their election prospects. If greater
contributions increase the probability of winning an election, an alternative
hypothesis is that firms make contributions to candidates they want to win.
Cooper, Gulen, and Ovtchinnikov (2010) study the relation between contri-
butions and election outcomes and find that there is positive relation
between firm financial performance and the number of candidates sup-
ported through contributions. One complication is that most contributions
go to incumbents and incumbents have a high probability of being
reelected. The relation between supported candidates and financial perfor-
mance is stronger the longer the period of support for a candidate and the
more powerful within Congress is the candidate. This suggests that the cau-
sation runs from the likelihood that a candidate will win to contributions.
Characterizing the nonmarket environments of countries is an important
research topic, and political risk is at the center of that topic. Aggregate data
can be used to assess the effect of the rule of law on foreign direct investment
and firm performance. Much of this research relies on characteristics of the
government system and institutions that protect property rights. An alterna-
tive is to rely on market data to assess the political risks present in a country.
Jensen (2008), for example, uses the prices of political risk insurance to
measure risks and relates the prices to the government institutions in a
country. He finds that democratic institutions are correlated with lower poli-
tical risk, and the most important features are constraints on the discretion
of the executive. Utilizing such market data can deepen the understanding of
both political risks and the effect of institutions on those risks.
Delving deeper into political risk in a country, Graham, Johnston, and
Kingsley (2013) identify a relatively prevalent form of political risk asso-
ciated with government restrictions on the repatriation of profits. They find
that transfer risk is more of a deterrent to foreign direct investment than is
the risk of expropriation and that there is less protection from the rule of
law for transfer risk than expropriation. A deeper understanding of the nat-
ure and magnitude of such political risks is an important research topic.
Most research on nonmarket strategy focuses on established firms, yet
innovation often begins with entrepreneurs that have new ideas for value
creation. Some entrepreneurial firms are able to operate without a nonmar-
ket strategy or to free-ride on the efforts of others, but many others face
nonmarket issues early in their lives. As considered in the section Public
Politics Example: Uber, Uber immediately faced regulations governing
taxi and limousine service, and taxi companies and their drivers worked for
more stringent regulations to limit its penetration of their markets.
Similarly, Tesla faced a host of local regulations on the marketing of its
automobiles, and well-organized auto dealers sought to erect new barriers.
Amazon and Airbnb initially were able to avoid collecting sales and value-
added taxes, but nonmarket forces developed to force collection, reducing
their competitive advantage over traditional retailers and hotels, respec-
tively. The study of nonmarket strategy for entrepreneurial firms is in its
infancy, providing many research opportunities.
Private Politics
Self-regulation has been studied theoretically, and self-regulation as
CSR has been studied extensively. Empirical research, however, on self-
regulation to forestall private and public politics is scant. An open question
both theoretically and empirically is the extent to which self-regulation by
an individual firm is effective or whether an industry has to self-regulate to
cause social pressure to dissipate. Malhotra, Monin, and Tomz (2014) con-
ducted an online experiment with members of the Audubon Society to
assess the effect of firm and industry self-regulation on members prefer-
ences for stricter government regulation. They found that environmental
self-regulation by a firm has little effect on preferences but self-regulation
by an industry has a significant effect. They conclude that the breadth of
the participation by companies is more important than the depth of
measures taken by firms. Such field experiments are an underutilized
approach to the empirical study of nonmarket strategy.
44 DAVID P. BARON
Baron, Neale, and Rao (2016) conducted an online experiment to

examine the matching of cooperative and confrontational activists with
profit-maximizing and a CSR firm. They find that subjects in the role of the
cooperative activist prefer to partner with a CSR firm, and anticipating
this, the profit-maximizer bids more for an engagement with the cooperative
activist than does the CSR firm.
An important line of research focuses on identifying the consequences of
intangibles such as reputation or policies such as corporate social responsi-
bility. Nyborg and Zhang (2013) find that firms in Norway with good repu-
tations for social responsibility paid lower wages than firms without such a
reputation. Similarly, Burbano (2014) and Cassar (2014) conduct Internet
and laboratory experiments, respectively, that show that workers are will-
ing to accept lower wages when they work for socially responsible firms.
This research cannot establish whether there was a selection effect where
the workers attracted were more or less qualified than the average, nor
does it indicate whether employee engagement was higher among the work-
ers attracted or whether those attracted were distracted by their interest in
social issues. Burbanos experiment, however, suggests that the workers
who accepted lower pay were more productive than other employees.
Further research on the effect of intangibles on employee motivation and
productivity would be worthwhile.
A major empirical challenge is to identify causal relations such as that
between CSR and financial performance. Insight can be obtained by disag-
gregating the data to examine more fine-grained relations. Minor (2013)
views CSR as providing protection against adverse events and distinguishes
ex ante CSR investments that reduce the likelihood of adverse events from
CSR actions that primarily bolster reputation. Using product recalls as the
adverse events, he finds that CSR investments are associated with the
avoidance of substantial losses from the events, whereas general CSR that
bolsters reputation provides no discernable protection. The latter finding is
consistent with the hypothesis that some CSR spending is motivated by
warm glow preferences of managers and allowed by agency problems.
Baron et al. (2011) find that CSR is increasing in social pressure suggesting
that firms respond to social pressure by increasing their CSR. Kotchen
and Moon (2012) also find that firms with poor CSR ratings in the KLD
categories of environment, community, and human rights respond by
improving their CSR on those issues. The relations identified in these
studies are best viewed as correlations, since there could be unobserved
characteristics of firms that affect CSR, financial performance, and
social pressure.
One unobserved characteristic is management. Hong and Minor (2013)

attempt to identify the effect of individual managers by first estimating a
manager fixed effect and then using it as an explanatory variable for finan-
cial performance and social performance. They find a negative relation
between managers effects on financial performance and on social perfor-
mance and that social performance is greater for managers with a smaller
share of incentive compensation. The latter finding is consistent with man-
agers having warm glow preferences for CSR, and the former is consistent
with financial and social performance being substitutes in production.
The effect of social pressure or CSR on firm performance is likely to be
issue-specific. In an event study, Epstein and Schnietz (2002) found that
firms targeted during the 1999 WTO protests in Seattle for their environ-
mental practices had significantly lower abnormal returns, whereas firms
targeted for their labor practices experienced normal returns. Kruger
(2015) studied the response of the financial markets to positive and negative
events pertaining to firms and found that negative events associated with
the environment or communities were associated with significantly negative
abnormal returns, but negative events for four other KLD categories had
no effect. Positive events were associated with small decreases in returns,
from which he concludes that the market was penalizing warm glow CSR
allowed by agency problems.
In such studies establishing causation remains a challenge. One informa-
tive approach is to ask whether the entry into or exit from a CSR index of
firms with strong social responsibility has an effect on performance.
Chatterji and Toffel (2010) and Becchetti, Ciciretti, Hasan, and Kobeissi
(2012) take this approach and find that firms exiting the KLD social index
experienced a significant negative abnormal return. Hong and Kacperczyk
(2009) report a similar finding. These findings have one or both of two cau-
sal interpretations. One is that exit from an index has information content,
and that content then must be that CSR and financial performance are
positively related. The other is that social investing has become sufficiently
pervasive that firms with worse CSR have lower demand for their shares
and hence lower prices.
Kruger (2015) argues that his multiple-event study allows the direction
of causality to be determined. Event studies should be free from reverse
causality, but those studies typically draw inferences from a single event.
Kruger collects data on over 2,700 events and relates each to one of six
KLD categories. He also disaggregates the events into those that are posi-
tive and those that are negative. He in effect conducts event studies on all
the events and finds that negative events relating to the environment and
46 DAVID P. BARON
communities are associated with strongly negative abnormal returns,

suggesting that negative events on those issues cause a decrease in returns.
This approach appears to be a fruitful way to address causality. Delving
into the relation between a financial and social performance remains a
fertile ground for research with the identification of causal effects the
principal challenge. Distinguishing among the various motivations for CSR
would be a major research accomplishment.
Incomplete information is a fundamental characteristic of the nonmar-
ket environment, particularly with respect to the social dimensions of a
firms operations. Certification of a firms environmental practices or a
cooperative engagement with an NGO can provide a signal to consumers,
business partners, and investors with social preferences. Determining the
value of the signal is a considerable challenge. Some evidence is provided
by the entry and exit into indices, but empirical research relating financial
performance to firms environmental practices that are and are not certified
would be useful.
Minor and Morgan (2011) and Becchetti and Ciciretti (2011) study the
role of CSR as insurance against adverse events, and CSR could also serve
as insurance to mitigate private politics pressure. For example, the public
could respond less to a campaign against a firm with a good reputation for
social responsibility than a campaign against a firm with a poor reputation.
This suggests that the ex-ante threat of private politics is less severe and
targeting by activists less likely. Theory (Baron, 2016), however, suggests
that a firm that is less vulnerable to private politics self-regulates less than
a more vulnerable firm. Whether activists target more or less vulnerable
firms and why remain attractive research questions.
The theory discussed in the section Self-Regulation predicts that firms
anticipate social pressure and (ex ante) self-regulate to forestall it or to
mitigate the probability of campaign success. If a firm is actually targeted
with a campaign, (ex post) self-regulation can be viewed as the outcome of
bargaining between the activist and the firm to resolve and end the cam-
paign. Empirical research that distinguishes between ex ante self-regulation
motivated by a threat and ex post self-regulation that results from the reso-
lution of a campaign would deepen our understanding of private politics
and self-regulation.
Firms and industries seek shields against private politics. Shields can be
provided by three principal means: (1) self-regulation, (2) cooperative part-
nering, and (3) multi-stakeholder organizations. Little is known about the
effectiveness of each of these for particular issues. Moreover, cooperative
partnering may provide a shield from moderate activists but not from
more radical activists, who may seek greater changes in the practices of
the firm. One purpose of industry-wide private governance organizations
is to strengthen their resistance to social pressure and prevent defections
so as to reduce the likelihood that the industry will be targeted and, if
targeted, to reduce the likelihood of success. The Sustainable Forestry
Initiative (SFI) has forestalled some pressure by the activists backing the
Forest Stewardship Council (FSC), but those activists continue to criticize
the SFI and work to convince retailers to sell only FSC-certified products.
If activists are impeded by a shield, they can use low cost alternatives
such as criticism through social media, although those alternatives likely
produce less social pressure. Similarly, in public politics a legislature may
be in gridlock, but officeholders can still press regulators to use their discre-
tion on a nonmarket issue. The effectiveness of a shield as a function of
how far a firms or industrys policy is from the demands of moderate and
more radical activists and what alternative sources of social pressure are
used are important research topics in private politics.
Normative
Although the perspective taken in this essay is positive, an important

normative question is whether private politics and the associated self-
regulation address a market or a government failure or whether the poli-
tics are the result of the intense preferences of segments of society.
Private politics can go beyond the economists concept of social efficiency
and embrace distributive and justice objectives, and activists can seek
more stringent regulation than justified by efficiency considerations. An
environmental standard that equates marginal social costs and marginal
social benefits achieves the socially efficient level of emissions, but the
remaining emissions could still pose a risk to the environment or human
health. Activists could seek to reduce or shift that risk even though it
would be socially inefficient to do so. That is, at the socially efficient level
of emissions, there remains uncompensated harm that can motivate pri-
vate or public politics actors. Data show that minorities and the poor are
over represented in areas where the remaining emissions are the highest,
raising environmental justice issues. Wolverton (2009) studied the original
siting of polluting facilities and found that minorities and the poor were
not disproportionately represented when the facilities were built. Instead,
48 DAVID P. BARON
facilities were sited where land was cheap and where other facilities
emitting toxic substances were already located. The low land prices subse-
quently attracted low cost housing, and the low cost housing attracted
low-income individuals, among whom minorities are disproportionately
represented. Whether this is the case more broadly warrants
further research.
More fundamentally, an important question in private politics is what
activists of various types want. Some may want social efficiency, but others
have distributive objectives, as evidenced by the campaign for a higher
minimum wage. The book Good Cop, Bad Cop edited by Thomas Lyon
(2010) provides some insight into the preferences and strategies of activists,
but rhetoric and reality can differ. Studies of the preferences of activists
and NGOs are needed to help assess whether their actions move firm prac-
tices toward or away from social efficiency.
Another approach to the question of whether private politics addresses a
government failure is to view a government failure as a political failure
rather than as a welfare issue. Gridlock could be viewed as a government
failure, and the question is whether social pressure-induced self-regulation
is a substitute for a blocked government action. Self-regulation can result
from either public or private politics as discussed in the section Self-
Regulation, with public politics-generated self-regulation moving policy to
the boundary of the gridlock interval but not into the interior. Private poli-
tics, however, can move policy into the interior of the gridlock interval. It
also can thwart the will of the majority by targeting individual firms that a
majority prefers not having targeted, since even small interest groups can
prevail in private politics.
A more limited question is what private politics can do more or less
effectively than government. For example, private politics might be more
effective at targeting an individual firm (public laws cannot single out a
firm or individual by name and hence can only legislate or regulate by char-
acteristics; private politics is not so-constrained), tailoring campaigns and
self-regulation to the specifics of a firm and its situation (avoiding one size
fits all), monitoring local compliance with regulation or cross-border self-
regulation, yielding results more quickly than government, and circumvent-
ing gridlock. Public politics might be more effective at enforcement,
measuring compliance and identifying hazards, covering an industry or
alike firms (with respect to some issue), regulating firms that are too hard
for private politics, subsidization (e.g., clean energy), and redistribution
(higher minimum wage).
NOTES
1. Integrated strategy is considered by Baron (1995, 2001), de Figueiredo

(2009), and Bonardi and Vanden Bergh (2015).
2. This paper does not review the research in the field. de Figueiredo (2009) and
Hillman, Keim, and Schuler (2004) provide reviews. Margolis, Elfenbein, and
Walsh (2007) and Orlitsky, Schmidt, and Rynes (2003) provide assessments of the
empirical research on the relation between corporate social and financial perfor-
mance. The citations used here are illustrative and do not necessarily represent the
most important contributions.
3. Early examples include Baysinger (1984), Keim (1981, 1985), Keim and
Zeithaml (1986), Keim, Zeithaml, and Baysinger (1984), Yoffie and Bergenstein
(1985), and Yoffie (1987).
4. For example, for strategic management Besanko, Dranove, Schaefer, and
Shanley (2013) provide economics-based models of particular strategy problems
and identify relevant strategies.
5. The legislation typically identified a resident by a shipping address in
the state.
6. Cited in Gordon Crovitz (2014).
7. This section relies in part on Stanford Graduate School of Business Case
P-81 Uber: 21st Century Technology Confronts 20th Century Regulation,
September 25, 2012, by David Hoyt and Steven Callander.
8. Other interests such as Zipcar are also affected but are unlikely to have a sig-
nificant influence on outcomes.
9. Uber lost in court and stopped operating in Germany.
10. Uber Policy White Paper 1.0, Principled Innovation: Addressing the
Regulatory Ambiguity Around Ridesharing Apps, posted April 12, 2013 by
Travis Kalanik.
11. The New York Times, August 10, 2014.
12. The New York Times, August 10, 2014.
13. Quoted in Jenkins (2014).
14. This section is based in part on Manheim (2013).
15. Fast Food, Poverty Wages, Center for Labor Research and Education,
University of California-Berkeley, and Department of Urban and Regional
Planning, University of Illinois at Champaign-Urbana, October 15, 2013. The data
period 20072011 covered the depth of the severe recession caused by the financial
crisis and the subsequent slow recovery. The study covered workers who worked
at least 10 hours a week for at least 27 weeks and did not include restau-
rant management.
16. The New York Times, January 14, 2014.
17. The New York Times, December 5, 2013.
18. San Jose Mercury News, September 5, 2014.
19. The Wall Street Journal, February 20, 2014.
20. Wal-Mart subsequently increased its minimum wage to $9.00 an hour and
$10.00 an hour the following year, but the increase affected only 6,000 of its 1.3 mil-
lion U.S. workers. Rather than higher wages, many workers wanted the opportunity
to work more hours.
50 DAVID P. BARON
21. The Wall Street Journal, September 5, 2014.

22. San Jose Mercury News, October 16, 2013.
23. The Wall Street Journal, February 20, 2014.
24. Wernerfelt (1984) presents a formalization of the resource-based approach, and
Makadok (2001) provides a model that incorporates both resources and capabilities.
25. See, for example, Barney, Ketchen, and Wright (2011).
26. Kingsley, Vanden Bergh, and Bonardi disaggregate the supply side by relat-
ing political competitiveness in a country to regulatory uncertainty and identify cor-
responding strategies in terms of their profile, coalitions, and political targets using
examples from the global telecommunications market.
27. Krehbiel (1998, 1999) and Baron (2014).
28. This view is taken by McWilliams and Seigel (2011).
29. Lyon and Salant (2014) provide a first look at activism and industry
transformation.
30. Baron (2014) considers self-regulation and lobbying in this context, along
with private politics social pressure. The effectiveness of nonmarket strategies in a
parliamentary versus a presidential system remains an open topic for research.
31. Baron (2015) presents a dynamic theory of policy choice in a proportional
representation parliamentary system with occasional crises, and equilibria yield tem-
porally durable policies.
32. Endogeneity of Wal-Marts choice of where to attempt a store opening
remains a concern.
33. de Figueiredo and Richter (2014) assess the challenges in conducting empiri-
cal research on lobbying.
34. Ansolabehere, de Figueiredo, and Snyder conclude that campaign contribu-
tions represent consumption for corporate executives.
REFERENCES
Ansolabehere, S., de Figueiredo, J., & Snyder, J. (2003). Why is there so little money in US
politics? Journal of Economic Perspectives, 17, 105130.
Argenti, P. A. (2004). Collaborating with activists: How Starbucks works with NGOs.
California Management Review, 47, 91116.
Bansal, P., & Hoffman, A. J. (2012). Oxford handbook of business and the natural environment.
Oxford: Oxford University Press.
Barney, J., Ketchen, D. J., & Wright, M. (Eds.). (2011). The future of resource-based theory:
Revival or decline? Journal of Management, 37(5), 12991315.
Baron, D. P. (1995). Integrated strategy: Market and nonmarket components. California
Management Review, 37 (2), 4765.
Baron, D. P. (2001). Private politics, corporate social responsibility, and integrated strategy.
Journal of Economics and Management Strategy, 10, 745.
Baron, D. P. (2012). The industrial organization of private politics. Quarterly Journal of
Political Science, 7, 135174.
Baron, D. P. (2013). Business and its environment (7th ed.). Upper Saddle River, NJ: Pearson.
Baron, D. P. (2014). Self-regulation in private and public politics. Quarterly Journal of

Baron, D. P. (2015). Elections and durable governments in a parliamentary democracy. Working
Paper. Stanford University.
Baron, D. P. (2016). The market for activism. Journal of Economics & Management Strategy,
Forthcoming.
Baron, D. P., & Diermeier, D. (2007). Strategic activism and nonmarket strategy. Journal of
Economics and Management Strategy, 16, 599634.
Baron, D. P., Harjoto, M. A., & Jo, H. (2011). The economics and politics of corporate social
responsibility. Business and Politics, 13(2), Article 1.
Baron, D. P., Neale, M. A., & Rao, H. (2016). Extending nonmarket strategy: Political
economy and the radical flank effect in private politics. Strategy Science, Forthcoming.
Baysinger, B. (1984). Domain maintenance as an objective of business political activity:
An expanded typology. Academy of Management Review, 9, 248258.
Becchetti, L., & Ciciretti, R. (2011). Stock market reaction to the global financial crisis:
Testing for the Lehman Brothers event. Giornale degli Economisti e Annali di
Economica, 70, 358.
Becchetti, L., Ciciretti, R., Hasan, I., & Kobeissi, N. (2012). Corporate social responsibility
and shareholders value. Journal of Business Research, 65, 16281635.
Becker, G. S. (1983). A theory of competition and pressure groups for political influence.
Quarterly Journal of Economics, 98, 371400.
Besanko, D., Dranove, D., Schaefer, S., & Shanley, M. (2013). Economics of strategy (6th ed.).
Hoboken, NJ: Wiley.
Bonardi, J.-P., Hillman, A., & Keim, G. (2005). The attractiveness of political markets:
Implications for firm strategy. Academy of Management Review, 30, 397413.
Bonardi, J.-P., Holburn, G., & Vanden Bergh, R. G. (2006). Nonmarket performance:
Evidence from U.S. electric utilities. Academy of Management Journal, 49, 12091228.
Bonardi, J.-P., & Vanden Bergh, R. G. (2015). Integrated market & nonmarket strategy:
Political knowledge and the resource-based view of the firm. In T. Lawton &
T. Rajwani (Eds.), Routledge Companion to Nonmarket strategy. London: Routledge.
Burbano, V. C. (2014). Corporate social responsibility and firm performance: Field experimental
evidence on the role of employee salary requirements and productivity. Working Paper.
University of California, Los Angeles.
Cassar, L. (2014). Contract design with a motivated agents: Experimental evidence. Working
Paper. University of Zurich.
Chatterji, A. K., & Toffel, M. W. (2010). How firms respond to being rated. Strategic
Management Journal, 31(9), 917945.
Cooper, M. J., Gulen, H., & Ovtchinnikov, A. V. (2010). Corporate political contributions
and stock returns. Journal of Finance, 65, 687724.
Crane, A., McWilliams, A., Matten, D., Moon, J., & Siegel, D. (2008). The Oxford handbook
of corporate social responsibility. Oxford: Oxford University Press.
de Figueiredo, J. M. (2009). Integrated political strategy. Advances in Strategic Management,
26, 459486.
de Figueiredo, J. M., & Richter, K. B. (2014). Advancing the empirical research on lobbying.
Annual Review of Political Science, 17, 163185.
de Figueiredo, R. J. P., Jr., & Edwards, G. (2007). Does private money buy public policy?
Campaign contributions and political outcomes in telecommunications. Journal of
52 DAVID P. BARON
Doh, J. P., Lawton, T. C., & Rajwani, T. (2012). Advancing nonmarket strategy research:
Institutional perspectives in a changing world. Academy of Management Perspectives,
26, 2239.
Eesley, C., & Lenox, M. J. (2006). Firm responses to secondary stakeholder action. Strategic
Management Journal, 27, 765781.
Epstein, M. J., & Schnietz, K. E. (2002). Measuring the cost of environmental and labor
protests to globalization: An event study of the failed 1999 Seattle WTO talks. The
International Trade Journal, XVI, 129160.
Foss, N. J. (2011). Why micro-foundations for resource-based theory are needed and what
they may look like. Journal of Management, 37, 14131428.
Gordon Crovitz, L. (2014). The Wall Street Journal, August 25.
Graham, B. A. T., Johnston, N. P., & Kingsley, A. F. (2013). Even constrained governments
steal: The domestic politics of transfer and expropriation risks. Working Paper.
University of Southern California.
Gupta, S., & Innes, R. (2011). Private politics and environmental management. Working Paper.
University of Florida.
Harrison, A., & Scorse, J. (2010). Multinationals and anti-sweatshop activism. American
Economic Review, 100, 247273.
Henisz, W. J., Dorobantu, S., & Nartley, L. J. (2014). Spinning gold: The financial returns to
stakeholder engagement. Strategic Management Journal, 35, 17271748.
Henisz, W. J., & Zelner, B. A. (2012). Strategy and competition in the market and nonmarket
arenas. Academy of Management Perspectives, 26, 4051.
Hillman, A. J., Keim, G. D., & Schuler, D. (2004). Corporate political activity: A review and
research agenda. Journal of Management, 30, 837857.
Holburn, G. L. F., & Zelner, B. A. (2010). Political capabilities, policy risk, and international
investment strategy: Evidence from the global electric power generation industry.
Strategic Management Journal, 31, 12901315.
Hong, B., & Minor, D. (2013). Choosing to be Good: How managers determine their impact
on financial and social performance. Working Paper. Western Ontario University.
Hong, H., & Kacperczyk, L. (2009). The price of sin: The effects of social norms on markets.
Journal of Financial Economics, 103, 119.
Ingram, P., Yue, L. Q., & Rao, H. (2010). Trouble in store: The emergence and success of
protests against Wal-Mart store openings in America. American Journal of Sociology,
116, 5392.
Jenkins, H. W., Jr. (2014). The Wall Street Journal, August 20.
Jensen, N. (2008). Political risk, democratic institutions, and foreign direct investment. Journal
of Politics, 4, 10401052.
Keim, G. D. (1981). Foundations of a political strategy for business. California Management
Review, 23(3), 4148.
Keim, G. D. (1985). Corporate grassroots programs in the 1980s. California Management
Review, 28, 110123.
Keim, G. D., & Zeithaml, C. P. (1986). Corporate political strategies and legislative decision-
making: A review and contingency approach. Academy of Management Review,
11, 828843.
Keim, G. D., Zeithaml, C. P., & Baysinger, B. D. (1984). New directions for corporate politi-
cal strategy. Sloan Management Review, 25(3), 5362.
Kingsley, A. F., Vanden Bergh, R. G., & Bonardi, J.-P. (2012). Political markets and regula-
tory uncertainty: Insights and implications for integrated strategy. Academy of
Management Perspectives, 5267.
Kotchen, M., & Moon, J. (2012). Corporate social responsibility for irresponsibility. Journal
of Economic Analysis & Policy, 12, 123.
Krehbiel, K. (1998). Pivotal politics: A theory of U.S. lawmaking. Chicago, IL: University of
Chicago Press.
Krehbiel, K. (1999). Pivotal politics: A refinement of nonmarket analysis for voting institu-
tions. Business and Politics, 1(April).
Kruger, P. (2015). Corporate goodness and shareholder wealth. Journal of Financial
Economics, 115, 304329.
Lawton, T. C., & Rajwani, T. (Eds.). (2015). The Routledge companion to nonmarket strategy.
London: Routledge.
Lenox, M. J., & Eesley, C. E. (2009). Private environmental activism and the selection and
response of firm targets. Journal of Economics and Management Strategy, 18, 4573.
Locke, R. M., Qin, F., & Brause, A. (2007). Does monitoring improve labor standards?
Lessons from Nike. Industrial and Labor Relations Review, 61, 331.
Lyon, T. P. (Ed.). (2010). Good cop, bad cop: Environmental NGOs and their strategies toward
business. Washington, DC: RFF Press.
Lyon, T. P., & Salant, S. (2014). Linking public and private politics: Activist strategy for indus-
try transformation. Working Paper. University of Michigan.
Makadok, R. (2001). Toward a synthesis of the resource-based and dynamic-capabilities views
of rent creation. Strategic Management Journal, 22, 387401.
Malhotra, N., Monin, B., & Tomz, M. (2014). Is corporate environmentalism profitable?
Experimental investigations of the effects of environmental corporate social responsibility
on political activity. Working Paper. Stanford University.
Manheim, J. B. (2013). The emerging role of worker centers in union organizing: A strategic
assessment. U.S. Chamber of Commerce: Workforce Freedom Initiative, November.
Margolis, J., Elfenbein, H., & Walsh, J. (2007). Does it pay to be good? A meta-analysis and
redirection of research on the relationship between corporate social and financial perfor-
mance. Working Paper. Harvard University.
McWilliams, A. (Ed.). (2015). The economics of corporate social responsibility. London: Edward
Elgar Publishing.
McWilliams, A., & Siegel, D. (2011). Creating and capturing value: Strategic corporate respon-
sibility, resource based theory and sustainable competitive advantage. Journal of
Management, 37, 14801495.
Minor, D. B. (2013). The value of corporate citizenship: Protection. Working Paper.
Northwestern University.
Minor, D. B., & Morgan, J. (2011). CSR as reputation insurance: Premium non nocere.
Nyborg, K., & Zhang, T. (2013). Is corporate social responsibility associated with lower
wages? Environmental Resource and Economics, 55, 107117.
Olson, M. J. (1965). The logic of collective action. Cambridge, MA: Harvard University Press.
Orlitsky, M., Schmidt, F. L., & Rynes, S. L. (2003). Corporate social and financial perfor-
mance: A meta-analysis. Organizational Studies, 24, 403411.
Porter, M. E. (1970). Competitive strategy. New York, NY: Free Press.
54 DAVID P. BARON
Roberts, J. (2004). The modern firm. Oxford: Oxford University Press.

Stigler, G. (1971). The theory of economic regulation. Bell Journal of Economics and
Management Science, 2, 321.
Wernerfelt, B. (1984). A resource-based view of the firm. Strategic Management Journal,
5, 171180.
Wolverton, A. (2009). Effects of socio-economic and input-related factors on polluting plants
location decisions. The B.E. Journal of Economic Analysis & Policy, 9. Article 14.
Yoffie, D. B. (1987). Corporate strategies for political action: A rational model. In A. A.
Marcus, A. M. Kaufman, & D. Beam (Eds.), Business strategy and public policy
(pp. 4360). New York, NY: Quorum Books.
Yoffie, D. B., & Bergenstein, S. (1985). Creating political advantage: The rise of the corporate
political entrepreneur. California Management Review, 28, 124139.
PART I
PUBLIC POLITICS
POLITICAL RISK AS A HOLD-UP
PROBLEM: IMPLICATIONS FOR
INTEGRATED STRATEGY
Kenneth W. Shotts
ABSTRACT
This paper explores firms strategic options when their investments are
subject to the threat of government expropriation. I develop a simple hold-
up model of political risk. In the model, a firm decides whether to invest
and then the government decides whether to expropriate the firms invest-
ment or to simply collect normal taxes on its profits. The government is
motivated by revenue and a wide range of nonpecuniary factors: its repu-
tation, electoral pressures, patronage opportunities, and pressure from
external actors. In the model, the likelihood of expropriation depends on
several factors: the firms profits, the amount of taxes it pays, the govern-
ments ability to operate the firms assets, and the governments political
incentives. Effective management of political risk requires an integrated
strategy, consisting not only of public and government relations efforts,
but also financial, value chain, and human resources strategies designed to
reduce the governments incentives for expropriation.
Keywords: Economic development; political risk; integrated strategy;
government relations; international business

ISSN: 0742-3322/doi:10.1108/S0742-332220160000034003
57
58 KENNETH W. SHOTTS
INTRODUCTION
Broadly defined, political risk is the possibility that a government will
change its policies in some way that is detrimental to a firms profits. Such
changes may range from minor alterations of tax or regulatory policy to
complete expropriation of a firms assets. The process behind policy
changes also varies widely. Some changes are implemented via institutional
rules that are fully legitimate and transparent. At the other end of the spec-
trum, the process may be completely illegitimate or corrupt. Scholars who
study political risk typically conclude that it is most prevalent in autocratic
countries, but that it is also present in democracies.
For global firms, as well as domestic firms in many developing countries,
effective assessment and strategic management of political risk are of fun-
damental importance. There are three main forms of political risk that
firms must contend with. The most obvious is direct nationalization or
expropriation of a firms assets. Less obvious is creeping expropriation,
that is, changes to tax law, intellectual property protection, or other regula-
tory policies that reduce a companys profits or make its operations more
difficult. A final type of political risk for multinational firms is imposition
of currency controls that limit their ability to take profits out of a country.1
All of these types of risk share important features in common;
namely, they represent substantial changes to government policies after
a company has made up front investments. Thus, political risk is an
example of a hold-up problem, that is, a situation in which one actor makes
a relationship-specific investment and then another actor has an opportu-
nity to expropriate part or all of the return on that investment. Economists
have explored a variety of solutions to hold-up problems, including vertical
integration, formal contracts, and informal relational contracts. However,
many of these solutions are difficult to apply to situations of political risk,
due to the power of sovereign nation states and the short time horizons of
many political actors.
In this paper, I develop a hold-up model of political risk as a game
played between a firm that can make an investment and a government that
can expropriate this investment. The key structural feature of the model is
that the government cannot make credible promises to respect the firms
property rights. The key characteristics of the actors payoffs are that the
firm tries to maximize profits, whereas the government is concerned about
tax revenue as well as a wide variety of political factors. The model is
substantially simpler than many previous game theoretic models of political
risk, because it is static rather than dynamic and it only incorporates very
Political Risk and Integrated Strategy 59
simple actions and payoffs for the actors.2 This simplicity is intentional,
because my primary goal is to provide a framework that can summarize
key factors determining the governments decisions, and thereby serve as a
useful springboard for analyzing strategies that the firm can use to protect
its investments.
The strategies that I discuss fall into four broad categories. First, I con-
sider ex ante tools to directly reduce the governments ability to expropriate
a firms investments. Second, I consider the possibility that a firm facing the
threat of expropriation may be better off making concessions and sharing
a greater proportion of its profits with the government. I then analyze
ways that a firm can build a coalition of supporters, either domestically or
internationally, who will convince the government not to expropriate its
investment. Finally, I analyze integrated strategies for political risk manage-
ment, including changes to the firms finance, human resources, operations,
and value chain strategies. Such strategies are important, because manage-
ment of political risk must be handled throughout the firms business, rather
than simply being a responsibility of the public relations or government
affairs group. A major theme of the strategies that I consider is that it is
crucial for a firm to ensure that the communities and countries in which
does business see a clear and direct benefit from its ongoing presence.
As I develop the model and its strategic implications, I draw on the
substantial empirical literature on political risk, governments incentives,
and firms investment decisions. I also illustrate my analysis with specific
examples of expropriation, as well as examples of strategic actions that
firms have taken to reduce the level of political risk that they face.
THE MODEL
I develop a simple theoretical model of strategic interactions between two
actors: a government and a firm. The government can be national, state, or
local and the firm can be either domestic or foreign. The firm moves first,
deciding whether to make a costly capital investment, for example, building
a factory. If the firm builds the factory, it makes short-run profits and
also may make long-run profits. However, the government can choose to
expropriate the firms investment, in which case the government operates
the factory and the firm doesnt get any long-run profits.
I use the notation ix to represent an actors payoffs. The superscript
represents the actor: i = f is the firm and i = g is the government. The
subscript represents the outcome of the game: x = 0 means the firm does
not invest, x = t means the firm invests and pays taxes (but isnt expro-
priated), and x = e means the firm invests and the government expropriates
the investment.
The firms objective is to maximize profits, which depend on several
factors. To build the factory, the firm must pay a fixed cost K. Its pretax
profits are P1 in the short-run and P2 in the long-run. The firm must pay
taxes on its profits, at an exogenously fixed rate t.3 Thus, if the government
does not expropriate, the firms net payoff from its investment is
ft 1 t P1 P2 K
On the other hand, if the government does expropriate, the firm only gets
the short-run revenue from its investment, and its net payoff is
fe 1 t P1 K
If the firm doesnt invest, its payoff is normalized to zero,

f0 0
At this point, three aspects of the model are worth discussing. First, the
model focuses on outright expropriation, as opposed to changes in tax
laws, regulatory requirements, or currency controls. However, the model
could also be interpreted as applying to any form of creeping expropriation
that essentially eliminates a firms long-run profits from its investment.
Second, the model assumes that the firm gets no compensation when the
government expropriates its investment. Of course, firms that are expro-
priated often receive some compensation. However, the level of compensa-
tion typically is insufficient to make up for the harm (foregone profits) that
the firm suffers as a result of expropriation. It would be straightforward to
incorporate an exogenously fixed level of compensation into the model,
without changing the main results.
Third, the model abstracts away from many factors that are part of stan-
dard financial valuation models. Such factors (discounting of profits from
multiple periods, the cost of capital, market risks, and technological risks)
also could be incorporated without changing the key results.
I now describe the incentives of the government, which is motivated by
revenues and also by a wide range of nonpecuniary factors. There are
several reasons that a government might care about revenues. One possibi-
lity is that its leaders are corrupt and want money for themselves. Another
possibility is that government leaders are good public servants who want to
provide public goods for their citizens and need revenue in order to do so.
A third possibility is that the governments goal is simply to stay in office;
in any political system, whether democratic or autocratic, a leader is more
likely retain power if the government is solvent and able to provide services
to its citizens.
There are two ways that the government can reap revenue. The first is to
collect taxes, at rate t, on the firms short-run and long-run profits. The
second is to expropriate the firms investment, and run the factory itself.
If the firm doesnt build the factory, the governments payoff is normalized
to zero,
g0 0
If the firm builds the factory and the government doesnt expropriate, the
revenue it collects is the tax rate t multiplied by the firms pretax profits in
the two periods
gt tP1 P2
If the firm builds the factory and the government expropriates the invest-
ment, it gets tP1 in the short run from taxes on the firms profits, as well as
aP2 in the long run, from operating the factory itself. Here, a, which is
assumed to be less than or equal to 1, parametrizes how effective the gov-
ernment believes it will be at running the factory. The value of a could
depend on technological factors such as complexity of the factorys opera-
tions, administrative factors such as the governments competence, and
even psychological factors such as politicians overconfidence in their own
ability to run the factory. A low value of a means that the factory is diffi-
cult for the government to operate and the government is aware of this
fact. However, there are several reasons that a could be close to 1: the
factory may be easy to operate, the government may have substantial tech-
nical and managerial expertise, or the government may mistakenly believe
it has the skills necessary to operate the factory. The value of a may also
depend on the factorys role in the value chain for example, if the factory
supplies specialized intermediate goods for the parent company, then the
government knows it wont be able to get much revenue if it expropriates
the factory. In contrast, if the factory produces a commodity for which
there is a thick global market, then the government is much more likely to
find buyers for its output.
In my model, the government doesnt just care about revenues if it

expropriates the firms investment, it also receives nonpecuniary political
benefits and pays political costs. As is standard in game theoretic models,
I assume that the government can put these factors on the same scale as
revenues, and that it makes rational decisions about tradeoffs between the
different factors that it cares about.
The political benefits, represented by the parameter B, can include a vari-
ety of factors. For example, the government may use the factory as a vehicle
for patronage, hiring family, friends, and political supporters. Or, from a
more sanguine perspective, it may want to operate the factory in the public
interest, exercising good environmental stewardship, creating safe working
conditions, or taking other actions that promote legitimate public policy
goals. A third possibility is that the mere act of taking the firms assets yields
political benefits for the government, by improving its electoral prospects.
For example, if voters believe that taking over the firm is a good idea, then
the government has an incentive to pander to them by expropriating the
firms investment, even if the government realizes it is likely to actually lose
long-run revenue in the process. Similarly, if the voters are concerned that
the government might be captured by business interests, then it can use
expropriation as a populist policy to signal that it cares about the masses.4
The political costs that the government incurs if it expropriates the firms
investment are represented by the parameter C, and can include a variety
of factors. If the firm is from a foreign country, the government has to
worry that expropriation will harm its relationship with that countrys
government. A domestic firm may similarly get support from groups within
its country that oppose expropriation, especially if the governing coalition
includes pro-business parties. Other stakeholders, such as the firms
employees, may also pressure the government to respect the firms invest-
ment if they expect that expropriation will negatively affect their interests.
In sum, the governments payoff if it expropriates the firms invest-
ment is
ge tP1 aP2 B C
ANALYSIS OF MODEL
Because the government cannot commit not to expropriate the firms
investment, I solve the model by backward induction. Suppose the firm has
built the factory will the government decide to expropriate? By comparing
the governments payoffs with and without expropriation ( ge and gt ), we

see that the government will expropriate if
aP2 B C > tP2 1
Equation (1) summarizes several factors that increase political risk. The
governments incentive to expropriate is highest if it expects to get high
political benefits and incur low political costs. Expropriation is also more
appealing if the firms technology is easy to master, or the government
believes it can operate the factory profitably. More subtly, low taxes make
expropriation more likely. This runs counter to the standard intuition that
low taxes give companies an incentive to invest. But, seen from the perspec-
tive of the government, the result is quite intuitive: if the benefits of an
investment are going solely to a company and not much is being shared
with the government, then expropriation is more appealing.
Finally, note that if the government believes it can run the factory rea-
sonably effectively (i.e., a > t, which means that the government receives
more revenue when it owns the factory than when it taxes the firms profits)
then high long-run profits P2 make expropriation more appealing to
the government.
Having characterized how the government will react if the firm invests,
I now turn to the firms decision about whether to invest. The firms
decision depends on whether it expects its investment to be expropriated. If
the firm does not expect to be expropriated, then a comparison of its payoff
ft with f0 = 0, shows that the firm will make the investment if the posttax
profits, in both the short and long run, exceed the cost of building the
factory, that is,
1 tP1 P2 > K 2
On the other hand, if the firm expects its investment to be expropriated,

then a comparison of its payoff fe with f0 = 0 shows that the firm must
recoup its costs very quickly, and it will only make the investment if its
short-run posttax profits exceed the cost of building the factory, that is,
1 tP1 > K 3
Obviously, investment is more appealing if the firm does not expect to be

expropriated. More subtly, when facing the threat of expropriation, an
investment is more appealing if it will yield substantial short-run profits P1
and if it doesnt require major capital investments K.
The following propositions give a nontechnical summary of the results

in Eqs. (1)(3). The first proposition characterizes the governments incen-
tives, and the second characterizes three patterns of optimal behavior for
the firm, depending on its expectations about the government.
Proposition 1. The government is more likely to expropriate the firms
investment when it believes it can run the factory effectively, when it
sees substantial political benefits and low political costs from expropriat-
ing, and when it doesnt receive much tax revenue from the firm.
Proposition 2. The firms investment decision depends on its profits as

well as the threat of expropriation.
Case 1: Investment absent political risk. If the government wont
expropriate, then the firm makes its investment decision solely based
on whether its investment will be profitable.
Case 2: Investment despite expropriation. If the government will expro-
priate and the firms short-run profits are sufficiently high to cover the
cost of building the factory, then the firm will invest even though it
expects to be expropriated.
Case 3: Investment deterred. If the government will expropriate and
the firms short-run profits dont cover the cost of building the fac-
tory, then the firm wont invest.
As an example of investment absent political risk (Case 1 of Proposition 2),

consider Siemenss decision to build in Singapore a center for research and
development of water purification technologies.5 The initial investment
decision in 2007 as well as Siemenss subsequent decision to sell its water
technologies division in 2012 were driven exclusively by assessments of the
market, as well as the synergies, or lack thereof, between the water technolo-
gies business and Siemenss other lines of business. The exceedingly unlikely
prospect of expropriation by the Singaporean government wasnt a part of
the calculation.
Even in unstable countries, it is possible for political risk to be low. For
example, Haber, Razo, and Maurer (2003) analyze the Mexican mining
and petroleum industries during a time of political upheaval (19111929)
and argue that during this period, companies were safe from expropriation,
due to factors incorporated in my model: (i) the government received a
substantial amount of tax revenue from the industries, (ii) mines, oil fields,
and refineries were difficult to operate and the government did not possess
the necessary technical expertise, and (iii) the industries were protected by
the U.S. government, which would have imposed costs on the Mexican
government in the event of nationalization.6
I now turn to Cases 2 and 3 of Proposition 2, which analyze investment
decisions by a company that faces high political risk. Examples of compa-
nies that make major investments in a country despite expecting to be
expropriated are rare. Possibly the best fit for my model would be consu-
mer products companies that can, at a reasonably low cost, build a distri-
bution network and reap substantial short-run profits from sales of
their products.
In contrast, there are countless situations where the threat of expropria-
tion deters investment. Indeed, along with other forms of uncertainty about
the business climate, this is arguably one of the most significant barriers
to economic growth in many countries. A recent example comes from
Bolivias handling of its immense lithium reserves in the Salar de Uyuni.
President Evo Morales has envisioned massive extraction of lithium, as
well as production of batteries and possibly even electric cars in Bolivia.
But foreign companies have been reluctant to make major investments. In
part, this is because President Morales has declared that the Bolivian gov-
ernment must play a major role in lithium production, stating that the
state will never lose sovereignty when it comes to lithium (Wright, 2010).
Moreover, given that Morales followed through on a campaign promise to
nationalize Bolivias natural gas industry after his first election in 2006,
companies have good reason to believe that any investments they make in
lithium extraction could likewise be expropriated.
Bolivia is far from alone in having investment deterred due to politi-
cal risk. For example, foreign investment in Ecuador fell dramatically
after President Rafael Correa expropriated assets of several multina-
tional oil companies in 2008.7 Later, after his re-election in 2013,
President Correa tried to court foreign investment, saying we cant be
beggars sitting on a sack of gold and claiming that the advantages of
our country for foreign investment [include] political stability (Garcia &
Ellsworth, 2013).
The fact that leaders who have engaged in expropriation seek new for-
eign investment, but struggle to obtain it, suggests an important possibility:
a government may actually be worse off as a result of its ability to expro-
priate. I now use my model to analyze how expropriation, or the threat
thereof, affects the payoffs of the firm and the government.
The firm obviously would prefer that the government not expropriate its
investment. Then it could get a profit that is greater than its profit from
being expropriated or its profit from not making any investment.8
The government, in contrast, can either benefit or be hurt as a result of

its ability to expropriate. If the government does not have an incentive to
expropriate, then the fact that it can expropriate has no effect on the firms
behavior. However, if aP2+BC > tP2 from Eq. (1), the government will
expropriate any investment the firm makes. The firm takes this into
account when making investment decisions.
If it expects an investment to be expropriated, the firm is only willing to
make the investment if, from Eq. (3), (1 t)P1>K. In this circumstance,
which corresponds to Case 2 of Proposition 2, the firm is undeterred, so
the government will take over the factory and will benefit from its ability
to do so.
On the other hand, if (1t)P1<K which corresponds to Case 3 of
Proposition 2, then the governments ability to expropriate deters the firm
from building the factory. Note that this is a bad outcome for both the firm
and the government. The firm makes no money and the government gets
no tax revenue. In contrast, if the government could commit not to expro-
priate, it would benefit, because it could get revenue gt = tP1 P2 from
taxes on the firms profits.
The existence of this no-investment equilibrium is of course driven by
the inability of the firm and the government to strike a binding contract,
for example, one requiring the firm to make the investment and the govern-
ment not to expropriate it. If such contracts were feasible, then the Coase
theorem suggests that the two parties would come to a mutually beneficial
agreement, in which the firm makes some profit and the government
receives some tax revenue. However, as argued by Acemoglu (2003) there
are good reasons to believe that the Coase theorem does not apply to politi-
cal economy, because a sovereign government cannot make binding com-
mitments. As pithily described by Vernon (1971, p. 47), any deal between a
company and a government is an obsolescing bargain, because almost
from the moment that the signatures have dried on the document, powerful
forces go to work that quickly render the agreements obsolete in the eyes
of the government.9
At a broader level, the fact that the government can be harmed by its
own ability to expropriate investments has profound implications for eco-
nomic development. In particular, many scholars have argued that contract
enforcement, the rule of law, political stability, and security of investments
are among the most important (or even the most important) factors that
affect the level of investment and growth in a countrys economy (see, e.g.,
Busse & Hefeker, 2007; Mauro, 1995; Rodrik, Subramanian, & Trebbi,
2004).10 Thus, the question of how best to manage and mitigate political
risk is not only important for companies; it is also vitally important for
governments and citizens.
STRATEGIES TO MITIGATE POLITICAL RISK

There are two basic questions that a firm must ask in the presence of politi-
cal risk. First, is it a good idea to make an investment? Second, what strate-
gies can be used to manage and mitigate political risk? This second
question is obviously relevant for a company that already has chosen to
make an investment. But it is equally important for a company that is in
the process of making an investment decision, because the firms ability to
devise an effective risk mitigation strategy affects its probability of being
expropriated, and hence the appeal of the investment.
I now use my model as a springboard to analyze a wide range of strate-
gies for dealing with political risk. A major theme of this analysis is that
political risk cannot simply be handled by the firms public and governmen-
tal affairs group. Rather, effective management of political risk requires an
integrated strategy. As defined by Baron (1995), an integrated strategy has
both market and nonmarket components, and is designed to take into
account crucial interactions between those components. An integrated strat-
egy for managing political risk is one that considers how a firms relation-
ship with the government will be affected by its decisions about operations,
human resources, and finance. An integrated strategy also can have multiple
different nonmarket components, including working with the government in
the country where the firm makes an investment, the governments of other
countries, and a variety of different stakeholder groups.
The strategies that I will analyze come in several different varieties. First
I analyze ex ante tactics that a firm could use to directly eliminate political
risk. Next, I discuss how financial concessions to the government can be
used to reduce the risk of expropriation. I then turn to an analysis of non-
market strategies for government and public relations. Finally, I discuss
ways that a firm can address political risk by altering components of its
market strategy: its human resources practices, ownership structure, capital
expenditures, and flow of profits.
GOVERNMENT COMMITMENT
The most obvious solution to a hold-up problem is to directly eliminate the
problem, that is, to change the structure of the interaction between the
company and the government, so that political risk is no longer an issue. In
particular, the fact that both the company and government can be harmed
as a result of the governments inability to commit not to expropriate sug-
gests a very natural remedy for political risk: the government should adopt
policies and procedures that limit its ability to expropriate the firms invest-
ments. If it can find a way to do this, the government will benefit because it
can collect taxes on the firms revenues. Of course, the firm also will benefit,
as it can make some profits.
The question, though, is not whether commitment is desirable, but
rather whether it is feasible. One way that a government might try to com-
mit is by agreeing to production sharing agreements or contracts that are
governed by international laws, arbitration bodies, or courts. However,
although international enforcement mechanisms sometimes influence coun-
tries behavior, this requires that the mechanisms be backed by a credible
threat of a substantial punishment.11 In situations where the punishment is
either small or lacks credibility, sovereign states can have substantial flex-
ibility to ignore international rulings.
A more subtle approach would be for the government to change its insti-
tutions to make policy changes more difficult. Intuitively, if there are multi-
ple actors who can block government decisions, including expropriation
decisions, then a firms investment is more secure. Indeed, the empirical lit-
erature shows that expropriation is less common in countries where
political authority is dispersed (Guriev et al., 2011; Henisz, 2000a; Jensen,
2006, 2008; Jensen et al., 2012; North & Weingast, 1989). And in a survey
of U.S. companies, Biglaiser and Staats (2010) find that investors pay close
attention to the rule of law and effectiveness of court systems. This suggests
a natural solution to the problem of political risk: creation of independent
judicial and regulatory agencies that are insulated from political pressures.
However, as noted by Manzano and Monaldi (2008) in their analysis of
the Latin American oil sector, nominally independent agencies can quickly
lose their autonomy if they are not supported by a countrys fundamental
institutional structure. And although a government seeking to make its
country more appealing to investors may try to reform its judicial or regu-
latory system, this is a process that is both slow and uncertain. Thus, in
countries where the risk of expropriation is high, it is unlikely that an indi-
vidual company will be able to convince a government to make the types of
basic institutional changes that are necessary to secure its investments.
The two types of commitment (via contracts and via institutions) that
I have discussed above involve formal procedures. A different possibility is
that commitment can be achieved informally, via implicit relational contracts.
As shown by Guriev et al. (2011), relational contracts are not only useful in
analyzing theories of the firm (the canonical application from Baker,
Gibbons, & Murphy, 2002), but are also naturally applicable to relationships
between firms and governments. The basic intuition is that a government
that expects a long-run flow of investments from a firm may find it more
appealing to refrain from expropriating the firm at any moment in time,
because it knows that if it chooses to expropriate, then it will lose future tax
revenue generated by future investments. The notion of commitment here is
endogenous and informal, that is, the government, as a sovereign state, is
able to expropriate the investment any time it wants, but it chooses not to
do so.
A simplified way of thinking about this sort of incentive for the govern-
ment is that it is a reputational cost part of the cost C that the government
incurs if it chooses to expropriate. This cost might result from the fact that
the firm being expropriated will no longer invest in the country, but it can
also include the fact that other firms may likewise refrain from investing.
The size of this reputational cost varies across countries, depending on
their current reputations. For example, if Singapore were to make a sur-
prising series of decisions to expropriate companies, it would lose a key
asset in obtaining future investment, namely, its strong reputation for hav-
ing a stable business environment. In contrast, once President Hugo
Chavez had spent several years expropriating Venezuelan and foreign com-
panies investments, the marginal harm to his governments reputation
from engaging in yet another expropriation was rather small.
The size of the reputational cost that a government incurs as a result of
expropriating also depends on the governments time horizon. In general,
relational contracts work best when both actors are motivated by the pro-
spect of future interactions; this means that governmental stability typically
reduces political risk.12 In a country with a great deal of turnover, a politician
has much less reason to worry about the future flow of investments, because
that is a problem his successor will have to deal with.13 The reputational cost
is also affected by how easy it is for firms to shift future investments to other
countries. To the extent that the government must compete against other
governments in luring investment, it has a greater incentive to develop and
maintain a reputation for sticking to deals that it makes with investors.
Note, however, that very few of the factors that make relational con-
tracts feasible as a solution to political risk are under an individual firms
control. From the perspective of a firm, the governments reputation and
time horizon are exogenous factors. Moreover, an individual firm typically
cannot influence how its competitors will react if it is expropriated.14
The one thing that a firm can potentially do, however, is to plan a sequence
of investments so that the government has an ongoing reason to expect
future benefits as long as it respects the firms property rights and profits
(Thomas & Worrall, 1994).
I now turn to a very different question: what can a firm do if it has

made an investment and realizes that the governments incentives make
expropriation appealing?
VOLUNTARY CONCESSIONS
If a government is considering expropriating a firms investment and its
motivation for doing so is primarily financial, there is a potential deal to be
struck, assuming that the firm can operate the factory more efficiently than
the government. In particular, it may be in both the firms interest and the
governments interest if, rather than the government expropriating the
factory, the firm continues to operate it, but pays a higher tax rate.
Recall that in my model, expropriation is appealing to the government
when aP2+BC > tP2. For the firm, expropriation is always a bad deal: it
gets a payoff of 0 in the second period and it would be better off paying a
higher tax rate and still making some posttax profits. In particular, as long
as BC isnt too large, then there is a higher tax rate t^ t; 1, such that
aP2 B C < t^P2 . With this higher tax rate, the government prefers simply
receiving tax revenue rather than expropriating,
and the firm is better off as
well, getting second-period profits 1 t^ P2 rather than 0.15
How might such a tax change be implemented in practice? The most
direct way would be for the firm to unilaterally offer to pay a higher share
of its profits to the government. This would be most natural if the firm
received special tax breaks or negotiated a highly favorable production
sharing agreement (PSA) when making its initial investment. Of course, a
decision to voluntarily hand over more money to the government would
run directly counter to the instincts of many businesspeople. But it is better
than having the entire investment expropriated.16 Although there are few
examples of companies voluntarily choosing to pay higher taxes, it is quite
possible that doing so could reduce the political risk they face. A more
common way of implementing this type of strategy is by initiating major
Corporate Social Responsibility programs that effectively share some of
a firms profits with local communities. Another way that a firm can
make voluntary concessions is by signing contracts with local companies,
including those controlled by the government, on terms that are favorable
to those companies.
A couple of subtle points about strategic concessions are worth noting.
First, the right time to make such a change is before the government
announces an expropriation, because once a politician makes a public
announcement, he is likely to bear a political cost if he flip-flops and

changes his mind, especially if he is a populist whose constituents expect
him to stand up to business interests.
A second subtle issue is that often the terms of a companys deal with a
country will be highly favorable precisely because of the threat of political
risk. Janeba (2002) develops a theory in which a government that cannot
make policy commitments offers up front subsidies or other incentives in
order to lure investment, and Li (2006) provides empirical evidence that tax
incentives are larger in countries with weak rule of law. As a concrete exam-
ple, consider Abdelals (2005) description of Royal Dutch/Shells invest-
ments in natural gas production on Sakhalin Island in the early 1990s, when
there was massive uncertainty about the future of Russias government. In
that context, a company investing in Russia needed the prospect of extra-
ordinary profits to justify taking extraordinary risks, and Shell was able to
negotiate a PSA that promised it the lions share of revenues from the
Sakhalin II project. But the fact that the agreement was so favorable to
Shell ultimately wound up putting the project at risk, and after a series of
disputes over the PSA and environmental issues, the projects assets were
acquired by Gazprom at a price below market value (Kramer, 2006).
This example illustrates the fact that although a favorable tax deal or
PSA is appealing to a company it is also a two-edged sword. For example,
a company that is granted a 10-year tax holiday will receive a larger stream
of profits provided it is not expropriated, but the fact that the firm is not
paying taxes further increases the companys risk of expropriation. Indeed,
based on interviews with World Bank experts on political risk, Jensen et al.
(2014, p. 24) conclude that unbalanced contracts [are] one potential trigger
for expropriation threats.
One way a company may seek to avoid this problem is by seeking fron-
tloaded subsidies of its capital investments rather than a favorable tax deal.
However, if the government is unable to afford such subsidies, but can only
offer a tax deal, then the company must always be ready to renegotiate. As
with any long-run interaction in the absence of enforceable contracts,
investment in an environment where political risk is present requires that
both parties continually benefit.
POLITICAL STRATEGIES
I now turn to other, nonfinancial, ways that a company can try to ensure
that the government doesnt have an incentive to expropriate. The strategies
I will discuss are political, in the sense that they are designed to influence
decisions made by government officials, either by decreasing the benefits B
or by increasing the costs C that the government incurs when expropriating.
However, to be clear, I will not discuss strategies in which a company
becomes directly involved in the political process by which the countrys lea-
ders are chosen.17 Of course, some companies have engaged in that sort of
direct political activity, via electoral politics in democracies, power struggles
in autocracies, or even by trying to undermine the stability of democratic
regimes. The most important reason to ignore such strategies is that they
often are unethical. Moreover, they arguably can be ineffective as well. For
example, a foreign companys attempt to manipulate an election can back-
fire if the countrys citizens learn about it. Similarly, a domestic company
that aligns itself tightly with a particular political faction runs a high risk of
losing its influence and being expropriated if a competing faction wins
power. The key question is what a company can do to achieve influence in a
way that is stable, durable, and legitimate.
The first natural instinct for a foreign company is to use its home gov-
ernment as a source of leverage, or as a guarantor of its interests. Many
host governments will respond to pressure from a foreign country; the
threat of retaliation by a country that is large, powerful, and wealthy can
increase the cost C of expropriation. Jensen et al. (2014) provide empirical
evidence that pressure from home country governments as well as the
World Bank can be an effective deterrent. And Asiedu et al. (2009) find
that firms are less concerned about expropriation risk when investing in a
country that receives foreign aid, which could be cut off.
However, as appealing as it may be for a company to rely on its home
government, there are limitations to this approach. The first limitation is
that such pressure may be perceived as illegitimate by the citizens and gov-
ernment of the country that is on the receiving end of the pressure. In such
circumstances, foreign pressure may increase the costs of expropriation,
but it also will increase the benefits B, because the government can say that
it is standing up against foreign interests. Which factor dominates, of
course, depends on the government, but standing up against foreign pres-
sure clearly is an effective political tactic for some leaders, for example,
President Correa of Ecuador, who has repeatedly taken stands against
companies from the United States, and who has intentionally irked
the U.S. government by allowing Wikileaks cofounder Julian Assange to
seek refuge in the Ecuadorian embassy in London.
More subtly, there is no guarantee that a countrys home government
will jump to promote its companies interests. A government may be
motivated by a wide variety of factors, and often its interests will not match
up with those of any particular company. This is even true for large state-
owned companies. For example, as described by Musacchio et al. (2009),
when President Morales of Bolivia nationalized Petrobrass natural gas
investments, Brazils President Lula de Silva, a populist who wanted to
establish himself as a leader across Latin America, chose not to pressure
Bolivia to reverse its course.
A second, and very different, political strategy that a company can use is
to develop a deep support base within the country where it makes investments
(whether this is its home country or a foreign one). This is not something that
can be accomplished easily or quickly; rather, the firm must consciously
decide to relate to society and local communities in a way that ensures that
they value its presence.18 If this is accomplished, then a government consider-
ing expropriating the firm will see lower political benefits B, because it will
not want to position itself as attacking a well-respected company. The govern-
ment might even incur additional costs C, if people who have benefitted from
the firms presence in their communities resent the governments actions.
Broadly speaking, this sort of community relations strategy falls under
the rubric of Corporate Social Responsibility (CSR). Of course, the exact
nature of the companys activities should depend on true community needs.
In developing countries, the needs might be roads, hospitals, or water
projects. In developed countries, it might make more sense for the company
to support civic or artistic organizations. And, as shown by Vekasis (2014)
analysis of Japanese firms operating in China, if a company faces political
risk due to contentious relationships between its home country and the
host country, it may find it useful to improve its own firm-specific image
with local citizens while also investing in general efforts to improve rela-
tions between the countries.
The exact nature of the projects that a firm should invest in also depends
on the firms market strategy and capabilities; as noted by Porter and
Kramer (2011), a companys CSR activities will yield higher benefits for
society, and better results for the firm, if it thinks carefully and strategically
about how to have the greatest positive impact. Finally, the companys
CSR projects should be developed in consultation with respected local
community leaders, both to ensure that the projects appropriately address
the communitys needs and to increase the chances that community leaders
will back the company if government officials consider expropriation.
I now turn to the topic of integrated strategy, that is, ways that a
firm can adapt its financial, operational, value chain, and human resources
strategies to better manage political risk.
INTEGRATED STRATEGY: HR, OPERATIONS,

AND VALUE CHAIN MANAGEMENT
As discussed in the previous section, CSR is one way to improve a companys
relationship with local communities. However, HR and value chain strategies
are even more important for achieving this goal. These strategies can take a
variety of forms: hiring local citizens as employees, training and promoting
them as managers, relying on local sources for inputs, and locating the firms
downstream value chain locally. These strategies are particularly important
for firms in extractive industries, because even ordinary people who are una-
ware of the voluminous academic literature on the resource curse are often
acutely aware of the possibility that a company could come in, take their
resources, ruin the local environment, and make massive profits, while pro-
viding little benefit for people living in the area. And if the local citizens
arent initially concerned about this possibility, politicians surely will have an
incentive to make sure they are aware of it.
Thus, it is important for a firm to build its relationship with a community
in a way that makes it clear that local citizens will benefit from its presence.19
As an example, consider the Brazilian company Vales investments in nickel
mining in Canada. Typically, Canada is not considered to be a high political
risk country. However, from the perspective of a mining company, which
must obtain permits from provincial governments, environmental regulators,
and other regulatory authorities, political risk is present everywhere, even in
Canada. In the 2000s, when Vale acquired rights to a mine in Voiseys Bay, in
the province of Newfoundland and Labrador, the company decided to build
a new refinery within the province, thereby benefitting its citizens, rather than
shipping ore to existing refineries in other Canadian provinces.
In terms of my hold-up model of political risk, a company that adapts its
HR and value chain strategies to benefit local communities is decreasing the
benefits B that a government can achieve by expropriating its investment,
because politicians have less incentive to try to claim credit for standing up
to the firm. Likewise, if the firm trains local employees and invests in their
human capital, this will increase the costs C that a government will incur if
it adopts policies that harm the firm, because the employees will worry
about job losses or reductions in their salaries. Similarly, local companies
that become part of the firms value chain have an incentive to encourage
the government to respect its investments, and thus governments are less
likely to interfere with the property rights of foreign firms that use local
companies as suppliers (Johns & Wellhausen, 2016).
However, my model also suggests a very different, and much more

Machiavellian strategy: rather than trying to develop good relationships with
the local community, the firm could instead try to ensure that a government
that expropriates its investment will be unable to operate the factory effec-
tively. In terms of the model, this scorched earth approach would mean
choosing HR, operations, and value chain strategies that drive down the
value of a (which represents how much revenue the government will get by
operating the factory itself).20 Recall that, as summarized in Proposition 1, a
low value of a makes expropriation less appealing for the government. There
are several ways of driving down the value of a: employing only expatriates,
keeping local employees in menial jobs rather than training them for high-
skill positions, using needlessly complicated operational procedures, and
sourcing from foreign companies that may choose not to do business with
the government if it takes over the factory.
This approach is unethical, because in pursuing its own self-interest the
firm would be intentionally taking actions that make others, particularly
local employees, substantially worse off, thus limiting the long-run benefits
that the community could get from its investment. The approach is also
very risky, because although it may reduce the governments ability to
operate the factory, it also will cause local citizens to resent the company,
thereby increasing the direct political benefits B that the government can
obtain by expropriating the companys assets.
INTEGRATED STRATEGY: OWNERSHIP

AND CONTRACTS
Given that expropriation directly reduces the wealth of a firms owners, a
company making investments in a risky environment should think strategi-
cally about the ownership structure for its projects.
One natural approach is to ensure that the firms equity partners include
individuals or companies that are politically powerful within the country
where the firm is investing. However, there also are risks associated with
this approach. First, it is risky to be tied too closely to any one faction, even
the current regime, which may well be out of power in the future.21 Second,
as noted by Henisz (2000b), a foreign company that forms a joint venture
with a domestic company in a country lacking strong institutions must
worry not only about the possibility of expropriation by the government
but also about the possibility of expropriation by its business partners.
Another approach that foreign companies can use is to acquire political

risk insurance. This could be direct insurance by the companys home
government or the World Bank. Or it could be implicit insurance obtained
by taking loans from host country banks, who would have an incentive
to protest any expropriation. Political risk insurance can also be achieved
indirectly via the structure of a firms contracts with customers. Any form of
political risk insurance effectively increases the cost C that a government
pays if it expropriates the companys investments, because the company has
a readymade coalition that will put pressure on the government.
Doh and Ramamurti (2003) give examples of companies that have miti-
gated political risk by using multilateral financing, partnerships with local
firms, and contracts with government-owned firms. A classic example of
this approach is how the Kennecott mining company structured its invest-
ments in the El Teniente mine in Chile. As described by Moran (1973,
p. 277), Kennecott lined up from as many directions as possible, interna-
tional supporters who would automatically share the Kennecott parents
outrage in case of nationalization. Kennecott also purchased political risk
insurance from the U.S. government, designed in a manner that would
induce multiple government agencies to come to its aid. (This subtlety is
important, because it takes into account the internal bureaucratic politics
of the U.S. government). Moreover, the company designed its contracts
with European and Japanese customers in a manner that would give them
an incentive to put pressure on the Chilean government. As a result of these
efforts, the company made profits off of its investments in the 1960s and
1970s, and even after President Salvador Allende was elected, the company
received substantial compensation in exchange for its assets.
INTEGRATED STRATEGY: TIMING OF PROFITS

A third type of integrated strategy that a firm can use to mitigate political
risk is to manage the stream of profits that it makes from an investment.
Most obviously, this can be done by holding down the level of initial capi-
tal investment and trying to ensure that revenues are frontloaded as much
as possible. In my model, such strategies are useful if a firm expects to be
expropriated, because a high level of short-run profits P1 and low level of
investment K can help ensure that the investment is profitable. More
importantly, the model also implies that frontloading profits makes expro-
priation less financially appealing for a government.22
Another financial strategy is to artificially reduce variance in the firms

profits. The intuition for why this makes sense stems from empirical evi-
dence that nationalizations in the oil industry are most common when
prices are high and oil companies are most profitable (Guriev et al., 2011;
Manzano & Monaldi, 2008). I now sketch a simple extension of my model,
allowing the firm to smooth its earnings. Suppose that second-period
high
profits will take one of two values, Plow 2 or P2 , depending on market con-
high high
ditions. If aP2 B C > tP2 , then the firm will be expropriated if its
investment turns out to be highly profitable. It may be better for the firm
to smooth its profits. Specifically, we can write the firms expected profits
high
as P2 = 1 Plow
2 P2 , where is the probability of good market con-
ditions for the firms products. If the average profit level isnt too high,
that is, aP2 B C < tP2 , then if the firm can sufficiently reduce variance in
its profits, it will no longer be vulnerable to expropriation. Indeed, the
firm may be willing to pay a substantial premium for variance reduction.
Of course, the types of financial management that I have described in this
section do not come without costs; indeed, the firm would be altering its
investment decisions and contract structures in ways that are not justified on
purely economic grounds. However, in the presence of political risk, such arti-
ficial alterations may make sense. For example, Durnev and Guriev (2011)
provide evidence that natural resource firms facing high levels of political risk
use a variety of financial strategies including negative accruals and opaque
accounting practices. The downside of such strategies is that they slow com-
panies growth and make it more difficult for them to acquire capital, but the
upside is that they make the firms less appealing as targets for expropriation.
CONCLUSION
In this paper, I have developed a simple hold-up model of political risk.
The model conceives of a sovereign government as being motivated both
by revenues it gets from taxing or owning a factory, as well as by a variety
of other political benefits and costs that it will incur if it expropriates a
firms investment. The firms investment decisions may depend on its expec-
tations about the governments future behavior. Most crucially, the govern-
ment cannot commit not to expropriate the firms investment.
The model provides a framework for thinking about several aspects of a
firms strategy. Two broad themes emerge. First, even for a purely profit-
motivated firm, it is essential to ensure that societal actors, including the
government and local communities, see the benefits of its presence. Second,
the tools that the firm can use to improve its relationships with these impor-
tant constituents are not just the province of the firms government and
community affairs group, but also include aspects of its market strategy.
This second observation has an important implication for implementa-
tion of effective strategies for dealing with political risk. It is natural that
each division of a firm will be more concerned about its own area than
the overall stability of the firms investment. Moreover, in a typical firm,
the government relations or CSR group does not have the leverage neces-
sary to induce other divisions to take costly actions that reduce political
risk. Thus, to implement an effective integrated strategy for political risk
management requires involvement by the firms most senior leaders, who
are best-positioned to assess the key tradeoffs. Also, there must be company-
wide awareness of the importance of developing good relationships with
local communities and governments; one way to accomplish this is by
developing a strong sense of the firms values for corporate citizenship.
I should also note that although this paper has focused on managing poli-
tical risk, the analysis has implications for assessment of political risk by
companies as they make investment decisions. Although the model treats risk
as being deterministic, real-world political risk always involves uncertainty,
and a company making an investment must assess the probability that it will
lose its investment. In many industries, a company cannot choose to avoid
any possibility of political risk, because to do so would be tantamount to
deciding never to make any investments and simply going out of business.
Often, when people think about political risk assessment, they focus on
the overall level of risk in a country, but political risk is actually a function
of the interaction between a particular firm and the particular government
officials that it deals with. A governments benefits B and costs C from
expropriating a firms investment may vary across companies. This differ-
ence may be in part due to the companies home countries for example, a
U.S. company doing business in Ecuador faces greater political risk than a
Chinese company, because President Correa has several ongoing disputes
with the U.S. government. Similarly, when it tried to sign contracts to oper-
ate ports in New York and New Jersey, the UAE-based company DP
World faced higher levels of political risk than many other foreign compa-
nies would have faced. When there is a market for control of assets, a com-
parison of the level of political risk that different companies face has
important implications for who should naturally hold a particular asset.
Sometimes, despite having strong market or operational advantages,
a company will decide that given the price for an asset, that asset is better
held by some other company that is less likely to be expropriated.
Finally, I will note a couple of challenges faced by firms that have little
experience dealing with political risk or that are moving into institutional
environments that are new and unfamiliar to them. The first problem is
simply one of awareness. The idea that governments should respect a firms
investments comes very naturally to businesspeople, who may thus see
commitment as the governments duty, rather than assessing how their own
actions affect the governments actions.
Similarly, for a profit-motivated firm, it is natural to think about finan-
cial motivations for government officials. However, for many government
officials, revenue is only a small component of the overall benefits and costs
of their decisions. In such situations, it is easy for businesspeople to con-
clude that politicians are crazy, inept, or uninterested in doing whats right.
But often what is really going on is that politicians are motivated by a
more complicated set of factors that are not immediately obvious to busi-
nesspeople. This means that an effective integrated strategy as well as an
accurate assessment of the level of political risk a company faces requires
a nuanced understanding of the political system and the incentives of indi-
vidual political decision-makers. The following examples give a small
sampling of incentives that motivate different government officials.
Social stability: As described by Baron (2013), when the government of

China shut down direct selling in 1998, its primary goal was not to
obtain financial benefits by seizing assets of companies like Avon and
Amway; rather, the benefits that the government saw from the policy
change included maintaining social stability, restricting the role of reli-
gion in society, and limiting mass gatherings of its citizens.
Interest group pressures: In the 1990s and 2000s, California agribusiness
companies planted hundreds of thousands of acres of new almond orch-
ards on land with very junior water rights. These major capital investments
were at risk of drought, because almond trees require a large amount of
water each year in order to survive. Moreover, water deliveries for the
almond growers were subject to political risk, because environmental
groups wanted to ensure adequate water supplies for endangered species
like the Delta smelt, and urban and suburban interests could put pressure
on the states Water Resources Control Board to cut agricultural water
deliveries rather than cutting deliveries to citizens elsewhere in the state. To
counteract these groups and promote their own interests, the agribusiness
companies funded groups like the California Latino Water coalition and
sponsored rallies in the state capital.
Nationalism: In 2009, Egypt and Algeria played a series of World Cup
qualifying matches, which were accompanied by nationalistic riots in
both countries. Soon afterwards, the Algerian government levied hun-
dreds of millions of dollars in fines on Djezzy, a subsidiary of the
Egyptian mobile company Orascom, alleging that it had failed to pay
taxes (El-Tablawy, 2010). Orascom tried to sell Djezzy to MTN, but this
move was blocked by the government. Ultimately, Orascom sold Djezzy
to VimpelCom, which subsequently sold Djezzy to the Algerian govern-
ment (Khrennikov, 2014).
Pandering: When Tata Motors encountered intense opposition to its
plans to build production facilities for the Nano microcar on prime agri-
cultural land in Singur, the source of political risk was not the West
Bengal state government, which was controlled by a nominally
Communist party. Rather, protests against the project were stoked by
Rabindranath Bhattacharya, a local politician from the Trinamool
Congress Party, who was facing a tight electoral race and who arguably
used the Nano controversy as an opportunity to pander to voters (Jha,
2013a). As the profile of the protests grew, the anti-Tata cause was taken
up by higher-level Congress Party politicians, including Mamata
Banerjee, who won the next round of statewide elections.
Given the enormous variety of political incentives for government decision-
makers, effective strategic management of political risk must be based on a
thorough understanding of the motivations, constituencies, and institutional
capabilities of the specific political actors whose actions can affect the firms
long-run success.
NOTES
1. Another form of country risk is the threat of political violence, which can dis-
rupt a companys operations. This risk, while important, falls outside my analysis,
which focuses on changes in government policies.
2. For dynamic infinite-horizon models, see Cole and English (1991), Thomas
and Worrall (1994), Asiedu, Jin, and Nandwa (2009), and Guriev, Kolotilin,
and Sonin (2011). For models with more complicated actions or payoffs, see
Eaton and Gersovitzs (1984) analysis of factor input decisions, Graham, Johnston,
and Kingsleys (2014) analysis of currency restrictions, and Jensen, Johnston,
Lee, and Sahins (2014) analysis of the effects of crises.
3. I assume these parameters take reasonable values: K > 0, P1 > 0, P2 > 0,

and 0 < t < 1.
4. For a theory of pandering, see Canes-Wrone, Herron, and Shotts (2001), and
for theories of posturing and populism, see Fox (2007), Fox and Shotts (2009), Frisell
(2009), and Acemoglu, Egorov, and Sonin (2013). An empirical analysis of populist
pressure and resource nationalism in Bolivia is given by Kohl and Farthing (2012).
5. For a discussion, see Waterworld (2012).
6. In the notation of Eq. (1), t was high, a was low, and C was high.
7. For a history of the expropriation, see Musacchio, Goldberg, and Reisen de
Pinho (2009).
8. In the model, as long as K isnt too high ft > fe > f0 = 0.
9. Obsolescing bargains also arise in what Baron (2001) calls private politics.
For example, if an activist group campaigns against a firm, then it may be in both
parties interest to strike a deal in which the firm changes its practices and the group
ends its campaign. However, the group can change its mind, and other activist
groups can also criticize the companys practices. Thus, a firm that is considering
making concessions may face a hold-up problem.
10. Disentangling causality is of course difficult when assessing the links between
institutions, investment, and growth, and other scholars (e.g., Glaeser, La Porta,
Lopez-de-Silanes, & Shleifer, 2004) have argued that it is actually growth that
causes institutional development.
11. For example, the Hickenlooper Amendment was designed to commit the
U.S. government to cut off foreign aid as a punishment for expropriation.
12. For a rare counterexample, see Guidolin and La Ferraras (2007) analysis of
how diamond company stock prices declined when the death of a key rebel leader
foreshadowed the end of the Angolas civil war.
13. On this point, see the informal argument in Olson (1993) as well as the formal
argument in Guriev et al. (2011).
14. Wellhausen (2015) shows that multinational companies investments in a
country are only affected by expropriation of companies from their home country;
in contrast, expropriation of a company from a different country typically doesnt
have any effect on a companys subsequent investment decisions.
15. Of course, if the government receives huge political benefits B from expro-
priation and only pays minor political costs C then even at a 100% tax rate the
government would prefer to expropriate.
16. The intuition of making concessions in order to avoid drastic policy changes
also arises naturally from political economy models of agenda setting and pivotal
politics (Baron, 1996; Krehbiel, 1998; Romer & Rosenthal, 1979).
17. Bucheli and Kim (2015) discuss historical examples of MNCs that sought to
control host countries political institutions and policymaking. They also analyze
factors that contribute to such firms success or failure.
18. An extreme example of the long-run nature of investments in good relations
is provided by Jhas (2013b) analysis of local institutions that provided a
stable business environment for Muslim traders doing business in India. These insti-
tutions date to the medieval era, but continue to have important effects today.
19. This is especially true in countries with political federalism. Jensen (2006)
shows that FDI is correlated with the amount of influence that political subunits
have over national policy. He argues that this is because employment benefits of
FDI are realized locally, and hence local officials have an incentive to protect firms
against adverse changes in national policy.
20. For example, in Eaton and Gersovitzs (1984) model, political risk induces a
foreign firm to skew factor inputs toward foreign managerial services and away
from capital investment.
21. For example, Fisman (2001) shows that rumors that President Suharto of
Indonesia was ill caused substantial declines in the value of firms connected to
his regime.
22. From Eq. (1), the government will expropriate if aP2+BC > tP2. Thus, if
a > t, that is, the government makes more money by owning the factory than by
taxing the firm, then shifting profits from P2 to P1 makes expropriation less appeal-
ing for the government.
ACKNOWLEDGMENTS
I received helpful comments from David Baron, John de Figueiredo, Sergei
Guriev, Nate Jensen, Dylan Minor, two reviewers, and participants at ISNIE
2015 and the 2015 Conference on Strategy & the Business Environment.
I also thank Stanford MBA and MSx students, and Executive Education
program participants who have shared with me their experiences and
thoughts on the topic of political risk.
REFERENCES
Abdelal, R. (2005). Journey to Sakhalin: Royal Dutch/Shell in Russia. Harvard Business
School Case 9-704-040.
Acemoglu, D. (2003). Why not a political Coase theorem? Social conflict, commitment, and
politics. Journal of Comparative Economics, 31, 620652.
Acemoglu, D., Egorov, G., & Sonin, K. (2013). A political theory of populism. Quarterly
Journal of Economics, 128, 771805.
Asiedu, E., Jin, Y., & Nandwa, B. (2009). Does foreign aid mitigate the adverse effect of
expropriation risk on foreign direct investment. Journal of International Economics,
78, 268275.
Baker, G., Gibbons, R., & Murphy, K. J. (2002). Relational contracts and the theory of the
firm. Quarterly Journal of Economics, 117, 3984.
Baron, D. P. (1995). Integrated strategy: Market and nonmarket components. California
Management Review, 37, 4765.
Baron, D. P. (1996). A dynamic theory of collective goods programs. American Political
Science Review, 90, 316330.
Journal of Economics & Management Strategy, 10, 745.
Baron, D. P. (2013). Business and its environment (7th ed.). Boston, MA: Pearson.
Biglaiser, G., & Staats, J. L. (2010). Do political institutions affect foreign direct investment?
A survey of U.S. corporations in Latin America. Political Research Quarterly,
63, 508522.
Bucheli, M., & Kim, M. (2015). Attacked from both sides: A dynamic model of multinational
corporations strategies for protection of their property rights. Global Strategy Journal,
5, 126.
Busse, M., & Hefeker, C. (2007). Political risk, institutions, and foreign direct investment.
European Journal of Political Economy, 23, 397415.
Canes-Wrone, B., Herron, M. C., & Shotts, K. W. (2001). Leadership and pandering: A theory
of executive policymaking. American Journal of Political Science, 45, 532550.
Cole, H. L., & English, W. B. (1991). Expropriation and direct investment. Journal of
International Economics, 30, 201227.
Doh, J. P., & Ramamurti, R. (2003). Reassessing risk in developing country infrastructure.
Long Range Planning, 36, 337353.
Durnev, A., & Guriev, S. (2011). Expropriation risk and firm control: A corporate transparency
channel. SSRN Working Paper No. 1020476. Retrieved from http://dx.doi.org/10.2139/
ssrn.1020476
Eaton, J., & Gersovitz, M. (1984). A theory of expropriation and deviations from perfect
capital mobility. Economic Journal, 94, 1540.
El-Tablawy, T. (2010). Egypts Orascom faces new Algeria tax bill. Associated Press. Retrieved
from http://www.washingtonpost.com/wp-dyn/content/article/2010/09/30/AR201009300
1533.html
Fisman, R. (2001). Estimating the value of political connections. American Economic Review,
91, 10951102.
Fox, J. (2007). Government transparency and policymaking. Public Choice, 131, 2344.
Fox, J., & Shotts, K. W. (2009). Delegates or trustees? A theory of political accountability.
Journal of Politics, 71, 12251237.
Frisell, L. (2009). A theory of self-fulfilling political expectations. Journal of Public Economics,
93, 715720.
Garcia, E., & Ellsworth, B. (2013). Ecuadors Correa enjoys re-election, seeks investment.
Reuters. Retrieved from http://www.reuters.com/article/2013/02/18/us-ecuador-election-
idUSBRE91G01K20130218
Glaeser, E. L., La Porta, R., Lopez-de-Silanes, F., & Shleifer, A. (2004). Do institutions cause
growth? Journal of Economic Growth, 9, 271304.
Graham, B. A. T., Johnston, N. P., & Kingsley, A. F. (2014). Even constrained governments
steal: The domestic politics of transfer and expropriation risks. Paper presented at the
2013 Conference on Strategy in the Business Environment. Retrieved from http://www.
uvm.edu/~akingsle/GJK2013.pdf
Guidolin, M., & La Ferrara, E. (2007). Diamonds are forever, wars are not: Is conflict bad for
private firms? American Economic Review, 97, 19781993.
Guriev, S., Kolotilin, A., & Sonin, K. (2011). Determinants of nationalization in the oil sector:
A theory and evidence from panel data. Journal of Law, Economics, and Organization,
27, 301323.
Haber, S., Razo, A., & Maurer, M. (2003). The politics of property rights: Political instability,
credible commitments, and economic growth in Mexico, 18761929. Cambridge:
Cambridge University Press.
Henisz, W. J. (2000a). The institutional environment for economic growth. Economics and
Politics, 12, 131.
Henisz, W. J. (2000b). The institutional environment for multinational investment. Journal of
Law, Economics, and Organization, 16, 334364.
Janeba, E. (2002). Attracting FDI in a politically risky world. International Economic Review,
43, 11271155.
Jensen, N. M. (2006). Nation-states and the multinational corporation: A political economy of
foreign direct investment. Princeton, NJ: Princeton University Press.
Jensen, N. M. (2008). Political risk, democratic institutions, and foreign direct investment.
Journal of Politics, 70, 10421052.
Jensen, N. M., Biglaiser, G., Li, Q., Malesky, E., Pinto, P. M., Pinto, S. M., & Staats, J. L.
(2012). Politics and foreign direct investment. Ann Arbor, MI: University of
Michigan Press.
Jensen, N. M., Johnston, N. P., Lee, C., & Sahin, A. (2014). Crisis and Contract Breach:
The Domestic and International Determinants of Expropriation. George Washington
University typescript. Retrieved from http://www.natemjensen.com/wp-content/
uploads/2014/09/Crisis-Expro-Nov11.pdf
Jha, S. (2013a). Analyzing political risks in developing countries: A practical framework for
project managers. Business and Politics, 15, 117136.
Jha, S. (2013b). Trade, institutions, and ethnic tolerance: Evidence from South Asia. American
Political Science Review, 107, 806832.
Johns, L., & Wellhausen, R. (2016). Under one roof: Supply chains and the protection of
foreign investment. American Political Science Review, forthcoming.
Khrennikov, I. (2014). VimpelCom sells 51% Djezzy stake to Algeria to reduce debt.
Bloomberg.com. Retrieved from http://www.bloomberg.com/news/articles/2014-04-18/
vimpelcom-to-sell-51-djezzy-stake-to-algeria-for-2-64-billion
Kohl, B., & Farthing, L. (2012). Material constraints to popular imaginaries: The extractive
economy and resource nationalism in Bolivia. Political Geography, 31, 225235.
Kramer, A. E. (2006). Shell cedes control of Sakhalin-2 to Gazprom. The New York Times.
Retrieved from http://www.nytimes.com/2006/12/21/business/worldbusiness/21iht-shell.
3981718.html?_r=0
Chicago Press.
Li, Q. (2006). Democracy, autocracy, and tax incentives to foreign direct investors. Journal of
Politics, 68, 6274.
Manzano, O., & Monaldi, F. (2008). The political economy of oil production in Latin
America. Economa, 9, 5998.
Mauro, P. (1995). Corruption and growth. Quarterly Journal of Economics, 110, 681712.
Moran, T. H. (1973). Transnational strategies of protection and defense by multinational
corporations: Spreading the risk and raising the cost for nationalization in natural
resources. International Organization, 27, 273287.
Musacchio, A., Goldberg, L. G., & Reisen de Pinho, R. (2009). Petrobras in Ecuador (A).
Harvard Business School Case 9-309-107.
North, D. C., & Weingast, B. R. (1989). Constitutions and commitment: The evolution of
institutions governing public choice in seventeenth-century England. Journal of
Economic History, 49, 803832.
Olson, M. (1993). Dictatorship, democracy, and development. American Political Science

Review, 87, 567576.
Porter, M. E., & Kramer, M. R. (2011). Creating shared value. Harvard Business Review,
89, 6277.
Rodrik, D., Subramanian, A., & Trebbi, F. (2004). Institutions rule: The primacy of institu-
tions over geography and integration in economic development. Journal of Economic
Growth, 9, 131165.
Romer, T., & Rosenthal, H. (1979). Bureaucrats versus voters: On the political economy of
resource allocation by direct democracy. Quarterly Journal of Economics, 93, 563587.
Thomas, J., & Worrall, T. (1994). Foreign direct investment and the risk of expropriation.
Review of Economic Studies, 61, 81108.
Vekasi, K. (2014). Transforming risk: Social strategies of multinational firms. Paper presented
at the 2014 Annual Meeting of the Midwest Political Science Association.
Vernon, R. (1971). Sovereignty at bay: The multinational spread of U.S. enterprises. New York,
NY: Basic Books.
Waterworld. (2012). Why Siemens is selling its water processing division. Retrieved from http://
www.waterworld.com/articles/wwi/print/volume-27/issue-6/regulars/water-leader-focus/
why-siemens-is-selling-its-water.html
Wellhausen, R. L. (2015). Investor-state disputes: When can governments break contracts?
Journal of Conflict Resolution, 59, 239261.
Wright, L. (2010). Lithium dreams. The New Yorker. Retrieved from http://www.newyorker.
com/magazine/2010/03/22/lithium-dreams
INCORPORATING LEGISLATIVE
EFFECTIVENESS INTO
NONMARKET STRATEGY:
THE CASE OF FINANCIAL
SERVICES REFORM AND
THE GREAT RECESSION
Craig Volden and Alan E. Wiseman
ABSTRACT
The field of nonmarket strategy has expanded rapidly over the past
20 years to provide theoretical and practical guidance for managers seeking
to influence policymaking. Much of this scholarship has built directly on
spatial and pivotal politics models of lawmaking. While extremely
helpful at identifying crucial targets for lobbying, these models treat all
policymakers as identical in their abilities to advance legislative agenda
items through various policymaking hurdles. We build upon these earlier
models, but include policymakers who vary in their relative effectiveness
at advancing measures through the legislative process. We identify how
the implications of our model deviate from those of conventional (pivotal
politics) analyses. We then present an empirical strategy for identifying

ISSN: 0742-3322/doi:10.1108/S0742-332220160000034004
87
88 CRAIG VOLDEN AND ALAN E. WISEMAN
effective Lawmakers in the United States Congress, and illustrate the

utility of this approach for managers developing nonmarket strategies in
legislative institutions, relying on the case of banking and financial
services reforms between 2008 and 2011.
Keywords: Legislative effectiveness; nonmarket strategy; financial
services reform
In many industries, whether a firm can secure and maintain a leading position
in the marketplace will depend on its ability to develop a market and a
nonmarket strategy. When a firms nonmarket strategy involves engaging
policymaking institutions, identifying policy entrepreneurs and particularly
effective lawmakers becomes central to strategic management. Yet, little is
known, theoretically or empirically, about such effective lawmakers how to
identify them, how they matter for policy outcomes, and how they fit into non-
market strategy. In this essay, we develop a theory of legislative effectiveness,
note how effective lawmakers can be identified, and illustrate their influence
over policies of great import to firms. In short, we here make the case for incor-
porating legislative effectiveness into nonmarket strategy, and into the study of
strategic management more broadly.
In his foundational work on competitive market strategy, Porter (1980,
pp. 2829) refers to the relevance of government when he notes that
government at all levels must be recognized as potentially influencing
many if not all aspects of industry structure and that no structural analy-
sis is complete without a diagnosis of how present and future government
policy, at all levels, will affect structural conditions. Yet it is usually
more illuminating to consider how government affects competition
through the five competitive forces [advanced by Porter] than to consider
it as a force in and of itself. While Porter does concede that strategy
may well involve treating government as an actor to be influenced, there
is effectively no guidance throughout his entire volume on the manner in
which such influence might be obtained.1 Needless to say, the field of
management strategy has expanded substantially over the past 35 years
to include a more formal engagement of the role of government on firms,
and how managers might seek to influence the policy outputs of govern-
ment institutions.
Legislative Effectiveness and Nonmarket Strategy 89
Yoffie (1988, p. 82) is blunt when he states that ignoring Washington until
you need it is a prescription for failure, and he illustrates how political
strategies can benefit seemingly disadvantaged firms and industries. Likewise,
Yoffie and Bergenstein (1985) explore how the strategies of corporate politi-
cal entrepreneurs created a favorable regulatory environment for firms as
diverse as American Express and MCI. Moving beyond these case-based
analyses, since the early 1990s, scholars of economics, politics, law, public
policy, and management have explored and refined various theories of
nonmarket strategy.
These theories have collectively explored how firms might (and should,
from a proscriptive perspective) interact with the media (e.g., Baron, 2005),
interest groups (e.g., Baron & Diermeier, 2007; Feddersen & Gilligan, 2002;
Lyon & Maxwell, 2004), and a wide range of political institutions, including
(but not limited to) legislatures, regulatory bodies, and electoral arenas
(e.g., Baron, 1999, 2001; De Figueiredo, 2009; De Figueiredo & Kim, 2004;
De Figueirdo & Tiller, 2002; Lyon & Maxwell, 1999).2 One of the most
prominent building blocks of contemporary nonmarket strategy is the Pivotal
Politics Theory (i.e., Krehbiel, 1996, 1998, 1999) that explores the strategic
implications of supermajoritarian institutions in lawmaking bodies for
managers.3 Building upon classic Median Voter Theories (e.g., Black, 1948;
Downs, 1957), pivotal politics theories demonstrate that when supermajorities
are required, policy change is limited, and particular actors (such as those
needed to overcome a veto or a filibuster) become pivotal.
In advancing his theory, Krehbiel (1999, p. 64) argues that effective
nonmarket strategy in the governmental arena consists of influence at
the margin. Therefore, the greater a managers understanding of essential
governmental processes, the greater are his or her prospects for effectiveness
at the margin. Krehbiels theory helps managers identify where the
margin is, and it provides a generic strategy for identifying pivotal voters
that might then be lobbied by managers to achieve their goals. Scholars of
nonmarket strategy have built upon the Pivotal Politics Theory to motivate
empirical analyses, and as a foundation for richer theories of nonmarket
strategy (e.g., Baron, 2005, 2014; Holburn & Vanden Bergh, 2004).4
While the rigor and real-world relevance of the pivotal politics theories
cannot be overstated, they collectively omit certain aspects of contemporary
lawmaking processes that might be deemed particularly relevant for managers.
Specifically, this body of theory essentially treats all actors in the legislature
(or any decision-making body, more generally considered) as substantively
identical in their abilities to advance measures through the legislative process.
While certain actors clearly benefit from possessing particular parliamentary
rights (e.g., the right to propose policies, veto rights, and the like), it is still the
case that the success or failure of a piece of legislation is highly dependent
how that bill comports with the preferences of various pivotal actors. Hence,
in such theories, any legislator with access to the agenda could propose the
same policy and experience the same prospects for success (or failure) as any
other actor with agenda access.
In contrast, we argue that a crucial feature of contemporary lawmaking
is that some legislators are better at making laws than others. Some law-
makers are better at identifying policy problems and coupling them with
viable solutions, better at identifying potential coalitions and the bargains
that might be necessary to bring them together, and better at packaging
their proposals in such a way that they appeal to a broad network of
policymakers perhaps independent of the ideological content of the
proposals. Media accounts, political biographies, and casual anecdotes
often point to a well-known collection of legislators in the U.S. Congress,
for example, who distinguished themselves across their careers as highly
effective lawmakers. Recent large-sample empirical analyses (Volden &
Wiseman, 2014) also demonstrate that some legislators are generally more
effective lawmakers than others; and that this variance in lawmaking skill
exhibits systematic patterns.
Little theoretical scholarship, however, has engaged the manner in which
lawmakers may vary in regards to their effectiveness; and this omission
is particularly profound when we consider its relevance for nonmarket
strategy. A managers knowledge of the relevant pivotal actor (such as a
veto player or subcommittee chair whose vote must be secured to ensure
passage) is a crucial step toward facilitating a successful nonmarket strategy;
but just as important is knowing who should be her firms advocate inside
the legislature. Not all advocates are equally skilled at advancing their
proposals; and selecting the wrong advocate can easily lead to legislative
failure, despite knowing which actors are pivotal.
While managers might be tempted to outsource political strategies to
lobbyists or other agents outside of the firm, there are several reasons to be
hesitant about such a plan. First, as noted by Yoffie and Bergenstein
(1985), it is often not cost-effective to retain an external lobbyist, depending
on a firms size. Likewise, Yoffie (1988, pp. 8889) also points to how
CEOs and lower-level managers are often more likely to gain meaningful
access to policymakers than lobbyists who might be representing their
firms interests. Hence, the burden is often upon managers to develop
realistic and meaningful nonmarket strategies with which to engage various
political arenas; and managers need to decide whether and how to spend
resources on their strategic objectives. Moreover, managers must decide

when to adopt an aggressive nonmarket strategy, when success can be
achieved with fewer resources, and when such efforts are not worth
the cost. We argue that none of those decisions can be wisely made without
fundamental knowledge of legislative effectiveness.
In what follows, we develop a theory that provides insights as to how
particularly effective lawmakers might succeed in bringing about policy
change where others might fail, and we provide guidance on how managers
can identify those most effective lawmakers, to help them advance their
nonmarket strategies. Therefore, our analysis illustrates the conditions
under which a misperception might arise (either from intuition or from
pivotal politics theories) that nothing can be accomplished, when close
work with highly effective lawmakers can, indeed, bring about policy
change. Likewise, we also identify conditions under which even a highly
effective lawmaker might be unable to secure policy change, and hence, a
manager might want to conserve her resources.
Moving beyond theory, we illustrate our empirical approach for identify-
ing highly effective lawmakers by presenting the legislators who were most
effective in the 110th House of Representatives (20072008) at advancing
legislation in a policy area that was particularly relevant to managers at that
time: banking and financial services policies. We also explore the role of one
particularly effective lawmaker in advancing financial services legislation
in the subsequent Congress (20092010), in the wake of the financial crisis.
Such analysis provides insight about the practical relevance of our argu-
ments to contemporary management strategy.
PIVOTAL POLITICS THEORIES

The Pivotal Politics Theory generically builds on a one-dimensional spatial
model of lawmaking in which actors have singled-peaked preferences and
policymaking is subject to an open amendment rule. As established by
Black (1948), if decisions are made by simple-majority vote, then the unique
equilibrium policy outcome will correspond to the ideal point of the legislative
median (i.e., the well-known Median Voter Theorem). The novel extension to
Blacks canonical framework that pivotal politics theories offer is to deviate
from the assumption of simple-majority voting. More specifically, in the con-
text of the U.S. Congress, the possibility of a presidential veto requires the
votes of 2/3rds of the legislature to secure override, and the possibility of a
filibuster requires the votes of 3/5ths of the legislature (U.S. Senate) to invoke
cloture. Together, these rules imply that policy gridlock can only be overcome
with the consent of a supermajority of voters.
The formal sequence of play in the Pivotal Politics Theory begins with
the legislative median effectively acting as an agenda setter, in which she
chooses which bill (if any) to propose as an alternative to the status quo
policy.5 If she proposes a bill, then a filibuster might ensue, in which
opponents of the bill attempt to engage in unlimited debate. If a filibuster
occurs, the filibuster pivot (positioned ideologically to represent approxi-
mately 2/5th of the Senate) will decide whether to allow the filibuster to
stand, or whether to invoke cloture, which would allow the legislative process
to progress. If a bill is ultimately adopted by the Congress, then a President
might choose to veto the bill; and if he does, then it will be up to the veto
pivot (positioned to represent 2/3rd of the legislature) to decide whether to
allow the veto to stand, or to vote in favor of an override, which would lead
to policy change.
The equilibrium results of the Pivotal Politics game identify how current
policies are mapped into new policy outcomes, as a function of the location
of the status quos and their spatial relationships to the preferences of the
different pivotal voters (i.e., the veto and/or filibuster pivot). A fundamental
result of the Pivotal Politics Theory is the pervasiveness of policy gridlock
defined as a situation in which the status quo is retained despite a simple
majority of voters who favor policy change especially in contrast to the
canonical Median Voter Theorem result (whereby all status quo policies
are predicted to move to the median voters ideal point). Moreover, the
Pivotal Politics Theory offers specific predictions regarding the scope of
policy change, as well as whose preferences are influential determinants of
said policy change when gridlock can be overcome.
With regards to nonmarket strategy, pivotal politics theories are particularly
insightful, as they compel managers to focus their attentions beyond the prefer-
ences of the legislative median, to consider other actors whose consent must be
secured to ensure that their legislative goals are recognized. While pivotal
politics theories offer much value for managers who seek to devise a nonmar-
ket strategy for navigating legislative institutions, they are collectively silent on
one important feature of contemporary lawmaking processes: the identity of a
legislative advocate. As noted above, pivotal politics theories essentially treat
the legislative median as the agenda setter, as a consequence of the legislative
process being subject to an open amendment rule. This assumption seems
entirely reasonable if legislators are essentially indistinguishable from each
other in their abilities to advance bills through the legislative process.
What would happen, however, if some legislators are generally better than
others at lawmaking? What would occur if some legislators, independent of
their ideologies, are more skilled than others at navigating the legislative
process and bringing about the policy compromises and bargains that were
necessary to secure legislative agreements? Pivotal politics theories are silent
about such scenarios, but they seem particularly relevant for managers who,
in addition to having to identify the scope of potential policy change as part
of their nonmarket strategy, must also identify an advocate who can advance
their causes in the legislature. The Legislative Effectiveness-Pivots Theory
developed below addresses precisely these questions.
A THEORY OF LEGISLATIVE EFFECTIVENESS

WITH PIVOTAL POLITICS
The model presented here, which we denote the Legislative Effectiveness-
Pivots Theory, is an extension of the Legislative Effectiveness Model (LEM)
advanced by Hitt, Volden, and Wiseman (2014). They analyze interactions
between a legislative agenda setter (denoted L, for the Lawmaker) and
various pivotal voters (e.g., the legislative median, committee chairs, etc.).
In their model, policies (and the bills that serve as the foundations for new
policies) are characterized by their spatial locations (in R1), and a quality
(or valence) term. Such bill quality might be thought of as arising from an
effective lawmaker matching a particularly relevant or successful solution
with a pressing policy problem, from striking a less-costly compromise that
eliminates opponents main objections, or from making a proposal in a
particularly attractive manner (electorally or politically), to name but a few
possibilities.6 One might interpret the bill quality dimension to capture
anything that policymakers agree they would like more of, while the spatial
dimension captures anything on which they differ.7 Put simply, bill quality
reflects the political and policy efforts of effective (or ineffective) legislative
proponents, where a higher quality can be attained at a lower effort cost by
a more effective lawmaker.
Analysis of the model (detailed below) yields several implications of
relevance to managers devising a nonmarket strategy for legislative insti-
tutions. First, highly effective lawmakers can cultivate proposals that
secure sufficient support in the chamber despite being clearly biased in
favor of the proposer (and therefore potentially more favorable to the
firm). Second, certain legislators will be better able to cultivate proposals
benefiting the interests of managers than others who are more closely aligned
with the firm but less effective as lawmakers. Third, effective lawmakers can
secure policy change (in a manner that benefits a managers interest) under
conditions where we might otherwise expect that gridlock would be obtained,
as described by conventional pivotal politics theories. Thus, managers whose
nonmarket strategies are overly beholden to traditional pivotal politics
viewpoints may miss out on opportunities to secure beneficial policy change
in collaboration with effective lawmakers.
Actors preferences in the model are defined by a quadratic loss in the
spatial distance between their ideal points and the policy outcome, and by
an additive (separable) benefit from the chosen policys quality. The median
legislators preferences, for example, can be represented by the following
utility function:
UM y; g xM y2 gy
where xM is the Medians ideal point, y {xb, xq} is the policy outcome in
the unidimensional space, and gy is the quality of the final policy (either the
bill or the status quo), which is good for all the actors. Similar to other
spatial models, we assume that xM = 0, so the Medians utility function
can be simplified to the following expression:
UM y; g y2 gy
We assume that the Lawmaker cares about policy location and quality,
and that she has the ability to add quality to a given policy proposal.
While we assume that it is costless to introduce a new policy, the
Lawmaker incurs a cost for any effort that she might exert to add quality
to a particular policy.8 More formally, Ls preferences can be represented
by the following utility function:
UL y; g xL y2 gy cgb
where xL is the Lawmakers ideal point (xL > xM = 0), and c 1 captures
the marginal cost that L must incur to add quality to a new bill (gb 0).9
This specification implies that there is a simple linear mapping between the
effort exerted by the Lawmaker and the level of quality that she adds to the
bill. Hence, c essentially captures the relative effectiveness of the Lawmaker
at producing bills that are generally attractive to all members, regardless of
their ideological content. If c is high, the Lawmaker is relatively ineffective at
lawmaking (because producing high-quality legislation is extremely costly),

whereas if c is low, the Lawmaker is relatively effective (because producing
high-quality legislation is not very costly). Note that our specification implies
that the Lawmaker does not incur any costs from the quality of the status
quo policy, and we assume that the Lawmaker values quality in a similar
manner to all other legislators. Note, also, that this specification does not
imply that all legislators will be in favor of a bill because it possesses positive
quality; rather, the addition of quality makes a bill somewhat more attractive
to all legislators, ceteris paribus. As we will see below, the legislative agenda
setter will leverage her ability to add quality to a bill to pass policies that
are spatially favorable to her interests, by providing just enough quality to
the bill to ensure that it gains sufficient support among pivotal voters in
the chamber.
To facilitate clear comparisons between the Legislative Effectiveness-Pivots
Theory and the standard Pivotal Politics Theory, we embrace identical
assumptions about the presence of supermajoritarian institutions in the legisla-
ture. That is, we assume that there is a President (with ideal point xP ) who has
veto power, for which a 2/3rd supermajority is required to override his veto (as
represented by the preferences of the veto pivot, with ideal point xV ), and that
the legislature allows for the possibility of unlimited debate filibusters
that require 3/5th supermajority (as represented by the preferences of the
filibuster pivot, with ideal point xF ) to invoke cloture and cut off debate. For
illustrative purposes, we assume that the configuration of the pivotal actors
ideal points is as follows: xP < xV < xM 0 < xF < 2xF < xL . That is, we assume
that there is a left-of-median President, and a right-of-median Lawmaker (with
ideal point xL ), where the Lawmaker is at least twice as far to the right of the
median as the filibuster pivot.10
Consistent with the Pivotal Politics Theory, we also assume that the
legislature operates under an open rule (meaning that all amendments are
allowed). However, only the Lawmaker can add a positive quality to a
proposal. All other proposals and the status quo are assumed to have quality
normalized to zero (g = 0). Similar to the Median, preferences of the
President, Veto Pivot, and Filibuster Pivot can each be represented by the
following utility function:
Ui y; g xi y2 gy ; i fF; P; Vg
The sequence of play is as follows. In stage 1, the Lawmaker offers a

proposal. If this proposal is accepted by the Filibuster and Veto Pivots, as
well as the Median in stage 2, it becomes the final policy outcome. If not,
the Median can offer an alternative proposal in stage 3, although with

quality set equal to zero, and the Pivotal Politics subgame ensues. Put
simply, the game presented here adds an earlier set of actions to the Pivotal
Politics model, which allows the Lawmaker to attempt to offer a quality
proposal preferred by all pivotal actors over the equilibrium proposal that
would otherwise follow from the Pivotal Politics Theory.
Given this game structure, the equilibrium can be described formally as
follows (with proof in the appendix):
Proposition 1. The unique subgame perfect equilibrium policy outcomes
of the Legislative Effectiveness-Pivots game consists of the following
spatial policy outcomes:
8 xL xV
>
> 0 if c > and xq 2xV or xq 2xF
>
> xV
>
>
>
> xL xV xL xV
>
> 2xV xq if c > and 2xV < xq < xV
>
>
>
> xV c
>
>
< xL xV xL xV
y 2xF xq if c >
2xF xq xV
and xF < xq < 2xF xV
c
>
>
>
> xL xV xL xV
>
> if c > and xV xq xF
>
> xq
>
> xq xV c
>
>
>
> xL xV
>
> xV otherwise
: c
While Proposition 1 appears somewhat cumbersome at first glance, it is

important to note that the first four policy options, y 0; 2xV xq ;
2xF xq ; xq g comport with the equilibrium policy locations of the Pivotal
Politics Theory. These all occur when c is large and thus it is too costly for
the Lawmaker to propose a quality alternative. The notable deviation from
the Pivotal Politics Theory occurs when the effective Lawmaker proposes a
bill with sufficient quality to yield a final policy that is located at
y xV xL c xV . This location is the weighted average between the
Lawmakers ideal policy and that of the Veto Pivot, where the weight is
determined by the relative effectiveness (c) of the Lawmaker. When the
Lawmaker is extremely effective (c very close to 1), the Lawmaker is able to
propose a policy close to her ideal point, and attach sufficient quality to it
that the Veto Pivot (as well as the Median and the Filibuster Pivot) would
prefer the Lawmakers proposal compared to what would be obtained in
the pivotal politics subgame. As the Lawmaker becomes relatively less effec-
tive, however (c increases above 1), her optimal proposal moves away from
her ideal point and closer to the Veto Pivot, who represents the binding
constraint in moving any policy away (to the right) from what would be
obtained in the pivotal politics subgame. Finally, as increasing bill quality
becomes prohibitively costly (i.e., c ), the Lawmaker essentially gives up,
in that she either makes no proposal, or perhaps proposes her ideal point
with zero quality (e.g., for position-taking purposes), which leads to the
pivotal politics equilibrium being obtained for all status quo policies in the
parameter space.
The intuition behind the equilibrium can be clearly gleaned by considering
Fig. 1, which identifies the relationship between status quo locations (the
x-axis) and equilibrium policy outcomes (the y-axis) from the Legislative
Effectiveness-Pivots game for the preference configuration that we consider
here (i.e., xP < xV < xM 0 < xF < 2xF < xL ), and for four kinds of lawmakers,
who vary in their relative degrees of effectiveness (as captured by the value
of c). To facilitate presentation, it is useful to focus attention on the two limit-
ing cases: the Highly Effective Lawmaker, as represented by the thick solid
Fig. 1. Equilibria in Legislative Effectiveness-Pivots Theory.

line at the top of the figure, and the Highly Ineffective Lawmaker, as repre-
sented by the thin solid line segments at the bottom of the figure. As
described above, we see that the policy outcome that corresponds to the
Highly Effective Lawmaker is constant in the status quo, in that for all status
quo locations, the Highly Effective Lawmaker is able to obtain a policy out-
come (y xV xL c xV ) quite near her ideal point. She couples this spatial
proposal with sufficient investment in bill quality as to make the Veto Pivot
at least indifferent between this proposal and the erstwhile result of the
pivotal politics subgame.
In contrast, for the Highly Ineffective Lawmaker, the final policy
outcome corresponds to what would be obtained in a Pivotal Politics equi-
librium. For relatively extreme status quo locations (xq 2xV or xq 2xF ),
the equilibrium outcome corresponds to the legislative medians ideal point.
For somewhat more moderate status quo locations (xq 2xV ; xV and
xq xF ; 2xF ), the equilibrium outcome corresponds to the reflection point
of the status quo around the relevant pivots ideal point (i.e., 2xV xq , and
2xF xq for the Veto and Filibuster Pivot, respectively). These are the clo-
sest possible policies to the Median that will be accepted by the (now indif-
ferent) relevant pivotal actor. Finally, for relatively centrist status quo
locations, xq xV ; xF , policy is gridlocked, meaning that it cannot be
moved, despite a (simple) majority preferring that it be changed.
Relative to the thick solid line near the top of the figure, as the costs (c)
of developing a high-quality proposal increase, the Highly Effective
Lawmaker invests less in bill quality. Instead, to continue to secure the
support of the Veto Pivot, she adjusts policy away from her own ideal point
toward that pivots preferred outcome. For sufficiently high costs, that
compromise becomes no better than what is attained in the Pivotal Politics
model without legislative effectiveness. For example, the Moderately
Effective Lawmaker, as represented by the thick dashed line, is able to
obtain y xV xL c xV for nearly all status quo policies, except for those
that correspond to a range of status quos somewhat between xM 0 and
2xF (centered around the filibuster pivots ideal point, xF ). For status quo
policies in this range, the policy outcome that would be obtained in a
Pivotal Politics equilibrium is actually more desirable to the Lawmaker
than what would be obtained from her own proposal.
Likewise, as Lawmakers become still less effective (c increases further),
y xV xL c xV moves farther away from the Lawmakers ideal point,
making the pivotal politics equilibrium outcome more desirable to the
Lawmaker than attaching additional quality to her policy proposal in an
effort to appease the veto pivot. Such a result can be illustrated by
considering the equilibrium policy locations that correspond to the Mostly

Ineffective Lawmaker, as represented by the thin dashed-and-dotted line.
For almost the entire parameter space, the equilibrium policy outcomes
mirror those that would be obtained in the standard pivotal politics analy-
sis. The exception to this pattern emerges for a small range of status quo
locations that are symmetric around the veto pivot, where the lawmaker is
able to (and prefers to) obtain y xV xL c xV along with an investment
in quality, rather than the Pivotal Politics equilibrium.
Hence, our analysis illustrates that, when little can be done to build up
common interests and improve the overall quality of proposals (and the
political benefits of supporting them), the Legislative Effectiveness-Pivots
model collapses to the Pivotal Politics Model as a special case. That said,
one would expect to be able to find some lawmakers and political entrepre-
neurs who are highly effective in many contexts; and they can therefore
help overcome the gridlock so common in such models. They include the
sort of policymakers who, like the late Congressman Dan Rostenkowsi
(D-IL), know how to cut through the bunk, make a deal, twist an arm, do a
favor, call in a chit, and move point A to point Z without a lot of philosophical
mumbo jumbo (Royko, 1994). Being able to identify and focus a nonmarket
strategy around such lawmakers may be extremely valuable in strategic
management, especially in an era in which policy gridlock is commonplace.
WHO ARE THE MOST EFFECTIVE LEGISLATORS?

In an effort to explore the causes and consequences of legislative effectiveness,
Volden and Wiseman (2012, 2014) developed a methodology for cardinally
ranking all members of the United States House of Representatives in their
proven ability to advance their legislative agenda items through several distinct
stages in the lawmaking process.11 More specifically, for each two-year
Congress, Volden and Wiseman identify the number of bills that each member
of the U.S. House sponsored (captured in a variable labeled BILL); and the
number of those bills that received action in committee (AIC), or action
beyond committee (ABC) on the floor of the House. For those bills that
received action beyond committee, they also identify how many of those bills
passed the House (PASS), and how many subsequently became law (LAW).
To account for variation in the substantive importance of legislation,
each bill is categorized as being either commemorative (C), substantive (S),
or substantive and significant (SS) based on the following coding protocol.
A bill is deemed substantive and significant if it had been the subject of an

end-of-the-year write-up in the Congressional Quarterly Almanac.12 A bill
was deemed commemorative if it satisfied any one of several criteria, such
as providing for a renaming, commemoration, private relief of an indivi-
dual, and the like. Finally, all other bills were classified as substantive.
Drawing on the 139,052 H.R. bills that were introduced between 1973
and 2008, across the 93rd through the 110th Congresses (8,478 of which
were commemorative, and 6,526 of which were substantive and significant),
Volden and Wiseman calculate Legislative Effectiveness Score (LES), for
each member i in each Congress t, as follows:
20 1 3
6B C 7
6B BILLCit BILLSit BILLSS C 7
6B it C 7
6B X N XN X N C 7
6@ A 7
6 BILLCjt BILLSjt BILLSS 7
6 jt 7
6 0 j1 j1 j1
1 7
6 7
6 7
6 B C 7
6 B AIC C
AIC S
AIC SS C 7
6 B it it it C 7
6 B X X X C 7
6 @ N N N
SS A
7
6 AIC C
AIC S
AIC 7
6 jt jt jt 7
6 7
6 0 j1 j1 j1
1 7
6 7
6 B C 7
6 B C 7
6 B ABCitC ABCitS ABCitSS C 7 N
6
LESit 6 B N C 7 5 7
6 @ X X X
N N
A 7
6 ABCjtC ABCjtS ABCjtSS 7
6 7
6 0 j1 j1 j1
1 7
6 7
6 7
6 B C7
6 B PASS C
PASS S
PASS SS C 7
6 B it it it C7
6 B X X X C 7
6 @ N N N 7
SS A 7
6 PASS C
PASS S
PASS
6 jt jt jt 7
6 0 j1 j1 j1
1 7
6 7
6 7
6 B C 7
6 B C 7
6 B LAWitC LAWitS LAWitSS C 7
6 B C 7
6 @ X N X N X N
A 7
4 LAW C LAW S LAW SS 5
jt jt jt
j1 j1 j1
where the five large terms represent the members fraction of bills (1) introduced,
(2) receiving action in committee, (3) receiving action beyond committee,
(4) passing the House, and (5) becoming law, relative to all N legislators.
Within each of these five terms, commemorative bills are weighted by ,
substantive bills by , and substantive and significant by . The overall
weighting of N/5 normalizes the average LES to take a value of 1 in each
Congress.13 Throughout their analysis in Volden and Wiseman (2012, 2014)
and in Volden, Wiseman, and Wittmer (2013), the authors assign = 1,
= 5, and = 10, signifying that substantive and significant legislation
exerts ten times the weight on the LES as commemorative legislation and
twice as much as normal substantive legislation. While such a weighting
scheme might seem arbitrary, as explicated in Volden and Wiseman (2014), the
weights were chosen by the authors to reflect the view that advancing a sub-
stantive and significant bill is more difficult than moving general substantive
legislation; and likewise, that advancing substantive legislation is a stronger
indicator of legislative effectiveness than is moving commemorative legislation.
Moreover, supplemental analysis by Volden and Wiseman demonstrate that
their core results are robust to alternative weighting schemes.14
Drawing on these Legislative Effectiveness Scores, Volden and Wiseman
demonstrate how the LES is positively correlated with a legislators party
status (i.e., whether she is in the majority party), seniority, committee and
party leadership status, and numerous other features that might be deemed
relevant to lawmaking processes. In addition, they also demonstrate (Volden &
Wiseman, 2014, pp. 5456) that accomplishments in a previous Congress,
even in earlier stages of the legislative process (e.g., bill introductions), are posi-
tively correlated with the number of laws produced by a legislator in a current
Congress. These results offer further justification for the decision to include all
five steps in the legislative process in the LES, rather than just focusing solely
on how many bills a legislator passes and/or gets signed into law.
While being able to identify which legislators are generally more successful
at advancing bills through the lawmaking process might be valuable in its
own right for managers who are seeking legislative advocates, Volden and
Wisemans methodology can be further refined in a manner to enhance its
relevance to managers. More specifically, Volden and Wiseman (2011, 2014)
draw on the Congressional Bills Project coding protocol (e.g., Adler &
Wilkerson, 2013) that categorizes bills into 1 of 19 policy areas, and then
employ their methodology to calculate Interest and Legislative Effectiveness
Scores (ILES) which are policy issue-specific legislative effectiveness scores.15
Drawing on these Interest and Legislative Effectiveness Scores, managers can
identify those lawmakers who are most successful (both in a particular
Congress, as well as previous Congresses) at advancing bills in specific policy
areas that might be relevant to the industries in which they operate.
AN ILLUSTRATION: BANKING AND FINANCE

IN THE GREAT RECESSION
To illustrate the efficacy and potential usefulness of our theoretical and
empirical approach, consider the state of the American economy in the
111th Congress (20092010). As a result of the Great Recession, a wide
range of firms, ranging from financial services to consumer retail, were
experiencing significant declines in their business, and were forced to lay
off sizable portions of their workforces. In such an environment, there was
substantial clamor among politicians (and the general public) for some
kind of financial services reform, to try to ensure that similar meltdowns
would not occur in the (near) future and to assist those individuals and
groups that had been most adversely affected by recent events. Given that the
United States had just elected a Democratic President (Barack Obama)
for the first time in eight years, and that both chambers of Congress were con-
trolled by the Presidents party, one might expect shifts in the identities of the
relevant pivotal actors, allowing for substantial policy change.
That said, the Legislative Effectiveness-Pivots Theory suggests that iden-
tifying the appropriate legislative advocate would still be of paramount
importance for managers and interest group leaders who sought to have
their agenda items advanced successfully through the legislative process.
Who should managers have looked to, if they hoped to achieve success in
legislative items related to banking matters, and financial services more
broadly considered? Some candidates for legislative advocacy are obvious.
Barney Frank (D-MA) was Chair of the House Financial Services
Committee in the 111th Congress, and would be expected to be at the fore-
front of advancing many banking and financial services bills. Likewise,
Nydia Velazquez (D-NY) was Chair of the House Small Business
Committee, and would therefore have significant influence over any legisla-
tion that dealt with the concerns of small businesses that were referred to
her committee. Given that the agendas of Frank and Velazquez were likely
to be clogged with a wide array of bills that constituted ordinary committee
business, however, one wonders whether other legislators, with less con-
strained agendas, could likely serve as viable advocates for a manager with
financial services interests?
In answering such questions, guidance might be gleaned by considering
the Banking Interest and Legislative Effectiveness Scores (B-ILES)
for those legislators who served in the preceding 110th Congress
(20072008). To illustrate this point, Table 1 presents the legislators who
had the 10 highest Banking ILESs in the 110th Congress, along with their
Table 1. Ten Highest Banking Interest and Legislative Effectiveness

Scores (B-ILES) in 110th Congress (20072008).
Name Party Banking ILES Chair Subcommittee Chair
Oberstar, James (MN-8) Democrat 23.53 Yes No

Rush, Bobby (IL-1) Democrat 22.35 No Yes
Chabot, Steve (OH-1) Republican 20.72 No No
Velazquez, Nydia (NY-12) Democrat 20.38 Yes No
Frank, Barney (MA-4) Democrat 19.61 Yes No
Capuano, Michael (MA-8) Democrat 19.08 No No
Kennedy, Patrick (RI-1) Democrat 18.87 No No
Maloney, Carolyn (NY-14) Democrat 15.91 No Yes
Kanjorski, Paul (PA-11) Democrat 11.61 No Yes
Conyers, John (MI-14) Democrat 10.20 Yes No
party affiliations, and their committee chair or subcommittee chair status

in that Congress.
As might be expected, Representatives who chaired committees that
dealt with banking and financial services matters, such as Barney Frank
(Financial Services) and Nydia Velazquez (Small Business) had notably
higher Banking ILESs in the 110th Congress than the average member of
the House (who had a Banking ILES of 1). We also see that James
Oberstar (D-MN), who was Chair of the Transportation and Infrastructure
Committee, as well as John Conyers (D-MI), who was Chair of the
Judiciary Committee, also had quite high Banking Scores, which is sensible
given that many bills flowing through their committees had relevance to
financial services.
Putting aside the more obvious candidates for potential advocates, the
Banking ILESs point to several other legislators who were also highly effec-
tive in advancing bills relating to banking and financial services. Three of
them (Bobby Rush, D-IL, Carolyn Maloney, D-NY, and Paul Kanjorski,
D-NY) held subcommittee chairs; two (Michael Capuano, D-MA, and
Patrick Kennedy, D-RI) were essentially rank-and-file members of the
majority party; and Steve Chabot (R-OH) was actually a minority party
member (and ranking member on the Small Business Committee). Hence,
even among the most highly effective legislators in banking and financial
services policy, the scope of success is distributed quite widely across differ-
ent institutional ranks and positions.
On a related note, Table 2 presents the minority party (Republican)
legislators who had the five highest Banking ILESs in that Congress, and
demonstrates that, despite lacking majority party status, several Republican
Table 2. Five Highest Banking Interest and Legislative Effectiveness

Scores (B-ILES) among Minority (Republican) Party in 110th
Congress (20072008).
Name Banking ILES LES
Chabot, Steve (OH-1) 20.72 1.76

Gallegly, Elton (CA-24) 9.54 1.14
Garrett, Scott (NJ-5) 7.98 1.06
Fallin, Mary (OK-5) 3.68 0.59
Castle, Michael (DE-1) 2.83 1.29
Representatives were quite successful at advancing bills relating to banking

and financial services matters. It is worth noting that with the exception of
Chabot, none of the top five minority party members were ranking members
of committees that would be expected to deal with financial services issues.16
Hence, if managers were focusing their attention on particular minority
party members, based solely on institutional positions, they would likely be
lobbying the wrong people.
Turning back to Table 1, several (non-committee-chair) Representatives
emerge as potential champions for firms seeking to advance their causes in
the 111th Congress. Patrick Kennedy (D-RI), Bobby Rush (D-IL), and
Carolyn Maloney (D-NY) all sponsored financial-services-oriented bills that
were all ultimately signed into law by President Bush (whereas the other
nonchairs that received high scores all fell short of producing new laws in
the 110th Congress). Of these three Representatives, Congresswoman
Maloney seemed like a particularly attractive target for lobbying on financial
serves matters.
First elected to Congress in 1992, Maloney was representing New Yorks
14th Congressional District in the 110th Congress, which covered large
portions of lower Manhattan, in addition to other parts of New York City.
Given the high density of banking interests that resided in her district,
Maloney had strong constituency-based motivations to maintain a focus on
legislative developments regarding financial services policies. While not
holding a committee chair in the 110th Congress, she was still a member
of the House Financial Services Committee, and was the Chair of the
Subcommittee on Financial Institutions and Consumer Credit, which
ensured that she had significant influence over bills dealing with a wide
range of issues relevant to banking and consumer credit interests. More
generally, Maloney had established a reputation over more than a dozen
years in office as being a focused legislator, obtaining successes through
perseverance when other lawmakers chose to focus their efforts on less

contentious matters.17
That she was such an effective lawmaker in the 110th Congress is parti-
cularly impressive given her ideological position vis-a-vis other members of
the House. Simply stated, Congresswoman Maloney was quite liberal, in
comparison to the Democratic caucus, and the chamber as a whole.18
Considering her ideology alone, one would not naturally expect her to
cultivate and advance sufficiently centrist legislation, so as to appease the
preferences of a more conservative-leaning President (i.e., George W. Bush).
In our view, based on the data presented here, Maloneys success came not
from her ideological position but from being a highly effective lawmaker,
similar to that illustrated in Fig. 1. Indeed, one might suspect that
Maloneys close constituency ties to banking interests likely provided her
with nuanced information about the industry that might have enhanced her
legislative effectiveness in this area. Therefore, managers seeking to advance
their causes in banking and financial services in the 111th Congress might
have been well served to consult with Representative Maloney about
their concerns.
As alluded to above, certain aspects of the political landscape in the
111th Congress suggested that significant policy changes might be obtained.
For the first time since 1994, the Congress and the White House were both
controlled by the Democratic Party. The Senate had shifted from being
essentially split between the parties (leaning slightly Democratic) to having
a solid Democratic majority. Moreover, the Democrats had expanded their
majority in the House by more than 20 seats, such that there was almost
an 80 seat difference between the parties. In the context of a pivotal politics
framework, a wide range of relatively right-leaning policies that had
been inherited from the previous Congress could be moved in a leftward
direction, as President Obama removed the credible veto threat of his
predecessor, President George W. Bush.
It should be no surprise that numerous banking and financial services
proposals were advanced by different legislators in the 111th Congress. By
the end of 2010, 607 bills had been introduced into the U.S. House that
dealt with banking and financial services, which was more than a 20%
increase over those introduced in 20072008 (and almost a 30% increase
over the number of banking-related bills introduced in 20052006). With
significant political pressure emerging to pass substantively meaningful
financial services reforms, it is clear that Congress was likely to do
something, and that managers would face significant competition for
access to the legislative agenda. Did the most effective lawmakers in the
Table 3. Ten Highest Banking Interest and Legislative Effectiveness

Scores (B-ILES) in 111th Congress (20092010).
Name Party Banking ILES Chair Subcommittee Chair
Frank, Barney (MA-4) Democrat 34.93 Yes No

Conyers, John (MI-14) Democrat 22.64 Yes No
Maloney, Carolyn (NY-14) Democrat 20.74 No No
Velazquez, Nydia (NY-12) Democrat 20.52 Yes No
Obey, David (WI-7) Democrat 20.08 Yes Yes
Waters, Maxine (CA-35) Democrat 16.38 No Yes
Rangel, Charles (NY-15) Democrat 15.06 Yes No
Altmire, Jason (PA-4) Democrat 12.59 No Yes
Kanjorski, Paul (PA-11) Democrat 9.61 No Yes
Oberstar, James (MN-8) Democrat 8.66 Yes No
110th Congress continue to be among the most effective lawmakers in the

subsequent Congress?
Turning to Table 3, which identifies those Representatives with the
10 highest Banking Interest and Legislative Effectiveness Scores in
the 111th Congress (20092010), we see that 6 of the 10 most effective
Banking and Financial Services lawmakers in the 111th Congress were
among the 10 most effective Banking and Financial Services lawmakers
in the 110th Congress. To some degree, this finding is unsurprising; after
all, four of these lawmakers continued to chair House committees, and
two of these Representatives (Frank and Velazquez) continued to chair
committees in the 111th Congress that had direct relevance to banking
and financial services matters (Financial Services Committee, and Small
Business Committee, respectively).19
What is notable, however, is that only one person on the top-10 list
held neither a chair nor a subcommittee chair in the 111th Congress; and
(ironically) she was also among the top-10 most effective banking and
financial services lawmakers in the 110th Congress: Carolyn Maloney.
Despite no longer chairing a subcommittee of the House Financial Services
Committee, Congresswoman Maloney continued to be among the most
effective lawmakers in this policy area. She was an active participant in the
debates surrounding the passage of Dodd-Frank, and she introduced a
dozen bills that dealt with banking and financial services matters. Many of
these bills received some sort of action in committee, and three of them
passed the House of Representatives. Maloneys crowning achievement was
H.R. 627, the Credit Card Accountability Responsibility and Disclosure Act
of 2009 (popularly referred to as the Credit Cardholders Bill of Rights),

which was signed into law by President Obama on May 22, 2009; and her
involvement in the passage of this bill serves as a nice illustration of the
strategic considerations that are present in our model.
The bill amended the Truth in Lending Act, the FTC Act, and the
Electronic Funds Act, to provide for numerous enhancements to standard
consumer protection measures for credit card issuance and usage, which
constrained creditors activities in regards to usage and late fees, as well as
mandating a wide range of information disclosures. In the context of our
model, the status quo policy was clearly located to the right of the House
median, which pivotal politics theories would suggest would correspond to
either no policy change, or a new policy that was located at, or to the right
of, the House median (depending on where the status quo was located
relative to the filibuster pivot). The Legislative Effectiveness-Pivots Theory,
in contrast suggests an even further policy change, due to the high effective-
ness of Rep. Maloney. While it is difficult to map the location of H.R. 627
with precision, it is worth noting that the measure had a profound impact on
the conventional practices of credit card companies and was the source of
intense lobbying by credit card and consumer interests. The bill clearly
represented a substantial leftward shift in policy from the status quo; and it
was very likely to the left of the House median legislator.
It is well documented that Maloney was the target of significant
lobbying efforts by banking interests on this measure. Those managers and
interest groups who had identified Maloney as a highly effective lawmaker
in this area may have been well served by their efforts. As an effective
lawmaker, the Congresswoman had significant leeway in designing the
policys specific provisions, and could therefore be open to a broad political
compromise. To an extent, willingness to reach such compromises may
help make a lawmaker effective in the first place. And such compromises
were evident on this particular bill. For example, to the displeasure of
consumer advocates, the bill allowed a nine-month lag time between when
the bill was signed, and when many of its provisions were scheduled to take
effect (in February 2010). Moreover, provisions relating to fees and
increases in credit card interest rates were not scheduled to be implemented
for at least 15 months after the passage of the Act. While several liberal
legislators, such as Barney Frank, were unhappy with how Maloney suppo-
sedly caved to the pressure of big banking interests (Kaiser, 2013, p. 101),
as an effective lawmaker, Maloney produced a bill with broad appeal even
across party lines, one that passed a typically highly polarized and partisan
House by a vote of 35770.
CONCLUSION
Conventional analyses of competitive strategy have focused on how managers
might optimize their firms positions given the market structure that they face.
A broader perspective, however, appreciates that managers can affect market
structure through their nonmarket actions. That is, by passing laws and
promulgating regulations, legislatures and bureaucratic agencies essentially
establish the rules of the game that managers play. Hence, strategies aimed at
securing and enhancing the welfare of a firm must focus on market and
nonmarket factors to ensure that the firm is well-positioned vis-a-vis its
competitors, both actual and latent.
In developing tools for nonmarket strategy, earlier theories have focused
on the roles of pivotal actors in lawmaking institutions, with an eye toward
cultivating sufficiently large (and often supermajority) support for a managers
policy proposals. Such theories emphasize institutional actors in committees
and political parties, as well as rules, such as those governing vetoes and
filibusters. As insightful as these theories have been, they have all neglected to
engage the relative efficacy of legislators as lawmakers. To the extent that one
might believe that legislators vary in their lawmaking competence, these pre-
vailing theories cannot speak to such variance. Hence, managers have received
little guidance as to whom they should recruit as policy advocates for their
respective causes.
In this essay, we have emphasized the salience of legislative effectiveness
for understanding the dynamics and determinants of contemporary law-
making. We have argued that managers could draw on these theoretical
insights to recognize potentially effective lawmakers. Moreover, we have
provided an empirical strategy for identifying highly effective lawmakers,
and illustrated the utility of our approach by considering the politics
surrounding a particularly contentious issue (and era). As demonstrated by
our analysis, differential effectiveness can be detected across lawmakers.
And lawmakers who are relatively effective now are likely to be effective in
the future. Therefore, managers who are seeking out advocates for their
firms causes in Congress today, for example, should look to previous
Congresses to identify those lawmakers who have a track record of being
successful in advancing similar causes.
Moving beyond our consideration of recent events, it is important to
note that the theoretical perspective we advance is not confined to analysis
of the U.S. Congress. In any collective decision-making body (regulatory
or legislative, partisan or nonpartisan, committee-based or broader), some
participants are more skilled than others at advancing their causes.
Managers who engage nonmarket institutions should recognize these differ-

ences; and they should understand when and where such highly effective
policymakers will help advance their strategic nonmarket objectives.
The theory, empirical approach, and examples advanced here illustrate the
benefits of incorporating legislative effectiveness into nonmarket strategy.
NOTES
1. More generally speaking, in the few instances that government is referred to in
Porters volume, he essentially treats it as a force that establishes exogenous barriers
that firms must take into account when devising a successful market strategy, rather
than considering the ways in which firms might (successfully) seek to influence the
government-mandated rules of the playing field.
2. Diermeier (2011), for example, presents a collection of theories and frameworks
for managing a companys reputational concerns, several of which provide explicit
guidance for how to navigate interactions with the media, interest groups, and various
political institutions and office holders.
3. Brady and Volden (1998, 2006) present a theory of lawmaking with pivotal
institutional actors, complementing and extending Krehbiels work.
4. Pivotal voters also figure prominently in Barons (2013, pp. 174178) discussion
of various majority-building strategies in government arenas.
5. For the purposes of illustration (and simplification), we analyze legislative
politics within only one chamber (implicitly the Senate), rather than both chambers
of the U.S. Congress.
6. A rapidly developing theoretical literature has begun to explore the role of quality
dimensions (i.e., valence) in electoral (e.g., Aragones & Palfrey, 2002; Ashworth &
Bueno de Mesquita, 2009; Groseclose, 2001; Meirowitz, 2008; Serra, 2010; Wiseman,
2005, 2006), and legislative (e.g., Hirsch & Shotts, 2012, 2013) politics.
7. An alternative way to model these types of interactions would be to assume
that agenda setters could provide targeted benefits to individual legislators in
exchange for their votes (i.e., Krehbiel, Meirowitz, & Wiseman, 2015; Snyder,
1991). While vote-buying might rightfully be considered among the tools available
to effective lawmakers, the model that we advance here is meant to capture the idea
that effective lawmakers have more tools at their disposal than only building support
for their proposals individually through particularistic benefits (e.g., Volden &
Wiseman, 2007).
8. Such costs might be related to the time and effort that a Lawmaker must
devote to bringing together pivotal decision makers to the bargaining table, gaining
policy expertise, engaging in research that is then publicized to emphasize the
positive aspects of the bill, and so on.
9. The assumption that c 1 implies that the marginal costs from producing
attractive legislation are at least as high as the marginal benefits that the Lawmaker
receives from said bills. If this assumption did not hold and c < 1, the Lawmakers
problem would be trivial, as she would seek to exert an infinite amount of effort to
maximize the quality of a new bill.
10. This is a nonessential technical assumption to allow us to focus on a relatively

limited number of cases. If we assumed that the Lawmaker were less extreme than
2xF , the results that emerge would be substantively similar to the analysis presented
here. Interested readers are referred to Hitt et al. (2014) for a more complete
consideration of ideal point configurations in the Legislative Effectiveness Model.
11. Other efforts to measure legislative effectiveness (building on Matthewss, 1960
study) have investigated which legislators (e.g., members of the U.S. House of
Representatives) pass the greatest number of laws (e.g., Anderson, Box-Steffensmeier, &
Sinclair-Chapman, 2003; Cox & Terry, 2008; Frantzich, 1979) or have the highest bill
passage rates (e.g., Bratton & Haynie, 1999; Hamm, Harmel, & Thompson, 1983)
Scholars (e.g., Meyer, 1980; Padro i Miquel & Snyder, 2006; Weissert, 1991) have also
analyzed how the reputational rankings of legislators perceived effectiveness
correlates with various personal and institutional considerations.
12. It should be noted that CQ Almanac stories are not ex ante measures of bill
significance, as bills that move further through the lawmaking process are much
more likely to be mentioned.
13. As Volden and Wiseman note, because their approach generates scores
separately within each Congress, over-time comparisons must be made with caution,
given different agenda sizes and productivity across Congresses.
14. More specifically, Volden and Wiseman (2014, pp. 5658) find that the
substantive effects regarding the correlates of the LES are essentially robust to
alternative specifications with < < .
15. The Congressional Bills Project draws on the coding protocol that was developed
by Baumgartner and Bryan (2002) as part of their Policy Agendas Project. The nineteen
major topic categories are (in alphabetical order): Agriculture; Banking & Commerce;
Civil Rights & Liberties; Defense; Education; Energy; Environment; Foreign
Trade; Government Operations; Health; Housing & Community Development;
International Affairs; Labor, Employment, & Immigration; Law, Crime, &
Family; Macroeconomics; Public Lands; Science & Technology; Social Welfare;
and Transportation.
16. Spencer Bachus (AL-6), the ranking member of the House Financial Services
Committee had a Banking ILES of 2.46 in the 110th Congress.
17. Related to this point, Volden et al. (2013, p. 333) discuss how Representative
Maloney continued to advocate for her legislative agenda items after the Republican
Party took over the House in the 104th Congress (19951996) while other
Democratic lawmakers (e.g., Chuck Schumer, NY) stopped sponsoring bills that they
had previously introduced (yet had been unsuccessful in advancing) when the
Democratic Party controlled the Congress.
18. Congresswoman Maloneys DW-NOMINATE score in the 110th Congress
was 0.442, which was more left-leaning than the House median (0.177) as well
as the Democratic Party median (0.406); and it placed her among the most liberal
quartile of the House in the 110th Congress.
19. Two new additions to the top-10 list, David Obey (D-WI) and Charles
Rangel (D-NY) held influential positions on two committees that would be expected
to deal with banking and financial services matters in some capacity
(Appropriations and Ways and Means, respectively).
20. Given the assumed preference alignment, the Filibuster Pivot will support the
Lawmakers proposal if it gains the support of these more distant actors.
ACKNOWLEDGMENTS
The authors thank David Baron, Keith Krehbiel, John de Figueiredo, and
an anonymous reviewer for feedback on an earlier version of this paper.
REFERENCES
Adler, E. S., & Wilkerson, J. D. (2013). Congress and the politics of problem solving.
New York, NY: Cambridge University Press.
Anderson, W. D., Box-Steffensmeier, J. M., & Sinclair-Chapman, V. (2003). The keys to legislative
success in the U.S. House of Representatives. Legislative Studies Quarterly, 28(3), 357386.
Aragones, E., & Palfrey, T. R. (2002). Mixed equilibrium in a Downsian model and a favored
candidate. Journal of Economic Theory, 103, 131161.
Ashworth, S., & Bueno de Mesquita, E. (2009). Elections and platform with valence competition.
Games and Economic Behavior, 67(1), 191216.
Baron, D. P. (1999). Integrated market and nonmarket strategies in client and interest group
politics. Business and Politics, 1(1), 734.
Baron, D. P. (2001). Theories of strategic nonmarket participation: Majority-rule and executive
institutions. Journal of Economics and Management Strategy, 10(1), 745.
Baron, D. P. (2005). Competing for the public through the news media. Journal of Economics
& Management Strategy, 14(2), 339376.
Baron, D. P. (2013). Business and its environment (7th ed.). New York, NY: Pearson.
Baron, D. P. (2014). Self-regulation in private and public politics. Quarterly Journal of
Political Science, 9(2), 231267.
Economics and Management Strategy, 16(3), 599634.
Baumgartner, F. R., & Bryan, D. J. (Eds.). (2002). Policy dynamics. Chicago, IL: University of
Chicago Press.
Black, D. (1948). On the rationale of group decision-making. Journal of Political Economy,
56(1), 2334.
Brady, D. W., & Volden, C. (1998). Revolving Gridlock: Politics and policy from Carter to
Clinton. Boulder, CO: Westview Press.
Brady, D. W., & Volden, C. (2006). Revolving Gridlock: Politics and policy from Jimmy Carter
to George W. Bush (2nd ed.). Boulder, CO: Westview Press.
Bratton, K. A., & Haynie, K. L. (1999). Agenda setting and legislative success in state legislatures:
The effects of gender and race. Journal of Politics, 61(3), 658679.
Cox, G. W., & Terry, W. C. (2008). Legislative productivity in the 93rd-105th Congresses.
Legislative Studies Quarterly, 33(4), 603618.
De Figueiredo, J. M. (2009). Integrated political strategy. Advances in Strategic Management,
26, 459486.
De Figueiredo, J. M., & Kim, J. J. (2004). When do firms hire lobbyists? The organization of
lobbying at the federal communications commission. Industrial and Corporate Change,
13(6), 883900.
De Figueiredo, J. M., & Tiller, E. H. (2002). The structure and conduct of corporate lobbying:
How firms lobby the federal communications commission. Journal of Economics and
Management Strategy, 10(1), 91122.
Diermeier, D. (2011). Reputation rules: Strategies for building your companys most valuable
asset. New York, NY: McGraw Hill.
Downs, A. (1957). An economic theory of democracy. New York, NY: Harper Collins.
Feddersen, T. J., & Gilligan, T. W. (2002). Saints and markets: Activists and the supply of
credence goods. Journal of Economics and Management Strategy, 10(1), 149171.
Frantzich, S. (1979). Who makes our laws? The legislative effectiveness of members of the
U.S. congress. Legislative Studies Quarterly, 4(3), 409428.
Groseclose, T. (2001). A model of candidate location when one candidate has a valence advantage.
American Journal of Political Science, 45(4), 862886.
Hamm, K. E., Harmel, R., & Thompson, R. (1983). Ethnic and partisan minorities in two
Southern state legislatures. Legislative Studies Quarterly, 8(2), 177189.
Hirsch, A. V., & Shotts, K. W. (2012). Policy-specific information and informal agenda power.
Hirsch, A. V., & Shotts, K. W. (2013). Competitive policy entrepreneurship. Unpublished
manuscript, Princeton University.
Hitt, M. P., Volden, C., & Wiseman, A. E. (2014). Spatial models of legislative effectiveness.
Unpublished manuscript, Vanderbilt University.
Holburn, G. L. F., & Vanden Bergh, R. G. (2004). Influencing agencies through pivotal political
institutions. Journal of Law, Economics, and Organization, 29(2), 458483.
Kaiser, R. G. (2013). Act of Congress: How Americas essential institution works, and how it
doesnt. New York, NY: Alfred A. Knopf.
Krehbiel, K. (1996). Institutional and partisan sources of Gridlock. Journal of Theoretical
Politics, 8(1), 740.
Chicago Press.
Krehbiel, K. (1999). Pivotal politics: A refinement of nonmarket analysis for voting institutions.
Business and Politics, 1(1), 6382.
Krehbiel, K., Meirowitz, A., & Wiseman, A. E. (2015). A theory of competitive partisan lawmak-
ing. Political Science Research and Methods, 3(3), 423448.
Lyon, T. P., & Maxwell, J. W. (1999). Corporate environmental strategies as tools to influence
regulation. Business Strategy and the Environment, 8, 189196.
Lyon, T. P., & Maxwell, J. W. (2004). Astroturf: Interest group lobbying and corporate strategy.
Journal of Economics and Management Strategy, 13(4), 561597.
Matthews, D. R. (1960). U.S. Senators and their world. Chapel Hill, NC: University of North
Carolina Press.
Meirowitz, A. (2008). Electoral contests, incumbency advantages, and campaign finance.
Journal of Politics, 70(3), 681699.
Meyer, K. (1980). Legislative influence: Toward theory development through causal analysis.
Padro i Miquel, G., & Snyder, J. M., Jr. (2006). Legislative effectiveness and legislative careers.
Porter, M. E. (1980). Competitive strategy: Techniques for analyzing industries and competitors.
New York, NY: The Free Press.
Royko, M. (1994, May 26). Big fish shouldnt be fried this time. Chicago Tribune, p. 3.
Serra, G. (2010). Polarization of what? A model of elections with endogenous valence. Journal
of Politics, 72(2), 426437.
Snyder, J. (1991). On buying legislatures. Economics and Politics, 3(2), 93109.
Volden, C., & Wiseman, A. E. (2007). Bargaining in legislatures over particularistic and collective
goods. American Political Science Review, 101(1), 7992.
Volden, C., & Wiseman, A. E. (2011). Breaking Gridlock: The determinants of health policy
change in Congress. Journal of Health Politics, Policy and Law, 36(2), 227264.
Volden, C., & Wiseman, A. E. (2012). Legislative effectiveness and representation. In L. C.
Dodd & B. I. Oppenheimer (Eds.), Congress reconsidered (10th ed.). Washington, DC:
CQ Press.
Volden, C., & Wiseman, A. E. (2014). Legislative effectiveness in the United States Congress:
The lawmakers. New York, NY: Cambridge University Press.
Volden, C., Wiseman, A. E., & Wittmer, D. E. (2013). When are women more effective lawmakers
than men? American Journal of Political Science, 57(2), 326341.
Weissert, C. S. (1991). Issue salience and state legislative effectiveness. Legislative Studies
Quarterly, 16(4), 509520.
Wiseman, A. E. (2005). Partisan strategy and support in state legislative elections: The case of
Illinois. American Politics Research, 33(3), 376403.
Wiseman, A. E. (2006). A theory of partisan support and entry deterrence in electoral competition.
Journal of Theoretical Politics, 18(2), 123158.
Yoffie, D. B. (1988). The politics of business: How and industry builds political advantage.
Harvard Business Review, 73, 8289.
Yoffie, D. B., & Bergenstein, S. (1985). Creating political advantage: The rise of the corporate
political entrepreneur. California Management Review, 28(1), 124139.
APPENDIX
Proof of Proposition 1. We derive the subgame perfect Nash equilibrium by

backwards induction. If the Lawmakers proposal is not accepted, then the
equilibrium of the subsequent subgame is the well-understood pivotal politics
equilibrium. That is, for xq 2xV and xq 2xF , the final policy location will cor-
respond to the Medians ideal point (xM = 0). For xq 2xV ; xV , or
xq xF ; 2xF , the final policy will correspond to the reflection of the status quo
around the relevant pivots ideal point (e.g., 2xV xq for the Veto pivot), and
for xq xV ; xF , policies will be gridlocked, meaning that the final policy will
be the same as the status quo.
Hence, given that xV < xM < xF < 2xF < xL , when the Lawmaker is considering
whether to make her proposal, she knows that whatever proposal (and
corresponding quality) that she might offer has to be weakly preferred by both
the Median and the Veto Pivot over the equilibrium policy outcome that will
ensue in the pivotal politics subgame.20
More specifically, if we consider any status quo location that corresponds
to a final outcome at xM in the pivotal politics subgame (i.e., for xq 2xV
and xq 2xF ), for any proposed bill, xb, with quality gb, it must be true that:
x2b gb 0
(i.e., the Median weakly prefers the legislative proposal to a policy located
at her ideal point). It must also be true that:
xV xb 2 gb x2V
(i.e., the Veto Pivot weakly prefers the legislative proposal to a policy
located at the Medians ideal point). Given that xV < xM 0, one of the
above expressions must be a strict inequality, in order for a new proposal
to defeat the status quo. More specifically, the Veto Pivots preferences
represent the binding constraint, which implies that for any bill, xb, that is
proposed, the bill quality, gb, must be equal to: x2b 2xV xb . Moreover, it
must also be true that the Lawmaker would prefer to propose the bill (with
such a quality) compared to simply proposing her ideal point with zero
quality, and ending up with the Medians ideal point as the final policy.
That is, it must be true that: xL xb 2 gb 1 c x2L .
Hence, the Lawmaker will choose xb, gb, to maximize:
xL xb 2 gb 1 c
such that: gb x2b 2xV xb
and xL xb 2 gb 1 c x2L :
Applying the calculus and solving for the optimal bill (and quality level)
yield the following equilibrium proposal, if the Lawmaker chooses to propose
a bill with nonzero quality (i.e., her participation constraint is not binding):

xL xV xV xL cxV cxV xL xV
xb ; gb xV ;
c c2
Moreover, we can identify that the Lawmakers participation constraint binds

at xL xV 1 c. More specifically, for xL < xV 1 c, the Lawmaker
would strictly prefer to leave the lawmaking to the Median. Hence, the
Lawmaker will only make the above proposal when xL > xV 1 c, and
otherwise will propose her ideal point with no quality attached (leading to a
final policy of xM = 0).
A similar logic follows for the other regions of the parameter space. As
noted above, for xq 2x
V ; xV , the equilibrium policy in the pivotal politics
subgame is 2xV xq ; and the Veto Pivots preferences are (again) the
binding constraint. Hence, for any proposed bill, xb, with quality gb, to be
passed it must be true that:
2
xV xb 2 gb xV 2xV xq
located at the reflection of the status quo around her ideal point). Given
that the above inequality will be binding in equilibrium, it must be true
that for any bill xb that is proposed, the attached level of quality, gb, must
be equal to: x2b 2xV xb 2xV xq x2q . Moreover, it must be true that the
Lawmaker would prefer to propose the bill (with quality attached)
compared to proposing
her ideal point with no quality attached, and ending
up with 2xV xq as the final policy. That is, it must be true that:
2
xL xb 2 gb 1 c xL 2xV xq
Hence, the Lawmaker will choose xb and gb to maximize:
xL xb 2 gb 1 c
such that: gb x2b 2xV xb 2xV xq x2q

2
and xL xb 2 gb 1 c xL 2xV xq :
yield the following equilibrium proposal, if the Lawmaker chooses to
make a proposal with nonzero quality (i.e., her participation constraint is
not binding):

xL xV xL xV 2 2
xb ; gb xV ; x V x q
c c2
Moreover, analysis reveals that the Lawmakers participation

constraintstarts
binding at xq xV xL c xV . Hence, whenever xq 2xV ; xV xL c xV , the
Lawmaker will propose her ideal point with no quality attached, which will lead
to the final policy being 2xV xq , whereas whenever xq xV xL c xV ; xV ,

2
the Lawmaker will propose xb ; gb xV xL c xV ; xL c2xV xV xq .
2
For xq xV ; xF , the equilibrium policy in the pivotal politics subgame is

xq, (and the Veto Pivots preferences are still the binding constraint).
Hence, for any proposed bill, xb, with quality level gb to be passed it must
be true that:
xV xb 2 gb xV xq 2
located at the status quo). Given that the above inequality will be binding
in equilibrium, it must be true that for any bill, xb, that is proposed, the
attached level of quality, gb, must be equal to: x2b 2xV xb 2xV xq x2q .
Moreover, it must be true that the Lawmaker would prefer to propose the
bill (with quality attached) compared to proposing her ideal point with no
quality attached, and ending up with the status quo as the final policy.
2
That is, it must be true that: xL xb 2 gb 1 c xL xq .
xL xb 2 gb 1 c
such that: gb x2b 2xV xb 2xV xq x2q

2
and xL xb 2 gb 1 c xL xq :
yield the following equilibrium proposal, if the Lawmaker chooses to
not binding):

xL xV xL xV 2 2
xb ; gb xV ; xV xq
c c2
Moreover, analysis reveals that the Lawmakers participation

constraint
starts
binding at xq xV xL c xV . Hence, whenever xq xV ; xV xL c xV , the
2
Lawmaker will propose xb ; gb xV xL c xV ; xL c2xV xV xq ,
2
whereas whenever xq xV xL c xV ; xF , the Lawmaker will propose her ideal

point with no quality attached, which will lead to the final policy being xq.
xq xF; 2xF , the equilibrium policy in the pivotal politics
Finally, for
subgame is 2xF xq ; and the Veto Pivots preferences are (again) the
binding constraint. Hence, for any proposed bill, xb, with quality level gb to
be passed it must be true that:
2
xV xb 2 gb xV 2xF xq
(i.e., the Veto Pivot weakly prefers the legislative proposal to a policy located
at the reflection of the status quo around the Filibuster Pivots ideal point).
Given that the above inequality will be binding in equilibrium, it must be true
that, for any bill xb that is proposed, the attached level of quality, gb, must be
equal to: x2b 2xV xb 4xV xF 2xV xq 4x2F 4xF xq x2q . Moreover, it
must be true that the Lawmaker would prefer to propose the bill (with quality
attached) compared to proposing her ideal point with no quality attached,
and ending up with 2xF xq as the final policy. That is, it must be
true that:
2
xL xb 2 gb 1 c xL 2xF xq
xL xb 2 gb 1 c
such that: gb x2b 2xV xb 4xV xF 2xV xq 4x2F 4xF xq x2q

2
and xL xb 2 gb 1 c xL 2xF xq :
yields the following equilibrium proposal, if the Lawmaker chooses to
not binding):

xL xV xL xV 2 2
xb ; gb xV ; xV 2xF xq
c c2
Moreover, analysis reveals that the Lawmakers participation constraint

starts binding at xq 2xF xV xL c xV . Hence, whenever xq

xF ; 2xF xV xL c xV , the Lawmaker will propose his ideal point with
no quality attached, which will lead to the final policy being 2xF xq ,

whereas whenever xq 2xF xV xL c xV ; 2xF , the Lawmaker will propose
2
xb ; gb xV xL c xV ; xL c2xV xV 2xF xq .
2
The c constraints in Proposition 1 follow immediately from the

Lawmakers participation constraints within each region of the equilibrium.
A UNIFIED MODEL OF
POLITICAL RISK$
Benjamin A. T. Graham, Noel P. Johnston

and Allison F. Kingsley
ABSTRACT
Political risk is a complex phenomenon. This complexity has incentivized

scholars to take a piecemeal approach to understanding it. Nearly all
scholarship has targeted a single type of political risk (expropriation)
and, within this risk, a single type of firm (MNCs) and a single type of
strategic mechanism through which that risk may be mitigated (entry
mode). Yet political risk is actually a collection of multiple distinct
risks that affect the full spectrum of foreign firms, and these firms vary
widely in their capabilities for resisting and evading these risks. We offer
a unified theoretical model that can simultaneously analyze: the three
main types of political risk (war, expropriation, and transfer restrictions);
the universe of private foreign investors (direct investors, portfolio equity
investors, portfolio debt investors, and commercial banks); heterogeneity
in government constraints; and the three most relevant strategic capabil-
ities (information, exit, and resistance). We leverage the variance among
$
Author order is alphabetical by convention. All authors contributed to this
paper equally.

ISSN: 0742-3322/doi:10.1108/S0742-332220160000034005
119
120 BENJAMIN A. T. GRAHAM ET AL.
foreign investors to identify effective firm strategies to manage political

risk. By employing a simultaneous and unified model of political risk, we
also find counterintuitive insights on the way governments trade off
between risks and how investors use other investors as risk shields.
Keywords: Political risk; foreign investment; strategy; emerging
markets; game theory
INTRODUCTION
Investors in Argentina are a bit like boat owners. The day they buy and the
day they sell are their two best days; in between theres often heartbreak.
This past century the Argentine government has serially defaulted on its
sovereign debt, expropriated countless foreign companies, repeatedly jacked
up its currency controls and transfer restrictions, and even invaded an island.
Yet numerous multinational companies, an array of institutional investors,
and even brave individuals have invested in Argentina. The ones who made
money have done so almost exclusively through strategic execution. Some
foreign investors partner with the government or influential local companies,
or gain privileged information from Argentine elites; many choose liquid or
hot investments, while others mount fierce resistance to any adverse gov-
ernment action. How can we understand which investors use what strategies
in the face of what type of political risk? When can being smart, fast, or
strong manage the risks of war, expropriation and transfer restrictions?
Existing research offers little satisfying theory to simultaneously
explain these observations about investment in a politically risk country
like Argentina. Admittedly political risk is a complex phenomenon. This
complexity has incentivized scholars to take a piecemeal approach to
understanding it. Nearly all scholarship has targeted a single type of
political risk and, within this risk, a single type of investor and a single
type of strategic mechanism through which that risk may be mitigated.
In this paper, we aim to develop a unified theoretical model of political
risk that better explains these heterogeneous observations about risks,
investors, and strategic capabilities.
While scholarship has focused primarily on expropriation risk, in the
case of Argentina and others, we observe that investors are exposed to
heavy losses from other political risks. Political risk is in fact a collection
A Unified Model of Political Risk 121
of multiple distinct risks, only one of which is expropriation. We begin

with a broad definition of political risk as the risk that the host government
fails to uphold the terms of an implicit contract between the host govern-
ment and an investor. This gives rise to a typology of political risk that
organizes the distinct ways in which this implicit contract may be violated.
We identify war risk, expropriation risk, and transfer risk as the three main
types of political risk.
Within this typology, the ravages of war constitute a violation of the
host governments commitment to protect the investors assets and person-
nel from violence. Expropriation risk encompasses more direct actions
taken by the host government to seize the assets of foreign investors,
including the outright nationalization of assets. We also focus on transfer
risk as a particularly common and costly form of creeping expropriation.
Transfer risk is the risk of restrictions on investors ability to convert and
transfer capital out of the host country. As the world has become gradually
more peaceful and as outright nationalization has become steadily less
common, transfer risk has emerged as the most pervasive and costly politi-
cal risk faced by foreign investors (Graham, Johnston, & Kingsley, 2015).
For instance, the Argentine government confiscates through exchange controls
in only a matter of weeks the equivalent of one massive, headline-grabbing
expropriation like YPF Repsol.1
While foreign direct investment (FDI) by multinational companies
(MNCs) dominates research on international investment in Argentina and
across emerging markets, other flows of private foreign capital are rising
in prevalence. We disaggregate foreign investment into four constituent
segments, each of which is associated with a particular type of firm.2 Direct
investors, usually MNCs, occupy controlling (>10%) ownership positions
in local companies. Portfolio equity and portfolio debt investors, usually large
institutional investors such as pension funds, mutual funds, and insurance
companies, own shares (<10%) or corporate bonds of local companies
purchased on local public exchanges. Bank lenders, that is, foreign commercial
banks, issue loans to diverse enterprises in the host country.
As Fig. 1 shows, FDI remains the single largest source of foreign capital
into emerging markets. However, in most years FDI accounts for less than
half of foreign investment.3 This suggests that the traditional focus on
MNCs limits academic understanding of foreign investors and may fail to
explain other substantively important global capital firms. Investigating
variation across the universe of private capital investors can thus provide
insight into the calculus of different types of firms and investors beyond
simply the MNC.4
600
Annual Net Inflows (Billions of USD)
FDI Bank Debt

Portfolio Equity Portfolio Debt
400
200
1978 1988 1998 2008 2012

Year
Fig. 1. Investment Volume into Emerging Markets.
In Argentina and beyond, we observe significant heterogeneity in inves-

tor strategies to mitigate and manage political risk. Research identifies
institutional characteristics of the host government, such as domestic
political constraints, that affect political risk (e.g., Delios & Henisz, 2003;
Henisz, 2000; Jensen, 2003, 2006; Li & Resnick, 2003; Weingast, 1995) but
political institutions are largely exogenous to investors and fail to fully
explain persistent economic institutions (Acemoglu & Robinson, 2006;
Alston, Harris, & Mueller, 2008) or many types of losses investors face,
notably transfer risk (Graham et al., 2015). Another rich body of work
exists on the roles of entry mode (e.g., Anderson & Gatignon, 1986;
Brouthers, 2002; Hennart, 2009) and manager experience or learning
(e.g., Delios & Beamish, 2001; Maitland & Sammartino, 2014) as means to
mitigate political risk. The strategy literature has further found that firms
develop nonmarket strategies to manage the political environment
(e.g., Baron, 1995; Bonardi, Hillman, & Keim, 2005; Hillman & Hitt, 1999;
Kingsley & Vanden Bergh, 2015; Kingsley, Vanden Bergh, & Bonardi,
2012), even as it changes over time (Emmons, 2000; Henisz & Zelner,
2005). We build on these existing insights, developing a structured and
comprehensive account of the wide array of relevant investor capabilities.
Here, we identify and analyze three main categories of investor capability:
level of information, ease of exit, and ability to resist. We leverage this
variance among our universe of investors to identify effective firm strategies
to manage political risk.
To integrate risks, investors, and capabilities, we construct a unified

model of political risk that simultaneously analyzes: the three main types of
political risk (war, expropriation, and transfer restrictions); the universe
of private foreign investors (direct investors, portfolio equity investors,
portfolio debt investors, and commercial banks); and the three most
relevant investor capabilities (information, exit, and resistance). We employ
this model to evaluate equilibria in which foreign investors will enter a host
country under threat of war, transfer restriction, and expropriation. The
game features two players: a host government and a foreign investor. We
evaluate the conditions under which: investment occurs; the government
expropriates; the government imposes transfer restrictions; and the foreign
investor chooses to expedite repatriation of assets. We model the behavior
of the average foreign investor in a given market. However, we specify
the model in such a way that we can assign varying capabilities to that
investor, which we match theoretically to different classes of investor, such
as direct investors or foreign banks. Similarly, we can vary the incentives of
the host government, allowing us to explore the implications of variation in
regime type or the existing portfolio of investors. Lastly, we can also vary
the risk of war. Thus, by assessing comparative statics of the model, we can
assess the changing implications for both government and investor behavior
as we vary risk-type, investor-type, and players capabilities endowment.
The paper proceeds, first, by developing a typology of political risk and
introducing the formal model. We then present theory articulating how
investor capabilities vary across classes of investment and we map these
capabilities to specific parameter values in the model. Next, we exploit
comparative statics from the model to analyze which capabilities mitigate
investors exposure to which political risks paying particular attention to
resistance capabilities. This analysis generates both testable hypotheses
for an empirical research program and direct implications for the strategy
of firms investing overseas. We conclude by discussing the (counterintuitive)
insights that a simultaneous and unified model of political risk offer scholars
of governments and firms.
THEORY: A UNIFIED MODEL OF POLITICAL RISK

Disaggregating Political Risk
Writing in the late 1970s, Stephen Kobrin bemoaned that definitions of the
term political risk were vague and over broad, with agreement between
scholars limited to an implication of unwanted consequences of political
activity (1979, p. 67). A consensus definition has continued to elude the

scholarly community because political risk is not, in fact, a single risk but
rather a diverse collection of related risks. In this project, we develop a
typology of political risk that reflects this heterogeneity, and then bring
these risks together under a single unified framework.
Establishing and enforcing a stable property rights regime can be difficult
within a sovereign state many governments fail to secure the property
rights of their own citizens but securing the property rights of foreign
investors is harder still. Within states, the central government functions as
the lawgiver, establishing the bounds of the rights that property holders may
expect to enjoy. In the international sphere, no such lawgiver exists and the
bounds of the property rights to which foreign investors are entitled remain
actively contested (e.g., Hadfield & Weingast, 2012, 2013). Thus, we do not
conceptualize of international property rights as a fixed set of universal
rights of property holders; instead we think of each property holder as
entering into an implicit contract with the government of the country in
which property is held.5 The nature of that contract, and hence the nature
of investors property rights, vary across countries, sometimes even across
investors within a given country, and often across time, especially as
reforms and emergent institutions unfold (Emmons, 2000).
The same institutions intended to secure the property rights of foreign
investors (i.e., reduce political risk) also attempt to shape the bounds of
those rights. The text of bilateral investment treaties (BITs), the terms of
political risk insurance contracts issued by state-backed political risk
insurers, and the opinions written by judges at the Hague and the district
court of New York can all be understood as various actors staking out
their position regarding the set of property rights to which foreign investors
should be entitled. This contestation over the terms of investors de jure
rights both affects and is affected by a parallel contestation regarding the
types of violations that investors, home-country governments, and host-
country publics will attempt to sanction via collective punishment, that is,
the terms of investors de facto rights (Alston et al., 2008). Thus, our model
of the conditions under which governments violate or respect international
property rights is informed by the ongoing process through which those
emergent rights are defined (Henisz & Zelner, 2005).
We conceive of political risk as the risk that the host government
violates the terms of its implicit contract with a foreign investor. In stylized
form, this contract commits the host government to refrain from the direct
seizure of assets; to honor explicit contracts; to provide equal protection
under the law; to allow the transfer of capital out of the host country at
market exchange rates; and to protect assets and personnel from violence.6
In return, the foreign investor commits to abide by domestic law. Thus, our
typology of political risk incorporates violations of any of these commit-
ments by the host government.
Because the terms of the contract between host government and investor
continue to evolve, no typology of political risk is set in stone for all time.
Similarly, any given set of host-government commitments can be usefully
grouped in several different ways. In this project, we adopt a market-based
typology in line with current practice in the broader political risk community,
which includes investors, insurers, ratings agencies, governments, and the
international legal community.7 Following these actors, we distinguish
between three types of political risk: war risk, expropriation risk, and
transfer risk (also known as inconvertibility).
The archetypical manifestation of expropriation risk is the nationalization
of property. Nationalization is easily observed and violates well-established
investor rights. For reasons we will discuss in more detail later, outright
expropriation has become increasingly costly for host governments and has
been declining in frequency since the early 1980s (Henisz & Zelner, 2010;
Minor, 1994).
As outright expropriation has declined in prevalence, it has been
supplanted by more subtle acts of creeping expropriation, which include
selective taxation, selective regulation, government breach of contract,
and transfer restriction (e.g., Kobrin, 1984; Wellhausen, 2013; Weston,
1975). Among these we focus on the risk of transfer restriction because
these restrictions are common, costly, and currently front and center in
debates over the boundaries of investor property rights (Graham et al.,
2015). The governments of major investment-sending countries like the
United States have been actively working to forge a consensus recognizing
transfer restrictions as a violation of investor property rights, while the
governments of investment-receiving countries have been pushing back
(e.g., Rose-Ackerman & Tobin, 2009). This contestation was particularly
acute surrounding failed attempts by the OECD in the late 1990s to create
a Multilateral Agreement on Investment that would have endowed foreign
investors with much stronger property rights than those available under
the 1994 Agreement on Trade Related Investment Measures (Deere, 2008).
Investment-sending countries have been more successful in enshrining
expansive property rights in BITs than in multilateral agreements; however,
in some ways this only underscores the lack of an international consensus
on investor rights vis-a-vis transfer restrictions and other types of creeping
expropriation.
Both expropriation and transfer restrictions are means through which

host governments extract wealth from foreign investors, causing costly
losses to foreign investors and potentially to the government itself. In
contrast, war is typically costly to both investors and host governments.
War frequently violates the host governments commitment to protect
assets and personnel (e.g., Ghobarah, Huth, & Russett, 2003; Oetzel &
Getz, 2012) and remains a salient political risk. From 1972 to 2002, losses
from war generated only a small number of claims with Berne Union
members (12% of the total), but the claims were large, accounting for half
of all claimed losses by value.
Modeling Each Risk
A formal model forces precision regarding the characteristics ascribed to

each type of political risk. To generate our modeling assumptions, we draw
on the existing scholarly literature as well as descriptive statistics from the
universe of political risk insurance claims that have been filed with MIGA,
OPIC, and the Berne Union.
War is not the most preferred outcome of any of the parties involved
and both its onset and its termination are difficult to predict (e.g., Fearon,
1995; Hegre, 2004). Thus, we model the (non)occurrence of war as the
result of a move by Nature, rather than as the result of a deliberate action
taken by the host government. This is consistent with the treatment of war
in commercial contracts as an event equivalent to acts of God. If war
occurs, it imposes costs on the foreign investor both by destroying assets
and by affecting the rate of return on assets that remain intact.
Expropriation is a deliberate strategy for seizing wealth from foreign
investors and typically targets only one or a few investors at a time. When
the assets to be seized are liquid, host governments typically take great care
to keep expropriations secret before they are enacted in order to prevent
investors from moving assets outside the host country before they can
be seized.8 Thus, we model the average investor as unable to predict
expropriation events before they occur.9
Transfer restrictions are more limited than expropriation in that they
only target a subset of investor assets those assets that the investor
wishes to move out of the host country. However, transfer restrictions are
typically economy-wide, hitting all foreign investors in the host country
simultaneously. Implementing such economy-wide restrictions is complex
and requires coordination between multiple branches of government. For

example, effective foreign exchange restrictions require mobilizing both the
central bank and private banks (to restrict the electronic transfer of funds);
customs enforcement (to prevent the physical movement of currency across
the border); and domestic law enforcement (to shut down black-market
currency exchanges). While governments aspire to implement such restrictions
in a manner unforeseen by foreign investors, doing so is inherently difficult.
Thus, investors typically enjoy a small window during which they know
that restrictions are imminent, but in which repatriation of assets is still
possible. This is reflected in our model by the assumption that, prior to
transfer restrictions taking effect, the average investor has the opportunity
to expedite repatriation of their assets that is, to slip some assets out of
the country before the restrictions take effect.
The characterizations of war, expropriation, and transfer risk introduced
in this section inform the theoretical model that follows.
THEORETICAL MODEL: AN EXTENSIVE-FORM GAME

WITH WAR, TRANSFER RESTRICTIONS, AND
EXPROPRIATION
To begin a discussion of the environment of political risk, we find conditions
under which it is optimal for a foreign investor to invest in a foreign country,
despite the risks of: violence or war breaking out; increased transfer costs for
the repatriation of capital; and outright expropriation of assets.
We characterize an equilibrium where this behavior occurs, and then
analyze how the equilibrium behavior changes with respect to different
types of investors.
We model the relationship between a host government and a foreign
investor as a five-move game. Define this investor as the average investor
over a range of investor types, investment sizes, and sectors.
Structure of the Game
In each round of play, a foreign investor (F) faces three distinct risks. First,
there is a risk of violence breaking out in the host country. If violence does
not break out, there remains the risk that the government (G) will choose
to extract rents from the foreign investor by increasing the rents gained
from F repatriating assets, or by expropriating assets.10 At the beginning of
the game, the foreign investor can either invest (I) or not invest (I). If F
invests, Nature (N) moves and determines two things: first, if the host
country breaks out into war with probability r, and second; the cost (CT)
associated with transfer rents.
If N selects war, the game ends. If not, the game proceeds. G directly
observes the value of CT that has been assigned, and based on that value,
the government can either uphold the investment contract by maintaining
the agreed-upon transfer rents, t0, or breach the contract by selecting some
t0 = t0 + , where > 0.11 The foreign investor is imperfectly informed
about the level of CT. She perceives, with probability p, that the value of
CT has been set such that the host government will breach its contract by
selecting t0 .12 Based on this perception, F selects what level () to expedite
repatriation before the new policy is announced. G then decides whether or
not to expropriate assets.13 Fig. 2 displays this five-move game.
Investor Incentives
As shown in Fig. 2, if the foreign investor plays I, both players receive

zero. Suppose F chooses to invest. Denote V as the amount they will invest
and as the expected rate of return if F plays I.14 If N selects war, the game
ends, with F losing a q (0, 1) amount of her original investment and
receiving an adjusted rate of return, c, on the remaining (1 q)
amount: (1 q)V( c) qV. If conflict does not break out in the host
country, the game proceeds. Denote V as the portion F intends to
repatriate ( [0, 1]), amid transfer restrictions t0, and as the amount of
repatriation F expedites, upon anticipating a transfer breach. Define
as the cost of expedited repatriation ( 0).15 If the investor plays I, she
receives a maximum of V(1 ) + V(1 t0). This occurs if G upholds
the investment contract and F plays = 0; the payoff is a weighted sum of
what she earns on her non-repatriated assets (1 ) and her repatriated
portion (, subject to t0). The investor receives a minimum of V, when
G plays E. Thus, while the investor prefers to invest with minimal transfer
restrictions and without the threat of expropriation, she may or may not
prefer intermediate transfer restrictions (or a chance of expropriation) to
the reservation payoff of zero, depending on the probability of transfer
breach (p), the probability of war (r), and how lucrative the investment
opportunity is ().
Fig. 2. A Two-Player Extensive-Form Game. Notes: Here a foreign investor (F)

chooses whether or not to invest; Nature (N) chooses war or no war (with
Probability r); if a war does not occur, a host government (G) chooses at what level
(t) to set transfer restrictions on that investment; F chooses, before the new policy is
announced (but seeing a p-probability of transfer breach), at what level to expedite
repatriation (); and G decides whether to expropriate or not. Here, if Nature
chooses War, the round of play ends. F loses part of her investment and any
residual may receive a lower rate of return.
Government Incentives
The host government (G) works in this game to balance trade-offs between
its desire for revenue in the current period with competing demands for
domestic political support, future revenue, and good diplomatic relations
with the governments of investment-sending countries. These complex
preferences are fed into the model in a relatively simple way, as factors
determining , which is the value to G of Fs assets if they are owned by
the government (i.e., expropriated outright), and CE and CT, which are the
costs of backlash G receives after expropriation or a unilateral increase in

transfer restrictions, respectively.
CE and CT capture a range of negative impacts accruing to the host
government as a result of its expropriation/transfer restriction actions.
Most important among these are: (1) Monetary settlements levied if the
host government is found in violation of its treaty commitments; (2) Lost
future investment; (3) Diplomatic costs imposed by the governments of
investment home countries; and (4) Domestic political costs imposed by the
domestic public (or a subset thereof).
The severity of settlement costs varies depending on the precise terms
of the treaties to which the host government is party to, that is, it depends
on the strength of investors de jure property rights. The severity of lost
future investment, diplomatic costs, and domestic political costs are deter-
mined by the respective abilities of investors, home-country governments,
and the domestic public to engage in collective punishment, that is, the
strength of investors de facto property rights. Collective punishment is
most effective when the community of potential punishers can coordinate
on what does and does not constitute a violation.16 Enshrining a particular
right in a treaty or in domestic law enhances collective punishment by
easing this coordination, but collective enforcement is possible even for
rights that do not exist in de jure form.
In related work, we develop theory regarding why the blowback costs
for expropriation (CE) increase as domestic political constraints increase,
while the blowback costs of transfer restriction (CT) are largely unaffected
by those same constraints. This distinction emerges primarily because
investors de facto rights are stronger with regard to expropriation, leading
to more effective censure of the government by the domestic public
(Graham et al., 2015). Both outright and (non-transfer) creeping expropria-
tion violate well-established and broadly accepted rights of investors, and
both are costly to domestic interests. In contrast, transfer restrictions
impose losses almost exclusively on foreigners,17 and the right to be free
from such restrictions is not universally acknowledged as a right that
foreign investors possess. Thus, coordination among potential punishers,
particularly members of the domestic public, is hindered by a lack of consensus
regarding what constitutes a violation.
There also exists a political component to , the governments valuation
of assets when owned outright.18 We refer to this political component of
the valuation as the political valence of the assets. For example, state
ownership of nationalized firms can provide opportunities for patronage
via employment and contracts with the now-state-owned enterprise.
Similarly, populist governments may enjoy increased support if they can
successfully claim acts of expropriation as acts of nationalism or indepen-

dence. Conversely, some fixed assets may be entirely unappealing to the
government, leading to a low value of . Note that in the case of revenue
seized via transfer restriction, G and Fs valuation of the revenue is identical
and there is no political component to valuation this revenue is fungible.
Returning to Fig. 2, we see that, like the investor, the host government
receives zero if F does not invest. Denote R as the governments share in the
investments value. R is a sum of the tax revenue and other benefits that
accrue to the government from the investors operation, and is the rate of
return on that investment.19 If war breaks out, the game ends, and G receives
R on the (1 q) undestroyed portion of the investment, minus a war cost,
Cw: R(1 q) Cw.20 If war does not break out, the game continues. If F
invests and G upholds the original investment contract (t0), G receives Vt0
on the portion that F repatriates and R(1 ) on the portion that F does
not: R(1 ) + Vt0.
If G breaks the contract, selecting t0 , it receives R(1 ) + Vt0
(1 ) CT with the new transfer restrictions and Vt0 on the amount that
F expedites before the policy shift: R(1 ) + V(t0 (1 ) + t0) CT.
Finally, if the host government expropriates, it receives CE, and, if it
expropriates after increasing transfer restrictions, it receives CE CT.
Notice that, without the prospect of backlash to a contract violation, the
government always prefers to either seize the maximum amount of transfer
rents or to directly expropriate, whichever offers the greater return. This
creates tension in the game between playing t0 or E, on the one hand, and
avoiding the backlash, on the other hand.
To analyze the strategy of investment amid transfer risk, we define a
subgame perfect Nash equilibrium (SPE) in which: the expropriated assets
are worth enough for G to expropriate if they do not breach the transfer
policy (condition 1); and too little if they do breach it (condition 2); the
cost of expediting repatriation () is sufficiently high so that F plays = 0
(condition 3); the cost of increasing transfer restrictions (CT) is sufficiently
low for G to consider playing t0 (condition 4); and F prefers to invest,
despite the r-probability of war and the p-probability of transfer breach
(condition 5). Put differently, we define an SPE where F strategy is to invest,
despite the risk of war, and the possibility of expropriation and transfer risk,
where war winds up not occurring, and where Gs strategy is to play {t0 , E}
when CT is sufficiently low, and {t0, E} otherwise. Formally:
Definition 1. A political risk equilibrium is an equilibrium in which F plays
{I, = 0}, and G plays {t0 , E} when CT R(1 ) + V(t0 + ) +
CE , and {t0, E} otherwise.
No
investment
War
Expropri No Expropri No Transfer Transfer Transfer Transfer

ation breaches ation breaches breach breach breach breach
and and
expropriation expropriation
Fig. 3. An Illustration of an Equilibrium Path in Proposition 1. Notes: Here, war

does not occur and the government (G) chooses to increase transfer restrictions but
not to expropriate.
Proposition 1. There is a political risk equilibrium when the following

conditions hold:
1. R1 Vt0 CE
2. R1 Vt0 CE
3. V
4. CT V
1 p p1 t0
5. r 1 p p1 t0 1 q c q :
See Appendix for proof. Fig. 3 displays an equilibrium path of

Proposition 1. Notice how each of the risks is present for the investor:
while war does not arise, G selects between {t0, E} (expropriation
outcome) and {t0 , E} (transfer breach outcome), ultimately choosing
the latter.
At this point, we have demonstrated a logic for how expropriation, war,
and transfer risks can accompany investment.
DYNAMIC INTUITIONS FROM THE STATIC MODEL

Using this unified framework, we can ask questions that would otherwise
be difficult to assess comprehensively. In particular, how do investors view
a countrys risk profile when there are multiple, distinct risks, and other
types of investors entering? Our logic suggests that, by influencing the
incentives of the government, the distribution of investor types currently in
place can create a risk shield for some prospective investor types, while
increasing the risks faced by other types.
The distribution of capital flows in an economy affects the host govern-
ments behavior in important ways. For example, countries that rely
primarily on bank debt (versus portfolio flows or direct investment) for
foreign investment may face a different set of incentives when it comes
to expropriation or imposing new transfer restrictions.21 But how? For
clarity, we focus on one simple distinction the level of an investors
repatriation and use it to demonstrate how a governments investment
portfolio can influence its risk portfolio (i.e., its willingness to expropriate
and impose transfer restrictions).
Let us first assume that investors may vary in their level of repatriation.
Portfolio debt and equity investors, for example, repatriate a large share of
their profits. While bank lenders also repatriate a material share of profits,
some funds may also be used to expand the banks lending portfolio, build
domestic branches, etc. Across types, however, direct investors, as a class,
repatriate the least from and reinvest the most in the domestic economy;
profits from the initial investment are often plowed back in to expand
and grow the affiliate over time, leading to long delays between the
initial investment and the eventual repatriation of profits. Thus, let us
assume that the amount repatriated () varies across investment types
(i.e., portfolio (p), bank (b), and direct (d) investors) in the following way:
p > b > d.
But how does this matter? Suppose that bank debt is currently the only
type of foreign investment in a host country. How might this affect a direct
investors decision to invest? Let us focus on investors intended levels of
repatriation; b and d, where b > d. Notice that, from Proposition 1 above,
is one of only two parameters that is common to all four equilibrium
conditions (the other being ), and thus changes in may have non-obvious
consequences for the behavior in equilibrium. We begin by looking at
Gs decision to expropriate or simply increase transfer restrictions.
Define b = d + b, where 0 < b 1 d.22 How does increasing
affect Gs decisions to expropriate or commit a transfer breach?
In Proposition 1, when N chooses CT R(1 ) + V(t0 + ) + CE

(call this the CT-condition), G plays {t0 , E} over {t0, E}. Solving for ,
the CT-condition becomes: R(1 ) + V(t0 + ) + CE CT. We
see that increasing (d b) will make this condition easier to satisfy when
(1 b) + bV(t0 + ) + CE CT > (1 d) + dV(t0 + ) + CE CT.
Substituting b = d + b, and solving for , this reduces to:
Vt0
< 1
R
Notice that conditions (1) and (2), which determine Gs expropriation deci-
sion (both on- and off-path), yield the same condition. Likewise, if the
-condition is satisfied, condition (4), which determines the feasibility of a
transfer risk breach, is easier to satisfy as increases. Less directly, we
can see that increasing makes expediting repatriation more attractive
(condition 3) and G more likely to invest (condition 5) (if q c c 1).
Overall, if the -condition is satisfied, we see that increasing makes the
expropriation conditions more difficult to satisfy, t0 more feasible, the expa-
triation condition easier to satisfy, and investment more likely to occur
(when q c c 1).
The CT -condition is particularly relevant because it determines whether
the game ends in transfer breach or expropriation.23 Notice the conditionality
in the -condition: increasing will make G more likely to behave in such a
way that leads to the transfer restrictions outcome ({t0 , E}) (over expropria-
0
tion) in certain circumstances (i.e., when < Vt R ), but increasing will have
0
the opposite effect under other circumstances (i.e., when > Vt R ). Intuitively,
this means that a predominance of foreign bank lending in the host economy
affects the governments relative preference for seizing assets via {t0 , E}
versus {t0, E}. And investors are not indifferent to this: some investors stand
to lose more to transfer risk than others. Specifically, when transfer restric-
tions increase, those which repatriate a larger share of their profits, stand to
lose more. If, for example, a direct investor were to reinvest all of her capital,
she would be unaffected by an increase in transfer risk. On the other hand,
that same direct investor remains vulnerable to expropriation risk. Thus,
our theory suggests that a predominance of bank debt in a host country
changes the relative level of transfer and expropriation risk and,0 through that
change, may further close the door to direct investors (if < Vt
R ) or open the
Vt0
door (if > R ). Thus, by affecting the host countrys political risk profile,
initial investment conditions influence future investment trends. It also
suggests that investors may benefit from being in the minority of investors; of
being outside of the predominant group upon which the host government
focuses its political risk policies.
But what does this mean for the prospective direct investor looking in?
There are two cases. In one case a high initial endowment of bank debt
raises the risk of expropriation relative to transfer breach and thereby
deters new entry by direct investors into the host economy. This occurs
0
when ( < Vt R ). In the second case, a high initial endowment of bank
debt has the opposite effect, dropping the0 risk of expropriation relative to
transfer breach. This occurs when ( > Vt R ). These are distinct and opposite
effects, meaning that the government can either earn enough on the
investment under foreign ownership () to simply skim off the top (with
transfer restrictions), keeping the production/ownership in the investors
hands, or just expropriate the investment if is sufficiently low in comparison
to its value under national ownership.
Notice that the opposite is implied for portfolio investors. Since portfolio
investors repatriate a higher share of their profits, they are more (less)
sensitive to transfer breach (expropriation). Thus, the same shift in risks that
deters FDI opens the door for portfolio investors, and the same conditions
that create a risk shield for direct investors may keep out portfolio investors.
The government adapts its strategy to seize assets efficiently from the type
of investor that is predominant in the economy. This shifts the political
risk profile in the host country and, in turn, affects the relative willingness of
different classes of foreign investor to enter. Table 1 summarizes this logic.
Notice that we have only looked at the case of bank predominance. If
we have three groups of investors, with respect to (portfolio, bank, and
direct), there are six other potential initial conditions (of predominance):
{portfolio}, {direct}, {portfolio & bank}, {portfolio & direct}, {bank &
direct}, and {portfolio & bank & portfolio}. We can perform a separate
analysis for each of the initial condition possibilities. This can, for example,
help us tailor the theory to specific country profiles (which may show
predominance with some investors over others), but can also help us
Table 1. Effect of Bank Predominance on Political Risk and

Future Investment.
-condition Effect on political risk Consequences for other investors
Vt0
> R
Transfer restrictions
Expropriation Prospective FDI
Prospective portfolio investment
Vt0
< R
Transfer restrictions
Expropriation Prospective FDI
Prospective portfolio investment
understand a countrys investment-related path dependence. The simple

takeaway is that, whether for investors inspecting potential investment
locations abroad, or host governments interested in attracting a particular
type of investment, our theory suggests that initial conditions matter; that
a countrys existing investment profile may already be leading to certain
future investment profiles, while creating barriers for other potentialities.
While simple, this takeaway is non-obvious and emerges only when
multiple investor types and multiple risks are considered jointly within a
unified model. In the following section, we move beyond variation in how
much investors repatriate and examine variation in investor capabilities.
CAPABILITIES OF INVESTORS
We model the behavior of the average foreign investor across a range of
investor types. However, investors vary in their capabilities for managing
political risk (e.g., Delios & Henisz, 2003; Kerner & Lawrence, 2014;
Khanna, Palepu, & Bullock, 2010; Wellhausen, 2013). In particular, we
focus on variation across investors with regard to: level of information;
ease of exit; and ability to resist. Each of these investor characteristics
corresponds to a parameter value in our model. Thus, we are able to exploit
the comparative statics in our model to assess the exposure of different types
of investors to different political risks. First we must establish theoretically
which capabilities are likely to be possessed by each type of investor.
An investors level of information, ease of exit, and ability to resist
jointly determine her vulnerability to adverse political events. To avoid
or limit losses, an investor must either: (a) prevent an adverse political
event from occurring; (b) shift assets out of the host country before losses
are incurred; or (c) obtain compensation for those losses after they
occur (Johnston, 2015). In the following sections, we outline how level of
information, ease of exit, and ability to resist allow investors to achieve
these favorable outcomes.
Ability to Resist
We model war, expropriation, and transfer restriction as ex post

irreversible investors cannot undue these events once they have occurred.
This does not mean, however, that investors are passive or powerless in the
face of adverse government action. Instead, investors engage in both ex

ante and ex post political risk management strategies to reduce the govern-
ments incentives to engage in expropriation or transfer restriction. In the
context of our model, an investors ability to resist is directly reflected in
the governments costs of expropriation (CE) and transfer restriction (CT).
The higher the investors ability to resist, the higher the costs of adverse
action against her. As CE and CT increase, it narrows the range of conditions
under which the government prefers to expropriate and impose transfer
restrictions, respectively.24
Investors ability to resist adverse government action is a function both of
their de jure power to seek legal redress in response to a host governments
violation of its treaty commitments and their de facto power to marshall
other actors to collectively punish the violating state. Investors increase
both of these types of power through ex ante and ex post political risk
management strategies.
An investors de jure power is determined by the precise terms of the
treaties to which the host government and the investors home government
are party. Thus, investors can and do increase their de jure power by lobbying
for specific treaties to be signed (an ex-ante strategy), but when they do so,
they empower not only themselves but also all other investors from their
home country. Ex post strategy for exercising de jure power primarily involves
litigation through international arbitration bodies.
The sources of an investors de facto powers of resistance are perhaps
more theoretically interesting. They include: (1) the relationship between
the investor and domestic stakeholders, especially joint-venture partners;
(2) the regime type in the host country; (3) the ability of the investor to act
collectively with other foreign investors to withhold future capital from the
host country; and (4) the relationship between the investors home govern-
ment and the host government, including military alliances, foreign aid,
and the presence of BITs or preferential trade agreements (PTAs).
The role of joint-venture partners in reducing political risk vulnerability is
well documented in the strategy literature (e.g., Delios & Henisz, 2003;
Henisz, 2000). Domestic partners are, themselves, stakeholders in the domestic
government, and harming them imposes direct political costs on the host
government. Additionally, to the extent that foreign investors hire local
employees, purchase inputs from local suppliers, supply difficult-to-replace
inputs to local customers, and contribute to the local community via
corporate social responsibility (CSR) efforts, they align their interests with
the interests of the domestic public more broadly effectively engaging the
domestic public as a shield (Baron, 2001; Henisz, Dorobantu, & Nartey, 2013).
If a foreign investor can increase the number of citizens that would be

harmed by government action against her, she has engaged in effective
ex ante positioning that increases the blowback the government will face for
that adverse action (e.g., Markus, 2012).
The ability of the domestic public to impose costs on the host
government is conditioned strongly by regime type in the host country.
Democratic regimes require the support of a majority of their citizens to
remain in office; non-democratic regimes may be able to stay in power with
the support of a much smaller winning coalition (e.g., Bueno de Mesquita,
Smith, Siverson, & Morrow, 2003). Thus, democratic regimes are expected
to be much more responsive to the concerns of the domestic public. We
expect that resistance strategies in which the domestic public plays a pivotal
role are most effective in countries where democratic accountability is
high.25 More targeted local-stakeholder strategies can also be effective in
an autocratic context, so long as the local stakeholder in question is part of
the political elite.26 Thus, regime type conditions the effectiveness of all ex
post political risk management strategies, and the decision to invest only
under certain types of regime could be considered an ex ante strategy.
Much of the blowback the host government faces following expropriation
and transfer restriction comes in the form of lost future investment. The
value of this lost future investment is higher when the harmed investors can
coordinate widely with other foreign investors. Broad coordination is best
achieved by investors in concentrated industries (e.g., Olson, 1965) and
investors whose co-nationals collectively account for a large share of inward
investment to the host country (Wellhausen, 2014). Thus, the effectiveness of
ex post collective punishment can be well predicted by the context in which
a firm chooses to invest (an ex ante decision).
Home governments may also intervene directly on behalf of their inves-
tors to impose costs on host governments, and mistreatment of investors can
have political repercussions across all facets of the bilateral relationship
between home and host government.27 Firms that can engage in effective ex
post lobbying efforts after they have been harmed can impose high costs on
the host government via their home government. In contrast to investors
utilizing BITs or other treaties to impose costs on the host government, we
would consider this type of lobbying and direct government to government
punishment an exercise of investors de facto power.
In addition to driving up the costs to the government (CE and CT), effective
resistance by investors sometimes results in the recovery of some of the
expropriated assets. This is particularly true of resistance based on de
jure power. In the interest of parsimony, we do not model this recovery
explicitly in the payoffs to the investor, but the addition of a recovery term
into the model would not alter the comparative statics of interest.28 We
consider this extension of the model a potentially fruitful avenue for
future research.
We now explore how the ability to resist affects the political risk equili-
brium. The blowback cost of transfer breach, CT, affects when the govern-
ment will choose an increase in transfer restrictions; it also affects when it will
choose expropriation. Rewriting the CT-condition in definition 1 for , recall
that G will choose transfer restrictions when R(1 ) + V(t0 + ) +
CE CT. It is straightforward to see that, as CT increases, the right side of
the inequality decreases, and the CT-condition becomes more difficult to
satisfy, making expropriation preferable over a larger range of investments.
Will the same be true for CE? Will it only affect the expropriation decision?
Notice from the model that each decision node accounts for both direct
and indirect factors. While transfer risk often leaves an investment intact,
outright expropriation does not; in our model, it is a final stop. Thus, while
CT only affects the transfer risk decision, CE affects both the decision to
increase transfer restrictions (through the CT-condition) and, ultimately, to
expropriate or not (conditions 1 and 2). For condition (1), R(1 ) +
Vt0 + CE, it is straightforward that as CE increases, the right side of the
inequality increases, and condition (1) is more difficult to satisfy. Analogous to
CTs impact on transfer risk, increasing CE makes expropriation less preferable
over a larger range of investments (the same obtains for condition 2).29
The CT-condition, R(1 ) + V(t0 + )+ CE CT, is
more nuanced. Here, G decides between expropriation and increasing
transfer restrictions (without expropriation). While increasing CE makes
the CT-condition easier to satisfy as expected, increasing the costs of
expropriation makes transfer restrictions more attractive, and increasing
CT makes it more difficult to satisfy again as expected, increasing the
costs of transfer breach makes transfer restrictions less attractive, notice
that CE and CT are oppositely signed. Thus, while {CE > 0, CT = 0} and
{CE = 0, CT > 0} produce intuitive comparative statics results, what if a
government will face costs in either scenario; what if {CE > 0, CT > 0}?
Consider that, while it is possible that only one cost is nonzero, it is more
often the case that costs, whether retaliatory or reputational, are triggered
by either expropriation or transfer breach.
So what happens, in this more realistic case? Condition (1) is unaffected
by this possibility: it only is affected by CE. The CT-condition however is
affected. In this SPE, increasing CT reduces the subset of investments over
which transfer restrictions are selected; however, if CE also increases, it
undermines this risk mitigation for transfer risk. If CE rises as much as CT,
then there may be no effect or even an increased size of the subset over
which G will select transfer restrictions! If scholars only look at transfer

risk, and the resistance costs associated, they might easily miss this nuance:
that simply increasing the blowback costs associated with transfer breach
does not mean that transfer breach is necessarily less likely! Indeed, once
we consider other risks, and other costs, our logic demonstrates that the
prevalence of transfer restrictions may even increase. This is particularly
likely to be the case if the same types of investors that can mount effective
resistance to transfer restrictions can also mount (even more) effective resis-
tance to expropriation. In the section Transfer Risk versus Expropriation
we argue that this is, in fact, a quite common set of conditions.
While the effect of CT on the willingness of G to engage in transfer
breach is conditional, the effect of CE on Gs willingness to expropriate is
straightforward. Looking at both conditions (1), (2) and the CT -condition,
we see that, as both CE and CT increase, G will be less likely to expropriate;
whether they increase transfer restrictions or not!30 Thus, it may be that
high CE,T (e.g., the ability to resist) is more valuable for investors that are
more concerned about expropriation (e.g., direct investors).
This section continues the process begun in the section Dynamic
Intuitions from the Static Model. Parameters in the model correspond to
the capabilities of investors; comparative statics from the model thus allow
us to explore that full strategic implications of variations in investor
capabilities. Direct investors may be dramatically, albeit indirectly, affected
by other foreign investors, through the political risk profiles that govern-
ments select, in response to those other investors; being a minority investor
may either entail a risk shield from the more predominant investor, or a
greater barrier to entry; and increasing the blowback costs to all political
risks, across the board, may not reduce all political risk, and may even
increase transfer risk (both relatively and absolutely). While it may be
tempting to take disparate intuitions like FDI should be concerned with
how a host treats other FDI and increasing costs of contract breach with
help prevent that breach, and apply them more generally to our under-
standing of risk phenomena, this section argues that, to do so, would be
too simplistic. Reality is more nuanced, and this section is a first attempt at
making that nuance tractable.
Level of Information
Access to information local and global, public and private is a critical

component of investors competitive advantage (e.g., Hirshleifer, 1971;
Rumelt, 1984; Vives, 1990). To identify opportunities and assess risks,

investors need information about: trends in the local and global economy;
pending changes in the local and global policy environment; the structure
of the local and global market; the identity and reliability of potential local
and global business counterparts; the preferences of relevant local and
global elites; and a range of other issues. Public information is available to
all investors in the market and can be obtained at relatively low cost, though
only some may have the capacity to gather, translate, and analyze it. Private
information is costlier to obtain and available to only some actors. Local
information about a host country tends to favor local investors, whereas
global information favors foreign (multinational) investors (Albuquerque,
Bauer, & Schneider, 2009).
In rich democracies, specifically, local public information about govern-
ment policy and economic fundamentals is abundant and easily accessible
from outside the country. However, in emerging markets government
transparency is often low, with limited and unreliable economic data
and press coverage, making local public information scarce (e.g., Adsera,
Boix, & Payne, 2003; Hollyer, Rosendorff, & Vreeland, 2014). The only
firms that can accurately identify opportunities and assess political risks are
those with access to local private information. This makes local private
information particularly valuable in emerging markets (e.g., Leuz &
Oberholzer-Gee, 2006).
In our model, the archetypical investor is assumed to have enough local
private information to foresee transfer events, but not to foresee expropria-
tion events.31 If an investor lacks sufficient information to foresee changes
to transfer risk, expedited repatriation (e.g., exit, to be discussed in follow-
ing section) becomes impossible and the value of , the value of assets repa-
triated early, is forced to zero; the investor loses her ability to exit before
the costly transfer breach. If a well-connected investor has sufficient private
information to foresee expropriation (we believe this is very rare), then that
investor would enjoy an option of costly repatriation similar to that we
model for transfer risk. In practice, different types of foreign investors vary
in their access to local private information, and therefore also vary in their
ability to foresee adverse political events.
Among the four investor classes we discuss, foreign commercial banks
possess the most local private information about the probability of changes
to foreign exchange restrictions. Because banks tend to have sizeable pre-
sences in the host country, both in terms of local branches and domestic
exposures in local and foreign currency, they build deep relationships with
political and economic elites who funnel non-public information and reveal
policy preferences (Cohen, 1996; Dinc, 2005; Faccio, 2006; Sengupta,

2007). From their information intensive lending, foreign banks also
gain detailed knowledge of and access to multiple parties across the local
economy (Beim & Calomiris, 2000).
Foreign direct investors are not as advantaged. Direct investors hold
corporate control positions that offer privileged principal information
about the firm or sector in which they are invested (Goldstein & Razin,
2006), which is often augmented by in-depth knowledge of their particular
investment and significant global private information about their product
or business line. However, the average direct investors local private
information is limited in scope to either their firm or the relevant sector in
which their firm operates. The willingness to develop broad political
connections may be similar to that of banks, particularly to the extent the
direct investor is resource-seeking, yet direct investors abilities are typically
limited due to the circumscribed nature of their engagement across the
host economy.
In contrast to banks and even direct investors, the average portfolio
debt and equity investors are large institutional investors with high levels of
global information but low levels of local information (Albuquerque et al.,
2009). Their advantage lies in information about market benchmarks,
relative value, future investment trends, and upcoming interest rate changes
in advanced countries; in short, information about the push factors that
explain much of variance in capital flows into emerging markets. But these
investors have neither the country-specific experience nor the ongoing
access to local elites necessary to obtain local private information. Indeed,
portfolio debt and equity typically function without any domestic presence,
short of an institutional prime broker that executes the trade. This lack of
local information explains much of the tendency of portfolio debt and
equity flows to exhibit herding behavior where investors cue off of a
first-mover with (possibly) superior knowledge (e.g., Avery & Zemsky,
1998; Banerjee, 1992; Lux, 1995; Shleifer & Summers, 1990) (Fig. 4).
Ease of Exit
Ease of exit refers to investors ability to cheaply and rapidly shift assets
out of the host country. Ease of exit is a necessary condition for investors
to be able to move assets before losses are incurred, but it is not sufficient.
Investors also need to know that an adverse event is pending hence the
importance of level of information, discussed above. In the model, ease of
Portfolio
Equity
Ease of Exit
Portfolio Bank Debt

Debt
Foreign Direct
Investment
Level of Information
Fig. 4. Summary of Investor Types. Note: A classification of foreign investors by

two capabilities: level of local private information (horizontal axis) and ease of exit
(vertical axis).
exit is reflected in the value of , which is the costs of early repatriation to

investors. High ease of exit equates to low costs of expedited repatriation
(i.e., low values of ).
Ease of exit varies across investors according to the liquidity of an inves-
tors stake, the investors time horizon, and the physical mobility of the
underlying assets. The liquidity of an investors stake determines her ability
to sell that stake before losses are realized; the investors time horizon
determines whether early liquidation of assets will induce losses and
whether losses can be avoided by waiting out the adverse event; and the
physical mobility of the underlying assets determines whether an investor
can move assets out of harms way.
The liquidity of investor stake varies substantially across classes of
investor. Direct investors are the least liquid because they typically own
large shares of the enterprise in question and are involved in management
of the firm in which they are invested (Goldstein & Razin, 2006; Vernon,
1971). Investments by foreign banks are markedly more liquid. Banks have
the capability to halt loans in progress and cancel future borrowings; in
certain circumstances they also have the right to accelerate outstanding loans
(Graham, Johnston, & Kingsley, 2014). Lastly, portfolio debt and portfolio
equity are the most liquid because they are traded in open markets.32
Portfolio equity markets are larger and deeper than public (or private) debt
markets, giving equity investors the highest levels of liquidity (Fig. 4).
The physical immobility of assets is central to the obsolescing bargain
in which foreign investors become vulnerable to host-government predation
once costs have been sunk into immobile assets such as mines or factories
(Kindleberger, 1969; Kobrin, 1984). However, as Kerner and Lawrence
(2014) demonstrate, even foreign direct investors (whose stakes are illiquid)
are capable of avoiding many political risks if the underlying assets they
own are not fixed. In other words, a direct investor whose assets are
primarily intangible (such as accounts receivable) or mobile (such as
product inventory) has high ease of exit. Empirically, this fixity can be
approximated by measuring plants, property, and equipment (PPE) as a
share of total affiliate assets, and this PPE share varies dramatically by
sector. For example, among the overseas affiliates of U.S. multinationals,
PPE accounts for half or more of total assets among mining and utilities
firms, while accounting for only 67% of total assets in the services and
wholesale sectors (Kerner & Lawrence, 2014, p. 115).
It is noteworthy that the ability of investors to shift assets out of the
host country before losses are incurred (a product of investors ease of
exit and access to information), is equivalent to the inability of the host
government to successfully seize assets from those investors. Thus, inves-
tors ability to avoid the consequences of an attempted seizure of wealth
deters the government from making that attempt.
Due to space constraints, we do not examine the comparative statics
with regard to information and liquidity (i.e., value of ) in the same degree
of detail that we devote above to variations in resistance capabilities. We
relegate this analysis to related work (Graham et al., 2015).
MAPPING CAPABILITIES TO RISKS

In the sections above, we outline how foreign investors vary in their
capabilities for managing political risk. However, some capabilities are
more relevant for certain types of political risk than for others. By mapping
different capabilities onto the risks they are most effective at managing, we
can develop detailed theory regarding how an investors capabilities shape
the universe of host countries in which they can operate effectively and,
conversely, how a countrys political risk profile affects the type, as well as
the volume of inward investment it attracts. It is important to note at the

outset of this section that our mapping of firm capabilities to specific risks
does not exhaust all political risk management strategies available to firms.
For example, one of the implications of our model is that investors can
limit their vulnerability to expropriation risk by assessing the political
valence of the assets in which they invest and selecting assets that have
relatively little value to the government (i.e., assets for which is small).
Thus, not all political risk management strategies depend on the three
capabilities we explore, but we expect these three capabilities explain a
substantial portion of the variation in firms political risk vulnerability.
War versus Expropriation and Transfer Restriction
War risk is unique in that it is not a deliberate strategy by the host

government to extract wealth from foreign investors; in terms of the implicit
contract, it is simply a failure to uphold the governments commitment to
protect the investors property and personnel from violence. Thus, war cannot
be effectively mitigated by strategies based on resistance war is already not
the governments most preferred outcome and it is unlikely even very capable
investors can prevent a war from occurring.
We acknowledge that in certain narrow contexts, mitigation of war risk
by select investors is possible. For example, foreign investment in oil
extraction in the Niger delta continued (sometimes profitably) in the face of
high levels of violence throughout the 1990s, with firms investing heavily in
a range of active risk-mitigation strategies, including private security forces.
However, these high-cost efforts are generally limited to investors engaged
in high-margin natural resource extraction (Graham, 2014).
Information-based strategies are similarly limited in the context of war.
Investors who carefully analyze local public information may have a more
precise estimate of the underlying probability of war than those who do
not, but the relevant local private information is unlikely to be available to
even the most connected investor. Military strategy for example, when
and where fighting will occur is a carefully guarded secret generally
unavailable even to most political elites. This levels the playing field
between investors with and without access to local private information.
While exit is the most productive, and usually the only, option available
to foreign investors to mitigate war risk, ease of exit itself is endogenous to
Bank Debt
H FDI
Portfolio Equity
Portfolio Debt
Ease of Exit
War Expropriation Transfer Risk
Gs Preparation for Breach
Fig. 5. Summary of Investor Ease of Exit, by Type of Risk. Notes: Here, the ease
to exit is ordered as low (L), medium (M), and high (H), and risk is ordered by how
intensively a government prepares prior to the breach.
the presence of war. War can cause markets to freeze and borders to close.
Previously liquid stakes become illiquid; previously mobile underlying
assets become immobile.
Thus, war is cataclysmic. It is the archetypical exogenous hazard. This is
reflected in the force majeure clauses of most commercial contracts which
explicitly carve out war as an extraordinary and unforeseeable event
beyond the control of either party, akin to acts of Godlike fire, floods,
earthquakes, and hurricanes.
Of course, to argue that investor capabilities are largely irrelevant is not to
argue that exposure to war risk does not vary across firms. For example,
investors with larger and more dispersed physical footprints in the host coun-
try are more vulnerable they have more physical property in more places
that can be destroyed, more personnel that can be injured or killed. Similarly,
investors that are more infrastructure-dependent are more likely to be harmed
when that infrastructure is destroyed (Collier, 1999). But these sources of
variation are not a reflection of capabilities, per se (Figs. 5, 6, and 7).
Bank Debt
FDI
H Portfolio Equity
Portfolio Debt
Access to Private
Information

Fig. 6. Summary of Investor Access to Private Information about the Investment

Climate, by Type of Risk. Notes: Here, the ease to exit is ordered as low (L),
medium (M), and high (H), and risk is ordered by how intensively a government
prepares prior to the breach.
Transfer Risk versus Expropriation
Strategies of resistance are more effective when the actions of the government
are universally accepted as violations of investor property rights. In this
context, it is easy to coordinate the domestic public and foreign investors
(including those not directly harmed) to collectively punish the government.
Coordinated resistance raises the costs of blowback, deterring the host
government from taking adverse action in the first place.
As noted earlier in the paper, the right of foreign investors to be free
from expropriation is accepted nearly universally, while the right to be
free from transfer restriction remains contested. If the transfer restrictions
become more broadly accepted as violations of investor property rights,
then CT will increase, investors resistance capabilities will become
more effective in reducing transfer risk, and ex post recovery of investor
losses will become more common. In the meantime, transfer restrictions
remain common and we expect that investors ability to engage in
Bank Debt
(= access to information ease of exit)
FDI
H Portfolio Equity
Portfolio Debt
Fs Ability to Mitigate Risk
Fig. 7. Summary of Investor Ability to Mitigate Risk, by Type of Risk. Notes:

Here, ability to mitigate risk is a composite of an investors ability to exit quickly
and their level of private information. It is ordered as low (L), medium (M), and
high (H). Risk is ordered by how intensively a government prepares prior to the
breach. Notice that the more steps the government must take (i.e., the more
comprehensive their preparation), the more ability investors have in general to
reduce the expected cost of the breach.
effective strategies of resistance remains limited. This is in stark contrast

to expropriation, against which we expect investor resistance to be
more effective.
Note that the factor that resistance by capable investors is likely to
increase CE than CT is precisely the condition we discussed in the previous
section, Ability to Resist. Under these conditions, an increase in CT is
not necessarily associated with a decrease in the range of conditions under
which transfer breach will occur. In related work, we test whether domestic
political constraints, which increase the accountability of the government
to the domestic public, are effective in reducing political risk. Consistent
with the theory outlined here, we find that political constraints reduce
expropriation risk but not transfer risk (Graham et al., 2015).
Hypothesis 1. Resistance is more effective at deterring governments from

engaging in expropriation than from imposing transfer restrictions.
Access to private information is potentially valuable in managing both
expropriation and transfer risk, but more valuable in managing transfer risk.
As noted previously, imposing economy-wide transfer restrictions requires
coordination across multiple branches of government. The government
attempts to keep pending transfer restrictions secret, and this need to involve
so many political actors in the decision-making process allows well-connected
investors (i.e., investors with access to local private information) to learn of
pending transfer restrictions before they occur. Thus, investors with access to
private information can anticipate transfer restrictions, those reliant on public
information cannot.
In contrast, the expropriation of assets from a single investor can be
accomplished by a single government entity with little outside coordination
and is therefore much easier for the government to keep secret until it
occurs. Even investors with high levels of private information are unlikely
to see it coming. This renders access to private information much less
valuable in managing expropriation than in managing transfer risk.
Hypothesis 2. Information is more useful in managing transfer risk than
in managing expropriation risk.
If investors cannot gain information about pending expropriation events,
then investors ease of exit is rendered less relevant you cant flee what
you cant see coming. We model the extreme case: if the government chooses
to expropriate, then the investor loses the full value of her investment,
regardless of capabilities. Partial ex post recovery of losses is possible, but
ex ante repatriation of assets is not.
Hypothesis 3. Exit is more effective at mitigating transfer risk than
expropriation or war risk.
To summarize, An investors exposure to expropriation risk is primarily
a product of the investors resistance capabilities. An investors exposure to
transfer risk, on the other hand, is affected by ease of exit (liquidity of
stake, time horizons), information, and resistance. This renders transfer
risk the most interesting subject of academic study: the losses it imposes on
foreign investors are large and growing, and investors abilities to manage
it vary across several different dimensions. Conversely, war risk is a purely
exogenous hazard and investor capabilities are largely irrelevant few if
any investors can manage war risk effectively (Fig. 8).
Classifying Risk by Amenability to Investor Capability
information- not information-

sensitive sensitive
resistance-
sensitive transfer risk expropriation
not resistance-
sensitive war
Fig. 8. Classifying Risk by Amenability to Investor Capability. Notes: This

figure categorizes risks by the degree to which they can be mitigated via resistance
or and information. Information here refers to access to private information and
ease of exit jointly.
CONCLUSION
Traditional analyses of political risk examine a single risk and a single
type of investor. In this paper, we bring together multiple risks and
investor types into a unified model, and we find reason to believe that
the insights derived from piecemeal analyses can be misleading. Direct
investors may be dramatically affected by other classes of investor,
through the risk environment that a government cultivates around those
investors; increasing a firms capacity to resist breaches of the implicit
contract does not necessarily reduce that firms vulnerability to all politi-
cal risks, and sometimes it may increase transfer risk (both relative to
expropriation, and absolutely); sometimes it is not the most predominant
investor class that is safest, but the minority investor that can use
the majority as a risk shield. Without looking at multiple risks, multiple
investors, and multiple types of risk-mitigation strategies, even simple
insights like these may go unseen, unintegrated into our theories, and
thus unaccounted for when trying to understand how political risk affects
foreign investment, and our globalizing world, more generally.
In developing a formal model of risks, investors, and capabilities, this
paper aims to generate precise and falsifiable predictions about investment
in emerging markets. We construct our model from: a more realistic
conceptualization of the complexity of political risk; the universe view of
private capital investors; and a broader assessment of the capabilities
investors employ to shield themselves from predatory, or simply unfortu-

nate, host-government behavior. Introducing this simultaneous model
allows scholars to better understand the counterintuitive trade-offs
governments make between types of sovereign theft and the costly
trade-offs investors make with regard to their location decisions and the
allocation of strategic resources.
On governments, our model shows that the traditional focus on political
constraints as an effective deterrent of sovereign theft is flawed. Even if
political constraints raise the costs of transfer breach, they may not reduce
the range of conditions under which transfer breach will occur. In more
democratic countries with higher levels of political constraints, outright
expropriation is a rare event but transfer and convertibility restrictions are
increasing and common. Political risk is not absent in these regimes; it just
occurs in altered form.
This work also has important implications for strategic management
scholars studying the firm. Although we focus heavily on the capabilities
associated with different classes of foreign capital (i.e., direct, portfolio
equity, portfolio debt, and bank debt investors), the model is directly
relevant to analyzing firm-level risk exposure and predicting firm strategic
behavior. Firms must assess their relative endowment of capabilities,
some of which may be severely constrained, to determine what advantages
they possess relative to competitors. If a firm has significant information
abilities (smart firms), perhaps born from previous experience or politi-
cal connections, entering markets with significant levels of transfer risk
give it a competitive advantage in opaque and non-transparent countries.
Firms with liquid assets and low levels of fixed property, plants, and
equipment (hot firms) may also be advantaged in such countries, yet
without information they are prone to making the costly error of rushing
to exit before they know theres a fire. Conversely, a firm with the ability
to resist and retaliate against host-country predation (strong firms),
perhaps due to significant home-country political support or capabilities
to put pressure on the host government, stands to gain from doing
business in expropriation-risky countries, which are typically shunned
by firms less endowed. A more extreme case is that miniscule pool of
investors with unique, highly sophisticated capabilities to manage war
risk; such investors should enter war-risky countries in which they will be
effectively awarded monopolies.
Drawing more directly on our model, our work further suggests that
firms must evaluate their relative position in the investor pool. For
instance, if direct investors observe that they are primarily surrounded by
portfolio debt and equity investors, they should be forewarned that transfer
risk is elevated, even as expropriation risk may be diminished. If banks
identify that the host economy is overpopulated with direct investors relative
to them, their transfer risk exposure decreases and the other investors
effectively act as a risk shield. In many ways, political risk is endogenous to
the capabilities of the current population of foreign investors operating in
the host economy, providing opportunities for much more sophisticated
risk assessment and opportunistic behavior by firms. Counterintuitively,
there can be value in being unimportant to the host government.
Thus, this unified model of political risk provides the necessary theoretical
foundation upon which to build a more nuanced and relevant analysis of
political risks, foreign investors, and strategic capabilities. The next steps in
the research program are to empirically test the theory using both country-level
capital flows data and firm-level data. We seek a predictive applied model of
global political risk.
NOTES
1. Authors calculation based on daily volumes of currency exchanged at the official
rate and the spread between the official rate and the blue market unofficial rate.
2. This typology has the added benefit of aligning with disaggregation by the
International Institute of Finance (IIF). This facilitates the use of IIF data in tests
of our theory.
3. Fig. 1 is based on data from the Institute for International Finance (IIF) and
covers investment into the 30 largest emerging markets.
4. This paper employs stylized facts about the universe of private foreign
investor types to motivate the formal models hypotheses and predictions about
firm-level strategy, as discussed in subsequent sections. We use investors and
firms as interchangeable terms, recognizing that some investors (e.g., portfolio
debt and equity investors) may represent multiple types of firms.
5. This notion of an implicit contract aligns with Frieden (1994) and Henisz and
Zelner (2005), among others.
6. Pursuing different research questions, Graham (2014) also examines
host-government commitments to enforce contracts between private parties.
7. In particular, our typology matches those of the World Banks Multilateral
Investment Guarantee Agency (MIGA); the Overseas Private Investment Corporation
(OPIC); the Berne Union; and the Credendo Group (formerly ONDD), the market
leader in private political risk insurance.
8. When investments are not liquid, governments can take time-consuming steps
such as passing laws authorizing and justifying the expropriation effectively taking
steps to make the expropriation more lawful and reduce the eventual costs of settlement.
9. We acknowledge that some extremely well-connected investors exist that can
foresee these events (and that some expropriations move slowly). In the model that
follows, if an investor could foresee expropriation, then she would face an option of
costly expedited repatriation similar to that we model for transfer risk. The more
liquid the underlying assets, the lower the costs of expedited repatriation.
10. We assume that the host government prefers more revenue to less, but cannot
control these mechanisms equally. War, for example, is seldom induced for the sake
of seizing rents from a foreign investor, but it can nevertheless lead to damages for
the investor. In this paper, we make no assumption about how governments intend
to use this revenue; however, we note that autocrats may value the personal wealth
created by rents, whereas democratic governments may value the political benefits
of redistributing the rents from foreign parties to domestic constituents.
11. G prefers to breach when the following condition is satisfied: CT R(1 ) +
V(t0 + ) + CE .
12. Note that t0 or t0 is a strategic choice made by G, and not directly by N. N
simply selects CT, which is, albeit, critical to informing Gs choice between t0
and t0.
13. If G expropriates, this model assumes that Fs intention to expedite repatriation
is unrealized that the assets will be seized in either case. This simulates the difference
between endogenous risk, which can be mitigated by informational advantages and
structural capabilities, and exogenous risk, which cannot. An alternative version could
allow F to salvage some of their assets even in the case of outright expropriation.
14. is a function of various investment indicators.
15. A future model could analyze a more general cost function, f(), such that
f(0) = 0 and f0 () 0.
16. For example, Hadfield and Weingast (2012).
17. We acknowledge that some domestic actors, including importers, are also
negatively affected, but it is foreign investors who are most likely to bear high costs.
18. Thus, V and may not be equal: G may value assets differently than F.
19. We envision as a composite of tax revenue from the international corporations,
profits from joint ventures, and income tax revenue from the public (we allow for
the possibility that foreign investment increases the domestic productivity of workers,
raising incomes, thus generating more income tax revenue for a given tax rate).
20. For an elaboration on the Cw cost parameter, see note 9.
21. In this example, we omit war because we assume that, in general, a government
does not choose war in order to increase revenue from foreign investment.
22. Notice that, since (0, 1), the upper bound of b will be 1 d.
23. For the remainder of this analysis, we focus on the -condition, which pertains
to conditions (1) and (2) and, most importantly, the CT-condition. While the other
conditions may change with , this analysis is more interested on the outcome of the
equilibrium (which is determined by the CT-condition.)
24. This emerges from the model via backward induction.
25. Fang and Owen (2011) find that international organizations can serve as a
substitute for domestic political constraints, helping autocratic countries credibly
commit to open-economy policies.
26. In the language of Bueno de Mesquita et al. (2003), part of the selectorate.
27. Such direct intervention by the home government has long been a focus of
the political risk literature (e.g., Frieden, 1994; Kahler, 1984; Kindleberger, 1969).
While early work focused on colonial and neo-colonial relationships, a recent
related literature is emerging around Chinas role in promoting and protecting

outward FDI (e.g., Shi, 2013).
28. One could model recovered funds as some proportion of the initial investment.
In the government payoffs, these funds could be separated from the rest of CE (and
would show up as a reduction in the government payoffs from expropriation), while
in the investor payoffs they would show up as a reduction in the investor loss from
expropriation.
29. Notice that increasing CT also makes condition (4) easier to satisfy, thus
making the transfer breach option more feasible for G.
30. Meaning that, when expropriation is preferable, the equilibrium in
Proposition 1 will be supportable only over a smaller subset of investments.
31. See the following section on mapping investor capabilities to risk.
32. See the discussion in the literature on hot capital, for example, Calvo,
Leiderman, and Reinhart (1996).
ACKNOWLEDGMENTS
We acknowledge helpful comments from the editors, two anonymous
reviewers, and panel participants at the American Political Science
Association Annual Meeting in Washington, DC, August 2014.
REFERENCES
Acemoglu, D., & Robinson, J. A. (2006). De facto political power and institutional persistence.
AEA Papers and Proceedings, 96(2), 325330.
Adsera, A., Boix, C., & Payne, M. (2003). Are you being served? Political accountability and
quality of government. Journal of Law, Economics, and Organization, 19(2), 445490.
Albuquerque, R., Bauer, G., & Schneider, M. (2009). Global private information in international
equity markets. Journal of Financial Economics, 94(1), 1846.
Alston, L. J., Harris, E., & Mueller, B. (2008). De facto and de jure property rights: Land settlement
and land conflict on the Australian, Brazilian and U.S. frontiers. NBER Working Paper
No. 15264.
Anderson, E., & Gatignon, H. (1986). Modes of foreign entry: A transaction cost analysis and
propositions. Journal of International Business Studies, 17(3), 126.
Avery, C., & Zemsky, P. (1998). Multidimensional uncertainty and herd behavior in financial
markets. The American Economic Review, 88(4), 724748.
Banerjee, A. V. (1992). A simple model of herd behavior. The Quarterly Journal of Economics,
107(3), 797817.
Baron, D. (1995). Integrated strategies: Market and nonmarket components. California
Management Review, 37(2), 4765.
Baron, D. (2001). Private politics, corporate social responsibility, and integrated strategy.
Beim, D., & Calomiris, C. (2000). Emerging financial markets (1st ed.). Boston, MA: McGraw-
Hill/Irwin.
Bonardi, J. P., Hillman, A. J., & Keim, G. D. (2005). The attractiveness of political markets:
Implications for firm strategy. Academy of Management Review, 30, 397413.
Brouthers, K. D. (2002). Institutional, cultural, and transaction cost influences on entry mode
choice and performance. Journal of International Business Studies, 33(2), 203221.
Bueno de Mesquita, B., Smith, A., Siverson, R., & Morrow, J. (2003). The logic of political
survival. Cambridge, MA: MIT Press.
Calvo, G., Leiderman, L., & Reinhart, C. (1996). Inflows of capital to developing countries in
the 1990s. The Journal of Economic Perspectives, 10(2), 123139.
Cohen, B. (1996). Phoenix risen: The resurrection of global finance. World Politics,
48, 268296.
Collier, R. (1999). Paths toward democracy: The working class and elites in Western Europe and
South America. Cambridge: Cambridge University Press.
Deere, C. (2008). Implementation game: The TRIPS agreement and the global politics of
intellectual property reform in developing countries. Oxford: OUP Oxford.
Delios, A., & Beamish, P. W. (2001). Survival and profitability: The roles of experience and
intangible assets in foreign subsidiary performance. Academy of Management Journal,
44(5), 10281038.
Delios, A., & Henisz, W. (2003). Political hazards, experience, and sequential entry strategies:
The international expansion of Japanese firms, 19801998. Strategic Management
Journal, 24(11), 11531164.
Dinc, S. (2005). Politicians and banks: Political influences on government-owned banks in
emerging markets. Journal of Financial Economics, 77(2), 453479.
Emmons, W. (2000). The evolving bargain: Strategic implications of deregulation and privatization.
Cambridge, MA: Harvard Business School Press.
Faccio, M. (2006). Politically connected firms. The American Economic Review, 96(1), 369386.
Fang, S., & Owen, E. (2011). International institutions and credible commitment of
non-democracies. The Review of International Organizations, 6, 141162.
Fearon, J. (1995). Rationalist explanations for war. International Organization, 49(3), 379414.
Frieden, J. A. (1994). International investment and colonial control: A new interpretation.
International Organization, 48(4), 559593.
Ghobarah, H., Huth, P., & Russett, B. (2003). Civil wars kill and maim people Long after
the shooting stops. The American Political Science Review, 97(2), 189202.
Goldstein, I., & Razin, A. (2006). An information-based trade off between foreign direct
investment and foreign portfolio investment. Journal of International Economics,
70(1), 271295.
Graham, B. (2014). Political risk and new firm entry. Retrieved from http://ssrn.com/
abstract=216613
Graham, B., Johnston, N., & Kingsley, A. (2014). Capital flows and political risk. Retrieved
from http://ssrn.com/abstract=2533651
Graham, B., Johnston, N., & Kingsley, A. (2015). Even constrained governments steal:
The domestic politics of transfer and expropriation risks. Presented at the 2013 annual
strategy and the business environment conference, Austin, TX. Retrieved from http://
papers.ssrn.com/sol3/papers.cfm?abstract_id=2106621
Hadfield, G., & Weingast, B. (2012). What is law? A coordination model of the characteristics
of legal order. Journal of Legal Analysis, 4(2), 471514.
Hadfield, G., & Weingast, B. (2013). Law without the state: Legal attributes and the coordina-
tion of decentralized collective punishment. Journal of Law and Courts, 1(1), 334.
Hegre, H. (2004). The duration and termination of civil war. Journal of Peace Research,
41(3), 243252.
Henisz, W. (2000). The institutional environment for economic growth. Economics and
Politics, 12, 131.
Henisz, W., Dorobantu, S., & Nartey, L. (2013). Spinning gold: The financial returns to
stakeholder engagement. Strategic Management Journal, 35(12), 17271748.
Henisz, W., & Zelner, B. (2005). Legitimacy, interest group pressures, and change in emergent
institutions: The case of foreign investors and host country governments. Academy of
Henisz, W., & Zelner, B. (2010). The hidden risks in emerging markets. Harvard Business
Review, 88(4), 8895.
Hennart, J. F. (2009). Down with MNE-centric theories! Market entry and expansion as
the bundling of MNE and local assets. Journal of International Business Studies,
40(9), 14321454.
Hillman, A. J., & Hitt, M. (1999). Corporate political strategy formulation: A model of approach,
participation, and strategy decisions. Academy of Management Review, 24(4), 825842.
Hirshleifer, J. (1971). The private and social value of information and the reward to inventive
activity. The American Economic Review, 61, 561574.
Hollyer, J., Rosendorff, B., & Vreeland, J. (2014). Measuring transparency. Political Analysis,
22(4), 413434.
Jensen, N. (2003). Democratic governance and multinational corporations: Political regimes
and inflows of foreign direct investment. International Organization, 57(3), 587616.
Jensen, N. (2006). Nation states and the multinational corporation: A political economy of
foreign direct investment. Princeton, NJ: Princeton University Press.
Johnston, N. (2015). Political risk and realpolitik: The politics of compensation for expropriation.
Retrieved from http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2134577
Kahler, M. (1984). Decolonization in Britain and France: The domestic consequences of international
relations. Princeton, NJ: Princeton University Press.
Kerner, A., & Lawrence, J. (2014). Whats the risk? Bilateral investment treaties, political risk
and fixed capital accumulation. British Journal of Political Science, 44(1), 107121.
Khanna, T., Palepu, K., & Bullock, R. (2010). Winning in emerging markets: A road map for
strategy and execution. Boston, MA: Harvard Business Review Press.
Kindleberger, C. (1969). American business abroad. New Haven, CT: Yale University Press.
Kingsley, A., & Vanden Bergh, R. (2015). How regulatory uncertainty drives integrated strategy
and non-market strategy. In T. Lawton & T. Rajwani (Eds.), The Routledge companion
to non-market strategy (pp. 4765). London: Routledge.
Kingsley, A., Vanden Bergh, R., & Bonardi, J. (2012). Political markets and regulatory uncertainty:
Insights and implications for integrated strategy. Academy of Management Perspectives,
26(3), 5267.
Kobrin, S. (1979). Political risk: A review and reconsideration. Journal of International
Business Studies, 10(1), 6780.
Kobrin, S. (1984). Expropriation as an attempt to control foreign firms in LDCs: Trends from
1960 to 1979. International Studies Quarterly, 28(3), 329348.
Leuz, C., & Oberholzer-Gee, F. (2006). Political relationships, global financing, and corporate
transparency: Evidence from Indonesia. Journal of Financial Economics, 81, 411439.
Li, Q., & Resnick, A. (2003). Reversal of fortunes: Democracy, property rights, and foreign
investment inflows in developing countries. International Organization, 57(1), 137.
Lux, T. (1995). Herd behaviour, bubbles and crashes. The American Economic Journal, 105, 881.
Maitland, E., & Sammartino, A. (2014). Decision making and uncertainty: The role of heuristics
and experience in assessing a politically hazardous environment. Strategic Management
Journal, 27, 149189.
Markus, S. (2012). Secure property as a bottom-up process: Firms, stakeholders, and predators
in weak states. World Politics, 64(2), 242277.
Minor, M. (1994). The demise of expropriation as an instrument of LDC policy, 19801992.
Journal of International Business Studies, 25(1), 177188.
Oetzel, J., & Getz, K. (2012). Why and how might firms respond strategically to violent
conflict? Journal of International Business Studies, 43(2), 166186.
Olson, M. (1965). The logic of collective action: Public goods and the theory of collective action.
Cambridge, MA: Harvard University Press.
Rose-Ackerman, S., & Tobin, J. (2009). Do BITs benefit developing countries? In R. P. Alford &
C. A. Rogers (Eds.), The future of investment arbitrations. Oxford: Oxford University Press.
Rumelt, R. P. (1984). Toward a strategic theory of the firm. In R. Lamb (Ed.), Competitive
strategic management (pp. 556570). Englewood Cliffs, NJ: Prentice Hall.
Sengupta, J. (2007). Dynamics of entry and market evolution. New York, NY: Palgrave Macmillan.
Shi, W. (2013). Risky business: A firm-level analysis of Chinese outward direct investment.
Paper presented at the international political economy society conference, Claremont,
CA, October 2526, 2013.
Shleifer, A., & Summers, L. (1990). The noise trader approach to finance. Journal of Economic
Perspectives, 4, 1933.
Vernon, R. (1971). Sovereignty at bay: The multinational spread of U.S. enterprises. New York,
NY: Basic Books.
Vives, X. (1990). Information and competitive advantage. International Journal of Industrial
Organization, 8, 1735.
Weingast, B. (1995). The economic role of political institutions: Market-preserving federalism
and economic development. Journal of Law, Economics, and Organization, 11(1), 131.
Wellhausen, R. (2013). Investor-state disputes: When can governments break contracts?
Journal of Conflict Resolution, 59(2), 239261.
Wellhausen, R. (2014). The shield of nationality: When governments break contracts with foreign
firms. New York, NY: Cambridge University Press.
Weston, B. (1975). Constructive takings under international law: A modest foray into the problem
of creeping expropriation. Virginia Journal of International Law, 16, 103.
APPENDIX
Proof of Proposition 1
Using the process of backwards induction, we begin with the final move of
the game; the governments expropriation decision. First, assume that the
investor (F) does not expedite the repatriation of his capital ( = 0) (we
provide this condition below, in A.3).
Suppose the government (G) chooses transfer breach (t0 ). G will play E
when his payoff for expropriation ( CE CT) is less than his payoff for
not expropriating (R(1 ) + V(t0 (1 ) + t0) CT). Solving for ,
and substituting = 0, this condition reduces to:
R1 Vt0 CE A:1
Suppose that the government chooses t0, instead. G will play E when his
expropriation payoff ( CE) is greater than his payoff for not expropriating
(R(1 ) + Vt0). Solving for , this condition reduces to:
R1 Vt0 CE A:2
Notice that, because t0 t0, this condition will be always be consistent with
condition A.1. Working backwards, we look at the investors decision to
expedite his repatriation at level .
The investor will select the amount to expedite which maximizes his
expected payoff. While G knows his transfer policy (t) before it goes into
effect, F only sees a probability p of a change to t0 . If G does not change
transfer policy, F will always prefer not to expedite repatriation:

V 0 A:3

(which is satisfied: 0 0). If G instead plays t0 , Fs preference is conditional.

For this SPE, we are looking for a condition under which F will play
= 0. We see that increasing decreases Fs payoff (V(1 ) + V[(1 t0 )
(1 ) + (1 t0)] ) when the first derivative with respect to is negative.
Solving for , this reduces to:
V A:4
Thus, when V, = 0 is optimal, regardless of p!

Continuing the backwards induction, with G playing E following t0 and
E following t0, and F selecting = 0 (for both cases), we now look at Gs
choice of transfer policy. He will choose t0 when the payoff for playing t0
( CE) is less than the payoff for playing t0 (R(1 ) + Vt0 CT).
Solving for CT, the condition reduces to:
CT R1 Vt0 CE A:5
Solving this condition for , we find: R(1 ) + Vt0 + CE CT.

Notice that this will be consistent with condition (A.1) (because CT 0), but
will only be consistent with condition (A.2) when R(1 ) + Vt0 + CE
CT R(1 ) + Vt0 + CE. Solving for CT, this condition reduces to:
CT V A:6
Working backwards, Nature moves and determines a value for CT. F does
not see this move for sure; he sees a p-likelihood of condition (A.5) being
satisfied. Then, working backwards again, Nature determines whether or
not war occurs. F does not see this move for sure: he sees an r-likelihood of
war. Finally, with conditions (A.1)(A.6), we now analyze the first move
of the game: Fs decision to invest or not.
F faces a CT-lottery and a war-lottery. He will play I when his expected
payoff for investing is greater than his break-even point of not investing
(0). Given the probabilities of transfer breach (p) and war (r), and the
expectation of the moves above, Fs expected payoff for investing is thus a
composite of two weighted averages: (1) of his payoff in the case of transfer
breach (V(1 ) + V(1 t0 )) and no breach (V): p[V(1 ) +
V(1 t0 )] + (1 p)(V), and (2) of his payoff in the case of war (r[(1 q)
V( c) qV]) and no war ([1 r][p[V(1 ) + V(1 t0 )] + (1 p)
(V)]). Altogether, his expected payoff of investing is: (r[(1 q)V( c)
qV]) + ([1 r][p[Vy(1 ) + V(1 t0 )] + (1 p)(V)]). Comparing
this weighted average to the payoff for not investing (0), and solving for r,
we see that F will play I when:
1 p p1 t0
r A:7
1 p p1 t0 1 q c q
In words, if F attributes the probability of war as greater than condition

(A.7), he will invest.
We conclude that if R (1 ) + Vt0 + CE (condition A.1),
R(1 ) + Vt0 + CE (condition A.2), V (condition A.4), CT
1 p p1 t0
V (condition A.6), and r 1 p p1 t0 1 q c q (condition A.7), a
political risk equilibrium (as defined in Proposition 1) exists for the game.
MOTIVATIONS FOR CORPORATE
POLITICAL ACTIVITY
Adam Fremeth, Brian Kelleher Richter

and Brandon Schaufele
ABSTRACT
Campaign contributions are typically seen as a strategic investment for

firms; recent empirical evidence, however, has shown few connections
between firms contributions and regulatory or performance improve-
ments, prompting researchers to explore agency-based explanations
for corporate politics. By studying intrafirm campaign contributions
of CEOs and political action committees (PACs), we investigate two
hypotheses related to public politics and demonstrate that strategic
and agency-based motivations may hold simultaneously. Exploiting
transaction-level data, with over 6.8 million observations, we show that
(i) when PACs give to specific candidates, executives give to the same
candidates, especially those who are strategically important to the firm;
and (ii) when executives give to candidates who are not strategically
important, PACs give to the same candidates potentially due to agency
problems within the firm.
Keywords: Corporate political activity; campaign contributions;
CEOs; political action committees (PACs); nonmarket strategy,
intraorganizational dynamics

ISSN: 0742-3322/doi:10.1108/S0742-332220160000034006
161
162 ADAM FREMETH ET AL.
INTRODUCTION
The public politics and corporate political activity (CPA) literatures have
sought parsimonious explanations for firm spending (Hillman, Keim, &
Schuler, 2004). Strategic investments in corporate politics are seen to open
doors, allow for information exchange, and buy policy outcomes (Austen-
Smith, 1995; Holburn & Vanden Bergh, 2014). Failing to link firms CPA
to political outcomes or performance benefits, researchers often attribute
the absence of results to agency-based explanations (Jensen & Meckling,
1976) for corporate politics where firms squander resources via political
involvement (e.g., Aggarwal, Meschke, & Wang, 2012; Hadani & Schuler,
2013). Both stories are theoretically plausible and conflicting empirical
evidence makes it challenging to give primacy to one explanation. The col-
lective tendency to celebrate the winner of a given race and to dismiss
others despite false dichotomies in empirical tests has impeded our ability
to build reliable insights into CPA and the effectiveness of nonmarket stra-
tegies. Instead it is reasonable to expect that both these explanations may
exist concurrently.
In the context of public politics, we show that both strategic and
agency-based motivations hold simultaneously for corporate-linked actors.
Rather than looking at organization-level political activity and its relation-
ship to firm outcomes, which is the focus of existing research, we emphasize
the internal dynamics of a firms CPA where multiple agents may act on a
firms behalf. We focus on transactions, which we define as the unique cam-
paign contributions from any firm-linked actor to a political candidate in a
given election cycle. Exploiting these firm-candidate pairs as the unit of
analysis enables us to explore the relationships between CEOs and political
action committees (PACs) and particular candidates. We claim that this is
the only level of analysis where we can find evidence for the simultaneity of
strategy and agency motivations as it does not necessitate mutually exclu-
sive interpretations of regression results.
Strategic motivations appear when the politician receiving a campaign
contribution is instrumental to an organization, while agency motivations
appear when the politician receiving a campaign contribution is personally
relevant to the organizations executives. An important reason for focusing
on intrafirm dynamics is that the magnitude of firm-level outcomes relative
to contributions poses challenges for accurate measurement. The amounts
spent on political investments are small relative to their expected payoffs,
even when taking into account the low probabilities of success (Kang, 2012).
Assume a firm, for example, were offered the following scenario: they could
Motivations for Corporate Political Activity 163
invest in a politically sensitive project that paid $2,000,000 with a 1% prob-

ability and $0 with 99% probability. If the firm spent $10,000 to engage in
public politics to obtain project approval (the maximum legal amount PACs
can contribute in an election cycle), the expected value of this project is
$20,000 and the expected rate of return is 100%. This successful project
would be difficult to detect empirically because it comprises less than one
twentieth of 1% of the revenue of the S&P 500s smallest firm. Hence, in our
analysis we juxtapose organizations political transactions directly against
their executives personal political transactions. This helps gauge the degree
to which one actor follows the others lead in taking either strategic or
agency-based actions. This represents a departure from the literature that
has generally focused on firm performance as the outcome and not the beha-
vior of other firm-linked actors.
Specifically, we examine what happens when either PACs or CEOs give
to particular candidates at legal limits that is, when they face exogen-
ously imposed constraints on how much they can contribute to any candi-
date. We do so in a novel dataset covering all campaign contributions
made by either S&P 500 firms or their CEOs to any Congressional candi-
date in election cycles between 1991 and 2008. Our data include 6,803,661
firm-candidate pairs. Once a PAC or executive reaches their campaign con-
tribution limit, they could continue to support a candidate by inducing
some other individual or organization to contribute. Our research design,
by recognizing these constraints, helps to unravel the dueling motivations
that underlie public politics because, presumably, if limits did not exist
both PACs and executives would be free to and hence would choose to give
more on their own. We find several key results that support both an agency
and strategy interpretation of CPA. For example, when PACs give to speci-
fic candidates at legal limits, rather than under them, executives are 8.3%
more likely to give to the same candidates and more likely to give more
money when those candidates are chairs of Congressional committees that
are relevant to the firm consistent with strategic motivations for cam-
paign contributions spilling-over from the PAC to the executive. Likewise,
we find that when executives give at legal limits to candidates who are not
strategically relevant committee chairs, their firms PACs are 45% more
likely to give to the same candidates and more likely to give more money,
results that are potentially consistent with agency explanations for
PAC giving.
A unique feature of our approach is that it examines the public politics
of all firms in the S&P 500. Analyses that focus on within industry data
may overlook the nuances of public politics that are systematic across
sectors. For instance, if we look at the proportion of CEO or PAC contri-

butions across industries that are at the statutory limits for election cycles
no clear trend emerges. PACs at firms in the Furniture and Fixtures indus-
try (SIC 25) hit their per candidate campaign contribution limits 15% of
the time when they make a contribution. In contrast, the PACs at firms in
the Oil and Gas sector (SIC 13), a sector generally considered to be politi-
cally active, only hit these limits 1% of the time. Further, an obsession with
outcomes and performance overlooks attempts to tease apart motivations
and drivers of public politics. As a result, this paper explores public politics
through executive-firm relationships within firms in all industries and sets
aside these outcome variables.
Cross-industry analysis also allows us to make more generalizable claims
about public politics. This is especially important as most research in this
area has focused on heavily regulated industries, such as finance and
energy. Firms in these industries have clear motivations for engaging politi-
cians and often do so in concert with each other and industry associations;
however, studying them may limit the generalizability of results beyond
these specialized sectors. We demonstrate the relevance of the intraorgani-
zational aspects of public politics and how executives and PACs respond to
one another. By digging into the particular transactions that are under-
taken by individuals and PACs, we uncover patterns in public politics that
have until now been unexplored. Further inquiries into the transaction level
data will help better identify discrepancies found in empirical results that
have relied on aggregated data.
Taken together our results suggest that much of the earlier literature on
public politics muddles a complex set of activities that cannot be fully
appreciated in isolation. Beyond a new set of empirical findings, a key
implication of our analysis is that researchers need to consider multiple
explanations for CPA whether strategy, agency, or some other alternative.
By adopting a pluralistic perspective, we are able to envision a richer set of
hypotheses for CPA as a part of nonmarket strategy.
TRANSACTIONAL NATURE OF CAMPAIGN

CONTRIBUTIONS
Before exploring motivations that might underlie the type of activities we
examine campaign contributions it is important to understand exactly
what they are, how they work in practice, and how they are regulated.
The extant literature often portrays the outcomes of public politics as a quid
pro quo between firms and politicians (e.g., see Grosser, Reuben, & Tymula,
2013), useful for obtaining access by opening doors (e.g., see Kalla &
Broockman, 2015), or alternatively channels through which managers exploit
agency rubbing shoulders, developing personal relationships, or seeking
post-CEO appointments (e.g., see Coates, 2012). For our empirical
approach, any or all of these may be valid and are not necessarily mutually
exclusive. A key advantage of focusing on intrafirm relationships is that we
have the advantage of being agnostic with respect to the outcomes that may
be sought.
Campaign contributions are the most widely studied form of CPA since
the Federal Election Commission (FEC) began making PAC summary
data available in the early 1980s (Milyo, Primo, & Groseclose, 2000).
Campaign contributions are monetary transfers between either individuals
or PACs, both of which can be linked to corporations, and accounts avail-
able to politicians which they can use to fund electoral campaign activities.
Campaign contributions are distinct from another prominent form of
corporate political activity, lobbying, for which disclosure records offer far
fewer details at the federal level in the United States (de Figueiredo &
Richter, 2014).
Some of the distinction in disclosure laws related to the two major types of
corporate political activity1 lobbying and campaign contributions come
from the nature of the activities themselves. It is easier to identify the target
for a campaign contribution, compared to lobbying. Only aggregate amounts
spent in a given period are required in lobbying disclosures rather than docu-
menting the counterparty for each dollar spent. Although rarely used in aca-
demic research to date, the FEC makes available not only summaries of how
much a given PAC contributes to political campaigns in aggregate, but also
data on each and every transaction a PAC initiates and the associated coun-
terparty. Transaction data are similarly available for individuals, including
executives, who contribute directly to politicians. As a result, it is possible to
use the campaign contributions data to know exactly how many dollars
are transferred from a given corporate-linked actor to a given politician.
Of course, the motivations for these monetary transfers can be complex.
Campaign contributions are protected by the First Amendment of the
U.S. Constitution as a form of freedom of speech; they are nevertheless
regulated, and specifically subject to counterparty limits given the potential
for appearance of corruption (Issacharoff, 2010; Persily & Lammie, 2004)
as the transfers could look like bribes if they were allowed in unlimited
amounts. PACs face limits on contributions at the threshold of $5,000 per
Table 1. Limits on PACs and Individuals Campaign Contributions to

Candidates per Election.
Period From PACs From Individuals
19912002 $5,000 $1,000

20032004 $5,000 $2,000
20052006 $5,000 $2,100
20072008 $5,000 $2,300
Source: Federal Election Commission (2015) (http://www.fec.gov/pages/brochures/contribli-

mits.shtml).
election and individuals in 2014 faced limits of $2,600 per election.2 The lim-
its on PACs have been constant at $5,000 since 1976 when they were fixed
in amendments to the Federal Election Campaign Act. The limits on indivi-
duals were fixed at $1,000 prior to the Bipartisan Campaign-Finance
Reform Act of 2002 at which time they were raised to $2,000 and bench-
marked to an inflation index allowing them to rise in the future. Table 1
illustrates this.
Even if individuals or PACs were to contribute the maximum allowable
amount in both a primary and general election, these amounts are designed
to be small relative to the average amount of funds candidates raise, which
Richter and Werner (2015) show is around $900,000 for the average major
party candidate over the 19912012 period. Given the relatively low legal
limits on contributions but the potentially unlimited desire for individuals
and corporations to make contributions, the limits are in fact binding for a
fair number of situations. Of course, depending upon whether the limits
bind on either individuals or PACs, what they are actually limiting will
differ: it could be an individuals attempt to exercise his freedom of speech
or a corporations attempt to influence politicians.
MOTIVATIONS FOR CORPORATE POLITICAL

ACTIVITY: STRATEGY OR AGENCY
The motivations underlying CPA have typically been considered to
improve firm performance or to advance personal goals of executives. We
take these ideas to a transaction level in the context of federal campaign
contributions to operationalize tests of strategy and agency. Our hypoth-
eses focus on the relationship between firms and their executives by
extrapolating from the proximate motivations for corporate-linked PACs and

CEOs to make campaign contributions to specific candidates. If a corporate-
linked PAC is motivated to give to a particular candidate for strategic
reasons, this may spillover into executives personal contribution decisions;
the same logic applies to spillovers into PACs contribution decisions based
on executives agency-based motivations. Recognizing that strategic motiva-
tions may explain some CPA transactions while agency may explain others,
we propose that they may occur simultaneously within firms.
Strategic Behavior of Executives
One view of CPA is that it is strategic in that its goal is to improve firm
performance as an element of a nonmarket strategy. Meta-analytic evi-
dence from 78 studies published between 1976 and 2010 supports this view
by finding a positive relationship between CPA and firm performance (Lux,
Crook, & Woehr, 2011).
A number of studies find similar relationships by taking more nuanced
approaches. Schuler (1996) identified that firms lobby for foreign trade pro-
tection when domestic demand is weak. Bonardi (2004) demonstrated that
lobbying increased when industries are deregulated and opened to foreign
competition. Beyond the issue of trade policy, firms that hire former federal
political appointees or members of Congress to serve in senior management
roles or on boards of directors enjoy positive abnormal returns in the range
of 1.62.7% (Hillman, Zardkoohi, & Bierman, 1999). Moreover, Stratmann
(1995) shows how the timing of firm-linked campaign contributions tends to
be related to politicians votes on bills and lead to positive legislative out-
comes for these special interests. Holburn and Vanden Bergh (2014) focus
on regulatory approvals for utility mergers and demonstrate the role of cam-
paign contributions as an important aspect of an integrated strategy that
incorporates both market and nonmarket features.
More recently, individuals campaign contributions, those of corporate
executives in particular, have been tied to strategic motivations. Gordon,
Hafer, and Landa (2007) show how executives with more incentive-laden
compensation schemes were more likely to contribute to federal candidates.
Fremeth, Richter, and Schaufele (2013) highlight how executives personal
contribution patterns change as individuals move up organizational hierar-
chies. Ovtchinnikov and Pantaleoni (2012) show how individuals target
politicians who have the greatest discretion over those policies specific
to the firms with which they are most closely associated. Nevertheless,
the recent literature fails to recognize the interplay between PAC campaign
contributions and those of executives. Strategic approaches to CPA must
take into account both firms and executives because transactions may be
coordinated among actors associated with a single firm. To this end, execu-
tives may want to contribute to candidates who advance their firms goals.
Moreover, executives might act on behalf of the firm when the firm faces
constraints, such as candidates who refuse contributions from PACs
(Richter & Werner, 2014).
In pursuit of strategic objectives, the CPA of a firm would utilize the
campaign contributions of both PACs and executives to advance their
agenda. This departs from previous approaches by explicitly identifying
multiple parties that are linked to a firm, each of whom has discretion over
how they deploy the funds they control. Executives are independently free
to make contributions to any candidate yet may choose to support firm
strategy by collaborating with the PAC.
The interplay between executives and PACs vis-a-vis potential political
counterparties can occur along two dimensions: (1) identifying which candi-
dates to jointly contribute to, and (2) determining the amount to contribute.
A key constraint encountered by firms using CPA for strategic pursuits is the
legal limit on how much they can contribute to any given candidate. Once
firms reach their statutory contribution limit, executives may tactically
supplement the firms contributions by contributing to these same candi-
dates. The basis for this argument is found within intrafirm dynamics: PACs
strategically select candidates to support and tapping into aligned executives
personal contributions effectively allows the PAC to circumvent its limits in
support of firm objectives.
Similarly, Cooper, Gulen, and Ovtchinnikov (2010) demonstrate that
not all candidates are of equal importance to firms and we would expect
that this internal ranking of candidates would discriminate based upon the
importance of a particular candidate to the firm. Vanden Bergh and
Holburn (2007) highlight how committees are differentially influential
relative to the whole legislative chamber (p. 10). They proceed to illustrate
how the accounting industry contributed more to relevant House and
Senate committees. As such, we define those candidates that chair, or sit
on, relevant Congressional committees as important to firms even if they
are not the most relevant politician to a particular vote or bill. Committee
chairs control the legislative agenda and governmental budget allocations
for specific industries. This includes scheduling hearings on the content of
proposed bills, determining who is invited to provide expert testimony
and allocating resources to research-specific policy topics. This makes
Congressional chairs especially important, even beyond mere committee
membership. Our empirical analysis includes both chairs and committee

members and expands the scope of analysis beyond highly regulated
sectors. Congressional committees oversee most industries and are not
exclusive to banking, utilities, and energy, areas that have generally been
the focus of research into public politics. For instance, the House
Agriculture committee oversees the agricultural, food, and tobacco indus-
tries and can be linked to Monsanto, Kraft, and Altria.
We emphasize the role of committee chairs and members rather than
pivotal politicians (Holburn & Vanden Bergh, 2002, 2004; Krehbiel, 1998)
as our predictions consider intrafirm dynamics across multiple issues and
industries. Rather than focus on specific votes or a specific bill being the
result of a transactional type of corporate political giving (Hillman & Hitt,
1999), our analysis expands the purview of nonmarket strategy to include
firms within industries irrespective of the degree of regulation and hence
focuses more on relational political strategies (Hillman & Hitt, 1999).
Executives are more likely to supplement the contributions of their PACs
to those candidates that hold these committee positions as they would be
of greatest strategic importance to the firm. Thus, we predict:
H1. Corporate executives will make campaign contributions consistent
with the strategic CPA objectives of the firms they lead:
(a) By giving to the same politicians and more to those politicians to
whom their corporate-linked PACs contribute
(b) By giving to the same politicians and more to those politicians to
whom their corporate-linked PACs are unable to legally contribute
more to given statutory constraints
(c) By giving to politicians and more to those politicians who chair or
are members of committees of importance to their firms
Agency Behavior of PACs
An alternative perspective on the use of campaign contributions is that it

represents an agency problem within the firm. Ansolabehere, de Figueiredo,
and Snyder (2003) provide an analytical literature review that establishes
links between CPA and the decisions of individual politicians are tenuous.
Hadani and Schuler (2013) propose that agency conflicts might explain why
the relationship between CPA and firm performance is not robustly positive
and perhaps negative. Aggarwal et al. (2012) find that firms whose PACs
contribute more in aggregate suffer financially and engage in other
behavior consistent with agency problems. However, agency conflicts may

or may not be the cause of a weak empirical relationship between CPA and
firm outcomes. The source of the agency conflicts, in theory, would be
preferences or priorities that motivate executives to engage in CPA but
which do not provide benefits to firms. Werner and Coleman (2015), for
example, find that in U.S. states where restrictions on campaign finance
were relaxed, legislators were more likely to pass anti-takeover laws that
were pro-management and anti-shareholder.
By looking at outcomes rather than behavior, these studies find CPA
reflects agency but say nothing about the agents themselves. For example,
a CEO that not only contributes a significant amount of her own funds to
political campaigns that are aligned with her personal interests but also
dedicates their time and that of their office to pursue these same ends could
be at odds with the strategic objectives of their firm. To this end, Coates
(2012) finds that CEOs who obtain postexecutive political appointments
are more likely to have led firms with active CPA efforts while in office.
Other motivations underlying agency behavior vis-a-vis CPA could include
whether a CEO has strong ideological views (Burris, 2001) or longstanding
social ties to politicians (Cohen & Malloy, 2014).
Despite the growing interest in an agency explanation for CPA, the
literature has not systematically tied PAC transactions to the preferences of
executives which underlie this perspective. Given agency motivations, to
whom the executive personally contributes may affect the firms CPA.
Executives with agency conflicts may coerce firm-linked PACs into con-
tributing to candidates that an executive has funded personally even when
those candidates have little strategic value to the firm. This internal
dynamic between executives and PACs enables us to measure a potential
agency conflict within firms. This extends previous literature by illustrating
executive motivations that are consistent with a narrow definition of agency
problems. Moreover, this can arise both in candidate selection and in the
amount of PAC funds allocated.
Like PACs, executives face constraints on how much they can personally
contribute to any given candidate. Consequently, when executives have met
their statutory giving limits, we are likely to see agency-motivated PAC con-
tributions to the same candidates; in these cases, the PAC would be supple-
menting the executives personal giving. While CEOs might be expected to
contribute for firm-related reasons (Fremeth et al., 2013), CEOs might also
contribute to specific candidates for personal reasons. In selected cases,
however, CEOs may encourage the PAC to make contributions unrelated
to the firms interests and to support their personal objectives. One example
highlighted by Clawson, Neustadtl, and Weller (1998) is a PAC that is

encouraged to give to the frat brother of the number-three person in the
company (p. 46). Further, we claim that any agency motivated contribu-
tion is unrelated to the strategic importance of the candidate to the firm. If
the CEOs agency motivations are a driver of PAC behavior, then money
should not flow to strategically important candidates (relevant committee
chairs or members) with the exception of cases the CEO contributes to these
individuals for personal reasons.
H2. Corporate-linked PACs will make campaign contributions consis-
tent with corporate executives agency objectives:
(a) By giving to the same politicians and more to those politicians to
whom corporate executives contribute
(b) By giving to the same politicians and more to those politicians to
whom corporate executives are unable to legally contribute more to
given statutory constraints
(c) By not giving to politicians and not more to those politicians who
chair or are members of committees of importance to their firms.
EMPIRICAL APPROACH
Our study focuses on campaign contributions by S&P 500 firm PACs and
CEOs in the 9 U.S. federal elections from 1991 to 2008. The level of analy-
sis is the PAC-candidate pair or CEO-candidate pair by election cycle; we
restrict the candidate counterparties to those who ran in general elections
for Congress but observe contributions made in both primary and general
elections within a cycle. Taking this transactional level approach to the
data enables us to test whether both strategy and agency interpretations are
consistent with the empirical evidence within the same model. This differs
from prior management studies, which tend to focus on firms aggregate
contributions over some time horizon, by expanding the number of obser-
vations in our study by the number of candidates present in each election.
Our dataset has 6,803,661 firm-candidate-election observations. Table 2
provides summary statistics.
Data
The FEC collects data on every campaign contribution to each candidate

over a $200 threshold. All of our campaign contributions data originally
Table 2. Summary Statistics for PAC and CEO Campaign Contributions.

Political action committees
Share of candidates receiving any contribution 3.0%
Of candidates receiving a contribution, share at the limit 14.4%
Share of committee chairs receiving any contribution 37.1%
Of committee chairs receiving a contribution, share at the limit 33.3%
Share of committee members receiving any contribution 20.5%
Of committee members receiving a contribution, share at the limit 16.8%
Chief executive officers
Share of candidates receiving any contribution 0.2%
Of candidates receiving a contribution, share at the limit 55.3%
Share of committee chairs receiving any contribution 4.0%
Of committee chairs receiving a contribution, share at the limit 70.6%
Share of committee members receiving any contribution 0.9%
Of committee members receiving a contribution, share at the limit 53.2%
Mean Standard Deviation
Political action committees

Contributions to candidates ($) 68.48 553.79
Contributions to committee chairs ($) 1,390.13 2,553.06
Contributions to committee members ($) 519.28 1,499.54
Chief executive officers
Contributions to candidates ($) 2.00 61.68
Contributions to committee chairs ($) 57.16 325.53
Contributions to committee members ($) 11.75 149.13
comes from FEC filings via the Center for Responsive Politics bulk data.
Data on firms and their CEOs are from COMPUSTAT. Our transaction
level dataset was constructed by establishing firm-candidate pairs for every
S&P 500 firm and every general election candidate for the 9 federal election
cycles. In general, there would be approximately 500,000 firm-candidate
pairs in a given cycle since there would be only 500 firms in the S&P 500
and about 1,000 candidates running in a general election. However, our
sample includes all firms that were in the S&P 500 at any point between
1991 and 2008 so there are 950 firms and the actual number of candidates
running in an election will vary as some seats are uncontested and others
have more than two candidates competing in the general election. As a
result, our sample includes closer to 700,000 firm-candidate pairs per cycle.
CEO contributions were then linked on a per candidate basis to this firm-
candidate pair unit of analysis. Fremeth et al. (2013) outline a small num-
ber of anomalies in the raw contribution dataseveral negative and over
limit contributions are recorded. As in that previous research, we recode all
contributions to ensure that all donations are greater than or equal to zero
and less than or equal to the FEC cycle limit. PACs directly linked to the
S&P 500 firms were identified and their per candidate contributions were
subsequently mapped onto these firm-candidate pairs. PAC-firm identifier
mappings were provided by Myers (2005).
Dependent Variables
The predictions of H1 focus on the propensity to give and the amount of
the contributions to particular candidates made by CEOs whereas those of
H2 focus on the same for PACs. CEO_Gaveijt is an indicator variable set
to 1 when a CEO from firm i contributes to candidate j in election-cycle t;
CEO_Amountijt is the amount given by those CEOs in dollars. Similarly,
PAC_Gaveijt is an indicator variable set to 1 when a PAC from firm i con-
tributes to candidate j in election-cycle t; PAC_Amountijt is the amount
given by those PACs in dollars. Table 2 shows that CEOs give $2.00 to the
average candidate and PACs give $68.48 to the average candidate. These
low numbers are driven by the large number of zeros: CEOs give to only
0.2% of all general election candidates for Congress while PACs give
to 3.0%.
Independent Variables
H1 and H2 are parallel in that they revolve around a similar set of indica-
tors for the actions of PACs or CEOs depending upon whether the lens
taken is one of strategy or agency. We employ a series of indicator vari-
ables that reflect the internal ranking of candidates revealed by the PAC
and CEO behavior. Within this ranking, the omitted categories for the
purposes of our empirical analysis are those candidates that receive no con-
tributions from the PAC in the case of H1 or the CEO in the case of H2.
Part (a) of H1 and H2 focuses on candidates who received contributions
below the limit. Hence, we use indicator variables PAC_Below_Limitijt and
CEO_Below_Limitijt each of which equals one when either the PAC or
CEO makes a contribution to a candidate that is greater than zero but
below the statutory limit. PAC_Below_Limitijt is equal to one when the
PAC makes a contribution from $200 to $4999 and CEO_Below_Limitijt is
equal to one when the CEO makes a contribution from $200 to $1 less than
the individual limits presented in Table 1. Part (b) of H1 and H2 focuses
on whether PACs or CEOs reach legal limits. Campaign finance laws dic-
tate the maximum amount individuals or PACs may contribute to a candi-
date in each election. For either the PAC or the CEO, candidates who have
received contributions at the limit are more important in satisfying either
strategy or agency motivations than those candidates who have not

received any contributions or those that receive below-limit contributions.
PAC_Limitijt is an indicator taking on a value of 1 when a PAC reaches
its statutory limit on giving to a particular candidate for an election;
CEO_Limitijt is the same for a CEO. Table 2 illustrates that CEOs reach
this contribution limit in 55.3% of the contributions that they make, while
PACs reach this contribution limit in 14.4% of their contributions.
Part (c) of H1 and H2 focuses on whether candidates chair or sit on
committees of importance to specific firms. Ovtchinnikov and Pantaleoni
(2012) create a mapping between Congressional committees and industries
over which they have jurisdiction based on the committees own statements
about the scope of their duties. Those authors specifically link Senate and
House committees by name to specific four-digit Standard Industrial
Classification (SIC) industry codes in Table B1 of that article; these identi-
fiers in turn can be linked to candidates in our datasets committee
membership using Stewart and Woons (2014) Congressional Committees
dataset3 and firms in our dataset using SIC codes found in COMPUSTAT.
Committee_Chairijt takes a value of 1 if candidate j chairs a committee
important to firm i given Ovtchinnikov and Pantaleonis (2012)
mapping during election cycle t. We explore an alternative definition of
important candidates which we define as Committee_Memberijt if candidate
j is a member of a committee important to firm i. Whether we examine the
role of chairs or all committee members depends upon the exact specifica-
tion. In order to test part (c), we need specifications where we interact these
variables with (i) whether these politicians received contributions and (ii)
whether those contributions were at the limit. As a result, we include the set
of interactions to further tease apart the internal ranking of candidates that
would result from either strategic or agency motivations. Table 2 demon-
strates that CEOs and PACs average contribution to important committee
chairs equals $57.16 and $1,390.13, respectively, notably more than the
amounts given to average candidates. A similar pattern appears in contribu-
tions to any relevant committee members. Further, when giving to important
chairs, CEOs reach limits in 70.6% of their contributions to strategic chairs
and PACs reach their limits on 33.3% of this same class of candidates.
Statistical Models and Their Interpretation
Linking our hypotheses to empirical tests is most straightforward when we

have a separate model for each hypothesis. We structure our empirical
specification such that we map specific parts of each of our two hypotheses
directly to sets of coefficients.
To test H1, we estimate a model specified as:
CEOijt 1 PAC Below Limitijt 2 PAC Limitijt 3 Committeeijt

4 PAC Below Limitijt Committeeijt
5 PAC Limitijt Committeeijt i t ijt 1
Our sub-hypotheses are tested using three F-tests with the following nulls:
H1a0 : 1 2 4 5 0
H1b0 : 1 4 2 5
H1c0 : 3 1 2 4 5
In an effort to isolate the strategic giving by the CEO, in model (1) we

focus on the CEOs giving in response to actions by the PAC. H1a states
that when PACs contribute to candidates CEOs are more likely to give
(more) to those same candidates. The first null, H1a0 ; tests the joint hypo-
thesis that CEOs respond to the PACs contributions. If we reject this null
and the CEO does respond to the PACs activity, we interpret this as evi-
dence of the CEO giving strategically on behalf of his firm. H1b posits an
increase in CEO contributions to those candidates for whom the PAC is
unable to make further contributions. The null, H1b0 ; contrasts the CEOs
response to the PAC making below-limit and at-limit contributions.
Rejection of this null could be interpreted as stronger evidence that the
CEO is acting strategically on behalf of his firm given additional logic:
when a PAC hits its statutory limit, an exogenously imposed constraint, on
giving to a particular candidate in support of the firms objectives, it has
done all that it legally can on its own to support its preferred candidate;
however, it could induce other agents, namely, the CEO, to contribute
additional sums. Finally, H1c examines whether the PACs contributions
to important committee chairs or members leads to a greater CEO response
to these individuals as demonstrated by testing the null H1c0 . We expect to
reject this null and find a positive and statistically significant relationship
between candidates of political value to a firm and the propensity of the
CEO to respond to the PAC vis-a-vis those candidates. Given how we
have defined Committeeijt as chairs or members of relevant Congressional
committees, these may not be the only candidates of strategic value to the
firm; however, they should be among the candidates with the greatest poli-
tical value.
The structure and logic for the tests of H2 are parallel to those for H1
although we reverse the roles of CEOs and PACs and switch strategic moti-
vations for agency motivations. In an effort to focus on agency-motivated
behavior of PACs, in model (2) we focus on PACs giving in response to
actions by the linked-firms CEO. To test H2, we estimate a model speci-
fied as:
PACijt 1 CEO Below Limitijt 2 CEO Limitijt 3 Committeeijt

4 CEO Below Limitijt Committeeijt
5 CEO Limitijt Committeeijt i t ijt 2
In this case, the sub-hypotheses are tested with the following nulls:
H2a0 : 1 2 4 5 0
H2b0 : 1 4 2 5
H2c0 : 3 4 5 > 0
H2a states that when CEOs contribute to candidates, PACs are more likely
to give (more) to those same candidates, so the first null, H2a0 , tests the joint
hypothesis that PACs respond to the CEOs contributions. If we reject this
null and find PACs do respond to CEOs giving based on personal prefer-
ences over candidates, it is suggestive of the PACs response to the CEOs
agency motivations. H2b posits an increase in PAC contributions to those
candidates for whom the CEO is unable to make further contributions.
Rejection of this null, H2b0 , implies that the PAC has an additional response
to the CEOs contributions when the CEO is no longer able to make larger
individual contributions. This could be interpreted as stronger evidence that
the PAC responds to the CEOs agency motivations if the CEO reaches his
personal contribution limit. Finally, H2c tests whether the PACs response
to CEOs contributions differs between those made to important committee
chairs or members and other candidates who the CEO may favor but are
not in such positions. We test this by examining the third null, H2c0 , that sets
the three relevant coefficients greater than zero. In general, the PAC can
respond to the CEOs contribution patterns by giving to either important
committee members or other candidates. If the response to CEOs contribu-

tions to noncommittee candidates is greater than the response to his contri-
butions to candidates who are members of important committees, then this
pattern suggests agency and we would reject the null.
In model (1), CEOijt represents either CEO_Gaveijt or CEO_Amountijt
depending upon whether we are testing for changes in the marginal prob-
ability of CEOs giving or changes in the dollar amount of giving.
Similarly, in model (2), PACijt represents either PAC_Gaveijt or
PAC_Amountijt depending upon what specifically we are testing. In both
models (1) and (2), firm-specific, time invariant characteristics such as
industry, location of head office, and corporate governance practices are
captured by i, the firm fixed effect. Time-varying factors that are common
to all firms such as FEC regulations or the nature of the political environ-
ment are contained in t. Respectively, vijt and ijt represent the error terms
in the models. When we are testing for changes in the marginal probability
of contributing to a particular candidate, we estimate a linear probability
model;4 otherwise we use a least-squares fixed effects estimator to control
for a range of unobservables.
EMPIRICAL RESULTS
Table 3 presents the results from empirical model (1) focused on how PAC
contributions influence CEO contributions. Results demonstrate that
executives make political contributions that are consistent with strategic
motivations of the firm, supporting H1. Columns I and II provide estimates
of the marginal probability that the average firms CEO responds to their
PAC by contributing to a particular candidate. Columns III and IV pro-
vide estimates of the amount of money contributed to a given candidate by
the average firms chief executive.
For candidates who are not on a committee and did not receive at-limit
contributions from a firms PAC, Columns I and II show that a CEO is
1.4% more likely to donate to them if the PAC had given to that candidate.
Columns III and IV show that under the same situation a CEO contributes
approximately an additional $17 to these candidates. When a firms PAC
reaches it legal limit (i.e., PAC_Limitijt=1, Committeeijt=0) for a candidate,
the effect on CEOs is larger: increasing the probability that he gives by
8% and the amount by $134. In this scenario, where the PAC has exhausted
its legally allowable capacity to give to a specific candidate, the CEO
Table 3. The Effect of PAC Campaign Contributions on CEO

Contributions.
Any CEO Contribution CEO $ to Candidate
to Candidate
I II III IV
PAC_Below_Limit 0.014*** 0.014*** 16.90*** 17.40***

(0.001) (0.001) (1.14) (1.19)
PAC_Limit 0.083*** 0.083*** 133.36*** 134.90***
(0.006) (0.007) (10.91) (11.42)
Committee chair 0.019*** 30.71***
(0.004) (7.21)
Committee member 0.004*** 4.64***
(0.001) (0.64)
PAC_Below_Limit*Committee chair 0.001 12.28
(0.008) (11.30)
PAC_Limit*Committee chair 0.047** 64.45*
(0.023) (37.66)
PAC_Below_Limit*Committee member 0.006*** 7.28***
(0.001) (1.90)
PAC_Limit*Committee member 0.001 7.35
(0.008) (14.11)
Election cycle fixed effects Yes Yes Yes Yes
Firm fixed effects Yes Yes Yes Yes
F-statistic 29.54*** 28.16*** 25.45*** 25.15***
Observations 6,803,661 6,803,661 6,803,661 6,803,661
F-stat for H1a 31.45*** 96.89*** 24.68*** 76.48***
F-stat for H1b 24.10*** 75.71*** 26.46*** 66.59***
F-stat for H1c 21.15*** 87.39*** 15.28*** 70.49***
***p < 0.01, **p < 0.05, *p < 0.10. Values in parentheses are clustered standard errors with
clustering on firms.
potentially provides a supplementary donation to that candidate from his

private bank account. Given that the average CEO only contributes $2 to
the average candidate the premium in the amount given by the average
CEO is 850% larger when his firms PAC contributes and 6,700% larger
when his firms PAC contributes at its legal limit. When we consider particu-
lar committee members of greater relevance to firms (i.e., Committeeijt=1),
the results suggest that a CEO is 1.9% more likely to contribute to candi-
dates that chair Congressional committees important to his firm when his
PAC has not contributed to this candidate; we estimate $30.71 in additional
CEO giving to these candidates. The results are weaker but statistically
significant when we consider all relevant committee members.
When we examine the patterns of CEO contributions in response to

PAC giving to committee chairs, we are able to compare the below-limit
and at-limit effects. CEOs are 3.4% more likely to give to committee chairs
if their PAC had made a below-limit contribution and 14.9% more likely if
that contribution was at the limit.5 These estimates are statistically signifi-
cant at the 1% level with corresponding standard errors of 0.007 and 0.022
on the (joint) marginal effects. Furthermore, CEOs are likely to give an
additional $35 when the PAC has made a below-limit contribution to the
relevant committee chair and $228 more when the PAC makes an at-limit
contribution. Both of these marginal effects are statistically significant at a
1% level with standard errors of 8.59 and 36.71, respectively. The values
for committee members are also statistically significant at the 1% level but
the estimates are of a smaller magnitude.
The bottom panel of Table 3 illustrates the F-tests of H1. We reject the
nulls for each sub-hypothesis and each model specification providing strong
suggestive evidence that CEOs make strategic contributions in response to
their PAC.
Table 4 presents the results from empirical model (2) focused on how
CEO contributions influence PAC contributions. In it, we flip the direction
of influence from Table 3, demonstrating that PACs make political contri-
butions that are consistent with agency objectives of the CEO. Ultimately,
we present evidence consistent with H2a but not for H2b or H2c. Columns
I and II provide estimates of the marginal probability that the average
PAC contributed to a particular candidate in response to giving by the
CEO. Columns III and IV provide estimates of the amount of money con-
tributed to a given candidate by the average firms PAC.
For candidates who did not receive an at-limit contribution from a CEO
and who were not members of a relevant committee, Columns I and II
show that a PAC is about 40% more likely to donate to them if the CEO
had given to that candidate at a lower level. Column III shows that under
the same situation, a PAC contributes an additional $1,995 to these candi-
dates and in Column IV an additional $1,898. When a firms CEO reaches
his legal limit (i.e., CEO_Limitijt=1, Committeeijt=0) to these candidates,
the effect on PACs is larger: increasing the probability that it gives to the
same candidate by 4.5% in both Columns I and II. Looking at the dollar
values the corresponding estimates are $2,416 and $2,339. In this scenario,
where the CEO has exhausted his legally allowable capacity to give to a
specific candidate who is not a member of an important committee,
the PAC potentially provides a supplementary donation from corporate-
linked accounts. These results are economically meaningful. Given that
Table 4. The Effect of CEO Campaign Contributions on PAC

Contributions.
Any PAC Contribution PAC $ to Candidate
to Candidate
I II III IV
CEO_Below_Limit 0.406*** 0.396*** 1,994.65*** 1,898.24***

(0.021) (0.021) (136.02) (136.75)
CEO_Limit 0.451*** 0.447*** 2,415.92*** 2,339.11***
(0.021) (0.020) (136.90) (135.58)
Committee chair 0.316*** 1,180.08***
(0.019) (94.29)
Committee member 0.171*** 425.39***
(0.011) (36.58)
CEO_Below_Limit*Committee chair 0.113 1,003.57
(0.088) (835.50)
CEO_Limit*Committee chair 0.162*** 522.13
(0.059) (475.54)
CEO_Below_Limit*Committee member 0.038 574.34**
(0.033) (258.13)
CEO_Limit*Committee member 0.079** 545.81*
(0.039) (280.62)
Election cycle fixed effects Yes Yes Yes Yes
Firm fixed effects Yes Yes Yes Yes
F-statistic 50.99*** 49.24*** 30.36*** 29.38***
Observations 6,803,661 6,803,661 6,803,661 6,803,661
F-stat for H2a 27.16*** 126.80*** 31.63*** 124.12***
F-stat for H2b 0.00 0.06 0.00 1.77
t-stata for H2c 0.32 0.89 2.66 3.46
***p < 0.01, **p < 0.05, *p < 0.10. Values in parentheses are clustered standard errors with
clustering on firms.
a
H2c involves a single restriction and a one-sided test, so a t-statistic is reported.
the average PAC contributes $68.48 to the average candidate the pre-
mium in the amount given by the average PAC is 2,900% larger when his
firms CEO contributes and 3,600% larger when the firms CEO contri-
butes at his legal limit.6
Next, we consider contributions from the PAC to candidates that chair
or are members of committees of importance to the firm. In Column IV
when the CEO gave, but not at their limit, to a committee member the
additional amount was $574 and the total marginal amount equaled $2,989
(s.e. $286). When the CEO gave at their limit to a committee member the
PAC contributed an additional $522 with a total marginal effect of $4,118
(s.e. $307). Therefore, when we assess H2a and H2b, we have conflicting
evidence. We are able to reject the null in H2a the PAC does respond to
CEO contributions.7 Although we do observe an increase in PAC contribu-
tions when the CEO reaches his limit, the increase is not statistically signifi-
cantly different than a below limit contribution as demonstrated by the
F-stat for H2b provided at the bottom of Table 4.
In the case when CEOs do not contribute to a chair of a relevant con-
gressional committee, their firms PAC still contributes $1,180, but when
the CEOs does give but below their statutory limit, their PAC contributes
$4,273. However, when the CEO is at their limit, the PAC actually contri-
butes marginally less at $4,118. H2c posits that committee membership
would be less meaningful if agency motivations are dominant; however, we
are unable to reject our one-directional statistical test of no difference in
contribution patterns between committee members and noncommit-
tee members.
The stability of our coefficient estimates in all of our models lends con-
fidence to the interpretation of our results. The hypotheses stated in H1
have empirical support in Table 3: corporate executives make campaign
contributions consistent with the objectives of their firms. The overall
hypotheses stated in H2 have weak support in Table 4: PACs do respond
to CEO contributions as outlined in H2a suggesting the potential for
agency. Yet, we find no evidence of agency in H2b or H2c. We have
shown that these motivations exist simultaneously given that our
empirical setup exploits variation in PAC or executive relationships with
candidates, and consequently does not rely on false dichotomies present in
past studies of agency- or strategy-motivated CPA. Given the conflicting
incentives of both actors, agency may be present but it appears to be
dominated by strategic motivations.
DISCUSSION
Simultaneous Presence of Strategic and Agency-Motivated Behavior
By developing theories about the inter-relationships between CEOs and

PACs campaign contributions at the transactional level, we contribute to
the literature by bridging disparate extant results underlying the motiva-
tions for public politics. We find evidence that: (i) agency is apparent within
PACs contributions, (ii) strategy is apparent within CEOs personal
contributions, and (iii) elements of both agency and strategy exist side by
side within the actors contributions.
Prior literature fails to recognize that multiple motivations underlying
CPA may be present simultaneously. This idea is most evident at a transac-
tional level: agency-based and strategic motivations could be present in the
same transaction if, for instance, an important committee member with
great strategic value to the firm was a college roommate of an executive at
a firm. The firm wants to contribute to a particular candidate because he
controls policy important to its performance and the executive wants to
give to this candidate because they are old friends. This action could be
interpreted in different ways depending upon the perspective. Recognizing
this duality leads us to the simple, but important, conclusion that agency
and strategy views of CPA are not mutually exclusive.
When applying an agency or strategy lens to CPA, the presence of the
other perspective constrains how far either can take us. We need to recog-
nize that the preferences of firm-linked actors may not perfectly correlate
with the positions of key political candidates. To illustrate this point and
the limits of strategically motivated contributions by corporate executives,
consider an example where an executive has strong pro-choice views: she is
unlikely to contribute personally to a staunch pro-life politician with her
own money no matter how important that politician is to her firm. To illus-
trate the limits of agency-motivated contributions by corporate-linked
PACs, consider an example where a candidate is in favor of raising the
minimum wage but happened to be the college roommate or golfing partner
of the CEO at a firm where the bulk of employees earn that wage: while the
CEO may contribute personally to his friend, it becomes strategically
inconsistent for the corporate-linked PAC to do the same. Hence, while
what is important to the firm may influence executives behavior, that
influence only extends so far and vice-versa. Our results suggest that this
interpretation has merit as agency and strategy coexist broadly across S&P
500 firms.
Interorganizational Dynamics in Public Politics
Shifting prevailing interpretations of CPA away from false dichotomies,

our results enable management scholars to devise more nuanced explana-
tions for observed behavior. Previous research ignores the complex realities
of CPA in practice. In particular, we highlight that the internal organiza-
tion of CPA activities and the extent to which they are coordinated among
corporate-linked actors have been understudied within the nonmarket

strategy literature. Moreover, that executives and PACs are separate actors
with their own motivations and interests has not been considered empiri-
cally. This is important because they each may contribute to political candi-
dates for the purpose of either short-term transactions or long-term
relationships (Hillman & Hitt, 1999). However, the degree of independence
remains an open question. Some CEOs, for example, may closely monitor
their PAC, while others do not. Sabato (1984) points out that the organiza-
tional structure of PACs can differ substantially and that while corporate
PACs are usually closely tied to the companys chief executive
office the CEO then delegates much of his authority to the public affairs
executive in the company and to the PACs governing board (p. 34).
Further investigation into internal structure and its implications for public
politics is warranted.
To the extent that prior research has examined coordination within the
CPA literature, it has been limited to coordination within industries
(Bombardini & Trebbi, 2012; Lenway & Rehbein, 1991; Olson, 1965),
rather than coordination among individuals associated with a firm. Shifting
to an intrafirm focus has two broad implications for further research: first,
the emphasis can be placed on actor behavior. This is particularly attractive
given the conflicting results within the existing literature and the long-term
orientation of nonmarket strategies. By tracking the behavior of firm-
linked actors other than the PAC, that is, by tracking the behavior of
executives linked to firms, we open the door to further research exploring
these interconnections; future research could examine interconnections
between lower level employees, lower level executives, board members, and
the lobbyists firms hire. Similarly, it would allow for better insight into the
interrelationships that are developed over time between firm-linked actors
and particular political or regulatory agents.
Second, moving from coordination within an industry to coordination
within a firm provides researchers with a perspective on nonmarket strategy
that mirrors market strategy in ways not previously considered. Firms may
compete for policy favors just as they compete in the market: focusing the
empirical analysis on internal organization enables researchers to compare
firms nonmarket strategies as they would with market strategies. Many of
the elements of nonmarkets strategies, such as who develops and directs the
strategy, how resources are marshaled, and how funds are allocated to spe-
cific activities, have received much attention in conventional strategy
research while nonmarket strategy has treated firms as a black box. As
we demonstrate, nonmarket strategies can have both strategic and agency
elements and we begin to open this box. Future research needs to investi-
gate the organizational features of strategic behavior and how firms man-
age agency conflicts when conducting CPA. Several prominent open
questions exist not only in the coordination among executives and PACs
but also in the coordination between executives and nonexecutive employ-
ees as an alternative approach to contravene campaign finance limits.
Managerial Implications
Increasingly, shareholders and the media are scrutinizing the political activ-
ity of the firm and its leaders. This has brought greater attention to the
political counterparties that are recipients of campaign contributions. In
April 2014, Brendan Eich, the CEO of Mozilla, came under fire for a string
of campaign contributions that had been made to Republican candidates
and social initiatives over the previous two decades. This led to a series of
internal resignations and external boycotts of Mozillas products.
Ultimately, Eich was forced to resign amid the controversy (Bilton &
Cohen, 2014). This recent episode demonstrates the blurred line between
what is personal (e.g., socially conservative positions) and what is job-
related when it comes to an executives political activities. The difference
between the two can become muddled when outsiders bring unanticipated
attention to the issue. The controversy may then spillover beyond an indivi-
dual executive as their firms face undue pressure that would interfere both
their market and nonmarket strategies. For instance, Mozilla encountered
customer boycotts from the controversy sparked by their CEOs political
contributions and it is conceivable that their future political overtures
would receive a cool reception as politicians avoid being tainted with the
controversy. As a result, managers must recognize that they are political
agents of the firm irrespective of what they choose to do with their pri-
vate dealings.
Likewise, increasing levels of CPA have yielded calls for corporate gov-
ernance reforms that would result in greater transparency and controls
over the direction of political spending (Bebchuk & Jackson, 2013). These
often take the form of shareholder initiatives and proxy votes at share-
holder meetings. For instance, Proctor & Gamble faced such a vote at its
2013 annual meeting when an institutional investor recommended that the
Board of Governors adopt a policy outlining the firms electioneering and
political contributions and communications activities. Among other
things, key provisions included a report and budget for political activities
and a declaration of the congruency of such activities with company values

and policies (Proctor and Gamble Company, 2014). Adoption of this or
similar initiatives requires managers to consider how the public or competi-
tors might respond to greater knowledge about their firms or their own
political actions.
Growing investor attention on the political activities of firms and grow-
ing media attention on individual mangers begs the question of the value in
alignment between the political activities of firms and executives. We
observe potentially conflicting motivations at play among S&P 500 firms;
however, we are unable to comment on how this shapes firm performance
or the success of particular initiatives in the nonmarket arena. Conceivably,
those firms that are able to better coordinate the contributions of the PAC
and their senior executives may be able to present a more cohesive position
vis-a-vis their political counterparties and at the same time avoid attacks
along ideological positions.
Highly versus Less Regulated Industries
The empirical approach taken in this paper departs from convention in the
study of public politics in many ways but the choice to focus on the beha-
vior of all S&P 500 firms stands out. To examine motivations, generally,
we assume that the degree of regulation that an industry faces should not
matter for the motivation underlying the relationship between an executive
and PACs contribution behavior vis-a-vis particular candidates. For
instance, a CEO with strong postemployment political aspirations would
manifest agency-like behavior that encumbers their firm in the heavily regu-
lated Tobacco Products sector (SIC 21) just as they would in the less regu-
lated Food Stores sector (SIC 54).
A closer examination of our data highlights this point as those sectors
that are dominated by transactions where the CEO gave at the limit to a
candidate but a PAC provided no contributions to that same candidate
include a diverse set of sectors, that are both more and less regulated. The
top decile includes financial services (SIC 64 and 67), manufacturing (SIC
39), apparel (SIC 23), leather products (SIC 31), heavy construction (SIC
16), and water transportation (SIC 44). Similarly, there would appear to be
no systematic trend in the data for those sectors where both the PAC and
the CEO gave at the limit to the same candidate. Here, the top decile
includes Amusement Services (SIC 79), Furniture (SIC 25), Personal
Services (SIC 72), Textile Mills (SIC 22), and Food Products (SIC 20).
The role of industry yields no clear trends via this casual empiricism. This
combined with the conflicting evidence when regressing industry-specific
roll call votes against campaign contributions suggests that there is likely
something else at play beyond the degree of regulation that explains activ-
ity in public politics. Our analysis demonstrates that a greater consideration
should be placed on the individuals that are making the choice to be politi-
cally active and the particular counterparties that they target when they
wade into public politics.
Limitations
There are several important limitations in the interpretation of our results.

First, the statutory limits are different on individuals and PACs implying
that the sizes of the coefficients across models have distinct proportional
meanings. Second, our models assume that behaviors are coordinated and
deliberate within the firm. It is still possible, though highly unlikely, that
the observed behavior is entirely coincidental. Thus, these results should
not be interpreted as causal, but are suggestive of key trends. Along the
same lines, our econometric models do not account for prospective slates
of candidates. It is possible that PACs, industries, or CEOs may group can-
didates when allocating campaign contributions rather than treating them
as individuals. To the extent that PACs target slates of candidates, our
econometric results may be biased. The existence of or potential for slates
opens a range of formal theoretical and empirical questions that merit
future analysis (Chamon & Kaplan, 2013). Another potential source of
bias is that we are unable to observe the precise timing of primary and gen-
eral election contributions and as a result there is a potential for a small
number of cases where one firm-linked party undertakes a contribution in
the primary election whereas the other firm-linked party acts in the gen-
eral election.
We defined important candidates as those who chaired or were members
of relevant Congressional committees. It is reasonable to expect that there
are many additional methods to classify important politicians. For example,
legislators who have demonstrated effectiveness at writing good policies
(Volden & Wiseman, 2016), legislators who sponsor or co-sponsor bills reg-
ulating the industry, or candidates whose constituencies are located near the
firms operations could each be considered relevant for firm performance
(Ovtchinnikov & Pantaleoni, 2012).
Finally, we have attempted to show one channel through which agency
may manifest. However, we must acknowledge that it is empirically
challenging to identify agency motivations and their alignment with firm

performance are fundamentally unobservable. As a result, researchers need
to use indirect proxies to provide evidence which suggests behavior that
encumbers firm performance. As PACs alter their behavior in response to
the CEOs choices, we do not observe a pure consumption effect. Similarly,
the negative coefficients on the interaction of CEO contributions and stra-
tegic, firm-linked actors hint that a frat brother effect may exist.
Implications for Future Research Design
Despite its limitations and despite our empirical work representing partial
correlations rather than getting to causality, our paper has implications for
new research designs in the public politics space. The limits on campaign
contributions, for example, may be useful in other research contexts as
may other features of the institutional environment that restrict corporate
involvement in politics. This paper only looks at federal politics in the
United States and hence only one institutional environment; however, pro-
mising future research may examine state level politics in the United States
or multiple countries as there would be a greater degree of variation across
more institutional settings. To some extent this variation across settings
may help future research get closer to the quasi-experimental ideal
for causality.
Another way that future research may be able to get closer to a quasi-
experiment would be to find better proxies for the firms and CEOs long-
standing political preferences. Variables we could use to proxy for CEOs
underlying preferences are easier to speculate about than for firms. CEOs
personal contribution data from the periods prior to them entering office
could be used (Fremeth et al., 2013). Another potential proxy for CEO pre-
ferences would be to dig into voterfile data that contain information on
individuals voter registration data including party affiliations and records
of whether individuals voted in a given election (Cooper, Haspel, &
Knotts, 2009).
A final way to get closer to causality in a research design on motivations
for giving would be to exploit the timing of contributions. PACs may con-
tribute before CEOs or vice-versa and this may be informative as to
whether each party is responding to one another. Such a research design
would require some further exogenous variation and one route would be to
examine how PACs behave when a CEO has announced their retirement.
The window following such an announcement would be insightful as
the outgoing CEO is acting as lame duck and there could be a de-linking
of the coordinating activities that we have identified above.
CONCLUSION
Despite the prior focus on external coordination, there are many interest-
ing, managerially relevant questions that remain unexplored about internal
coordination, such as the role of the CEO as an independent political actor,
the roles of employees in public politics, and whether firm-linked PACs
always pursue goals that are in the best interests of the corporate entity.
Exploiting firm-specific transactional data is likely to be a fruitful arena for
public politics research as management researchers proceed to uncover the
multifaceted realities of the relationships between firms, PACs, CEOs, and
politicians.
NOTES
1. A more recent type of corporate political activity is independent expenditures

where organizations can allocate funds toward advertising on issue-based initiatives
or in support of individual candidates provided the advertisements are not coordi-
nated with the election campaigns of the candidates themselves. The development
of this activity is outside our empirical window. Moreover, there is little empirical
evidence that this is a widely used channel by corporations (Bonica, 2014) or that it
is effective (Werner, 2011).
2. The recent 2014 McCutcheon v. the Federal Election Commission Supreme
Court decision overturned aggregate limits on individual contributions that would
prevent individuals from giving the maximum possible to each candidate, but did
not overturn the legality of limiting amounts on the transactions we consider here.
3. Stewart and Woons Congressional Committees dataset is available at: http://
web.mit.edu/17.251/www/data_page.html
4. Because we are estimating nearly 1,000 fixed effects and because we care about
the marginal effect rather than the ability to predict, we use a linear probability
model (Angrist & Pischke, 2009).
5. These are the joint effects when either PAC_Limitijt=1 or PAC_Below_
Limitijt =1 and Committeeijt=1.
6. These estimates can be considered a lower bound on the agency effects if
CEOs are able to exploit their PACs without contributing from their personal bank
accounts or if their preferred candidate may also be strategically important to the
firm so they dont need to contribute to them personally.
7. An alternative test of H2a would test the null of 1 + 2 = 0. We reject this
possibility in all models, for example an F-stat for Column II is equal to 507.49.
REFERENCES
Aggarwal, R. K., Meschke, F., & Wang, T. Y. (2012). Corporate political donations:
Investment or agency? Business and Politics, 14, Article 3.
Angrist, J. D., & Pischke, J. (2009). Mostly harmless econometrics. Princeton, NJ: Princeton
University Press.
Ansolabehere, S., de Figueiredo, J. M., & Snyder, J. (2003). Why is there so little money in
U.S. politics? Journal of Economic Perspectives, 17, 105130.
Austen-Smith, D. (1995). Campaign contributions and access. American Political Science
Review, 89, 566581.
Bebchuk, L. A., & Jackson, R. J. (2013). Shining light on corporate political spending.
Georgetown Law Journal, 101, 923947.
Bilton, N., & Cohen, N. (2014). Mozillas chief felled by view on gay unions. New York Times.
Retrieved from http://bits.blogs.nytimes.com/2014/04/03/eich-steps-down-as-mozilla-
chief/?_r=0
Bombardini, M., & Trebbi, F. (2012). Competition and political organization: Together or
alone in lobbying for trade policy? Journal of International Economics, 87, 1826.
Bonardi, J. P. (2004). Global and political strategies in deregulated industries: The asymmetric
behaviors of former monopolies. Strategic Management Journal, 25, 101120.
Bonica, A. (2014). Avenues of influence: On the political expenditures of corporations and their
directors and executives. Retrieved from http://ssrn.com/abstract=2313232
Burris, V. (2001). The two faces of capital: Corporations and individual capitalists as political
actors. American Sociological Review, 66, 361381.
Chamon, M., & Kaplan, E. (2013). The Iceberg theory of campaign contributions: Political
threats and interest group behavior. American Economic Journal: Economic Policy,
5, 131.
Clawson, D., Neustadtl, A., & Weller, M. (1998). Dollars and votes: How business campaign
contributions subvert democracy. Philadelphia, PA: Temple University Press.
Coates, IV, J. C. (2012). Corporate politics, governance, and value before and after citizens
united. Journal of Empirical Legal Studies, 9, 657696.
Cohen, L., & Malloy, C. J. (2014). Friends in high places. American Economic Journal:
Economic Policy, 6, 6391.
Cooper, C. A., Haspel, M., & Knotts, H. G. (2009). The value of voterfiles for U.S. state
politics research. State Politics & Policy Quarterly, 9, 102121.
Cooper, M. J., Gulen, H., & Ovtchinnikov, A. V. (2010). Corporate political contributions
and stock returns. The Journal of Finance, 65, 687724.
de Figueiredo, J. M., & Richter, B. K. (2014). Advancing the empirical research on lobbying.
Annual Review of Political Science, 17, 163185.
Fremeth, A. F., Richter, B. K., & Schaufele, B. (2013). Campaign contributions over CEOs
careers. American Economic Journal: Applied Economics, 5, 170188.
Gordon, S. C., Hafer, C., & Landa, D. (2007). Consumption or investment? On motivations
for political giving. Journal of Politics, 69, 10571072.
Grosser, J., Reuben, E., & Tymula, A. (2013). Political quid pro quo agreements: An experi-
mental study. American Journal of Political Science, 57, 582597.
Hadani, M., & Schuler, D. (2013). In search of El Dorado: The elusive financial returns on
corporate political investments. Strategic Management Journal, 34, 165181.
Hillman, A. J., & Hitt, M. A. (1999). Corporate political strategy formulation: A model of
approach, participation and strategy decisions. Academy of Management Review,
24, 825842.
research agenda. Journal of Management, 30, 837857.
Hillman, A. J., Zardkoohi, A., & Bierman, L. (1999). Corporate political strategies and firm
performance: Indications of firm-specific benefits from personal service in the U.S.
government. Strategic Management Journal, 20, 6781.
Holburn, G. L. F., & Vanden Bergh, R. G. (2002). Policy and process: A game-theoretic
framework for the design of non-market strategy. Advances in Strategic Management,
19, 3366.
Holburn, G. L. F., & Vanden Bergh, R. G. (2004). Influencing agencies through pivotal politi-
cal institutions. Journal of Law, Economics, and Organization, 20, 458483.
Holburn, G. L. F., & Vanden Bergh, R. G. (2014). Integrated market and nonmarket strate-
gies: Political campaign contributions around merger and acquisition events in the
energy sector. Strategic Management Journal, 35, 450460.
Issacharoff, S. (2010). On political corruption. Harvard Law Review, 124, 118142.
Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm: Managerial behavior, agency
costs, and ownership structure. Journal of Financial Economics, 3, 305360.
Kalla, J. L., & Broockman, D. E. (2015). Campaign contributions facilitate access to congres-
sional officials: A randomized field experiment. American Journal of Political Science,
In Press.
Kang, K. (2012). Policy influence and private returns from lobbying in the energy sector.
Working Paper, Tepper School of Business, Carnegie Mellon. Retrieved from http://
repository.cmu.edu/cgi/viewcontent.cgi?article=2440&context=tepper
Chicago Press.
Lenway, S. A., & Rehbein, K. (1991). Leaders, followers, and free riders: An empirical test of
variation in corporate political involvement. Academy of Management Journal,
34, 893905.
Lux, S., Crook, T. R., & Woehr, D. J. (2011). Mixing business with politics: A meta-analysis
of the antecedents and outcomes of corporate political activity. Journal of Management,
37, 223247.
Milyo, J., Primo, D. M., & Groseclose, T. (2000). Corporate PAC campaign contributions in
perspective. Business and Politics, 2, 7588.
Myers, B. (2005). Corporate political activity and asset pricing. Working Paper, Anderson
School of Management, University of California-Los Angeles. Retrieved from http://
papers.ssrn.com/sol3/papers.cfm?abstract_id=950104
Olson, M. (1965). The logic of collective action: Public goods and the theory of groups. Boston,
MA: Harvard University Press.
Ovtchinnikov, A. V., & Pantaleoni, E. (2012). Individual political contributions and firm
performance. Journal of Financial Economics, 105, 367392.
Persily, N., & Lammie, K. (2004). Perceptions of corruption and campaign finance: When
public opinion determines constitutional law. University of Pennsylvania Law Review,
153, 119180.
Proctor and Gamble Company. (2014). Annual report. Retrieved from http://www.pginvestor.
com/interactive/lookandfeel/4004124/PG_Annual_Report_2014.pdf
Richter, B. K., & Werner, T. W. (2014). Campaign contributions from corporate executives in
lieu of political action committees. Working Paper. Retrieved from http://papers.ssrn.
com/sol3/papers.cfm?abstract_id=2237685
Richter, B. K., & Werner, T. W. (2015). Sources of congressional candidates funds: Does inter-
est group money dominate? In A. Cigler, B. Loomis, & A. Nownes (Eds.), Interest
group politics (9th ed.). Washington, DC: Congressional Quarterly Press/Sage.
Sabato, L. J. (1984). PAC power: Inside the world of political action committees. New York,
NY: W.W. Norton & Company.
Schuler, D. A. (1996). Corporate political strategy and foreign competition: The case of the
steel industry. Academy of Management Journal, 39, 720737.
Stewart, C., & Woon, J. (2014). Congressional committees, modern standing committees,
103rd112th Congresses dataset. Retrieved from http://web.mit.edu/17.251/www/data_
page.html
Stratmann, T. (1995). Campaign contributions and congressional voting: Does the timing of
contributions matter? Review of Economics and Statistics, 77, 127136.
Vanden Bergh, R. G., & Holburn, G. L. F. (2007). Targeting corporate political strategy:
Theory and evidence from the U.S. accounting industry. Business and Politics, 9, 131.
Volden, C., & Wiseman, A. E. (2016). Incorporating legislative effectiveness into nonmarket
strategy: The case of financial services reform and the great recession. Advances in
Strategic Management: Strategy beyond Markets, 34.
Werner, T. (2011). The sound, the fury, and the non-event: Business power and market
reactions to the citizens united decision. American Politics Research, 39, 118141.
Werner, T., & Coleman, J. J. (2015). Citizens united, independent expenditures, and agency
costs: Reexamining the political economy of state antitakeover statutes. Journal of Law,
Economics, and Organization, 31, 127159.
THE MARKET FOR
LEGISLATIVE INFLUENCE
OVER REGULATORY POLICY
Rui J. P. de Figueiredo, Jr. and Geoff Edwards
ABSTRACT
We show that, in the US telecommunications industry, market partici-

pants have a sophisticated understanding of the political process, and
behave strategically in their allocation of contributions to state legisla-
tors as if seeking to purchase influence over regulatory policy. We find
that interests respond defensively to contributions from rivals, take into
account the configuration of support available to them in both the legisla-
ture and the regulatory commission, and vary their contributions accord-
ing to variations in relative costs for influence by different legislatures.
This strategic behavior supports a theory that commercially motivated
interests contribute campaign resources in order to mobilize legislators
to influence the decisions of regulatory agencies. We also report evidence
that restrictions on campaign finance do not affect all interests equally.
The paper therefore provides positive evidence on the nature and effects
of campaign contributions in regulated industries where interest group
competition may be sharp.
Keywords: Non-market strategy; campaign finance; political economy;
telecommunications regulation; regulation; political institutions

ISSN: 0742-3322/doi:10.1108/S0742-332220160000034007
193
194 RUI J. P. DE FIGUEIREDO, JR. AND GEOFF EDWARDS
INTRODUCTION
One of the central questions in electoral politics is how money, influence,
and policy outcomes are related. While substantial scholarship on cam-
paign contributions has developed over the last couple of decades, direct
evidence on whether contributions and legislative behavior are related is
inconclusive, although the public and media perception of a relationship is
strong. A less direct, but no less instructive approach to answering the
question is to ask whether contributors act strategically in their allocation
of campaign resources to politicians. In other words, do interests tailor
their contribution strategies to the political and institutional environment
in a manner consistent with maximizing the political return on contribution
dollars? While some progress has been made from this angle, the existing
literature has yet to fully incorporate a theory of contributions in exchange
for legislative influence over regulatory policy.
We study a six-year dataset of campaign contributions to state legisla-
tors from competing interests in the local telecommunications industry. We
show that the interests in this industry are sophisticated and strategic in
their contribution decisions. Furthermore, the contribution patterns we
observe are consistent with a theory that the interests seek to purchase leg-
islative influence over the decisions of independent regulatory commissions.
In essence, we describe a market for the purchase of legislative influence
over regulatory policy.
The existing literature on campaign contributions can be classified into
two streams: studies of whether contributions affect legislative behavior; and
studies of the determinants of contribution patterns, where contributions
become the dependent variable. A substantial body of literature has devel-
oped on whether campaign contributions affect legislative behavior, but
unfortunately, the evidence to date is mixed. Much of the prior research in
this area has examined effects of contributions on roll call votes by members
of Congress.1 Ansolabehere, de Figueiredo, and Snyder (2003, hereafter
ADS) survey 36 empirical studies of contributions and roll call votes and
conclude that the weight of the evidence so far favors the view that contribu-
tions are unrelated to voting behavior.2 Combined with evidence that the
vast majority of contributions come in small sums from individual donors
rather than through organized Political Action Committees (PACs), ADS
propose that most contributions reflect the consumption value individuals
receive from giving to campaigns, rather than expectations of private returns
on investments. While plausible for the bulk of contributions, ADS acknowl-
edge that their theory is less applicable to contributions from corporate and
The Market for Legislative Influence over Regulatory Policy 195
industry (economic) interests and note that we may find the effects in
domains other than direct legislative action such as regulatory arenas. This
exception alluded to by ADS is of particular interest since the 2010 Citizens
United Supreme Court decision lightened constraints on these contributions.
Even if the majority of contributions are motivated by the consumption
value of consumer contributors, and are consequently benign, the minority
of contributions from organized interests with economic motives might still
significantly affect legislative behavior, in which case normative concerns for
the system of campaign finance remain.
A plausible reason why many studies have failed to find strong evidence
of a link between contributions and roll call votes is that votes may be a
poor currency of exchange. Herndon (1982) and Hall and Wayman (1990)
have argued that the gains for interests from vote buying are marginal,
while the risks of suspicion of improper influence on both political careers
and the reputations of interest groups are considerable, particularly as both
contributions and votes are on the public record. But as many political
scientists have observed, it is not necessary to imagine a situation of out-
right vote buying in order to suspect that contributors act strategically with
a view to securing the services of legislators. Non-vote related legislative
services are perhaps both more important and more readily exchangeable.
An important function of contributions from interest groups is to purchase
access to legislators and to keep both doors and minds open to the merits
of their positions (Herndon, 1982). Beyond mere access, Denzau and
Munger (1986) and Hall and Wayman (1990) propose that interest groups
provide political resources in an implicit exchange for policy relevant ser-
vices or effort from legislators.3 Hall and Wayman (1990) provide evidence
that contributions are allocated by interests in order to mobilize otherwise
latent legislative support in the committee stage of the legislative process,
and speculate that similar mobilization is likely to occur where the legisla-
tive service required by interests is influence over executive agencies.4
A clear advantage of these forms of legislative service is that the often
informal nature of committee processes and interactions with regulatory
agencies limits the scope for public scrutiny of legislators responsiveness to
contributing interests, even where the issues involved are highly salient
(Hall & Wayman, 1990).
Despite these persuasive arguments, and their long grounding in the politi-
cal science literature, very few studies have tested for an effect of contributions
on non-vote related behavior of legislators. Few studies that have examined
this relationship support the proposition that private money influences non-
vote related legislative behavior. In addition to Hall and Waymans (1990)
study of participation in Congressional committees, Hansen and Park (1995)

and de Figueiredo and Edwards (2007) assume contributions mobilize legis-
lators to influence the decisions of independent regulatory agencies, and
test for a relationship between contributions to legislators and regulatory
outcomes: the former at the Federal level, concerning decisions of the
International Trade Administration; the latter at the state level, concerning
regulatory outcomes in the telecommunications industry the context of
our current research. Both studies find contributions to legislators to be
predictive of regulatory decisions. Together, these three studies provide
consistent support for the proposition that, vote-buying aside, contribu-
tions purchase legislative services and effort, including legislative influence
over regulatory policies set by independent regulators.5
Turning from studies of the effects of contributions to studies of the
determinants of contribution patterns, there is consistently strong evidence
that interests behave as rational investors in legislative outcomes. Again,
the work in this area is dominated by studies of contributions to members
of the US Congress. Important empirical studies in this regard owe theore-
tical debt to Denzau and Munger (1986) and include: Grier and Munger
(1986, 1991, 1993); Poole, Romer, and Rosenthal (1987), Snyder (1990,
1992), Stratmann (1991, 1992, 1995, 1996), Endersby and Munger (1992),
Romer and Snyder (1994), and Kroszner and Stratmann (1998). These stu-
dies suggest that interest groups allocate contributions with regard to char-
acteristics of legislators that affect either their willingness to provide
services to interest groups (determined by their ideology, the preferences of
their geographic reelection constituency, and the intensity of electoral com-
petition they face) or their productivity in providing such services (e.g.,
membership on a relevant committee and seniority).6 Most of these studies
assume interests representing relatively homogenous industries that seek
legislative services unopposed but for unorganized constituents. Less com-
mon has been work that examines the effects of organized interest group
competition on contribution patterns, as will be necessary in the context of
our research on the telecommunications industry. Austen-Smith and
Wright (1994) present and test a model of counteractive lobbying behavior
by competing interests, and Kroszner and Stratmann (1998) model the con-
tribution behavior of competing interests in the course of testing a positive
theory that Congressional committees exist to foster repeated dealings
between interests and committee members.7 Both papers report that, in the
context of interest group competition over legislative policy, competing
interests respond defensively to the non-market activities of their rivals.8
Complementary to this literature has been the very robust if more recent
strand of scholarship on campaign contributions from the perspective of the
non-market strategy of firms. In a series of papers on campaign contribu-

tions, Fremeth, Richter, and Schaufele (2013) connect contribution activity
to strategic behavior of executives and firms. Macher and Mayo (2012, 2015)
examine cross-national data to highlight how market and non-market com-
petition are highly important in understanding political influence activities of
firms in general and in regulatory arenas in particular. Further, Holburn and
Vanden Bergh (2014) note that particular events such as mergers and acquisi-
tions prompt substantially higher contributions from firms attempting to pro-
tect potential rents created by successful prosecution of these transaction.
Perhaps closest to the study here is Holburn and Vanden Berghs extensive
theoretical and empirical work tracing the same pathway of influence studied
here: namely from firm to legislator to regulator. Holburn and Vanden
Bergh (2004, 2008) first develop a series of theoretical predictions (consistent
with many of the results developed below) about when interest groups influ-
ence regulatory policy indirectly through the legislature. In subsequent work
(Holburn & Vanden Bergh, 2014; Vanden Bergh & Holburn, 2007), they test
these results indicating that the channel we explore here appears in other
regulatory politics settings by showing that contributions are indeed strategic.
The primary difference between this work and the present study is that while
they examine cases of client politics (i.e., a single interest group attempting
to influence a legislator and regulator), we examine a setting where there
is competition between groups, or so-called interest group politics. In
addition, considered together with the related work in de Figueiredo and
Edwards (2007), the work here also provides subsequent linkages to
non-market outcomes which has been elusive in other work.
Our current research adds to the literature on the determinants of contri-
bution patterns. Intense interest group competition in the industry we study
allows us to contribute to the less developed empirical research in this area.
Our research is also novel in two important respects. First, we study patterns
of contributions in the context of an industry in which legislative influence
over regulatory policy is likely to be the main focus of non-market activities
by competing interests, and we reveal patterns in contributions that are parti-
cular to that form of legislative service. We provide strong evidence that inter-
ests in the telecommunications industry contribute strategically to legislators
with a view to influencing the decisions of independent regulatory agencies.
In this sense, the current study complements recent research that is itself novel
in demonstrating a link between the relative contributions of competing inter-
ests and important regulatory outcomes (de Figueiredo & Edwards, 2007).
Second, we study patterns of contributions to state legislators and test
the generalizability to the state level of the extensive learning on contribution
patterns derived from analyses of representatives of the US Congress.
Our research is, to our knowledge, the first to test theories of the strategic
determinants of contribution behavior at the state level. The state level offers
great potential for research in this field due to the scope to develop both time
series and cross-sectional variation in dependent and explanatory variables.
Both of these contributions fit squarely in the broader, if newer, litera-
ture on the non-market strategy of firms. As noted in the introduction to
this volume (de Figueiredo, Lenox, Oberholzer-Gee, & Vanden Bergh,
2016), this burgeoning field requires a greater understanding of the linkages
between the factors that affect firms strategic choices, the way those trans-
late into non-market outcomes, and how those connect to firm perfor-
mance. In this context, this paper provides the underpinnings for the first
two of those links. First, it demonstrates the import of strategic factors in
environments where firms compete in the non-market arena. This stands in
contrast to the literature as noted above which often studies contexts when
firms within an industry either collaborate on non-market strategy or
are unopposed in the sense that Wilson (1990) termed client politics.
Second, the paper provides an important underpinning for the second
link between strategies of firms and outcomes by providing the strate-
gic foundation for the policy outcomes studied in de Figueiredo and
Edwards (2007).
We proceed as follows. The section Research Design introduces the
research design and explains the choice of empirical context for this study.
In the section Theory and Hypotheses, we draw from a broad range of
literature to develop conjectures regarding the strategic nature of contribu-
tions by interests seeking influence over regulatory policy, and provide pre-
dictions as to how contribution patterns are determined by elements of the
political and institutional environment. These hypotheses provide a foun-
dation for documenting patterns which we can evaluate in the data, our
primary purpose. In the section Econometric Specification, Data and
Measures, we set out our empirical model and describe our data and mea-
sures. The next section reports results, and the final section concludes,
including a discussion of some future directions for research on patterns of
contributions to state legislators.
RESEARCH DESIGN
We examine patterns of contributions from rival interests in the local tele-
communications industry to candidates for state legislatures. We assume
that the primary objective of legislators is to maximize votes and prospects
of reelection (Fiorina, 1977; Mayhew, 1974); that legislators maximize
votes by providing services to geographic constituents, and raising

campaign resources (which can be used in advertising and campaigning to
purchase votes or defend votes from challengers); that as a quid pro quo
for campaign resources, legislators provide services to contributing private
interests (Chappell, 1982; Denzau & Munger, 1986; Grier & Munger, 1991;
Welch, 1974); and that these services extend beyond activities related to the
drafting and voting on legislation (Hall & Wayman, 1990; Kroszner &
Stratmann, 1998). In particular, we assume that legislators are capable of
influencing the decisions of independent regulatory commissions.9
The empirical context for our research concerns the contribution pat-
terns of rival entrant and incumbent firms in the local telecommunications
industry since the introduction of a regime permitting entry and competi-
tion in 1996. We choose this industry focus for several reasons. First, this
industry is well suited to examination of hypotheses regarding the contribu-
tion behavior of interest groups seeking to purchase legislative influence
over regulatory policy. Since 1996, a focus of the industry at the state level
has been on regulatory policy, and the main service legislators can offer in
this context is influence over the decisions of the independent regulatory
commissions (also known as Public Utility Commissions or Public Service
Commissions).10 Second, the regulatory battle over entry into the local
telecommunications industry offers the chance to study the contribution
patterns of competing interests, rather than an homogenous industry
group. The interests are well organized and funded on both sides and the
regulatory battle is essentially a zero-sum game a decision benefiting
entrant firms imposes an equally large cost on incumbents, and vice versa.
Third, with the decisions of the regulatory commissions carrying enormous
potential to determine the prospects for entry and competition in local
telecommunications, and the profitability of the rival interests, effective
non-market strategies are particularly important in this industry. Finally,
deliberations on regulatory policy and the interactions between legislators and
regulators are typically of low saliency to the general constituency, providing
scope for influence activities in the absence of close public scrutiny. Overall,
we suspect that if strategic contribution behavior to influence regulatory
outcomes is to be revealed anywhere, it will be in a context in which a set of
conditions such as these prevail.
THEORY AND HYPOTHESES

As noted earlier, our purpose in this paper is to document the degree of and
factors which may lead to counteractive campaign contributions by
competitive firms in a regulatory context. To that end, we draw on the exist-

ing literature to derive and test six empirical conjectures or, more loosely,
predictions concerning the strategic contribution behavior of the rival inter-
ests in the local telecommunications industry. The first prediction concerns
contribution behavior in the context of interest group competition for legis-
lative services. Two further predictions concern contribution patterns under
varying political compositions of the regulatory commission and the legisla-
ture. The final three predictions concern relationships between contributions
and determinants of the price contributors face for the supply of legislative
influence services.
Responsive Contributions: All else equal, we expect that interests will
respond defensively to the contributions of their rivals, consistent with a
number of models and prior tests of the non-market strategies of competing
interests (Austen-Smith & Wright, 1992, 1994; Baron, 2001; Bernheim &
Whinston, 1986; Grossman & Helpman, 1994; Kroszner & Stratmann, 1998).
In our study, as we look at the overall pattern of contributions to state legisla-
tures rather than contributions to individual legislators; it is not necessary that
matching contributions go to the same politicians. Contributions to one legis-
lator could be matched by a rival interests contributions to a second legislator
in an attempt to mobilize the second politician to defend the rivals position
against the influence activities of the first.
Hypothesis 1. All else equal, interests contribute defensively in response
to contributions from rival interests.
Party Ideologies: We suspect that the party compositions of the regulatory
commission and the legislature will be important for the contribution stra-
tegies of incumbent and entrant firms. As noted elsewhere (see, e.g., de
Figueiredo & Edwards, 2007), Republicans are most likely to support the
interests of regulated incumbent telecommunications firms, while
Democrats, interested in the benefits for consumers of greater competition
in telecommunications, are more likely to support the interests of
entrants.11 Contributions to purchase legislative influence over regulatory
outcomes will be most valuable where the regulatory commission is
not ideologically aligned with the interests position. Indeed, where the
regulatory commission is already aligned, purchasing legislative influence
might be unnecessary. For example, incumbents will be more likely to
contribute to legislators when Democrats control the regulatory commis-
sion than when the commission is controlled by Republicans. Conversely,
entrant contributions should be higher when the regulatory commission
is Republican.
Hypothesis 2. All else equal, contributions will be greater when the party
ideology of the regulatory commission is not aligned with the contribut-
ing interest.
In addition, we expect that the dominant party ideology in the legislature
will affect contribution strategies in an interactive way. Interestingly the lit-
erature is somewhat divided on when and to whom groups will contribute
to; this is particularly true in the context of competitive interest groups. On
the one hand, in the context of one-sided vote buying, Snyder (1991) posits
that a group will contribute to marginally opposed policymakers. On the
other hand, there are at least two reasons to believe in a competitive context
we may see that friendly legislators are the objects of focused influence.
First, strategically, when there are competitive groups, it may be that groups
will act defensively and support friendly policymakers to make purchase
of these voters more difficult (see, e.g., Groseclose & Snyder, 1996). Second,
if contributions are intended to mobilize legislative support in the context of
a broad array of choices for even a friendly legislator, they will tend to be
directed to legislatures predisposed to support the interests position, and
that consequently offer low supply prices for the provision of legislative
services on behalf of that interest (Denzau & Munger, 1986; Grier &
Munger, 1991; Hall & Wayman, 1990).12 Which of these two views holds is
essentially an empirical question.
Based on these factors, we test the degree to which the latter argument
holds when a regulator is misaligned with the group (vs. the null that the
opposite holds as with Snyder, 1991). Specifically, we evaluate whether con-
tributions are greatest where the regulatory commission is not aligned with
the contributing interest, but the legislature is aligned. For example, when
Democrats control the regulatory commission, are incumbents more likely
to contribute to purchase legislative influence from friendly Republican
legislatures than from less friendly Democrats?
Hypothesis 3. All else equal, given an unaligned regulatory commission,
interests will contribute more to aligned (friendly) legislatures than to
unaligned legislatures.
Constituencies and Electoral Competition: Next we explore further the idea
that interests are more likely to contribute to friendly legislatures
(in contrast to regulators). The literature has identified three factors that
determine the willingness of legislators to provide services: their ideology;
the preferences of their geographic reelection constituency; and the degree
of electoral competition they face. The effects of alternative party
ideologies on contribution patterns in the local telecommunications indus-

try have just been mentioned: we expect incumbent firms to contribute
more to Republican legislatures and entrant firms to contribute more to
legislatures controlled by Democrats.13 Constituency characteristics have
also been shown to determine costs for legislative services.14 Legislatures
with constituencies that have more (less) reason to favor competition and
entry in telecommunications have lower supply costs for servicing entrant
(incumbent) firm interests and are consequently more (less) likely to receive
contributions from entrants than from incumbents.15
Hypothesis 4. All else equal, interests will contribute more when consti-
tuency characteristics are more favorable.
The literature also predicts that the willingness to supply legislative services
will increase, and contributions will increase, as electoral competition
intensifies legislators will be more willing to provide services in exchange
for campaign resources in tight electoral battles.16
Hypothesis 5. All else equal, interests will contribute more as electoral
competition intensifies.
Elected Commissions: Supply prices for legislative services depend not only on
the willingness of legislators to provide services, but also on how effective leg-
islative effort may be. Legislators will be less willing to supply service when
the effectiveness of legislative effort to influence regulators is lower.17 An ele-
ment of the institutional environment of telecommunications regulation that
could affect the degree of legislative influence is whether regulatory commis-
sioners are elected or appointed.18 We propose that elected commissions enjoy
greater independence from legislative influence, and legislators are therefore
less productive in seeking to influence regulatory decisions.19 Costs for legisla-
tive services are consequently higher, and interests will contribute less.
Hypothesis 6. All else equal, interests will contribute less when regula-
tory commissioners are elected.
Controls: In order to test these predictions, we include important controls
in our analysis. A further aspect of the institutional environment is particu-
larly relevant: campaign finance laws. We control for the effects of statu-
tory prohibitions on corporate contributions, and limits on corporate and
PAC contributions. We expect that the existence of prohibitions and limits
on contributions will reduce contributions from the rival interests, although
ADS (2003) note that contribution limits for Congressional candidates are
rarely binding, and Che and Gale (1998) argue that, in the context of
interest group competition, tighter limits could in fact increase aggregate

contributions by constraining contributions from high value interests but
encouraging low value interests to contribute more aggressively. We also
control for effects on entrant (incumbent) contributions of the presence of
a major entrant (incumbent) firm headquarter office in a state,20 and, as the
largest incumbent firms are geographically delimited into four separate
regions, we control for the possibility that these firms employ different con-
tribution strategies in their respective regions.21 Finally, we expect that
both entrants and incumbents will contribute most aggressively in the lar-
gest states, where the potential gains from favorable regulatory decisions
will be greatest (Tripathi, 2000). We therefore control for state size using
Gross State Product (GSP).22
ECONOMETRIC SPECIFICATION, DATA

AND MEASURES
As we study the contribution strategies of competing interests, we estimate
a simultaneous equation model of contribution behavior summarized in
Eqs. (1) and (2):
C E it E E C I it E Xit E Z E it E it 1
C I it I I C E it I Xit I Z I it I it 2
where CitE and CitI are contributions from the rival interests in state i and
election cycle t; X is a vector of exogenous political and institutional vari-
ables that affect the contribution patterns of both interests; ZE and ZI are
vectors of exogenous variables that are peculiar to entrant and incumbent
contribution patterns; and E and I are error terms for each equation.
We utilize a panel dataset of the 50 US states over three electoral cycles
(19971998, 19992000, and 20012002),23 so the unit of analysis is a
state-cycle.24 Descriptive statistics of the variables included in the analysis
are provided in Table 1. This table indicates that there is substantial varia-
tion in all variables in our analysis. A correlation matrix is presented in
Table 2. State-by-state variable means are displayed in Table 3.
Table 1. Descriptive Statistics.

Variable Observations Mean Standard Minimum Maximum
Deviation
Entrant contributions 145 7.170 10.218 0 66.026

($ per 1,000 capita)
Incumbent contributions 145 14.409 15.665 0 100.933
($ per 1,000 capita)
Legislative ideology 150 0.493 0.430 0 1
Commission ideology 150 0.64 0.460 0 1
Metropolitan population (%) 150 0.678 0.205 0.278 1
GSP in FIRE 150 5.545 2.857 2.322 18.404
($ thousands per capita)
Electoral competition 147 0.127 0.088 0.393 0.005
Gross state product 150 0.179 0.215 0.016 1.260
($ trillions)
Elected commission 150 0.233 0.424 0 1
Prohibition on corporate 150 0.353 0.480 0 1
contributions
Limits on corporate and PAC 150 1.6 0.655 0 2
contributions
Entrant headquarters 150 0.067 0.250 0 1
Incumbent headquarters 150 0.267 0.444 0 1
Qwest 150 0.28 0.451 0 1
SBC 150 0.24 0.429 0 1
Verizon 150 0.26 0.440 0 1
Data on campaign contributions from the telecommunications industry

to candidates for state legislatures (lower and upper houses) were obtained
from the Institute on Money in State Politics (www.followthemoney.org).25
Complete data was available for most states for each of the three electoral
cycles.26 Contributions at the state level can come directly from corpora-
tions, from Political Action Committees (PACs) or from individuals.27 We
therefore measure contributions from entrant and incumbent interests as
the sums of contributions direct from entrant and incumbent corporations,
from PACs associated with each type of corporation, and from individual
employees of each type of corporation.28 In all, the contribution dataset we
use consists of 54,649 contributions from more than 1,000 different contri-
butors, totaling just over $23 million. Statewide totals of contributions on
behalf of each interest were calculated for each electoral cycle. These totals
were then normalized for state size using state population in thousands.
The units of our final measures of contributions from the rival interests are
therefore contribution dollars per 1,000 capita.
Table 2. Correlation Matrix.
Entrant Incumb Leg Id Com Id Metro GSP Elect GSP Elected Prohib Limit Entrant Incumb SBC Qwest Verizon
in FIRE Comp HQ HQ
Entrant contributions 1
Incumbent contributions 0.4672 1
Legislative ideology 0.1788 0.2234 1
Commission ideology 0.0695 0.1902 0.0669 1
Metropolitan population 0.1775 0.1012 0.1167 0.1435 1
GSP in FIRE 0.0576 0.2264 0.0856 0.086 0.5636 1
Electoral competition 0.0998 0.0746 0.0931 0.0598 0.0943 0.0989 1
Gross state product 0.0913 0.0677 0.0716 0.1771 0.5573 0.3189 0.2306 1
Elected commission 0.2747 0.1486 0.0914 0.0045 0.2176 0.3174 0.0168 0.2032 1
Prohibition on corporate 0.2629 0.1326 0.1139 0.1385 0.1914 0.0164 0.1385 0.233 0.0363 1
contributions
Limits on corporate and 0.129 0.2336 0.1085 0.1163 0.0154 0.1386 0.1937 0.1924 0.0066 0.1598 1
PACs contributions
Entrant HQ 0.0666 0.0603 0.1296 0.0462 0.0198 0.155 0.0897 0.1822 0.007 0.1977 0.1014 1
Incumbent HQ 0.0296 0.0228 0.1203 0.1889 0.3932 0.3796 0.1668 0.5508 0.1893 0.1182 0.1184 0.2407 1
SBC 0.2997 0.2754 0.0474 0.1489 0.1846 0.1512 0.2238 0.3522 0.1759 0.2991 0.0914 0.0524 0.2561 1
Qwest 0.0045 0.0068 0.4518 0.0789 0.2845 0.2046 0.0357 0.276 0.1548 0.2669 0.2516 0.1572 0.2501 0.3457 1
Verizon 0.2365 0.4118 0.176 0.0669 0.179 0.5623 0.1767 0.0047 0.3072 0.0816 0.148 0.109 0.1959 0.3395 0.3588 1
206
Table 3. Variable Means by State (Averaged over Three Electoral Cycles).
State Leg Com Metro GSP Electoral GSP Elected Prohib Limits IXCHQ ILECHQ Qwest SBC Verizon Entrant Incumb
Id Id % in FIRE Compet Com Corp Corp&PAC Contrib Contrib
0.16
RUI J. P. DE FIGUEIREDO, JR. AND GEOFF EDWARDS

AK 1 0.33 0.42 4.22 0.02 0 1 2 0 0 0 0 0 21.91 66.43
AL 0 0.67 0.70 3.42 0.15 0.11 1 0 1 0 0 0 0 0 0.77 30.07
AR 0 0.67 0.49 2.66 0.25 0.06 0 0 2 0 1 0 1 0 1.02 31.35
AZ 0.83 1 0.88 5.02 0.09 0.14 1 1 2 0 0 1 0 0 1.00 1.00
CA 0 0.67 0.97 7.79 0.11 1.20 0 0 0.67 0 1 0 1 0 17.06 19.54
CO 0.83 0.67 0.84 6.16 0.06 0.15 0 0.67 2 0 1 1 0 0 9.24 12.41
CT 0 1 0.96 12.80 0.10 0.15 0 1 2 0 1 0 0 1 9.31 0.47
DE 0.5 0.33 0.81 16.67 0.15 0.03 0 0 2 0 0.33 0 0 1 5.04 5.53
FL 1 0.67 0.93 5.73 0.12 0.43 0 0 2 0 0 0 0 0 11.95 18.09
GA 0 1 0.69 5.03 0.08 0.26 1 0 2 0 1 0 0 0 0 0.63
HI 0 0 0.73 7.34 0.28 0.04 0 0 2 0 0 0 0 0 11.12 9.49
IA 1 0.33 0.45 4.21 0.06 0.09 0 1 1 0 0 1 0 0 8.50 33.52
ID 1 1 0.39 2.96 0.34 0.03 0 0 2 0 0 1 0 0 14.21 24.07
IL 0.5 1 0.85 7.20 0.04 0.43 0 0 0 0 1 0 1 0 21.01 39.50
IN 0.5 0 0.72 3.75 0.07 0.18 0 0 1 0 0 0 1 0 8.93 19.60
KS 1 1 0.57 3.67 0.18 0.08 0 0 2 0 0 0 1 0 14.62 23.92
KY 0.17 0 0.49 2.89 0.10 0.11 0 1 2 0 0 0 0 0 1.36 11.11
LA 0 0 0.75 3.64 0.16 0.13 1 0 2 0 0 0 0 0 1.75 9.05
MA 0 1 0.96 9.94 0.34 0.26 0 1 2 0 0.33 0 0 1 0.14 2.00
MD 0 0 0.93 6.70 0.21 0.17 0 0 2 0 0.33 0 0 1 3.65 6.18
ME 0.17 0.5 0.36 4.73 0.03 0.03 0 0 2 0 0 0 0 1 3.27 4.26
MI 0.67 0.83 0.82 4.27 0.06 0.30 0 0.33 2 0 0 0 1 0 2.48 12.14
MN 0.17 1 0.70 6.37 0.07 0.17 0 1 2 0 0 1 0 0 0.01 3.15
MO 0.17 0 0.68 4.50 0.05 0.16 0 0 1.33 1 1 0 1 0 8.46 21.91
0.16
The Market for Legislative Influence over Regulatory Policy

MS 0 0.67 0.36 2.45 0.06 1 0 1 1 0 0 0 0 1.26 12.33
MT 1 0.67 0.34 3.00 0.11 0.02 1 1 2 0 0 1 0 0 1.34 2.67
NC 0.33 0 0.67 5.51 0.06 0.24 0 0.33 2 0 0 0 0 0 8.37 25.21
ND 1 1 0.44 3.70 0.18 0.02 1 1 1 0 0 1 0 0 0.73 1.10
NE 0.5 1 0.52 4.46 N/A 0.05 1 0 2 0 0.67 1 0 0 0.81 2.27
NH 1 0.33 0.60 7.98 0.15 0.04 0 0.33 2 0 0 0 0 1 0.05 0.87
NJ 1 1 1 9.02 0.07 0.31 0 0 2 0 0.33 0 0 1 1.61 2.01
NM 0 1 0.57 3.50 0.09 0.05 0.33 0 0 0 0 1 0 0 7.93 12.51
NV 0.5 0.67 0.87 6.05 0.09 0.07 0 0 2 0 0 0 1 0 59.80 34.69
NY 0.5 1 0.92 13.62 0.13 0.74 0 0 2 1 1 0 0 1 4.64 2.83
OH 1 1 0.81 4.79 0.11 0.35 0 0 2 0 0.67 0 1 0 5.40 18.98
OK 0 1 0.61 2.92 0.10 0.08 1 0 2 0 0 0 1 0 1.12 10.89
OR 1 0.17 0.73 4.45 0.06 0.12 0 0 0 0 0 1 0 0 14.80 32.02
PA 1 1 0.85 5.44 0.05 0.37 0 1 1 0 1 0 0 1 4.55 10.59
RI 0 1 0.94 8.55 0.35 0.03 0 1 2 0 0 0 0 1 0 0.22
SC 0.67 0 0.70 3.51 0.06 0.10 1 0 2 0 0 0 0 0 3.72 15.71
SD 1 0 0.34 5.47 0.19 0.02 1 1 1 0 0 1 0 0 5.99 21.39
TN 0 1 0.68 4.08 0.07 0.16 0.33 1 2 0 0 0 0 0 2.86 10.72
TX 0.5 1 0.85 4.64 0.05 0.67 0 0 1 0 1 0 1 0 7.48 31.06
UT 1 1 0.77 4.93 0.21 0.06 0 0 1 0 0 1 0 0 11.62 6.80
VA 0.67 1 0.78 5.80 0.06 0.22 0 0 0 0.33 0.33 0 0 1 8.29 10.34
VT 0.17 0.5 0.28 4.83 0.07 0.02 0 0 2 0 0 0 0 1 1.41 0
WA 0.83 0 0.83 5.87 0.03 0.20 0 0 2 0 0 1 0 0 16.72 20.57
WI 0.5 1 0.68 4.58 0.05 0.16 0 1 2 0 0 0 1 0 0.82 6.01
WV 0 0.33 0.42 2.37 0.24 0.04 0 1 2 0 0.33 0 0 1 0 3.99
WY 1 1 0.30 4.09 0.21 0.02 0 1 1 0 0 1 0 0 3.39 19.45
Total 0.49 0.64 0.68 5.55 0.13 0.18 0.23 0.35 1.6 0.07 0.27 0.28 0.24 0.26 7.17 14.41
207
From Table 1, we note that on average, incumbent interests give a lot

more than entrants (roughly double in fact). There are many possible reasons
for this. One possibility is that entrants are more capital-constrained than
incumbents. Later in this section, we discuss one possible form of capital
constraint: entrants appear to enjoy less comprehensive or effective PAC
machines through which to organize and deliver contributions. Another
possibility is that, on average, propensity for the provision of legislative
services differ between incumbents and entrants. A third possibility is that
incumbents have cause to contribute on a wider range of issues than
entrants. For example, incumbents must not only consider the threat of
entry to their markets, but also the method by which their retail prices are
determined. Contributions from incumbents could in part represent
attempts to influence regulation of retail rates, among other things.29
Table 4 presents summary information on the patterns of contributions
from the three classes of contributors: corporations; associated PACs;
and associated individuals.30 All data in Table 4 are averages over the 145
state-cycles in our sample. Table 4 reveals significant differences in the
Table 4. Summary (Means) of Contributions to State Legislators

(Statewide Totals and per 1,000 Capita): from Entrant and Incumbent
Corporations, Associated PACs and Associated Individuals; and to
Democrats and Republicans.
Total Contributions ($) Contributions per 1,000
Capita ($)
Entrants Incumbents Entrants Incumbents
From
Corporations (direct) 31,395 25,698 4.55 3.99
(65.0%) (28.4%)
Associated PACs 14,756 61,450 2.62 9.00
(30.6%) (67.8%)
Associated individuals 2,119 3,461 1.90 1.41
(4.4%) (3.8%)
To
Democrats 23,806 43,321 3.89 6.23
(49.3%) (47.8%)
Republicans 24,310 46,988 5.11 8.07
(50.4%) (51.9%)
Other 153 301
(0.3%) (0.3%)
Total 48,269 90,610
sources of contributions on behalf of entrant and incumbent firms, but little

difference in their destinations. On average across our sample, entrant con-
tributions are predominantly direct from entrant corporations, whereas
most contributions on behalf of incumbent firms come from associated
PACs. Indeed, although total contributions on behalf of incumbent inter-
ests are roughly twice those on behalf of entrants, entrant firms give more
direct contributions than incumbent firms. Nearly two-thirds of entrant
contributions come direct from entrant corporations. Incumbents dominate
PAC giving, with incumbent-associated PACs giving more than four times
as much as entrant-associated PACs. This is likely a reflection of the larger
employee bases of the incumbent firms in most states. More than two-
thirds of incumbent contributions come from PACs.31 Contributions from
individual employees of the rival firms are relatively trivial in the data. As
noted above, for the purpose of the main analysis in this paper, we measure
contributions from entrant and incumbent interests as the sums of contri-
butions from all three sources, without distinction. Nonetheless, some
implications of the different sources of contributions on behalf of entrant
and incumbent firms can be derived from our results, and we discuss these
implications in the next two sections. Table 4 also presents the average split
in contributions from each interest to Democrat and Republican legisla-
tors. There is no clear evidence here that entrant firms give more to
Democrat legislators or that incumbent firms give more to Republicans.
On average, both interests divide their contributions more or less equally
between Democrat and Republican legislators (Republican legislators enjoy
a very slight advantage).
To test our hypotheses, we develop measures of the dominant party
ideologies in the state legislatures and regulatory commissions, relevant
constituency characteristics, the intensity of electoral competition, and rele-
vant institutional features. We measure the dominant party ideology of the
legislature as a categorical variable coded zero if both houses are
Democrat-controlled, one if Republicans control both houses, and 0.5 if
the houses are divided, there is no clear majority in one house or the legisla-
ture is non-partisan (Nebraska).32 Regulatory commission ideology enters
the analysis as a categorical variable coded zero if the majority of commis-
sioners were Democrat, one if Republican, and 0.5 if the commission was
evenly divided or entirely composed of Independents.33 Our empirical
analysis includes an interaction term between these measures of legislative
ideology and regulatory commission ideology. We employ two measures of
constituency characteristics: percentages of state populations living in
metropolitan areas a measure of the urbanization of state populations;34
and per capita Gross State Product in the Financial, Insurance and Real
Estate sector (GSP in FIRE per capita) a measure of the strength of
business customer demand for telecommunications services.35 Highly
metropolitan states and states with stronger business demand should favor
greater competition and entry in telecommunications.36 The intensity of
electoral competition in state legislatures is measured as minus one multiplied
by the average of the absolute values of the current shares of Democrat seats
minus 0.5 in the upper and lower houses, respectively.37 This measure is very
similar to a Ranney Index.38 We multiply by minus one so that higher values
in our measure reflect closer electoral races.39 A dummy variable is included
for states that elect their regulatory commissioners.40 Eleven states elected
their commissions throughout our study period. Analysis of unconditional
correlations suggests a significant negative relationship between this variable
and our measures of entrant and incumbent contributions.41 Another dummy
measures whether there is a statutory prohibition on direct contributions to
legislators from corporations, and a categorical variable has been created
to represent the presence of upper limits on the size of corporate and PAC
contributions.42 This latter variable is coded zero where there are no limits,
one if contributions from either corporations or PACs are limited, and two
if contributions from both corporations and PACs are limited. In Table 3,
we can see that most states limit at least one of these sources, exceptions
being Illinois, New Mexico, Oregon, and Virginia. Two further dummy
variables denote the presence of a major entrant or incumbent firm
headquarter office in a state,43 and a final set of dummy variables code
the states into the respective regions of the four largest incumbent firms
(the Regional Bell Operating Companies (RBOCs): BellSouth, Qwest,
SBC, and Verizon).44
We can now rewrite Eqs. (1) and (2) in expanded form:45
Entrant contributionsi;t E E Incumbent contributionsit

E1 Legislative ideologyit
E2 Commission ideologyit E3 Interactionit
E4 Metropolitan populationit E5 GSP in FIREit
E6 Electoral competitionit
E7 Elected commissionit
E8 Prohibition on corporate contributionsit
E9 Limits on corporate and PAC contributionsit
E1 Entrant headquartersit Eit
3
Incumbent contributionsit I I Entrant contributionsit

I1 Legislative ideologyit
I2 Commission ideologyit
I3 Interactionit
I4 Metropolitan populationit
I5 GSP in FIREit
I6 Electoral competitionit
I7 Elected commissionit
I8 Prohibition on corporate contributionsit
I9 Limits on corporate and PAC contributionsit
I1 Incumbent headquartersit I2 Qwesti
I3 SBCi I4 Verizoni Iit
4
This model reflects our theoretical setup. Contributions from each interest
are modeled as a function of rival contributions. Contributions from each
interest are also functions of aspects of the political environment that feature
in our hypotheses (the dominant party ideologies in the legislature and the
regulatory commission and an interaction term between these; constituency
characteristics; and the degree of electoral competition) and the institutional
environment (elected commissions; and the existence of statutory prohibi-
tions and limits on contributions). In addition, contributions from entrants
are a function of the presence of a major entrant firm headquarter office in a
state; and incumbent contributions are determined by the presence of a
major incumbent firm headquarter office and the peculiar strategies of the
four major incumbent firms in their respective regions.
RESULTS
We estimate our model using both Ordinary Least Squares (OLS) and
Two-Stage Least Squares (2SLS) methods.46 Data is pooled over all three
electoral cycles. All estimations include GSP to control for effects of
state size, but estimated coefficients suggest, somewhat surprisingly, that
state size is not a significant determinant of per capita contribution
patterns.47 All estimations also include year dummies to control for nation-
wide differences between cycles. All standard errors reported in this paper
are consistent in the presence of heteroskedasticity using the Huber-White

robust covariance estimator (White, 1980).
Columns 1 and 2 of Table 5 present OLS estimates of Eqs. (3) and (4),
respectively. The results from these estimations are mostly supportive of
our hypotheses, although many coefficients in the entrant contribution
equation lack statistical significance at conventional levels.
First, we find strong evidence in both equations that interests respond
positively to increasing contributions from rivals, consistent with our first
hypothesis that predicts defensive and responsive contribution strategies.
Second, for the incumbent contribution equation, coefficients on legisla-
tive ideology, commission ideology, and the interaction term between
these reveal that incumbents give more when the regulatory commission is
controlled by Democrats, but only to Republican legislatures, consistent
with both hypotheses concerning ideology.48 For the entrant contribution
equation, however, we report insignificant coefficients on each of the
ideology variables.
Third, all four coefficients on our two measures of constituency
characteristics are signed as predicted by our hypothesis concerning the
propensity to provide legislative services. Entrant contributions tend to
respond positively to higher metropolitan populations and GSP in FIRE
per capita. Conversely, incumbent contributions are lower in states with
constituencies more favorable to competition and entry in local telecommu-
nications. In terms of statistical significance, only the coefficient on GSP in
FIRE in the incumbent contribution equation is significant at conventional
levels. The remaining three coefficients nonetheless provide weak additional
support for this hypothesis.49
Fourth, testing for the effects of electoral competition returns a result
contrary to our predictions. While we find that entrants tend to give
more in closely contested electorates (although again, this coefficient is not
statistically significant), we find that incumbents appear to give less. One
possible explanation for this unexpected result draws again on the concept
that contributions depend on the relative costs and thus willingness for
legislators to supply legislative influence, but in a manner that we did not
at first predict. According to Denzau and Munger (1986), more informed
voters increase the willingness to provide for policies that voters favor, and
vice versa for policies that voters oppose. We imagine that greater electoral
competition will cause voters to become better informed on issues that
concern them, including telecommunications prices and the extent of com-
petition and entry in the industry. The majority of voters in most states are
likely to benefit from greater competition and entry and lower prices for
Table 5. Ordinary Least Squares and Two-Stage Least Squares Panel Estimations of Entrant and
The Market for Legislative Influence over Regulatory Policy

Incumbent Contributions ($ per 1,000 capita) to State Legislators.
Ordinary Least Squares Two-Stage Least Squaresa Two-Stage Least Squaresb
Entrant Incumbent Entrant Incumbent Entrant Incumbent

contributions contributions contributions contributions contributions contributions
(1) (2) (3) (4) (5) (6)
Incumbent contributions 0.311*** 0.389*** 0.372***

($ per 1,000 capita) (0.092) (0.108) (0.079)
Entrant contributions ($ per 0.488*** 0.575 0.722***
1,000 capita) (0.157) (0.811) (0.209)
Legislative ideology 1.466 21.521*** 3.055 21.171** 4.851 20.577***
(2.927) (7.244) (3.052) (8.254) (2.986) (6.262)
Commission ideology 1.453 3.459 1.569 3.531 2.232 3.654
(2.321) (3.664) (2.249) (3.345) (2.170) (3.460)
Leg Id Com Id 4.654 18.864** 6.192* 18.795*** 6.444* 18.679***
(3.276) (7.383) (3.619) (6.934) (3.415) (6.659)
Metropolitan population (%) 9.530 9.685 9.475* 10.258 8.511* 11.228*
(5.795) (6.683) (5.372) (7.983) (4.693) (6.790)
GSP in FIRE ($ thousands 0.375 0.784** 0.530 0.854 0.921** 0.971**
per capita) (0.283) (0.378) (0.341) (0.704) (0.413) (0.391)
Electoral competition 10.052 29.688*** 11.596 29.677*** 11.375 29.659***
(8.075) (10.844) (7.699) (9.952) (7.107) (9.898)
Gross state product 8.340 1.483 9.279 2.282 7.392 3.636
($ trillions) (7.010) (6.532) (6.464) (11.002) (5.718) (6.506)
Elected commission 3.310** 7.246** 2.662* 6.632 2.684* 5.592**
(1.452) (2.796) (1.459) (6.940) (1.399) (2.499)
Prohibition on corporate 4.491*** 1.538 4.231*** 1.884 4.040*** 2.471
contributions (1.269) (2.504) (1.273) (3.438) (1.120) (2.518)
Limits on corporate and PAC 0.350 5.216*** 0.068 5.056** 0.479 4.785***
213
contributions (1.171) (1.650) (1.167) (1.969) (1.055) (1.589)
Table 5. (Continued )
214
Ordinary Least Squares Two-Stage Least Squaresa Two-Stage Least Squaresb
Entrant Incumbent Entrant Incumbent Entrant Incumbent

contributions contributions contributions contributions contributions contributions
(1) (2) (3) (4) (5) (6)
Entrant headquarters 2.461 2.457 2.165
RUI J. P. DE FIGUEIREDO, JR. AND GEOFF EDWARDS

(1.640) (1.577) (1.620)
Incumbent headquarters 5.865** 5.946*** 2.848 6.082***
(2.357) (2.271) (1.887) (2.180)
Qwest 14.936** 14.764** 4.605** 14.471***
(5.965) (6.165) (1.937) (5.422)
SBC 4.610 4.854 4.702 5.267
(4.440) (4.180) (2.999) (4.298)
Verizon 16.110*** 15.336* 14.024***
(4.180) (9.020) (3.881)
Constant 2.613 23.920*** 3.938 23.617*** 5.416 23.104***
(4.000) (4.141) (3.452) (4.945) (3.682) (3.867)
Year dummies Yes Yes Yes Yes Yes Yes

Observations 142 142 142 142 142 142
Adjusted R2 0.30 0.50 0.29 0.50 0.32 0.49
First stage F-test of excluded F(4,121) F(1,121) F(2,120) F(2,120)
instruments = 10.92 = 0.59 = 65.50 = 25.61
Hansens J test of 13.96 N/A 0.44 0.03
overidentifying restrictions (0.003) (0.508) (0.866)
Robust standard errors in parentheses.

a
Excluded instrument for entrant contributions: entrant headquarters. Excluded instruments for incumbent contributions: incumbent
headquarters, Qwest, SBC, and Verizon.
b
Excluded instruments for entrant contributions: entrant headquarters and a rank instrument. Excluded instruments for incumbent
contributions: Verizon and a rank instrument.
*Significant at 10%; **significant at 5%; ***significant at 1%.
telecommunications services. As these voters become better informed, legis-

lators should be more disposed to performing services for entrant firms and
less disposed to assisting incumbents the costs of influence on behalf of
entrants will fall, and the costs for providing services to incumbents will
rise. If this is the case, entrants will contribute more and incumbents less as
electoral competition intensifies.
Finally, we find that both interests give fewer contributions when regula-
tory commissions are elected rather than appointed. This supports our
proposition that as elected commissions are less susceptible to influence from
legislators, influence services by legislators are less productive, and interests
consequently provide fewer contributions in exchange for legislative services.
In an industry context in which interests seek influence over regulatory policy,
elected commissions appear to provide some insulation from the legislative
avenue for such influence.
Turning to the controls in our analysis, we find that prohibitions on
corporate contributions have a significant impact on entrant contributions,
but no significant effect on incumbent contributions; while limits on corpo-
rate and PAC contributions clearly impede incumbent contributions
without affecting entrant contributions. We interpret these results having
regard to our preliminary summary of the sources of contributions from
the rival interests. Entrants, without established operations and employee
bases in most states, appear to lack PAC machines to match those of
the incumbent firms. To compensate, entrants rely very much on direct
corporate contributions. Prohibitions on corporate contributions therefore
significantly constrain the ability of entrant interests to contribute in those
states. Meanwhile, with contributions on behalf of incumbents on average
nearly double those on behalf of entrants, limits on corporate and PAC
contributions are binding on incumbent contributions, but do not appear
to limit contributions from entrants. We also report evidence consistent
with our expectations that incumbent contributions are higher in states in
which there is an incumbent firm headquarter office, and the different
RBOCs employ different contribution strategies between their respective
regions, with Qwest and Verizon in particular appearing to be less aggres-
sive contributors than the omitted RBOC, BellSouth.
In a simultaneous equation model, OLS estimates are likely to suffer
from simultaneity bias, as all factors that predict entrant contributions also
predict incumbent contributions (and vice versa). In the present context if
simultaneity does exist, the bias could go in either direction depending on
whether contributions crowd out or crowd in competing groups contribu-
tions. We therefore employ a 2SLS (Instrumental Variables) estimation
technique to control for simultaneity. In the first stage, we estimate contri-

butions from each interest as functions of all exogenous variables in the
model. Identification of the second stage structural equations in Eqs. (3)
and (4) requires that we include in each first stage (reduced form) equation
at least one instrument that has a non-zero coefficient in the first stage and
is excludable from the corresponding second stage. In other words, for
validity, we need instruments that are both strong (correlated with the
endogenous second-stage regressor) and excludable (uncorrelated with the
error term in the second stage). The natural choice for excluded instru-
ments are those built into our simultaneous equation model: entrant head-
quarters is an excluded instrument for entrant contributions in Eq. (4);
incumbent headquarters and the RBOC regions are excluded instruments
for incumbent contributions in Eq. (3). We then use the predicted values of
entrant and incumbent contributions from the first stage as independent
variables, along with the common and peculiar exogenous variables, to esti-
mate Eqs. (3) and (4) in the second stage. With valid instruments, this
method allows us to generate consistent estimates of the effects on each
interests contributions of exogenous changes in contributions by the rival
interest. Columns 3 and 4 of Table 5 report results from 2SLS estimations
of Eq. (3) and (4) using the modeled instruments.50
Before discussing results of these estimations, it is important to conduct
specification tests to assess the validity of our instruments. First, weak
excluded instruments lead to bias in estimated 2SLS coefficients (Bound,
Jaeger, & Baker, 1995; Staiger & Stock, 1997).51 The incumbent headquar-
ters dummy and the RBOC region dummies are, together, reasonably
strong predictors of incumbent firm contributions in the first stage (the first
stage F-test on these instruments is F(4,121) = 10.92, p-value < 0.0001).
Unfortunately, the entrant headquarters dummy is, on its own, too weak
to qualify as a valid instrument for entrant contributions (the first
stage F-test is F(1,121) = 0.59, p-value = 0.442). In the absence of a
suitable naturally occurring excludable instrument for entrant contribu-
tions, we re-estimate Eq. (4) supplementing the entrant headquarters
dummy with a rank-based instrument, following Evans and Kessides
(1993) and Kroszner and Stratmann (1998). We construct this instrument
by sorting the observations in our sample from lowest entrant contributions
to highest and assigning ranks (1, 2, and 3, respectively) to observations in
the smallest, middle, and largest thirds of the sample. By construction, this
instrument is correlated with entrant contributions (the first stage F-test with
the inclusion of the rank instrument is F(2,120) = 25.61, p-value < 0.0001).
Under a reasonable set of assumptions, the rank instrument is also orthogo-
nal to the error in Eq. (4), as shown by Wald (1940) and Koenker and
Bassett (1978).52 Results for the incumbent contribution Eq. (4) using this
alternative instrumentation method are reported in column 6 of Table 5.
Second, valid instruments must be uncorrelated with the error term in the
structural equation. As our model includes more than one excluded instru-
ment for incumbent contributions, we can perform an overidentification
test a test of the joint null hypothesis that the excluded instruments are
uncorrelated with the error term in the structural Eq. (3) and correctly
excluded from it. A rejection casts doubt on the validity of the excluded
instruments.53 For the entrant contribution equation estimated in column 3,
we reject the null hypothesis that all our excluded instruments for incumbent
contributions (incumbent headquarters and the RBOC regions) satisfy the
exclusion restriction at least one of these may not be valid.54 To control for
this concern, we construct another three level rank instrument, this time for
incumbent contributions. As discussed above, we are confident on theoretical
grounds that this instrument is both strong and excludable. Overidentification
tests of each of the originally excluded instruments with the rank instrument
suggest that of these, only Verizon is properly excluded from Eq. (3). We
therefore re-estimate Eq. (3) using only Verizon and the rank instrument as
excluded instruments, and report the results in column 5 of Table 5.
Columns 5 and 6 of Table 5 therefore present our preferred 2SLS estima-
tions of Eqs. (3) and (4), respectively.55 In the remainder of this section, we
discuss the results of these estimations and ignore the poorly specified 2SLS
estimations in columns 3 and 4. DurbinWuHausman specification tests56
of whether there are systematic differences in the coefficients in the OLS and
2SLS estimates report weak evidence of endogeneity. For example, for the
incumbent contribution Eq. (4), the DurbinWuHausman 2 statistic with
one degree of freedom is 2.24 with a p-value of 0.135.57 We would reject the
null hypothesis that the OLS estimation of Eq. (4) yields consistent estimates
at the 15 percent confidence level. We therefore consider that, while sacrifi-
cing some efficiency, it is prudent to prefer our 2SLS estimates to ensure
consistent estimates of the effects of rival contributions.
The results of our 2SLS estimations in columns 5 and 6 are similar to, but
generally stronger than our OLS results, and are again generally supportive
of our predictions. There are essentially four differences between the OLS
and 2SLS results. First, while the 2SLS results again confirm our first hypoth-
esis on matching contributions, the coefficient on entrant contributions in the
incumbent contribution equation is now much larger. It is interesting to us
that even after controlling for potential simultaneity bias, we find that
entrants and incumbents do not appear to match dollar for dollar and nor do
they defend equally. For each extra dollar of entrant contributions (per 1,000
capita), incumbents match with $0.72; but in response to an extra incumbent
dollar, entrants match with only $0.37. Similar results have been reported in
two previous studies of interest group competition.58
Second, while our 2SLS results again report that incumbents give signifi-
cantly more to aligned (Republican) legislatures when the regulatory commis-
sion is ideologically unaligned (Democrat), we now also find a statistically
significant positive coefficient on the interaction term in the entrant contribu-
tion equation. Consistent with our second hypothesis, this implies that
entrants give more when the regulatory commission is ideologically unaligned
(Republican). But contrary to our third hypothesis, when Republicans control
the regulatory commission, entrants prefer to purchase influence services from
Republican rather than Democrat legislatures. In other words, it would
appear that Republican legislatures are, for some reason, more willing than
Democrats to perform influence services for interests that find themselves
faced with unsympathetic regulatory commissions. While our findings are not
quite as expected (we find consistent support for our second hypothesis, but
conflicting results for our third) there is little doubt that the party ideologies
of both regulatory commissions and legislatures are important considerations
for the contribution strategies of both interests.
Third, our 2SLS estimates are generally more significant than our OLS
estimates. For example, all four coefficients on our measures of constitu-
ency characteristics are now statistically significant at the 10 percent level
or lower. And while the coefficient on electoral competition in the entrant
contribution equation is not significant at conventional levels, it is positive
as predicted, and the probability of a coefficient of this magnitude or
greater is just 0.11.
Finally, the entrant contribution equation now includes three additional
variables that were previously excluded from Eq. (3): incumbent headquar-
ters, Qwest, and SBC. The Qwest dummy is significant at the 5 percent level
and the incumbent headquarters and SBC dummies are not convincingly
close to zero.59 This helps us understand a little better the rejection of the
overidentification test in our first attempt at instrumentation in column 3:
it seems probable that each of these variables is properly included rather
than excluded from the structural equation for entrant contributions.
DISCUSSION AND CONCLUSION

We have extended the literature on strategic contribution behavior beyond
Congress, to examine patterns of contributions by competing interests to
state legislatures. And we have elicited and tested predictions of contribu-

tion behavior from the extant literature that are peculiar to a context in
which regulatory (rather than legislative) outcomes feature. Our empirical
context is the battle between opposed interests over regulatory policy in
the local telecommunications industry. With regulatory policy the main
focus of the interests in this industry, it might surprise some that the inter-
ests bother to contribute at all to candidates for state legislatures. A con-
sumption theory of contributions that interests contribute purely for the
satisfaction of political participation would predict that contributions
are mostly benign, varying mainly with income. An alternative theory
that the interests contribute in strategic fashion with a view to purchasing
influence by legislators over regulatory policy would predict that rival
contributions, the configuration of support for each interest in the domi-
nant party ideologies of the regulatory commission and the legislature, and
relative costs of legislative services will each be important determinants of
contribution patterns in this industry.
We report results generally supportive of our proposition that interests
in the local telecommunications industry contribute strategically to state
legislators with a view to purchasing legislative influence over decisions of
independent regulatory commissions. We demonstrate that interests contri-
bute in defensive patterns in response to contributions from rival interests.
We report that the dominant party ideologies of regulatory commissions
and state legislatures are important considerations for the contribution
strategies of both interests. And we show that contributions follow the
comparative advantages of legislatures in providing legislative services to
the rival interests. For example, the intensity of electoral competition drive
contributions from the competing interests higher, consistent with contribu-
tions being positively related to the willingness (predisposition) of legisla-
tors to perform influence services on behalf of those interests. And the
interests give less when regulatory commissions are elected and legislatures
are likely to be less productive in efforts to influence those commissions.
Most notable, perhaps, is that we find elements of the political and institu-
tional environment of telecommunications regulation to be significant
determinants of patterns of contributions to state legislators. In other
words, the interests act as if the regulatory environment matters for their
contribution strategies. We suggest this is compelling evidence that there
exists a competitive market for the purchase of legislative influence over
regulatory policy. Moreover, these results are suggestive of an important
normative element to our work, in the spirit of the literature discussed ear-
lier on non-market strategy. Following on the puzzles introduced by ADS,
our paper is suggestive of the possibility that influencing votes may be diffi-
cult but instead providing resources to legislators to influence actors down-
stream in the policy process (i.e., at the agency level) may be more fruitful.
In addition to tests of the strategic nature of contributions, our study
reports interesting results on our controls for campaign finance laws. We
find that alternative approaches to restricting contributions affect the two
main interests in the telecommunications industry in different ways, distort-
ing the competitive outcome in each case. Prohibitions on contributions
direct from corporations (mirroring the prohibition at Federal law under
the Federal Election Campaign Act of 1971) restrict the ability of entrant
interests to compete for legislative services, without significantly impeding
the flow of incumbent contributions. Conversely, limits on corporate and
PAC contributions appear binding on the contributions of incumbents
without constraining contributions from entrants. In short, a states choice
of campaign finance laws appears to do more than affect the total level of
contributions in the state different laws impact on the activities of some
interests more than others. These results have normative implications for a
states choice of campaign finance laws as well as more recent developments
at the federal level given Citizens United, and raise the positive question
whether private interest beneficiaries of the alternative laws are influential
in the choice of those laws.
Our analysis perhaps raises as many questions as it has been able to
answer. With respect to the empirical results, there are also a number of
open issues fur future research. One question arises out of our failure to
find consistent support for our third hypothesis on the effects of party
ideologies in the legislature. Our results imply that both sides of the indus-
try seek out Republican legislatures to provide influence services when the
ideology of the regulatory commission is unsympathetic. In a recent study
of Congressional votes on financial services legislation, Stratmann (2002,
p. 360) similarly found Republicans to be more responsive than Democrats
to changes in contribution levels. Future research might examine whether
Republican legislators are systematically more responsive than Democrats
to quid pro quo service provision in exchange for contributions. Another
possible explanation concerns a potentially important omitted variable in
our analysis that is the governorship. Because appointed elected officials
are most likely influenced by the governor, not controlling or potentially
interacting governors ideology creates faulty interpretation on the legis-
lative variables. Unfortunately because of the fact that in our dataset here
there are only three electoral cycles of which there were at most one
gubernatorial change per state, there was not sufficient variation to identify
such interactions between the governor and legislature. That said, in future
research, with longer panels it may be possible to disentangle these effects.
With regard to the theory of responsive contributions, the paper highlights
a need for future theory development as well as a number of open puzzles. At
a general level, the results in this paper highlight the need to more explicitly
develop an integrated model of legislative and executive politics (a la Baron,
2001) that provides sharper comparative statics when interest groups are
competing for influence of regulators. Such a model would incorporate the
executive, legislator and multiple interest groups with differential valuations
over policy outcomes (de Figueiredo, Jacobi, & Weingast, 2006). Although
no small task, the progress we have seen on the theory side in analyzing the
components provides an opportunity for such a valuable integration, which
in turn may provide the linkages between strategy, non-market outcomes and
firm performance envisioned by de Figueiredo et al. (2016). In addition, the
paper leaves for future research two persistent puzzles in the empirical litera-
ture on interest group competition that arose again in our analysis: why
defensive and responsive contributions do not appear to be dollar for dollar,
and why interests differ in their propensities to respond. We find
that incumbents are roughly twice as responsive to entrant contributions
as entrants are to contributions by incumbents. One possible explanation
worthy of further exploration is that defensive responses by entrants are
capital constrained, perhaps in part due to the absence of adequate PAC
machinery in many states. Another possibility is that there are not common
valuations of the outcomes leading to some but not perfect responsiveness.
Still another possibility is that defensive contributions to friends carry
different prices than other contributions.
Finally, as noted earlier, the paper provides an important extension of
ADS (2003) in that it does show that unlike in other contexts, in settings such
as ours where there is interest group competition between profit-motivated
actors in a regulatory domain there do appear to be important strategic
factors affecting behavior. But at the same time, our results show that despite
what may be important stakes the dollar values as ADS point out are quite
small, even in such a setting. This leaves open the question of why even in
these special cases, as ADS ponder, there is so little contributed.
NOTES
1. Studies that find in the affirmative, that contributions affect voting records,
include: Silberman and Durden (1976), Chappell (1981), Kau, Keenan, and Rubin
(1982), Coughlin (1985), Wilhite and Theilmann (1987), Langbein and Lotwis
(1990), and Stratmann (1991, 1995, 2002). Studies that find no effect of contribu-
tions on votes include: Chappell (1982), Evans (1988), Grenzke (1989), Wright
(1985, 1990), and Langbein (1993).
2. This claim is based on a finding that in three out of four instances in the
literature, coefficients on contributions were either insignificant or the wrong sign.
A conclusion on this basis alone is a little hasty. In counting the number of signifi-
cant coefficients out of all coefficients tested, the result is biased by several studies
that tested a very large number of coefficients. An alternative assessment of the lit-
erature would be to ask how many of the studies reported significant coefficients for
at least half of those tested. On this assessment, exactly half the studies support the
existence of an effect of contributions on roll call votes. An alternative conclusion
might be that the literature is evenly balanced on this question.
3. Examples of non-vote related support that legislators can provide to a contri-
buting interest include: influence over the form of legislation developed at the
committee stage through the drafting, amendment, or vetoing of bills; efforts to
negotiate with other legislators to win their support; efforts to rally popular support
through the media; intervention with bureaucrats; and the application of implicit or
explicit pressure on independent regulatory agencies to arrive at decisions favorable
to the interest (Hall & Wayman, 1990; Kroszner & Stratmann, 1998; Snyder, 1992).
4. Mobilization might include not simply a financial encouragement to act on an
interests behalf, but also the simultaneous provision of information and arguments
that the legislator requires to make a persuasive case in support of the interests pre-
ferred position. In this sense, contributions and the provision of information (lobbying)
go hand in hand, consistent with findings of a close correspondence between these two
non-market activities for corporate interests (Ansolabehere, Snyder, & Tripathi, 2002).
5. In addition to the empirical literature we focus on here, there has been more
recent theoretical work more directly linking contributions to legislators and regulatory
outcomes. Prominent examples include: Holburn and Vanden Bergh (2004), who
examine the allocation of resources by interest groups across multiple institutional
options (e.g., legislature or administrative agency) depending on the configuration of
preferences among public officials and interest groups; de Figueiredo and de
Figueiredo (2002) present a model which considers how interest groups allocate
resources around administrative rulemaking given the shadow of the courts; and
Gordon and Hafer (2005) develop a model of strategic signalling in which contribu-
tions provide credible information to agencies about interest groups willingness to con-
test rulemaking activities.
6. In addition, Grier, Munger, and Roberts (1991, 1994) report that industry
structure is an important determinant of whether and how much industries contri-
bute, and Tripathi (2000) finds that the size of the governments defense budget is
an important driver of political activity in the defense industry.
7. See also Stratmann (2002) and Bombardini and Trebbi (2011).
8. These results are consistent with theoretical models of interest group competi-
tion for public policy (Baron, 2001; Bernheim & Whinston, 1986; Grossman &
Helpman, 1994). These models predict that the interests play a prisoners dilemma
game, with each interest preferring to give less, but forced in equilibrium to choose
its non-market strategy in a defensive fashion, to counteract the non-market activ-
ities of its rival.
9. This follows a substantial literature on the theory of legislative control of

regulatory policy (Calvert, McCubbins, & Weingast, 1989; de Figueiredo, 2002; de
Figueiredo, Spiller, & Urbiztondo, 1999; de Figueiredo & Vanden Bergh, 2004;
Fiorina, 1979; McCubbins, 1985; McCubbins, Noll, & Weingast, 1987, 1989;
McCubbins & Schwartz, 1984; Vanden Bergh & de Figueiredo, 2003; Weingast, 1984;
Weingast & Moran, 1983). This theory proposes that, even if less than perfect, legisla-
tures can exert influence over regulators using coercive mechanisms that include proce-
dural requirements, oversight, and budgetary and appointment decisions. Oversight
can be direct (e.g., through committee hearings in which the regulator must demon-
strate in a transparent manner that it has properly exercised its mandate) or indirect
(e.g., through interest group feedback to legislative committee members).
10. The Telecommunications Act of 1996, and the First Report and Order of the
Federal Communications Commission (FCC), confers on independent regulatory
commissions in each state the responsibility for a wide array of decisions with the
potential to substantially determine the prospects for entry and competition in local
telecommunications, and the profitability of incumbent and entrant firms. This
includes settling the terms and conditions on which incumbent firms must provide
interconnection, resale services, and unbundled network elements.
11. Support for these alignments abounds. For example, see Teske (1991) and de
Figueiredo and Edwards (2007).
12. Empirical confirmations of a positive relationship between contributions and
the predisposition of Congressional legislators to support particular interests (implying
a negative relationship with supply prices for legislative services and effort) include:
Jacobson (1980), Chappell (1982), Poole et al. (1987), Grier and Munger (1986, 1991,
1993), Stratmann (1991, 1992, 1995), and Kroszner and Stratmann (1998).
13. In our current research, we examine contributions to state legislatures rather
than individual legislators. We therefore proxy for ideology using the ideology of the
dominant party in the state legislature. Ideology has long been considered determina-
tive of contribution patterns from rival interests, whether measured using roll call vote
scores for individual legislators (e.g., Americans for Constitutional Action ratings or
Chamber of Commerce vote scores) or party dummies. For example, in the battle
between corporate interests and labor unions, Chappell (1982) and Grier and Munger
(1986) report, respectively, that corporate interests tend to give more to conservative
and Republican legislators, while unions tend to give overwhelmingly to liberals and
Democrats. See also, Grier and Munger (1991) and Stratmann (1996). It is also worth
noting that we examine a linkage which is implicit in the sense of regulators ultimately
determining policy at the behest of the legislature. As noted earlier, this linkage relies
on a well-grounded literature on political control of the bureaucracy by elected
officials (e.g., legislators and/or executives). For a deeper discussion on this topic
and potential microfoundations, respectively, see de Figueiredo and Edwards (2007)
and Holburn and Vanden Bergh (2004, 2008).
14. A large number of studies demonstrate the effects of constituency characteris-
tics on contribution patterns. Contributions are consistently greater to legislators
with constituencies that are supportive of the contributing interests position. For
example, see Chappell (1982), Poole et al. (1987), Grier and Munger (1986, 1991,
1993), Stratmann (1991, 1992, 1995), and Kroszner and Stratmann (1998).
15. We expect that urban constituencies will favor greater entry and competition
in local telecommunications, as competition places pressure for the rebalancing of
retail prices that have traditionally seen urban constituents subsidize rural constitu-
ents. Rural constituencies will be less enamored by the supposed benefits of entry and
competition most entry will take place in urban areas, and the subsidies that rural
constituents have traditionally enjoyed will be at threat.
16. For example, see Jacobson (1980), Chappell (1982), Poole et al. (1987), Grier
and Munger (1986, 1991, 1993), and Stratmann (1991, 1992). Notably, a maintained
assumption here is that demand for services is sufficiently elastic such that aggregate
contributions are increasing with supply shifts.
17. Existing literature suggests that legislators with greater productivity in providing
legislative services (e.g., through membership on a relevant committee or some
leadership position) will have lower supply costs and receive greater contributions
(Denzau & Munger, 1986). For confirmatory empirical evidence of this relationship,
see Chappell (1982), Poole et al. (1987), Hall and Wayman (1990), Grier and Munger
(1986, 1991, 1993), Stratmann (1991, 1992, 1995), Romer and Snyder (1994) and
Kroszner and Stratmann (1998), but see Gopoian (1984) and Wright (1985, 1990).
18. Appointments are typically by the governor, with legislative assent.
19. Theoretically, Snyder and Weingast (2000) among others have shown that
political appointees to regulatory commissions are highly responsive to the chief
executive and relevant legislative chambers. In contrast, when the regulator is
elected, they are more responsive directly to constituents. Empirically, the hypothesis
proposed here is consistent with the empirical evidence for this institutional setting
examined in de Figueiredo and Edwards (2007).
20. We expect that the presence in a state of a headquarter office of a major firm
will increase contributions by virtue of the number of interested employees resident
in that state and likely to contribute to PACs that are active in that state.
21. These firms are the Regional Bell Operating Companies (RBOCs). In 1985,
US West (now Qwest) was the most politically active of the (then) seven RBOCs,
requiring Teske (1991) to include a special dummy for the presence of this company
in his study of the determinants of state regulatory policy. The political strategies of
the (now) four RBOCs have likely changed significantly since 1985, and while we
expect variation in strategies persists, we do not presume that Qwest is the most
active in our study period.
22. In robustness tests, we found that measures of per capita income (GSP per
capita, and average disposable income) were insignificant as determinants of contri-
bution patterns from the telecommunications industry to state legislators. Although
individuals, through PACs, make a large amount of contributions in this context,
insignificant coefficients on measures of per capita income suggest that consumption
value theory (ADS, 2003) is not predictive of contribution patterns in this context.
We also tested for an effect of the size of state legislatures (expecting that larger
legislatures receive more contributions) but did not find a significant effect.
23. Five states Kentucky, Louisiana, Mississippi, New Jersey, and
Virginia have odd cycles (cycles that conclude in odd years) and we account for
this in the data construction. Excluding these states does not significantly alter our
results. Also given its unicameral legislature we exclude Nebraska in our regression
results which explains the total observations of 142 in the tables that follow.
24. This unit of analysis is necessary in order to study the effects of the political
and institutional environment of utility regulation on contribution patterns.
Using state-cycles permits the examination of variables that are not usually included
in studies of contribution patterns, such as regulatory commission ideology and
whether the regulatory commission is elected or appointed. Unfortunately, the
choice of this unit of analysis involves a trade off, and precludes us from studying
here the effects of individual legislator characteristics, such as committee member-
ship and seniority.
25. A detailed description of the Institutes process for collecting and entering
data into its database is available at http://www.followthemoney.org/Institute/
about_data.phtml
26. No data was available for five states in the 1997/1998 cycle. This reduced the
number of available observations for the study from 150 to 145. In the later regres-
sion analysis, we also exclude the non-partisan Nebraska legislature reducing the
observations in those models to 142.
27. While the Federal Election Campaign Act of 1971 prohibits corporate contri-
butions direct to parties and candidates at the federal level, many states still permit
direct corporate contributions to candidates for state offices.
28. Entrant and incumbent corporations were identified using the FCCs
Telecommunications Provider Locator (October 2000 and February 2003 editions),
Hoovers Online, Internet searches, and other industry resources. Corporations
listed as incumbent local exchange companies (ILECs) were classified as incumbent
firms; corporations listed as inter-exchange companies (IXCs) were classified as
entrants. PACs were associated with entrant or incumbent corporations using PAC
names. Individuals were associated with entrant or incumbent corporations using
information on their employer provided in the contribution data from the Institute.
It is worth noting that we aggregate contributions into incumbents and entrants
in aggregate. In general, this poses little issue for the analysis of the incumbent
ILECs which were generally operating in non-overlapping geographic areas. For
IXCs it does meant that the maintained hypothesis is that aggregating these contri-
butions together is sensible on the basis that entrant interests were sufficiently
aligned in comparison to their alignment with the incumbent. In practice, this was
less of an issue since many of the IXCs were entering also in non-overlapping ways
during the period under study given the regulatory hurdles (i.e., single incumbent
entrant combinations in a state for some period).
29. In robustness tests, the method of retail rate regulation (price caps or rate of
return regulation) was not a significant determinant of contribution patterns.
30. While we consider Table 4 to be broadly indicative, we caution that identifying
PAC contributions separately from corporate contributions is, unfortunately, impre-
cise. The Institute gathers information on contributions by reviewing contribution
disclosure filings by each candidate for state office, and there is unavoidable variability
in the description of contributors in these filings. In particular, it is possible that some
contributions attributed as direct from corporations were actually provided by PACs
associated with those corporations.
31. Two paired t-tests comparing the means of percents of contributions in each
state-cycle (1) direct from entrant firms and direct from incumbent firms, and
(2) from PACs on behalf of entrant firms and from PACs on behalf of incumbent
firms, confirm that the sources of contributions on behalf of entrant and incumbent
firms vary systematically as described in the text.
32. This measure is similar to Teske (1991). An alternative approach to measur-

ing legislative ideology was tested for robustness. Two dummy variables were
created one for Democrat control of both houses and another for Republican
control of both houses. Substituting these alternative measures makes little
difference to our results and does not add additional insight. Data on the party
composition of the state legislatures is from annual editions of the US Census
Bureaus Statistical Abstract of the United States (19972003).
33. Data on the party ideologies of the regulatory commissioners were obtained
from annual membership directories of the National Association of Regulatory
Utility Commissioners (1996, 2000, 2001, and 2002).
34. Data is from the US Census Bureaus Statistical Abstract of the United States
(19972003). In an earlier study, Chappell (1982) similarly used urban population
percentage as a measure of constituency preferences.
35. The FIRE sector is a major consumer of telecommunications services (Teske,
1991). Data were obtained from the Regional Economic Accounts provided by the
US Department of Commerces Bureau of Economic Analysis (http://www.bea.gov/
bea/regional/gsp/).
36. Entry and competition will tend to occur first in high margin metropolitan
areas and in the provision of services to business customers, placing pressure on cross-
subsidies in incumbent firm retail pricing structures from metropolitan customers to
non-metropolitan customers and from business customers to residential customers.
Metropolitan and business customers will favor greater entry and competition in the
expectation that this will not only drive down retail prices they face, but also enhance
service quality and expand product options available to them.
37. This is most accurately described as a measure of the intensity of electoral
competition in the prior electoral cycle, rather than the current one. Nonetheless, this
measure is preferred due to concerns of endogeneity when using a measure of the
intensity of electoral competition in the current cycle: electoral competition in
the current cycle is potentially partly determined by contribution levels. This measure
is less accurate as a measure of electoral competition than measures used in studies
where the units of analysis are individual legislators. In those studies, the margin
of votes over 50 percent earned by a legislator is a good measure of the intensity of
electoral competition that legislator faces. In our study, our measure of electoral
competition is less than ideal, as the seatsvotes curve is likely to vary from state to
state: for example, 51 percent of the vote might translate into 51 percent of the seats
in some states, but 80 percent of the seats in others.
38. A Ranney Index of the intensity of party competition averages together the
proportion of seats won by Democrats in the state house and senate elections along
with the Democratic percentage in the gubernatorial election.
39. For example, where Democrats hold all seats in both houses, our measure of
electoral competition is 0.5. Where Democrats hold half the seats of both houses,
electoral competition is measured as zero. And where Democrats hold no seats in
either house, our measure of electoral competition is again 0.5. In our sample, the
least competitive state-cycle (0.393 by our measure) was Idaho in 2001/2002, in
which Democrats held only 7 of 70 lower house seats and just 3 of 35 senate seats.
The most competitive state-cycle (0.005 by our measure) was Washington in 2001/
2002, in which Democrats held exactly half (49) of the 98 lower house seats and
25 of the 49 senate seats.
40. Data on elected commissions is from annual editions of the Council of State
Governments The Book of the States (19962002).
41. In states with elected commissions, entrant (incumbent) contributions average
1.77 (9.45) per 1,000 capita. For states with appointed commissions, the average
value is 8.76 (15.87). The unconditional correlation coefficient between elected com-
missions and entrant (incumbent) contributions is 0.29 (0.17) with a p-value of
0.0004 (0.038).
42. Data on prohibitions and limits on contributions to candidates for state legis-
latures is from Feigenbaum and Palmer (1998, 2000, 2002).
43. Information on the location of major entrant (IXC) and incumbent (ILEC)
firm headquarter offices is from various editions of the FCCs Statistics of
Communications Common Carriers (19972002). This dataset records the locations
of headquarter offices only for firms with greater than $100 million in operating rev-
enues. Firms with less than $100 million in operating revenues are relatively trivial
in the context of the telecommunications industry, and we do not expect their
exclusion to bias results in any significant way.
44. In our estimations, the dummy for BellSouth is omitted to avoid collinearity.
45. For notational purposes, GSP has been omitted here, but is included in all
estimations. Data on GSP were obtained from the Regional Economic Accounts pro-
vided by the US Department of Commerces Bureau of Economic Analysis (http://
www.bea.gov/bea/regional/gsp/).
46. With dependent variables bounded below by zero, it is sometimes preferable to
perform a Tobit procedure rather than linear regressions. However, the lower bound
on contributions is unlikely to be quantitatively important in our data, as only a small
fraction of observations lie at this bound (entrant contributions were zero in 18 of 142
state-cycles; incumbent contributions were zero in just six state-cycles). In robustness
tests, we alternatively estimated Tobit models and found very similar results.
47. This is surprising as it suggests that states offering bigger potential profits do
not necessarily attract more political activity per capita (see, e.g., Tripathi, 2000).
48. Note that the interaction term between legislative ideology and commission
ideology almost completely offsets the main effect on legislative ideology, suggesting
that when Republicans control the regulatory commission, incumbent interests do
not discriminate between contributions to Democrat and Republican legislatures.
49. The coefficient on metropolitan population percent in the entrant contribu-
tion equation is positive as predicted, and the probability of a coefficient of this
magnitude or greater is just 0.103. The coefficient on GSP in FIRE in the entrant
contribution equation is also positive as predicted and would be significant at the
10 percent level on a one-sided test. Finally, the coefficient on metropolitan popula-
tion percent in the incumbent contribution equation is negative as predicted and
would also be significant at the 10 percent level on a one-sided test.
50. We report only the second-stage results of 2SLS estimations, for simplicity of
presentation.
51. A common rule of thumb is that, for a single endogenous regressor, an
F-statistic below 10 is cause for concern (Staiger & Stock, 1997).
52. If a change in entrant contributions does not alter the rank, then the rank is
independent of the error term. This condition will be violated only for observations
near the thresholds between the ranks, so we have chosen a small number of ranks
to reduce the likelihood of changes in ranks.
53. We test the exclusion restriction using Hansens J statistic (Hansen,

1982) a test of overidentifying restrictions that is robust to heteroskedasticity in
the errors.
54. Hansens J statistic (distributed as 2 with three degrees of freedom) is 13.96,
p-value 0.003.
55. The instruments used in these estimations appear both strong and excludable.
For the entrant contribution Eq. (3), the excludable instruments we use for incum-
bent contributions (Verizon and the rank instrument) also appear to be strong
instruments (the first stage F-test is F(2,120) = 65.50, p-value < 0.0001). For the
incumbent contribution Eq. (4), the instruments we use for entrant contributions
(entrant headquarters and the rank instrument) appear to be excludable (Hansens
J statistic with one degree of freedom is 0.03, p-value 0.866) as well as strong.
56. Durbin (1954), Wu (1973), and Hausman (1978).
57. For the entrant contribution Eq. (3), there is, surprisingly, no evidence of
endogeneity, even at the conservative 15 percent level. The DurbinWuHausman
2 statistic with one degree of freedom is 0.80 with a p-value of 0.3701. We cannot
reject the null of no endogeneity. This suggests that our OLS estimates of this
equation are consistent, assuming we have validly instrumented for incumbent con-
tributions in our 2SLS estimation.
58. In similarly constructed 2SLS estimations, Austen-Smith and Wright (1994,
pp. 4041) and Kroszner and Stratmann (1998, p. 1177) also find that interests
respond defensively to their rivals, but not dollar for dollar and not equally. In a
study of the lobbying behavior of organized interests involved in the confirmation
battle over Robert Borks nomination to the US Supreme Court in 1987, Austen-
Smith and Wright find that the interests lobbied predisposed legislators in a defen-
sive fashion (in response to lobbying from their rivals) but the response of one of
the interests was 1.34, more than counteractive, while the response of the other was
just 0.12. Kroszner and Stratmann report that in a competition between banking
PACs and securities PACs, banking PACs matched an extra dollar from securities
PACs with $0.80, but securities PACs matched an extra dollar of contributions
from banking PACs with only $0.38.
59. For the incumbent headquarters dummy, the probability of a coefficient of
2.848 or larger is just 0.13. For the SBC dummy, the probability of a coefficient of
4.702 or larger is just 0.12.
ACKNOWLEDGMENTS
The authors are grateful to Followthemoney.org for the provision of data;
and David Baron, Dino Falaschetti, Paul Gertler, Anne Karing, Orie
Shelef, Jason Snyder, Pablo Spiller, Richard Vanden Bergh and seminar
participants at the ISNIE conference and UC Berkeley Political Economy
seminar for helpful suggestions. All faults are solely the authors. The
author order is alphabetical and does not reflect relative contributions.
REFERENCES
Ansolabehere, S., de Figueiredo, J., & Snyder, J. (2003). Why is there so little money in U.S.
politics? Journal of Economic Perspectives, 17(1), 105130.
Ansolabehere, S., Snyder, J., & Tripathi, M. (2002). Are PAC contributions and lobbying
linked? New evidence from the 1995 public lobby disclosure act. Business and Politics,
4, 131155.
Austen-Smith, D., & Wright, J. (1992). Competitive lobbying for a legislators vote. Social
Choice and Welfare, 9(3), 229257.
Austen-Smith, D., & Wright, J. (1994). Counteractive lobbying. American Journal of Political
Science, 88(1), 2544.
Baron, D. (2001). Theories of strategic nonmarket participation: Majority rule and executive
institutions. Journal of Economics and Management Strategy, 10, 4789.
Bernheim, B., & Whinston, M. (1986). Menu auctions, resource allocation, and economic
influence. The Quarterly Journal of Economics, 101(1), 132.
Bombardini, M., & Trebbi, F. (2011). Votes or money? Theory and evidence from the US
Congress. Journal of Public Economics, 95(7), 587611.
Bound, J., Jaeger, D., & Baker, R. (1995). Problems with instrumental variables estimation
when the correlation between the instruments and the endogenous explanatory variable
is weak. Journal of the American Statistical Association, 90, 443450.
Calvert, R., McCubbins, M., & Weingast, B. (1989). A theory of political control and agency
discretion. American Journal of Political Science, 33(3), 588611.
Chappell, H. (1981). Campaign contributions and voting on the Cargo preference bill:
A comparison of simultaneous models. Public Choice, 36(2), 301312.
Chappell, H. (1982). Campaign contributions and Congressional voting: A simultaneous
Probit-Tobit model. The Review of Economics and Statistics, 64(1), 7783.
Che, Y., & Gale, I. (1998). Caps on political lobbying. American Economic Review,
88(3), 643651.
Coughlin, C. (1985). Domestic content legislation: House voting and the economic theory of
regulation. Economic Inquiry, 23(3), 437438.
de Figueiredo, J. M., & de Figueiredo, R. J. P., Jr. (2002). The allocation of resources by
interest groups: Lobbying, litigation and administrative regulation. Business and
Politics, 4(2), 161181.
de Figueiredo, J. M., Lenox, M., Oberholzer-Gee, F., & Vanden Bergh, R. G. (2016).
Introduction. Advances in Strategic Management, 34.
de Figueiredo, R. J. P., Jr. (2002). Electoral competition, political uncertainty, and policy
insulation. American Political Science Review, 96(2), 321333.
de Figueiredo, R. J. P., Jr., & Edwards, G. (2007). Does private money buy public policy?
Campaign contributions and regulatory outcomes in telecommunications. Journal of
de Figueiredo, R. J. P., Jr., Jacobi, T., & Weingast, B. R. (2006). The new separation of
powers approach to American politics. In The Oxford handbook of political economy.
Oxford: Oxford University Press.
de Figueiredo, R. J. P., Jr., Spiller, P. T., & Urbiztondo, S. (1999). An informational perspective
on administrative procedures. Journal of Law, Economics and Organization, 15, 283305.
de Figueiredo, R. J. P., Jr., & Vanden Bergh, R. G. (2004). The political economy of state-level
administrative procedure acts. Journal of Law and Economics, 47, 569588.
Denzau, A., & Munger, M. (1986). Legislators and interest groups: How unorganized interests
get represented. American Political Science Review, 80(1), 89106.
Durbin, J. (1954). Errors in variables. Review of the International Statistical Institute, 22, 2332.
Endersby, J., & Munger, M. (1992). The impact of legislator attributes on union PAC
campaign contributions. Journal of Labor Research, 13(1), 7997.
Evans, D. (1988). Oil PACs and aggressive contribution strategies. Journal of Politics,
50, 10471056.
Evans, W., & Kessides, I. (1993). Localized market power in the U.S. airline industry. Review
of Economics and Statistics, 75(1), 6673.
Feigenbaum, E., & Palmer, J. (1998). Campaign finance law 98: A summary of state campaign
finance laws with quick reference charts. Washington, DC: Federal Election Commission.
Fiorina, M. (1977). Congress: Keystone of the Washington establishment. New Haven, CT:
Yale University.
Fiorina, M. (1979). Control of the bureaucracy: A mismatch of incentives and capabilities. In
W. Livingston, L. Dodd, & R. Schott (Eds.), The Presidency and Congress: A shifting
balance. Austin, TX: LBJ School of Public Affairs.
Fremeth, A., Richter, B. K., & Schaufele, B. (2013). Campaign contributions over CEOs
careers. American Economic Journal: Applied Economics, 5(3), 170188.
Gopoian, J. (1984). What makes PACs tick? An analysis of the allocation patterns of economic
interest groups. American Journal of Political Science, 28(2), 259281.
Gordon, S., & Hafer, C. (2005). Flexing muscle: Corporate political expenditures as signals to
the bureaucracy. American Political Science Review, 99(2), 245257.
Grenzke, J. (1989). PACs and the Congressional supermarket: The currency is complex.
American Journal of Political Science, 33, 124.
Grier, K., & Munger, M. (1986). The impact of legislator attributes on interest-group
campaign contributions. Journal of Labor Research, 7(4), 349361.
Grier, K., & Munger, M. (1991). Committee assignments, constituent preferences, and
campaign contributions. Economic Inquiry, 29(1), 2443.
Grier, K., & Munger, M. (1993). Comparing interest group PAC contributions to House and
Senate incumbents, 19801986. The Journal of Politics, 55(3), 615643.
Grier, K., Munger, M., & Roberts, B. (1991). The industrial organization of corporate
political participation. Southern Economic Journal, 57(3), 727738.
Grier, K., Munger, M., & Roberts, B. (1994). The determinants of industrial political activity,
19781986. The American Political Science Review, 88(4), 911926.
Groseclose, T., & Snyder, J. M. (1996). Buying supermajorities. American Political Science
Review, 90, 303315.
Grossman, G., & Helpman, E. (1994). Protection for sale. American Economic Review,
84(4), 833850.
Hall, R., & Wayman, F. (1990). Buying time: Moneyed interests and the mobilization of bias
in Congressional committees. The American Political Science Review, 84(3), 797820.
Hansen, L. (1982). Large sample properties of generalized methods of moments estimators.
Econometrica, 50, 10291054.
Hansen, W., & Park, K. (1995). Nation-state and pluralistic decision making in trade policy:
The case of the international trade administration. International Studies Quarterly,
39, 181211.
Hausman, J. (1978). Specification tests in econometrics. Econometrica, 46(6), 12511271.

Herndon, J. (1982). Access, record, and competition as influences on interest group contributions
to Congressional campaigns. The Journal of Politics, 44(4), 9961019.
Holburn, G., & Vanden Bergh, R. (2004). Influencing agencies through pivotal political
institutions. Journal of Law, Economics and Organization, 20(2), 458483.
Holburn, G., & Vanden Bergh, R. (2008). Making friends in hostile environments: Political
strategy in regulated industries. Academy of Management Review, 33, 521554.
Holburn, G., & Vanden Bergh, R. (2014). Integrated market and nonmarket strategies:
Political campaign contributions around merger and acquisition events in the energy
sector. Strategic Management Journal, 35(3), 450460.
Jacobson, G. (1980). Money in congressional elections. New Haven, CT: Yale University.
Kau, J., Keenan, D., & Rubin, P. (1982). A general equilibrium model of Congressional
voting. The Quarterly Journal of Economics, 97(2), 271293.
Koenker, R., & Bassett, G. (1978). Regression quantiles. Econometrica, 46(1), 107112.
Kroszner, R., & Stratmann, T. (1998). Interest-group competition and the organization of
Congress: Theory and evidence form financial services political action committees.
American Economic Review, 88(5), 11631187.
Langbein, L. (1993). PACs, lobbies, and political conflict: The case of gun control. Public
Choice, 77(3), 551572.
Langbein, L., & Lotwis, M. (1990). The political efficacy of lobbying and money: Gun control
in the US House, 1986. Legislative Studies Quarterly, 15, 413440.
Macher, J. T., & Mayo, J. W. (2012). The world of regulatory influence. Journal of Regulatory
Economics, 41(1), 5979.
Macher, J. T., & Mayo, J. W. (2015). Influencing public policymaking: Firm-, industry- and
country-level determinants. Strategic Management Journal, 36(13), 20212038.
Mayhew, D. (1974). Congress: The electoral connection. New Haven, CT: Yale University.
McCubbins, M. (1985). The legislative design of regulatory structure. American Journal of
McCubbins, M., Noll, R., & Weingast, B. (1987). Administrative procedures as instruments of
political control. Journal of Law, Economics and Organization, 3, 243277.
McCubbins, M., Noll, R., & Weingast, B. (1989). Structure and process, politics and policy:
Administrative arrangements and the political control of agencies. Virginia Law
Review, 75, 431508.
McCubbins, M., & Schwartz, T. (1984). Congressional oversight overlooked: Police patrols
versus fire alarms. American Journal of Political Science, 28(1), 165179.
Poole, K., Romer, T., & Rosenthal, H. (1987). The revealed preferences of political action
committees. The American Economic Review, 77(2), 298302.
Romer, T., & Snyder, J. (1994). An empirical investigation of the dynamics of PAC contributions.
Silberman, J., & Durden, G. (1976). Determining legislative preferences on the minimum
wage: An economic approach. Journal of Political Economy, 84, 317329.
Snyder, J. (1990). Campaign contributions as investments: The U.S. house of representatives,
19801986. The Journal of Political Economy, 98(6), 11951227.
Snyder, J. (1991). On buying legislatures. Economics and Politics, 3, 93109.
Snyder, J. (1992). Long-term investing in politicians; or, give early, give often. Journal of Law
and Economics, 35, 1543.
Snyder, S. K., & Weingast, B. R. (2000). The American system of shared powers: The
President, Congress and the NLRB. Journal of Law, Economics and Organization,
16, 269305.
Staiger, D., & Stock, J. (1997). Instrumental variables regression with weak instruments.
Econometrica, 65(3), 557586.
Stratmann, T. (1991). What do campaign contributions buy? Deciphering causal effects of
money and votes. Southern Economic Journal, 57(1), 606620.
Stratmann, T. (1992). Are contributors rational? Disentangling strategies of political action
committees. Journal of Political Economy, 100(3), 647664.
Stratmann, T. (1995). Campaign contributions and Congressional voting: Does the timing of
contributions matter? Review of Economics and Statistics, 72(1), 127136.
Stratmann, T. (1996). How reelection constituencies matter: Evidence from political action
committees contributions and Congressional voting. Journal of Law and Economics,
39(2), 603636.
Stratmann, T. (2002). Can special interests buy Congressional votes? Evidence from financial
services legislation. Journal of Law and Economics, 45, 345373.
Teske, P. (1991). Interests and institutions in state regulation. American Journal of Political
Science, 35(1), 139154.
Tripathi, M. (2000). PAC contributions and defense contracting. Business and Politics,
2(1), 5373.
Vanden Bergh, R. G., & de Figueiredo, R. J. P., Jr. (2003). Political uncertainty and
administrative procedures. In B. Burden (Ed.), Uncertainty in American politics.
Cambridge: Cambridge University.
Vanden Bergh, R. G., & Holburn, G. (2007). Targeting corporate political strategy in pivotal
political institutions: Theory and evidence from the US accounting industry. Business
and Politics, 9.
Wald, A. (1940). The fitting of straight lines if both variables are subject to error. Annals of
Mathematical Statistics, 11, 284300.
Weingast, B. (1984). The Congressionalbureaucratic system: A principalagent perspective
(with applications to the SEC. Public Choice, 44(1), 147191.
Weingast, B., & Moran, M. (1983). Bureaucratic discretion or Congressional control?
Regulatory policymaking by the Federal trade commission. Journal of Political
Economy, 91, 765800.
Welch, W. (1974). The economics of campaign funds. Public Choice, 20, 8397.
White, H. (1980). A heteroskedasticity-consistent covariance matrix estimator and a direct test
for heteroskedasticity. Econometrica, 48, 817838.
Wilhite, A., & Theilmann, J. (1987). Labor PAC contributions and labor legislation: A
simultaneous logit approach. Public Choice, 53(3), 267276.
Wilson, J. Q. (1990). Bureaucracy. Cambridge, MA: Harvard University.
Wright, J. (1985). Political action committees, campaign contributions, and roll calls: An
organizational perspective. American Political Science Review, 79, 400414.
Wright, J. (1990). Contributions, lobbying, and committee voting in the U.S. house of
representatives. American Political Science Review, 84, 417438.
Wu, D. (1973). Alternative tests of independence between stochastic regressors and
disturbances. Econometrica, 41(4), 733750.
PART II
PRIVATE POLITICS
CORPORATE REPUTATIONAL
DYNAMICS, PRIVATE
REGULATION, AND
ACTIVIST PRESSURE
Jose Miguel Abito, David Besanko and

Daniel Diermeier
ABSTRACT
We model the interaction between a profit-maximizing firm and an activist

using an infinite-horizon dynamic stochastic game. The firm enhances its
reputation through self-regulation: voluntary provision of an abatement
activity that reduces a negative externality. We show that in equilibrium
the externality-reducing activity is subject to decreasing marginal returns,
which can cause the firm to coast on its reputation, that is, decrease the
level of externality-reducing activity as its reputation grows. The activist,
which benefits from increases in the externality-reducing activity, can take
two types of action that can harm the firms reputation: criticism, which
can impair the firms reputation on the margin, and confrontation, which
can trigger a crisis that may severely damage the firms reputation. The
activist changes the reputational dynamics of the game by tending to keep
the firm in reputational states in which it is highly motivated to invest in

ISSN: 0742-3322/doi:10.1108/S0742-332220160000034008
235
236 JOSE MIGUEL ABITO ET AL.
externality-reducing activity. Criticism and confrontational activity are

shown to be imperfect substitutes. The more patient the activist or the
more passionate it is about externality reduction, the more likely it is to
rely on confrontation. The more patient the firm and the more important
corporate citizenship is to firms brand equity, the more likely that it will
be targeted by an activist that relies on confrontation.
Keywords: Private regulation; activists; campaigns; dynamic stochastic
game; Markov perfect equilibrium
INTRODUCTION
The regulation of economic activity is one of the main arenas of political
competition. The impetus for changes to regulatory regimes frequently origi-
nates with concerned citizens, often motivated by social or ethical concerns.
Traditionally, concerned citizens have used public institutions such as legisla-
tures, executive agencies, and courts to advance their agenda. In recent years,
however, many activists have concluded that public processes respond too
slowly and can be blocked too easily by special interests. In response they
have turned to private politics instead. Private politics refers to actions
by private interests such as activists and NGOs that target private agents,
typically firms, often in the institution of public sentiment (e.g., Baron, 2001,
2003; Baron & Diermeier, 2007; Feddersen & Gilligan, 2001; Ingram,
Yue, & Rao, 2010; King & Pearce, 2010; McDonnell & King, 2013). Issues
have included, among others, environmental protection, human rights,
discrimination, privacy, safety of employees and customers, endangered
species, and animal welfare testing. The activists explicit or implicit goal is
private regulation, that is, the voluntary adoption of rules that constrain
certain company conduct without the involvement of public agents.1 Michael
Brune, executive director of the Rainforest Action Network (RAN), a lead-
ing global activist group, commented that companies were more responsive
to public opinion than certain legislatures were. We felt we could create
more democracy in the marketplace than in the government (Baron &
Yurday, 2004). If successful, such strategies may lead to alternative, private
governance mechanisms. Examples are the Equator Principles or the
Sustainable Forestry Initiative.2 Private regulation is also particularly widely
used in cases where public institutions are missing or are underdeveloped.
Corporate Reputational Dynamics 237
One such example is the attempt to reduce the availability of conflict

diamonds, which are used to fund civil wars in West Africa.
The actions of activist groups center on a corporate campaign, the orga-
nizational framework for satisfying activists goals. In a campaign an acti-
vist group tries to affect the business practices of a target firm through a
combination of threatened harms and promised rewards (e.g., Baron &
Diermeier, 2007). Harm usually takes the form of damage to the companys
reputation, though more direct actions (e.g., disrupting certain operations)
are not uncommon. Similarly, rewards may come in the form of endorse-
ments that enhance a companys reputation.
Activists pursue different goals and use different tactics. While some are
quite radical and use confrontational means, others are more moderate and
use a combination of criticism and engagement such as letter writing and
shareholder resolutions (Baron, 2012). One such confrontational approach
is to try to create a reputational crisis that has a significant impact on the
companys image, as in the confrontation between the activist group
Greenpeace and Royal Dutch/Shell over the disposal of the Brent Spar oil
storage buoy (Diermeier, 2011). In that example Greenpeace occupied the
platform in the middle of the North Sea. Shell then decided to clear it using
water cannons, which led to a media storm and drops in sales of up to
40 percent across Northern Europe.
This paper focuses on modeling a campaign in a dynamic context
between a firm that cares (to some extent) about its image as a good citizen
and an activist that seeks to tarnish the firms image to advance its own
agenda. The firm can enhance its image by engaging in private regulation,
modeled here as voluntary activity aimed at curbing a negative externality,
above and beyond that required by public policy. The activist group can
target the firm in two ways. It can engage in an effort that counteracts the
firms efforts to improve its reputation through private regulation. Or it
can try to trigger a crisis that can drastically harm the firms reputation.
We model the interaction between the firm and the activist as a discrete-
time, infinite-horizon dynamic stochastic game involving a firm and an acti-
vist. Though we provide some analytical characterization of the Markov
perfect equilibrium of this game, we rely heavily on computational methods
because, as we show, the activists impact on firm behavior is complex and
subtle. In particular, we employ the homotopy method utilized in Besanko,
Doraszelski, Kryukov, and Satterthwaite (2010) to show how equilibrium
behavior is affected by the fundamentals of the model, including the effi-
cacy of the activist, the discount factors, and the returns to corporate citi-
zenship. The efficacy of the activist is affected by two factors: the saliency
of the activists campaign to call attention to the firms shortcomings (what

we call criticism) and the newsworthiness of the activists efforts to provoke
a crisis (what we call confrontation). Criticism harms the firms reputation
at the margin, while confrontation if it becomes newsworthy can dras-
tically hurt the firms reputation. Increases in salience (holding newsworthi-
ness constant) are shown to induce a substitution, in the long run, of
confrontation for criticism, while increases in newsworthiness (holding
salience constant) induces substitution of effort in the opposite direction.
Thus, fundamentally, criticism and confrontation are substitutes.
We further show that an activist with a higher discount factor that is,
a more patient activist tends to rely to a greater extent on confrontation
than one with a lower discount factor. Since the discount factor is driven,
in part, by the likelihood that the activist persists over time, our model sug-
gests that, all other things equal, crisis provocation is a tool more likely to
be used by a secure and well-funded activist than by one who may not be
around in the future. We also find that the firms long-run value declines
significantly as the activists discount factor increases. Moreover, a firm
with a higher discount factor is more susceptible to a crisis than a firm with
a lower discount factor and that the activist derives more long-run value
when the firm has a high discount factor than when it has a low discount
factor. Thus, the most dangerous adversary for a firm is a patient activist,
and the most inviting target for an activist is a patient firm.
Our paper focuses on the positive implications of the interaction between
a firm and an activist.3 As just noted, the model illustrates circumstances
under which an activist would tend to rely more on confrontation or criti-
cism, and thus it provides a positive theory of activist behavior. In addition,
the model helps resolve a puzzle: why do firms like Coca Cola, Cisco
Systems, or McDonalds firms with established brands and multi-billion
market capitalizations devote the same or increasing resources, year after
year, to voluntary efforts that address social problems such as carbon emis-
sions or obesity, even when it is hard to imagine how such activity could
make their very strong brands even stronger? It seems plausible that this
activity would reach a point of diminishing marginal returns, making
increased levels of it inconsistent with shareholder value maximization.
One way to resolve this puzzle is to invoke a theory of moral management
of the kind developed by Baron (2009a). Another way to resolve it is through
agency arguments: senior managers who authorize spending on corporate
citizenship do so to burnish their own private reputations, rather than
enhance shareholder wealth, and equilibrium contracting may be unable
to eliminate this form of perquisite consumption entirely. Our model, by
contrast, provides a different explanation for the puzzle that does not require
abandoning the assumption of shareholder wealth maximization. Specifically,
it suggests that in a modern media and communications environment, cor-
porate reputation is subject to both countervailing pressures and drastic
shocks that, at least to some extent, can be triggered by activists. These
pressures and shocks boost private regulation by keeping a firm in situa-
tions where the accumulation of additional reputational capital has signifi-
cant value. A potential social value of an activist, then, is to keep the firm
well below the point at which diminishing marginal returns would induce it
to scale back its voluntary activity.
The organization of the remainder of this paper is as follows. The next
section describes the model of competition between the firm and the acti-
vist. The following section presents the general conditions for equilibrium.
The section Computational Approach explains the computational
approach we employ and the baseline parameters we use in those computa-
tions. The section Analysis provides some analytical characterization of
the Markov perfect equilibrium and also presents the results of our compu-
tational analysis. The final section summarizes and concludes. Throughout
this paper, we distinguish between Propositions that are established
through formal arguments and Results, which either establish a possibi-
lity through a numerical example or summarize a regularity revealed
through a systematic exploration of the parameter space. Proofs of all pro-
positions are in the appendix.
THE MODEL
Model Structure
The basic structure of our model is one of competition between a firm and
an activist group. Put simply, the firm seeks to enhance its reputation for
corporate citizenship, while the activist takes steps to undermine that
image.4 We model this competition as an infinite horizon dynamic game.
The Firm
The firm produces a single product which has a demand curve qt = et pt,
where qt is quantity at time t, pt is price, and et is the strength of the firms
overall brand equity. We assume that brand equity is given by et e0 Rt ,
where e0 is a fixed component (determined by factors such as product
performance or design), and Rt is the firms reputation for corporate
citizenship (hereafter referred to as reputation). The importance of repu-

tation is captured by the parameter > 0.
Our
model 5is set in discrete time and has a discrete
6 state space
R 1; ; R . The firms initial reputation is R0 1; R . We assume
that Rt is influenced by activities of the firm and the activist group and
evolves according to the following stochastic process:

Rt F~ t A~t if ~t 0

Rt1 ~t 1 1
f1; ; maxfRt 1; 1gg if
where F~ t , A~t , and ~ t are random variables taking on values {0, 1}. As
Fig. 1 illustrates, the firms reputation for corporate citizenship evolves in
two ways. If ~ t 0, reputation evolves incrementally, moving up or down
by one unit depending on the realizations of A~t and F~ t (both of which we
discuss below). By contrast, if ~ t 1, the firms reputation can drop preci-
~
pitously. In particular, if t 1, reputation will fall to a particular value
between 1 and max{Rt 1,1} according to a uniform distribution. We char-
acterize this event as a crisis. A crisis can cause a firms reputation for citi-
zenship to take a potentially significant one-time hit, which equals R2t on
average. This formulation captures a bigger they are, the harder they fall
Rt+1 = R+1
Ft = 1
At = 0
Ft = 0
Rt+1 = R
Ft = 1
t = 0
At = 1
Ft = 0
Rt = R Rt+1 = R 1
Rt+1 = R 1
t = 1
Rt+1 = R 2
Rt+1 = 1
Fig. 1. ~
Stochastic Process for R.
property: firms with greater reputations for citizenship will, on average,

experience a greater absolute drop in reputation as a result of a crisis. This
is consistent with the observation that companies with the strongest reputa-
tions for citizenship tend to receive the greatest scrutiny by activists and the
media, and thus seem to have the greatest vulnerability in the event of a cri-
sis.7 An alternative view (e.g., Alsop, 2004; Dowling, 2002; Minor, 2010) is
that firms that have invested heavily in building a reputation for citizenship
may have a bank account that can cushion the impact of the reputational
shock from the crisis. In the model this is captured by the feature that the
proportionate drop in a firms reputation is independent of R (on average it
is 50 percent). Thus, a firm with a strong reputation is cushioned to some
extent from the impact of a crisis.8
The firms production process creates a negative externality that is neither
taxed, priced in the market, nor regulated. The firm can positively affect its
reputation by voluntarily abating some of the externality, and we let xt denote
the level of the firms externality-reducing activity. The situation we are mod-
eling is especially pertinent to firms whose supply chains are located in parts
of the world where conventional policy interventions for negative externalities
are either ineffective or unavailable. As such, externality-reducing activity
occurs in the realm of what Baron (2003) and Baron and Diermeier (2007)
refer to as private politics, and we refer to it as private regulation.
Potential consumers do not observe xt, but xt generates an imperfect sig-
nal X~ t which, if it exceeds a threshold X0, creates awareness among consu-
mers that the firm is voluntarily taking steps to reduce an externality.9
From this, the firm gets credit for being a good citizen, which incrementally
enhances its reputation, that is, F~ t 1. We assume X~ t has a logistic distri-
bution with mean ln(xt) and scale parameter 1, so
xt
F xt Pr F~ t 1jxt
1 xt
and exp(X0) > 0. The function F() takes on values between 0 and 1
for all positive values of xt; it is strictly increasing and concave, and
approaches 1 as xt becomes arbitrarily large.
We assume that externality-reducing activity does not affect the perfor-
mance, quality, or appearance of the firms products. Therefore, although
the activity creates a direct social benefit, it has no direct consumer benefit
and thus does not enter the firms demand function. The provision of xt is
assumed to increase the firms total costs, which are given by cqt + kxt,
where c (0, e0) is the marginal cost of output and k > 0 is the marginal
cost of externality-reducing activity.
The Activist
Unlike the firm, the activist internalizes direct benefits from x. The social
benefit of x is denoted by w(x) and is given by:
8 1 w0
>
> w x w1 x2 x
>
< 0 2 w1
wx
>
> w20
>
: otherwise:
2w1
The activists private utility is given by u(x)=w(x), where 0 is a para-

meter that measures the activists passion for the social benefits created
by x. If > 1, the activist is so passionate that it over-internalizes the social
benefits of externality reduction.
Offsetting the firms efforts to build its reputation, the activist can harm
the firms reputation in two distinct ways. First, the activist group can
engage in criticism, denoted by z. Criticism comprises things such as letters
to editors, op-ed pieces, letter-writing campaigns, share-holder resolutions,
Facebook groups, and blogs that publicly call attention to the firms short-
comings and which may, therefore, counteract the firms attempt to burnish
its image through private regulation. In each period, criticism influences a
signal Z~t which can damage the image of the firm if it exceeds a threshold
Z0 beyond which the activists criticism penetrates the public consciousness.
Thus, A~t 1 if and only if Z~t Z0 . We assume that Z~t has a logistic distri-
bution with mean ln(zt) and scale parameter 1, so
zt
A zt Pr A~t 1jzt
1 azt
where exp(Z0) > 0.
Second, the activist group can engage in confrontation, denoted by d.
Confrontation is deliberately aimed at provoking a reputational crisis. In each
period, confrontation creates a level D~ t of potentially newsworthy negative
publicity about the firms activities. The publicity need not be accurate; what
matters is that it is potentially newsworthy enough to attract mass media atten-
tion. Once D~ t exceeds a newsworthiness threshold D0, the publicity blows up
and develops into a crisis.10 Thus, ~ t 1 if and only if D~ t D0 . We assume
that D~ t has a logistic distribution with mean ln(dt) and scale parameter 1, so
dt
dt Pr ~ t 1jdt
1 dt
where exp(D0) > 0.
Criticism and confrontation work through different channels. The for-

mer is more constructive and intended on changing business practices. It
frequently does not generate significant coverage by the mass media.
Religious organizations and pensions funds, for example, TIAA-CREF,
commonly pursue this approach, but it is also in the arsenal of many
activist groups (e.g., Eesley & Lenox, 2006). By contrast, confrontation is
intended to generate significant mass media coverage, critical to the firm.
The general idea is to create a spectacle through acts of civil disobedience
(e.g., occupying an installation or throwing a pie at the CEO), theatrical
protests (e.g., dressing up as a polar bear to protest global warming), and
other forms of confrontation (e.g., posting wanted posters in the CEOs
residential community).11 In his handbook for activists, San Francisco low-
rent-housing advocate Shaw (1996) summarized the approach as follows:
Ideally, tactical activists should use the media both to generate a scandal and then to
demand a specific, concrete result (p. 155).
The parameters and capture the efficacy of each type of activity.12

The parameter depends on the salience of the activists efforts to draw
attention to the firms shortcomings using nonconfrontational tools such as
blogs or op-ed pieces. It would thus be a function of the activists skill in
developing a persuasive narrative that counters the firms efforts to burnish
its reputation, as well as its effectiveness in mobilizing a community of fol-
lowers to disseminate that narrative. It would also depend on the salience
of the activists issue in a given market-place.13 By contrast, the parameter
depends on the mass media environment. For example, it would reflect
the extent to which mass media outlets are inclined to provide coverage
of the actions the activist group takes to provoke a crisis. The likelihood
that the media will provide such coverage may depend on many factors
such as the issue environment in a given country, the skill of the activists in
generating media coverage, and the structure of the media, for example, the
importance of state-owned media.14 In a world without mass media, we
have = 0. Given the distinction between these parameters, we refer to
as the salience parameter and as the newsworthiness parameter.
Both z and d are costly to the activist, and the activists cost function is
given by bzz + bdd, where bz > 0 and bd > 0 are constants. We normalize bd
by assuming that bd bd R R2 bz . This specification ensures that the cost
to the activist of obtaining one unit of reputation reduction through criti-
cism is equal to the cost of obtaining, on average, one unit of reputation
reduction through confrontation.15 We adopt this specification so as not to
bias the choice between the two activities solely due to differences in their
marginal costs. Any difference in the intensity of the activists critical and
confrontational activities will be due to differences in and or to the intrin-

sically different ways that the two activities affect reputational dynamics.
Comments on the Model Specification

1. The firms objective is the maximization of the discounted value of its
profits. The firm thus has no intrinsic preference for engaging in extern-
ality-reducing activity. It does so only to improve its reputation or to
blunt the effort of the activist.
2. The activist receives no intrinsic utility from harming the firms reputation:
it cares only about the level of x provided by the firm. In this respect, the
activist is pragmatic. Its benefit from harming the firms reputation is to
keep the firm motivated to supply higher levels of x. In practice, activists
may have ideological interests that translate into a direct utility for hurting
firms reputations or financial conditions. Still, the pragmatic activist
specification is plausible because we believe that to be effective an activist
must be pragmatic to some extent. The pragmatic activist model is also a
useful benchmark because it highlights the role of the activist as a strategic
player in the firms reputation management process. (By contrast, purely
ideological activists would merely be noise.)
3. R is assumed to be observable to the activist, and thus the activist can
condition its actions on it. This may appear to be a strong assumption.
Unlike a firms physical capital, installed base of customers, or cumula-
tive experience, R is not a standard metric that would be followed by
investment analysts. However, the media does give attention to firms
reputations for corporate citizenship (often in the form of rankings).
For example, for many years, Fortune has published a list of Americas
Most Admired Corporations, and one component of that ranking
(used in empirical work on corporate reputation) is Responsibility to
the community and environment.16 In addition, effective activists are
likely to be skillful at sensing public sentiment about companies and tai-
loring their efforts to that sentiment. Finally, tools from computational
linguistics and computer science provide technologies that enable indivi-
duals and groups to perform highly sophisticated content analysis of
media and analyst coverage of firms to determine how their public image
is being portrayed. In light of these considerations, the assumption that
R is observable to the activist strikes us as plausible.
4. The firm cannot take actions ex ante to reduce the likelihood of a crisis.
All the firm can do is to plug away and attempt to build its reputation
over time (which, as noted above, provides a cushion in the event of a
crisis). A hamburger restaurant chain, for example, can improve its
animal welfare standards but cannot give up serving meat entirely with-
out abandoning its business model. The inability of the firm to take
actions to reduce the likelihood of a crisis can be motivated by the fol-
lowing view of a crisis. A crisis is primarily a phenomenon that arises
within, and plays itself out, in the context of the media. Within that
realm, there are notable asymmetries between what activists and firms
can do to provoke or prevent a crisis. Activists may be able to draw
attention to problems that can provoke media scrutiny, but firms typi-
cally have less ability to influence the media narrative (Bond &
Kirshenbaum, 1998; Dennis & Merrill, 1996). This arises because good
news that a firm might want to highlight to prevent a crisis (e.g.,
Toyota solving problems with its accelerators) is typically less news-
worthy than bad news that an activist might highlight to trigger a
crisis (e.g., car crashes traceable to faulty accelerators).
5. The firm and the activist group are assumed to be unable to contract on
the provision of x, z, and d.17 In practice, of course, bargaining between
activists and firms sometimes does occur, but there are various reasons
why bargaining solutions may be infeasible. For example, some activist
groups may be unwilling to strike deals with firms lest their volunteers
or donors see them as selling out. This effect will be particularly pro-
nounced if the activist group competes in a market for donors or volun-
teers with other groups less willing to compromise. Such competition
may also make the enforcement of any agreement between a firm and an
activist group impossible.18
EQUILIBRIUM CONDITIONS
We restrict attention to the Markov perfect equilibrium (MPE) in which
the state
variable is the firms reputation
R. An MPE is a vector of strate-
gies x R; z R; d R; R R such that:

For each state R 1; ; R , x*(R) maximizes the discounted present
value of the firms expected
profits, given the activists strate-
gies z R; d R; R R .

For each state R 1; ; R , (z*(R), d*(R)) maximizes the discounted
present value of the activists expected utility, given the firms strat-
egy fx R; R Rg.
Firms Bellman Equation and KuhnTucker Conditions
With linear demand and constant marginal cost, the firms

per-period
profit
e R c
2 c
1
contribution in state R is 0 4 . We assume that 0; 22ec0 , which
2e0
implies that single-period profit is strictly concave in R.

Let VF R denote the present value of the firms expected profit in state
R in equilibrium. It is defined by the Bellman equation
2
e0 R c
VF R
4
8 2 8
939
>
> >
>F x 1 A R VF R 1
>
> >
>
>
> 6 >
> x
R x
R
R >
>7>
>
>
< 6 >
< 1 F
2 F V >
=7>
=
6

1 x
A
RV
R 1
A F
7
max kx F 6 1 R P
F A F 7
x0 >
> 6 >
> >
>7>
5>
maxfR1;1g
>
> 4 >
> VF r >
> >
>
>
: >
: R r1
>
; >
;
R1
2
where F 0; 1 is the firms discount factor, A R A z R, and
R d R. In writing this expression, it is understood that in state
R R, VF R 1 VF R and in state R = 1, VF R 1 VF R.
Straightforward algebra reveals that the firms continuation value (the
term in large square brackets in Eq. (2)) is a function of (among other things)
VF R VF R 1 and VF R 1 VF R. Following Cabral and Riordan
(1994), we refer to these differences as prizes. A prize is the increment to the
firms long-run value due to a one-step increase in its reputation and thus
represents the marginal benefit of reputation enhancement to the firm.
The KuhnTucker conditions for the firm are:
MBx k; x 0; MBx kx 0 3
where MBx is the firms marginal benefit of externality-reducing activity

given by:

F 1 R
MBx x; VF R; A R; RjR
1 x2

1 A R VF R 1 V
F R
4
A R VF R VF R 1

where VF R VF 1; ; VF R .19
The marginal benefit function, which is strictly decreasing in x, is the

firms demand curve for externality reduction: it shows how much
externality-reducing activity the firm purchases at price k. The activist
shifts this demand curve both directly (through and A) and indirectly,
through the impact of the activist on VF R.
Activists Bellman Equation and KuhnTucker Conditions
Analogous to the Bellman equation for the firm, the Bellman equation for
the activist gives us the present value of the activists utility in state R in an
equilibrium:
VA R max
z 0; d 0
8 9
>
>
> 2 8 ux R bz z bd Rd
93>
>
>
>
>
<

F R 1 A z VA R 1
=>>
>
>
>
< 6 1 d 1 F R A z 2F RA z VA R 7>
=
6 :
7
;7
6 1
x zV
R 1
>
> A 6 PmaxfR1;1g F A A 7>
>
> 6 7>
>
>
> 4 d r1 VA r 5>
>
>
>
: >
;
R1
5
where A 0; 1 is the activists discount factor; F R F x R,
and (analogous to before) it is understood that in state R R,
VA R 1 VA R and in state R = 1, VA R 1 VA R. The
KuhnTucker conditions are:
MBz bz ; z 0; MBz bz z 0 6
MBd bd R; d 0; MBd bd Rd 0 7
where MBz and MBd are the marginal benefits of criticism and confronta-
tion, respectively, and are given by:

A 1 d
VA R 1
MBz z; d; VA R; F RjR F R V A R

1 z2 1 F VA R 1 VA R
8
A
MBd d; z; VA R; F RjR
1 d2
8 9
> PR1 1
>
>
> r1
V r VA R >
>
< R1 A =

9
>
>
F R 1
>
A z VA R VA R 1
>
>
: 1 R z V R 1 V R ; >
F A A A
Like MBx, MBz, and MBd depend on prizes that result from changes in the
firms reputation.
Equilibrium Conditions

An MPE is a collection of five R 1 vectors VF ; VA ; x ; z ; d satisfying
the five sets of equilibrium conditions20 for each of the R values of R:
Eqs. (2), (3), (5)(7).21 The KuhnTucker conditions are complementary
slackness conditions, so for the computational analysis below, it is useful to
reformulate these conditions as a system of equations. To illustrate, con-
sider Eq. (3). Following Borkovsky, Doraszelski, and Kryukov (2010,
p. 1127), we can rewrite Eq. (3) as a pair of equations involving two vari-
ables, x and x

n
MBx k max 0; x 0 10

n
x max 0; x 0 11
where n N is a large positive integer. The system (10) and (11) can be
shown to be equivalent to Eq. (3) when
8 1
>
> n
>
< MBx k if MBx k < 0;
x 1
12
>
>
n
: x
> if x > 0;
0 if MBx k x 0
Moreover, conditions (10) and (11) are n 1 continuously differentiable

with respect to x and x. Transforming the other KuhnTucker conditions
in this fashion, let H(VA,VF, x, z, d|) denote the system of equations

defining an MPE where is a vector of parameters. An MPE VF ; VA ;
x ; z ; d ; x ; z ; d thus solves

H VF ; VA ; x ; z ; d ; x ; z ; d j 0 13

where H VF ; VA ; x ; z ; d j 0 is a system of 8R nonlinear equations
in 8R unknowns.22 Condition (13) forms the basis of the computational
analysis below.
COMPUTATIONAL APPROACH
Our objective is to develop a comprehensive intuition about equilibrium
interactions between a forward-looking firm and a forward-looking activist.
To do this, we rely on a partial analytical characterization of the MPE,
supplemented by computations of the MPE for a large set of parameter
values. This section sets the stage for the computational analysis.
Baseline Parameterization
Table 1 shows the baseline parameterization used to compute the show-

case equilibrium. While the baseline parameterization is not intended to
be representative of any particular industry, it is neither entirely unrepre-
sentative nor extreme. To put these parameters in perspective, we note that
the growth in the firms reputation for corporate citizenship can potentially
increase the firms brand equity from e = (150.25)100 197 at its initial
value of R = 15 to e = 100(300.25) 234, while brand equity could poten-
tially fall to e = 100 if R = 1. Given the baseline demand and cost para-
meters, a crisis in the initial state that crashed the firms reputation by the
expected amount would cause its per-period profit contribution to fall by
about 30 percent, while the worst possible crisis would cause per-period
profit contribution to fall by about 80 percent. To put this in perspective,
when Extra Strength Tylenol was implicated in six deaths in suburban
Chicago area in 1982, Tylenols sales dropped by about 87 percent (Lewin,
1982). In our model, a shift in the demand curve sufficient to cause an 87
percent decline in sales revenue (given an optimal pricing response pre- and
post-shift) would decrease the per-period profit contribution by about 98
percent. Thus, the worst possible crisis under our parameterization would
be on a par with a Tylenol-style reputational crisis.23
Table 1. Baseline Parameter Values.

Parameter Description Baseline Value
R Size of state space 30

R0 Firms initial reputation 15
c Marginal cost of output 20
k Marginal cost of externality-reducing investment x 100
w0 Intercept of marginal social benefit curve for x 125
w1 Slope of marginal social benefit curve for x 0.5
Activists passion 2
e0 Brand equity for firm with R = 1 100
Elasticity of brand equity with respect to reputation for citizenship 0.25
Salience parameter: externality-reducing activity 0.20
Salience parameter: criticism 0.20
Newsworthiness parameter: confrontation 0.20
bz Marginal cost of criticism note: bd R bz R2 150
F Firms discount factor 0.95
A Activists discount factor 0.95
i
The discount factor can be thought of as i 1r , where r > 0 is the per-
period discount rate and i (0,1] is the exogenous probability that the
agent survives from one period to the next. This interpretation is especially
relevant for the activist, who may operate on a very tight budget, and who
may suddenly disappear as a result of shocks to its funding. Consequently,
our baseline parameterization corresponds to a variety of scenarios that
differ in the length of a period. For example, it corresponds to a period
length of one year, a yearly discount rate of 5.3 percent, and certain survi-
val. But it also corresponds, for example, to a period length of one month,
a monthly discount rate of 1 percent (which translates into a yearly
discount rate of 12.68 percent), and a monthly survival probability of 0.96,
which translates into an expected life span of about 26 months.
Computational Analysis
We perform two types of computational analyses. First, to generate insight

about possible regularities of the equilibrium, we compute equilibria over a
grid G of parameter values given by:
8 9
< ; ; ; ; A j f0:10; 0:20; 0:30; 0:40g; f0:10; 0:20; 0:30; 0:4g; =
G f0:5; 2:0; 4:0g; A f0:80; 0:95; 0:99g; F f0:80; 0:95; 0:99g;
: ;
f0:001; 0:15; 0:25; 0:35; 0:40g
where in defining G it is understood that all other parameters are fixed at

baseline levels. The grid is designed to determine how the equilibrium varies
as we vary all the parameters (, , , A) that determine the activists
incentives, as well as the parameter that determines the returns to the
firm from corporate citizenship.
Second, we change key parameters on a one-at-a-time basis to isolate
how each parameter affects the equilibrium. Any parameter not varied is
set at its baseline level: The parameters varied are:24 (1) salience of criti-
cism: [0, 0.40]; (2) newsworthiness: [0, 0.04]; (3) activists discount
factor: A [0, 0.999]; (4) firms discount factor F [0, 0.999]; (5) returns
to corporate citizenship: [0, 0.40]; activist passion: [0.5, 4]. The
method used for these computational exercises is described in the appendix.
ANALYSIS
Equilibrium Behavior with no Activist
To establish a benchmark, we describe the outcome when there is no activist.

This corresponds to the case in which = = 0.
Proposition 1. In the absence
of an activist,
the firms externality-reducing
effort and value function x R; VF R are found by solving the follow-
ing system of equations for (VF, x) recursively:

x R 0
h
i2
e0 R c
VF R
4 1 F
and for R < R

" 1 ! #
1 F
2
x max VF R 1 VF 1 ;0 14
k

2
e0 R c kx2
VF 15
4 1 F 1 F

For any R 1; ; R 1 , the firms value function is strictly increasing
and strictly concave in R, that is, VF R 1 > VF R and VF R 1 VF R <
VF R VF R 1. The firms level of externality-reducing activity is non-
increasing in R, that is, x*(R + 1) x*(R), and the inequality is strict in any
state in which x*(R) > 0.
Proposition 1 implies that in the absence of an activist, reputation
enhancement is valuable to the firm, but it is subject to diminishing mar-
ginal returns. The firm thus reduces its externality-reducing activity as its
reputation grows, that is, it coasts on its reputation. Because Proposition
1 holds for an arbitrary end state, R, which can be made arbitrarily large,
the concavity of the value function is attributable to fundamentals (princi-
pally, the concavity of single-period profit in R), not to end effects due to
a finite R.
Fig. 2 shows the equilibrium in the no activist case for the baseline para-
meter values. In this case, x*(R) (depicted in the middle panel) decreases
monotonically from about 15 in the lowest state of R = 1 0 in R = 30.
Given the assumed parameter values, this implies that there is a 0.58 prob-
ability of reputation growth in the initial state R = 15, but this declines
over time as the firms externality-reducing activity diminishes. This process
of reputation growth can be shown to take about 40 periods on average.
V*F (R) x*(R) p*(R)

x105
2.4 15 130
2.2 120
2 110
10
1.8 100
VF
p
x
1.6 90
5
1.4 80
1.2 70
1 0 60
0 10 20 30 0 10 20 30 0 10 20 30
R R R
Fig. 2. No Activist Equilibrium: Baseline Parameter Values. Note: The three

figures show the equilibrium value function of the firm, its externality-reducing
activity, and product price as functions of firms reputation when there is
no activist.
As the firms reputation grows from R = 15 to R = 30, it is able to raise its

price by about 17.5 percent (from about 108 to 127, as depicted in the
right-hand panel), and its value grows by about 23 percent (from about
186,000 to 229,000, depicted in the left-hand panel). Thus, as the firms
reputation grows, it coasts, and eventually draws its externality-reducing
activity down to zero.
Equilibrium Behavior with an Activist
The Role and Impact of the Activist: A Preliminary Cut

The firms incentive to coast provides the basis for reputation-impairing
action by the activist. If the firm did not coast (either globally or
locally) that is, if x*(R) monotonically increased in R the interests of
the firm and the activist would be aligned. Both parties would benefit from
a growth in the firms reputation, and the activist would have no reason to
block the firm from increasing R. We show, though, that it in equilibrium
x*(R) cannot be monotonically increasing in R. Thus, even when the firm
faces an activist, it still coasts on its reputation to some extent. This, in
turn, provides a potential motivation for the activist to choose positive
levels of either criticism or confrontation.
Proposition 2. The firms equilibrium level of externality-reducing activ-
ity cannot be monotonically increasing in R; that is, there exists states
R and R + 1 such that x*(R + 1) x*(R). Thus, the firm (weakly)
coasts on its reputation in at least some states.
The activists behavior has a potentially complex set of effects on the
firms decision to invest in externality reduction. A useful starting point is to
perform the following thought experiment. Start with a situation in which
the activists criticism and confrontation are zero and then consider a small
exogenous increase in confrontation d in a single focal state Rn. This gener-
ates an exogenous perturbation in the probability n d Rn of a crisis
in the focal state, but keeps and A fixed at zero in all other states. As
the firm adjusts to this perturbation
optimally,
it will alter
the profile of
values, so VF R VF Rjn VF 1jn ; ; VF Rjn . This one-state
perturbation, though far simpler than what actually happens in equilibrium,
provides a relatively clean way to isolate how confrontation shifts the
firms demand curve for externality-reducing activity at various points in the
state space.
With A held at zero, the marginal benefit of externality reduction in the

focal state Rn becomes:
F
MBx x; VF Rjn ; n jRn 1 n VF Rn 1jn VF Rn jn

1 x 2
and thus

dMBx x;VF Rj n ;n jRn F

dn 1 x2
8
9
< VF Rn 1j j
n n
>
V F R n
>
=
1 n VF Rn 1j VF Rn j
n n
16
>
: >
;
n
In all other states R Rn, the marginal benefit of externality reduction is
F
MBx x; VF Rjn ; n jR VF R 1jn VF Rjn

1 x2
and thus

dMBx x; VF Rjn ; n jR F VF R 1jn VF Rjn

17
dn 1 x2 n
In the focal state Rn, the firms demand curve depends on n in two
ways. First, an increase in n directly decreases the marginal benefit of x in

state Rn. Holding the prize VF Rn 1jn VF Rn jn fixed, this effect

would unambiguously shift the demand curve leftward in state Rn in a man-
ner akin to the impact of ad valorem sales tax on a consumer demand
curve. We call this the direct effect of confrontation on the firms demand
curve for externality reduction.25 Note that there is no direct effect in states
other than Rn.
Second, an increase in n affects the prize itself, both in state Rn, as well
in other states R Rn. We call this the prize effect of confrontation, and it
V R1jn VF Rjn
is given by the sign of F n , which is not obvious. We gain

insight into it by performing a comparative statics analysis on n , allowing
n to go to 0 (which places us at the no-activist equilibrium). The following

lemma characterizes the prize effect as the perturbation in confrontation
goes to 0 in the limit.
Lemma 1. Starting from the no-activist equilibrium, consider a one-state
perturbation
n in
the probability of a crisis in state Rn < R: (a) In states
R Rn 1; ; R , as n 0, the prize effect for confrontation is zero;
(b) In the focal state Rn, as n 0, the prize effect is strictly positive;
and (c) in states R {1,, Rn 1}, as n 0, the prize effect is
strictly positive.
Lemma 1, in conjunction with Eqs. (16) and (17), leads immediately to:
Proposition 3. Starting from the no-activist equilibrium, consider a one-
state perturbation n in the probability of a crisis in state Rn < R. As
n 0, the impact of the perturbation on the firms externality-reducing
activity is as follows: (a) In states R > Rn, the perturbation has no effect;
(b) In the focal state R = Rn, the perturbation has an ambiguous effect
(a negative direct effect may or may not be offset by a positive prize
effect); (c) In states R < Rn, the perturbation has a positive effect (due to
the positive prize effect).
However, if the perturbation is large, the direct effect must dominate
the prize effect in the focal state.
Proposition 4. Starting from the no-activist equilibrium, consider a one-
state perturbation n in the probability of a crisis in state Rn < R, and
suppose that in this state, the firm would have invested a positive
amount in externality-reducing activity. As n 1, the direct effect in
state Rn dominates the prize effect and has an unambiguously negative
effect on the firms externality-reducing activity in state Rn.
Propositions 3 and 4 hint at the complexity of the activists role in shap-
ing the firms behavior. In equilibrium, single-state perturbations occur
in all states simultaneously, and none of these perturbations are necessa-
rily small or large. Moreover, they interact with perturbations in
criticism z, which was fixed in our thought experiment. Moreover, in equili-
brium the perturbations are themselves endogenous, so the firms equili-
brium decisions feedback and affect them. Finally, both the direct and
prize effects are static phenomenon in the sense that they relate to the
impact of the activist on the intensity of the firms incentives in particular
states. They do not speak to how the presence of the activist will change
the dynamics of how the firms reputation evolves over time.
Still, this analysis, limited though it is, provides a helpful insight: the
activists impact on the firms externality-reducing activity is not unambigu-
ously positive. That is, the presence of an activist may have (for the activist)
the unintended consequence (through the direct effect of confrontation) of
suppressing the thing that the activist wants, namely, an abundant supply
of x.
We can also conduct an analysis of a small perturbation in criticism,
yielding to a positive probability nA in focal state Rn. With held to 0,
the expression for marginal benefit (4) in the focal state is:
2 3
V F R 1jnA VF RjnA

4
MBx x; VF RjnA ; nA jR F
VF Rjn VF R 1jnA
5
1 x2 A V R A 1jn V Rj
n
n
F A F A

V F Rj

n
VF R 1jnA
The direct effect of A is given by the term

n A ,
VF R 1jnA VF RjnA
which depends on the concavity of the firms value function. Our computations
reveal that the firms equilibrium value function is not necessarily concave when
an activist is present. However, as nA 0,

A V F R 1jA
VF0 R VF0 R 1
n n
F Rj
V

VF R 1jnA VF RjnA VF0 R 1 VF0 R
and Proposition 1 established that this latter expression is non-negative.

This gives us the following result:
Proposition 5. In the focal state Rn, the direct effect of criticism for small
perturbations is non-negative.
Proposition 5 highlights that criticism and confrontation can have differ-
ent effects on the firms marginal benefit of externality reduction. The
proposition, to be sure, provides only limited insight (e.g., it says nothing
about the prize effect of criticism, or the direct effect for nonsmall
changes, both of which appear to be generally ambiguous). But like
Propositions 3 and 4, it points to the complex impact that the activist can
have on the firms equilibrium behavior.
Fig. 3 summarizes the implications of the preceding propositions.
Confrontation in a given state may shift the firms demand curve leftward
or rightward, depending on the strength of the direct and prize effects.
Confrontation in higher states unambiguously shifts the demand curve in a
MBx
Direct effect, small levels of critical activity in state R

Prize effect, small levels of confrontational activity in state R > R
Prize effect, small levels of confrontational activity in state R (possibly)
Direct effect, confrontational activity in state R Dx in state R

Prize effect, small levels of confrontational activity in state R (possibly)
x
Fig. 3. How the Effects of Activist Behavior Can Shift the Firms Demand
Curve for Externality-Reducing Investment.
given state leftward due to the prize effect. Small levels of criticism in a
given state have a direct effect of shifting the firms demand curve right-
ward in a given state, but an ambiguous prize effect. What happens in
equilibrium is a complex amalgam of these various shifts.
Computational Results: Baseline Parameterization

Fig. 4 shows the value functions (upper panel) and policy functions (lower
panel) for this showcase equilibrium.26 The equilibrium level of externality-
reducing activity, x*(R), generally decreases as the firms reputation grows,
but not everywhere. Thus, there is coasting, but the coasting is not global.
A comparison of Figs. 2 and 4 indicates that in any state R, the firms
externality-reducing activity is less when there is an activist than when there
is not. The direct effect of confrontation, discussed above, is one driver of
this, though it may not be the only one.
However, even though the activist induces a reduction in the intensity
of externality reduction state-by-state, it does not follow that over time
V*F(R) V*A(R)
x105 x104
1.6 3.5
1.4 3
2.5
1.2
VF
VA
2
1
1.5
0.8 1
0.6 0.5
0 10 20 30 0 10 20 30
R R
x*(R) z*(R) d*(R)

10 1.5 1.5
8
1 1
6
d
x
4
0.5 0.5
2
0 0 0
0 10 20 30 0 10 20 30 0 10 20 30
R R R
Fig. 4. Equilibrium Policy and Value Functions with an Activist: Baseline

Parameters. Notes: These graphs shows equilibrium values and policies as functions
of firms reputation. The top figures show the equilibrium value functions of the
firm and the activist. The bottom three figures show the externality-reducing
activity of the firm, and the activists choice of criticism and confrontation.
the firm will invest less when there is an activist. This is because, as can be
seen in the lower panel of Fig. 4, the activist generally engages in positive
amounts of both criticism and confrontation, with the mix of the two
activities varying with R. Therefore, unlike the no-activist case, the firm
will, in all likelihood, not reach states in which its externality-reducing
activity falls to 0.
To expand on this point, it is useful to describe the dynamics of the
model. The equilibrium actions of the firm and activist generate a Markov
E[R] E[x*(R)]
16 7.5
14
7
E[R]
E[x]
12
6.5
10
8 6
0 20 40 60 80 100 0 20 40 60 80 100
t t
E[z*(R)] E[d*(R)] E[p*(R)]

0.18 0.75 110
0.16
0.7 105
0.14
E[p]
E[d]
E[z]
0.12 0.65 100

0.1
0.6 95
0.08
0.06 0.55 90
0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100
t t t
x105 E[V*F(R)] x104 E[V*A(R)]

1.15 2.8
1.1 2.7
2.6
1.05
E [VA]
E[VF]
2.5
1
2.4
0.95 2.3
0.9 2.2
0 20 40 60 80 100 0 20 40 60 80 100
t t
Fig. 5. Equilibrium Dynamics with an Activist: Baseline Parameters. Notes: These

figures plot the expected reputation, policies, price, and values, where expectations
are taken with respect to the time t distribution.
process. Given any starting state, this process implies a transient probabil-
ity distribution over R, x*(R), z*(R), and d*(R) for any time period t.
Using these distributions, we can construct expectations over the firm and
the activists equilibrium behavior, as well as the firm and activists value,
for any time T = t. Fig. 5 illustrates the path of these expectations assum-
ing R0 = 15. The upper panels show how the firms expected reputation
and externality-reducing effort vary over time. For example, by T = 20, the
firms expected reputation E20[R] is approximately equal to 11, and as time
passes, reputation is expected to fall to slightly less than 10. Due the
activists efforts to impair the firms reputation, the firm experiences a
gradual decline in reputation from the initial state.27 However, unlike the
no-activist case, the firm does not, in the long run, stop investing in extern-
ality reduction. Indeed, there is sharp contrast in the time path of the firms
externality-reducing activity without and with an activist. Without an
activist, the time path of externality reduction declines over time; with an
activist, the firms expected externality-reducing activity would actually rise
over time, settling into an expected level of a little over 7 in the long run.
But this expectation actually disguises the fluidity of the firms situation.
As time passes both the firm and the activist continue to invest, which
causes small increases and decreases in the firms reputation, as well as an
occasional crisis which causes reputation to fall dramatically. Fig. 6 shows
the transient distributions over the firms reputation at three points in time:
T = 4, 8, and 16. It also shows the limiting distribution over R, which we
use to characterize the long-run dynamics of the game. In the long run, the
firms reputation could range from 1 to 30, with (as indicated earlier) an
expectation a little less than 10 and a mode of about 2. Thus, the interac-
tion between the firm and the activist gives rise to a dynamic in which the
firm occasionally manages to increase its reputation, but because x*(R)
T= 4 T= 8
0.25 0.14
0.12
0.2
0.1
0.15 0.08
Density
Density
0.1 0.06
0.04
0.05
0.02
0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
R R
T = 16 T=
0.07 0.09
0.08
0.06
0.07
0.05
0.06
0.04
Density
Density
0.05
0.03 0.04
0.03
0.02
0.02
0.01
0.01
0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
R R
Fig. 6. Transient Distributions Over the Firms Reputation R: Baseline Parameters.

Note: These figures show the distribution of reputation as the system evolves.
decreases in R throughout most of the state space, each time it increases its
reputation, it reduces its externality-reducing activity. From time to time,
the activists criticism reduces the firms reputation, and sometimes, the activist
provokes a severe crisis that causes the firms reputation to collapse drastically.
In the aftermath of these episodes, the firms motivation to enhance its reputa-
tion increases and it steps up its externality-reducing activity.
Computational Results: Grid Search Over G

The baseline parameters represent a single point in parameter space. To
explore the generalizability of the insights generated by this example, we
turn to the grid search over G.28
We begin by summarizing the extent to which equilibria have certain
properties in common. Table 2 reports percentages of the parameterizations
for which various properties were true in particular states, while Table 3
reports the percentage of parameterizations in which the equilibrium had
particular dynamic properties.29 (In the table and throughout the remain-
der of this paper, the subscript 0 refers to the no-activist case.)
From Tables 2 and 3, we can draw a number of conclusions. The first
property in Table 2 compares the firms equilibrium level of externality-
reducing activity in a given state R to the level that would have prevailed in
the absence of an activist:
Result 1. For each state R R, x R x0 R for over 70 percent of the
equilibria in the grid, and in 29 of 30 states, x R x0 R for over 80 per-
cent of the equilibria in the grid. Thus, the presence of the activist
is often associated with a reduction in the firms externality-reducing
activity in a given state.
We also see that, Proposition 2 notwithstanding, Proposition 1 does not
extend to the equilibrium with an activist:
Proposition 6. In contrast to the equilibrium in the absence of an activist,
the firms equilibrium level of externality-reducing activity is not monotoni-
cally decreasing in R; that is, there exist states R and R + 1 such that x*
(R + 1)>x*(R). Thus, the firm does not globally coast on its reputation.
On the other hand, coasting does occur to some extent, and it ensures
that in the wake of a crisis of expected severity, a firm may often increase
its level of externality-reducing activity:

Result 2. For each state R 6, x R < x R2 or x R1
2 for over 60 per-
cent of the equilibria in the grid. Thus, for a firm with a sufficiently
262
Table 2. Properties of Equilibria, State-by-State, in the Parameter Grid.
Proportion of Equilibria in G with Property in State R
Properties 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15
1. x R x0 R 0.95 0.89 0.89 0.85 0.84 0.86 0.87 0.88 0.86 0.90 0.86 0.90 0.85 0.87 0.88
2. x R xR 1 0.64 0.62 0.44 0.40 0.38 0.38 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37
3. x R < x R2 or x R1
2 0.36 0.38 0.48 0.53 0.60 0.61 0.62 0.63 0.63 0.64 0.64 0.64 0.64 0.64
4. VF R VF0
R 0.68 0.68 0.68 0.67 0.67 0.66 0.66 0.65 0.65 0.65 0.65 0.65 0.65 0.65 0.65
5. VA R VA0
R 0 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
6. VF R 1 VF R 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
7. VA R VA R 1 1.00 1.00 0.87 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
Properties 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
1. x R x0 R 0.88 0.88 0.88 0.85 0.83 0.84 0.87 0.85 0.85 0.85 0.83 0.82 0.81 0.72 0.86
2. x R xR 1 0.37 0.37 0.37 0.36 0.37 0.36 0.37 0.37 0.37 0.37 0.37 0.37 0.53 0.36
JOSE MIGUEL ABITO ET AL.

3. x R < x R2 or x R1
2 0.64 0.64 0.64 0.64 0.64 0.64 0.64 0.64 0.64 0.64 0.64 0.64 0.64 0.64 0.64
4. VF R VF0
R 0.65 0.65 0.64 0.64 0.64 0.64 0.64 0.64 0.64 0.64 0.64 0.64 0.64 0.63 0.63
5. VA R VA0
R 0 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
6. VF R 1 VF R

1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
7. VA R VA R 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
Table 3. Expected and Modal Long-Run States for Equilibria in the

Parameter Grid.
Property Proportion of Equilibria in G
E R < R 30 0.964
E[R] < 20 0.712
E[R] < 10 0.297
Mode[R] < 30 0.730
Mode[R] < 20 0.694
Mode[R] < 20 0.321
strong reputation, a crisis that reduces the firms reputation to the

expected post-crisis level will tend to motivate the firm to increase its
externality-reducing activity.
We also see from Table 2 that while the firm benefits from enhancing its
reputation, the activists value often declines when R increases.
Result 3. The firms value function is increasing in R. By contrast, the
activists value functions usually though not always is decreasing in
R. Thus, the firm benefits from an improvement in its reputation, while
the activist often benefits by hurting the firms reputation.
Table 3 provides insight into the nature of the dynamics in the model
and shows that in contrast to the no-activist case, the firms reputation
would rarely be expected to grow to the maximal attainable level:
Result 4. The firm rarely (in fewer than 4 percent of computed equilibria)
would be expected to attain the maximum possible reputation state R in
the long run, and often (more than two out of three cases) would be
expected to attain a long-run reputational state less than R = 20.
We next classify equilibria into three categories: (1) Diversified activist;
(2) specialized activist; and (3) ineffective activist. A diversified activist
engages in positive amounts of criticism and confrontation for long-run
relevant values of R, where long-run relevant means that the probability
of that state in the limiting distribution is at least 0.10.30 A specialized acti-
vist engages in just criticism or just confrontation in the long-run relevant
states. An ineffective activist does not engage in either activity in the long-
run relevant states. Table 4 characterizes the equilibria according to this
taxonomy, while Table 5 shows how the type of equilibrium correlates with
short-run and long-run levels of externality-reducing activity (measured by
Table 4. Taxonomy of Equilibria.

Equilibrium Type Proportion of Cases in G Representative Parameterization
Diversified activist 0.063 Baseline parameterization

Specialized activist
Confrontation 0.059 large relative to
Criticism 0.029 large relative to
Ineffective activist 0.523 Low value of , , A, , and/or
Table 5. Average Level of Short-Run and Long-Run Externality-

Reducing Investment by Type of Equilibrium.
Equilibrium Type E4[x*(R)] E[x*(R)]
Diversified activist 10.29 10.83

Specialized activist
Confrontation 9.95 9.51
criticism 9.17 2.98
Ineffective activist 2.52 0
the average values of E4[x*(R)] and E[x*(R)] over each equilibrium in

the category).
About 52 percent of the equilibria in the grid involve ineffective activists.
This is because for particularly low values of (like = 0.001), the firm
does not invest in externality reduction because a reputation for corporate
citizenship has very limited value, so, in turn, there is limited benefit to
activities aimed at compromising the firms reputation. Ironically, then,
firms that are apathetic to building a reputation are not inviting targets for
activists. Table 5 indicates that externality-reducing activity is generally
highest in both the short and long run when there is a diversified activist
and generally the lowest when there is an activist that specializes in criti-
cism. We summarize these results as follows:
Result 5. The highest levels of externality-reducing activity in both the
short run and the long run are generally associated with diversi-
fied activists.
Results 15 provide a strong suggestion of the forces at work in the
model and the trade-offs they create. On the one hand, positive levels of
criticism and confrontation change the firms incentives for externality-

reducing activity in a given state through the direct and prize effects.
In particular, the tendency for x*(R) to be less than x0 R in a given
state R (identified in Result 1) is a footprint of the negative direct effect of
confrontation. On the other hand, the activists presence changes the evo-
lution of states in the game. The clearest manifestation of this is Result 2,
which indicates that in the wake of a crisis the firm would, more often
than not, step up its level of externality-reducing activity. The activist
thus provides an antidote to coasting: its actions shape the dynamics of
the game in a way that tends to keep the firm in states in which it is more
motivated to invest in x.
In the introduction, we raised the question of why firms with well-
established brands (e.g., Coca Cola) seem to devote significant resources
to voluntary efforts that address negative externalities or other social
problems, even when it is hard to imagine how such activity could make
their very strong brands even stronger. Or, put differently, why dont
firms with strong reputations seem to reach a point of diminishing
returns that makes it worthwhile to coast on their reputations? Our
model resolves this puzzle by highlighting the role of activists and the
damage they can do to corporate reputations. In particular, the activist
tends to keep a targeted firms reputational capital at levels at which the
accumulation of additional reputational capital has significant value. For
this to occur, the activist must use confrontation and criticism as
more than just threats; there must be a positive probability along the
equilibrium path that the activist actually harms the firms reputation
from time to time.31
Comparative Statics of Long-Run Outcomes
An equilibrium is a vector in a large multi-dimensional space, so describing

in a compact way how it changes with changes in underlying parameters is
difficult. To simplify this task, we focus on how changes in parameters affect
the long-run equilibrium behavior, summarized by E[x*(R)], E[z*(R)],
and E[d*(R)];h long-run i performance, summarized by E[R] and

E p R E e0 R2 c ; and long-run value, summarized by E VF R

and E VA R .
E[x*(R)] E[z*(R)] E[d*(R)]

7.9 1.4 1.4
7.8 1.2 1.2

7.7
1 1
7.6
0.8 0.8
7.5
0.6 0.6
7.4
0.4 0.4
7.3
7.2 0.2 0.2
7.1 0 0
0 0.2 0.4 0 0.2 0.4 0 0.2 0.4

Fig. 7. How E[x*(R)], E[z*(R)] and E[d*(R)] Vary with . Notes:

Figures show how expected equilibrium policies (at the limiting distribution) vary
with salience of activist criticism. The dotted line indicates the baseline
parameter value.
Activist Efficacy: Variations in and

Figs. 7 and 8 summarize how expected long-run equilibrium outcomes
change as we vary the saliency parameter , holding all other parameters
fixed at baseline levels. (The dotted line identifies the baseline value of the
focal parameter, in this case .)32 Fig. 7 illustrates that as increases, the
activist tends to engage in more criticism and less confrontation in the long
run. This suggests that the activists policy tools are substitutes. If the
activists criticism has salience just a little below the baseline level (which,
recall, is where = ), then in the long run it stops relying on this
instrument and specializes in fomenting crises. Thus, z and d appear to be
imperfect but close substitutes.
Fig. 8 illustrates that an increase in salience has an ambiguous impact
on the firms long-run reputation and its value. Initially, long-run reputa-
tion and value rises with an increase in . This is because the increase in
criticism induces a substitution away from confrontation, which reduces
the probability of crises, limiting the frequency with which the firms
reputation crashes. However, as salience increases even more, the firms
expected long-run reputation and value fall. This is associated with an
increase in the level of the firms externality-reducing activity.
E[R] E[p*(R)]
9.8 94.5
9.7
9.6
9.5
94
9.4
9.3
9.2
93.5
9.1
9
8.9
8.8 93
0.1 0 0.1 0.2 0.3 0.4 0.5 0.1 0 0.1 0.2 0.3 0.4 0.5

4 E[V*F(R)] 4 E[V*A(R)]
x10 x10
9.7 3.1
9.65 3.05
3
9.6
2.95
9.55 2.9
9.5 2.85
9.45 2.8
2.75
9.4
2.7
9.35 2.65
9.3 2.6
0.1 0 0.1 0.2 0.3 0.4 0.5 0.1 0 0.1 0.2 0.3 0.4 0.5

Fig. 8. How E[R], E[p*(R)], E VF R , and E VF R Vary with . Notes:
Figures show how expected equilibrium reputation, prices and values (at the
limiting distribution) vary with salience of activist criticism. The dotted line
indicates the baseline parameter value.
Figs. 9 and 10 summarize the impact on the equilibrium of changes in

the newsworthiness parameter . This analysis reinforces the insight that
criticism and confrontation are imperfect but close substitutes. If
increases just a little above the baseline level, E[z*(R)] falls to 0, while if
increases just a little below the baseline level, E[d*(R)] falls to 0. An
empirical implication is that if otherwise similarly situated activists have
different s or s due to idiosyncratic reasons, we would expect to see the
activists specialize in one tactic or the other.
Changes in have an ambiguous effect on the firms long-run externality-
reducing activity, and that effect differs from the effect of changes in . If the
newsworthiness parameter increases beyond a certain point, the long-run
E[x*(R)] E[z*(R)] E[d*(R)]

7.5 3 3
7
2.5 2.5
6.5
2 2
6
5.5 1.5 1.5
5
1 1
4.5
0.5 0.5
4
3.5 0 0
0 0.2 0.4 0 0.2 0.4 0 0.2 0.4

Fig. 9. How E[x*(R)], E[z*(R)], and E[d*(R)] Vary with . Notes:

with newsworthiness of activist confrontation. The dotted line indicates the baseline
parameter value.
level of x declines. This reflects a sufficiently powerful direct effect of con-

frontation discussed earlier.
E[d*(R)] is also nonmonotonic in ; as it becomes sufficiently easy
to provoke a crisis, the activists expected confrontation is scaled back
in equilibrium.
Finally, the activists long-run value E VA R may decrease in . This

is a consequence of the decline in E[x*(R)] for sufficiently large values of
. If provocative activity is highly newsworthy, the activist is actually hurt.
The logic is that in a media environment in which crises are very easy to
provoke, a firm simply gives up hope that it can sustain a good reputation
for corporate citizenship and scales back the externality-reducing activity
that the activist values. We can summarize the results of this analysis
as follows:
Result 6. (i) For the activist, criticism and confrontation are imperfect
substitutes; (ii) Increasing the newsworthiness of the activists efforts to
provoke a crisis does not unambiguously increase the firms long-run
externality-reducing activity, nor does it necessarily increase the
intensity of the activists confrontation in the long run; (iii) By contrast,
E[R] E[p*(R)]
30 130
125
25
120
115
20
110
15 105
100
10
95
90
5
85
0 80
0.1 0 0.1 0.2 0.3 0.4 0.5 0.1 0 0.1 0.2 0.3 0.4 0.5

E[V*F(R)] E[V*A(R)]
x105 x104
2.5 3
2.5
2
2
1.5
1.5
1
1
0.5 0.5
0.1 0 0.1 0.2 0.3 0.4 0.5 0.1 0 0.1 0.2 0.3 0.4 0.5

limiting distribution) vary with newsworthiness of activist confrontation. The
dotted line indicates the baseline parameter value.
increasing the saliency of the activists criticisms of the firm does increase
the firms long-run externality-reducing activity (over the range where
the activist engages in positive amounts of criticism).
Activist and Firm Patience: Variations in A and F

Figs. 11 and 12 summarize the impact on long-run equilibrium outcomes of
varying A, holding all other parameters fixed at baseline levels. A moder-
ately impatient activist (A between about 0.60 and 0.90) is a minor threat
to the firm; it engages in small amounts of criticism, but no confrontation
in the long run. A highly impatient activist (A less than about 0.60) is no
E[x*(R)] E[z*(R)] E[d*(R)]

8 1.4 1.5
7 1.2
6
1
1
5
0.8
4
0.6
3
0.5
0.4
2
1 0.2
0 0 0
0 0.5 1 0 0.5 1 0 0.5 1
A A A
Fig. 11. How E[x*(R)], E[z*(R)], and E[d*(R)] Vary with A. Notes:
Figures show how expected equilibrium policies (at the limiting distribution) vary with
the activists discount factor. The dotted line indicates the baseline parameter value.
threat at all; it engages in no equilibrium activity of any kind. The most

dangerous activist, from the firms perspective, is a patient one. As A
increases, the activist aggressively substitutes confrontation for criticism
(except for the very highest values A, at which point the activist increases
both activities in tandem).33
Recall that the discount factor reflects both the time preference of
the activist, as well as its survival probability. The computational results
indicate that a well-funded activist with strong survival prospects is
more likely to attempt to provoke a crisis, while an activist with a lower
survival probability will tend to engage in criticism. This is not because
confrontation is less expensive or more efficacious for the well-funded
activist (efficacy and cost are being held fixed in this analysis), but rather
because the payoff from the activists two instruments have different
dynamic implications. Inducing a crisis that crashes the firms reputation
has a potentially big payoff to the activist since the firm, in the wake of
the crisis, significantly increases x to rebuild its image. However, it takes
time for the activist to trigger a crisis of sufficient impact to really matter,
so a less well-funded, and therefore more impatient, activist may forego
crisis provocation altogether and instead seek to motivate the firm by
E[R] E[p*(R)]
30 140
130
25
120
20
110
15 100
90
10
80
5
70
0 60
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
A A
E[V*F(R)] E[V*A(R)]
x105 x105
2.5 15
2
10
1.5
5
1
0.5 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
A A
Fig. 12. How E[R], E[p*(R)], E VF R , and E VF R Vary with A. Notes:
limiting distribution) vary with the activists discount factor. The dotted line
activities that marginally chip away at its reputation. We summarize these

insights as follows:
Result 7. A more patient activist tends to rely on confrontation to a
greater degree, and on criticism to a lesser degree, than a less patient
activist. Above a threshold value of A, the firms value declines precipi-
tously as A increases.
Figs. 13 and 14 summarize the impact on long-run equilibrium out-
comes by varying F, holding all other parameters (including A) fixed
E[x*(R)] E[z*(R)] E[d*(R)]

10
0.8 0.8
8
0.6 0.6
6
0.4 0.4
4
2 0.2 0.2
0 0 0
0 0.5 1 0 0.5 1 0 0.5 1
F F F
Fig. 13. How E[x*(R)], E[z*(R)], and E[d*(R)] Vary with F. Notes:
Figures show how expected equilibrium policies (at the limiting distribution) vary with
the firms discount factor. The dotted line indicates the baseline parameter value.
at baseline levels. There are two noteworthy implications of this analy-

sis. First, a more patient firm is more vulnerable to a crisis than a less
patient firm. Second, an activist prefers to interact with a more patient
firm. These implications arise because a more patient firm derives a
bigger prize from building reputation than a less patient firm, which
makes the more patient firm more willing to invest in reputation-building.
This directly benefits the activist. When a crisis occurs, the more
patient firm has a greater motivation to rebuild its reputation than the
less patient firm, which makes crisis provocation a particularly attrac-
tive strategy against a patient firm. This latter implication suggests that
if the activist must choose among potential targets, a financially sound
firm would be a more attractive target than a marginal firm. This is
consistent with empirical evidence about activist behavior (Eesley &
Lenox, 2006). We summarize the insights from this part of the analysis
as follows:
Result 8. A more patient firm is more susceptible to crises than a less
patient firm. The activist prefers a more patient target to a less
patient target.
E[R] E[p*(R)]
30 130
125
25
120
20 115
110
15 105
100
10
95
5 90
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
F F
x105 E[V*F(R)] x104 E[V*A(R)]

10 4
9 3.5
8
3
7
6 2.5
5 2
4 1.5
3
1
2
1 0.5
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
F F
Fig. 14. How E[R], E[p*(R)], E VF R , and E VF R Vary with F. Notes:
limiting distribution) vary with the firms discount factor. The dotted line indicates
the baseline parameter value.
Returns to Reputation for Corporate Citizenship: Variations in

Figs. 15 and 16 summarize the impact on long-run equilibrium outcomes
by varying , holding all other parameters fixed at baseline levels. The
greater the impact of reputation on brand equity, the more the firm invests
in externality-reducing activity. However, the activists behavior is not
monotonic in . For 0.125, the firm does not invest in externality reduc-
tion, and the activist accordingly chooses no activity of either kind. Once
exceeds 0.125, there is an upward jump in criticism, but as increases
E[x*(R)] E[z*(R)] E[d*(R)]

25 0.35 1
0.3
20 0.8
0.25
15 0.2 0.6
10 0.15 0.4
0.1
5 0.2
0.05
0 0 0
0 0.2 0.4 0.6 0 0.2 0.4 0.6 0 0.2 0.4 0.6

Fig. 15. How E[x*(R)], E[z*(R)] and E[d*(R)] Vary with . Notes:
with the elasticity of brand equity with respect to reputation. The dotted line
between 0.125 and 0.25, the activist decreases criticism and substitutes con-
frontation for it. As increases further above 0.25, increases in elicit
more of both types of activities. Criticism is apparently the more attractive
tool for the activist when it faces a firm that has only a modest concern
with using corporate citizenship to build brand equity. By contrast, when
corporate citizenship has a large effect on brand equity, crisis provocation
becomes increasingly attractive. Thus, a firm for whom an image of good
corporate citizenship is particularly important would be especially vulner-
able to confrontational tactics by an activist.34 We summarize this part of
the analysis as follows:
Result 9. If is below a threshold level, the firm does not engage in
externality-reducing activity in the long run, and the activist does
not engage in any effort to harm the firms reputation. Above that
threshold, as increases, the firms long-run externality-reducing
activity increases, as does the activists levels of confrontation,
increasing the likelihood of crises. The activists long-run level of
criticism initially falls as increases above the threshold, but it even-
tually begins to increase.
E[R] E[p*(R)]
30 200
25
150
20
15
100
10
5 50
0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.1 0 0.1 0.2 0.3 0.4 0.5 0.6

E[V*F(R)] E[V*A(R)]
x105 x104
5 15
4.5
4
10
3.5
3
2.5 5
2
1.5
0
1
0.5
0 5
0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.1 0 0.1 0.2 0.3 0.4 0.5 0.6

limiting distribution) vary with the elasticity of brand equity with respect to
reputation. the dotted line indicates the baseline parameter value.
Activist Passion: Variations in

Figs. 17 and 18 summarize how variations in the activists passion affect
long-run equilibrium behavior and performance. If is slightly less than 1,
it is completely ineffective: it engages in no criticism or confrontation. If
the activists objective function is social welfare ( = 1, indicated by the
dotted line in the left of each panel), it engages in positive, but very small,
amounts of criticism and confrontation. Thus, an activist that sees itself
merely acting on behalf of the general public interest will hardly make a
difference in the long run. Only if the activist is sufficiently passionate will it
E[x*(R)] E[z*(R)] E[d*(R)]

8
1.2 1.2
7
1 1
6
5 0.8 0.8
4 0.6 0.6
3
0.4 0.4
2
0.2 0.2
1
0 0 0
0 2 4 0 2 4 0 2 4

Fig. 17. How E[x*(R)], E[z*(R)], and E[d*(R)] Vary with . Notes:
with the activists passion. The dotted line at = 2 indicates the baseline parameter
value, while the value = 1 represents perfect alignment of the activists passion
with social welfare.
be motivated to take actions that lead to significant amounts of externality-

reducing activity in the long run. However, beyond a certain point ( slightly
less than 2) increases in induce a decline in externality-reducing activity.
This is because the passionate activist is very keen to provoke a crisis
(long-run confrontation monotonically increases in ). However, because of
the direct effect of confrontation this will induce the firm to scale back its
externality-reducing activity. Increasing the passion of the activist makes the
activist a more dangerous adversary for the firm, but may not advance the
social interest. We summarize this analysis as follows:
Result 10. If the activist is insufficiently passionate, it engages in no crisis
or confrontation of any kind and is thus ineffective. As the activists pas-
sion increases beyond this threshold, the long-run level of confrontation
rises monotonically, while the long-run level of criticism initially decreases
but then increases in . The firms long-run externality-reducing activity
increases as the activists passion increases beyond the threshold, but
eventually it falls. The firms long-run value monotonically decreases as
activist passion increases.
E[R] E[p*(R)]
30 130
125
25 120
115
20 110
105
15 100
95
10 90
85
5 80
0 0.5 1 1.5 2 2.5 3 3.5 4 0 0.5 1 1.5 2 2.5 3 3.5 4

E[V*F(R)] E[V*A(R)]
x105 x104
2.5 6
5
2
4
1.5 3
2
1
1
0.5 0
0 0.5 1 1.5 2 2.5 3 3.5 4 0 0.5 1 1.5 2 2.5 3 3.5 4

limiting distribution) vary with the activists passion. The dotted line at = 2
indicates the baseline parameter value, while the value = 1 represents perfect
alignment of the activists passion with social welfare.
SUMMARY AND CONCLUSIONS

We model the interaction between a firm and an activist using a discrete-
time, infinite-horizon dynamic stochastic game. The firm is assumed to be
profit maximizing while activists care about reducing a socially inefficient
externality. The firm can engage in activity that reduces the externality and,
with some probability, will receive an improvement in its corporate reputa-
tion, which enhances consumer demand. Activists can engage in two forms
of costly activity: they can criticize the firm, which, with some probability,
has a marginally negative impact on the firms reputation, or they can trigger
a crisis which crashes the firms reputation.
While the firm has an incentive to invest in externality-reducing activity
without the existence of an activist, this effort is subject to decreasing
marginal returns in equilibrium. The incentive to coast when reputational
equity is high creates a conflict between firm and activists. To prevent the
firm from coasting, activists engage in a combination of criticism and crisis
inducing behavior. The activists efforts prevents the firm from coasting.
That said, state-by-state the activists presence functions like a tax and
depresses firms incentives to engage in corporate citizenship, which serves
neither the firms nor the activists interests. However, the activists activ-
ities keep the firm motivated to supply externality reduction even in the
long-run. This raises the possibility that the impact of activist campaigns
on long-run social welfare could be ambiguous.
However, to the extent that such forms of private politics present an
alternative regulatory mechanism any welfare comparisons need to be dis-
cussed in a broader context (Abito et al., 2014). On the one hand, we can
consider such mechanisms in cases where traditional conditions for public
regulation do not hold. For example, Pigouvian taxes or subsidies may be
infeasible; collective action and information problems may make Coasian
bargaining impractical; or governance problems may undermine public
regulation. Such a perspective may be especially appropriate for globally
operating firms with business operations in countries with weak or nonexis-
tent regulatory mechanisms. In that case, activist pressure would serve as a
(partial) substitute for public regulation. That said, activists also operate in
mature economies with fully developed legal, political, and regulatory insti-
tutions. Still, activists increasingly have resorted to directly targeting firms
to change business practice. Moreover, activists have stated publicly that
such private politics campaigns are more effective than the traditional
channel of pressuring elected officials (e.g., Baron & Diermeier, 2007). To
assess such claims, a proper comparison would move beyond a traditional
welfare analysis and compare mechanisms based on private politics with
political economy models of public regulation, where public policy is the
consequence of competition among politicians, interest groups, and voters
in public arenas.
From a positive point of view, we investigated how the nature of the
equilibrium depends on the parameters of the model such as the relative
effectiveness of the two activist activities, the returns to corporate citizen-
ship, the discount parameters and so forth. For example, companies for
whom corporate citizenship has a higher value are more inviting targets for
activists. Moreover, activists are more effective when they both criticize
and try to trigger crises. More patient activists rely more on confrontation,
and more patient firms are more vulnerable to crises. Criticism and
confrontation, however, are imperfect substitutes, and only in the case of
criticism does effectiveness of that activity necessarily increases long-run
externality-reducing activity by the firm.
The analysis generates a variety of empirical implications:
1. NGOs will target companies even if such companies engage in corporate

social responsibility and self-regulation.
2. Large, well-financed, highly visible firms are better targets for activists
and are more vulnerable to crises.
3. Boycotts and other similar forms of NGO activities negatively impact
shareholder value. First, the impact of criticism or a reputational crisis
leads to a decrease in the reputation of the company and thus lower
profits. Second, once targeted the company will invest more heavily in
costly private regulation. Both activities negatively impact shareholder
value in the short-run.
4. Diversified activists (i.e., NGOs that use both criticism and confronta-
tion) are most successful in inducing firms to choose high levels of
externality reducing activities in both the short and the long-run. Better
financed activists tend to rely more on confrontation.
5. For activists, criticism and confrontation are imperfect substitutes.
Increasing the salience of an issue increases the firms long-run invest-
ment in abatement activities, while increased newsworthiness has mixed
effects: in intermediate ranges increasing newsworthiness increases
abatement, but once newsworthiness reaches a critical value activists are
hurt, since firms simply give up if a crisis can be triggered too easily.
6. Firms for whom an image of good corporate citizenship is particularly
valuable engage in higher levels of abatement, but are also more fre-
quent targets for confrontational tactics.
7. Only activists that are more passionate than the general public have
any impact on the long-run behavior of firms. If effective activists are
also more successful in obtaining resources in a market for donation
and volunteers this implies that ceteris paribus activists should have
more extreme preferences that the public. Yet, by the same logic, acti-
vists will not be too extreme either, since very passionate activists are
less effectively, as they tend to engage too much in confrontation
which reduces the incentives for firms to invest in abatement activities
in the long run.
Our implications are consistent with various findings in the empirical

literature. First, negative media coverage and activist targeting have a
negative impact on firm financial performance. Karpoff, Lott, and Wehrly
(2005), Konar and Cohen (1996), and Beatty and Shimshack (2010) show
that bad news about a companys environmental performance reduces their
share-price. King and Soule (2007) and Vasi and King (2012) provide evi-
dence that protests and boycotts lower corporate financial performance.
King (2008) shows that the financial impact of boycotts varies with media
attention as captured in our newsworthiness parameter .
Second, various studies have found that CSR activities lack a positive
financial impact on firms and may have even have a (moderately)
negative impact (Fisher-Vanden & Thorburn, 2011; Jacobs, Singhal, &
Subramanian, 2010). Such findings have been viewed as puzzling as we
may expect corporate social responsibility to be positively correlated
with profits, for example, by appealing to socially conscious customers,
employees and share-holders. But our model points out that CSR and self-
regulation may be necessary as a defensive tactic against activist threats.
The benefit to the firm then results in risk mitigation. This is consistent
with recent evidence. Kotchen and Moon (2012) show that companies tend
to engage in more reputation-enhancing activities if they had suffered more
reputational damage in the past. McDonnell and King (2013) as well as
McDonnell, King, and Soule (2015) show that firms that were targeted by
a boycott respond with increased pro-social activities to mitigate negative
reputational impact.
Third, there is considerable evidence that activists select firms strategi-
cally. In the context of environmental issues. Eesley and Lenox (2006)
show that firm characteristics such as cash-flow, assets, and advertising
intensity are positively correlated with an increased likelihood of being tar-
geted in a boycott. Such measures partially capture our notions of news-
worthiness, salience and firm patience.
While the broad alignment with the existing empirical literature is
encouraging, many of the empirical implications of the model have not
been tested. For example, we are not aware of studies that assess whether
confrontation and criticism are indeed imperfect substitutes. Similarly, the
model points out that newsworthiness has a nonlinear effect on activist
success, while existing empirical studies tend to focus on linear influence.
More generally, the interaction between firms and activities is rich in strategic
complexity that should be captured in empirical studies. This suggests that
there is an important role for statistical approaches that explicitly capture
strategic interaction such as structural estimation.35
Our analysis also has some practical implications for managers and acti-
vists. In the Introduction we stated the puzzle why well-known companies
such as McDonalds or Coca Cola continue to invest in corporate social
responsibility activities and self-regulation even though their brand equity
is unlikely to be improved by such activities. Our model implies that
activists will continue to target such companies to reduce coasting, which
creates incentives for companies to continue to invest in private regulation.
Companies will continue to experience criticism and confrontation despite
adopting responsible business practices, and from time to time such activ-
ities will negatively impact the companys reputation. Following Argenti
(2004) we can call this phenomenon the Starbucks Paradox: a company
that engages in CSR activities continues to be targeted by activists asking
management to do ever more. Managers are often puzzled by such findings
and expect that they should be rewarded for their good deeds by lowering
the likelihood of an activist campaign. Such a mind-set rests on a static
notion of preemption. That is, private regulation works to forestall activist
actions. But a dynamic view of corporate reputation shows that this
reasoning is incomplete. Decreasing dynamic returns to reputation create
incentives for companies to coast which activists can counter by engaging
in sporadic campaigns. Such campaigns, if successful, destroy some of a
firms reputational capital which keeps the firm hungry, that is, incentivized
to continue to invest in self-regulation.
Even high levels of private regulation and corporate social responsibility
do not forestall NGO activity. Passionate activists always want the com-
pany to do more. Companies thus will experience episodes of criticism and
confrontation and rarely achieve the maximal reputational state and,
despite ongoing investments in CSR, will hover around a moderate level of
reputation. The presence of criticism and confrontation by activists and
NGOs is not an aberration, but should be viewed as a normal and expected
part of a companys business environment and should be managed as such.
This is especially true for well-known companies where maintaining a
strong reputation is highly valuable.
Our approach left out many of the complexities regarding the interaction
between firms and activists. For example, activists were limited to inflict
harm on the firm, an assumption that, while empirically supported, ideally
would be derived in equilibrium of a richer model. Correspondingly, it
would be worthwhile to consider a socially motivated firm. Other natural
extensions would allow for bargaining between firms and activists and con-
sider multiple, competing firms and activists. That said, even in the simple
model, the dynamic interactions between the firm and the activist proved
surprisingly rich. We hope that more complex approaches can be developed

on its foundation.
NOTES
1. Maxwell, Lyon, and Hackett (2000) call this self-regulation. Vogel (2010)
presents the closely related idea of civil regulation. In some cases (e.g., Maxwell
et al., 2000) self-regulation may be motivated by the desire to forestall more demanding
public regulation. In our model public regulation is either highly unlikely (perhaps
because of effective industry lobbying) or practically infeasible (as in the case of conflict
diamonds). See Egorov and Harstad (2012) and Baron (2013) for recent models that
investigate the interaction between private and public politics.
2. For an overview of such governance models see Koppell (2010).
3. Abito, Besanko, and Diermeier (2015) studies the normative aspects of this
interaction.
4. In the model, activists can only harm the firms reputation, they cannot
improve it, for example, by endorsing the firms business practices and products.
Much of the empirical literature on activists has pointed out that activists focus on
inducing harm (e.g., Friedman, 1999). While we follow this approach here, future
work may enlarge the strategy set for activists to include providing benefits for the
firm and then provide an equilibrium analysis to explain the prevalence of harm.
For a discussion and a static model with both harm and benefits, see Baron and
Diermeier (2007).
5. The upper bound R contributes to diminishing marginal return to investments
in reputation building, but as illustrated below it is not the only source of diminishing
marginal returns in the model. Diminishing marginal returns to investment in reputation
building is supported by empirical evidence; see, for example, Lev, Petrovits, and
Radhakrishnan (2010) on the impact of corporate charitable contributions on
sales growth.
6. R0 does not affect the equilibrium, but it does affect the transient (short-run)
dynamics implied by the equilibrium.
7. See Argenti (2004). Dean (2004) and King and McDonnell (2013) provide
empirical evidence that supports this phenomenon.
8. Note that by changing assumptions on the distribution over 1 and
max{R 1,1} in the event of a crisis, we can change the implicit strength of the
bigger they are, harder they fall property and the bank account property.
9. We do not model the process by which the signal X~ t is generated. This process
may involve costs, such as advertising or public relations efforts that are independent
of the direct costs of providing xt.
10. For a discussion of the underlying processes, see Baron (2009b) and
Diermeier (2011), especially Chapters 13.
11. For examples of such tactics, see Diermeier (2011), Chapter 3.
12. For empirical studies of the impact of different forms of activism on firm
behavior in the context of environmental pollution, see Eesley and Lenox (2005),
Eesley and Lenox (2006), and Lenox and Eesley (2009).
13. For example, it has been hypothesized that a countrys concern about animal
welfare may be systematically related to its economic growth. See Frank (2008).
14. See Baron (2009b, 2007) and Diermeier (2011) for details.
15. Strictly speaking, this would be correct only if = . By normalizing the cost
parameters, we ensure that the difference between reputation-impairing and crisis-
inducing effort is due either to differences in and , or to fundamental differences in
the nature of reputation-impairing activity and crisis-inducing activity.
16. This ranking is now called the Worlds Most Admired Companies.
17. In this respect, our model differs from that of Baron (2003), which assumes
that the target firm and the activist can negotiate over the firms provision of
externality-reducing activity and the activists undertaking of a campaign against
the firm.
18. For a discussion of activist commitment, see Baron and Diermeier (2007).
19. We condition on R in writing MBx() because MBx does not depend on the
entire vector VF R but rather on just parts of it in a manner specific to the state R.
20. We check second-order conditions in the appendix.

21. Thus, for example, VF VF 1; ; VF R , with the other terms in
VF ; VA ; x ; z ; d defined in the same way.
22. These equations are the two Bellman equations (2) and (5); and six reformulated
KuhnTucker conditions, that is, Eqs. (10) and (11) applied to Eqs. (3), (6), and (7),
using (12) for x, z, and d.
23. Of course, this effect would be transitory since, in equilibrium, the firm would
then take steps to rebuild its reputation, as Johnson and Johnson did during the
Tylenol crisis. Activist campaigns can have a similar (short-term) impact on sales.
During Greenpeaces 1995 campaign against Shell over the Brent Spar Platform
Shells sales in Germany fell by 40 percent. For a discussion of both cases, see
Diermeier (2011), Chapters 1 and 3.
24. We take a finer grid than G in the comparative statics exercise since equilibria
are easier to compute in this case using the homotopy method.
25. The direct effect is analogous to the compensated effect of price on quantity
demanded in consumer theory.
26. Though we cannot rule out the possibility of multiple equilibria, we were
unable to find more than one equilibrium in this case.
27. Of course, these dynamics are contingent on the starting state. If the starting
state had been less than R = 10, there would have been a gradual rise in the firms
reputation to about 10.
28. The grid contains 720 unique parameter combinations. We were able to
compute equilibria for 641 of these, or about 89 percent of the grid. To reflect the
79 parametrizations for which we could not compute equilibria, we can construct
bounds for the computed proportions for each property in the subsequent tables.
Specifically, if p is the proportion of computed equilibria satisfying some property,
then the corresponding proportion over the entire grid is inside the interval
[0.89p, 0.89p + 0.11]. Of the 79 cases for which equilibria could not be computed,
60 of these involve = 4, while the rest involve = 2. For those with = 2, there
is one case with = 0.35 and 18 cases with = 0.4. These cases are associated with
high returns to reputation for the firm.
29. At all points in G, we did not identify cases of multiple MPE.
30. We also used a cutoff of 0.05 to define long-run relevant states. The lower cutoff
results in more non-classifiable equilibria, but the breakdown between diversified,
specialized, and ineffective activists was about the same.
31. This effect could also operate if a firm faced an exogenous probability of either
a reputational crisis or a marginal diminution of its reputation. Thus, activists per se
are not critical to resolving the puzzle set out in the introduction. Still, the forces that
enable activists to harm corporate reputations, such as the nature of the mass media
environment, are much the same forces that allow exogenous events, such an oil spill
or a major product defect, to become newsworthy enough to impair a companys
reputation. Moreover, activists are often quick to exploit such exogenous events. The
key point of our theory is that periodic destruction of reputational capital, whether
activist-induced or exogenous, can keep a firm sufficiently motivated to enhance it,
thereby explaining why some companies never seem to reach the point at which it
makes sense to coast.
32. For each comparative statics analysis in this section, we present two figures.
The first shows how long-run equilibrium behavior (x, z, d) varies with the focal
parameter, and the second shows how long-run performance (reputation, price, and
firm and activists values) vary with the focal parameter.
33. Thus, for extremely patient activists, reputation-impairing and crisis-inducing
effort become complementary.
34. For empirical evidence supporting this claim see, for example, Eesley and
Lenox (2006).
35. See Baron, Harjoto, and Jo (2011) for the use of a simultaneous equation
model to address these concerns.
36. Throughout the proof, we suppress dependence of VF() on n where there is
no ambiguity.
REFERENCES
Abito, J. M., Besanko, D., & Diermeier, D. (2015). Private politics and public interest: NGOs,
corporate campaigns, and social welfare. Mimeo.
Alsop, R. J. (2004). The 18 immutable laws of corporate reputation: Creating, protecting, and
repairing your most valuable asset. New York, NY: Free Press.
Argenti, P. (2004). Collaborating with activists: How Starbucks works with NGOs. California
Baron, D. P. (2003). Private politics. Journal of Economics and Management Strategy,
12(1), 3166.
Baron, D. P. (2009a). A positive theory of moral management, social pressure, and corporate
social performance. Journal of Economics and Management Strategy, 18(1), 743.
Baron, D. P. (2009b). Business and its environment (6th ed.). Upper Saddle River, NJ:
Prentice Hall.
Baron, D. P. (2012). The industrial organization of private politics. Quarterly Journal of
Political Science, 7(1), 135174.
Baron, D. P. (2013). Self-regulation in private and public politics. Mimeo, Stanford University.
Baron, D. P., Harjoto, M. A., & Jo, H. (2011). The economics and politics of corporate social
performance. Business and Politics, 13(2).
Baron, D. P., & Yurday, E. (2004). Strategic activism: The rainforest action network. Case
P-44. Stanford, CA: Graduate School of Business, Stanford University.
Beatty, T., & Shimshack, J. P. (2010). The impact of climate change information: New evidence
from the stock market. The B.E. Journal of Economic Analysis & Policy, 10(1), Article 105.
Besanko, D., Doraszelski, U., Kryukov, Y., & Satterthwaite, M. (2010). Learning-by-doing,
organizational forgetting, and industry dynamics. Econometrica, 78(2), 453508.
Bond, J., & Kirshenbaum, R. (1998). Under the radar: Talking to todays cynical consumer.
New York, NY: Wiley.
Borkovsky, R. N., Doraszelski, U., & Kryukov, Y. (2010). A users guide to solving dynamic
stochastic games using the homotopy method. Operations Research, 58(4), 11161132.
Cabral, L., & Riordan, M. (1994). The learning curve, market dominance, and predatory
pricing. Econometrica, 62(5), 11151140.
Dean, D. H. (2004). Consumer reaction to negative publicity: Effects of corporate reputation,
response, and responsibility for a crisis event. Journal of Business Communication,
41(2), 192211.
Dennis, E. E., & Merrill, J. C. (1996). Media debates: Issues in mass communication. White
Plains, NY: Longman.
Diermeier, D. (2011). Reputation rules: Strategies for building your companys most valuable
asset. New York, NY: McGraw-Hill.
Dowling, G. (2002). Creating corporate reputations: Identity, image, and performance.
New York, NY: Oxford University Press.
Eesley, C., & Lenox, M. (2005). Firm responses to secondary stakeholder action. Best paper
proceedings of the 2005 Academy of Management conference, Honolulu, HI.
Eesley, C., & Lenox, M. (2006). Secondary stakeholder actions and the selection of firm targets.
Best paper proceedings of the 2006 Academy of Management conference, Atlanta, GA.
Egorov, G., & Harstad, B. (2012). Private politics and public regulation. Mimeo,
Feddersen, T., & Gilligan (2001). Saints and markets: Activists and the supply of credence
goods. Journal of Economics and Management Strategy, 10(1), 149171.
Fisher-Vanden, K., & Thorburn, K. (2011). Voluntary corporate environmental initiatives and
shareholder wealth. Journal of Environmental Economics and Management,
62(3), 430445.
Frank, J. (2008). Is there an Animal Welfare Kuznets Curve? Ecological Economics,
66, 478491.
Friedman, M. (1999). Consumer boycotts. New York, NY: Routledge.
Ingram, P., Yue, L., & Rao, H. (2010). Trouble in stores: Probes, protest, and store openings
by Wal-Mart, 19982007. American Journal of Sociology, 116, 5392.
Jacobs, B. W., Singhal, V. R., & Subramanian, R. (2010). An empirical investigation of
environmental performance and the market value of the firm. Journal of Operations
Karpoff, P., Lott, J., & Wehrly, E. (2005). The reputational penalties for environmental violations:
Empirical evidence. Journal of Law and Economics, 48, 653675.
King, B. (2008). A political mediation model of corporate response to social movement activism.
Administrative Science Quarterly, 53(3), 395421.
King, B., & Pearce, N. (2010). The contentiousness of markets: Politics, social movements and
institutional change in markets. Annual Review of Sociology, 36, 249267.
King, B., & Soule, S. A. (2007). Social movements as extra-institutional entrepreneurs:
The effect of protests on stock price returns. Administrative Science Quarterly,
52(3), 413442.
Konar, S., & Cohen, M. (1996). Information as regulation: The effect of community right to
know laws on toxic emissions. Journal of Environmental Economics and Management,
32, 109124.
Koppell, J. G. S. (2010). World rule: Accountability, legitimacy, and the design of global governance.
Chicago, IL: University of Chicago Press.
Kotchen, M., & Moon, J. J. (2012). Corporate social responsibility for irresponsibility. The BE
Journal of Economic Analysis & Policy, 12(1).
Lenox, M., & Eesley, C. (2009). Private environmental activism and the selection and response
of firm targets. Journal of Economics & Management Strategy, 18(1), 4573.
Lev, B., Petrovits, C., & Radhakrishnan, S. (2010). Is doing good good for you? Yes,
charitable contributions enhance revenue growth. Strategic Management Journal,
31, 182200.
Lewin, T. (1982). Tylenol Posts an Apparent Recovery. New York Times, December, 25.
Maxwell, J. W., Lyon, T. P., & Hackett, S. C. (2000). Self-regulation and social welfare: The
political economy of corporate environmentalism. Journal of Law and Economics,
43(2), 583618.
McDonnell, M. H., & King, B. (2013). Keeping up appearances: Reputation threat and prosocial
responses to social movement boycotts. Administrative Science Quarterly, 58, 387419.
McDonnell, M. H., King, B., & Soule, S. (2015). A dynamic process model of private politics:
Activist targeting and corporate receptivity to social challenges. American Sociological
Review, 80(3), 654678.
Minor, D. (2010). CSR as reputation insurance: Theory and evidence. Working Paper.
Shaw, R. (1996). The activists handbook: A primer for the 1990s and beyond. Berkeley, CA:
University of California Press.
Vasi, I. B., & King., B. (2012). Social movements, risk perceptions, and economic outcomes.
American Sociological Review, 77, 573596.
Vogel, D. (2010). The private regulation of global corporate conduct: Achievements and
limitations. Business and Society, 49(1).
APPENDIX
Second-Order Conditions
Second-order conditions generally depend on x, z, and d, and thus we can

only check that the second-order sufficient conditions are satisfied locally.
Showing this for the firms problem is straightforward. However, if the
activist uses both instruments, then we have to check numerically if these con-
ditions are satisfied. These sufficient conditions are satisfied for all parametri-
zations in the grid, including our baseline case. In what follows, we derive the
second-order conditions assuming the nonnegativity constraints do not bind.
Firms Problem
Differentiating the first-order condition of the firms problem with respect
to x yields
2
MBx
1 x
Since MBx = k > 0, then

2
MBx < 0
1 x
Activists Problem
The Hessian of the activists problem is given by

M11 M12
M
M21 M22
where
2
M11 MBz < 0
1 z
2
M12 MBd < 0
1 d

A F R VA R
VA R 1
M12 M21
1 z2 1 d 2 1 F VA R 1 VA R
The Hessian is negative definite if M11 < 0 and M11M22 M12M21 > 0.
The condition on the determinant can be rewritten as
41 z1 > 2
where
8 9
>
>
PR1 1
>
>
>
< r1 VA r VA R >
=
R1
1
>0
>
>
F R 1
>
A z VA R VA R 1
>
>
: 1 R z V R 1 V R ; >
F A A A

FR VA R
VA R 1
2 > 0:
1 F VA R 1 VA R
While M11 < 0 can be readily seen, we have to compute the condition on
the determinant. We check whether 4(1 + z)1 > 2 holds at the equili-
brium solution for every parametrization in the grid. Our numerical results
show that this is true for all parametrizations in the grid.
Computational Method
For each analytical experiment, we explore the graph of the equilibrium

correspondence as the relevant parameters vary using the homotopy
algorithm discussed in Besanko et al. (2010). This natural-parameter
homotopy gives us a convenient method for computing equilibria and
succinctly summarizing how the equilibrium correspondence varies as we
change the parameters of the model (i.e., comparative statics). To explain
the algorithm, let X VF ; VA ; x ; z ; d denote the equilibrium vector,
and let H1 = {X*|H(X*|)=0} denote the equilibrium correspondence.
To explore this correspondence, we follow paths along the surface by
varying a single parameter, such as . The parameter that is varied is
known as the homotopy parameter. The homotopy method starts with a
pair of functions (X*(s), (s)) H1 given parametrically by a scalar s,
which implies H(X*(s)|(s), /)=0, where / is the set of all parameters
remaining fixed as varies. To remain on an equilibrium path, it is

necessary that:

H X sjs; = dX s H X sjs; = 0
s 0 A:1
x ds
HX sjs;= HX sjs;=

where
x is the 8R 8R Jacobian, dXdss and
are 8R 1 vectors, and 0 (s) is scalar valued. This is a system of 8R 1
differential equations that must be solved in order to identify a path.
The homotopy algorithm is not guaranteed to find all the MPE. This is
because H1 may contain equilibria that are off the main path. To identify
additional equilibria, we use the Pakes-McGuire algorithm at a variety of
different starting values. In addition, we can choose other parameters
besides to be the homotopy parameter. By using, for example, as the
homotopy parameter for a fixed value of , we can crisscross the
parameter space by using equilibria on the paths to generate paths with
respect to . A path must either intersect with all paths, or they will
lead us to additional equilibria that in turn can give us an initial condition
to generate an additional path.
To compute equilibria, we use Hompack written in Fortran 90. Our
programs are available upon request.
Proofs of Propositions
Proof of Proposition 1:
Characterization of the solution to the firms problem and proof that the
solution is unique:
When there is no activist, Eq. (2) simplifies to
2

e0 R c
xF
1 F VF R max kx VF R 1 VF R

4 x0 1 x
A:2
and the KuhnTucker condition becomes
F
VF R 1 VF R k A:3
1 x2

which holds with equality if x*(R) > 0. Now, at R R, VF R 1 VF R
2

e0 R c
so x R 0. Substituting x = 0 into Eq. (A.2) implies VF R 4 1 .
F
The derivation of Eqs. (14) and (15), and the proof that the MPE
is unique, is by induction.
Consider an arbitrary R < R, and suppose
x*(R + 1) and VF R 1 are the unique (x, VF) satisfying satisfy Eqs. (14)
and (15). Now, for state R, consider the maximization problem in
Eq. (A.2). The solution to this maximization problem is:
8
>
>
>
> 0 if F VF R 1 VF < 1
>
> k
>
< 0 1
8 91
x
B<
=
2
> C
> 1 B F VF R 1 VF
>
> 1C if F VF R 1 VF 1
>
> @
: k ; A k
>
:
A:4
but this expression is equivalent to Eq. (14). If we substitute Eq. (A.4) into
Eq. (A.2) and rearrange terms, we get Eq. (15).
To prove that the solution to Eqs. (14) and (15) is unique, notice that
Eq. (A.4) which, recall, is equivalent to Eq. (14) traces out a locus in
(x, VF) space. This locus has two pieces. For VF > VF R 1 k , this locus
F
coincides with the vertical axis. For VF VF R 1 k ; the locus is traced
F
out by the equation VF# x VF R 1 k1x

2
F , which is strictly decreasing

in x when x 0. Condition (15) also traces out a locus in (x, VF) space, and
e R c
2
this locus, denoted by V^ F x 4 0 1 1 kx2

, is monotone increasing in
F F
e R c
2
x and takes on a value of 4 0 1 when x = 0. There are two possibilities.

F
e0 R c
2

If 4 1 >VF R 1 , the intersection of the two loci occurs at x = 0 and
k
F F
e0 R c e R c e R c
2 2 2
VF 4 1 , so x*(R)=0 and VF R 4 0 1 . If 4 0 1 <VF R 1 k ,

F F F F
the intersection of the two loci occurs at x such that where V^ F x

e R c
2
kx2
VF# x 4 0 1 1 k1x 2
VF R 1 0. It is straightforward to
F F F
establish that this quadratic equation has a unique positive root, x*(R).
This, in turn, implies that VF R is unique.
Proof that the firms value function is strictly increasing in R: Suppose, to
the contrary, that VF R 1 VF R. From the KuhnTucker conditions
in Eq. (A.3), it would follow that x*(R)=0, and from Eq. (15) in state R,
we would have:

2
e 0 R c
VF R A:5
4 1 F
Now, since x*(R + 1) 0, Eq. (15) in state R + 1 implies that

2
e 0 R 1 c
VF R 1 A:6
4 1 F
Comparing Eqs. (A.5) and (A.6), we have VF R 1 > VF R, a

contradiction.
Proof that the firms externality-reducing activity is nonincreasing in R:
The proof is by induction. Note that 0 x R x R 1, establishing
the result at R R 1. Assume, then, x*(R + 2) x*(R + 1). There are
two cases to consider: x*(R) > 0 and x*(R)=0. Consider the first case,
x*(R) > 0. In this case, we want to establish that x*(R + 1) < x*(R).
Suppose, contrary to what we want to prove, that x*(R + 1) x*(R). This,
then, implies x*(R + 1) > 0, so x*(R + 1) must therefore satisfy Eq. (A.3)
with equality in state R + 1:
F
VF R 2 VF R 1 k A:7
1 x R 12
Similarly, since x*(R) > 0:
F

VF R 1 VF R k A:8
1 x R 2
Now, from Eq. (15) we have
VF R 2 VF R 1

2
2
2
e0 R 2 c e 0 R 1 c k x R 2 kx R 12

4 1 F 4 1 F 1 F 1 F
A:9
VF R 1 VF R

2
2
e 0 R 1 c e0 R c kx R 12 kx R2
A:10
4 1 F 4 1 F 1 F 1 F :
Substitute Eq. (A.9) into the left-hand side of Eqs. (A.7) and (A.10)
into the left-hand side of Eq. (A.8), equate the resulting expressions, and
rearrange terms to get:
2 (
2 ) 3
2 e0 R 2 c
6 1 x R 1
2 7
6 7
6 ( e0 R 1
)c 7
6 2 7
4 e R 1 c 5
1 x R2 0

2
e0 R c
2 3

2 x R 1
2

6 1 x R 7
x R2
4k6 4
7
5 A:11
2 x R 2
2

1 x R 1
x R 12
Now, by assumption, x (R + 1) x*(R), so [1 + x*

2 2
(R + 1)] [1 + x*(R)] . Moreover, given our parameter conditions,
(e0Rc)2 is an increasing, strictly concave function of R, so
(
2 ) (
2 )
e0 R 2 c e0 R 1 c
0<
2 <
2
e 0 R 1 c e 0 R c
Thus, the left-hand side of Eq. (A.11) is strictly negative, so

2 x R 12 2 x R 22
1 x R < 1 x R 1 0
x R2 x R 12
A:12
where the second inequality in Eq. (A.12) follows because, by the induction
hypothesis, x*(R + 2) x*(R + 1). But Eq. (A.12) implies x*(R + 1) <
x*(R), which contradicts the assumption that x*(R + 1) x*(R). Thus, it
must be the case that x*(R + 1) < x*(R) for the case of x*(R) > 0.
Consider, now, the second case: x*(R)=0. In this case, we want to establish
that x*(R + 1) x*(R), which could only hold if x*(R + 1)=0. So, suppose,
to the contrary, that x*(R + 1) > 0. Since x*(R)=0, it follows from Eq. (15)
e R c2 e R1 c
2

that VF R 4 0 1 . Moreover, VF R 1 04 1 kx1
R12
. Thus,
F F F
we get the following chain of implications:
8 9

< e R 1 c
2 kx R 12 e R c
2 =
0 0
F VF R 1 VF R F
: 4 1 F 1 F 4 1 F ;
8 9
< e R 1 c
2 kx R 22 e R c
2 =
0 0
F
: 4 1 F 1 F 4 1 F ;
8 9
F < e R 1 c
2 kx R 22 e R c
2 =
0 0
>
1 x R 1 : 4 1 F 1 F 4 1 F ;
8
2
2 9
F < e R 2
c
kx R 2 2
e R 1
c =
0 0
>
1 x R 1 : 4 1 F 1 F 4 1 F ;
8
2
9
>
>
> e0 R 2 c kx R 22 > >
>
>
> >
>
F < 4 1 F 1 F
=
>
1 x R 1 >
> e0 R 1 c
2
kx R 12 > >
>
> > >
>
: >
;
4 1 F 1 F
F

VF R 2 VF R 1
1 x R 1
k
The inequality in the second line follows from the induction hypothesis
that x*(R + 1) x*(R + 2). The inequality in the third line follows because
F > 1xFR1

, since x*(R + 1) > 0 by assumption. The inequality in the
fourth line follows because [e0R c]2 is strictly concave in R. The inequality
in the fifth line follows because x*(R + 1) > 0. The equality in the sixth line
follows from Eq. (15), while the equality in the last line follows from the
first-order condition for x in state R + 1. But the implication of this chain of
inequalities is that F VF R 1 VF R > k, which contradicts the fact

that when x*(R)=0,
the KuhnTucker condition would imply
F VF R 1 VF R k. Summarizing, we have shown that if x*(R)>0,
then x*(R) > x*(R + 1), and if x*(R)=0, then x*(R + 1)=0. This is what we
wanted to prove.
Proof that the firms value function is strictly concave in R
From the KuhnTucker condition (A.3)
k1 x R2
VF R 1 VF R A:13
F
k1 x R 12
VF R VF R 1 A:14
F
There are three cases to consider. First, suppose x*(R 1) and x*(R) are
both positive. Then, the above conditions hold with equality. We proved
above that x*(R 1) > x*(R), which immediately implies
VF R VF R 1 > VF R 1 VF R.
Second, suppose that x*(R)=0, but x*(R 1) > 0. Then, Eq. (A.14)
holds with equality, while Eq. (A.13) holds with inequality. This

implies VF R VF R 1 k1x R1
> k1x R
2 2
VF R 1 VF R.
F F
Third, suppose that x*(R 1)=x*(R)=0. We have established that x*() is
nondecreasing, so it would follow that x R 1 x R x R 1
x R 0. In this case, then from condition (15), we have VF R 1
e0 R c e0 R1 c
2 2
e0 R1 c2
41F
, V F R 41F
, and V F R 1 41F
. Given our assump-
2
tions on , e0 R c is a strictly concave function in R, so
VF R 1 VF R < VF R VF R 1 in this case as well.

Suppose, to the contrary, that x*(R) is strictly increasing in R for all R. We
will show that the solution to this problem is z = d = 0 in all states, which,
in turn, will imply that x*(R) could not be strictly increasing in R. We
begin by noting that if the activist sets z = d = 0 in all states, then Eq. (5)
implies that the activists value, denoted by VA0 R, is given by the recursion:

VA0 R ux R A F x RVA0 R 1 1 F x RVA0 R A:15

At R R,

u x R
VA0 R A:16
1 A

since VA0 R 1 VA0 R . Now, in state R 1, the recursion in Eq. (A.15)
is given by:

VA0 R 1 u x R 1 A F x R 1 VA0 R 1 F x R 1
VA0 R 1
Rearranging terms gives us:

u x R 1 A F x R 1 VA0 R
VA R 1
0

1 A 1 F x R 1 1 A 1 F x R 1

Substituting Eq. (A.16) in place of VA0 R in the above expression, and
rearranging terms, gives us:

1
VA0 R 1 1 Ru x R 1 1 1 R u x R A:17
1 A
where 1 R 1A
1A A F x R1
0; 1. Since u(x) is nondecreasing
and because we assumed x*(R1) < x*(R),
have Eq. (A.17)

ux R
implies VA0 R
1 VA0 R
1A
Consider, now, the recursion for VA0 in state R 2:

VA0 R 2 u x R 2

A F x R 2 VA0 R 1 1 F x R 2 VA0 R 2
Rearranging terms gives us

1
VA0 R 2 2 Ru x R 2 1 2 R VA0 R 1
1 A
where 2 R 1A
1A A F x R2
. Substituting Eq. (A.17) into the above

expression for VA R 2 yields
0
2 3
2 Ru
1 x R 2
VA R 2 4 5 A:18
1 Ru x R 1
0
1 A 1 2 R
1 1 R u x R

R0 2 ux R 1 ux R, Eqs. (A.17) and (A.18) imply
Since ux
VA0 R 2 VA R 1 . Reasoning inductively in this fashion for all R tells
us that when z = d = 0 for all R,

VA0 1 VA0 R
Given this, along with Eqs. (8) and (9), the activists marginal benefit for
z is nonpositive for all z > 0, and the activists marginal benefit for d is also
nonpositive for all d > 0. This implies that z*(R)=d*(R)=0, for all R.
Thus, if x*(R) is strictly increasing, the activist will not engage in criticism
or confrontation in any state.
However, if the activist sets z = d = 0 in all states, the firms maximiza-
tion problem is solved by choosing the level of externality-reducing activity
as in the no-activist case. By Proposition 1, we have seen that x*(R) in that
case is nonincreasing, which contradicts our assumption that x*(R) is
monotone increasing in R.
Proof of Lemma 1
With z(R)=0 and an exogenous perturbation n > 0 in state Rn, the firms
optimization problem in state Rn can be written as
2
e0 Rn c
VF Rn max kx F 1 n VF Rn
x0 4
n RX n 1
F 1 n F xVF Rn 1 VF Rn F VF r
Rn 1 r1
A:19
The firms optimization in a nonfocal state R f1; ; Rn1 ; Rn 1; ;

R 1g is:
2
e 0 R c
VF R max kx F VF R F F xVF R 1 VF R
x0 4
A:20
The firms optimization in state R is
2

e0 R c
VF R max kx F VF R A:21
x0 4

The solution in state R is x R 0, and from this it follows that
2
e0 R c VF R
VF R 4 1 ; which is independent on n . Thus, 0.36
F n

Now consider the firms optimization in a nonfocal states R > Rn and

R + 1 > Rn. In both states, differentiate condition (A.20) with respect to
n , and evaluate at n 0. Utilizing the envelope theorem and rearranging
terms gives us:

VF R F F x0 R VF R 1 VF R

n n 0 1 F n n
0

VF R 1 F F x0 R 1 VF R 2 VF R 1
n n 1 F n n
0
0
where x0 R is the level of externality-reducing activity chosen by the firm

when there is no activist. Subtracting these expressions and rearranging
terms gives us:

VF R 1 VF R F F x0 R 1 VF R 2 VF R 1
n n 1 1 x R
n n
0
F 0 F 0

A:22
Evaluating this at R R 1 gives us

VF R VF R 1 F F x0 R
n n 1 1 x R 1

0 F 0 F

VF R 1 VF R
n 0
n 0

where the second equality follows from the fact that x0 R 0. Using
Eq. (A.22), we can reason recursively and deduce that
VF R1VF R n
n 0 0 for all R > Rn. This establishes part (a) of Lemma 1.
Now in the focal state Rn, differentiate Eq. (A.19) with respect to n ,
utilize the envelope theorem, and evaluate at n 0 to get:
8 2 3
< PRn 1
VF Rn 4 V r 5 VF Rn
F VF0 Rn r1 F0

n n 0 : Rn 1 n n 0

F x0 Rn VF Rn 1 VF Rn
9
=
VF Rn 1 VF Rn
F x0 Rn A:23

n n ;
0
Now, do the same for state Rn + 1 and rearrange terms:

VF Rn 1 F F x0 Rn 1 VF Rn 2 VF Rn 1
n 0
n n 0 1 F n 0
where the equality to zero follows from the earlier result that
VF R1VF R
0 for R > Rn. Using this, and rearranging terms in
n
Eq. (A.23) implies

VF Rn
n n 0

PRn 1

VF0 r
F F x0 Rn VF Rn 1 1 F x0 Rn VF0 Rn r1 Rn 1

<0
1 F 1 F x0 Rn
PRn 1

VF0 r
since F x0 Rn VF Rn 1 1 F x0 Rn VF0 Rn r1 Rn 1 > 0 from

the result in Proposition
1 that V
F0 nR is monotone increasing in R. Thus,
VF Rn 1VF Rn n
n 0 > 0 , establishing part (b) of Lemma 1.

Finally, consider states R < Rn. Differentiating Eq. (A.20) with respect to
n in these states and using the envelope theorem gives us, as before,

VF R 1 VF R F F x0 R 1 VF R 2 VF R 1

n 1 1 x R
n
n 0 F 0 F n 0

Evaluating this at R = Rn 1 yields

VF Rn VF Rn 1
n n
0

F F x0 Rn VF Rn 1 VF Rn

n >0
1 1 F x0 Rn 1 F n 0

VF Rn 1VF Rn n
since we have already established that n 0 > 0. Recursively
applying this in all states below Rn establishes part (c) of Lemma 1.
Note that if n 1, then MBx x; VF Rjn ; n jRn < 0. This implies
x*(Rn)=0. Since, by assumption, x0 R > 0, the perturbation unambiguously
reduces the firms choice of x in this state.
This result follows directly from the data in Property 2 of Table 2.
SELF-REGULATION AND
REGULATORY DISCRETION:
WHY FIRMS MAY BE RELUCTANT
TO SIGNAL GREEN
Thomas P. Lyon and John W. Maxwell
ABSTRACT
A large literature studies why firms self-regulate and signal green.

However, it has ignored that regulators have enforcement discretion, and
may act strategically. We fill this gap. We build a game theoretic model of
whether a firm should signal its type through substantial self-regulation.
We find self-regulation is a double-edged sword: it can potentially preempt
legislation, but it can also lead regulators to demand higher levels of compli-
ance from greener firms if preemption fails. We show how self-regulatory
decisions depend upon industry characteristics and political responsiveness to
corporate environmental leadership. We have made a number of simplifying
assumptions. We assume activist groups cannot challenge regulatory flexibil-
ity in court, and that regulatory penalties are fixed and are not collected by
the regulator. Firms with low compliance costs confront a tradeoff regarding
self-regulation. They can blend in with the rest of the industry, and take
few self-regulatory steps. This reduces the risk of regulation somewhat, and
preserves their ability to obtain regulatory flexibility should regulation be

ISSN: 0742-3322/doi:10.1108/S0742-332220160000034009
301
302 THOMAS P. LYON AND JOHN W. MAXWELL
imposed. Alternatively, they can step up with substantial self-regulation.

This better mitigates the risk of regulation, but at the risk of signaling low
costs and becoming a target for stringent enforcement should regulation
pass. Recent work has found negative market reactions to corporate claims
of voluntary emissions reductions, despite the conventional wisdom that it
pays to be green. We offer a new explanation to scholars and managers:
regulatory discretion may undermine the ability of industry self-regulation
to profitably preempt mandatory regulatory requirements.
Keywords: Self-regulation; signaling; regulatory discretion;
asymmetric information; private governance
JEL classifications: D83; L31; M14; Q56
INTRODUCTION
It has become a commonplace among consultants and popular business
writers that it pays to be green (Esty & Winston, 2006; Porter & Van der
Linde, 1995), and indeed there is a significant academic literature that finds
a positive association between environmental and financial performance
(Klassen & McLaughlin, 1996; Margolis & Walsh, 2001). However, the real
question for managers is when does it pay to be green? (King & Lenox,
2001), and scholars are still working out the answer. Recent work has
found insignificant average market reactions to recognition granted
by third parties for environmental performance, and negative reactions to
corporate claims of voluntary emissions reductions and awards from
nongovernmental sources (Fisher-Vanden & Thorburn, 2011; Jacobs,
Singhal, & Subramanian, 2010; Lyon, Lu, Shi, & Yin, 2013). However, the
explanation for these varying outcomes remains unclear. We focus on a
factor that has received relatively little attention: regulatory discretion.
Much of the literature on industry self-regulation argues that it can profit-
ably preempt mandatory regulatory requirements (Delmas & Montes-Sancho,
2010; King & Lenox, 2000; Manzini & Mariotti, 2003; Maxwell, Lyon &
Hackett, 2000; Segerson & Miceli, 1998; Short & Toffel, 2010). In addition,
some lines of work have emphasized the private benefits for firms who partici-
pate in voluntary self-regulatory schemes in order to signal that they are
green (Darnall & Carmin, 2005; Potoski & Prakash, 2004; Prakash & Potoski,
2005), or elaborated the various ways in which self-regulation can provide
Self-Regulation and Regulatory Discretion 303
private benefits by changing the nature of regulatory threats or their enforce-

ment (Decker, 2003; Delmas & Terlaak, 2001; Gray & Shimshack, 2011;
Heyes, 2005; Innes & Sam, 2008; Maxwell & Decker, 2006; Morantz, 2009).
Regardless, self-regulation is typically seen as a profitable way to influence the
business environment of the firm, which makes negative market responses to
it a puzzle for management scholars.
We argue that regulatory discretion holds a key to unlocking this puzzle.
Most of the literature on self-regulation assumes that the stringency of a
regulatory threat is fixed and known to the firm, even if there is uncertainty
about whether enabling legislation will be passed. In practice, however,
regulators have substantial discretion in how they implement and enforce
legislation (Besley & Coate, 2003; Lyon & Mayo, 2005; Olson, 1995;
Shimshack & Ward, 2005). Firms that signal green may thus be treated
differently by regulators from other firms, and from the perspective of
strategic management it is important to understand whether this special
treatment is advantageous or not. As mentioned above, a significant strand
of prior work has argued that regulators treat environmental leaders more
favorably. However, there is also a risk regulators which will infer that
leaders have low costs of abatement, and expect higher performance from
them, a phenomenon sometimes referred to as the ratchet effect (Laffont &
Tirole, 1988; Weitzman, 1980). Our paper explores how managers should
approach the decision to self-regulate when regulatory discretion influences
the way regulatory threats will actually be enforced.
Regulatory discretion arises in many different regulatory settings.
Discretion in environmental enforcement has perhaps attracted the most
attention (Gray & Deily, 1996; Shimshack & Ward, 2005). However, health
and safety regulations also allow for substantial discretion (Olson, 1995;
Verkuil, 1982, p. 249), commentators note that OSHA, like other regulators,
can (and already does) induce additional compliance by reducing penalties it
has levied against an employer in return for the firms agreement to undertake
additional actions to protect workers (Shapiro & Rabinowitz, 2000, p. 150).
In addition, financial regulations include flexibility provisions for small or
community banks.1
Regulators often pay special attention to the needs of small- and medium-
sized enterprises (SMEs), for whom regulation is thought to be especially
burdensome.2 Indeed, sometimes regulators are required to offer flexible
enforcement, as under the U.S. Regulatory Flexibility Act of 1980 (RFA),
which codifies a federal preference for regulatory flexibility in order to protect
small businesses from the potentially high costs of compliance with regulation
(Regulatory Flexibility Act, 5 U.S.C. 601612). Regulatory discretion is far
from a uniquely American phenomenon. In 2003, the EU took steps along

lines similar to the RFA, creating a European Business Test Panel to
examine the potential impact of legislative and regulatory initiatives on
small businesses, followed, in 2007, by an Action Programme for Reducing
Administrative Burdens in the European Union, which includes a series of
specific tests of those burdens on SMEs. Similar policies have been studied by
the OECD, and are being discussed in the organization for Asia-Pacific
Economic Cooperation (APEC) and the Association of Southeast Asian
Nations (ASEAN).
There are many different ways to implement regulatory discretion. It is
well documented that regulators are less likely to impose penalties on firms
that are major employers in a given area or that are experiencing financial
duress (Gray & Deily, 1996). Another familiar form of discretion is the
bubble concept utilized by the EPA, which allows businesses some flex-
ibility in reducing total emissions by placing an imaginary bubble over an
entire plant and demanding only that overall emissions levels meet estab-
lished standards rather than requiring each stack within the plant to be of
the most efficient design. (Verkuil, 1982, p. 226). As Maloney and Yandle
(1980) point out, bubbles can be crafted in a variety of ways, and can be set
at the level of an entire company as well as at the level of a single plant.3
Another structured approach is the use of regulatory tiering, which treats
firms differentially based on their size. Tiering can be achieved in numerous
different ways, including less frequent inspections, lighter fines for non-
compliance, exemptions, waivers, reduced requirements, or simpler report-
ing requirements for certain types of firms (Brock & Evans, 1985).
Prior theoretical work shows how regulatory discretion can be used to
improve regulatory outcomes (Brock & Evans, 1985; Harrington, 1988;
Heyes & Rickman, 1999). However, this work has focused on the regulators
perspective and ignored the implications of regulatory discretion for strategic
management. In this paper, we extend the literature on self-regulation by
studying a setting with regulatory discretion, in which the public and private
benefits of self-regulation may be in tension. We characterize when these
conflicting incentives motivate firms to undertake substantial self-regulation,
and when they lead instead to modest levels of self-regulation. In contrast to
prior work, we show that signaling green may subject low-cost firms
to more stringent regulation, so they may prefer to pool with brown (i.e.,
high-cost) firms.4
Our analysis has important managerial implications for firms considering
self-regulation. As mentioned above, the existing literature has emphasized
that self-regulation can stave off stakeholder pressure and potentially
preempt regulation. These benefits suggest that managers should find self-
regulation, a very attractive strategy. Our analysis shows that matters are
more complicated. We recognize that preemption does not always succeed,
and we show that if it fails then self-regulation may influence the implementa-
tion of regulation in a way that raises costs for the self-regulating firm. In parti-
cular, low-cost firms confront a tradeoff regarding self-regulation when most
firms in the industry have high compliance costs. On one hand, they can blend
in with the rest of the industry and take modest self-regulatory steps. This does
little to reduce the risk of regulation, but it preserves their ability to take
advantage of regulatory discretion should regulation be imposed. On the other
hand, they can step up and take substantial self-regulatory steps. This reduces
the risk of regulation, but at the cost of signaling that the firm can meet
stringent regulations (signaling green) and thereby subjecting it to tougher
regulatory enforcement should regulation pass. Overall, then, our results point
out that the self-regulatory decision is complex, and explain why managers
may not always be rewarded for self-regulation.
The remainder of this paper is organized as follows. The next section
lays out the basic structure of the game between the regulator and the firm.
The subsequent three sections then work through the stages of the game in
reverse chronological order, starting with the regulators decision regarding
how to enforce the law, moving on to the regulators decision regarding
how much effort to put into achieving passage of the law in the first place,
and then the firms incentives to engage in self-regulation. After that, we
present a simple example with functional forms that allow for closed-form
solutions, discuss our results in the context of the literature more generally,
and conclude.
THE MODEL
In this section, we present a simple model of regulatory discretion, within a
setting where firms may choose to undertake self-regulatory measures to
reduce the probability of legislation. If legislation is passed, the regulator
has discretion regarding whether to enforce it in a rigid or a flexible fashion.
Industry expectations about enforcement affect firms incentives for self-
regulation, and in turn, industry self-regulation influences the effort with
which the regulator pursues passage of legislation. Thus, the presence or
absence of regulatory discretion can ultimately affect the likelihood that
legislation is passed. To fix ideas we write as if the regulation under consid-
eration requires the abatement of pollution.
We assume the imposition of regulation is uncertain and occurs with a

probability that is a function of regulatory effort. One interpretation is that
enabling legislation may not require or even authorize regulatory action,
especially if the regulator puts forth little effort to ensure that it does. For
example, the Supreme Courts decision in Massachusetts v. EPA, 549 U.S.
497 (2007), left the decision whether to regulate greenhouse gases as an air
pollutant in the hands of the EPA. A second interpretation is that even
though legislation has been passed, the regulator may lack the resources to
enforce it in all applications, so firms are uncertain whether it will be
enforced against them. For example, the Clean Water Act of 1972
mandated that all navigable waters of the U.S. must be fishable and swim-
mable by 1983, but by the mid-1990s the number of river miles meeting
these requirements had increased by only 6.3 percent and 4.2 percent,
respectively (Freeman, 2002); the Environmental Protection Agency (EPA)
had nowhere near enough resources to achieve full implementation. A third
interpretation is that even though legislation is passed, and the regulator
has adequate resources for implementation, court decisions may block the
imposition of regulation in certain circumstances. For example, questions
over what constitutes a navigable waterway have hampered the ability of
the EPA to apply the Clean Water Act to many bodies of water within the
United States (Duhigg & Roberts, 2010).
The model involves two strategic players: a polluting firm (he) that
desires to minimize costs and a regulator (she) that desires to maximize
environmental quality. Following Heyes and Rickman (1999), we assume
the firm has environmental impacts on two domains, j = 1, 2, and imposes
environmental damage d in each domain unless he undertakes abatement.
These domains can represent two different pollutants, two different plants, or
two different media. (In the case of occupational safety and health, the
domains might be different types of injuries or safety requirements.)
Regardless, this multi-faceted character is an important feature of most real-
world environmental legislation that is typically ignored in theoretical models.
Heyes and Rickman (1999), in contrast, present a model of multi-dimensional
regulation in which firms face a penalty F for noncompliance on each dimen-
sion. Firms differ in their costs of compliance, which are private information,
and in the model high-cost firms opt to pay a penalty rather than comply with
regulations. The regulator aims to reduce environmental damage, and may
offer firms a deal in which they are forgiven for being out of compliance on
one dimension if they comply on the other dimension. (Gray & Scholz, 1993)
find that over 50% of firms inspected by OSHA have violations, but that only
about 30% are actually assessed penalties.) This discretion increases overall
compliance if the share of high-cost firms is large enough.5
Firms have either low costs or high costs of abating their emissions.
Low-cost firms comply with regulations on both domains and high-cost
firms do not comply on either domain unless they are offered regulatory
flexibility. Flexibility allows the high-cost firms to escape the full burden of
the regulation while still inducing some compliance. Assuming compliance
costs are equal in both domains is the simplest way to achieve this setup.6
Firms of type i = L have low cost cLj=cL per unit of abatement for j = 1, 2,
and firms of type i = H have high cost cHj = cH per unit of abatement for
j = 1, 2. The fraction of firms that have low costs is . The ex ante value of
is known to the regulator, but the compliance costs of each individual firm
are private information.7 Let (qi1, qi2) be a vector of abatement decisions for
a firm of type i with qij [0, d] the level of abatement achieved on domain j.
Regulation (if enforced inflexibly) requires a firm to eliminate emissions on
both domains, and there is a penalty F for noncompliance onPeach domain.
A firms objective is to minimize the sum of abatement costs 2j 1 ci qij plus
any penalties due to noncompliance if regulation is enforced. Thus, a firm
complies on domain j if Cijd F. We assume cLd < F < cHd, so a low-cost
firm complies on both dimensions, incurring cost 2cL, but a high-cost firm
complies on neither, incurring penalties 2F. Under inflexible regulation, the
expected amount of abatement is 2d.
The Pregulators
objective
is to minimize
P expected
environmental damages
D 2j 1 d qLj 1 2j 1 d qHj . An important feature of
our model is that the regulator does not include the penalty revenue from
noncompliance in her objective function; she is focused solely on environmen-
tal compliance plus her cost of effort. This may seem like a strong assumption,
but it is also realistic environmental enforcement agencies typically do not
keep fines they impose on firms that are out of compliance. Even if fines could
be designated for environmental remediation, our assumption captures the
regulators preference for mitigation over remediation.
The regulator can choose to implement the regulation inflexibly, in the
traditional manner, or can choose to grant the firm flexibility in complying
with it. Following Heyes and Rickman (1999), we model this flexibility as
allowing the regulator to waive penalties if the firm complies on at least one
domain. We assume cHd < 2F, so the high-cost firm prefers to comply on
one dimension rather than pay the penalty on two dimensions. Why would
the regulator ever find this type of flexibility desirable? If the regulator knew
that the firm was of type L, then of course she would never offer flexibility.
But if the firm is known to be of type H, then without flexibility the firm will
fail to abate on either domain, while with flexibility the firm complies on one
domain, generating environmental improvement d. When the regulator is
uncertain of the firms type, she can choose between traditional enforcement,
which generates expected abatement 2d, or regulatory flexibility, which gen-
erates abatement d for certain. Thus, regulatory flexibility is environmentally
preferable if < 1/2.
Of course, the firm is also free to undertake voluntary abatement in an
attempt to reduce the likelihood of legislation. The linearity of the cost
function implies that his costs are the same for any emissions reductions
R1 and R2 on the two domains that sum to R. (We show below that the
firm may have strategic incentives to choose a heterogeneous mix of abate-
ment across the two domains.) Then firm is minimized cost of meeting a
given abatement target is simply ciR. With some level of abatement granted
voluntarily by the firm, the regulator may decide that other priorities are
more urgent and reduce her efforts to implement rules in a way that would
produce even greater reductions.
The game unfolds in three stages that occur subsequent to natures
initial move determining the firms cost of compliance.
1. The firm can self-regulate at a level [R1, R2].
2. The regulator attempts to infer the firms type, and nature reveals
whether regulation will be imposed.
3. If regulation is imposed, the regulator decides whether to offer regulatory
flexibility. If she does not, then the firm either complies at abatement level
d on each domain or pays a fine F on each domain on which he is out of
compliance. If she does offer flexibility, compliance is waived on one
domain and the firm complies on the other domain.
As is typical in game theoretic models, we solve the model in reverse chron-
ological order so as to obtain a Perfect Bayesian Equilibrium.
REGULATORY FLEXIBILITY AND FIRM COMPLIANCE

At stage 3 of the game, the regulator can choose to enforce the legislation
in the traditional manner, which means inspecting the firm and imposing a
penalty of F for each failure to comply. Alternatively, the regulator can
give the firm flexibility to comply on the domain of its choice while the
regulator agrees to waive the penalty on the other domain. As noted above,
the regulator will not offer flexibility to a firm known to be of type L, but
will offer flexibility to a firm known to be of type H. If the regulator is
uncertain of the firms type, traditional enforcement generates expected abate-
ment R + [2dR] = 2d + (1)R, while flexibility generates abatement
d + min[R1, R2] for certain. Then flexibility is environmentally preferable if
d + min[R1, R2] > 2d + (1)R, or if
< ~ d maxR1 ; R2 =2d R 1
Thus, we have the following proposition.

Proposition 1. At stage 3, if the regulator is certain of the firms type
then she offers flexibility iff the firm is of type H. If the regulator
is uncertain of the firms type then she offers flexibility
iff < ~ d maxR1 ; R2 =2d R 1=2:
Proof. Suppose that ~ > 1=2. In this case, we would have [2dR] 2d
2 max[R1, R2]), which simplifies to 2 max[R1, R2] R or max[R1, R2] R/2.
But R = R1 + R2, so max[R1, R2] R/2. Hence, ~ > 1=2 is false, and we
must have ~ 1=2.
The regulators decision to offer flexibility depends on the mix of the
firms self-regulatory abatement choices across the two domains. If the firm
does not invest in self-regulation, so that R = 0, the expected benefit of reg-
ulation is 2d without flexibility, and d with flexibility. Thus, ~ 1=2 when
R = 0 and the regulator offers flexibility if and only if < 1/2. If the firm
increases his self-regulatory abatement equally across the two domains,
then it remains true that the regulator will offer flexibility if and only if
< 1/2, because ~ remains equal to 1/2. Now consider what happens if the
firm allocates his investment in self-regulation differentially across the two
domains. As R increases, the change in ~ is given by
8
> d maxR1 ; R2
>
> if Ri Rj
~ < 2d R2

Ri >
> 1
>
: if Ri > Rj
2d R
Note that the regulator responds very differently to increases on the two
domains, raising ~ when self-regulation is increased on the lesser domain
(the domain with less abatement), and lowering ~ when self-regulation is
increased on the greater domain. The reason is that if regulation is imposed

and flexibility is granted, the firm can strategically take advantage of an
unequal division of abatement. With flexibility, the firm will elect to comply
on the domain with greater self-regulation, which means that the regulator
gains less on the compliance domain than if self-regulation were split
equally. As a result, when self-regulation is divided unequally across the
domains, the regulator has less incentive to grant flexibility, and restricts
the conditions under which she will do so, that is, she reduces the critical
threshold ~ below which she grants flexibility when the firms type is
unknown. Thus, as the firm shifts his self-regulatory abatement toward the
greater domain, this can cost him the possibility of regulatory flexibility,
since the range of values of for which the regulator grants flexibility
shrinks. At the same time, however, the regulators incentive to press for
regulation falls as the firm shifts his self-regulation toward the greater
domain, thereby reducing the likelihood that regulation will be imposed at
all. We examine the tradeoff between these two factors, and its impact on
the firms self-regulation decision, in the Self-Regulation section.
LEGISLATIVE STAGE
At stage 2, nature determines whether or not regulation will be imposed.
This probability is decreasing in the firms self-regulatory investment R. As
shown in Proposition 1, the regulators decision regarding flexibility depends
upon whether she knows the firms costs or not, that is, on whether stage 1
resulted in a separating equilibrium or a pooling equilibrium. Since we
are solving the game in reverse chronological order, we must consider both
possibilities at stage 2.
We assume the probability with which regulation is imposed is a function
of R, the firms level of self-regulation. The two types of firm either choose
the same level of self-regulation, in which case a pooling equilibrium occurs,
or they choose different levels, leading to a separating equilibrium. We
assume 0 (R) < 0 and (R) 0, with limR2d (R)=0. Thus, the probability
of regulation is decreasing in the level of industry self-regulation, but at a
decreasing rate.
An important element of the analysis to follow is the inverse hazard rate
1 R
0 R
The concept of a hazard rate derives from industrial engineering, where it

indicates the probability a machine will fail at time t given it has survived
up to that point. In our case, we can think of a failure from the firms
perspective as the imposition of regulation. Then (R)=1(R) is the
probability that the firm survives unregulated at abatement level R, and the
corresponding density function at R is 0 (R)=0 (R). We will assume this
density function is log-concave, in which case Theorem 3, Corollary 2
of Bagnoli and Bergstrom (2005) implies that the inverse hazard rate is
monotone increasing in R. This property is known as the monotone likeli-
hood ratio property (MLRP).
If both types of firm choose the same R at stage 1, then at stage 2
the regulator is uncertain of the firms type, and chooses whether to offer
flexibility based on the criterion in Lemma 1.
SELF-REGULATION
At stage 1, the firm must decide how much self-regulation to undertake
before legislation is proposed. There are two possibilities the firm must con-
sider, one in which the regulator offers flexibility and one in which she does
not. As shown above, the regulator prefers to offer flexibility to a high-cost
firm, but not to a low-cost firm. Because the firm knows its type, but the
regulator does not, the regulator must try to infer the firms type from its
level of self-regulation.
This is obviously a signaling problem, for which there can be different
types of equilibria. An equilibrium is a strategy for each type of firm, and a
set of beliefs on the part of the regulator that are consistent with these
strategies. A pooling equilibrium exists if each type of firm takes the same
strategy, and the regulator is unable to update her prior beliefs conditional
on the firms actions. This occurs when one type of firm prefers to mimic
the choice of the other type, so that the two types are treated identically,
rather than to make a choice that would reveal its type. A separating equili-
brium exists if each type of firm takes a different strategy, and the regulator
updates her beliefs, conditional on the firms actions, so that she knows the
firms type with certainty.
The nature of the equilibrium depends crucially on what the regulator
does if she is uncertain of the firms type, which will be the case at stage 1 if
both firm types choose the same level of self-regulation. As shown in
Proposition 1, the regulator offers flexibility in this situation if < ~ and
~ Thus, in the remainder of this section, we

refuses to offer flexibility if > .
analyze both these possibilities and characterize fully the conditions under
which each type of equilibrium exists.
~
Uninformed Regulator Offers Flexibility <
To begin, suppose that < . ~ Thus, if after self-regulation the regulator

remains uncertain of the firms type, then if legislation passes the regulator
offers flexibility to both types of firm. If a high-cost firm anticipates flexibility,
his expected profits are
HF R cH R RcH d maxR1 ; R2 2
The expression for expected profits implies that the marginal cost of invest-
ing in R1 and R2 is the same, but the marginal benefit of investing in
max[R1, R2] is greater than the benefit of investing in min[R1, R2] because
it reduces the incremental cost of complying with regulation, should it be
imposed. The first-order conditions are
d H R
cH 0 RcH d maxR1 ; R2 0 3
d minR1 ; R2
d H R
cH 0 RcH d maxR1 ; R2 RcH 0 4
d minR1 ; R2
Since 0, the second-order conditions are negative. Obviously, the two

first-order conditions cannot both hold. Investing in min[R1, R2] reduces
the likelihood of regulation, but does not change the incremental cost of
regulation, should it be imposed, since the firm will comply on the domain
in which it has done the most self-regulation. Investing in max[R1, R2]
also reduces the likelihood of regulation, but has the additional benefit of
reducing the incremental cost of regulation, should it be imposed. Thus,
the marginal benefit of abatement on the domain with more self-regulation
is always greater than the marginal benefit on the domain with smaller
self-regulation. We have thus established the following.
Lemma 2. The high-cost firm chooses min[R1, R2] = 0.
Without loss of generality, we will henceforth assume R2 R1 = 0 for
the high-cost firm, which means the firm elects to comply on domain 2 if
granted regulatory flexibility.8 Note that unlike the high-cost firm, the
low-cost firm is indifferent with regard to allocating self-regulatory effort
between the two domains. The reason is that the low-cost firm will
be required to comply with regulation on both dimensions, so that his
self-regulatory choices do not affect his costs at stage 3, unless there is a
pooling equilibrium. If the low-cost firm decides to pool with the high-cost
firm, he undertakes all self-regulation on domain two, just as the high-cost
firm would. Thus, we can rewrite the firms expected profits as
HF R2 cH R2 R2 cH d R2
Solving first-order condition (4) implies the solution is
1 RHF
RHF d 5
0 RHF
where the subscript H indicates the high-cost firm and the subscript F indi-
cates that the firm received flexibility.9 Denote the associated maximized
level of profits by
HF RHF cH RHF RHF cH d RHF 6
Substituting Eq. (5) into Eq. (6) and rearranging terms yields

1 RHF 2
HF RHF cH d 7
0 RHF
The low-cost firm could mimic abatement level RHF, receive flexibility,
and earn
LF RHF RHF cL d 1 RHF cL RHF
Alternatively, the low-cost firm could opt not to mimic. In this case the regula-
tor infers that the firm has low costs and requires compliance on both domains
if legislation passes. The low-cost firm would then earn expected profits
LN R R2cL d 1 RcL R
cL R N RcL 2d R
In a separating equilibrium, the low-cost firms first-order condition is
LN R
cL 1 0N R2d R N R 0 8
R
The optimal level of self-regulation thus balances the marginal cost of self-
regulation, cL (the increase in expected costs if regulation does not occur),
against the marginal benefit of self-regulation (the reduction in expected
costs if regulation is imposed). The marginal benefit has two components,
a reduction in the probability of legislation, and a reduction in the incre-
mental amount of abatement required if legislation passes. The optimal
solution is
1 RLN
RLN 2d 9
0 RLN
where the subscript L indicates the low-cost firm and the subscript N
indicates that there is no regulatory flexibility. An immediate observa-
tion follows.
Proposition 2. RLN > RHF.
Proof. We seek to compare the solutions to Eqs. (5) and (9), which can
be written as R = f(R) d + [1(R)]/0 (R) and R = g(R)=d + f(R),
where f0 (R)>0 due to the MLRP. Since g(R) > f(R)R, if we consider
the solution to Eq. (5) and shift f(R) upwards by d, the solution to
Eq. (9) moves along the 45 line and thus must be greater.
Intuitively, a firm that is known to have low costs will face more strin-
gent enforcement of regulation, and thus has stronger incentives to try and
preempt that regulation. A high-cost firm can expect more flexible and leni-
ent treatment by the regulator, and thus has less incentive to preempt. This
intuition is consistent with the empirical literature on voluntary approaches
to regulation, which consistently finds that large firms (which presumably
have lower costs) are more likely to participate in voluntary programs
(Lyon & Maxwell, 2004, 2007, 2008).
The corresponding maximized level of profits is
LN RLN RLN 2cL d 1 RLN cL RLN 10

Substituting Eq. (9) into Eq. (10) yields

1 RLN 2
LN RLN cL 2d 11
0 RLN
If self-regulation was completely ineffective at reducing the probability of

regulation (R), so that 0 (R)=0, then the firm would make no attempt to
preempt regulation, and would eschew self-regulation altogether by choos-
ing RLN = 0. Since this is not the case, the low-cost firm chooses a strictly
positive level of self-regulation, which is increasing in the level of damages
d as can be seen in Eq. (9). However, the firm always chooses a self-regulatory
level that is strictly less than 2d since otherwise the second term on the
right-hand side of Eq. (9) would be negative, requiring that (RLN) be
greater than one, which is impossible.
As mentioned above, for a pooling equilibrium to exist, the low-cost
type must prefer to receive flexible regulation and pool with the high-cost
type, that is we require
LN RLN < LF RHF 12
Intuitively, the firm prefers the pooling equilibrium if the cost of the
increase in upfront self-regulation outweighs the reduction in expected
incremental costs of regulation. As we have noted before, the latter has two
components, the change in the probability of regulation and the change in
the incremental abatement required if regulation occurs. Importantly, the
separating equilibrium actually imposes a greater demand for abatement,
in the amount 2d rather than d. This additional cost will only be incurred
by the low-cost firm if the (R) function is sufficiently elastic with respect
to self-regulatory investment. The following proposition characterizes the
low-cost firms choice regarding self-regulation.
Proposition 3. If a high-cost firm choosing RHF would be offered
flexibility, then the low-cost firm prefers to mimic the high-cost firms
self-regulatory choice, that is, LN(RLN) < LF(RHF).
Proof. From Eqs. (7) and (11),

1 RLN 2 1 RHF 2
L RLN L RHF cL d
0 RLN 0 RHF
The first two terms inside the braces on the right-hand side of the equation
are negative, but the third is positive. Proposition 2 shows that RLN > RHF,
and the monotone likelihood ratio property implies that [1(RLN)]/
0 (RLN) is increasing in R. Thus the sum of the second two terms inside
the braces is negative, making the entire right-hand side negative,
so L(RLN) < L(RHF).
The proposition shows that the low-cost firm prefers to pool with the
high-cost firm whenever it can. Intuitively, losing flexibility is very costly
for the low-cost firm, as it will have to spend an additional dcL on compli-
ance in the event regulation is imposed. Although it can offset this cost to
some extent by increasing its investment in self-regulation, this is not
enough to make it profitable to separate from the high-cost type, if we
make the mild assumption of log-concavity of the regulatory density
function in conjunction with limits on its magnitude as regulation is
completely preempted. The following proposition fully characterizes the set
~
of equilibria when < .
~ there is a unique pooling equilibrium in which
Proposition 4. If < ,
both types choose RHF and the regulator offers flexibility.
Proof. Proposition 1 shows that if < ,~ then a regulator who does not
know the firms type will offer flexibility. Lemma 4 shows that the low-
cost firm prefers to pool with the high-cost firm if he can thereby obtain
flexibility. Thus, no separating equilibrium exists, but the conditions for
a pooling equilibrium at RHF are met. A pooling equilibrium at another
~ can exist if the regulator holds off-equilibrium path beliefs
level, R;
that any deviation from R~ must be committed by a low-cost type firm.
However, equilibria at any R~ RHF are strictly dominated for both types
of firm.
The proposition establishes that when there is a large enough proportion
of high-cost firms, the only equilibrium involves pooling at the profit-
maximizing level of self-regulation RHF. This equilibrium is supported by
the regulators preference to offer flexibility when she is uncertain of the
firms type. As mentioned earlier, the experience with the EU Emissions
Trading System (ETS) provides support for the notion that firms may have
incentives to avoid disclosing their types to the regulator prior to the impo-
sition of regulation. Similarly, many electric utilities avoid improving their
older coal plants too much in order to avoid triggering New Source Review
requirements.
We turn next to the case where she prefers not to offer flexibility when
uncertain of the firms type.
~
Uninformed Regulator Does Not Offer Flexibility >
Turn now to the case where > . ~ A pooling equilibrium would exist if both
types of firm prefer to take the same choice of R conditional on the regula-
tors anticipated optimal response. When > , ~ Lemma 1 implies that the
regulator refuses to offer flexibility if both types of firm take the same
choice of R. As shown in the previous section, a low-cost firms expected
profits are maximized at RLN when he does not expect to receive flexibility.
Now consider the high-cost firms incentives. If legislation passes, he must
now either abate or pay the penalty F on each dimension. However, by
definition a high-cost firm prefers to pay the penalty rather than incur the
cost cHd. In this case, the high-cost firms expected profits are simply.
H R R2F cH R
Then differentiating its profits with respect to R yields
H R
0 R2F cH 13
R
Would the low-cost firm prefer to pool at this level? If he pools, his
expected profits are
L R RHN 2dcL 1 RHN cL RHN
In contrast, if he does not pool, his expected profits are
L R R2dcL 1 RcL R
R2d RcL cL R
which is maximized at RLN defined by Eq. (9). Since RLN RHN, there is
no pooling equilibrium when > .~
A separating equilibrium would exist if the two types of firms choose dif-
ferent levels of R, and the regulator tailors her response to the firms type,
offering flexibility to a high-cost firm but not to a low-cost firm. The analy-
sis in the previous section then shows that conditional on this regulatory
response, the low-cost firm would choose RLN and the high-cost firm would
choose RHN. Thus a separating equilibrium does exist.
The foregoing analysis has established the following result.
Proposition 5. If > ~ there is a unique separating equilibrium, with the
high-cost type choosing RHF and the low-cost type choosing RLN. The
regulator offers flexibility to the high-cost type and not to the low-
cost type.
Together, Propositions 4 and 5 completely characterize the nature of
equilibrium in the game. They are illustrated in Fig. 1, which plots the
firms choice of R on the vertical axis at the left-hand side of the figure and
, the probability the firm has low costs, on the vertical axis at the right-
~ there is a unique pooling equilibrium in
hand side of the figure. For < ,
which both types choose RHF and the regulator allows flexibility in enforce-
~ there is a unique
ment; it is indicated by the thick solid gray line. For > ,
separating equilibrium, with the high-cost type choosing RHF and the
low-cost type choosing RLN; these two strategies are indicated by the thick
dashed gray lines. In the separating equilibrium, the regulator offers
flexibility to the high-cost type but not to the low-cost type. In both types
of equilibria, a high-cost firm chooses the same level of self-regulation, so it
Separating RLN
Equilibrium
Pooling
Equilibrium
RHF
Fig. 1. Self-Regulatory Strategies.

is the strategic behavior of a low-cost type that is of primary interest. In

the first type of equilibrium, a low-cost firm prefers to stay in step with the
rest of the industry through modest levels of self-regulation, and thereby
ensure he will receive flexible regulatory treatment should legislation pass.
In the second type of equilibrium, a low-cost firm strategically signals his
type through substantial self-regulation, even though he foregoes regula-
tory flexibility should his self-regulatory efforts fail to preempt legislation.
Interestingly, this signaling benefits the entire industry by reducing the
probability of legislation. Furthermore, a high-cost type is assured of
receiving flexible regulatory treatment should legislation pass, so a low-cost
types signaling provides strictly positive gains to a high-cost type. This last
observation suggests a high-cost firm might have incentives to try and con-
vince the regulator that is high, so as to drive a low-cost type to undertake
more self-regulation. Although such persuasive efforts go beyond the scope
of our model, they may be of interest for future research.
NUMERICAL EXAMPLE
In this section, we present a numerical example of our model in order to
illustrate its equilibrium behavior. We are particularly interested in how
equilibrium behavior changes with ex ante changes in , the mix of firm
types, and with changes in the level of damages d.
Recall from our model that there are two key possible levels of self-
regulation. The high-cost firm always receives regulatory flexibility, and
chooses RHF defined by Eq. (5). If the low-cost firm decides to separate, its
optimal level of self-regulation is RLN defined by Eq. (9). These choices of
self-regulation, and the decision whether or not to mimic, will depend on
the how responsive the threat of regulation is to self-regulatory behavior,
as captured in the function (R). For this example, we shall assume

R
R 1 14
2d
Thus, (0)=1, (2d)=0, 0 (R) < 0, and (R) > 0. In particular,

R 1
0 R 1 15
2d 2d
Letting = 2, and substituting Eq. (14) into Eq. (5) we have
2 p
1 1 R2dHF 2d 6
RHF d 1 2d :36701d 16
d 1 2d
RHF 3
and substituting Eq. (14) into Eq. (9) we have
2 p
1 1 R2dLN 2d 3
RLN 2d 1 2d :8453d 17
d 1 2d
RLN 3
Now we must check whether, in our example, the low cost firm would rather
choose RLN, or mimic the high cost firm and choose RHF and thereby receive
flexible regulation. Recall from the Self-Regulation section above that
L RHF cL RHF RHF cL d RHF
which after a series of definitional substitutions and algebraic calculations is

p
12 2 6
L RHF cL d
9 9
Similarly, the section on Self-Regulation also shows that
L RLN cL RLN N RLN cL 2d RLN
which after a series of definitional substitutions and algebraic calculations is
p
4 3
L RLN cL d 2
9
Thus we can confirm that for this example the low cost firm will prefer to
mimic since
8 2 9 8 9
< p3= <
2 p3=
12 2 6 4 3
L RHF L RLN cL d 4 5 cL d 4 2 5
: 9 9 ; : 9 ;
3:9708cL d > 0
18
Thus, if the regulator will offer flexible regulation when the two firm types
pool, a low cost firm will indeed want to pool and choose RHF = .36701d.
Recall from the Regulatory Flexibility and Firm Compliance section
that granting flexibility is preferable for the regulator if < ~ where
d RHF
~
2d RHF
Using Eq. (16) we can compute that
d :36701d
~ :38763
2d :36701d
Thus, when < ~ :38763 the firms will pool and the regulator will offer flex-
ibility. Above this level, the firms will separate with the low cost firm choosing
RLN = .8453d and the high-cost firm continuing to choose RHF = .36701d.
DISCUSSION
Our analysis has highlighted an important trade-off facing a firm with low
compliance costs in an industry facing a regulatory threat: stay in step with
high-cost members of the industry so as to take advantage of regulatory
discretion if regulation passes, or step out and signal green in order to reduce
the risk of regulation, even though this means foregoing flexibility should
regulation pass. This decision depends crucially on ex ante regulatory beliefs.
If these beliefs would lead the regulator to offer flexible regulation, then by
lowering its level of self-regulation to match that undertaken by a high-cost
firm, a low-cost firm can ensure a flexible implementation of the regulation
should it pass. The increased likelihood of a more flexible, and therefore less
costly, regulation may be more attractive to a low-cost firm than the alter-
native of a less likely but stricter, and therefore more costly, regulation.
Previous work has examined the potential interplay between the non-
market strategy of self-regulation and regulatory decisions, but the settings
examined differ from ours in important respects. Much of the literature
ignores the possibility that firms know more about their cost structure than
do regulators (Manzini & Mariotti, 2003; Maxwell et al., 2000; Segerson &
Miceli, 1998). However, two prior papers have explored self-regulation in a
context of asymmetric information. Denicolo (2008) has shown that if firms
compete in the same market, low-cost firms may use self-regulation to

signal that the cost of regulation to the industry is low and thereby induce
regulation on high-cost rivals. If the total costs to industry are sufficiently
low, the regulator will not mind imposing the regulation on the high-cost
firms. However, Denicolo (2008) ignores the possibility of regulatory
preemption, and also assumes that the regulator adopts a one size fits all
policy that applies to both types of firms. Thus, the only role of self-regulation
in his paper is to induce regulators to impose new rules. In contrast, we
allow for self-regulation to preempt regulation, and we allow for regulatory
flexibility. Our analysis highlights the differences in strategic behavior when
legislators and regulators take into account that regulation has differential
cost impacts on firms, and therefore tailor regulations to avoid politically
unacceptable costs on certain classes of firms. In such situations, low-cost
firms may not benefit from imposing regulation on their high-cost rivals,
since their rivals may be held to weaker standards; instead low-cost firms
focus on preempting regulation altogether.
Heyes (2005), like Denicolo (2008), assumes that preemption is impossible;
in addition, he assumes self-regulation involves a reversible short-run cost
(e.g., an effort cost) rather than a long-run commitment. He shows that
high-cost firms may use self-regulation to obtain tailored, softer regulatory
standards. In his setting, high-cost firms are willing to undertake more self-
regulation than low-cost rivals if the costs of regulation, once imposed, will
rise with type in a sharply nonlinear manner. In contrast, we explore the case
where self-regulation involves an irreversible long-run commitment (e.g., a
capital investment) that can preempt legislation. In our setting, low-cost firms
can always afford to mimic the self-regulatory efforts of their high-cost rivals,
frustrating attempts by high-cost firms to signal their type.
Our results are consistent with the long-held notion that strong regula-
tory threats are needed to induce high levels of self-regulation. The threat
of inflexible regulation (as would be applied to firms known to be low-cost)
does induce greater industry self-regulation. However, our analysis points
to the need to expand the notion of what constitutes a strong regulatory
threat, because that threat may endogenously change depending on firm
actions. Traditionally, a strong regulatory threat has been modeled by the
notion of expected regulatory costs, with each component (the likelihood
of regulation and the cost of regulation to the firm) of the expected costs
viewed as exogenously fixed. Here we have shown that firm self-regulatory
actions may alter both the form and the likelihood of regulation. As a
result, firm decisions shape rather than simply respond to the regulatory
threat. Ironically, this ability to potentially shape the form of a regulation
can lead to lower levels of self-regulation than might arise otherwise.

Low-cost firms may cut back on self-regulatory efforts that would provide
industry-wide benefits (a reduced regulatory probability) in pursuit of
private advantage (more flexible regulation).
Empirical evidence suggests that many self-regulatory actions involve very
modest amounts of voluntary improvement. A number of empirical studies
have found that participants in voluntary programs undertake self-regulatory
actions that differ little from nonparticipants (King & Lenox, 2000; Rivera &
De Leon, 2004). Explanations for the weak behavior of participants range
from rules governing industry trade associations (Manzini & Mariotti, 2003)
to free riding (Delmas & Montes-Sancho, 2010). Our analysis suggests an
additional explanation. Firms may be reluctant to undertake too much self-
regulation if this raises the likelihood that they will be targeted for stringent
enforcement if regulatory requirements are imposed.
Our work highlights the role of asymmetric information in shaping both
self-regulatory actions and the potential sanctions firms might face. We
have focused on government regulations, but our insights also apply to
potential societal sanctions, or civil regulation. This means that asymmetric
information may play an important role in understanding firm choices
between substantive self-regulation and lower levels of self-regulation that
may be viewed as merely symbolic. Much of the prior literature examines
self-regulatory decisions when firms cost conditions are known (Manzini &
Mariotti, 2003; Maxwell et al., 2000; Segerson & Miceli, 1998). Not surpris-
ingly, self-regulatory decisions become more complex for industries, or
industry segments, where cost conditions are not known to regulators. If
the cost structure of a subset of firms in an industry is common knowledge,
those firms will have incentives to engage in self-regulation according to the
principles laid out in these earlier works. However, as we argued in the
Introduction, there are also likely to be other industry members, especially
smaller firms, whose costs are only known to the firms themselves. For
these firms, self-regulation involves a tradeoff between providing industry-
wide benefits (via strong preemptive self-regulation) and private gains
(from pooling with firms that have high compliance costs).
CONCLUSIONS
We have studied the strategic role of signaling green via self-regulation in
a setting with regulatory threats that can be preempted or implemented
with discretion if passed. The possibilities of regulatory preemption and
enforcement discretion mean that firms with low compliance costs confront
a tradeoff regarding self-regulation when most firms in the industry have
high compliance costs. On one hand, they can blend in with the rest of
the industry and take modest self-regulatory steps. This reduces the risk of
regulation somewhat, and preserves their ability to obtain regulatory
flexibility should regulation be imposed. On the other hand, they can step
forward and take substantial self-regulatory steps. This is more effective in
mitigating the risk of regulation, but at the risk of signaling low costs and
becoming a target for stringent enforcement should regulation pass.
Our analysis significantly extends the small theoretical literature on sig-
naling green (Denicolo, 2008; Heyes, 2005), which ignores the possibility of
regulatory preemption. When preemption is impossible, Heyes (2005)
shows that high-cost firms might paradoxically have incentives to signal
brown by self-regulating, in order to obtain lenient regulatory treatment;
however, this only works if self-regulation involves short-term effort rather
than long-term capital investment. Denicolo (2008) shows that low-cost
firms might signal green via self-regulation in order to convince regulators
to impose regulations on their higher-cost rivals, but he ignores the possibility
that self-regulation might instead lead to regulatory preemption. In contrast,
we show that low-cost firms may signal green in order to preempt
legislation, even though this may cost them a chance at benefiting from
regulatory discretion.
An important feature of our model is that the design of regulation has
important effects not just on the firms incentives for compliance, but
also on the likelihood of legislation itself. This would certainly come as no
surprise to anyone involved in politics. Moreover, the economic literature
on tradable permits has recognized this concern explicitly, both from an
empirical perspective (Joskow & Schmalensee, 1998) and from the perspec-
tive of policy design (Stavins, 1998). Likewise, the legal literature has
demonstrated the importance of grandfathering as a tool for making such
legislation at the Clean Air Act of 1972 politically palatable (Ackerman &
Hassler, 1981). An important contribution of our approach is to link regu-
latory design to the regulators incomplete information about firm types.
We have made a number of simplifying assumptions that could be
relaxed in future work. For example, we have assumed that activist groups
are not able to challenge regulatory flexibility in court. Heyes and Rickman
(1999) present an extension of their model of regulatory flexibility in which
an activist group is included, and show that regulatory flexibility may be
constrained in this setting. We have also assumed that regulatory penalties
are fixed, and are not collected by the regulator. If the regulator were able
to allocate penalty revenues toward environmental remediation, this would
make enforcement against high-cost firms more attractive, and reduce the
value of regulatory flexibility. Nevertheless, we expect that our results
would continue to hold qualitatively as long as mitigation is preferred
to remediation. In addition, we have assumed that compliance costs are
linear, and that within each firm the costs of compliance are the same on
each dimension; our results could be made more general by relaxing these
assumptions.
NOTES
1. The Dodd-Frank Act provided for tiered regulation in several areas including
an exemption for community banks from Consumer Financial Protection Bureau
examination and enforcement, and indemnification of community banks from
FDIC premium increases that will result from increasing the Deposit Insurance
Fund minimum reserve ratio from 1.15% to 1.35%. Other examples of tiered regu-
lation can be found in the final Basel III rule including allowing community banks
to continue using the Basel I mortgage risk weights, the exclusion of accumulated
other comprehensive income (AOCI) from the definition of regulatory capital, and
the grandfathering of tier one treatment of trust preferred securities (TruPS) for banks
with assets under $15 billion. See http://www.cbai.com/index.php/governmental-
relations-3/80-alerts-announcements/979-tiered-regulation-regulatory-relief
2. Congress can choose to prohibit regulatory agencies like the EPA from
exercising enforcement flexibility. In overturning part of the EPAs Final Rule on
greenhouse gases, the D.C. Circuit Court of Appeals rebuked EPA for proposing to
exempt certain types of facilities from enforcement, warning that [T]here exists
no general administrative power to create exemptions to statutory requirements
based upon the agencys perceptions of costs and benefits. See Center for
Biological Diversity v. EPA and Lisa Perez Jackson, 2013.
3. Their analysis found that a regional bubble over a companys entire operations
yielded significantly more clean air than source standards and still generated cost
savings of considerable size (Maloney & Yandle, 1980, p. 51).
4. The closest work to ours is the small literature in economic theory that
examines self-regulation as a signal. Heyes (2005) shows that a high-cost firm may
self-regulate to signal that its abatement costs are high and thereby weaken the
stringency of future regulation. However, he assumes the probability of legislation
is unaffected by self-regulation, whereas we explore the more realistic case where
self-regulation can reduce the likelihood of legislation. Denicolo (2008) shows that
a low-cost firm may self-regulate to signal that abatement costs are low, thereby
inducing the regulator to impose regulation raising the costs of its higher-cost
rival. However, he also ignores the possibility of preemption, and in addition
he assumes away the possibility of regulatory discretion. Thus, neither paper
addresses the relationship between regulatory discretion and preemptive self-regu-

lation that is at the heart of our analysis.
5. In practice, regulatory flexibility can mean either applying different abatement
requirements to different firms, or applying the same requirement to all firms but
granting firms flexibility in how they choose how to meet it. For example, the
former version of flexibility can be implemented through the grandfathering of
incumbent firms who face less stringent standards than do new entrants; the latter
can be implemented through the use of tradable permits instead of best available
technology requirements. One might argue that the regulator already knows
the cost structure of incumbent firms, but in reality incumbent firms are not all the
same and there will be some uncertainty about which firms really have costs so
high they need to be exempted from compliance. Regardless of how flexibility is
implemented, however, it reduces the cost of compliance and thus reduces industrys
incentive to resist regulation.
6. In Heyes and Rickman (1999), costs of abatement are drawn from a continuous
distribution. This results in nine possible types of firms, depending upon whether for
each dimension j we have cij < F, cij (F, 2F), or cij > 2F. However, all the actions in
the paper involve tradeoffs between the likelihood of two types: firms that would com-
ply on neither domain without a regulatory deal, and firms that would comply on both
domains without a regulatory deal. Thus, assuming firms have identical costs on both
domains captures the key insight of the original paper.
7. For example, the Regulatory Flexibility Act requires both an initial and a final
regulatory flexibility analysis for any new rule unless the promulgating agency certifies
that it will not have a significant economic impact on a substantial number of small
entities (605b). What constitutes a significant impact and a substantial number
of entities, however, remains vague. This leaves considerable uncertainty regarding
how many firms actually suffer high costs of compliance, and exactly how high those
costs may be.
8. Note that the high-cost firm does not have to do as much additional abatement
as he would if R = 0. Hence, if the fine was enough to motivate compliance on one
domain when R = 0, it is surely enough to motivate compliance when R > 0.
9. One might be concerned that if R2 is so large that (dR2) < F, then the regulator
will revoke the promise of flexibility since the firm will comply on one dimension even
without flexibility. However, since the regulator does not keep the penalty revenue, she
is indifferent between two outcomes that both offer compliance on one dimension, and
she is assumed to continue to offer flexibility regardless of the high-cost firms choice
of R2.
ACKNOWLEDGMENTS
We thank the editor and an anonymous referee for helpful suggestions, as
well as audience members at the Istanbul World Congress of Environmental
and Resource Economists.
REFERENCES
Ackerman, B. A., & Hassler, W. T. (1981). Clean coal dirty air: Or how the clean air act became
a multibillion-dollar bail-out for high-sulfur coal producers and what should be done about
it. New Haven, CT: Yale University Press.
Bagnoli, M., & Bergstrom, T. (2005). Log-concave probability and its applications. Economic
Theory, 26, 445469.
Besley, T., & Coate, S. (2003). Elected vs. appointed regulators: Theory and evidence. Journal
of the European Economic Association, 1, 11761206.
Brock, W. A., & Evans, D. S. (1985). The economics of regulatory tiering. The Rand Journal
of Economics, 16, 398409.
Center for Biological Diversity v. EPA and Lisa Perez Jackson. (2013, July 12). No. 111101,
U.S. Court of Appeals for the District of Columbia Circuit.
Darnall, N., & Carmin, J. (2005). Greener and cleaner? The signaling accuracy of US
voluntary environmental programs. Policy Sciences, 38, 7190.
Decker, C. S. (2003). Corporate environmentalism and environmental statutory permitting.
Journal of Law and Economics, 46, 103129.
Delmas, M., & Montes-Sancho, M. (2010). Voluntary agreements to improve environmental
quality: Symbolic and substantive cooperation. Strategic Management Journal,
31, 575601.
Delmas, M., & Terlaak, A. (2001). A framework for analyzing environmental voluntary
agreements. California Management Review, 43, 4463.
Denicolo, V. (2008). A signaling model of environmental overcompliance. Journal of Economic
Behavior & Organization, 68, 293303.
Duhigg, C., & Roberts, J. (2010). Rulings restrict clean water act, hampering E.P.A. The
New York Times, March 1, p. A1.
Esty, D. C., & Winston, A. S. (2006). Green to gold: How smart companies use environmental
strategy to innovate, create value, and build competitive advantage. New Haven, CT:
Yale University Press.
Fisher-Vanden, K., & Thorburn, K. (2011). Voluntary corporate environmental initiatives and
shareholder wealth. Journal of Environmental Economics and Management, 62, 430445.
Freeman, III, A. M. (2002). Environmental policy since earth day I: What have we gained?
The Journal of Economic Perspectives, 16, 125146.
Gray, W. B., & Deily, M. E. (1996). Compliance and enforcement: Air pollution regulation in the
US steel industry. Journal of Environmental Economics and Management, 31, 96111.
Gray, W. B., & Scholz, J. T. (1993). Does regulatory enforcement work? A panel analysis of
OSHA enforcement. Law and Society Review, 27, 177213.
Gray, W. B., & Shimshack, J. P. (2011). The effectiveness of environmental monitoring and
enforcement: A review of the empirical evidence. Review of Environmental Economics
and Policy, 5(1), 324.
Harrington, W. (1988). Enforcement leverage when penalties are restricted. Journal of Public
Economics, 37, 2953.
Heyes, A. G. (2005). A signaling motive for self-regulation in the shadow of coercion. Journal
of Economics and Business, 57, 238246.
Heyes, A., & Rickman, N. (1999). Regulatory flexibility Revisiting the Harrington paradox.
Journal of Public Economics, 72, 361378.
Innes, R., & Sam, A. G. (2008). Voluntary pollution reductions and the enforcement of envir-
onmental law: An empirical study of the 33/50 program. Journal of Law and Economics,
51, 271296.
Jacobs, B. W., Singhal, V. R., & Subramanian, R. (2010). An empirical investigation of
environmental performance and the market value of the firm. Journal of Operations
Joskow, P., & Schmalensee, R. (1998). The political economy of market-based environmental
policy: The US acid rain program. Journal of Law and Economics, 41, 81135.
King, A. A., & Lenox, M. J. (2000). Industry self-regulation without sanctions: the chemical
industrys responsible care program. Academy of Management Journal, 43, 698716.
King, A. A., & Lenox, M. J. (2001). Does it really pay to be green? An empirical study of firm
environmental and financial performance. Journal of Industrial Ecology, 5, 105116.
Klassen, R. D., & McLaughlin, C. P. (1996). The impact of environmental management on
firm performance. Management Science, 42, 11991214.
Laffont, J. J., & Tirole, J. (1988). The dynamics of incentive contracts. Econometrica,
56, 11531175.
Lyon, T. P., Lu, Y., Shi, X., & Yin, Q. (2013). How do investors respond to green company
awards in china. Ecological Economics, 94, 18.
Lyon, T. P., & Maxwell, J. W. (2004). Corporate environmentalism and public policy.
Cambridge: Cambridge University Press.
Lyon, T. P., & Maxwell, J. W. (2007). Environmental public voluntary programs reconsidered.
Policy Studies Journal, 35, 723750.
Lyon, T. P., & Maxwell, J. W. (2008). Corporate social responsibility and the environment:
A theoretical perspective. Review of Environmental Economics and Policy, 2, 240260.
Lyon, T. P., & Mayo, J. W. (2005). Regulatory opportunism and investment behavior:
Evidence from the electric utility industry. RAND Journal of Economics, 36.
Maloney, M. T., & Yandle, B. (1980). Cleaner air at lower cost: Bubbles and efficiency.
Regulation, 4, 4952.
Manzini, P., & Mariotti, M. (2003). A bargaining model of voluntary environmental agreements.
Journal of Public Economics, 87, 27252736.
Margolis, J., & Walsh, J. (2001). People and profits? The search for a link between a companys
social and financial performance. Mahwah, NJ: Laurence Erlbaum Associates.
Maxwell, J. W., & Decker, C. S. (2006). Voluntary environmental investment and responsive
regulation. Environmental and Resource Economics, 33, 425439.
Maxwell, J. W., Lyon, T. P., & Hackett, S. C. (2000). Self-regulation and social welfare:
The political economy of corporate environmentalism. Journal of Law and Economics,
XLIII(2), 583618.
Morantz, A. D. (2009). Has devolution injured American workers? State and federal enforcement
of construction safety. Journal of Law, Economics and Organization, 25, 183210.
Olson, M. (1995). Regulatory agency discretion among competing industries: Inside the FDA.
Journal of Law, Economics and Organization, 11, 379405.
Porter, M. E., & Van der Linde, C. (1995). Toward a new conception of the environment-
competitiveness relationship. The Journal of Economic Perspectives, 9, 97118.
Potoski, M., & Prakash, A. (2004). Regulatory convergence in nongovernmental regimes?
Cross-national adoption of ISO 14001 certifications. Journal of Politics, 66, 885905.
Prakash, A., & Potoski, M. (2005). Green clubs and voluntary governance: ISO 14001 and
firms regulatory compliance. American Journal of Political Science, 49, 235248.
Rivera, J., & De Leon, P. (2004). Is greener whiter? Voluntary environmental performance of
western ski areas. Policy Studies Journal, 32(3), 417437.
Segerson, K., & Miceli, T. (1998). Voluntary environmental agreements: Good or bad news
for environmental protection? Journal of Environmental Economics and Management,
36, 109130.
Shapiro, S. A., & Rabinowitz, R. (2000). Voluntary regulatory compliance in theory and
practice: The case of OSHA. Administrative Law Review, 52, 97155.
Shimshack, J., & Ward, M. B. (2005). Regulator reputation, enforcement, and environmental
compliance. Journal of Environmental Economics and Management, 50, 519540.
Short, J. L., & Toffel, M. W. (2010). Making self-regulation more than merely symbolic: The
critical role of the legal environment. Administrative Science Quarterly, 55, 361396.
Stavins, R. N. (1998). What can we learn from the grand policy experiment? Lessons from
SO2 allowance trading. The Journal of Economic Perspectives, 12, 6988.
Verkuil, P. R. (1982). A critical guide to the regulatory flexibility act. Duke Law Journal,
1982, 213276.
Weitzman, M. (1980). The ratchet principle and performance incentives. Bell Journal of
PRIVATE POLITICS DAILY: WHAT
MAKES FIRMS THE TARGET OF
INTERNET/MEDIA CRITICISM?
AN EMPIRICAL INVESTIGATION
OF FIRM, INDUSTRY, AND
INSTITUTIONAL FACTORS
Dominik Breitinger and Jean-Philippe Bonardi
ABSTRACT
Private politics refers to situations in which activists or NGOs try to push

firms to conform to social standards (regarding, for instance, human
rights and environmental protection) without public policy intervention.
The existing literature on private politics has focused on large campaigns
such as consumer boycotts, and looked at the impact of those boycotts on
firms financial performance and on the likelihood that firms comply with
activist demands. Even though these large campaigns are important,
focusing on them leads to neglecting the fact that a large portion of the
time and resources that activists consecrate to private politics is used
to monitor firms and criticize them through Internet posting and media
statements, rather than to launch high profile campaigns. Little is known,

ISSN: 0742-3322/doi:10.1108/S0742-332220160000034010
331
332 DOMINIK BREITINGER AND JEAN-PHILIPPE BONARDI
however, about what drives these activists when they criticize companies,
why they target certain companies and not others, and whether this
criticism should be considered as a primary step in the production of
full-fledged campaigns. In this paper, we fill this gap by exploring a
unique international database of CSR-based criticisms against Fortune
500 companies for the 20062009 period. This database allows us to
look at the impact of a broad range of factors including industry differ-
ences, country/institutional differences and firm-specific dimensions, on
the likelihood that a certain firm will be targeted by activist critique.
Results indicate that criticism is driven by strategic intents. Similar to
previous literature, large and visible firms in certain industries are more
targeted than others. In addition, these firms also tend to come from
countries with strong institutions and high standards of living.
Keywords: Private politics; activists; Internet/media criticism
INTRODUCTION
Within the field studying firms nonmarket strategies (Baron, 2003), private
politics in which activists launch campaigns and target firms in order to get
them to change their practices has become more and more important (Beck,
2000; King & Pearce, 2010). Whereas public politics in which activists and
other interest groups compete against firms in lobbying for public policy
changes (Kingsley, Vanden Bergh, & Bonardi, 2012) is often perceived as
uncertain and difficult to manage, private politics has indeed proven to be an
effective way for activists to reach their goals and push firms to self-regulate
(Baron & Diermeier, 2007; Bunn, 20032004). These activists have evolved
into well-organized civil society groups that diligently scrutinize business
practices (Yaziji & Doh, 2009). To make uncovered social/environmental
issues salient, NGOs dispose of an arsenal of weapons to target or hit firms,
ranging from letter writing campaigns to media criticism (Bonardi & Keim,
2005), boycotts (Friedman, 1985) and public, sometimes violent, campaigns
(Teegen, Doh, & Vachani, 2004). NGOs, therefore, make use of radio, televi-
sion and newspaper ads to condemn practices of particular firms, organize
boycotts, sit-ins, customer confrontations; and employ face-to-face challenges
in the form of blockades, protests, banner-hangs, and so on (Sasser, Prakash,
Cashore, & Graeme, 2006, p. 06).
What Makes Firms Targets of Internet/Media Criticism? 333
The academic literature has recognized this growing trend, and the number
of studies regarding private politics has risen significantly in recent years.
Theoretical pieces argue that activists, when they implement anticorporate
campaigns, pick their targets strategically based on the cost of consumers
switching to other products, as well as on how costly the campaign will be for
the target (Baron & Diermeier, 2007). In particular, companies with a strong
brand name or a high level of reputation are more likely to be targeted by
activists campaigns (King, 2008). Empirical studies on the phenomenon
confirm that target selection is heavily influenced by target characteristics:
companies that are large, visible, and financially successful seem to be pre-
ferred targets (King & Soule, 2007). Some studies also argue that an affiliation
to certain industries, particularly polluting industries, also matters (Lenox &
Eesley, 2009). Rehbein, Waddock, and Graves (2004) analyze 1,944 labor-
and environmental-related shareholder resolutions against 600 US firms in the
1990s. They provide evidence that companies with inferior social performance
significantly receive more shareholder resolutions. Likewise, companies that
have produced products that have negative contingencies are preferred
stakeholder targets Rehbein et al. (2004, p. 261). The study by Lenox and
Eesley (2009) examines the selection (targeting) and response strategies of 552
environmental campaigns organized against 273 US firms between 1988 and
2003. They find, similar to Rehbein et al., that smaller, less visible firms are
less likely to be targeted.
One limitation of this literature, however, is that it tends to concentrate
on large and very visible events such as boycotts (John & Klein, 2003;
King, 2008; Koku, Akhigbe, & Springer, 1997) or large campaigns (Eesley &
Lenox, 2006). In doing so, this literature neglects the fact that most of what
activists do is often more mundane and related to monitoring and
criticizing firms over the Internet or in the media, but without necessarily
launching these large campaigns. A significant part of activist behavior in
private politics has therefore not been studied empirically. Several questions
emerge as a result: Do activists pick corporate targets in the same way for
mundane criticism as they do for large campaigns? What drives activist
behavior for daily private politics, and what determines the intensity of the
pressure these activists put on firms?
Beyond studying why certain firms are targeted, this paper thus aims at a
better understanding of what drives activist behavior. Activists are presented
in the literature as playing two roles regarding firms and environmental or
social issues. First, activists can play a monitoring role: scan firms activities
around the world and act as whistleblowers when bad behaviors are spotted.
This role is seen as particularly important when multinational firms, using
globalization to their advantage, operate in countries in which institutional

structures are too weak to prevent bad behaviors (Scherer & Palazzo, 2011).
In this case, we should observe that activist targeting activities focus on
firms, mostly from developed countries, creating environmental and social
issues abroad. Second, activists can also use firms to gather attention to their
cause and build reputation or legitimacy (Den Hond & de Bakker, 2007).
This is what we call here the strategic motive for activism. In this case,
they will target well-known and visible firms, and even potentially CSR-
oriented firms, as a way to generate attention. One of the purposes of the
empirical section of this paper will be to provide insights regarding these
two mechanisms.
Note that this question leads to a deeper one that has to do with the
social production of corporate targets and shaming campaigns (Bartley &
Child, 2014). Even if criticizing is a different activity than running a
full-fledged campaign, both might be connected and criticism might in fact
be an early stage of private politics that might later lead to full campaigns.
So, are criticism and campaigns connected or are they two separate aspects
of private politics? The first potential view on this question is that criticism
is not strategic, and that the strategic decision-making (targeting) for the
activist comes later. Activists would monitor a large pool of environmental
and social issues, criticize and then strategically pick those that will become
targets in full anticorporate campaigns. In that case, criticism will be
mostly about random monitoring of companies, and will not display
the types of strategic behaviors highlighted by previous research on private
politics. The other view is that monitoring and criticizing firms are already
strategic behaviors. As such, there is no such thing as random criticism and
criticism is generally the first stage of an anticorporate campaigns.
Understanding this distinction is important for two reasons. First, managers
must understand how they need to react to criticism (Lamin & Zaheer, 2012).
If criticism is strategic, then managers need to pay attention to it, even at an
early stage and even if the criticism is not very severe. Second, understanding
whether criticism is strategic is also important to account for the welfare
impact of activism. Activism, in effect, is often presented as an effective
counter-power to firms in a globalizing world, in which governments no
longer have the ability and authority to supervise questionable corporate
practices (Scherer & Palazzo, 2011). However, if even criticism is strate-
gic, activism will probably be an imperfect way of supporting deficient
institutions in identifying environmental and social problems, in that this
identification will be skewed toward certain firms in certain places and
will leave many others without much surveillance.
The second limitation of the current literature on private politics is that

it does not explore how country or institutional factors might have an
impact on the occurrence of public criticism or on its intensity. There are at
least two good reasons why these country factors might matter for criticism
and private politics in general. On the one hand, if public criticism is strategic,
then criticism will likely target firms in countries with higher standards of
living and stronger institutions. On the contrary, if criticism were not strategic,
but mostly about sheer monitoring to identify issues with corporate practices,
then we should expect it to target companies from and in countries with lower
standards of living and weaker institutional systems.
This discussion speaks to a broader debate about behavioral changes
regarding environmental or social issues. In particular, we should expect to
observe importance high impact of institutional factors on Internet/media
criticism if hypotheses such as the Environmental Kuznets curve bear some
truth. This theory, which has received some empirical support (Stern,
Common, & Barbier, 1996), suggests that people in countries with a certain
level of income per capita backed up by the right institutional environment
will give higher value to environmental quality and will take actions
accordingly. Higher standards of living and stronger institutions should
therefore have two effects: (1) push people to take action, become activists
themselves and monitor firms behaviors, and (2) create a fertile ground for
public criticism and campaigns against polluting or questionable corporate
behaviors. As a result, firms should receive more criticism the higher the
standards of living and the stronger the countries institutions. This would
bring support to the idea that human societies might self-regulate over time
and find ways to engage in development modes that are more sustainable.
In this paper, we explore these questions by building on a global dataset
of CSR-related criticism that includes Internet/media criticism for 451
multinational companies between 2006 and 2009. These companies are
incorporated in 26 countries and have been criticized for their operations in
114 countries. We also exploit these institutional variations and explore
whether country-level factors contribute to explain why certain firms
become private politics targets, and explanations for the intensity of the
criticism these companies face.
Our results suggest that criticism is indeed strategic. Activists use
mundane Internet/media criticism in a similar way to what has been
uncovered by previous literature for large campaigns and boycotts: large
and visible companies operating in close-to-consumer (food/beverage;
personal goods/textile) and controversial industries (e.g., tobacco; mining)
are more likely to be targeted. Moreover, new evidence is provided for
the importance of country factors: companies that are incorporated in

countries of high educational level, environmental attitude, standard of
living, and low unemployment rates are preferred targets of NGO/media
activism, which is overall in line with the Environmental Kuznets Curve
approach and the idea that activists operate mostly in countries in which
there is a favorable audience for their claims.
The remainder of this paper is divided into four parts. In the next section,
we explore the differences but also the potential connection between
criticism and activist campaigns, and build on these to develop testable
hypotheses in the following section. The section Results develops the
empirical tests. The last section discusses these results and concludes.
INTERNET/MEDIA CRITICISM AND

ACTIVIST CAMPAIGNS
Internet/Media Criticism and Campaigns as Different Aspects of
Private Politics
In this section, we highlight the differences between campaigns, which

has been the main focus of the existing literature, and the more mundane
criticism that constitutes the day-to-day activity of many activist organi-
zations. The first difference has to do with the content itself and where it
comes from. The theoretical literature on private politics models anticor-
porate campaigns as a demand by the activist costly for the firm to
implement associated with a threat of harm and a reward condition if
the activists demand is met (Baron & Diermeier, 2007). For example, in
the Rainforest Action Network case against Citigroup, which started
in 2000, the explicit demand by the activist was to get the bank to stop
financing activities that could endanger ecosystems. This demand was
originally made through a letter sent by the activist organization to
Citigroup and stating that: Citigroup will have to extract itself from
unsustainable investments in fossil fuels to move its capital support to
renewable energy; they will have to cease funding destruction of primary
forest for timber or mining or oil exploration and transition that support to
alternative building supplies and paper materials (Yurday & Baron,
2004). Internet/media criticism, on the other hand, is generally anonymous,
that is, it does not involve a well-identified activist group and articulates
neither a concrete demand nor a precise threat.
In the same way, Internet/media criticism does not call for a precise
response from the firm, and its reputational implications are likely to be
quite different from campaigns both for the activist and for the firm. For
the firm, campaigns can have a highly negative impact on its reputation,
which even materializes in drops in share price (King & Soule, 2007). The
opposite is true for activists, who can derive important reputational bene-
fits from successful campaigns, especially in cases in which these activists
do not wish to get one firm to concede but rather a whole industry (Baron &
Diermeier, 2007). Internet/media criticism, on the other hand, typically
generates lower reputational impact and would probably have to be fol-
lowed by subsequent steps for concrete reputation loss to take place. Part
of the issue here also comes from the fact that criticism has a much lower
cost than full-fledged campaigns; a characteristic that could associate them
more with cheap talk, therefore generating a lower direct impact on
firms. Table 1 summarizes the key differences between public criticism and
activist campaigns.
The conclusion of this analysis is that Internet/media criticism and
corporate campaigns have different characteristics, and therefore should not
necessarily be expected to display the same characteristics as full activist
campaigns do. If corporate targets have been shown to be strategically
chosen in previous literature (King & McDonnell, 2015), criticism could be
of different nature. In particular, criticism could be considered to be mostly
about random monitoring, with activists thus playing a key role in scanning,
identifying, and revealing objectionable behaviors around the world.
Another view, however, could be that criticism and campaigns might be
different but related parts of private politics. From that angle, criticism would
already be strategic and would support the strategic targeting identified in the
previous literature for full-fledged campaigns. We explore this idea in the
next section.
Table 1. Public Criticism versus Activist Campaigns.

Criticism Campaigns
Generally anonymous Led by a well-identified activist group

No well-articulated demand Includes a specific demand
Lower cost to organize Higher cost to organize
Lower reputation impact both Higher reputation impact both
for the firm and the activist for the firm and the activist
Does not necessarily call for Calls for a response by the firm
a response by the firm
Internet/Media Criticism as Early Stage of Activist Campaigns
There are, in fact, two key theoretical reasons why Internet/media criticism
might already be strategic activities and therefore might be directly con-
nected to strategic targeting as in full campaigns. The first reason is that
activist organizations have limited attention capabilities (Ocasio, 1997) and
therefore need to selectively allocate these attention capabilities. This selec-
tion allocation could depend on many different factors associated with the
goals of the specific activist organization, but it is reasonable to assume
that this allocation follows a targeting approach that is akin to what
activists do when they target firms in full campaigns. The second reason to
believe that Internet/media criticism is strategic, more fundamental in nature,
is that environmental and social issues do not objectively exist but are rather
social constructions instrumented by activist organizations (Bartley & Child,
2014). From this perspective, the key question becomes about the role played
by Internet/media criticism in this social construction.
Conceptually, private politics can be considered to work as a chain
composed of different stages, with Internet/media criticism being one of the
early stages, in which the most attractive issues to pursue are selected, and
full-fledged campaigns being the last and triggering open confrontations
with firms.1 The sequence of the production chain of anticorporate
campaigns would thus start with the selection of a type of issue the activist
wants to focus on, followed by some media/Internet criticism against
targeted firms and potentially ending up with activists fighting a campaign
against this firm. This private politics chain, leading to the production of
full-fledged campaigns, is described in Fig. 1.
Assuming that activists behave strategically all along, the first step of
campaigns might be the selection of the countries in which to operate.
These countries will be the ones in which activists demand are the most
likely to be heard and supported by the local population. Similarly,
activists will probably select industries that will be the most likely to attract
attention. Based on these choices, activists will then pick the best target
firms to create attention to their cause, and then will monitor them closely
and will criticize them when the opportunity will appear. Among the firms
criticized, some will then be targeted with a specific demand and a threat of
harm, leading to a full campaign.
Note that considering criticism as an early stage of an activist campaign
also fits with the theoretical literature in a different way. In effect, the theo-
retical literature suggests that both firms and activists have in fact strong
incentives to agree on conditions that would avoid a costly campaign for
Reputational risk for the

firm
Selection by
activists of Monitoring,
countries to surveillance, Criticism of Official
operate in and of scanning of firms on the targeting and
industries to demand with Full
potential issues Internet or in campaign
target the media a threat of
harm
Strategic importance for the

activist organization
Fig. 1. The Production Chain of Anticorporate Campaigns.
both (Baron & Diermeier, 2007). In other words, in equilibrium full cam-
paigns should not occur and, in many cases, campaigns should stop early
in the production chain described in Fig. 1. On the other hand, criticism
should take place and often constitute the first step through which activists
create bargaining power in order to negotiate the settlement that will
prevent a full campaign. In fact, existing evidence suggests that whereas
relatively few full anticorporate campaigns take place, many others could
have the potential to be launched. For instance, Wright and Boudet (2012)
study different contexts in which campaigns could be expected to take
place. They show that only half of them actually do, in spite of problematic
environmental issues, organized interests monitoring them and early criticism
being voiced.
The implication of this analysis is that Internet/media criticism and full
campaigns could be directly connected, in which case they could be driven by
a set of similar factors. Another way of considering this is that public criticism
is a strategic activity, that is, it would not be about random monitoring but
would imply a primary selection of the right targets for activists. If this is the
case, then we should expect public criticism to be driven by a similar set of
factors to what has been identified in previous literature regarding full
campaigns. In what follows, we explore this theory and develop a set of
testable hypotheses that will guide our empirical analysis.
HYPOTHESES
Based on our concept of private politics production chain, criticism should
be driven by country-specific factors, industry factors and firm-specific
ones. The last two are actually in line with the existing literature on private
politics and firm-activists relationships. Exploring country-specific or insti-
tutional factors, on the other hand, is one of the contributions of this
paper. We consider them in turn and derive hypotheses that will guide our
empirical analysis.
Firm-Specific Factors
In the context of the production of anticorporate campaigns, activists

might first focus their criticism on wealthy firms. Lenox and Eesley (2009)
argue that activists selectively target resource-poor firms (low cash flows),
as the cost to an activist to deliver a certain level of harm to a firm, the
greater the firms financial capital (2009, p. 50). Firms with higher cash
flows at their disposal might exert greater corporate opposition as they
are able to support dedicated legal and public relation staff (ibid.).
Nonetheless, opposite views claim that financially successful firms that
is, wealthy in terms of assets and sales trigger media attention because
they have the resources to meet activists demands (Yaziji, 2004). Therefore,
a wealthy firm can more easily implement the requested changes, under the
assumption that it does not employ its resources to buffer, the activism
raised against it. Whether resource-rich or -poor corporations attract
Internet/media attention will be left to the empirical analysis.
H1. The wealthier the firm, the higher the likelihood of becoming a
target of Internet/media criticism.
Further, it has been argued in the targeting literature that firm size alone
may be a reason why activists target companies (Graves, Rehbein, &
Waddock, 2001; Rehbein et al., 2004, p. 250). This is based on the general
observation that firm size is usually highly correlated with firm visibility,
which in turn creates higher media attention for the stakeholders campaign
(Meznar & Nigh, 1995). Additionally, targeting scholars claim that
[L]arge, visible firms are attractive targets as campaigns against them are
more likely to garner attention from the media and the general public
(Lenox & Eesley, 2009, p. 50). If criticism is strategic, we should therefore
expect visible firms to be targeted.
In a similar vein, Baron & Diermeier posited that companies with pri-
mary and prominent brand(s) (2007, p. 612) are more likely to be selected
as activists expect to derive publicity for themselves. Therefore, well-known

firms might be preferred targets as contentions with these firms are more
visible and make the activists themselves better known. The better
known a company is, the juicier the target it makes (Yaziji, 2004, p. 111;
see also Diermeier & van Mieghem, 2005; Hendry, 2006). Moreover, contri-
butions to social movement theory associate such potential threats of
negative publicity (Den Hond & de Bakker, 2007, p. 911) with highly
exposed, well-known companies and/or brands. Following this reasoning,
we derive the following hypothesis:
H2. The more visible the firm, the higher the likelihood of becoming a
target of media/Internet criticism.
Finally, the CSR (i.e., social or environmental) performance of a company
has also been discussed as a potential factor that attracts or buffers activist
attention. In particular, socially responsible firms might receive more activist
attention particularly because of their emphasis on social responsibility
(Argenti, 2004, p. 111). Since firms are thought to be sympathetic (Baron &
Diermeier, 2007, p. 612), that is more receptive to social and environmental
concerns, such truly socially responsible companies (Argenti, 2004, p. 111)
are more likely to become targeted (King & McDonnell, 2015). In this
vein, it has been further argued that good corporate students are
evaluated more strictly in such a way that they keep up their good perfor-
mance. Luo, Meier, and Oberholzer-Gee (2010) suggest that the media are
more likely to report on corporate environmental sins if target firms indicate
better environmental records, and provide higher degrees of CSR transpar-
ency (disclosure of environmental performance data). If criticism is strategic,
firms that have presented themselves as CSR-oriented firms could therefore be
attractive targets for activists.
H3. Companies that are perceived as more socially/environmentally
responsible are less likely to be targeted by Internet/media criticism.
Industry Factors
In the context of the production chain of anticorporate campaigns presented

in Fig. 1, activists should be expected to pick firms in industries that might be
well known from the public and that can be easily associated to questionable
behaviors. Yaziji (2004), for instance, posits that companies are particularly
exposed to external criticism if they sell unhealthy products (tobacco, alco-
holic beverages), offer dangerous services (gambling, pornography), or even
produce dangerous goods (weapons, nuclear energy). Studies on corporate
reputation emphasize the inherent dangers to which a high-reputation firm is

exposed if it operates in a per se relatively controversially perceived industry
(see Barnett & Hoffmann, 2008).
As indicated previously, Rehbein et al. (2004) provide evidence for such
industry-affiliation claims. They find that operating in industries that are
likely to produce negative contingencies (p. 249) provokes NGO/media
attention. King and Soule (2007), as well as King (2008), provide further
empirical evidence that supports this believed causality between industrial
controversy and stakeholder activism. Similar to full campaigns, activists
might thus pick targets in industries that will easily generate interest and
support from the public.
H4. Companies that operate in controversially perceived industries are
more likely to be targeted by Internet/media criticism.
Similarly, high consumer exposure (Weber & Marley, 2012), that is operating
close to final customers, may increase the targeting likelihood (Bartley &
Child, 2014). Baron and Diermeier (2007) reason that it may be relatively
low cost for an activist (p. 614) to damage the reputation of a consumer
product company () whereas harming an industrial products company
may be quite costly (ibid.). This argument is not only explained by higher
customer visibility. In fact, it has also been associated with the existence of
substitute products. For example, the boycott against Shells Brent Spar
oilrig in 1995 was highly successful for Greenpeace because final consu-
mers, who followed the NGOs boycott call, incurred comparatively low
switching costs by easily getting fuel elsewhere (see Diermeier & van
Mieghem, 2005). Resource-dependence theory explains such stakeholder
movements according to which consumers can successfully withhold
resources from a firm if they possess power (e.g., knowledge) over this
company (Pfeffer & Salancik, 1978). This is the case for corporations whose
products are close to the final consumer. Thus, if criticism is strategic, we
hypothesize that:
H5. Companies that operate in close-to-consumer industries are more
likely to be targeted by Internet/media criticism.
Country-Specific and Institutional Factors
Country-specific and institutional factors have not been considered yet in

the private politics literature. Yet, they should have an impact and could
help identify whether criticism are driven by strategic intent or by sheer

monitoring. Within the production chain of anticorporate campaigns, acti-
vists are likely to issue criticism against firms from countries in which the
population will be sympathetic to the cause that these activists want to
defend. Previous literature suggests, in fact, that the level of intensity and
severity of activists or NGOs criticism vary across countries (Wapner,
2009). A firm like Nestle from Switzerland, for instance, has been much
more criticized for using palm oil than one of its large rival, Unilever, from
the United Kingdom. Doh and Guay claim that activists search for particu-
lar access points (2006, p. 52) per country due to national particularities.
Additionally, Diermeier and van Mieghem (2005), with regard to the Brent
Spar case, illustrated how Greenpeace chose not to target Shell Plc.
in England, but instead concentrated its activism on Shells German
subsidiary because Germany was deemed more receptive and sensible to
environmental concerns.
Socioeconomic conditions have also been shown to matter to influence
citizens attitudes toward firms, toward corporate social responsibility
activities, and toward the potential need for self-regulation (Matten &
Moon, 2008). In that spirit, socioeconomic conditions that structurally
shape behaviors, for instance, through a countrys level of education or
its employment rate (Beck & Levine, 2005; La Porta, Lopez-de-Silanes,
Shleifer, & Vishny, 1998, 2000) should matter in the context of activists
targeting decisions. Firms incorporated in a country with a high level of
education could be more strictly scrutinized than companies incorporated
in countries where citizens are less knowledgeable and aware of social and
environmental issues.
The literature on sustainable development and growth also supports the
idea that individuals in a society should display a higher willingness to pay
for better environmental conditions and cleaner air the higher the standard
of living in the country. In particular, the Environmental Kuznets curve
theory argues that environmental degradation first increase with income
and then reaches a threshold after which environmental degradation
decreases with income (Stern et al., 1996). The factors driving this inverted-U
relationship between per capita income and pollution are generally con-
sidered to be related to the evolution from industry-oriented economies to
more information-intensive and service-oriented ones, as well as to a
higher environmental awareness, stronger environmental regulations and
more advanced green technologies. However, in addition to these factors,
a stronger scrutiny on business practices and more pressure on polluting
firms should also take place, once the income threshold has been reached,
in order for the downward part of the curve to emerge. From this perspec-
tive, companies incorporated in countries with higher standard of living
should also be criticized more often and more intensely than companies
incorporated in countries that have not reached that threshold.
Similarly, other measures of a countrys wealth such as the unemploy-
ment level of unemployment or a countrys level of education should be
associated with a higher level of activism and therefore more criticism
against firms.
H6. The better the socioeconomic situation of a country, the higher the
likelihood for firms incorporated in that country to be targeted by
Internet/media criticism.
Second, institutional characteristics might matter as well. More democratic
and less corrupt institutions create plenty of opportunities for outside parties
to voice their concerns not only about what government officials are doing
but also about what firms and managers do (Dyck & Zingales, 2002).
Institutions include constitutions, laws, policies, [hence] any formal agree-
ments (Doh & Guay, 2006, p. 52) that are (formally) set up in the national
context to provide the background conditions against which the actions of
individuals and associations take place (ibid.; see also Jones, 1999). The
stakeholder literature has emphasized, albeit without much real empirical
analysis, the variation across national political environments. Doh and Guay
(2006) and De Bakker and den Hond (2008), for instance, highlight such
governmental variance, notably between Europe and the United States, by
referring to: the opposition of several European countries to the marketing
of genetically modified crops and food in Europe; the insistence of US-firms
and Government to enforce patent protection for Aids/HIV medication
in developing countries; as well as the differing national commitments to
CO2-emission reductions of the Kyoto Protocol (ibid.).
Institutions favoring a strong governance and control over government
officials, for instance, through less corruption and more democracy, should
also offer more opportunities for activists to voice their concerns and criti-
cize business practices. As governments seem to influence companies in their
behaviors and decisions, they might also influence activists in their targeting
strategies. Therefore, favorable country contexts might actively encourage
stakeholder activism and strategic criticism for firms for certain countries
(see, e.g., Jones, 1999).
H7. The less corrupt and more democratic the institutions of a country,
the higher the likelihood for firms incorporated in that country to
become targeted by Internet/media criticism.
DATA AND METHODS

To explore our hypotheses empirically, we built a unique dataset of firms
that were exposed to CSR-related criticism from 2006 to 2009. The firm
sample consists of the worlds most admired firms, annually ascertained by
Fortune. To become eligible, corporations must surpass $10 billion in
revenue and rank among the largest firms within their industry peers
(by revenue). We chose this sample as the firms included are typically large
multinationals, which exposes them to various different sociocultural
contexts and stakeholder scrutiny. Data for these firms are also relatively
easy to access. The Fortune survey for the period 20062009 contains
642 companies, which have been part of the ranking at least once, and we
obtained detailed data for 451 of them. We matched this sample with
another database covering Internet/media criticism. The database on
CSR-related criticism was, in its original form, provided by the Swiss-based
consultancy Reprisk, whose mission is to capture the impact of external
factors on a firms reputation. It includes criticism information collected
for many firms part of the ranking since 2006. Each firm-specific criticism
contains detailed information about the content, initiator, date, and location
of the reported stakeholder action, as well as its level of severity.
We tried to make sure that our database contained only criticism, and
not other stages of the campaigns production chain in two ways. First,
we considered only the first criticism and avoided any further one, or any
escalation of the same criticism: once a criticism has been made, it is not
considered a second time in our data. Second, we looked at all the criticism
one by one and excluded the ones that were referring to larger actions such
as boycotts or organized actions.
The constructed panel dataset merges the data on firms and criticism
exposure for the 20062009 period, and comprises 8,054 Internet and
media criticism observations for the Fortune-rated companies. Criticism is
considered from a large number of different sources (both Internet and
media sources) expressed in nine languages.2 The collapsed dataset gener-
ates 1,419 firm-level observations once missing data are taken into
account. This results in 4.7 criticisms per firm and year on average (see
Table 2), but with significant variations across firms; some firms being
criticized often and others hardly or never. Further, the severity of each
criticism, as evaluated by Reprisk, varies between three degrees (13, see
Fig. 2), with a value of 1 indicating light criticism (e.g., the firm being
criticized for being slow at issuing an environmental report), whereas
3 indicates very severe criticism for the company reputation (e.g., the firm
being accused of jeopardizing customers health). These different levels of

severity are coded for each criticism reported by two independent analysts
of the consultancy, based on several predefined criteria that determine the
extent and severity of the accusation. Tables 2 and 3, as well as Figs. 2 and
3, provide a description of the criticism included in our database: per indus-
try, per country, number of criticism and severity.
Dependent Variable(s)
Because our empirical approach is exploratory, we consider different esti-

mation methods using both the firm sample and the criticism samples.
First, we look at the probability of a firm being criticized in our sample,
and consider probit models in which the dependent variable takes a value
of 1 if a firm has been criticized in a given year. Second, we take a first
look at the intensity of the criticism and consider whether there are dif-
ferences in the factors that lead firms to be targeted by severe criticism,
Table 2. Number of Criticisms Collected 20062009.

Variable/Year 2006 2007 2008 2009
Number of criticisms p.a. (TCQ) 604a 1,657 3,269 2,524

Number of firms included in the criticism database 451 451 451 451
Average number of criticisms per firm 1.34 3.67 7.25 5.60
a
The database for CSR criticism was only established in 2006; this explains the fewer number
of firm-criticism observations for 2006, provided in Tables 1 and 2.
Table 3. Number of Criticism per Country of Origin of the Firms.

Country Total Country Total
USA 3,741 Australia 128

UK 1,043 Others 121
Germany 777 Brazil 86
France 541 Luxembourg 84
Netherland 486 Russia 84
Switzerland 419 Canada 70
Japan 318 Finland 66
China 185 Italy 66
Korea 154 Belgium 52
Spain 136 Indonesia 41
Fig. 2. Number and Severity Levels of Corporate Criticisms.
instead of the overall criticism. We therefore compute a binary outcome

variable (severe criticism) that indicates whether a company has been
exposed to severe stakeholder activism (categories 2 and 3 in our data)
over a year. Again, we just use simple probit models. Third, we move
from the firm level to the criticism level and create an ordinal-dependent
variable that includes all four possible different levels of harm/critique a
corporation can be exposed to: from zero (i.e., zero harm or no criticism)
to 3, with 3 being the most severe level of criticism activity. We then use
ordered probit to explore the drivers for the intensity of criticism. For all
these models, we include year dummies and compute robust standard
errors by clustering at the firm level.
Firm-Level Variables
To test the hypotheses H1H4 assessing the influence of corporate factors
on critique exposure, we first estimate the influence of financial dimensions:
firm assets, considered in logarithm to account for skewness, and cash flow,
as proxies of resource-richness. To account for firm profitability, we further
include Return on assets (ROA). Sales also in logarithm are inserted
Fig. 3. Number of Criticisms per Industry.
as a measure of firm size. All financial data were obtained from Thomson-
Datastream and, for the sake of comparability, considered in logs. To
approximate firm visibility in terms of brand awareness, we compute
another binary variable that indicates whether a firm possesses top brands
(annually ranked by Interbrand)3 in its corporate portfolio. The absolute
brand values were not included because of insufficient data.
With regard to variables regarding a firms CSR performance (as
hypothesized in H4), we include a binary variable that distinguishes
companies that belong to the Dow Jones Sustainability Index (DJSI).4 As
an approximated measure of ecological/environmental performance, we
construct a dummy variable that indicates whether the firm disclosed its
carbon footprint (proxied by the data from the carbon disclosure project,
CDP).5 Absolute scores were not taken into consideration as they would
have restrained the estimation sample significantly. We finally compute a
binary variable that controls whether a company is publicly listed.
Industry-Level Variables
To identify industry-specific effects, as hypothesized in H5, we compute
dummy variables for each industry sector based on the sector-classification
of the criticism database.
Country-Level Variables
To evaluate whether an NGO/medias firm selection depends on the
particularities of the country in which the firm is established, we inserted
variables that mirror the (institutional) quality of a country as well as its
socioeconomic situation following measures used in previous studies (see,
e.g., Gallup, Sachs, & Mellinger, 1999; Kaufmann, Kraay, & Mastruzzi,
2011; Sachs, 2003).
First, to estimate the influence of socioeconomic factors as hypothesized
in H6, we include the GDP (Gross Domestic Product) annual growth
rate per capita and the countrys annual unemployment rate as economic
measures. To account for sociological factors, we insert a countrys
education index, as well as an indicator of its standard of living. The data
for all three variables was obtained from the United Nations Development
Program (UNDP) database.6 Further, to take into account the growing
importance of environmental awareness in many nation-states (Halme &
Huse, 1997; Philippe & Durand, 2011), we also include a measure indicating
a countrys environmental consciousness, proxied by the environmental
performance index (EPI).7
Second, to capture the effects of institutional factors as hypothesized in
H7, we include governance indices that constitute approved proxies for
indicating a countrys institutional quality (see, e.g., Kaufmann et al., 2011;
Sachs, 2003). Such governance data was obtained from the Worldwide
Governance Indicators (WGI) project.8 Therefore, we consider four
variables: Voice and accountability indicates to what extent a countrys
citizens are able to participate in selecting their government, as well as free-
dom of expression, freedom of association, and a free media (Kaufmann,
Kraay, & Mastruzzi, 2009, p. 6). Political stability measures the perceptions
of the likelihood that the government will be destabilized or overthrown by
unconstitutional or violent means, including domestic violence and terrorism
(ibid.). Finally, control of corruption mirrors perceptions of the extent to
which public power is exercised for private gain, including both petty and
grand forms of corruption, as well as capture of the state by elites and
private interests (ibid.).
RESULTS
Table 4 presents descriptive statistics and pairwise correlations. Results of
our empirical investigation are provided in Table 5. We first discuss the
results of the targeting likelihood estimations (Models 12), and then
move to explore the determinants of criticism intensity (Model 3).
Results for Model 1 General Criticism Sample
In the first column of Table 5, we estimate the likelihood that firms will
become a target of Internet/media criticism. Among the corporate factors,
the two variables signaling firm size and visibility sales and brand
awareness have a significant impact on the targeting likelihood, providing
support to H2. Likewise, for stock-listed corporations, belonging to the
DJSI increases the targeting likelihood (supporting H3). This implies that
corporations with a stronger CSR orientation are more likely to be targeted,
a result that is in line with what was found by Luo et al. (2010) regarding
media coverage of oil spills. These first results are strongly in line with
the idea that Internet/media criticism is already a strategic stage, that is,
that activists pick the corporate targets that are the most likely to attract
attention to their cause.
With regard to industry effects, service industries such as travel/leisure,
insurance/banking, telecom, and media particularly reduce the targeting
likelihood. This is in line with the argument that service industries are
perceived as less polluting industries than those of the manufacturing sectors
(see, e.g., Cole, 2000). Interestingly, low statistical support could be provided
for H4 and H5: among the controversial industries, only Mining is more
likely to be targeted, and traditional B2C sectors, for example, Food and bev-
erage or Automotive, do not seem to affect the targeting likelihood either.
Regarding country effects, our results suggest, on the one hand, that the
targeting likelihood is high for firms incorporated in countries for which insti-
tutions are less stable and democratic. The coefficients of voice/accountability
and political stability as proxies of political/institutional influence (H7)
Table 4. Descriptive Statistics and Correlation.
Mean S.D. Criticism Criticism Cash ROA Sales Assets DJSI Envt. Notlisted Top GDP Unemployment Life Educ. Envt. Voice Pol. Anticorrupt
Intensity flow Disclos. Brand Growth Quality Consc. Account Stab.
Criticism 0.72 0.23

Criticism 1.36 0.37 1.00
Intensity
Cash flow 8.52 3.08 0.07 0.12 1.00
ROA 0.06 0.02 0.03 0.05 0.23 1.00
Ln Sales 17.8 12.05 0.29 0.32 0.24 0.10 1.00
Ln Assets 10.2 5.3 0.24 0.26 0.00 0.12 0.67 1.00
DJSI 0.51 0.12 0.01 0.02 0.02 0.02 0.07 0.12 1.00
Envt. Disclo. 0.69 0.53 0.25 0.27 0.11 0.11 0.45 0.42 0.15 1.00
Notlisted 0.04 0.08 0.15 0.12 0.06 0.00 0.15 0.17 0.05 0.31 1.00
Top Brand 70.3 0.18 0.04 0.07 0.02 0.00 0.14 0.16 0.03 0.09 0.10 1.00
GDP Growth 1.45 0.89 0.04 0.05 0.05 0.07 0.01 0.02 0.15 0.18 0.19 0.07 1.00
Unemployment 4.61 0.95 0.08 0.04 0.14 0.06 0.18 0.25 0.30 0.10 0.06 0.03 0.07 1.00
Life Quality 77.4 0.48 0.04 0.05 0.06 0.07 0.03 0.06 0.02 0.08 0.18 0.01 0.43 0.25 1.00
Education 77.2 0.18 0.03 0.04 0.07 0.01 0.10 0.14 0.11 0.05 0.06 0.03 0.31 0.43 0.50 1.00
Envt. Consc. 81.9 0.26 0.10 0.12 0.04 0.06 0.07 0.14 0.25 0.20 0.21 0.05 0.57 0.33 0.18 0.02 1.00
Voice Account 85.7 0.17 0.06 0.04 0.04 0.01 0.09 0.11 0.27 0.31 0.24 0.11 0.73 0.25 0.55 0.24 0.51 1.00
Pol. Stab. 66.4 0.75 0.02 0.05 0.01 0.05 0.08 0.07 0.27 0.18 0.12 0.04 0.36 0.22 0.32 0.21 0.43 0.53 1.00
Anticorrupt 87.4 0.34 0.04 0.05 0.02 0.04 0.05 0.05 0.25 0.27 0.18 0.09 0.66 0.08 0.57 0.28 0.52 0.59 0.52 1.00
First column (Criticism): Firm sample, n = 1,419; Other columns: Full issue sample, n = 8,054.
Table 5. Likelihood of a Firm Being Targeted.

Variables Probit Model (1) Probit Model (2) Model (3)
General criticism Severe criticism Criticism intensity
Probit Probit Ordered probit
Corporate influence
Ln cash flow 0.025 0.045 0.0380

(0.0698) (0.082) (0.054)
ROA (return on assets) 0.310 0.137 0.710
(0.878) (0.930) (0.841)
Ln sales 0.471*** 0.359* 0.401***
(0.099) (0.101) (0.075)
Ln assets 0.066 0.211* 0.054
(0.096) (0.114) (0.072)
Social performance (DJSI) 0.180* 0.152 0.099
(0.112) (0.189) (0.175)
Environmental disclosure 0.0170 0.120 0.130
(0.120) (0.119) (0.085)
Nonpublic listed dummy 1.007 1.196 1.047
(1.224) (1.404) (1.359)
Top brand dummy 0.655*** 0.240** 0.337***
(0.128) (0.099) (0.083)
Country/institutional factors
GDP growth rate per capita 0.0221 0.020 0.067*

(0.051) (0.077) (0.029)
Unemployment 0.160 0.279*** 0.011
(0.193) (0.451) (0.009)
Life quality 0.055** 0.029 0.022***
(0.0242) (0.026) (0.0072)
Level of education 0.010 0.284*** 0.058*
(0.041) (0.072) (0.003)
Environmental consciousness 0.143*** 0.061** 0.042***
(0.035) (0.029) (0.013)
Voice and accountability 0.081** 0.191*** 0.0028
(0.0375) (0.068) (0.0082)
Political stability 0.0355* 0.008 0.022***
(0.022) (0.028) (0.0036)
Control of corruption 0.101*** 0.081** 0.0171*
(0.032) (0.044) (0.008)
Industry influence
Aerospace/defense 0.490 0.312 1.092**

(0.577) (0.355) (0.527)
Automotive 0.302 1.508* 0.720*
(0.371) (0.692) (0.376)
Chemicals 0.555 0.011 1.052**
(0.689) (0.422) (0.529)
Table 5. (Continued )
Variables Probit Model (1) Probit Model (2) Model (3)
General criticism Severe criticism Criticism intensity
Probit Probit Ordered probit
Construction 0.90 1.112** 1.398***

(0.575) (0.488) (0.527)
Financial industry 1.098* 1.428** 1.366**
(0.610) (0.644) (0.571)
Food/beverage 0.0833 1.677** 1.181**
(0.362) (0.712) (0.343)
Insurance 1.429** 1.326*** 1.978***
(0.609) (0.279) (0.579)
Mining 1.126** 1.241** 1.211*
(0.402) (0.474) (0.544)
Oil/gas 0.528 0.367 0.942*
(0.599) (0.389) (0.521)
Personal-/household goods 0.232 1.478** 0.419
(0.356) (0.752) (0.557)
Pharmaceuticals 0.390 0.590 0.397
(0.355) (0.665) (0.429)
Retail 1.002*** 1.101** 0.893**
(0.359) (0.347) (0.406)
Tobacco 0.345 1.901** 1.131*
(0.658) (0.788) (0.272)
Utilities 0.399 0.417 1.058**
(0.555) (0.525) (0.518)
Computer/electronics 0.371 1.053 0.619
(0.367) (0.678) (0.489)
Telecom/media 1.402*** 1.634*** 1.632***
(0.360) (0.377) (0.370)
Travel/leisure 0.800** 1.112 1.033***
(0.361) (0.373) (0.411)
Raw materials 0.555 1. 283** 1.027**
(0.572) (0.552) (0.510)
General industrials 1.165** 1.751*** 1.541***
(0.558) (0.563) (0.566)
Constant 12.82*** 9.48*** Cut1 6.287***
(4.122) (4.553) Cut2 7.016***
Cut3 9.863***
Year dummies Yes Yes Yes

Wald Chi2 306.62*** 238.94*** 206.71***
Pseudo R2 21% 25% 19%
Observations 1,419 1,419 8,054
Robust standard errors in parentheses.

***p < 0.01, **p < 0.05, *p < 0.1.
are both negative and significant. These results are further supported by the
socioeconomic influence factors hypothesized in H6: firms incorporated in
countries of low-living standards are preferred targets. Together, these results
do not support the idea of strategic criticism and would rather suggest that a
part of activists behaviors is about monitoring of corporate issues for firms
from countries in which issues are likely to occur.
On the other hand, evidence is also provided for superior country variables.
Companies that are incorporated in countries of high environmental
consciousness (H6) attract unwanted NGO/media attention. This finding
is endorsed by the corruption indicator, a further measure of institutional
quality (H7): the better a firms home-country controls corruption, the
higher the companys targeting likelihood becomes. These results on envir-
onmental consciousness and control of corruption are in line with what
we would expect if activists select firms from countries that are positioned
high up in the Environmental Kuznets Curve.
Overall, these results are thus somewhat two-sided and seem to suggest
that activists adopt complementary targeting behaviors. They talk to a cer-
tain audience by targeting large firms in countries in which CSR is perceived
important. However, activists also play a monitoring role and uncover
dubious behaviors for firms from countries in which institutions are weak.
With regard to the actual strength of the effects, the estimated coeffi-
cients in probit, nonlinear regression models cannot be interpreted directly
and we thus calculated marginal effects. These marginal effects for firm
sales and firm (brand) awareness are 14% and 17%, respectively. Further,
the marginal effect of social performance accounts for a 5% increase in the
probability of becoming a target. Companies incorporated in countries of
high environmental attitude (consciousness) increase the targeting likeli-
hood by 5%. Companies incorporated in low-living quality countries
increase their targeting likelihood by 2%. The institutional parameters
voice/accountability and political stability increase the targeting likelihood
by 2% and 1%, while with growing control of corruption the targeting like-
lihood increases by 3%. The largest marginal effects were calculated for
the service industries that range between 28% and 52%.
Model 2: Likelihood of Being Severely Criticized
Regarding the likelihood of being severely targeted by Internet/media

activism (Model 2 in Table 5), results regarding corporate factors are overall
in line with what was observed for the General criticism sample, except
that the coefficient regarding Assets is now significant, but only at the
10% level.
Additionally, in contrast to what was observed for the General criticism

model, industry effects appear more impactful as drivers of severe criticism. In
line with H5, individual industry-fixed effects indicate that close-to-consumer
operating industries attract unwanted Internet/media attention. Firms
producing food/beverages, as well as Personal/household goods, which
include the textile/fashion sector, or Automotive, are more likely to be
targeted. Moreover, operating in controversially discussed industries
such as mining and tobacco positively influences the likelihood of severe
criticism, and therefore lends some support to H4.
On the country level, five variables seem to encourage criticism activity
relative to firms: companies are especially targeted if they are incorporated in
countries of high environmental attitudes, efficient corruption control, high
educational levels and low unemployment rates. Again, these results seem
very much in line with the idea that activists pick targets from countries that
tend to be high in the Environmental Kuznets Curve, that is, countries in
which citizens are likely to be sympathetic to the activists cause.
With regard to the magnitude of the coefficients, the largest marginal
effects were again estimated for the significant industry variables that range
between 31% and 45%. On the country level, low unemployment rates and
high education levels increase the targeting likelihood by 10%, supporting
the idea that activists target firms to fit with a certain audience of educated
and relatively wealthy readers. The remaining significant socioeconomic
and institutional parameters show marginal effects between 2% and 6%.
Model 3: Determinants of Criticism Intensity
Moving now to criticism as dependent variable, the results of the ordered

probit specifications of Model 3 are presented in the third column of
Table 5. The three cut-off levels are significant, suggesting that the proposed
criticism thresholds constitute different severity levels of critical allegations.
A positive sign of the coefficients implies a higher likelihood that the raised
critique is harsh or very harsh. Regarding firm-specific factors, positive and
significant coefficients can be observed for firm sales, and brand awareness,
in line with H2. Results are also consistent with our previous estimates
regarding industry effects: as for the likelihood of being severely targeted,
criticism are more likely to be harsh for controversial sectors, like Mining or
Tobacco, and for close-to-consumers sectors such as Food & Beverage or
Automotive. This is again consistent with H4. Together, these results support
the idea of a relatively close connection between mundane and more severe
criticism, and our concept of production chain for anticorporate campaigns.
On the country level, factors such as companies being incorporated in

countries with high environmental attitudes, education and anticorruption
levels, and GDP growth rates raise the likelihood of more intense activism.
At the same time, the negative signs of the two significant coefficients living
quality and political stability indicate that inferior life quality also implies
that Internet/media criticism will likely be severe. Again, our results regarding
country-specific and international factors are thus not one-sided. Activists
seem to engage both in strategic behaviors by making harsher criticism
against firms from countries in which socioeconomic and institutional condi-
tions will create a fertile ground for activists claim, but activists also monitor
and make harder criticism against companies from countries that are less
stable institutionally and poorer. Both strategic and monitoring behaviors
therefore seem to be at play.
DISCUSSION CONCLUSION
This paper makes three contributions to the private politics literature.
First, our results suggest that media/Internet criticism against firms seems
to be driven by factors similar to what has been previously observed for
full-fledged activist campaigns. As such, this study provides support to the
idea that public criticism can be seen as the first step of a chain leading to
the production of these campaigns. Depending on the model, firm sales,
assets, and brand awareness positively influence the targeting likelihood.
This is in line with studies showing that activists pick targets not only to
highlight issues but also to attract attention and create legitimacy, which is
more likely with large and visible firms (King & Soule, 2007; Lenox &
Eesley, 2009). Internet/media criticism should therefore be seen as one of
the first steps in private politics, potentially leading subsequently to larger
campaigns, rather than standalone activities with their own logic.
Similarly, some of our results indicate that a firms CSR orientation, as
measured through its inclusion in DJSI, makes it more likely to be targeted.
This is also in line with social movement/stakeholder theory in that transpar-
ently operating firms disclosing CSR-data offer more room for criticism (see,
e.g., Luo et al., 2010). These findings also support the strategic accounting
literature, which associates the provision of corporate information with
a competitive disadvantage for the disclosing firm (see, e.g., Verrecchia,
2001). Alternatively, these firms might be particularly targeted by activists
who suspect firms of green-washing Laufer (2003).
Additionally, we also found that industry effects matter, especially in

explaining more severe criticism. Companies operating in industries that
are controversial (mining, tobacco) or close to final customers (e.g., food/
beverage, textile sector) seem to be preferred targets for activists when
engaging in severe media criticism. These results indicate that, in these
industries, activists might better emphasize the urgency and legitimacy of
their claims and protests as suggested by stakeholder theory (Mitchell,
Agle, & Wood, 1997). Moreover, these findings endorse explanations
stating that activists select these industries where they expect higher returns
in terms of corporate compliance/responses to the activists requests
ex post (Baron & Diermeier, 2007; King & McDonnell, 2015). These results
regarding industry effects are also in line with the boycott literature:
consumers are willing to support a boycott as long as affordable substitute
products are available (Diermeier & van Mieghem, 2005). This is certainly
the case in the food/beverage sector, and with personal/household goods
and clothes.
One important implication of these results is that managers should not
take criticism lightly. If criticism is already a strategic stage that might lead
to the production of full campaign against their firm, managers need to
anticipate and think about possible reactions even at this early criticism
stage. This idea speaks to two streams of literature on private politics.
First, it is in line with existing empirical results that show that negative
public exposure, both in the media and in the Internet, has a significant
and positive impact on the development of CSR activities among firms that
are exposed (Zyglidopoulos, Carroll, Georgiadis, & Siegel, 2012). Second,
our analysis is also in line with the theoretical literature on private
politics, which tend to suggest that full campaigns are often preempted
and do not occur because both the activists and the firm have incentives
to negotiate at early stages and avoid the costs for both of running the
full campaign (Baron & Diermeier, 2007). Criticism might thus be an
important early point of contact between firms and activist, leading to the
development of more sustainable practices even before any major conflict
has emerged.
The second contribution of this paper to the literature has to do with
how country-specific and institutional differences influence private politics,
which have not been explored in previous literature. We find that both the
occurrence and the intensity of criticism against firms are, at least partially,
driven by these country-specific and institutional factors. Activists locate or
mediatize their campaigns against firms in countries that are extreme in
terms of institutional quality, as well as socioeconomic performance. Either
they target firms incorporated in healthy and wealthy economies where an

environmental consciousness is already highly advanced (DiMaggio & Powell,
1983; Matten & Moon, 2008), or they select companies that operate in
unstable and poorer economies with less developed ecological sensitivity
(Kaul, Conceicao, Le Goulven, & Mendoza, 2003). In this latter case, NGOs/
media might specifically target firms operating in such countries to exert
pressure on the firms taking on (more) social/environmental responsibility.
Overall, these results seem to suggest that activists pursue a twofold
targeting strategy: selecting firms incorporated in high welfare-states where
the audience is likely to be supportive, while at the same time also providing
some monitoring of firms from countries with weaker economic contexts
and less institutional stability. So, regarding their welfare impact, should we
expect activists to provide the monitoring and whistle-blowing function that is
necessary for private regulation to emerge? Overall, not entirely. Activists,
even when they engage in criticism, are already working on the manufacturing
of anticorporate campaigns and are therefore going to strategically select
the countries, the industries and the firms they go after. So, should we be
pessimistic about the role that activism can play in achieving sustainable
development? Probably not either. Activists do play their role of agents of
change in countries and industries for which, in line with the Environmental
Kuznets Curve approach, there will be a favorable audience for their
activities. As other countries continue to develop, the role that activists can
play in sustainable development should expand as well.
Limitations
Despite the interesting and new findings presented and discussed above,
this study is certainly subject to limitations. First, regarding the connection
between our conceptual framework and our data, we did not test the
sequential dimension of our campaign production chain. In particular, we
did not study empirically the ordering going from the selection of countries,
industries and firms to criticize and then ultimately leading to full-fledged
campaigns. Even if it seems unlikely, it might be that campaigns come first
and criticism heat up after that. Anecdotal evidence seems to speak against
this (Wapner, 2009). However, this is something that should certainly be
studied empirically as it would reinforce our understanding of the manufac-
turing of campaigns and its implication on firms. In the same way, it would
be interesting to study why some criticism lead to full-fledged campaigns
while others do not (Wright & Boudet, 2012).
Second, there are limitations to the data used here. By focusing on firms
present in the Fortune survey, we chose to explore Internet/media criticism for
very large companies that are likely to be primary targets for activists and
providing country/institutional variations. It would be interesting to replicate
the analysis for a broader sample of firms. That being said, it is unclear
whether including smaller firms would dramatically change our results.
Considering only large firms reduces the variation in the sample and should
make it more difficult to obtain results regarding firm-specific dimensions.
Similarly, regarding institutional effects, we focused on institutional
dimensions for firms home countries. It would be, however, interesting
to also explore what drives Internet/media criticism in host countries. If
activists play a monitoring role and become substitutes for deficient
governance in certain countries as well as counter-balance the influence of
multinationals (Scherer & Palazzo, 2011), then this role should be observed
empirically in the foreign countries where these multinationals operate.
This certainly should be explored in future research.
We also do not know if the higher scrutiny and severity of criticism in
some countries is related to the fact that activists or NGOs play tougher
or whether there are just more activists in these countries, and therefore
more criticism. In that sense, future studies should explore how country-
specific factors and institutional conditions affect the supply of activism
and how this supply of activism determine the number of criticism and of
full campaigns that take place.
There are also potential endogeneity issues with some of our measures.
For instance, our measure of environmental awareness might in fact be
driven by previous activism. A larger dataset including many more years of
observations should allow to address this type of concern in the future.
Fourth, one of the limitations of the data conducted here is that this
data did not allow for a clear identification of who the criticizers were.
Most of the criticism clearly comes from activists acting through the media
and the Internet, but it is often very difficult to pinpoint where the criticism
started and who started it. To better understand the drivers of private politics,
this origin of criticism is something that needs to be studied in more details.
It is also difficult to differentiate Internet versus media criticism.
Publications such as The Economist, for instance, could be on paper or
online. It would be interesting, however, to see whereas Internet criticism,
as they can go viral, tend to become activists preferred option and if they
have a stronger impact on firms than traditional media.
Last, it would be interesting to study firm responses to the criticism they
are confronted with (Lenox & Eesley, 2009). As suggested earlier, one
difference between criticism and full campaigns is that criticism does not
necessarily call for a direct response. However, it does not mean that firms
should not take actions. Should firms provide a public response, for
instance through the media? In which circumstances would such a public
response to criticism be the most appropriate? Conversely, should firms
engage in self-regulation for instance through CSR investment once
they have been criticized in order to prevent the occurrence of a full
campaign? This is in fact what the theoretical literature would suggest.
More theoretical and empirical work is warranted to provide answers to
these questions.
To conclude, this study provides a first exploration of how activists select
the firms they criticize and how this criticism potentially differs but also
interacts with broader campaigns as studied in previous literature. Many
questions, however, remain to be answered regarding the role that activists
play in monitoring firms and the environmental or social issues these firms
create, as well as, more generally, regarding the role that these activists can
play in creating self-regulating and sustainable economic systems.
NOTES
1. Naturally, a private politics chain leading to a full campaign against one firm
might also lead subsequently to another private politics chain targeting another
firm, as suggested by the idea of sequential targeting (Baron & Diermeier, 2007).
2. Criticism regarding companies comes from various sources (i.e., media,
Internet postings, news agencies, NGO reports, etc.) reporting in English, Spanish,
German, French, Italian, Portuguese, Mandarin, Korean, and Russian.
3. http://www.interbrand.com/en/best-global-brands/best-global-brands-2008/
best-global-brands-2010.aspx
4. http://www.sustainability-index.com/07_htmle/data/djsiworld.html
5. https://www.cdproject.net/en-US/Results/Pages/reports.aspx
6. http://hdr.undp.org/en/data/build/
7. http://epi.yale.edu/Countries
8. http://info.worldbank.org/governance/wgi/index.asp
REFERENCES
Argenti, P. A. (2004). Collaborating with activists: How Starbucks works with NGOs.
Barnett, M. L., & Hoffmann, A. E. (2008). Beyond corporate reputation: Managing
reputational interdependence. Corporate Reputation Review, 11(1), 19.
Baron, D. P. (2003). Private politics. Journal of Economics and Management Strategy,

12, 3166.
Bartley, T., & Child, C. (2014). Shaming the corporation: The social production of targets and
the anti-sweatshop movement. American Sociological Review, 79, 653679.
Beck, T., & Levine, R. (2005). Legal institutions and financial development. In C. Menard &
M. Shirley (Eds.), Handbook of new institutional economics (pp. 251278).
Dordrecht: Springer.
Beck, U. (2000). What is globalization? Cambridge: Polity Press.
Bonardi, J. P., & Keim, G. (2005). Corporate political strategies for widely salient issues.
Academy of Management Review, 30(3), 555576.
Bunn, I. D. (20032004). Global advocacy for corporate accountability: Transatlantic
perspectives from the NGO community. American University International Law Review,
19, 12651306.
Cole, M. A. (2000). Air pollution and dirty industries: How and why does the composition of
manufacturing output change with economic development? Environmental and Resource
Economics, 17(1), 109123.
De Bakker, F. G. A., & den Hond, F. (2008). Introducing the politics of stakeholder
influence A review essay. Business & Society, 47(1), 820.
Den Hond, F., & de Bakker, F. (2007). Ideologically motivated activism: How activist groups influ-
ence corporate social change activities. Academy of Management Review, 32(3), 901924.
Diermeier, D., & van Mieghem, J. A. (2005, August). A stochastic model of consumer
boycotts. Working Paper (COSM-05-001), Kellogg School of Management,
DiMaggio, P. J., & Powell, W. W. (1983). The iron cage revisited: Institutional isomorphism
and collective rationality in organizational fields. American Sociological Review,
48, 147160.
Doh, J. P., & Guay, T. R. (2006). Corporate social responsibility, public policy, and NGO
activism in Europe and the United States: An institutionalstakeholder perspective.
Journal of Management Studies, 43(1), 4773.
Dyck, A., & Zingales, L. (2002). The corporate governance role of the media. In The right to
tell. The role of mass media in economic development (pp. 107137). Washington, DC:
World Bank Institute.
Eesley, C., & Lenox, M. J. (2006). Firm response to secondary stakeholder action. Strategic
Friedman, M. (1985). Consumer boycotts in the United States, 19701980: Contemporary
events in historical perspective. Journal of Consumer Affairs, 19, 96117.
Gallup, J. L., Sachs, J., & Mellinger, A. D. (1999). Geography and economic development.
International Regional Science Review, 22(2), 179232.
Graves, S. B., Rehbein, K., & Waddock, S. (2001). Fad and fashion in shareholder activism:
The landscape of shareholder resolutions, 19881998. Business and Society Review,
106(4), 293314.
Halme, M., & Huse, M. (1997). The influence of corporate governance, industry and country
factors on environmental reporting. Scandinavian Journal of Management, 13(2), 137157.
Hendry. (2006). Taking aim at business: What factors lead environmental non-governmental
organizations to target particular firms? Business & Society, 45, 4786.
John, A., & Klein, J. G. (2003). The boycott puzzle: Consumer motivations for purchase
sacrifice. Management Science, 49, 11961209.
Jones, M. T. (1999). The institutional determinants of social responsibility. Journal of Business
Ethics, 20(2), 163179.
Kaufmann, D., Kraay, A., & Mastruzzi, M. (2009). Governance matters VIII: Aggregate and
individual governance indicators, 19962008 (pp. 1103). World Bank Policy Research
Working Paper No. 4978.
Kaufmann, D., Kraay, A., & Mastruzzi, M. (2011). The worldwide governance indicators:
Methodology and analytical issues. Hague Journal on the Rule of Law, 3(02), 220246.
Kaul, I., Conceicao, P., Le Goulven, K., & Mendoza, R. U. (Eds.). (2003). Providing public
goods. Managing globalization. New York, NY: Oxford University Press.
King, B., & McDonnell, M. H. (2015). Good firms, good targets: The relationship between
corporate social responsibility, reputation, and activist targeting. In K. Tsutsui &
A. Lim (Eds.), Corporate social responsibility in a globalizing world. Cambridge:
King, B. G. (2008). A political mediation model of corporate response to social movement
activism. Administrative Science Quarterly, 53, 395421.
King, B. G., & Pearce, N. A. (2010). The contentiousness of markets: Politics, social movements,
and institutional change in markets. Annual Review of Sociology, 36, 249267.
King, B. G., & Soule, S. A. (2007). Social movements as extra-institutional entrepreneurs: The
effect of protests on stock price returns. Administrative Science Quarterly, 52, 413442.
Kingsley, A., Vanden Bergh, R., & Bonardi, J. P. (2012). Political markets and uncertainty:
Insights and implications for integrated strategy. Academy of Management Perspectives,
26, 5267.
Koku, P. S., Akhigbe, A., & Springer, T. M. (1997). The financial impact of boycotts and
threats of boycotts. Journal of Business Research, 40, 1520.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. (1998). Law finance. Journal of
Political Economy, 106, 11131155.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. (2000). Investor protection and
corporate governance. Journal of Financial Economic, 58, 327.
Lamin, A., & Zaheer, S. (2012). Wall street vs. main street: Firm strategies for defending legitimacy
and their impact on different stakeholders. Organization Science, 23(1), 4766.
Laufer, W. S. (2003). Social accountability and corporate greenwashing. Journal of Business
Ethics, 43, 253261.
Lenox, M. J., & Eesley, C. (2009). Private environmental activism and the selection and
response of firm targets. Journal of Economics & Management Strategy, 18(1), 4573.
Luo, J., Meier, S., & Oberholzer-Gee, F. (2010, May). Whats news? CSR and corporate reputation
management. Working Paper.
Matten, D., & Moon, J. (2008). Implicit and explicit CSR: A conceptual framework for a
comparative understanding of corporate social responsibility. Academy of Management
Review, 33(2), 404424.
Meznar, M. B., & Nigh, D. (1995). Buffer or bridge? Environmental and organizational
determinants of public affairs activities in American firms. Academy of Management
Journal, 38(4), 975996.
Mitchell, R. K., Agle, B. R., & Wood, D. J. (1997). Toward a theory of stakeholder identification
and salience: Defining the principle of who and what really counts. Academy of
Ocasio, W. (1997). Towards an attention-based view of the firm. Strategic Management

Journal, 18, 187206.
Pfeffer, J., & Salancik, G. (1978). The external control of organizations: A resource dependence
perspective. New York, NY: Harper & Row.
Philippe, D., & Durand, R. (2011). The impact of norm-conforming behaviors on firm reputation.
Strategic Management Journal, 32, 969993.
Rehbein, K., Waddock, S., & Graves, S. B. (2004). Understanding shareholder activism:
Which corporations are targeted. Business & Society, 43(3), 239267.
Sachs, J. (2003). Institutions dont rule: Direct effects of geography on per capita income. NBER
Working Paper Series No. 9490, Columbia University.
Sasser, E. N., Prakash, A., Cashore, B., & Graeme, A. (2006). Direct targeting as an NGO
political strategy: Examining private authority regimes in the forestry sector. Business
and Politics, 8(3), Article 1.
Scherer, A., & Palazzo, G. (2011). The new political role of business in a globalized world:
A review of a new perspective on CSR and its implications for the firm, governance,
and democracy. Journal of Management Studies, 48(4), 899931.
Stern, D., Common, M., & Barbier, E. (1996). Economic growth and environmental degradation:
The environmental Kuznets curve and sustainable development. World Development,
24(7), 11511160.
Teegen, H., Doh, J. P., & Vachani, S. (2004). The importance of Nongovernmental
Organizations (NGO) in global governance and value creation: An international business
research agenda. Journal of International Business Studies, 35, 463483.
Verrecchia, R. E. (2001). Essays on disclosure. Journal of Accounting and Economics,
32, 97180.
Wapner, P. (2009). Environmental activism and world civic politics. Albany, NY: State
University of New York Press.
Weber, J., & Marley, K. A. (2012). In search of stakeholder salience: Exploring corporate
social and sustainability reports. Business & Society, 51(4), 626649.
Wright, R., & Boudet, H. (2012). To act or not to act: Context, capability and community
response to environmental risk. American Journal of Sociology, 118(3), 728777.
Yaziji, M. (2004). Turning gadflies into allies. Harvard Business Review, 82, 110115.
Yaziji, M., & Doh, J. (2009). NGOs and corporations: Conflict and collaboration. Cambridge:
Yurday, E., & Baron, D. P. (2004). Anatomy of a corporate campaign: Rainforest Action
Network and Citigroup. Stanford Graduate School of Business, Case: P-42A.
Zyglidopoulos, S. C., Carroll, C. E., Georgiadis, A. P., & Siegel, D. S. (2012). Does media
attention drive corporate social responsibility? Journal of Business Research,
65, 16221627.
PART III
INTEGRATED
POLITICAL STRATEGY
NAVIGATING NATURAL
MONOPOLIES: MARKET
STRATEGY AND NONMARKET
CHALLENGES IN RADIO AND
TELEVISION AUDIENCE
MEASUREMENT MARKETS
Hillary Greene and Dennis A. Yao
ABSTRACT
This paper explores how firms within the audience measurement industry,
specifically its radio and television markets, have navigated myriad
market and nonmarket challenges. The market strategies and the non-
market forces that constrain those strategies are largely defined by two
features: the delineation of its geographic markets by political bound-
aries and markets that have natural monopoly characteristics. While the
pre-monopoly stage or periods of competition may be comparatively
short-lived, they are still telling. Monopolists undertake market strate-
gies designed to ensure that they are not supplanted and nonmarket
actions geared to avoiding undesirable constraints and reputational
damage. Depending on their legal and regulatory environment, customers

ISSN: 0742-3322/doi:10.1108/S0742-332220160000034011
367
368 HILLARY GREENE AND DENNIS A. YAO
of the measurement services have used both market and nonmarket

actions to mitigate the market power of the audience measurement
firms. This paper focuses primarily on the U.S. radio and television
audience measurement markets that Arbitron and Nielsen, respectively,
have dominated for decades. Non-U.S. markets, which frequently feature
Americas foremost firms, illustrate alternatives to Americas largely
laissez-faire approach.
Keywords: Monopoly; nonmarket strategy; television ratings industry
INTRODUCTION
Television audience measurement in the United States has long been almost
the sole province of A. C. Nielsen. Nielsens dominance which began during
the industrys formative years has persisted despite decades of oftentimes
profound changes in what was being measured (e.g., size of the television
viewing audience, the advent of cable television with its explosion of chan-
nels, invention and diffusion of home recording devices, and, recently,
cross-platform viewing) and how it was measured (e.g., measurement meth-
odologies and devices, computer-facilitated data analysis). Each of these
changes provided prospective entrants an opportunity to undermine
Nielsens dominance, yet none has done so. Radio audience measurement in
the United States also became dominated by a single firm while in Europe,
nearly every television ratings market and most radio ratings markets now
have a single provider (Commission of the European Communities, 2008).
Few companies have experienced as much sustained success as Nielsen
in TV ratings and American Research Bureau (ARB), now Arbitron, in
radio ratings. This paper explores how these dominant firms have
attempted to maintain market hegemony. We do not exhaustively chronicle
their strategies or government and legal constraints to the exercise of these
strategies. Given the radio and television rating industries important com-
monalities including a certain shared history, they are usefully considered
in tandem. ARBs and Nielsens experiences will be the central focus of
these business histories.1 Instead, we focus primarily on a selected set of
illustrative events in the United States. We also focus on the economic
function of ratings in terms of offering quantification of the value of adver-
tising and do not directly consider their interplay with various public values
Navigating Natural Monopolies 369
including cultural diversity and programming quality, nor do we focus on

the accuracy of various measurement methodologies employed or consider
media advertising more generally.
The audience measurement industry provides a useful context for exam-
ining market and nonmarket actions that have important parallels to other
markets with a strong tendency toward monopoly (e.g., computer operat-
ing systems, various parts of an electric utility value chain). Because the
markets long-term tendency toward monopoly would likely be anticipated
by firms competing in the early stages of the market, the evolution of such
markets typically consists of a period in which firms compete for market
dominance followed by a period of monopoly exploitation and mainte-
nance. Firms competing for dominance may be tempted to act aggressively
with respect to the gray areas of the law, in the hope that any dominance
secured will be sustained in the future. Once dominant, firms typically
leverage their position to increase value capture and to ensure continued
dominance. But, when dominance approaches monopoly proportions, a
different set of nonmarket constraints become particularly salient.
Despite the U.S. acceptance of monopolies obtained through superior
performance, once achieved, monopoly status attracts both governmental
and private interest which could, in turn, restrict the monopolists actions
in ways that decrease its ability to capture value or even create conditions
that make replacement of the incumbent monopolist more likely. In some
cases, this monopoly outcome leads to direct governmental regulation,
while in other cases, indirect pressure is applied through the use of voice
(Hirschman, 1970) and the application of general laws intended to protect
fair competition or achieve some public interest purpose. That is, in a
market structure where the fear of losing competitors to rivals is dampened
because of the lack of actual and potential competition, buyers and users
of the ratings increase their efforts to effect change from the monopolist
using voice rather than exit. An important channel for voice is the use
of governmental processes, the media, and industry associations.
The U.S. government has repeatedly declined to directly regulate the
audience measurement industry. Oversight broadly defined has occurred
in response to challenges to the industrys legitimacy, accuracy, and
fairness. These measures, oftentimes related, have taken the form of
Congressional hearings, public and private lawsuits, and industry self-
regulation. These nonmarket activities have prompted the exit of fraudu-
lent raters, some industry-initiated reforms, and for some complainants,
accommodations to better serve those complainants individual concerns.
Worldwide, one sees an assortment of actions that did not occur in the
U.S. context, including the creation of joint industry committees that set
terms and negotiate prices through competitive bidding, promulgation of
government-mandated minimum specifications regarding how the ratings
are produced, and production of ratings by industry consortia.
The next section provides background regarding the audience measure-
ment industry and highlights key episodes within its history before discuss-
ing the core economic features of the radio and television audience
measurement markets. That discussion provides the foundation for under-
standing the market strategy implications of these features considered in the
section Firm Strategies, Nonmarket Constraints, and Industry Dynamics
which also examines the legal constraints associated with actions employed
by firms in this industry. The use of voice is then illustrated with two
examples in the Threats of Governmental Intervention, Self-Regulation,
and Negotiation section. A study in contrast follows with an examination
of several non-U.S. and decidedly more interventionist government policies
in the section Alternative Models of Television Audience Measurement
Markets. Final remarks are offered in the conclusion.
MARKET STRUCTURES OF THE RADIO AND

TELEVISION RATINGS MARKETS
This section describes audience measurement for radio and television as a
product. It then provides an overview of the evolution of these markets
and highlights features that have contributed to their strong tendency
toward a monopoly structure.
Radio and Television Audience Ratings
Radio and television advertising is valued by the size and other characteris-
tics of the audience reached. Measurement of the audience in the form of
audience ratings is crucial to such media because it provides a currency
for buyers and sellers of advertising airtime, in addition to evaluating the
popularity of programs more generally. At a gross level, undertaking such
estimates is much more challenging compared with other media, for exam-
ple, print, for which circulation size or subscription numbers are objective
measures of reach, though getting a fine-grained idea of what is read,
watched, or listened to remains challenging in all cases.
The primary purchasers of these audience measurements are advertisers,

advertising agencies (which design and place advertisements for the adverti-
sers), and individual radio and television stations and networks that sell
airtime. Each of these potential buyers has a different willingness-to-pay
(WTP) for ratings information depending on which audiences are targeted
in the case of the buyers of advertising, and what audiences are reached in
the case of the sellers of advertising. Individual radio and television broad-
cast stations provide local coverage (e.g., Boston), while network affilia-
tions allow airtime to be sold that gives one-buy access to regional and
national audiences. Audiences are estimated for both local and national
markets. Different analytical categories such as by region are also some-
times analyzed.
An audience rating is developed by sampling the relevant target popula-
tion to learn what programs the households (or, preferably, its individuals)
are listening to or viewing, and then extrapolating from that sample to a
local or national rating. Beville (1988) describes the methodology for gener-
ating ratings:
1. The instrument the questionnaire, diary, or mechanical meter that records the
original viewing or listening data and the accompanying material and instructions
that determine how certain responses are recorded.
2. The data collection procedures personal interview, telephone, mail, meter with
cartridge or leased line, interactive cable; interviewer training, control, and supervi-
sion; verification procedures, etc.
3. The sample design universe, basic frame, randomness, sampling error (statistical
reliability), degree to which sample plan is achieved, nonresponse and its effects.
4. Operational and administrative procedures procedures used in processing data,
standards for handling of ambiguous and incomplete responses, weighting, projec-
tions, printing, report composition, quality control.
The fundamental features have remained constant over several decades,

though there have been great advances in technology for both the underlying
media and associated measurement devices. Every methodology has advan-
tages and disadvantages. For example, early devices that specifically recorded
how a radio was tuned did not capture who, if anyone, was actually listen-
ing. Moreover, while the ratings based on such devices were not cheap,
households without telephones could be included in the sample. Diary meth-
ods suffer from problems associated with faulty recall (Beville, 1988) while
telephone-based surveys were limited to individuals that had telephones.
Furthermore, the extent of the issues frequently differs across different
population segments.
Developing audience ratings is both an art and a science. The underlying

statistical foundation for sampling is relatively straightforward. In practice,
however, numerous confounding factors complicate that analysis and affect
the ratings quality.2 These problems are exacerbated by the cost generally
required to increase sample size or accuracy coupled with a frequent unwill-
ingness by buyers, especially those with limited budgets, to finance such
enhancements. The net result is that different rating methodologies fre-
quently produce different results.
Ultimately, the technical questions raised in audience measurement are
arguably superseded by the deceptively simple issue of what to measure.
[Television audience measurement technology] is, by no means, simply measuring
something out there. A complicated set of socio-technical conventions has to be agreed
upon, which can change according to the relations of the partners in the industry ,
the different stages in television history, the technology available and the viewer
habits and behaviours. The audience can never be considered independently of the
instruments used to measure it what is considered as television viewing? In
many ways, the answer to the question is tautological: television viewing according
to [a firm] is what [that firm] defines as television viewing at a given moment of history.
(Bourdon & Meadel, 2014, pp. 100101)
Overview of the Evolution of Audience Measurement Markets

in the United States
In 1920, the advent of a new industry, commercial radio broadcasting,

spawned the growth of the related industry for audience measurement.
In the intervening years, the audience measurement industry (AMI) has
evolved in fits and starts in response to not only radio, but also television,
and more recently the Internet. Despite experiencing such profound
changes, the industrys defining characteristics including its tendency
toward natural monopolies have remained constant.
Competition within the AMI context has been a competition to control
the ratings market itself, rather than a competition solely within the market.
As will be developed below, a combination of factors have resulted in
AMIs demonstrated structural tendency toward monopolization.
Notwithstanding that tendency, this industry has experienced spates of
actual or potential competition that are often fueled by advertisers, broad-
casters, or intermediaries (i.e., advertising agencies). Not surprisingly, local
(as opposed to national) ratings, whether in radio or in television, have
exhibited a somewhat greater tolerance for the persistence of multiple
ratings providers. Most of this competition has arisen either relatively early
in the industrys history or accompanied changes in either the underlying
media technology (radio, television, internet), its use (at home, mobile), or
the associated measurement technology or methodology (telephone, diary,
meter). During AMIs earlier years, such contests could result in a new mar-
ket leader emerging. Over the past 50 years, however, their outcome has
almost invariably been to further ensconce the existing monopolist.
At the risk of oversimplifying, the 1930s witnessed dueling radio ratings
whose competition reflected the perennial challenge associated with develop-
ing a ratings methodology that best combines accuracy and cost-effectiveness.
The initial standard based on the telephone recall method, developed by
Archibald Crossley and financed by the participants in the advertising indus-
try, would ultimately be supplanted by Hoopers telephone coincidental
method. Hoopers triumph was short-lived given the then ever-changing
media marketplace. Two increasingly important demands concerned a
national measurement of radio audiences and the measurement of televi-
sion audiences. Despite certain virtues, Hoopers telephone coincidental
methodology was limiting because it required phone service and, as such,
potentially overstated the habits of certain populations (city, television)
relative to others (rural populations, radio).
In the late 1930s, Nielsen acquired the patent rights for the audimeter.
This device automatically recorded how a radio was tuned during a given
period and as later developed would not require phone service or a visit
from a Nielsen representative. Nielsen readily adapted this technology to
measure television viewing and developed a formidable patent portfolio
regarding audience measurement devices more generally. In 1950, Nielsen
acquired Hoopers national radio and national TV rating services and, in
so doing, became the sole provider of national ratings.
Not surprisingly, as television became increasingly important so too did
competition within that sector. Just as Nielsen was prevailing over Hooper,
Nielsens next major competitor ARB was founded specifically to mea-
sure television audiences. ARBs initial advantage over Nielsen was not its
technology but, instead, its pioneering use of the decidedly non-technical
diary system which was a written journal detailing programming viewed,
and the introduction of an improved pricing model which focused on
broadcasters as the primary source of revenues. Nielsen would ultimately
adopt a diary approach in local markets because those markets could not
support the expense of its audimeter (Buzzard, 1990). ARBs attempt to
enter the national market with its own meter device (the Arbitron) was
hampered by Nielsens patent suit against the company (Seiler testimony
in Harris Hearings, 1963, p. 1699). With regard to ARBs national ratings

based on diaries, its participation in that market became no longer profit-
able after Nielsen offered aggressive price reductions.
Dissatisfaction with ratings quality and with some aggressive business
practices led to a number of industry and governmental initiatives which
culminated in congressional hearings in 1963 (the Harris Hearings) and
investigations by the Federal Trade Commission (FTC). The Harris
Hearings painted an unattractive picture of the industry and led to a shake-
out among existing providers as well as some industry self-regulation. In
the subsequent decade, Nielsen and Arbitron in television and radio rat-
ings, respectively, became the sole providers of the core national and local
audience measurement reports.3 Having established these monopoly posi-
tions, the two firms interests moved from competing for the market to
exploiting and maintaining their respective monopolies. Over the subse-
quent decades both have been able to defeat a handful of entry challenges
which typically involved the introduction of improved technology not cur-
rently deployed by the incumbents.
Economic Characteristics of the Radio and Television Ratings Markets
The tendency toward a monopoly market structure derives from four

important market characteristics: the presence of scale economies, the need
for comparability, the significance of information imperfections, and par-
tially conflicting buyer preferences.
Economies of Scale
Audience measurement involves a large up-front investment in choosing,
sampling, and analyzing the relevant market, even for firms that have tech-
nical expertise from related environments (Commission of the European
Communities, 2004, 2008). There are also fixed costs associated with
improving ones sampling methods and changing the sample panel. The
fixed costs associated with measurement systems involving recording
devices are much greater than those associated with interview-oriented sys-
tems because of R&D related to the underlying technology, the cost of the
technology and its deployment, and the maintenance of the system. Once
these fixed costs are incurred, the cost of delivering an additional report to
another user in the market is relatively small. There are, of course, variable
costs associated with selling the reports to potential buyers and with creat-
ing customized reports. However, because the core costs associated with
audience measurement can be spread across multiple buyers, raters with

high market shares can recoup their costs with lower prices than can firms
with low market shares.
The strategic importance of fixed costs depends heavily on potential
buyers WTP and the size of the potential markets. Small markets with low
WTP potential buyers (e.g., secondary city radio markets) may be unable
to support more than one rating service, if that. Such markets have particu-
larly strong natural monopoly tendencies.4 If fixed costs are not so great
relative to the revenue size of the market, then scale economy considera-
tions alone would not preclude the possibility of multiple competitors.
Value of Comparability
In markets with multiple services, ratings users must compare potentially
disparate results because raters typically employ different methodologies,
samples, protocols, and even definitions of the local market. This compari-
son problem is compounded because those characteristics are continually
evolving and have frequently been largely proprietary.
From a user/buyer perspective, the value of some standardization across
ratings can be thought of in terms of demand side increasing returns related
to comparability (or compatibility) issues (Saloner, Shepard, & Podolny,
2001). Standardization reduces the negotiation costs associated with agree-
ments concerning advertising or sponsorship prices and eases assessment of
performance trends of a program or a station. The value of comparability
or standardization is described in terms of the advantages of having a
single currency for transactions.
Unless the buyers of the ratings can be persuaded that one rating has a
significant quality advantage over another or a particular advantage to
them, they will resist switching rating systems, especially if relative ratings
are more important than absolute ratings. Unfortunately, as will be consid-
ered next, assessing the quality of ratings is very difficult, which increases the
suspicions of advertisers that observed switches by stations or networks may
have been made in response to (potentially) unjustified increases in ratings.
Information Imperfections
During the 1963 Harris Hearings, numerous witnesses voiced skepticism
regarding the quality of radio and television ratings. This skepticism per-
sists today.5 Unlike goods or services where quality is observed before or
during use, ratings users can only assess quality through indirect means.
Accuracy, for example, might be inferred through comparisons of ratings
provided by different firms at the same time or across ratings of a single
firm over time. It can also be assessed by examining the methods, samples,
and procedures used by the raters or through independent statisti-
cal studies.6
Ratings have frequently differed, sometimes by significant amounts.
When there has been competition among raters, for example, the historical
record is replete with divergence among the ratings. For example, one radio
station owner testified in the 1963 Harris Hearings that Hooper, Pulse, and
Nielsen all well-established rating firms reported audience shares for
the same time and station that were 42%, 29%, and 5%, respectively
(McClendon testimony in Harris Hearings, 1963, p. 278), and that such dif-
ferences were not uncommon.7
In the more recent monopoly era, comparison evidence is largely based
on method or device changes by the incumbent. After Nielsen switched to
its local people meter technology in the United States around 2004, some
local television market ratings experienced dramatic shifts away from
broadcast channels and toward cable channels.8 Arbitron also experienced
significant changes in ratings when it shifted to its portable people meter
system (Napoli, 2014).9
Ratings reports buyers have also learned about the quality of ratings and
potential improvements through their own research. Typically, such research
efforts were conducted under the auspices of one or more industry or profes-
sional groups. See Beville (1988) for a detailed discussion of these studies.
Bourdon and Meadel (2014) note one problem with organizing and conduct-
ing industry-funded studies is that there are conflicts of interest across the
various funders of these research efforts.10
On the whole, observers now seem to believe that the skepticism about
ratings quality is a reflection of the difficulty of the problem and disagree-
ments about the appropriate audience to be measured and not some sys-
tematic problem with the raters themselves. Beville (1988, p. 307) sums this
up by stating [o]ver the years the methods and the accuracy of ratings sys-
tems have frequently been challenged. However, the integrity, objectivity,
and fairness of the people delivering the audience numbers have rarely been
questioned.
Partially Conflicting Buyer Preferences

In most markets, buyer preferences are unlinked in the sense that one
buyers purchase does not affect another buyers utility. The primary excep-
tions to that independence are markets that involve externalities, such as
positive network externalities (e.g., platform markets) or negative conges-
tion externalities. Relative rankings and measurement markets such as the
audience measurement market have the feature that the comparative por-
tion of the measurement has a zero-sum quality: a viewer in a particular
time slot either views one broadcasters program or a different broadcas-
ters program. The importance of preference conflict for market dynamics
depends on which users are the primary buyers of the product or service. In
the U.S. audience measurement markets the bulk of the revenues were gen-
erated from the broadcasters even though advertisers, advertising agencies,
and broadcasters all depend on ratings (Boyer, 1988). This situation affects
rating firms incentives to make changes to its system based on accuracy
alone because the broadcasters have a joint preference for high ratings for
their media relative to other media (e.g., radio versus print) and a strong
self-interest within the media class for high relative ratings.
Summary
In this section, we examined how economies of scale, value of comparabil-

ity via a common currency, information imperfections, and conflicting
buyer preferences create a natural monopoly market in equilibrium. Both
economies of scale and a large value given to a common currency could
individually push a given market toward a tight oligopoly or even a mono-
poly. Together the two factors make a monopoly market outcome even
more likely. Information imperfection in the form of uncertainty about
quality has the effect of increasing the relative weight that buyers will place
on a common currency, thereby again pushing toward a monopoly out-
come. It, like conflicting buyer preferences, also makes displacing entry less
likely. This combination of characteristics is somewhat unusual and
explains why we observe not only a single rating firm in each market in the
United States, but one whose identity has remained the same.
FIRM STRATEGIES, NONMARKET CONSTRAINTS,

AND INDUSTRY DYNAMICS
Both the economics and history of the audience measurement industry
reveal a very strong tendency toward natural monopoly. As such, evolution
within the ratings markets naturally divides into two phases. The first
entails competition for the market and the second, after the emergence of a
monopoly firm, entails the exploitation and maintenance of that monopoly.
The relevant business strategies and tactics employed will vary depending
upon the industry phase because activities of monopolists, or even firms
with market power, may be subject to nonmarket forces. This section
examines how firms navigate these forces and how industry direction is
shaped particularly when either substantial legal risk is involved or the gov-
ernment or interest groups initiate nonmarket actions.
Competition for the Market
The goal of firms during the competition phase is to win the market. This
phase entails oftentimes brutal competition for market share in which firms
invest in market share through pricing that is aggressive relative to the qual-
ity provided. This competition takes place within the context of conventional
strategy decisions including decisions regarding firm positioning and scope,
capability development, and alliances and acquisitions. Some of these decisions
may involve actions that involve legal risk which the firms balance against
potential gains. The victors goal then becomes maintaining its monopoly.
Positioning
Ratings firms position themselves to offer an attractive value proposition
to buyers and to achieve a sustainable competitive advantage for them-
selves. Elements of that decision include choices regarding product and
geographic markets, customer base, and the activities through which the
firm creates value. The firms would decide, for example, which media to
measure (e.g., radio, television) and whether to focus on local and/or
national markets. Another key decision concerns measurement methodol-
ogy. For example, some leading audience measurement firms focused on
audience measurement in local radio and television markets using a diary
recall method or telephone interviews, while for national television markets
Nielsen relied primarily on passive monitoring of home television set use.
Because audience measurement is multifaceted, firms may also make deci-
sions about customization and whether their service is positioned as provid-
ing core market figures or as a complement to those estimates.
Relative to passive monitoring through a device, diaries had the advan-
tage of lower capital requirements for entry coupled with the disadvantage
of easier imitation and, hence, less distinctiveness. Further, the actual
device used for passive monitoring was amenable to intellectual property
protection. Firms choosing an easily imitated methodology could still
obtain a sustainable advantage based on scale economies and the
importance to buyers of a single ratings currency though this strategy

depended on gaining an early market share advantage. A technology-
oriented approach, on the other hand, would still reap the advantage from
an early market share advantage while being less susceptible to imitation.
The problem with a technology strategy was its larger initial cost and possi-
bly a more complex start up.
Another important aspect of positioning concerns the choice of ratings
quality. This choice is complicated in markets in which information
imperfections that interfere with a buyers ability to assess quality exist.
Assuming that buyers desire higher quality assessments, the ambiguity char-
acterizing such assessments should encourage sellers to undertake strategies
designed to develop a positive quality reputation. Such investments create
value by partially solving the users information problem and allow the
investing firm to capture value because the reputation is firm-specific
(Oberholzer-Gee & Yao, 2013). One important consequence is that rivals
will find it more difficult to gain headway even with an allegedly superior
offering. A positive reputation may result from signals of actual quality and
marketing and customer-facing activities directly demonstrating quality,
including claims regarding superiority of method and widespread usage.
Buyers difficulties in assessing audience measurement quality make it
difficult for even a superior service to demonstrate its superiority. This
dynamic arguably contributes to a general downgrading of the importance
of quality to buyers relative to other rating characteristics such as the
ability to offer a standardized currency. Then, one buyers decisions are
increasingly influenced by the others decisions which creates a herding
effect where the ratings are accepted because other users have accepted
them. Alternatively, stations and networks may find it natural to make
credible arguments supporting measurement approaches that favor their
own self interests. When quality is difficult to assess, more weight may also
be placed on pricing.
Nielsens actions in the ratings markets are instructive. During the compe-
tition for the market phase, Nielsen benefited from its existing excellent repu-
tation as a market research firm, based its method around the Audimeter, a
patented and distinct technology that its rivals lacked, and invested heavily in
R&D and its patent portfolio. Its name became synonymous with television
ratings in the United States, and it became the dominant firm in the television
audience measurement market in the United States.
The reputation benefit that Nielsen received from its high standing in
the market research business is an example of leveraging a corporate
resource across businesses (Collis & Montgomery, 2005). Such positive
reputation externalities could, however, morph into unwelcome tainting:

poor quality in one product line has negative externalities on other product
lines. For example, if radio audience measurement is inherently more diffi-
cult than television audience measurement, having a radio ratings business
could affect perceptions of the quality of ones television ratings.11 Nielsen
described the role of such considerations in his decision to exit the radio
business in 1964.
[A.C. Nielsen] has recently been lambasted unmercifully by broadcasters, and others.
[A] rather substantial portion of the criticism was predicated on the limitations (actual
and/or alleged) in the techniques we are using for radio, both network and local; since
we were losing money on these services and saw no prospect of ever eliminating the
losses, I felt that we could not justify to our stockholders a course which would have
continued to subject us to criticism which tends to damage our professional reputa-
tion . (Beville, 1988, p. 38)
Nielsens technology-oriented approach differentiated it from most other

competitors. Nielsen sought to use this approach as tangible input evidence
of its superior quality of service. In addition, Nielsen built a patent portfo-
lio around its audience measurement technologies that allowed Nielsen to
sue rivals who adopted similar technologies, thereby raising its rivals costs
and slowing entry or deployment of their systems.
Finally, as will be discussed in detail in a later section, concerns about
quality of audience measurement led to a number of industry and govern-
mental initiatives which attempted to address, in part, the information
imperfections problem.
Pricing and Deep Pockets

Nielsens exit from radio ratings not only highlights reputational concerns
regarding quality, but also provides insight into the nature of the competi-
tion to win the local and national radio audience measurement markets at
that time. During the 1950s and 1960s, much of the competition for the
ultimate control of these markets in the United States was being fought.
While rivals in most markets always seem quick to complain about bloody
below-cost price wars, both the historical record and the market economics
suggest extremely fierce competition accurately characterized the audience
measurement industry during this period.
In theory, a superior competitor should be able to attract capital during
a war of attrition for a market that allows it to prevail against inferior com-
petitors. But in audience measurement markets, the quality assessment pro-
blem undermines the efficiency of the capital markets and their ability to
finance a superior competitor. Hence, a firms willingness and ability to
sustain losses can augment the value to tactics designed to drain competitor
resources or even play a pivotal role in determining the outcomes of a battle
to win a market. Nielsen with its market research business had an advantage
in financing its competition for the market as it could internally access
capital needed to sustain short-term losses associated with the television
ratings market during its leaner, earlier years (Buzzard, 1990, pp. 2627).
Acquisitions and Alliances

Through consolidation a firm can access general resources and competen-
cies that it lacks and use its newly expanded resource set to create addi-
tional value. Many firms, for example, adroitly secure technology through
acquisition and licensing rather than internal development. Others engineer
acquisitions or alliances to quickly enter new product or geographic
markets. Mergers, acquisitions, and strategic alliances can also be used to
take advantage of scale economies (Doz & Hamel, 1998).
In emerging markets such as television ratings in the mid-20th century,
rapidly changing market conditions, new technologies, and lack of capital
combined to make mergers, acquisitions, and strategic alliances common-
place. Given intense competition and significant fixed costs, markets could
rarely support two or three profitable firms.12 Acquisition targets some-
times involved former competitors, turned failing firms. An early example
of this approach was Nielsens 1950 acquisition of Hoopers national radio
and television audience measurement assets. This acquisition gave Nielsen
close to monopoly positions in the national radio and television audience
measurement markets as Hooper was Nielsens primary competitor in those
markets. More recently, there have been a number of mergers, joint ven-
tures, and acquisitions in this industry (e.g., 2013 merger between Nielsen
and Arbitron) that were presented as combinations to enhance efficiency as
the entities involved were not failing firms.
While consolidations potential benefits include increased economies of
scale and a more widespread single currency, harms may also accompany
it. Consolidation may create market power, increasing price to buyers, or
slowing the development and deployment of innovation. Hence, most
developed countries have competition laws enabling the government to
prohibit or dissolve unlawful proposed or consummated mergers, respec-
tively. In the United States, for example, the Clayton Act designates unlaw-
ful combinations as those that may substantially lessen competition.13
While the benefits of acquisitions in the form of efficiencies are a part of
the antitrust merger analysis, the obstacles to recognition of such efficien-
cies in merger law are quite substantial.
Nielsens 1950 acquisition of its principal competitor, Hooper, Inc.,

would seem to have courted considerable antitrust risk. In this case,
Nielsen was either lucky or prescient. It wasnt until 1963 long after the
acquisition had shaped the industry that the acquisition was found to be
illegal by the FTC. In the complaint accompanying the order, the FTC
alleged a number of offending actions which included acquisition of the
customers and trade names from Hooper along with an agreement that
Hooper would not reenter the market for a substantial length of time. The
resulting FTC consent order imposed certain obligations including among
other things prior approval by the FTC for any acquisition of business,
physical assets, or goodwill in the ratings industry (FTC, 1963).
Contemporaneous press accounts indicated that the FTC was even consid-
ering ordering a divestiture of 50% of Nielsens national clients to try to
redress the competitive harm brought on by Nielsens action, but such dra-
conian actions were never undertaken. It was also reported that Nielsen
threatened to exit the national television ratings market if such a divestiture
were ordered.14
More recently, a number of mergers, acquisitions, and joint ventures in
this industry have attracted antitrust attention. In 2008, for example, the
European Commission allowed WPP, a 50% owner of AGB Nielsen, the
leading television audience measurement firm in Europe to acquire TNS,
which had about a 2030% overall share in those markets, on the condi-
tion that one of the two television audience measurement services would be
divested (Commission of the European Communities, 2008). Then, in 2013,
the United States permitted, with a required sale/license of some audience
measurement assets, a merger between Arbitron, who had a monopoly on
local radio audience measurement, and Nielsen, who had a monopoly in
television audience measurement (U.S. Federal Trade Commission [FTC],
2014). Although Arbitron and Nielsen were once competitors in their
respective markets, the FTC apparently did not see them as potential future
rivals.
Exploiting Monopoly
Many aspects characterizing competition for the market reappear in the

same or similar form as strategies to exploit or maintain a monopoly.
Firms with considerable market power, however, have additional opportu-
nities to capture value from buyers and suppliers because of their relatively
strong bargaining position (Porter, 1980). These firms may also offer
greater efficiencies in the buyerseller relationships and transactions

through economies associated with scale or through joint investments that
reduce costs or increase quality. With or without the creation of efficien-
cies, most firms find some exercise of market power to capture value from
suppliers and buyers irresistible. In some industries this exercise of market
power is either partially short-circuited through direct regulation or
restrained through the exercise of antitrust laws which depends on the
nature of the industry and the conduct of the firm involved.
Firms in a monopoly position with a weak threat of entry, which argu-
ably characterizes Nielsen and Arbitron in their respective U.S. radio and
television ratings markets for the last several decades up until the internet
age, are able to set monopoly prices.15 Antitrust law and policy has
wrestled with what to do with inefficiently high pricing. However, antitrust
intervention to lower prices is seen as problematic because (1) the mono-
poly may have been achieved through superior performance (e.g., efficien-
cies greater than that of rivals or from the firms innovative investments)
and (2) the remedy to high pricing is to regulate price or to facilitate entry.
As a result of these difficulties, antitrust law in the United States does not
interfere with monopolies that are obtained through superior performance.
The focus of the law is, instead, on preventing the leveraging of market
power to achieve or maintain monopoly positions.16 As such, the antitrust
laws in the United States provide no recourse against actual monopoly
pricing. If the government wishes to regulate pricing, it must pass legisla-
tion or introduce rules with that specific intent. As discussed in the next
section, the U.S. Congresss interest in this market did not seriously include
governmental regulation of the industry.
In radio and television ratings markets charges of noncompetitive
pricing by firms with market power were commonplace. The 1963 FTC
complaint (U.S. FTC, 1963), for example, stated that Nielsen was engaged
in noncompetitive pricing.17 Pauley, President of ABC Radio Network,
recounted in the 1963 Harris Hearings that ABC radios subscription to the
NRI (Nielsen Radio Index) was canceled when it refused to purchase an
additional set of ratings which was bundled with the previous service
(Harris Hearings, 1963, p. 231), and the Radio Association of Broadcasters
(Radio Advertising Bureau [RAB], 1982) claimed that Arbitron charged
for its ratings based on a radio stations advertising rates.
Joint activities among competitors, particularly those which are poten-
tially market-power enhancing, invite antitrust scrutiny. As such the Radio
Advertising Bureau (RAB) sought a business review letter from the U.S.
Department of Justice Antitrust Division (DOJ) regarding its proposal to
establish a buyer-side consortium to offset seller power in the radio audi-

ence measurement market. It proposed a joint negotiating group in which
an RAB Committee would negotiate a contract with Arbitron which would
be nonbinding on their members. The RAB proposal was a comparatively
weak form of a joint buying group. Because joint buying groups pose anti-
trust risk through the artificial creation of buyer market power, RAB asked
the U.S. Department of Justice Antitrust Division to review whether the
proposal would pass antitrust law muster. The DOJ effectively denied the
request when it replied it is unable to state a present intention not to chal-
lenge the RABs proposed joint negotiating group because DOJ was not
convinced that Arbitron had a natural monopoly given services including
the Birch Report, a firm that DOJ thought might continue to coexist with
Arbitron going forward, and because the buyer cartel could have the poten-
tial of generating its own anticompetitive effects (U.S. DOJ Response
Letter to RAB, 1984).
The U.S. antitrust agencies general reluctance to allow the creation of
countervailing buyer power precluded a buyer-based solution to the pro-
blem posed by natural monopoly in the audience measurement markets. In
contrast, as we will discuss later, most other countries have evolved varia-
tions on a buyer consortium model that allows the buyers (or a national
agency) to put out audience measurement for competitive bid.
Maintaining Monopoly
A primary interest of this paper is the exploration of actions a monopolist

uses to maintain its market position. Such actions are largely directed
toward suppressing or undermining potential entry which is attracted by
the profit potential of gaining a monopoly. When there are different ways
to satisfy market demand, an entrant chooses whether to enter as an
imitator or to enter using a different approach (i.e., choosing a differ-
ent position).
The difficulties associated with entry in audience measurement markets
were summarized in a 1982 business review letter from a joint committee of
radio stations to the Antitrust Division of the DOJ.
[a]ny new service must overcome the tendency for the existing service to be viewed as a
norm especially by advertising agencies against which to gauge the credibility of
the new service. If the new services data are different, it may be seen as undermeasuring
or overmeasuring radio listening as compared to the existing service, which will impair
the new services credibility. Therefore to displace the present service, a new entrant
must establish credibility, offer a significantly better product, educate time buyers and
sales reps in becoming familiar with using its format, and provide complete market and
nationwide coverage. A minimum of three to five years of sustained effort would be
required . (RAB, 1982, p. 10)
As just discussed, while DOJ may have been sympathetic to the entry
problem, it nonetheless believed viable entry was possible and the risks
of buyer coordination regarding price posed an anticompetitive problem of
its own. This position contrasted with that of the European Commission
which, in discussing European ratings markets, concluded that [t]he
most likely route to replace an incumbent supplier is either through the
formation of a [joint buying group] or by initiative of a large national TV
station (Commission of the European Communities, 2004, p. 7). As just
discussed, neither approach has emerged in the United States, though the
approaches are common elsewhere.
Given their market power, it is not surprising that incumbents have
adopted a number of business practices which increase the difficulty of
entry by foreclosing buyers from the entrant or by increasing buyer switch-
ing costs. In this subsection, we illustrate the general monopolist strategies
with examples most of which relate to Nielsen.
Business Practices to Limit Available Customers or Increase Switching Costs

In general, a monopolist may leverage its market power by forcing its
buyers (and suppliers) to accede to a number of business practices that
increase buyer switching costs. For example, the monopolist may force its
buyers to sign contracts that directly or indirectly impact the attractiveness
of purchasing from its rivals (e.g., long-term contracts with large early ter-
mination fees, pricing that is tied to sales of other products, pricing that
involves volume purchase discounts); products may be designed to create
intentional incompatibilities with competitor products; monopolists may
refuse to provide rivals access to essential facilities; may hold informa-
tion valuable for making comparison with entrant products secret, and so
on. Individually, each of these actions may seem relatively small, but the
sum of such practices can significantly increase the barriers to entry both in
terms of direct switching costs and the speed of entry.
Within the audience measurement context, a number of such practices
are commonplace. Nielsen, for example, has been accused of attempting to
maintain its monopoly through acquiring its competitors, securing multi-
year staggered contracts with ABC, CBS, NBC, and FOX Broadcasting,
imposing steep financial penalties on customers who refuse to enter into
long-term agreements, and engaging in predatory pricing against a
potential entrant, erinMedia, in 2005 (erinMedia, 2005). Staggered con-

tracts reduce the number of customers that would not have to pay termina-
tion fees in a given year. Similar accusations were raised in 2011 by
Sunbeam, a local television broadcast station owner in Miami-Ft.
Lauderdale, regarding the anticompetitive aspects of staggered contracts
and the requirement that subscribers return historical ratings data when
their subscription ended (Sunbeam Television Corp., 2011). Nielsen denied
these allegations and pointed to legitimate business reasons for their
actions. None of these cases went to a full trial on the merits.
Other complaints by potential competitors and disfavored customers
involved ratings incumbents favoring key customer groups to discourage
their defection to entrants. Sometimes these allegations were direct
(Sunbeam Television Corp., 2011) and other times the allegations came
from disfavored customers who claim to be unfairly treated because of the
impacts that proper treatment would have on the customers larger rivals
(Woodlinger testimony in Harris Hearings, 1963 p. 207; FAIR Hearings,
2005). Comparing ratings systems, let alone discerning rating bias, is diffi-
cult given the different methodologies employed. For example, compari-
sons become quite difficult if the rating systems employ different
geographic and demographic contours as part of their methodologies.
Antitrust law not only addresses mergers that create additional market
power, it also constrains the conduct of parties that possess such power
when that conduct is used to create an unfair market. In practice, how-
ever, exclusionary conduct which has a non-pretextual business justification
and which does not have a less restrictive alternative is less likely to be
condemned by antitrust law under a rule of reason analysis (Greene,
2015). Thus, for example, a contract provision requiring the return of the
previous ratings data provided by the incumbent (upon switching to
an entrant) could be justified as protecting proprietary information
from rivals.
Innovation Competition and Displacement of the Incumbent

In dynamic markets, monopolists are at greater risk of being supplanted by
entrants who are better positioned to meet the needs of the market or
entrants with (arguably) superior technology. An entrant has an opportu-
nity to adopt a superior position because the needs of the market have
shifted substantially from the position of the incumbent, perhaps because
of preference shifts. The entrant might also possess superior technology to
meet the needs of the market. An entrant with a new technology attempting
to dislodge an entrenched monopolist faces several challenges. First, new
technologies are unlikely to dominate mature technologies across the entire

buyer population. Hence, the market factors contributing to the existing
rating monopoly present additional challenges for entrants. In a conven-
tional oligopoly market, a new technology deemed to be superior by a
quarter of the buyers would likely be sufficient to support successful entry.
However, in a monopoly market attracting a much greater market share
may be necessary to support entry. Second, as discussed earlier, the super-
iority of the new technology is often difficult to ascertain, particularly
before or at initial introduction. Finally, even if the new technology is
superior, questions may arise regarding whether others, especially the
incumbent, may imitate that technology, or even whether the entrants
development and use of the technology infringes others intellectual prop-
erty rights.
An important factor with respect to imitation is its speed. For example,
if the new entrant exercises sufficient control over the intellectual property
rights regarding both a valuable new technology and its close substitutes,
so as to exclude the incumbent, then displacement of the incumbent is pos-
sible and, perhaps, to be expected. If, however, imitation can be rapid, then
buyers may prefer, given the other advantages of the incumbent, to wait
for the incumbent to catch up. Moreover, if the new technology will be
used in conjunction with pre-existing assets then the incumbents advantage
over an entrant would allow them to more easily undercut entrant pricing.
When technological change necessitates the replacement of existing hard-
ware, however, an important part of the incumbents cost advantage will
be nullified, giving the entrant with a superior technology a reasonable pro-
spect of displacing the incumbent even without strong intellectual property
protection.18
The strategic use of patent litigation depends not only on the underlying
merits of the patent dispute, but also on the litigants relative positions in
terms of their abilities to bear the uncertainties and costs of litigation. The
property rights issue involves more than just having an incumbent greet
device-based entry with patent infringement lawsuits. Patents grant exclu-
sivity over the claimed inventions. But the value of such grants may be
ambiguous because the scope of patent coverage and sometimes even the
validity of the patent itself, is frequently unclear. Furthermore, outcomes
from such disputes run the gamut from injunctions against use, to licensing
or damages settlements, to invalidating the patent itself. Thus, patents are
more than just licenses to sue in a narrow sense, they are also opportu-
nities to leverage the uncertainty, time delay, and costs of high-stakes legal
proceedings into strategic weapons (Antitrust-Intellectual Property Report,
2007). Uncertainty regarding the outcome of a patent infringement suit

usually has a greater impact on potentially infringing firms, especially when
they are in need of additional infusions of support either in the form of capi-
tal or the development of complementary products. Hence, the stronger
firm may gain additional advantage from prolonging the litigation, which
has the effect of both depleting the alleged infringers resources directly and
undermining its ability to attract additional resources.
As previously discussed, buyers in the audience measurement markets
heavily value a common currency of ratings and have heterogeneous and
often conflicting interests concerning rating systems which are likely to
favor one broadcaster at the expense of another.19 Furthermore, it is diffi-
cult to directly determine the quality of a rating system. These factors sug-
gest that entry into audience measurement appears considerably more
difficult than entry into most, even monopoly, markets: need for a common
currency, quality uncertainty and conflicting interests make it difficult for
an entrant or a group of buyers to coordinate a sufficiently large defection
from the incumbent. Furthermore, these difficulties are likely to slow entry,
creating a window in which the incumbent could imitate or otherwise catch
up. The incumbent also has market knowledge and existing relationships
that would speed implementation of a fast-follower system. As an example
of these market dynamics, in 1985 Audits of Great Britain (AGB)
attempted to bring to the United States a variant of its advanced people
meter technology which it had deployed elsewhere. This technology allowed
demographic data to be combined with program viewing information;
something Nielsen could not then do. AGB apparently misjudged the U.S.
market and was slow to get its system adapted. While AGB was market
testing its system, Nielsen quickly developed a similar system and then per-
suaded the buyers to remain with Nielsen. AGB exited the U.S. market in
1988 (Buzzard, 1990).20
Nielsens actions to defend its monopoly position provide interesting
insights into the challenges entrants face regarding intellectual property.
Because Nielsens position in audience measurement depended on protecting
their proprietary technology, Nielsen had long invested in building a strong
patent position. Nielsen allegedly attempted to undermine entry by
Sindlinger in 1952 by asserting its patent rights and using the uncertainty
regarding Sindlingers patent position as an argument to dissuade invest-
ment in Sindlingers firm and as arguments to get Sindlingers current
investors to force a merger between Sindlingers firm and Nielsen. Lack of
investor support appeared to have convinced Sindlinger to settle with
Nielsen and the group never formed (Harris Hearings, 1963, pp. 1593,
1597, 1603, 1675, 1678, 1688). A patent infringement suit brought by

Nielsen also arguably deflected ARB from competing head-to-head with
Nielsen in network television ratings with its Arbitron instantaneous audi-
ence measurement system in the late 1950s.21 More recently, to deal with
measurement of cross-platform viewing, Nielsen has used acquisition to
build its intellectual property portfolio and continues to use litigation to
slow down its competitors.22,23
While the outcome of patent disputes may be uncertain, litigation or licen-
sing undertaken to protect and/or monetize patents does not itself present
unusual legal risk to the patent holder. Such risk does adhere, however, if
the patent was fraudulently acquired or was leveraged to extend beyond its
rightful metes and bounds. The FTC found Nielsen guilty of such miscon-
duct leading to a 1963 consent decree. Two of Nielsens business practices
that violated the FTC Act were (1) the systematic use of patent proceedings
to discourage potential and actual competitors from developing and
using devices for measuring audiences, and (2) Disparaging and
hindering competitors efforts to develop devices and has attempted to
impede and sabotage the financing of these competitive efforts (U.S. FTC,
1963). The subsequent consent order required Nielsen to make its audience
measurement patent portfolio available nonexclusively and royalty-free for
four years and nonexclusively and for reasonable royalties after that.
Summary
This section examined the dynamics of audience measurement markets in

which competition for the market eventually gave way to exploitation/
maintenance of the monopoly by the winner of the initial competition. In
the competition for the market phase, pricing competition is fierce given
competitors willingness to absorb losses as an investment in winning the
monopoly prize. First mover advantages associated with a large early mar-
ket share exacerbate the intensity of the competition, and the significance
of information imperfections affecting the external capital market gives
advantages to firms who can rely on internal capital to withstand initial
losses. These are not unusual features of competition advantages stemming
from a dominant position which occur in a wide range of markets exhibit-
ing large production or network economies. Attempts to establish market
share advantages through acquisition are a common strategy also
common across industries and are now a common concern of antitrust
authorities.
When a natural monopoly is regulated by the government (or the creation

of countervailing buyer power), monopolists undertake business practices to
increase barriers to entry. These business practices increase the costs of entry
by increasing buyer switching costs or by increasing the cost of entry
through patent litigation and the like. Such business practices may be
imposed, in part, because of the market-power leverage the monopolist has
over its buyers. This is a common feature of these markets but one which
conservative use of antitrust law will find difficult to correct given the osten-
sible legitimate business justifications that exist for actions that have the
salutary effect of raising switching costs. Again, the actions and practices of
the firms in the audience measurement industries are not unusual for firms
that possess market power. The primary difference is that the characteristics
of these markets lead to unusually high barriers to entry.
THREATS OF GOVERNMENTAL INTERVENTION,

SELF-REGULATION, AND NEGOTIATION
The structure of the radio and television ratings markets has allowed the
incumbent monopolist in their respective U.S. markets to resist displace-
ment. Here, the interests of the buyers and that of society more generally
are arguably only weakly promoted through the usual forces of market
competition. As a result, action outside the market the use of voice
rather than exit increased in relative importance. This section examines
two Congressional investigations, roughly 50 years apart, of the U.S.
ratings industry. The first investigation concerns hearings undertaken in
1963 entitled, The Methodology, Accuracy, and Use of Ratings in
Broadcasting, which are commonly known as the Harris Hearings for
the Committee Chairman Representative Oren Harris. Those hearings built
upon a 1961 Report, Evaluation of Statistical Methods in Obtaining
Broadcast Ratings, that Chairman Harris had previously requested of the
American Statistical Association. The second investigation concerns a ser-
ies of hearings during the first decade of the twenty-first century. While
these hearings addressed the perennial challenges plaguing ratings accu-
racy, they were specifically tailored to address questions of accuracy with
regard to minorities as well as other populations such as rural communities.
For example, Senate hearings were conducted regarding the proposed
FAIR Ratings Act (Fairness, Accuracy, Inclusivity, and Responsiveness
in Ratings Act of 2005) and in 2009 the House of Representatives
conducted hearings entitled, Will Arbitrons Personal People Meter

Silence Minority Owned Radio Stations?
While each investigation carried at least a potential threat of government
regulation, it would not be forthcoming. Each Congressional hearing pro-
vided a platform to give voice to the interests, including powerful (televi-
sion networks) and less powerful (minority stations) consumers of ratings
and public interest groups, who felt underserved, if not ignored, by increas-
ingly monopolistic ratings firms.
The 1963 Harris Hearings
Despite increased consolidation in certain industry sectors such as

national television and national radio ratings, the 1950s witnessed
ongoing competition particularly at the local level as well as an ultimately
abortive effort to introduce meaningful competition in the national televi-
sion ratings market. Increasing attention naturally focused upon the
ratings industry as its economic significance grew. For example, when
Nielsen acquired Hooper in 1950 there were six million televisions in the
United States. By 1960, the number had grown to more than 60 million.
The economic significance of audience measurement grew and with that
so did its role in influencing the development or selection of programming
content. In the late 1950s, a scandal surrounded the news that the
outcomes of several enormously popular television quiz shows were
rigged. Even this would be laid, in large part, at the feet of the ratings
industry as politicians claimed the rigged outcomes were undertaken to
increase television viewing ratings.
In the late 1950s, FCC and Congressional investigations into the rat-
ings industry prompted a technical examination (1961 Madow Report) of
the approaches used to measure radio and television audiences and an
investigation by the FTC regarding whether the claims made by the measure-
ment firms were deceptive. Further Congressional hearings (Harris Hearings)
in 1963 built an evidentiary case about the need for reform. The Harris
Hearings, as well as a further FTC investigation into the business practices,
focused much attention on Nielsen.
The Madow Report, prepared under the auspices of the American
Statistical Association, found that the rating services are, on the whole,
doing a reasonably good technical piece of work for the purposes to be
served but went on to suggest that a survey is an economic product in
which [d]ifferent purchasers will find different products adequate for their
uses (Madow, Hyman, Jessen, Sheatsley, & Wright, 1961, p. 12). The
report did raise a wide range of concerns about the quality of the ratings
reports and the ways in which users understand and use the ratings, given
the methodologies inherent limitations. The picture painted is one of
inevitable compromises that will affect some results and a lack of under-
standing stemming from insufficient research on methods, audience compo-
sition, field interview and data problems, and so on (Madow et al., 1961,
pp. 1925). While silent with respect to a summary statistic opinion on the
quality of the ratings, the report ventured that small market areas may
have samples too small or too poorly planned for the uses made of them
(Madow et al., 1961, p. 28).
The 1963 Harris Hearings, which involved a substantial investigative
phase, presented an in-depth and unflattering view of the audience mea-
surement industry.
The disturbing pictures that came into greater focus during the hearings
were (1) an industry whose product, even under the best of circumstances,
was fraught with oftentimes inescapable ambiguity or uncertainty, (2) firms
(Videodex and Conlan) who did little of the field research that they claimed
to have done, and (3) a monopolist that was insensitive to questions of
accuracy because of its size and undertook legally questionable activities to
maintain its market position.
During the hearings, the audience measurement industry was assailed
from every direction. Ratings systems were criticized for failing to keep
pace with changes in technology (e.g., the shift from fixed in-home to
mobile listening, the growth of FM radio). Ratings methodologies were cri-
ticized (e.g., the definition of geographic markets and other sampling tech-
niques). Minority groups claimed they were underrepresented in the
ratings. Evidence emerged that broadcasters indirectly gamed (e.g., special
promotions during ratings periods) or directly manipulated (e.g., contact
with participants in the ratings sample whose identities should be confiden-
tial) the ratings.
The purchasers of these ratings, who played a major role in financing
the industry, were highly critical of the ratings themselves while acknowled-
ging their dependence on the oftentimes unresponsive producers of ratings.
The network television broadcasters, for example, criticized Nielsens unre-
sponsiveness to quality concerns as well as to information requests by
which broadcasters sought to assess quality. Radio broadcasters thought
that their audiences were severely underestimated relative to television
viewing in part because mobile listening was underrepresented. This pano-
ply of concerns reflects, in part, frustration with the inherent difficulty
of measuring audiences. But the inherent difficulties do not excuse over-

claiming and lack of transparency that characterized many of the rating
services at the time.
And, of course, there were a multitude of complaints about price.
Nielsens business practices were singled out for special attention.
A. C. Sindlinger, a very disgruntled former competitor of Nielsen,
explained how his business was undermined through what he describes as
back channel interference by Nielsen intended to undermine the confidence
of Sindlingers investors (Sindlinger testimony in Harris Hearings, 1963,
p. 1675). The most telling part of Sindlingers testimony, however, were his
claims regarding a Nielsen master plan to dominate the ratings industry
(Sindlinger testimony in Harris Hearings, 1963, pp. 16761677). Sindlinger
provided to the Harris committee a copy of a 1951 subpoena to Nielsen
asking, among a large set of other documents, for the Nielsen master plan,
which included:
A. The development of a patent program designed to block any future competition
and to hinder and/or eliminate any existing competition in the measurement of
national radio and subsequently television audiences, by mechanical, electronic or
automatic devices.
B. A merger with the C. E. Hooper Co.
C. The development of a program of long-range contracts to hinder and to eliminate
any future competition.
D. A program of gradually increasing fees.
E. A program of capturing through a form of tie-in sales and/or agreements, advertis-
ing agencies and clients (Sindlinger testimony in Harris Hearings, 1963, p. 1680).
Nielsen executives denied the existence of the document referred to by

Sindlinger and, in any event, Nielsen never responded to the subpoena as
the parties settled their litigation. Nonetheless, in its 1963 complaint the
FTC referred to Nielsens actions from 1946 to 1963 as constituting a pro-
gram, the purpose of which has been and is now to monopolize, attempt to
monopolize and to restrain trade . (U.S. FTC, 1963, p. 1083).
Almost all observers view the 1963 hearings as a watershed event for the
radio and television audience measurement industry. The public interest
reform agenda of the hearings organizers nicely translated into political
capital given the publics keen interest in radio and television ratings.
While the hearings did not result in legislation or governmental regulation,
the revelations had many welfare-improving effects, though the underlying
structural problems of the industry remained.
Two small firms, Videodex and Conlan, were pushed to exit, while
Pulse, the leading local radio audience ratings firm, had its methods
severely criticized. Nielsens radio audience measurement also came under

substantial fire. Some observers believed that Arbitron was able to enter
and displace a weakened Pulse in the local radio market because of the
reputational damage suffered by Pulse in the hearings and because Nielsen,
which was a strong second in this market, exited in 1964 because of a com-
bination of factors that included its inability to cost-effectively measure
portable use and reputational damage from the hearings (Beville, 1988;
Buzzard, 2012, p. 45; RAB, 1982).
The FTCs consumer protection investigation into deceptive claims
resulted in an order that required more complete disclosures from the three
major rating services. Nielsen, ARB, and Pulse.24 Further, as discussed
above, the FTCs antitrust investigation led to a consent order giving
Nielsens rivals access to Nielsens audience measurement patents and
requiring that Nielsen obtain prior approval from the FTC for future
acquisitions over the following 10 years.
In partial response to calls by the hearings organizers for self-regulation
of the industry, the National Association of Broadcasters created the
Broadcast Ratings Council (BRC), a self-regulatory group composed of
members from the broadcasters, advertisers, and advertising agencies,
which would audit and certify rating services. Compliance was voluntary.
A number of industry research efforts on measurement methods were also
started which gave the users a better understanding of the strengths and
weaknesses of various ratings and provided ammunition with which the
users could pressure the raters to make improvements (Beville, 1988).
There are ironies regarding voluntary industry self-regulatory certifica-
tions and audits in a market involving a monopoly seller. A strong incum-
bent may be ambivalent regarding self-regulatory certification. On the one
hand, an extensive set of standards, for example, imposing minimum sam-
ple sizes and so on, increases entry costs, while certification gives credibility
to those who have actually entered. Furthermore, certification of rating
systems is not binding and hence dominant firms can (and eventually did)
proceed at times without certification.
Switching to Local People Meters in the United States, 20022009
Another notable threat to the radio and television audience measurement

firms occurred in the mid-2000s in the wake of shifts by both Nielsen and
Arbitron to people meter technology. This technology allowed the firms
to determine the viewing and listening patterns of individuals in local
markets. Nielsen had been using this technology for its national ratings,
but was now rolling it out in local television markets, while Arbitron
replaced its primarily diary-based system with its new meter-based system.
Any major shift in methods, samples, or technologies creates shared uncer-
tainties regarding accuracy of the ratings and their costs. Perhaps the big-
gest concern from this shift, however, arose from the heterogeneous effects
the change had on those being rated.
The initial impact of the change in the television ratings market was dra-
matic, with many programs experiencing drops of greater than 30% and
some greater than 60% (Napoli, 2014; Sunbeam Television Corp., 2011).
Nielsens local people meter (LPM) technology, perhaps because it elimi-
nated a recall bias which would favor the better known broadcast channels,
increased cable channel ratings at the expense of broadcast station ratings.
While the mechanism differed, under Arbitrons new portable people meter
(PPM), overall ratings of radio stations declined by 1530% and those of
some stations declined even more. Most of the criticism of these new sys-
tems was directed to defects in the sampling process (Napoli, 2014).
Critically, the impact of both of these shifts appeared to disfavor minority-
targeted programs and stations.
Although disgruntled stations chose to cancel their subscriptions to the
Nielsen ratings, this exit option was not effective against a monopolist pro-
vider of an essential service. As a result, many unhappy stakeholders took
their case outside the market in hopes of obtaining changes through public,
governmental, and legal pressure.25 The public interest issues to which they
attached their diverse interests were the impact of inaccurate ratings on the
diversity of programming and long-standing concerns about monopoly
unresponsiveness.
In the television audience measurement context, a key interest group
championing the diversity issue was Dont Count Us Out. Dont Count
Us Out played a role in attracting legislative interest to the issue which, in
turn, led to congressional hearings on S.1372 FAIR Ratings Act of 2005.
The bill sought to give the Media Ratings Council (formerly the BRC) for-
mal accreditation power over rating systems. Accreditation would be based
on principles of accuracy, equal representation, and disclosure of methods.
This bill was introduced by Senator Conrad Burns (R-MT) who previously
had founded a network of radio and television stations.
The FAIR Ratings Act hearings focused attention on the impact of rat-
ings on diversity of television programming, monopoly unresponsiveness,
and reduced innovation. Testimony was introduced that strongly suggested
that LPM caused ratings of minority-targeted programming to fall. During
the hearings, attention was directed to Nielsens decision to commercialize

its LPM in several cities even after being denied accreditation by MRC,
though MRC gave conditional or full accreditation within several months
in each of these cases (FAIR Hearings, 2005, pp. 1012).26 Nielsen argued
that mandatory accreditation would slow ratings innovation to a
crawl (FAIR Hearings, 2005, p. 14) and could inhibit entry. Nielsen also
voiced a need to support the MRC by agreeing to a new voluntary audit
and accreditation standard that will enable measurement services to
respond more quickly to dynamic changes. Nielsens critics responded
that the pace of innovation under Nielsen couldnt have been slower and
that Nielsen innovations in the past were responses to entry (Metzger testi-
mony in FAIR Hearings, 2005, pp. 3545). Nielsen also apparently threa-
tened that mandatory accreditation would lead it to discontinue its Nielsen
Hispanic Station Index (FAIR Hearings, 2005, p. 20). Ultimately, the bill
did not make it out of committee and Dont Count Us Out was later
revealed to be supported by News Corp (a broadcaster and station owner
that was disadvantaged by the change). Dont Count Us Out was dis-
banded in 2006 when News Corp signed a deal with Nielsen.
Similar events characterized Arbitrons implementation of its portable
people meter (PPM) for radio audiences in 2007. A range of minority-rights
interest groups organized resistance to the PPM claiming that the new mea-
surement technology undercounts minority listeners. In response to a
petition of the PPM Coalition, the Federal Communications Commission
explored whether the Arbitron service should be formally investigated,27
while the House Oversight and Governmental Reform Committee also
investigated Arbitrons actions relating to its PPM (Arbitron Hearings,
2009). The PPM Coalition and others also encouraged governmental offi-
cials in various states and localities to bring lawsuits against Arbitron seek-
ing damages and injunctions against unaccredited PPM implementation
(Napoli, 2014). Agreements with the interested parties caused the FCC
investigation and the various lawsuits to be dropped.28
A cynical interpretation of the broader events surrounding the chal-
lenges to and eventual acceptance of the PPM and LPM is that they were
part of nonmarket strategies undertaken by parties who would be eco-
nomically disadvantaged by the technology changes and who tried to
mitigate the effects of those changes through nonmarket threats. Buyers
did this by organizing and then cooperating with various interest groups
which represented minority groups, encouraging government bodies to
bring lawsuits, and by bringing their own lawsuits. The combination of
perceived arrogance and nonresponsiveness of the monopolists Nielsen

and Arbitron along with the public interest in maintaining the diversity of
media program content, provided an attractive basis for governmental
attention. This attention is perhaps best thought of as the use of voice
to prod a monopolist to take accommodating actions as the threat of
actual legislation was likely quite tenuous given the powerful interests
represented by those who were advantaged by the changes in ratings
and who wished to avoid ratings turmoil. Also, as it turned out, neither
the FCC nor the FTC engaged in formal investigations of Nielsen
or Arbitron.
The use of regulatory or litigation channels has the advantage of circum-
venting majoritarian politics as does the general public pressure created by
the activist groups (Baron, 2003). The basis of the lawsuits against the
deployment of the people meters was that the new system was represented
as improving accuracy, but actually decreased accuracy to the detriment of
the suing parties. Both Nielsen and Arbitron strengthened this claim by
running ahead of the self-regulatory accreditation process. Unsurprisingly,
Nielsen and Arbitron settled their lawsuits with each offering some conces-
sions to their respective plaintiffs complaints.29
Summary
The two governmental hearings discussed in this section illustrate how pub-
lic attention can be focused on particular markets. The Harris Hearings
were a response to a wide range of concerns held by buyers, users, rivals,
and the public at large regarding the poorly perceived performance of the
audience measurement markets. While it does not appear that the threat of
substantial government intervention in these markets was serious, the hear-
ings seemed an effective use of voice which surfaced a range of embarras-
sing conduct by many firms in the audience measurement markets. These
hearings likely hastened the demise of fraudulent ratings firms, precipitated
a shakeout among industry participants, perhaps even providing one
factor that persuaded Nielsen to exit the radio market, and led to some
self-regulation of quality on the part of the industry.
The FAIR hearings illustrated a more focused use of voice in which var-
ious interest groups brought pressure on Nielsen and Arbitron that ulti-
mately (along with some public and private lawsuits) led the two firms to
negotiate accommodations to the complaining parties concerns. Here, the
interest groups concerned had no chance to create change via actual legisla-
tion, but could use the hearings and lawsuits to pursue their narrow
interests.30
ALTERNATIVE MODELS OF TELEVISION AUDIENCE

MEASUREMENT MARKETS
Earlier we noted that restrictions on coordinated buyer action were dis-
couraged by antitrust concerns in the United States. Outside of the United
States, joint buyer action is common and, given the various state policies
regarding the development of radio and television, different models of audi-
ence measurement have evolved. In this section, we focus on television
audience measurement and examine joint buyer groups in the United
Kingdom and Australia and recent developments in India which have led
to government certification of raters and the emergence of a new joint,
industry-run rating organization.
Worldwide, audience measurement markets (which are geographically
defined at the country level because broadcast stations are licensed by
countries), are viewed as organized around one of the three models: the
own-service (OS) model in which ratings are provided by third-party pri-
vate-sector firms; the joint industry committee (JIC) model where a com-
mittee (usually representing broadcasters, advertisers, and advertising
agencies) specifies its criteria and then negotiates with firms to provide the
data; and the media owner contract (MOC) model where one or more
broadcasters delegate their own surveys or ratings research. There are a
few examples where an industry group undertakes its own ratings as well.31
The United States is an OS model country. The United States takes a
noninterventionist approach to potential inefficiencies posed by the radio
and television audience measurement industry. It relies on within-industry
self-regulation to solve specific problems raised in these markets and
across-industry law enforcement to police general competition and consu-
mer protection excesses. In contrast, most other countries have opted for
the JIC model (Bourdon & Meadel, 2014; Commission of the European
Communities, 2008). The development of these models of organizing the
ratings markets have also been shaped by the long-time dominance of
broadcast models which were subscription or state-funded rather than
advertising-funded as well as the stance of the various countries regarding,
otherwise problematic, activities to offset the power of a dominant firm.
A primary motivation behind the JIC approach is to control what is

measured and how it is measured. Subsequent bidding to provide the
services themselves should induce greater innovation and shift more of
the surplus toward the buyers. Moving these decisions away from a single
authority to a consortium, however, potentially makes decisions more
political which may slow decision making and create a status-quo basis
reflecting the (satisfied) self-interest of many of the member firms.
Furthermore, the effectiveness of these franchise-bidding competitions
depends on whether the bidders perceive that non-incumbents have a realistic
prospect of unseating the incumbent. This, in turn, depends on how dissatis-
fied the users appear to be with the incumbent and the history of who wins
the competitions.
Television Audience Measurement in the United Kingdom
The Broadcasters Audience Research Board (BARB) in the United

Kingdom is a good example of the JIC approach. Initial television audience
measurement in the United Kingdom was developed as a tool for BBC to
understand its audience and later to make programming choices
(Schwarzkopf, 2014). When commercial television was authorized in the
United Kingdom in 1955, it was regulated by the Independent Television
Authority (ITA). Commercial television (ITV) used third-party raters that
focused on advertising as opposed to programming. Competition between
BBC and ITV spilled over into ratings where wildly conflicting figures
were produced by each groups raters (Annan Report, 1977). In response
to the Annan Report which, within a general review of broadcasting in the
United Kingdom, recommended a single rating system for all broadcasters,
BBC and ITV created BARB, which contracted with Audits of Great
Britain (AGB) for quantitative research while taking input from the BBC
research office regarding programming (Schwarzkopf, 2014).
The television audience measurement portion of BARB is a joint indus-
try committee (JIC) self-regulatory group which selects and manages the
methods and samples used in the rating and puts out to bid exclusive long-
term contracts to provide television audience measurement services. BARB
is a non-profit company owned by the broadcasters and the Institute of
Practitioners in Advertising. Controversy regarding appropriate methods
takes place within BARB as a means for revising BARBs choices. This
contrasts with the U.S. case where pressure is brought to bear directly on
the rater itself, either through attempts to supplant the incumbent or
through stakeholders which use legal and legislative processes to pressure

for change (Balnaves, 2014).
Television Audience Measurement in Australia
OzTAM, which is owned by the three major television networks in

Australia, is a good example of a MOC. OzTAM provides its subscribers
with television audience ratings for five metropolitan areas and for sub-
scription television across Australia. It contracts with an audience measure-
ment firm to generate data according to its specifications and standards
and then sells the data to its subscribers. The contract with an audience
measurement firm runs for 57 years, but has been extended during the
middle of the contract. Revenues come from subscriptions to the data ser-
vice.32 A portion of the revenues OzTAM receives is used to fund audience
measurement research. The Australian Association of National Advertisers
and other user groups have observer status at OzTAM board and techni-
cal committee meetings.
In most of these bidding competitions the incumbent, frequently the
local company when there is one, is favored. For example, in Germany,
GfK, the German audience measurement firm, has supplied television audi-
ence measurement for more than 20 years.33 Some upheaval in the firms
chosen to do audience measurement has, however, occurred in recent years.
For example, in the United Kingdom, BARB switched its television
audience measurement contract from Nielsen to TNS;34 in Australia,
Commercial Radio Australia shifted its radio audience measurement con-
tract from Nielsen to GfK.35
The JIC or MOC approaches are not without social welfare risk. While
the banding together of users is a valuable tool to offset the monopoly
power of a rating firm, such an organization, especially if it is composed of
a small subset of the entire group of users, may merely move the monopoly
surplus from a firm to this subset of users.36 To mitigate such possible
anticompetitive possibilities, OzTAM, for example, agreed with the
Australian Competition & Consumer Commission to supply Elemental
Data on reasonable commercial terms to any [qualified] third party
(OzTAM, 2000). There may also be antitrust impediments to joint negotia-
tion arrangements as in the United States.
If the decision making group running a JIC or MOC is composed
of firms whose individual interests conflict, there is the possibility that
competition created by the competitive bidding process may not produce
welfare-enhancing outcomes. This situation is similar to what might be faced

by industry technical standards committees in which a new standard might
be delayed because of the interests of committee members who might try to
delay adoption of a new standard because they prefer the current standard.
Television Audience Measurement in India, 20082014
The evolution of the television audience measurement industry in India

provides an important point of reference against which to compare evolu-
tion of the U.S. television audience measurement industry. Until recently,
the ratings industry in India used an own-service firm model similar to that
of the United States. Since 2008 the Indian model has evolved to a hybrid
government certification/JIC model and may end up with joint industry
provision of ratings.
Television audience measurement did not exist in India until 1988.
After a brief period of competition, TAM Media Research, a joint
venture between Kantar Media Research/IMRB and Nielsen Media
Research, became the dominant firm (TRAI, 2008b). In 2008, the
Ministry of Information & Broadcasting (MIB) directed the Telecom
Regulatory Authority of India (TRAI) to assess the landscape and to
provide recommendations regarding what, if any, action was warranted.
TRAI subsequently issued a Consultation regarding television ratings
services which emphasized the enormous importance of ratings data,
the challenge of generating such data with sufficient accuracy, and the
practical reality that the industry tends almost invariably toward a mono-
poly owing to the importance of having a single currency with which to
buy and sell advertising (TRAI, 2008a, 2008b).
The central question TRAI posed was whether there is a need for
the Government to regulate the system of Television Rating Points (TRP),
or whether this should be left to be decided by industry initiative .
TAMs response was clear. In response to the question of [s]hould
Government intervene into an industry created TV Ratings system? The
answer is an emphatic: No. TAM observed that its parent companies
run TV audience measurement services across 40 + countries and in almost
all countries, they work with the local industry body. There is no govern-
ment involvement. With regards to TRAIs concern regarding monopoly
status, TAM responded that, [a]n audience measurement service is unlike
other industry services where the output is used as a trading currency for
airtime. Hence the need for a singular service is essential (TAM, 2008).
Following its Consultation, TRAI recommended self-regulation by the

Broadcast Audience Research Council (BARC), a recently formed volun-
tary effort of industry stakeholders to oversee and control the TV audi-
ence measurement system in India, or its equivalent. TRAI further noted
that a guideline framework is needed to maintain quality standards and miti-
gate the deficiencies observed in the current ratings service (TRAI, 2008b).
The dissatisfaction that arguably fueled the 2008 consultation and
recommendations appeared to have intensified over the following years.
The sole, albeit fledgling, competitor to TAM, exited the industry in 2011.
TAM claims to have been responsive to recent buyer concerns, but some
key buyers completely disagree. New Delhi TV argued, for example, that
TAMs priority was staving off regulation.37 Meanwhile, BARC was beset
with numerous governance and funding problems. In 2012, MIB asked
TRAI to undertake another Consultation because a self-regulated televi-
sion rating system has failed to take off, as BARC has been unable to take
any credible action on the recommendations made by TRAI. In 2013,
TRAI and MIB, while being highly critical of BARC, remained in favor of
a self-regulatory approach (TRAI, 2013).
As a result of this consultation, the contours of what would constitute
industry self-regulation have changed. In particular, whereas previously no
government registration of rating firms with the government was sought,
firms must now register and be approved by the government as having met
a set of requirements that include, for example, minimum sizes of the sam-
ple panel, speed with which the panel is rotated, composition of the panel,
disclosures. The rating firms continued operation would then be predicated
on their ongoing compliance with those specific regulations.
This outcome stands in stark contrast to an industry-based scheme in which
the free market or the industry itself plays the central role in generating the
accreditation requirements or where accreditation is implicit in the specifica-
tions for audience measurement put out for bid. Additionally, it appears that
operation consistent with the regulatory guidelines is not voluntary. By con-
trast, the comparable entity in the United States is a voluntary organization.
While MIB has not seen fit to incorporate the ratings function into a
government agency, it should be noted that the TRAI consultations and
recommendations and MIBs statements specifically reserve the right to
revisit this issue. It remains to be seen whether, in fact, BARC, which
appears to be angling toward running its own rating system, can implement
the system with which it has been tasked. BARC has been aggressive in
selecting technology and partners with which to generate the ratings infor-
mation that it will own. TAM, at least for now, is not one of the partners.
Although much of the criticism has been directed toward BARC, one
also wonders if TAM could have played a more active role in making the
original and less threatening conception of BARC more successful. In any
event, it remains to be seen if BARC can manage audience measurement
on its own. If not, perhaps TAM will eventually reemerge as the dominant
player in audience measurement in India.
DISCUSSION: NO EXIT AND THE POLITICAL

ECONOMY OF COMPLAINTS
A monopolist in audience measurement services is in an enviable position:
its ratings are essential, it has significant advantages over entrants, and the
market itself is unlikely to support more than one ratings firm. Individual
buyers of such services, on the other hand, are in a weak position because
they have no exit option, so must either band together or exercise their
voice (Hirschman, 1970) directly by complaining to the seller or indir-
ectly through the political system, the legal system, or the media.
While monopoly markets obviously differ in their particular charac-
teristics and, hence, in the strategies that a monopolist can be expected to
pursue, there are a number of lessons from the radio and television rating
markets that seem generally applicable. The first set of lessons concerns
actions to limit the exit options available to buyers and the second set
involves navigating the complications caused by voice.
Navigating the exit option is most difficult when buyers have formed
an industry-wide buying group that puts out for bid the monopoly
franchise. When the request for bid does not contemplate major improve-
ments, the incumbent has a number of advantages over those seeking
to displace it, including lower costs and lower uncertainty regarding future
outcomes. But when the bid is oriented toward new technologies, the
incumbents natural cost advantage is reduced because a new technology,
if adopted, would force the winning bidder, including the incumbent, to
make major investments. Both the conditions of possible displacement the
desired specifications and the decision regarding displacement depend on
how such decisions are made by this buyer group. An important feature of
these buyer groups is the extent to which their interests vis-a-vis the status
quo and the contemplated changes are heterogeneous. If interests are hetero-
geneous, as is the case in the ratings markets, and, particularly if each buyer
strongly believes that its individual interests are best served with a single
supplier, then the monopolist can be expected to shape its products or ser-
vices to ensure that the majority of the decision makers in the buyer group
will be unwilling to shift to a new supplier. Here, the governance structure of
the buyer group may play an important role in determining, for example, if a
contract is put up for bid, the nature of the specifications that are put out
for bid, and how the various bids are evaluated. Thus, buyers who anticipate
such actions by the incumbent should commit to decision rules to reduce the
role individual self-interest will play in decision making (e.g., by removing
some discretion from the process or by delegating more power to parties
whose self-interest cannot be as easily manipulated).
In some political jurisdictions (e.g., the United States), the formation of
such buyer groups is prevented by antitrust or competition laws. There, a
priority objective for the monopolist is to prevent legislative or regulatory
action that would allow buyer groups to form. Such a nonmarket tactic
may involve working with firms in other industries which also wish to resist
the relaxation of antitrust laws along such lines or, as in the United States,
it may be largely given. Then, assuming the joint buyer group problem is
avoided, the incumbent faces the difficult, albeit conventional, task of
avoiding being supplanted by an entrant claiming a superior value proposi-
tion. But this improvement in the value proposition will need to be substan-
tial because successful entry in a monopoly market likely requires buyers to
believe that the entrant will win the monopoly in a relatively short time and
that belief, in turn, requires that the value proposition (after the incumbents
response) is preferred by a majority of the buyers, not just a modest share,
as would be the case in markets that can support several competitors.
Furthermore, one expects that the incumbent has leveraged its market
power to raise switching costs for most buyers and may have more flexibility
than the entrant to offer lower pricing, especially in the situation where the
entrant offers higher quality rather than the same quality at a lower cost.38
Where the formation of joint buyer groups is circumscribed and no sub-
set of buyers is willing to sponsor entry, buyers primary means to influ-
ence the monopolist is through voice. Voice manifested as nonmarket
threats involving lawsuits, legislation, or public embarrassment can result
in many types of outcomes. At one extreme, voice could lead to a govern-
ment takeover. While the goals may differ, if private-sector stakeholders
are united in their distaste for governmental regulation, intervention would
also have to overcome these relatively focused business interests. At the
other extreme, voice is merely a means to increase the responsiveness of
monopoly firms and, perhaps, to galvanize users to undertake joint actions
that improve rating quality or usefulness.
Monopoly firms may navigate these threats by making concessions

needed to satisfy individual parties, thereby draining any larger movement
toward government or even self-regulatory action by its major supporters.
In the audience measurement industry, we see this in settlements of various
lawsuits by both Nielsen and Arbitron. Threats by monopoly providers to
withhold services or even exit have also been used to scare government
away from large-scale intervention.
The individual concession approach has been relatively successful, in
part, because the most aggrieved parties are primarily motivated by specific
rather than system-wide concerns, politicians are content with gaining poli-
tical capital by declaring a quick victory even while the true public interest
objective remains unmet, and the interests of those desiring major change
are frequently in conflict. This last point connects market interests to non-
market possibilities. Even after Nielsen and Arbitron had established their
hegemonies in their respective U.S. markets, nonmarket actions failed to
weaken their hold on those markets. Perhaps a key to their success is that
audience measurement is a monopoly industry where the users are inher-
ently divided: exit is quite difficult and voice is not harmonious.
NOTES
1. Arbitron grew out of ARB (American Research Bureau) while Nielsen oper-
ated under various names such as A. C. Nielsen and Nielsen Media Research.
2. In addition to disagreement regarding the audience composition, other chal-
lenges include selecting an appropriate sample given the constantly changing demo-
graphics and size of the relevant population, knowing who is watching or listening,
various actions to manipulate or hype audience levels during sampling periods,
changes in the modes of viewing or listening (e.g., radio listening in automobiles
rather than at home), identifying faulty data, as well as particular weaknesses asso-
ciated with different measurement systems (e.g., recall of survey respondent,
prompting techniques, inattention to making diary log entries, keeping television on
while not viewing, device failures) (Beville, 1988).
3. As the industry matured, the markets became increasingly concentrated.
Even by the early 1960s, the U.S. national and local television markets were consid-
ered to be near monopolies held by Nielsen with some competition by ARB (Harris
Hearings, 1963, p. 1113); while Nielsen held a national radio market share of
8085% (Harris Hearings, 1963, p. 1113) and shared the local radio market with
market leader Pulse and some small players (Harris Hearings, 1963, p. 1280).
4. But see Balnaves, ORegan, and Goldsmith (2011) who describes the
Australian radio audience measurement market as one in which two firms continued
to compete until 1973 when they merged.
5. Victor Sholis, a vice president of a local radio station, offered a typical senti-
ment along these lines. there has been no way to ga[u]ge validity, reliability or
integrity of the radio ratings (Harris Hearings, 1963, p. 192). In the 1985 FAIR
Ratings Act Hearings, Gale Metzger former President of Statistical Research, Inc.
which conducted research for the industry stated Are Nielsens new systems better
than the old? Are the audience estimates more accurate? The truth is, no one
knows [because][c]lients cannot get to information they need for decisions
(FAIR Hearings, 2005, p. 33).
6. Assessing ratings quality through the connection between ratings and sales is
another possibility, but one confounded by many factors involving pricing and
other marketing efforts associated with the product being sold.
7. Ratings differences increase the pressure on raters to provide information
needed to understand a rating systems strengths and weaknesses and competition
induces the raters to critically assess each others approaches and explain such
assessments to the buyers. Despite these advantages to competition, few users in the
United States have consistently supported a second rating system and most markets
have only one ratings firm. The free-rider problem creates obstacles for funding a
second service as a check on the first. At times, the various television networks
encouraged entry, but such support did not persist, presumably in part because the
incumbent made concessions that made the second rater less valuable.
8. Comparison problems also emerged in other country settings. A 2002 switch-
over to a new panel by BARB in the United Kingdom led to suspension of reports
because of the substantial apparent drop in viewership (BBC News, 2002).
9. Much of the problem with ratings dispersion can be traced to changes in
audience demographics and usage patterns and the effectiveness of various rating
systems for capturing these changes. Sometimes the mismatch between the audience
that the advertisers want to measure and the audience that the rating system is mea-
suring is quite evident. In the middle of the 20th century, huge increases in
portable and mobile radio listening created major challenges to existing ratings ser-
vices that were built around home listening. FM radio listenership was not well
measured in its infancy. The shift to cable television in the 1970s and 1980s exacer-
bated the recall problem associated with diary-based or telephone survey methods
(Rubens, 1984). Now, the shift to cross-platform viewing is the challenge along
these audience usage pattern lines. It is also useful to note that changes in rating
systems not only reflect changes in the effective sample, but also differences in view-
ing routines (Bourdon & Meadel, 2014). But see Beville (1988) for a view that dis-
persion is less a problem than many think it is.
10. Buyer assessment of rating methodologies is difficult in part because key
methodological elements are unobservable and very reluctantly disclosed. For
example, examination of the inputs requires knowledge regarding the sampling
algorithm or the procedures through which data is generated, cleaned, or rejected.
Raters have been reluctant to fully disclose the weaknesses and caveats of their
studies, especially in written materials, or to share raw data with the users. This
stance is easier for monopolists to take. Although the rating firms claim to have
been straightforward with respect to the statistical reliability of their ratings, users
have long complained about the transparency of the rating methods. This concern
led to settlements in the early 1960s between leading ratings firms and the FTC
regarding appropriate disclosure and caveats in materials pertaining to

raters services.
11. These concerns also have impacts on geographical expansion. In the local
radio market, for example, locations with fewer stations and smaller population
bases may be insufficient to support a local sample size that would produce high
quality ratings. Ratings firms choosing to operate in such locales would necessarily
produce lower quality ratings which, in turn, could have a negative impact on parts
of the business where higher quality ratings are produced.
12. Antitrust analysis has been lenient with acquisitions of failing firms.
However, some acquisitions were arguably of firms that, while currently not profit-
able, had a positive profit future.
13. Strategic alliances, including licensing, are also subject to scrutiny under the
Clayton Act. Alliances can be treated more leniently than acquisitions because the
alliances are typically time limited, narrowly focused, do not involve substantial
integration of coordination between the parents, and they usually have strong effi-
ciency justifications. Access to otherwise exclusively held intellectual property is
also a strong justification for alliances (see, e.g., Yao, 1997).
14. Billboard (1963); Washington Week, November 4, 1963.
15. Monopolies sometimes engage in price discrimination and bundled pricing.
16. United States v. Grinnell Corp. (1966).
17. Fierman (1985) reports that [i]ndustry watchers maintain that Nielsens
profits in the ratings business are already as much as 60% of sales before taxes.
18. Interview with Doug Pieffer (CEO, OzTAM), October 15, 2014.
19. Beville (1988, p. 69), for example, notes that deployment of metering technol-
ogy that would better measure cable channel viewing would need to be supported
by the less well-established cable interests as non-cable broadcasters expected to be
disadvantaged by such a change. See also Eaman (1984, p. 131) who cites the con-
cerns of an ABC executive regarding the expected substantial loss of overall viewers
relating to a measurement system switch.
20. Much criticism has been leveled at Nielsen for being slow to innovate once it
had established its monopoly. It is argued that Nielsen was a fast second in
response to threatened technology-based entry (Buzzard, 2002). In addition,
Nielsens deployment of new technologies in the United States appeared to lag that
deployment in some other countries. Finally, slow innovation or at least slow
deployment may also result in part from the somewhat zero-sum nature of ratings.
Buyers who think their ratings will decline under a new system may resist
improved rating systems.
21. ARB and Nielsen settled their patent dispute. The settlement resulted in
ARB licensing Nielsen patents for 5% of revenue going forward (Harris Hearings,
1963, p. 1570) and providing royalty-free licenses to ARB inventions relating to
receiver monitoring equipment (Harris Hearings, 1963, pp. 1570, 1697).
22. It was reported in 2007 that Nielsen brought an infringement lawsuit against
TNS which had just won the BARB measurement contract in the United Kingdom.
Nielsen claimed that the technology TNS put forth, infringed on Nielsen patents
(The Guardian, December 18, 2007).
23. For example, the European Commission report on the VNU/WPP proposed
joint venture noted the importance of Nielsens Program Signal Identification
Data Collector patent (Commission of the European Communities, 2004). More

recently, the primary condition that the FTC required to avoid challenging the
Nielsen-Arbitron merger was the divestiture and sale of intellectual property assets
relating to cross-platform audience measurement services. FTC Press Release,
FTC Approves Final Order Settling Charges that Nielsen Holdings N.V.s
Acquisition of Arbitron, Inc. Was Anticompetitive, February 28, 2014.
24. Despite the drubbing the industry took as a result of the hearings, Nielsen
could not help itself from commenting that FTCs disclosure order addressed fine
print and that Nielsen will continue to act as we have always done.
25. The interests arrayed against the LPM also attempted to halt the rollout of
LPM in Los Angeles through court order.
26. In these hearings, the MRC testified that Nielsen had commercialized its
LPM service in Boston in 2002 even after being denied accreditation by MRC
(received nine months later). Nielsen commercialized its New York LPM in 2004
also after having been denied accreditation (though it received conditional accredi-
tation four months later). A shorter delay between commercialization and deploy-
ment occurred in Los Angeles; Chicago, Philadelphia, San Francisco, and
Washington were commercialized prior to an MRC audits (FAIR Hearings,
2005, pp. 1012).
27. FCC Media Bureau Action, PPM Coalition Files Petition Seeking Commission
Inquiry Pursuant to Section 403 of the Communications Act (47 U.S.C. 403),
September 4, 2008. The PPM Coalition consisted of the National Association of
Black Owned Broadcasters, Spanish Radio Association, Minority Media and
Telecommunications Council, American Hispanic Advertising Association, Border
Media Partners, Entravision, ICBC Broadcast Holdings, Spanish Broadcasting
System, and Univision.
28. Bachman (2010).
29. See, for example, Bachman (2010).
30. See, for example, the use of private politics (Baron, 2003).
31. BBM in Canada is an industry-sponsored committee which itself managed a
rating service. This service is largely funded by broadcasters by subscription but is
controlled equally by advertisers, advertising agencies, and broadcasters (Eaman,
1994, p. 59).
32. Telephone interview with Doug Pieffer, CEO OzTAM, October 15, 2014.
33. GfK website, References-Some Exemplary Projects, Accessed on October
17, 2014.
34. Sweeney (2007).
35. GfK, Press Release: GfK Expands Audience Measurement and Wins Radio
Contract in Australia, February 25, 2013.
36. In some settings, incorporation of the self-regulatory organization as a non-
profit is helpful.
37. New Delhi TV took the unusual step of filing litigation against Nielsen in
New York State. It claims in its complaint that when it met with TAM following
the 2008 India government consultation, TAM emphasized avoiding regulation.
38. Offering lower prices to users is easy for the incumbent given its already
amortized fixed costs, production economies, and compatibility advantages. It also
helps that the ratings business has many complex components which need to be
mastered to offer a good service (Buzzard, 2002).
ACKNOWLEDGMENT
The authors thank Young Hou and Lydia Kim for research assistance and
Erin Rubin for editorial support.
REFERENCES
Bachman, K. (2010). Arbitron settles with PPM coalition. Adweek, April 22.
Balnaves, M. (2014). Still the British model? The BARB versus Nielsen. In J. Bourdon &
C. Meadel (Eds.), Television audiences across the world: Deconstructing the ratings
machine (pp. 5368). New York, NY: Palgrave Macmillan.
Balnaves, M., ORegan, T., & Goldsmith, B. (2011). Rating the audience: The business of
media. London: Bloomsbury Academic.
Baron, D. P. (2003). Private politics. Journal of Economics & Management Strategy,
12, 3166.
BBC News. (2002). TV ratings spat rumbles On. BBC News, January 31.
Beville, H. M., Jr. (1988). Audience ratings: Radio, television, and cable (Rev. ed.). Hillsdale,
NJ: Lawrence Erlbaum Associates.
Billboard. (1963). Investigators to listen to Collins plea. Billboard, May 18, p. 10.
Bourdon, J., & Meadel, C. (2014). The monopoly that wont divide: Frances mediametrie. In
J. Bourdon & C. Meadel (Eds.), Television audiences across the world: Deconstructing
the ratings machine (pp. 88101). New York, NY: Palgrave Macmillan.
Boyer, P. J. (1988). The media business: British People Meter firm suspends its U.S. opera-
tion. The New York Times, August 1.
Buzzard, K. (1990). Chains of gold: Marketing the ratings and rating the markets. Metuchen,
NJ: Scarecrow Press.
Buzzard, K. (2012). Tracking the audience: The ratings industry from analog to digital.
New York, NY: Routledge.
Buzzard, K. S. F. (2002). The peoplemeter wars: A case study of technological innovation and
diffusion in the ratings industry. Journal of Media Economics, 15, 273291.
Collis, D., & Montgomery, C. (2005). Corporate strategy: A resource-based approach (2nd ed.).
New York, NY: McGraw-Hill/Irwin.
Commission of the European Communities. (2004, September 15). Case No. COMP/M.3512-
VNU/WPP/JV.
Commission of the European Communities. (2008, September 23). Case No. COMP/M.5232-
WPP/TNS.
Doz, Y., & Hamel, G. (1998). Alliance advantage: The art of creating value through partnering.
Boston, MA: Harvard Business School Press.
Eaman, R. A. (1994). Channels of influence: CBC audience research and the Canadian public.
Toronto: University of Toronto Press.
ErinMedia, L.L.C. v. Nielsen Media Research, Inc., 401 F. Supp. 2d 1262 (M. D. Fla. 2005).
Fierman, J. (1985). Television ratings: The British are coming. Fortune, April 1.
Greene, H. (2015). Muzzling antitrust: Information products, innovation, and free speech.
Boston University Law Review, 95, 35107.
Hirschman, A. O. (1970). Exit, voice, and loyalty: Responses to decline in firms, organizations,
and states. Cambridge, MA: Harvard University Press.
Madow, W. G., Hyman, H. H., Jessen, R. J., Sheatsley, P. B., & Wright, C. R. (1961). Report
of the American Statistical Association Technical Committee on Broadcast Ratings to
provide information requested by the Subcommittee on Legislative Oversight of the
House Committee on Interstate and Foreign Commerce (Madow Report).
Napoli, P. M. (2014). The local peoplemeter, the portable peoplemeter, and the unsettled law
and policy of audience measurement in the United States. In J. Bourdon & C. Meadel
(Eds.), Television audiences across the world: Deconstructing the ratings machine
(pp. 216233). New York, NY: Palgrave Macmillan.
Oberholzer-Gee, F., & Yao, D. A. (2013). Market imperfections and sustainable competitive
advantage. In C. Thomas & W. Shughart II (Eds.), Oxford handbook of managerial eco-
nomics (pp. 262277). New York, NY: Oxford University Press.
OzTAM Pty Limited. (2000). Undertaking by OzTAM Pty Limited to Australian competition &
consumer commission concerning agreement with ATR Australia Pty Limited.
Porter, M. (1980). Competitive strategy. New York, NY: Free Press.
Radio Advertising Bureau. (1982). Re: Analysis of and request for business review letter for a radio
industry negotiating committee regarding audience measurement services, B.R.L. 845.
Report of the Committee on the Future of Broadcasting. (1977). Command 6753. HMSO
(Annan Report).
Rubens, W. S. (1984). High-tech audience measurement for new high-tech audiences. Critical
Studies in Mass Communication, 1, 196.
Saloner, G., Shepard, A., & Podolny, J. (2001). Strategic management. New York, NY: Wiley.
Schwarzkopf, S. (2014). The politics of enjoyment: Competing audience measurement systems
in Britain, 19501980. In J. Bourdon & C. Meadel (Eds.), Television audiences across
the world: Deconstructing the ratings machine (pp. 3352). New York, NY:
Palgrave Macmillan.
Sunbeam Television Corp. v. Nielsen Media Research, Inc., 763 F. Supp. 2d 1341, 135253
(S.D. Fla. 2011), affd, 711 F. 3d 1264 (11th Cir. 2013).
Sweeney, M. (2007). TNS swipes Barb contract from Nielsen. The Guardian, December 18.
TAM Media Research. (2008, April 18). TAMs response to TRAI consultation paper.
Telecom Regulatory Authority of India. (2008a, July 24). Draft recommendations on policy
and operational issues for Television Audience Measurement (TAM)/Television Rating
Points (TRP).
Telecom Regulatory Authority of India. (2008b, March 28). Consultation paper on policy
guidelines for and operational issues for Television Audience Measurement (TAM)/
Television Rating Points (TRP).
Telecom Regulatory Authority of India. (2013, September 11). Recommendations on guide-
lines for Television Rating Agencies.
United States v. Grinnell Corp., 384 U.S. 563 (1966).
U.S. Department of Justice. (1984). Response letter: Radio Advertising Bureau, B.R. L. 845.
U.S. Department of Justice and U.S. Federal Trade Commission. (2007). Antitrust enforce-
ment and intellectual property rights: Promoting innovation and competition: A report
issued by the U.S. Department of Justice and the Federal Trade Commission.
(Antitrust-Intellectual Property Report)
U.S. Federal Trade Commission. (1963, October 23). In the matter of A. C. Nielsen Company.
63 F.T.C. 1082 (consent order).
U.S. Federal Trade Commission. (2014, February 28). FTC approves final order settling
charges that Nielsen holdings N.V.s acquisition of Arbitron, Inc. was anticompetitive
(press release).
U.S. House of Representatives, Committee on Oversight and Government Reform. (2009).
Will Arbitrons personal people meter silence minority owned radio stations?
(Arbitron Hearings).
U.S. House of Representatives, Subcommittee of the Committee on Interstate and Foreign
Commerce. (1963). Hearings on the methodology, accuracy, and use of ratings in broad-
casting (Harris Hearings).
U.S. Senate, Committee on Commerce, Science, and Transportation. (2005). Hearings on
S.1372 the FAIR Ratings Act (FAIR Hearings).
Yao, D. A. (1997). Antitrust constraints to competitive strategies. In G. Day & D. Reibstein
(Eds.), Wharton on dynamic competitive strategy (pp. 313337). Hoboken, NJ: Wiley.
THE ORGANIZATION OF
NONMARKET STRATEGY
Dylan Minor
ABSTRACT
The purpose of this paper is to explore how firms organize to engage in

nonmarket strategy. To achieve this end, we explore the organization of
nonmarket strategy via a formal model of the firm. The model is moti-
vated by a qualitative study of the organization of nonmarket strategy of
25 large, US firms. Firms either integrate nonmarket strategy activities
throughout the firm or create stand-alone business units that specialize in
nonmarket strategy activities. We find that the advantage of integration
over specialization is U-shaped in the importance of nonmarket strategy
to the firms market strategy. We identify several other factors that
predict the advantage (and disadvantage) of integration over specialization.
The value of this paper is that it is (to the best of our knowledge) the
first to identify the factors that should cause a firm to either integrate or
specialize the organization of its nonmarket strategy. It also develops an
original typology of the organization of nonmarket strategy.
Keywords: Nonmarket strategy; corporate social responsibility;
organizational design

ISSN: 0742-3322/doi:10.1108/S0742-332220160000034012
413
414 DYLAN MINOR
INTRODUCTION
One of the most important points to be made up front is that there is no single univer-
sally accepted method for designing a CSR structure
Business for Social Responsibility (2002)
How a firm organizes is a fundamentally important question. There has been

a plethora of work exploring how to organize a firm in terms of what belongs
in and outside of a particular firm. Early work includes Coase (1937),
Williamson (1975), and Grossman and Hart (1986). More recently, the grow-
ing field of organizational economics has focused on organizing within
a firm. Primary areas of exploration have included corporate hierarchy
(e.g., Alonso, Dessein, & Matouschek, 2008; Garicano, 2000; Harris &
Raviv, 2002), task design (e.g., Holmstrom & Milgrom, 1991; Laux, 2001;
Schottner, 2008), delegation (e.g., Alonso & Matouschek, 2008; Krishna &
Morgan, 2008; Prendergast, 1995), and incentive design (see Bolton &
Dewatripont, 2005 for a survey). This paper continues in this stream of
literature by exploring how to organize nonmarket activities within a firm. As
will be discussed, there is clear heterogeneity in how firms currently organize
nonmarket operations. The purpose of this paper is to propose a typology
not only to help identify different types of nonmarket organizational struc-
tures, but also to understand why different organizational structures arise.
Nonmarket strategy is a broad term that refers to a firms activities outside
of the marketplace that can help it gain competitive advantage (Baron, 2009).
This includes both public politics strategies (e.g., lobbying and engaging with
regulators) and private politics strategies (e.g., engaging with activists).
Meanwhile, corporate social responsibility (CSR) can be a part of both of
these types of political strategies. CSR can be used to increase returns to
lobbying and prevent or soften future regulation (e.g., Baron, 2009; Hong &
Minor, 2014; Lyon, 2004; Minor & Morgan, 2011). CSR can also be used to
appease activists and possibly avoid future adverse activist actions (e.g., see
Barnett, 2007; Baron, 2009; Godfrey, 2005). Typically, firms do not have a
division called non-market strategy. Instead, firms have divisions that carry
out some of these nonmarket strategy functions, but these groups are often
referred to as Corporate Social Responsibility, Sustainability, or some similar
name. Consequently, for this paper, we will focus on CSR divisions to help
illuminate how firms organize their nonmarket strategy functions though
we do consider the organizational consequences of adding non-CSR elements
to a CSR group in Section titled Non-CSR Nonmarket Functions. For
the balance of the paper, we refer to nonmarket strategy organization as
The Organization of Nonmarket Strategy 415
organizing CSR strategy. To the best of our knowledge, this is the first paper
to explicitly explore the organization of these groups.
We begin in the next section by developing a typology of four types of
CSR strategy organization. We then use four firms to illustrate these cate-
gorizations. This qualitative analysis motivates our next section, where we
utilize organizational economics to explore when a firm should choose a
particular CSR strategy organizational form. In particular, we develop a
model where a firm must decide on organizing market strategy activities
and nonmarket strategy activities in separate units versus having both units
engage in both activities. There is a trade-off in this decision in that there is
an opportunity cost for a single unit to engage in both activities; however,
when both units engage in both activities, the firm receives outputs on both
domains from two units as opposed to a single output toward each domain
from a single, specialized unit. In this sense, the model is related to the
extant multi-tasking literature (see Holmstrom & Milgrom, 1991; Laux,
2001; Prendergast, 2002). However, there are some additional elements that
distinguish it from this literature. First, it is assumed that there are some
complementarities between market and nonmarket strategy activities. That
is, engaging in one activity can help the performance of the other. Second,
and more novel, is that there are externalities in our model, since we are
dealing with (possible) social output. In particular, both the manager and
the firm may value CSR beyond its ability to enhance financial performance.
This introduces the possibility of different outcomes organizationally than if
such externalities were not present.
We find that the organization of the firm is generally non-monotonic in
the importance of CSR to the firm where importance comes from the
degree of financial and social performance complementarities, the firms
value of CSR, and the managers value of CSR. In particular, for low and
high values of CSR importance, it is best for the firm to integrate its CSR
strategy business units should be engaged in both market and nonmarket
activities. However, for intermediate values of CSR importance, it is best
for the firm to organize CSR strategy activity into a stand-alone business
unit that specializes in CSR activities. The intuition is as follows. For high-
importance CSR settings, the synergy of CSR and financial performance
overcomes the (potentially high) cost of multi-tasking within a business
unit. In contrast, for low-value CSR firms, the return to having two units
both work on market strategy activities overcomes the multi-tasking cost,
which is small since business units minimally engage in CSR when it is not
very important. However, for intermediate values of CSR, these forces net
out in the opposite direction: CSR strategy activities are sufficiently
416 DYLAN MINOR
valuable to warrant a significant level of activity, but not valuable enough

to overcome the (increased) cost of distracting a unit from market strategy
activities. Hence, having CSR strategy activities located in a specialized
CSR unit is the best organizational design for this setting.
Related to our study is the organizational design literature using agent-
based simulation. Owing to the difficulty of finding closed-form solutions
in the analysis of organizational design while using multiple variables,
some have used agent-based simulation to generate large-sample-size
numerical examples to provide evidence of optimal organizational design.
For a recent example of this technique, see Claussen, Kretschmer, and
Stieglitz (2014). They study the trade-off of commitment and flexibility
within an organization. Similarly Rivkin and Siggelkow (2003) explore the
trade-off of organizational search and stability. Siggelkow and Rivkin
(2005) extend Rivkin and Siggelkow (2003) to allow for dynamic environ-
ments. Given a particular distributions of decisions across managers or
departments, these papers focus on how the primitives of the model predict
which organizational form converges to a superior (i.e., locally optimal)
performance level. In contrast, this current paper focuses on how to get the
distribution of choices right when a particular manager might also be
making choices on the same dimensions as another manager or business
unit. In addition, these agent-based simulations generally assume mean zero
performance complementarities between different decision dimensions, as
payoffs are randomly assigned assuming i.i.d. That is, they do not explicitly
consider the case where, for example, a nonmarket strategy can have a sys-
tematic bias to enhance market strategy. Finally, all choices in these models
are binary, which disallows the analysis of magnitudes of choices and their
relationship to organizational form. Thus, this paper complements this
strand of literature by exploring aspects of organizational design that are not
the focus of the extant papers. This paper also complements these organiza-
tional design papers by explicitly analyzing firm-level nonmarket strategy
choice and implementation. Baron (2016) argues that more attention should
be given to this level of nonmarket strategy analysis.
We now begin with our typology.
A TYPOLOGY OF NONMARKET ORGANIZATION

Based on personal discussions with senior executives of CSR units at
several SP500 firms and data obtained from 2013 sustainability reports
available on company websites, we identified two primary dimensions that

differentiate CSR units. One dimension is the degree of integration of a
firms CSR unit across the rest of the organization. One measure of the
degree of integration is to what extent the firm has a particular business
group specialize in CSR strategy activities. Another marker of the degree
of integration is how many levels of the organizational hierarchy decide on
CSR strategies. Some firms concentrate almost all decisions in the C-suite,
whereas others tend to push decisions all the way down to front-line
employees. In an intermediate form of CSR integration, each separate busi-
ness unit acts as a separate division in terms of deciding on CSR strategy;
however, these decisions primarily take place at the head of the respective
business unit. The degree of integration of CSR activities can then be con-
ceptualized as the procedure of the firms CSR activities.
On one end of the spectrum of integration, a company like Intel embeds
its CSR activities within and across all of its business units. It also incenti-
vizes all of its workers to engage in sustainability: Intel links a portion of
every employees variable compensation from front-line staff to our
CEO to environmental sustainability metrics. On the other end of the
integration spectrum is a company like Starbucks. In 1999, Starbucks
formed a stand-alone CSR department. This department is led by a Senior
Vice President who focuses on business practices, environmental issues,
community affairs, corporate giving, and the Starbucks Foundation. The
CSR unit operates similarly to any other important business unit of the
firm. Thus, whereas Intel has to a large degree integrated its CSR strategy
making throughout its organization, Starbucks has a separate division that
specializes in CSR strategy making.
A second dimension differentiating CSR units is the degree of profit
alignment in their CSR activities. Some types of CSR, such as some energy
costs savings measures, simply pay for themselves financially. Another
example is consumer-facing firms that engage in CSR and in response enjoy
an increase in customer sales that exceed such CSR costs. These types of
CSR would be appropriate in a Milton Friedman world of CSR, where a
managers sole responsibility is to maximize the profits of the firm
(Friedman, 1970). However, other types of CSR are not expected to gener-
ate a full financial return of its cost. One feature of this latter type of CSR
is that it often is not well related to the market strategy activities of the
firm. Another feature is that the CSR may be done in such a way that it is
difficult to recoup many of the costs. Thus, this second dimension can be
conceptualized as the purpose of the firms CSR activities.
418 DYLAN MINOR
On one extreme toward CSR profit alignment, there is Halliburton,

which although it appears in the 2013 Dow Jones Sustainability Index lists
financial Performance as the primary reason for its CSR efforts. In its
sustainability report, Halliburton further explains that its CSR issues
impact shareholder value and are, therefore, important to the company.
On the other end of the spectrum is Patagonia, which reorganized under a
Benefit Corporation charter to make legally explicit its objective of social
performance and that it is not maximizing shareholder value based solely
on financial outcomes.
From these two dimensions, we can now map CSR organizational forms
into four types. We denote a firm that is profit-centric in its CSR and has its
CSR strategic activities largely carried out by a specialized business unit as
Strategic. These firms are generally engaging in CSR primarily for profit and
are implementing their CSR activities by means of this stand-alone CSR
business unit. At the other extreme of both dimensions is the organizational
form we call Mission. Firms in this category are engaging in CSR from less
of a profit motive. An additional sign of this type of organizational form is
that the CSR activities are devised and engaged in at possibly all levels of
the firm. Most benefit corporations would also fit in this category.
In contrast, those firms that disperse their CSR strategic activities
broadly throughout the organization but are more profit-driven in their
CSR pursuits are categorized as Integrated. Finally, the opposite CSR
organizational form of Integrated is Foundation. These firms carry out
CSR activities in a more stand-alone-unit fashion and are not purely profit-
driven in their CSR.
In Fig. 1, we plot four firms according to this CSR organization typol-
ogy. Of course, in practice, there are degrees of integration and degrees
of profit-centeredness. Thus, if we plotted a variety of firms, they would
portray a distribution of outcomes. In this spirit, we randomly selected a
sample of 24 companies from the Dow Jones Sustainability Index. We then
hand collected data from each firms most recent sustainability report and
graded them on each dimension of our typology.1 In addition to these
DJSI companies, we added Patagonia as an example of a benefit corpora-
tion, and a far north-east reference point, creating a total sample of
25 firms. In Fig. 2, we report the results of this qualitative analysis. We
then divide the resultant space into quadrants to map these firms into our
typology introduced with Fig. 1.
It should also be noted that this chart represents firms stated organiza-
tional form. We are unable to test the actual organizational form of all of
these firms. However, there is some comfort in knowing that for those firms
Integrated Mission
Intel Patagonia
CSR is more integrated within the firm
Strategic Foundation
Halliburton Starbucks
CSR is less profit-centered
Fig. 1. Organizational Form Typology.
for which we interviewed top executives, we found similar proclamations

from the executives as those in their sustainability reports.
Although an exact plotting is admittedly subjective, it is apparent from
this chart that there is heterogeneity across firms. We next turn to identify-
ing the environments in which a firm should operate by means of a given
organizational form to illuminate the origins of this heterogeneity.
MODEL
To model the problem of organizational design, we assume that the firm
chooses integration or specialization for two units, or groups.2 Integration
means that both units will perform both market and nonmarket activities.
Specialization means that one unit handles only market activities and the
other handles nonmarket activities. To match the reality that much of the
nonmarket activity at firms that have a nonmarket-specific unit involves
corporate social responsibility activities, and to simplify exposition, we
refer to nonmarket activities in the model as CSR. For simplicity and to
abstract away from a team problem, we refer to each unit interchangeably
420 DYLAN MINOR
Integrated
10 Integrated Mission
Intel
Patagonia
HP
Procedure
Cisco
Gap
Autodesk Quest Diagnostic
Disney
5
ConAgra Food Northern Trust
Target
Strategic Foundation
3M
Ball Corp Whirlpool Starbucks
McDonalds
Rockwell Collins
General Mills
BNY Mellon H&R Block
Citigroup Hershey
Halliburton Xcel Energy Travelers
Specialized 10
0 More Profit-centric 5
Purpose Less Profit-centric
Fig. 2. Sample of Organizational Forms.
as a manager. In particular, we consider two risk-neutral managers,

1 and 2, each managing its respective unit.
The firm uses business units 1 and 2 to collectively implement a level of
CSR activity indexed by level S and a level of market activity indexed by
level M.3 Hence, we explicitly consider both market and nonmarket choices
without taking the other as given, as is often the case in the extant literature
(see Baron, 2016).
Total output Y is a function of both of these factors and is written as
Y M S 1
where M = M1 + M2, S = S1 + S2, and R captures CSR outputs

contribution to overall output.4 Thus, measures how important CSR is in
the firms production process. For example, a firm that faces consumers
who place a greater value on CSR, enjoys a greater . In contrast, when
= 0, CSR activities do not help the production process at all. We do not
consider < 0. This would mean that a firm is engaging in CSR that hurts
overall firm performance. A firm should not engage in this type of CSR.
Note that overall firm performance can also mean social performance in
addition to financial performance (see section ( )).
It is costly to implement activities, and this cost is a function of the level
of activity. In particular, a managers cost of production is
Mi2 S2
Ci i Mi Si 2
2 2
for manager i {1, 2}. The parameter R captures the degree of economy
of scope of a mangers production process. In particular, when 0 < < 1, as
commonly assumed (see Bolton & Dewatripont, 2005), increasing one level
of production increases the marginal cost of an additional unit of produc-
tion on both dimensions. This case can be thought of as a negative effort
spillover from one task to another; as one becomes exhausted from
one activity, increasing activity on either dimension is more costly. If
instead, < 0, increasing production on one dimension, reduces the cost of
increased production on the other dimension. For example, if engaging in
more CSR enabled a manager to produce market activities more cheaply,
then < 0. This could also be thought of as a learning spillover effect:
as one becomes better at CSR, one can more effectively generate market
output (and or vice versa).
Finally, if = 0, then there is no difference between having one worker
do two tasks or two workers each do one task. In other words, any link
between tasks is assumed away.
Manager is payoffs are
i Y Si 3
where i is the managers financial payment,5 which is calculated as share of

total output, and R is a managers valuation of personally producing
CSR. With this latter term, we allow for a manager to be intrinsically moti-
vated to produce some level of CSR. We will assume that i . This
means that managers receive no less financial benefit from market activities
422 DYLAN MINOR
than utility from social activities. Recall that our analysis is focusing on
profit-seeking firms and their workers. However, we will later consider
firms that care about CSR beyond its contribution to total output in sec-
tion ( ), as well as managers with high values of > i in section ( ). With
these basic ingredients we can identify the managers activity-level choices
as a function of organizational form.
Integration
Assuming the managers are engaged in both market and CSR activities in
their respective units, we obtain the following production levels:
Lemma 1. If manager i engages in both market and CSR activities,

outputs are Mi i 1 2 i and Si 12i for i{1, 2}.6
Proof. See Appendix A.

We then calculate the total market output as
M M1 M2

1 2 2 1 2 4

1 2
and social output as
S S1 S2

2 1 2 5
1 2
Thus, total output is

1 2 2 2 1 2 1 2
Yintegrated
1 2
6
Specialization
When managers only engage in one activity, we get the following outputs,
assuming manager 1 does M and manager 2 does S:
Lemma 2. If manager 1 engages in market activity and manager 2
engages in CSR activity, outputs are M1 = 1 and S2 = 2 + .

Thus, the firms total output is simply

Yspecialized 1 2 7
The next natural question is, when does a firm prefer specialization to inte-
gration? Fortunately, the analysis is simplified in that we do not need to
worry about comparing profitability or net output of different organiza-
tional forms but instead we can simply compare which form provides the
greatest output, as given by our next Lemma.
Lemma 3. If Yintegrated > (<) Yspecialized, integration (specialization) is the
optimal organizational form.

We now use this Lemma to identify when one organizational form is
preferred to another.
Optimal Organizational Form
The optimal organizational form then depends on the primitives of the

model. We will consider each primitive in turn in order to identify when
firms should choose one organizational form over another.
Economy of Scope
First consider the case as economy of scope approaches zero (i.e., 0).
Then, we have

2 1 2 1 2 > 1 2 8
424 DYLAN MINOR
This means that integrated generates greater total output than specialized.
Intuitively, when there is sufficient economies of scope across both activities,
having both units engage in both strategies dominates being specialized, since
two rather than one units generate similar levels of output on both dimen-
sions. Similarly, and more extreme, is when (1, 0). Here, Yintegrated is
further increased compared to when = 0, while Yspecialized does not change,
thus we still have Yintegrated > Yspecialized. That is, if a manager becomes more
effective in engaging in one activity because of engaging in the other, then
market and CSR activities should naturally be integrated within the organi-
zation. In contrast, as +, it can be shown that the inequality becomes
the opposite: specialized is the preferred organizational form. Intuitively,
when it becomes costly enough to engage in both activities simultaneously,
the firm is better off having specialized units; one business unit should specia-
lize in CSR activities and the other in market activities. We label this
increased cost as the multi-tasking cost. To explore the effects of the other
model parameters, we now consider, as commonly assumed (see Bolton &
Dewatripont, 2005), more moderate economy of scopes, such that 0 < < 1.
Importance of CSR for Output

Recall that measures how much CSR activities contribute to overall output.
First consider the case when CSR is not very important in the production
2
process (i.e., as 0). Then, we have Yintegrated = 1 1 2 2 and Yspecialized =
1. Since i and < 1, we have Yintegrated > Yspecialized, and the firm will
choose to have integrated units.7 When CSR matters little in total output, the
manager is also receiving very little incentive to participate in CSR since she
is paid a share of output; thus, she will choose relatively little CSR invest-
ment. This also means there will be little multi-tasking cost to offset the
advantage of having both units engage in both activities, yielding greater out-
put from units being integrated. In short, when CSR does not matter, it
makes little sense to have a separate unit engaging in CSR activities.
If instead CSR activities are very important in producing output (i.e., as
+), it is also the case that Yintegrated > Yspecialized. Here, the synergy
of CSR activities and market activities in generating output overcomes any
multi-tasking cost of having managers engage in both activities. In practice, this
means that when nonmarket activities are important for market performance, it
is critical to have the market and nonmarket strategies integrated locally, which
is best achieved by forcing business units to engage in both types of activities.
However, for intermediate values of and 8 it can be shown that
the opposite is true: Yintegrated < Yspecialized. Intuitively, there are two
forces that determine the overall output of a firm with an integrated
organizational form: the importance of CSR to overall output and the

cost of multi-tasking . When the multi-tasking cost of a manager enga-
ging in both CSR and market activities is sufficiently greater relative to
the benefit it provides to overall output, increased importance of CSR
(i.e., increased ) still induces the integrated firm to produce more CSR.
However, this greater CSR production is done with relatively poor effi-
ciency due to a higher , which yields a reduction in overall output for such
a firm. Eventually, however, CSR is valuable enough to production that
it adequately offsets the multi-tasking cost , which yields an increase
in overall output as the increased importance of CSR (i.e., increased )
induces greater firm production of it.9
Meanwhile, a specialized organizational form does not face such a
trade-off, as each unit only engages in only one activity, thus avoiding
a multi-tasking cost . Hence, Yspecialized is always increasing in .
Consequently, the relative advantage (or disadvantage) of the integrated
organizational form compared to the specialized organizational form is
then U-shaped in the importance of CSR the firms production process.
For an example of this U-shape of relative organizational form advantage
over values [0, 1], assume the following parameters: 1 = .25, 2 = .25,
= .2, and = .75. This yields Fig. 3, showing that intermediate values
of predict specialization (i.e., the gray region), whereas low and high values
of predict integration.
The x-axis of the graph represents [0, 1]. Given a level of , the
dashed line is the output from the organizational form of integrated,
Specialization
Output
Integration
Importance of CSR for Overall Output
0.2 0.4 0.6 0.8 1.0
Fig. 3. Integration and the Importance of CSR.

426 DYLAN MINOR
whereas the solid line is the output from the organizational form of
specialized. Whichever organizational form output is higher identifies the
preferred organizational form. Thus, for the gray region, which is approxi-
mately (.25, .85), specialized is preferred. Otherwise, integrated
provides superior total output.
Manager Valuation of CSR

Now consider increasing . We find that as , Yintegrated > Yspecialized if
and only if > 1 22 : Thus, if CSR is sufficiently important in the produc-
tion process, then given the manager cares enough about CSR, integration
is preferred. In this case, the firm and managers are aligned in producing
higher levels of both market and CSR activities. Otherwise, if CSR is not
as important, and given the manager cares enough about CSR, specializa-
tion is preferred. Intuitively, if a manager cares a lot about CSR, but CSR
is not that important for production, the firm is better of having that man-
ager produce CSR alone rather than having both managers simultaneously
engaging in CSR.
Firm Valuation of CSR

Now we consider the case where the firm values CSR beyond its contribu-
tion to output. In particular, assume that the firms objective is
to maximize
Y M S S
where is the firms additional valuation of social output over its value to
financial output. However, note that we can rewrite this as
~
Y M S
where ~ . Hence, our previous analysis of applies here as well.

That is, we witness a U-shaped relationship between the relative advantage
of integrated compared with specialized as a function of the firms valua-
tion of CSR.
Non-CSR Nonmarket Functions

Those firms with substantial nonmarket activities divorced from CSR strat-
egy can also be nested within our model and typology. These non-CSR
nonmarket activities generally include actions such as influencing legisla-
tion, influencing regulations, and litigation actions (see Baron, 2009). If we
assume a manager does not value their personal engagement in these activ-
ities outside of its value in increasing financial performance, we simply set
manager nonmarket production valuation = 0 Now consider what hap-
pens when = 0; this yields

1 2 1 2 1 2
Yintegrated
1 2
and
Yspecialized 1 2 2
As can be shown, the inequality Yintegrated > Yspecialized always holds with
= 0. In this world, under an integrated form, the manager provides the
exact division of labor between market and CSR activities as preferred by
the firm. Since the firm is receiving market activity level effort from two
managers rather than only one (i.e., when specialized), this double activity
more than offsets the cost of multi-tasking. Here, we are again assuming
moderate multi-tasking costs of 0 < < 1. Of course, as shown in section
( ), if is too great, the specialized organizational form will again domi-
nate, even with = 0.
In practice, nonmarket strategy often entails both CSR and non-CSR
nonmarket activities. Hence, for our model, we could redefine manager is
payoff as
~ i
i Y S
where ~ , with 0 < < 1. The term then captures the fraction of non-
market activities that are represented by CSR activities, for which manager
i values her personal production at rate . With this new notation, we can
reassess all of our previous analysis in terms of , ~ manager is effective
valuation of personal nonmarket activity production. For a given level of
nonmarket activity, those firms with a relatively high share of CSR activ-
ities (i.e., high ) and managers with high valuation of CSR activities (i.e.,
high ) are likely to be integrated, assuming sufficient returns to nonmarket
strategy given the firms economy of scope i:e:; > 1 22 , as given by our
previous analysis in section ( ).
428 DYLAN MINOR
Relating Theory to Organizational Form in Practice

In practice, as shown in Fig. 2, many (large) firms have some degree of inte-
grated organizational form rather than being fully integrated or fully spe-
cialized. We can conceive of firms that are partially integrated and partially
specialized as follows: consider a firm that has a combination of integrated
and specialized units. For example assume a firm has 10 units. Different
parts of the firm could have different features (e.g., different values of
and ) that dictate some units to be specialized and some to be integrated.
Perhaps, the marketing unit is fully integrated and the finance and opera-
tion units are specialized. On average, we would expect those firms that
have a greater total degree of firm-wide , to also have a greater degree of
firm-wide integration (i.e., have more integrated units), for example.
Hence, with this conceptualization, we can still make predictions about the
degree of integration as a function of firm-wide parameters.
From section ( ), we showed that the incidence of integrated organiza-
tional form is U-shaped in the degree that a firm cares about CSR beyond
any financial benefit that it might yield. This suggests that the degree of
integration should also be U-shaped in the degree a firm cares about CSR
beyond profits.
We can then relate the pattern of the data from section ( ) to our theore-
tical analysis. To do so, we fit a quadratic model to the 25 observations. In
particular, we regress the degree of integration (i.e., integer values 1
through 10) on a quadratic equation of the degree of being less profit-
centric with CSR (i.e., integer values 1 through 10). In Fig. 4, we show the
resulting estimated polynomial:
Degree of Integration
5
1
More Profit-centric Less Profit-centric
2 4 6 8 10
Fig. 4. Quadratic Fit to Sample.

With so few observations and the fact that the data was qualitatively
obtained, we must view this chart with care and merely consider it sugges-
tive, albeit consistent with our theory. In the end, the hope is that this
papers theoretical predictions will spur future rigorous, quantitative
exploration into the organization of nonmarket strategy.
Once a firms degree of integrated form is determined, its second dimen-
sion, the importance of CSR beyond profits, identifies its placement in one
of our four quadrants of our typology. From the analysis in the previous
section, we expect the firms where CSR has the greatest and least value to
be integrated and thus to be Mission and Integrated firms, respectively. In
contrast, the firms that are slightly more moderate in their valuation of
CSR are predicted to fall either into the Strategy or Foundation category,
depending on how important CSR is to them. Of course, if a firm faces
extremely great (or poor) economies of scope (i.e., low (high) multi-tasking
costs), it will necessarily be integrated (or specialized). For these cases, it is
also the firms value of CSR that determines its ultimate location in our
typology. Hence, assuming a reasonable proportion of firms with moderate
economies of scope, we expect in practice the location of firms in our typol-
ogy to trace through the four quadrants in a U-shaped manner, as found in
the suggestive chart above. The greater the proportion of firms with moder-
ate economies of scope, the greater the strength of the U-shape.
CONCLUSION
We created a typology for nonmarket organizational design. In particular,
we identified two dimensions of categorization: the degree that CSR activity
is integrated with market activities within a firm and the importance of CSR
strategy in a firms production process. We dubbed these resulting four
organizational types as Integrated, Strategic, Mission, and Foundation.
We then used a simple model to identify when firms are likely to choose
each of these organizational forms. Our primary finding is that the relative
advantage of firms with integrated over specialized organizational forms is
U-shaped in the importance of CSR: those firms for which CSR has either
a high or low level of importance whether it is because of the importance
of CSR strategy on market outcomes, a firms value of CSR beyond its
financial value, or a managers high valuation of CSR are most likely to
choose an integrated organizational form, combining market and CSR
activities throughout the firm. In contrast, those firms where CSR is
430 DYLAN MINOR
moderately important in the production process are more likely to choose

to specialize their CSR operations in a stand-alone business unit.
Empirical implications of these results are severalfold. First, once
proxies for the importance of CSR to the firm are identified (i.e., measures
of managerial valuation of CSR, measures of the degree of CSR comple-
mentaries with financial performance, and the preferences of the board of
directors and shareholders concerning CSR), the U-shaped relationship
between these and the probability of CSR integration can be assessed.
Second, using a proxy for the proportion of non-CSR nonmarket activities
(e.g., political contributions and lobbying) can be used to test an increase
of such a proxy being associated with reduced incidence of CSR specializa-
tion. Third, measures of the economy of scope of a firm should also predict
organizational form. At the extreme, strong (poor) economies of scope pre-
dict integration (specialization).
We note that our analysis did not formally explore more than two divi-
sions. Although we trust the intuition will carry over to n divisions, we
leave this analysis to future research. Similarly, we did not simultaneously
explore the joint effects of myriad forces, as is typical of the agent-based
simulation literature. This paper suggests future avenues of research by
means of this methodology: numerical simulations could determine organi-
zational outcomes based on the assumption that certain strategic activities
are correlated in terms of payoffs. This would be a departure from the stan-
dard approach of assuming i.i.d. draws of payoff profiles. However, it
could provide rich insights into the integration of market and nonmarket
strategy and its implications on the organization of nonmarket strategy.
NOTES
1. See the Appendix B for information on our method of categorization.

2. Note that Fig. 2 provides a qualitative example of firms with varying degrees
of integration and profit-centeredness. To simplify the analysis and make sharp
predictions, our theoretical analysis focuses on the decision to choose integrated or
specialized organizational forms for a given pair of units. That is, we are exploring
the world of Fig. 1. See section ( ) for a discussion of how to link the resultant
theory to the sample of firms in Fig. 2.
3. Of course, rather than a scalar choice, strategy activities are often multi-
dimensional and can be represented as a vector of choices. However, we could pro-
ject a scalar output variable onto multiple dimensions to capture complex activities.
Thus, M and S can be thought of as the output for a given multiple-dimension
choice, which in sum represents a strategy. We can then order these complex tasks
by their output levels, creating a one-to-one mapping between complex activities
and scalar outputs. Since we do not introduce noise into our model, the scalar out-
put choice then becomes the scalar input choice.
4. We could add noise to the production process but since business units are risk-
neutral, it would not change the results and would just add additional notation. If
we instead assume the business units are risk-averse, similar results to those pre-
sented obtain. However, notation and exposition are greatly complicated.
5. We consider the optimal organizational form holding compensation structure
fixed. This allows us to isolate organizational form effects from differential compen-
sation effects. We leave exploring optimal pay structures to future work.
6. Note that without additional assumptions, it is possible for Mi and Si to take
on negative values for a given set of parameters. However, given such parameters, it
simply means that Specialized, the alternative organizational form, is the preferred
one; this form always generates positive output, as shown in the next section.
7. The inequality Yintegrated > Yspecialized is true if and only if (2 + 12) > 2.
To see this is the case, note that (2 + 12) min (1 + 2) (1 + 2) > 2,
where min is the minimum of 1 and 2. The final inequality follows since (l + 2) >
2 for all l.
8. From the previous section, regardless of , we know that low and high values
of result in optimal organizational forms of integration and specialization, respec-
tively. Hence, we must consider intermediate values of to explore the effects of .
9. Formally, this can be seen by noting that Yintegrated is first decreasing and then

increasing in . In particular, we see that Yintegrated 1 2 2 1 2 .

Thus, Yintegrated > 0 if and only if + ( ) (1 + 2) > 0. If we fix sufficiently
greater than 0 (note that 12 1 2 > , since i ), then low values of generate

Yintegrated < 0 and higher values of generate Yintegrated > 0. In other words,
Yintegrated is U-shaped in .
ACKNOWLEDGMENTS
I Thank Bryan Hong and two anonymous referees for their helpful
comments. I also thank Sangeeta Khemani and Romain Sinclair for their
excellent research assistance.
REFERENCES
Alonso, R., Dessein, W., & Matouschek, N. (2008). When does coordination require centrali-
zation? The American Economic Review, 98(1), 145179.
Alonso, R., & Matouschek, N. (2008). Optimal delegation. The Review of Economic Studies,
75(1), 259293.
Barnett, M. L. (2007). Stakeholder influence capacity and the variability of financial returns to
corporate social responsibility. Academy of Management Review, 32(3), 794816.
Baron, D. P. (2009). Business and its environment (6th ed.). Englewood Cliffs, NJ:
Prentice Hall.
432 DYLAN MINOR
Baron, D. P. (2016). Strategy beyond markets: A step back and a look forward. In J. M. De
Figueiredo, M. Lenox, F. Oberholzer-Gee, & R. G. Vanden Bergh (Eds.), Strategy
beyond markets (Vol. 34, pp. 411434). Advancements in Strategic Management.
Bingley, UK: Emerald Group Publishing Limited.
Bolton, P., & Dewatripont, M. (2005). Contract theory. Boston, MA: MIT Press.
Business for Social Responsibility. (2002). Designing a CSR structure. Retrieved from www.bsr.org
Claussen, J., Kretschmer, T., & Stieglitz, N. (2014). Vertical scope, turbulence, and the benefits
of commitment and flexibility. Management Science, 61(4), 915929.
Coase, R. H. (1937). The nature of the firm. Economica, 4(16), 386405.
Friedman, M. (1970). The social responsibility of business is to increase its profits. New York
Times, September 13, pp. 3233, 122, 124, 126.
Garicano, L. (2000). Hierarchies and the organization of knowledge in production. Journal of
Political Economy, 108(5), 874904.
Godfrey, P. C. (2005). The relationship between corporate philanthropy and shareholder
wealth: A risk management perspective. Academy of Management Review,
30(4), 777798.
Grossman, S. J., & Hart, O. D. (1986). The costs and benefits of ownership: A theory of verti-
cal and lateral integration. The Journal of Political Economy, 94(4), 691719.
Harris, M., & Raviv, A. (2002). Organization design. Management Science, 48(7), 852865.
Holmstrom, B., & Milgrom, P. (1991). Multitask principal-agent analyses: Incentive contracts,
asset ownership, and job design. Journal of Law, Economics, & Organization, 7(Special
issue), 2452.
Hong, B., Li, Z., & Minor, D. (2015). Corporate governance and executive compensation for
corporate social responsibility. Journal of Business Ethics, 115.
Krishna, V., & Morgan, J. (2008). Contracting for information under imperfect commitment.
The RAND Journal of Economics, 39(4), 905925.
Laux, C. (2001). Limited-liability and incentive contracting with multiple projects. The RAND
Journal of Economics, 32(3), 514526.
Lyon, T. P. (2004). Corporate environmentalism and public policy. Cambridge: Cambridge
University Press.
Minor, D., & Morgan, J. (2011). CSR as reputation insurance: Primum non nocere. California
Prendergast, C. (2002). The tenuous trade-off between risk and incentives. Journal of Political
Economy, 110(5), 10711102.
Prendergast, C. J. (1995). A theory of responsibility in organizations. Journal of Labor
Economics, 13(3), 387400.
Rivkin, J. W., & Siggelkow, N. (2003). Balancing search and stability: Interdependencies
among elements of organizational design. Management Science, 49(3), 290311.
Schottner, A. (2008). Relational contracts, multitasking, and job design. Journal of Law,
Economics, and Organization, 24(1), 138162.
Siggelkow, N., & Rivkin, J. W. (2005). Speed and search: Designing organizations for turbu-
lence and complexity. Organization Science, 16(2), 101122.
Williamson, O. E. (1975). Markets and hierarchies. New York, NY: The Free Press.
APPENDIX A: PROOFS
Lemma 1. If manager i engages in both market and CSR activities, out-

puts are Mi i 1 2 i and Si 12i for i {1, 2}.
The managers problem is

2
max Mi S2i
Y Si Mi Si
M i ; Si i 2 2
Obtaining the first-order conditions yields the following:
FOCMi : i Mi Si 0
Mi i Si
FOCSi : i Si Mi 0
Si i Mi
Using the system of four resultant equations yields

1 1
M1
1 2
1
S1
1 2

2 2
M2
1 2
2
S2
1 2

Lemma 2. If manager 1 engages in market activity and manager 2
engages in CSR activity, outputs are M1 = 1 and S2 = 2 + .
Assume manager 1 does M and manager 2 does S.
Manager 1s problem is

max M2
iY 1
M1 2
434 DYLAN MINOR
which yields optimal choice of level
M1 1
Manager 2s problem is

max S2
2 Y S2 2
S2 2
and the managers CSR choice becomes
S2 2

Lemma 3. If Yintegrated > (<) Yspecialized, integration (specialization) is the
optimal organizational form.
By optimal organizational form, we mean the greatest output net of the
costs of creating such output (i.e., greatest net output). For the purely
profit-driven firm (i.e., the firm that does not care about CSR output
beyond its contribution to financial performance), net output is simply
financial profit. The net output derived from either nonmarket organiza-
tional form (i.e., integrated or specialized) is simply the share of output
after paying managers for their efforts. Thus, we write the firms net
output as a function of each organizational form as:

integrated 1 1 2
2 3
1 2 2 2 1 2 1 2
4 5
1 2

specialized 1 1 2 1 2
We then see that integrated > specialized if and only if Yintegrated > Yspecialized:
integrated > specialized

1 1 2
2 3
1 2 2 2 1 2 1 2
4 5
1 2

> 1 1 2 1 2

1 2 2 2 1 2 1 2

1 2

> 1 2
Yintegrated > Yspecialized

436 DYLAN MINOR
APPENDIX B: ORGANIZATIONAL FORM

TYPOLOGY CODING
To categorize our sample of 25 firms, we first studied if the firms procedure
for engaging in CSR was more driven by an integrated or specialized organi-
zational form approach. Specifically, using each firms 2013 Sustainability
report, we analyzed how CSR-related decisions were made and implemented.
Many reports also provided schematics on the flow of decision making and
implementation of CSR strategy. We further explored if strategy was gener-
ated from within a central CSR unit versus created autonomously within
separate business units, and if generated from a specialized unit, how much
influence outside units had on the specialized units activities. Based on this
qualitative review, we rated firms by assigning them an integer between 1
and 10, inclusive, where 10 was having CSR fully integrated throughout the
firm and 1 was fully specialized. Ranks were then compared to assess relative
differences across firms and verify the appropriate rating of each of the
25 firms. To do so, two research assistants (RAs) independently created a
rating for each firm, as well as a summary of their qualitative research to
justify each of their ratings. We then all met to discuss the ratings and ratio-
nales to arrive at a final rating for each firm.
To score firms on the dimension of how profit-centered their CSR activities
were, we followed the same process by having each RA score firms and pro-
vide rationale, as well as engage in a discussion. However, for this dimension
we explored the purpose of the firms CSR activities. If a CEO stated that CSR
was used because it increases shareholder value or makes economic sense, for
example, this generated a more profit-centric rating (i.e., a lower number) than
if the CEO instead said CSR is engaged in because it helps make the world a
better place. Overall, this purpose rating was formed primarily by reviewing
statements from top management that argued for why their company engages
in CSR activities. The more statements that linked profit and CSR, the more
profit centric their rating, and the more statements that linked CSR and non-
(direct)profit motives, the less profit centric their rating. Here, we gave the
most profit-centric a rating of 1 and the least profit-centric a rating of 10.
Other firms were assigned an integer in between these extremes.
In the end, the purpose of the qualitative coding exercise was to develop
a sense of the types of organizational approaches firms are currently taking
when engaging in CSR by providing a simple, simultaneous snapshot of
their efforts. From this analysis, it became clear that there is significant
heterogeneity on both dimensions of CSR purpose and procedure. This
finding motivated our theoretical analysis to help understand the origins of
these differences.
COMPLEMENTARITY IN FIRMS
MARKET AND POLITICAL
CAPABILITIES: AN INTEGRATED
THEORETICAL PERSPECTIVE
Nan Jia and Kyle Mayer
ABSTRACT
We examine how a firms market-oriented capabilities (in areas such

as R&D or marketing) and consumer focus (business-to-business or
business-to-consumer) foster its effectiveness in pursuing corporate political
activities. We then explore the sustainability of any advantage that firms
may gain from their political activities. We develop a conceptual frame-
work to propose that a firms political capabilities to implement different
political tactics, such as information provision and constituency building,
are a product of how related these tactics are to different market-
oriented capabilities and to the skills needed to serve different types of
customers. Finally, we propose that the integration of market strategies
and political strategies provides new insight into the sustainability of the
advantages that a firm might gain through political activities.
Keywords: Market capabilities; political capabilities; complementarity;
customer focus; political tactics; sustainable advantage

ISSN: 0742-3322/doi:10.1108/S0742-332220160000034013
437
438 NAN JIA AND KYLE MAYER
INTRODUCTION
Strategic management scholars have increasingly recognized that a firms
likelihood of succeeding in a particular market is a function of not only its
market-related resources and capabilities, such as patents, brands, and/or
its marketing or R&D capabilities (e.g., Barney, 1986, 1991; Peteraf, 1993;
Wernerfelt, 1984), but also its political capabilities to influence the rules of
the game in its favor through political strategies (e.g., Hillman & Hitt,
1999; Oliver & Holzinger, 2008). Governments drive the framework of leg-
islative and policy constraints within which market competition occurs,
thereby shaping the opportunities and threats that firms face in their busi-
ness operations (Baysinger, 1984; Ring, Bigley, DAunno, & Khanna,
2005). An emerging body of literature has started to investigate firms poli-
tical capabilities defined as the firms ability to effectively identify and
utilize relevant political tactics to achieve political goals (Fremeth &
Shaver, 2014; Holburn & Zelner, 2010).
While scholars have generally recognized the necessity for firms to simul-
taneously possess both strong market capabilities and strong political
capabilities in order to achieve higher performance (Li, Peng, Macaulay,
2013), these two types of capabilities have been largely treated as indepen-
dent assets, both of which firms need to develop. The linkage between
market capabilities and political capabilities, however, has rarely been
discussed. For example, the literature on firms political capabilities has
primarily focused on how prior political activities (in home countries or
different countries) enable future political activities (e.g., Garcia-Canal &
Guillen, 2008; Henisz & Delios, 2002; Henisz & Zelner, 2012; Holburn &
Zelner, 2010), implying that the foundation of political capabilities is the
prior experience in the same or other policy environments. Moreover, some
hold the view that a substitutive relationship exists between market capabil-
ities and political capabilities, arguing that political strategy may shield a
firm from market competition and thus reduce the firms incentives to
invest in certain market strategies, thereby creating a substitution effect
(e.g., Lenway, Morck, & Yeung, 1996; Morck, Sepanski, & Yeung, 2001).
Such a strategy, however, is highly dependent on whether the government
limits market competition over the long term in specific ways that favor the
focal firm(s), which may be difficult to sustain.
By contrast, this paper seeks to advance our understanding of this issue
by highlighting that a firms market capabilities may influence what politi-
cal actions it undertakes and how effectively the actions are executed.
Using capabilities and resources that are central to the firms competitive
Complementarity in Firms Market and Political Capabilities 439
advantage in the market and applying them to political tactics to achieve cer-
tain policy ends creates a complementary relationship between market strat-
egy and political strategy. In this way, firms can develop political capabilities,
a process that is typically facilitated by drawing on the resources and capabil-
ities that the firm has developed in competing with rivals in the market. We
develop a conceptual framework to generate testable propositions regarding
how market capabilities (in areas such as marketing or R&D) and customer
focus (i.e., business-to-business or business-to-consumers) can lead to the
development of political capabilities and facilitate the use of particular politi-
cal tactics of information provision and constituency building. In addition,
we highlight the value of both internal and external alignments as key drivers
of the sustainability of advantages from political activities.
Many interesting phenomena of how firms pursue corporate political
activities motivated us to link a firms capabilities developed in markets to
its corporate political activities. While many companies are politically
active, some seem to be particularly good at or more prone to use certain
political tactics than others. For example, many internet-based high-tech
firms in the Silicon Valley appeared particularly skillful at motivating
grass-root support and organizing political campaigns than firms in tradi-
tional media and telecom industries with whom they battle over issues such
as antipiracy and network neutrality,1 but the formal lobbying efforts of
the internet-based firms have lagged behind traditional industries, and
some large high-tech firms, such as Google, Facebook, and Apple, are still
learning to play the lobbying game in the Capitol.2 While firms do learn to
use different political tools even those that they are not initially familiar
with over time, as evidenced in those in more traditional industries that
have longer history of dealing with politics, the possibility that some skills
learnt by the firms in running businesses in market settings may facilitate
the utilization and success of certain tactics in political settings (i.e., politi-
cal tactics) is quite interesting but under-explored.
In order to examine how firms market capabilities influence the effica-
cious use of political capabilities, our theorizing combines elements from
research on firms political capabilities with the knowledge developed by
the rich body of literature on the resource-based view (RBV) of the firm.
The RBV focuses on how firms utilize resources and capabilities to build
sustainable competitive advantage (e.g., Barney, 1986, 1991; Peteraf, 1993).
Work on the RBV, however, has primarily focused on market competition
and typically regards the institutional environment as exogenous while
examining how firms can outperform other firms in terms of profitability,
market share, and growth. While research on political capabilities is very
consistent with the underlying premise of the RBV that firms must build
unique and difficult-to-imitate resources and capabilities to overcome com-
petition, to date, there has been relatively little work on political capabil-
ities from the RBV perspective. We seek to address this important gap in
the literature because of the importance of the institutional environment to
the value of a firms market resources and capabilities; for example, changes
in the law can influence the value of patents, trademarks, and brand names,
as well as technical capabilities associated with the development of research
and development (R&D). Thus, firms often must learn to develop political
capabilities, which we argue may flow from their market capabilities and
resources, to protect their profit stream. We further argue that seeking to
develop complementary political strategies and market strategies is likely to
provide a more substantial advantage by helping firms overcome competi-
tion while influencing the rules of the game in their favor.
This study provides an important extension of prior research on corpo-
rate political strategy, which has focused on how other corporate character-
istics, such as financial resources, financial slack, and firm size, influence a
firms use of various political tactics (e.g., Hillman & Hitt, 1999; Lenway &
Rehbein, 1991; Schuler, 1996; Schuler, Rehbein, & Cramer, 2002). By inte-
grating ideas from the RBV, we seek to highlight the key role played by a
firms existing portfolio of market resources and market capabilities in its
attempts to create political advantage.
We begin by investigating the capabilities needed to effectively imple-
ment political tactics, based on how relevant market capabilities and custo-
mer focus are for developing political capabilities. We then discuss how the
internal and external consistency of the firms political strategy influences
the sustainability of the advantages gained from using political tactics.
POLITICAL TACTICS AND MARKET CAPABILITIES

The central question that we address in this paper is which market capabil-
ities cause firms to be more or less likely to pursue, and more effective in
pursuing, different types of political strategies. A central element contribut-
ing to a firms effectiveness in pursuing political strategies is its ability to
implement the appropriate political tactics. To achieve a particular political
goal, firms must utilize political tactics, such as lobbying and campaign
contributions, to influence political decisions and public policy. In this
section, we examine two major political tactics that have been the subject
of extensive prior research: information provision and constituency build-

ing. We argue that successfully utilizing these tactics requires particular
capabilities. We then examine how the political capabilities to use certain
political tactics can originate from or be enhanced by various market cap-
abilities. Therefore, a firms market capabilities influence how effective the
firm is in implementing different political tactics.
Common Political Tactics and Underlying Political Capabilities
In seeking to use political strategy to achieve private goals, firms may

deploy a broad range of specific political activities, such as lobbying by
providing information, testifying at government hearings, offering
campaign contributions, and forming political action committees (PACs).
Firms can also extend personal services, such as political network building,
appoint officials as board members, form collective organizations, directly
participate in political processes, and engage in constituency building
(e.g., Grier, Munger, & Roberts, 1994; Hillman & Hitt, 1999; Hillman,
Zardkoohi, & Bierman, 1999; Oliver & Holzinger, 2008). Researchers
typically divide common political activities into three general categories:
information provision, constituency building, and financial contributions
(e.g., Hillman & Hitt, 1999).
Hillman and Hitt (1999) provide detailed definitions of the three generic
tactics. Information provision occurs when firms seek to influence political
decision making by providing information that is relevant to political actors,
such as regulators and legislators. For example, firms may lobby by supply-
ing position papers or technical reports, providing research findings and
survey results on the subject matter, or testifying as expert witnesses.
Constituency building occurs when firms attempt to shape public policies by
obtaining support from individuals and constituents whose policy prefer-
ences and votes matter to politicians. For example, firms may mobilize
grass roots such as consumers, employees, shareholders, and commu-
nities, run advocacy advertising and political education programs, or engage
in public relations (PR) building through, for example, press conferences.
Thus, information provision is used to directly influence policy makers,
whereas constituency building is used to motivate policy makers indirectly
by having their constituents pressure them to take a particular action.
To effectively deploy these political tactics, firms must have certain poli-
tical capabilities. Political capability refers to the firms ability to know
when and how to use particular political tactics to achieve a specific
political outcome. For example, the ability to identify and engage key poli-
tical actors has been found to contribute to a firms success in handling
political risks and obtaining favorable policy outcomes (Holburn &
Vanden Bergh, 2002, 2008; Holburn & Zelner, 2010). In addition, a firms
ability to identify different types of political issues and to take different
actions at varying stages as political issues develop is a critical factor
influencing the firms gains or losses in the political arena (Bonardi &
Keim, 2005).
We begin by focusing on the specific political capabilities that are needed
to most effectively implement information provision and constituency
building. We first discuss information provision. We make an important
note that despite what the name suggests, information provision does not
merely require an ability to effectively transfer information; instead, it
requires a much richer set of capabilities and skills. Influencing public pol-
icy through information provision demands the use of individual persuasion,
that is, the use of deep knowledge of technology and context to persuade
individual politicians or regulators that it is advisable to (or not to) take a
particular course of action this is essentially a fact-based mechanism to
influence the target audience. To effectively use information provision,
firms should have the ability to identify what information is most impor-
tant to target users (i.e., policy makers), to conduct relevant research that
can generate critical information that is needed to convince the target audi-
ence, to have a deep understanding of the technology and context, to effec-
tively convey key elements of this information to the target audience, and
to have the flexibility to respond to requests and questions that are raised
from various perspectives.
Constituency building essentially aims at mass persuasion, which requires
the organizational ability and skills to motivate many people and constituent
organizations to pressure politicians for particular outcomes. Constituency
building goes beyond providing information to the target large-scale audience
to sway their rational thought processes but, more importantly, often requires
building connections with the audience and appeal to them at a personal or
emotional level this is essentially an emotion-based mechanism to influence
the target audience. This individual appeal helps firms win popular support,
which requires a different set of skills and capabilities than persuading an
individual (or a small group of) politicians through information provision.
Specifically, the firm must be able to connect with individual society members
and groups to form stakeholder coalitions, to advertise its preferred policy
position to align its interests with those of the broader group so that it
can win popular support, and to engage the media to build positive public
opinion. That opinion, to a large extent, depends on firms ability to appeal

to the hearts, or the emotional side, rather than to the heads, or the
intellectual side, of the target audience with comprehensive and sophisti-
cated information.3 For example, the U.S. public accounting profession
effectively engaged in public relation campaigns and advocacy advertising,
including running advertisements in major newspapers, to inform the
general public and certified public accountants (CPAs) of the liability issues
facing the profession in order to push for liability reform legislation
(Roberts, Dwyer, & Sweeney, 2003).
Building Political Capabilities from Market Capabilities
The political capabilities that are needed to successfully utilize a political

tactic may be gained in multiple ways. Two main mechanisms by which
firms may develop political capabilities in using political tactics include
(1) prior experience using a particular political tactic to equip the firm with
the skills needed to use the same tactic in the future, and (2) market
capabilities to generate the underlying abilities that can be applied in the
political arena to assist in the execution of relevant political tactics.
Research has found that prior political experience in using political stra-
tegies increases the effectiveness of using such strategies in the future
(e.g., Bonardi, Holburn, & Vanden Bergh, 2006; Frynas, Mellahi, &
Pigman, 2006) because firms learn to better use political strategies from their
own experience (Holburn & Zelner, 2010). Therefore, developing the politi-
cal capability to utilize a particular political tactic is similar to building a
market capability, as described in research on organizational learning and
the RBV (e.g., Leiblein & Miller, 2003). The more frequently a firm engages
in a particular activity, the better it tends to become at performing that
activity. The value of direct experience has been established in a wide variety
of fields (e.g., Cyert & March, 1963; Delios & Henisz, 2003; Holburn &
Zelner, 2010; Levitt & March, 1998) and typically forms the basis on which
RBV scholars determine which firms have stronger capabilities. That is,
firms with more experience in a particular activity are thought to have
greater capabilities in that area. Therefore, firms develop capabilities in
using a particular political tactic as they gain direct experience in using
that tactic.
While experiential learning clearly occurs in the context of political strat-
egy, we focus on the alternative means by which firms develop their politi-
cal capabilities the influence of the firms market capabilities and market
resources, which we argue offer greater potential to provide a sustainable

competitive advantage because imitating a firms market capabilities and
market resources is more difficult than simply gaining experience with a
particular tactic. A firms capabilities and skills that have accumulated
through its market operations are likely to influence its political capabilities
to utilize specific political tactics. Building on the firms existing market
capabilities not only leads to more advanced political capabilities but also
enables the firm to build such capabilities more quickly and potentially
more cost effectively than if it were starting without a relevant underlying
skill set from the related market capability.
Building political capabilities from firms market capabilities has another
advantage. Political activities and market activities compete for limited
internal resources (Bonardi, 2008); thus, if a firm can build on existing mar-
ket capabilities to develop a political capability, as compared to developing
an unrelated political capability, the relatedness of the market capability
with the political capability decreases the cost of developing the political
capability, as fewer resources are necessary for the task. This in turn allevi-
ates concerns about resource constraints. Therefore, building political cap-
abilities that draw upon existing market capabilities enables firms not only
to build stronger political capabilities but also to develop these capabilities
in a more cost effective manner than if it had to develop an unfamiliar
underlying knowledge base.
While we investigate how market capabilities generate the underlying
abilities that can be applied in the execution of relevant political tactics, we
do not assume that policy makers have exactly the same needs as a firms
customers. For example, policy makers value attributes such as fiscal
revenue, local employment, and investment, which are different from the
quality and price of products and services that customers value. While pol-
icy makers and customers may value different things, the key is the similar
skills and capabilities that are used to persuade them to adopt a particular
course of action.
Information Provision
The capabilities that are needed for effective information lobbying may be
different depending on the type of information that is provided. We argue
that the capability to provide information through lobbying about techni-
cal aspects of the business is directly related to a firms technical capabil-
ities associated with the development of R&D. Successful development of
R&D typically calls for firms to know how to conduct research to generate
the information and technologies needed to meet their customers need.
Moreover, it also calls for the skills of translating their technical knowledge
into the information that their customers can understand and appreciate.
For example, firms with strong technical capabilities typically rely on
experienced product managers to bridge the gap between technical knowl-
edge and customers needs, which enables the firms to take full advantage
of their strength in R&D.
Strong technical abilities associated with R&D experience make infor-
mation lobbying, particularly providing information on the technical
aspects of the firms business, a more natural option to meet the firms
political needs, for the following reasons. First, the aforementioned
experience associated with successfully developing R&D gives the firms
deep knowledge of the technologies and the contexts that are integral to
building strong capabilities in information provision. For example, firms
with greater technological capabilities tend to be more experienced in
conducting or commissioning the research projects that are needed to gen-
erate results that are relevant to policy making and in generating technical
reports to influence policy makers. Second, the experience of successfully
generating and applying technical knowledge to meet clients needs tea-
ches these firms to not simply put an engineer in a room with a politician
(as they would not do to a client); rather, they have the engineer work
with the personnel who are more skilled in communication to help bridge
the gap between the knowledge of the engineer and the knowledge of
the politician (as their product managers do with a client). Finally, exten-
sive experience and strong technological capabilities also shape the firms
perspectives on political issues, because prior experience and focus may
have a cognitive imprinting effect that shapes the firms analysis of future
situations (e.g., Holburn & Zelner, 2010). Therefore, we generate the first
proposition.
Proposition 1. Firms with greater technical capabilities associated with
research and development (R&D) tend to have stronger political cap-
abilities in utilizing information provision related to policies that would
affect the technical aspects of their business than firms with weaker
technological capabilities.
While greater technological capabilities enable firms to more effectively
target political decision makers by generating highly relevant and in-depth
information related to the technical aspects of their business, the effective-
ness of utilizing these political tactics will be further enhanced if the firm
can also communicate technical information to the target audience in the
most accessible way.
First, if a firm has strong capabilities in marketing, this is a strong signal

that it is capable of explaining the products that are created by its engineers
to its customers. For example, Intel has complemented its strong R&D cap-
abilities with well-developed marketing and sales capabilities. Pharmaceutical
firms have not only strong R&D capabilities but also strong capabilities in
distribution and sales (i.e., relationships with doctors and healthcare organi-
zations) that allow them to take full advantage of their technological
strengths. The combined capabilities of these firms, that is, the strong techno-
logical capabilities with the complementary skills that are needed to success-
fully translate the technical skills into a competitive advantage, will facilitate
the use of information provision to influence political outcomes.
Second, firms with a competitive advantage related to marketing are
more likely to be experienced in identifying the needs of customers, in
designing a compelling product to fit those needs, and in determining the
business impact of technological innovations. This further helps to effec-
tively communicate and pitch the technologies to the target policy audience.
Third, effective communication becomes particularly important as some
political decision makers whom firms target by utilizing information provi-
sion have less technological expertise and are less able to handle technical
information than others (Hillman & Keim, 1995). When political decision
makers are less familiar with the technical aspects of a firms business, an
ability to clearly communicate with the target audience when reporting
commissioned research results, testifying as an expert witness, and present-
ing position papers and technical reports will strengthen the effectiveness of
the firms information lobbying.
Finally, a firms experience in marketing provides it with greater exper-
tise to effectively communicate with its customers, a process that is not a
one-way provision of information from the firm to its customers. Instead,
firms need to collect inputs from customers and to reiterate modifications
to prototypical products and services based on feedback from representative
samples of customers (e.g., Kotler & Keller, 2008); in recent years, with
the increased prevalence of social media and online tools, user-generated
content and direct engagement with consumers has greatly enhanced two-way
communication between firms and customers in marketing (e.g., Scott, 2010).
These underlying capabilities enable a firm to more effectively communicate
with its target audience in the political arena, by responding to questions and
requests raised by policy makers in an iterative process, which helps to
further convince the policy makers. Therefore, a stronger capability in
marketing, coupled with stronger technological capabilities, will further
enhance the effectiveness of utilizing information provision related to policies

that affect technical aspects of the firms business.
Proposition 2. The positive relationship between firms technological cap-
abilities and their political capabilities in utilizing information provision
related to policies that would affect technical aspects of their businesses
is stronger when firms have greater marketing capabilities.
Some information lobbying may focus on the social and political impact
of policies on firms and on other stakeholders. Instead of relating the criti-
cal technical information that is relevant to policy makers, under some
circumstances, lobbying firms may have to focus on the social and political
aspects of their business and public policies, such as arguments centered
on how a firms success in a local district may affect jobs and local eco-
nomic growth. The capabilities and skills underlying a firms effective use
of information provision on the social and political aspects of its business
can be driven by the firms extensive experience in business-to-business
(B2B) marketing.
Although marketing capabilities, in general, enable firms to more effec-
tively communicate with their customers to influence customer purchasing
behavior and decisions, different skills matter differentially to effectively
convince different types of target audiences. For example, some firms with
capabilities in marketing directly to consumers understand the psychology
of purchase decisions and how to use a variety of tactics (e.g., how to play
on emotions and build a bond between the consumer and the product) to
influence customers behavior (for a review, see Rucker & Petty, 2004).
Other firms often need to target large corporate clients and have thus devel-
oped the ability to persuade a particular powerful entity to make a pur-
chase or take a course of action in B2B settings. We argue that the logic
behind the link between strong marketing capabilities and the use of infor-
mation provision regarding the social and political aspects of firms busi-
ness further rests on the skills that firms develop in persuading corporate
clients in B2B marketing, based on the following reasoning.
Firms that focus on B2B transactions are more likely to be accustomed
to communicating with (and convincing) business customers by presenting
information to suit their customers specific business contexts. It is critical
to understand that B2B firms extensive experience with selling to industrial
buyers typically involves a data-driven sales process in which firms need
not only to communicate the technical information of their products and
services, but also to consider the social and political factors that affect their
corporate clients purchasing decisions. A corporate purchasing decision is

rarely just about the product. Instead, a variety of other sociopolitical
factors can influence corporate purchasing decisions; these factors include
disputes over who has the authority to make the decision, hidden agendas
about what is actually required, and a requirement that a firm must con-
vince the decision maker that the decision will not have negative conse-
quences in the long term. As an example, in the early to mid-1980s, Apple
computers were more powerful and easier to use than those made by IBM,
but corporations tended to stick with IBM not because of the computers
performance reasons but because of risk aversion. There was an old saying
in the information technology world no one ever got fired for buying an
IBM. There was simply too much risk in buying then-unknown Apple
computers, despite the superiority of the product.
Firms that sell to other corporations typically become skilled at addres-
sing these issues. These skills in marketing are based on knowing how to
persuade a decision maker who represents an organization to make a sig-
nificant purchase decision after considering various technical, social, and
political factors within the organization. These skills have much in common
with firms ability to gain support from the key political decision makers
whom they must convince of the social and political aspects of public
policies and their businesses. Persuading a potential customer to buy a pro-
duct is similar to persuading a politician to support (or oppose) a particular
policy. Although customers and policy makers demand different things and
may have different decision criteria, an ability to understand a decision
makers needs and constraints, and a plan for how to address them in a
way that creates value for the decision maker is a capability that can gener-
alize beyond marketing to influencing policy makers in the political arena.
For example, two U.S. firms bid on a contract to build a new refueling tan-
ker for the U.S. government: Boeing and Northrop Grumman. Part of
Boeings successful strategy in securing the contract was to position the
Northrop Grumman bid as being spearheaded by EADS, a European
entity, creating the appearance of a U.S. versus foreign bid for a U.S.
Defense Department contract. This strategy was implemented even though
Northrop Grumman would have been the lead contractor and Boeing also
had foreign partners as part of its bid. Boeing was also able to get gover-
nors of states where it had facilities to lobby the U.S. government on its
behalf by convincing them that the contract would result in additional jobs
for their states. The marketing capabilities that Boeing gained in part from
selling commercial aircraft to airlines around the world served them well in
persuading a large political entity to decide in their favor.4
Along the same vein, when Boeing selected suppliers and subcontractors
for the Dreamliner (787), one criterion was whether the country that the
supplier was from had a large commercial aviation business, making it
more likely to purchase the final product. Having local suppliers is a
good selling point when dealing with airlines in many countries. While
these examples involve purchase decisions, the same issues hold for policy
decisions related to issues such as energy, import constraints, and patents
(e.g., Baron, 2012).
Therefore, the underlying capabilities and skills of marketing to a speci-
fic powerful entity developed through B2B marketing will facilitate a firms
ability to effectively lobby to a few key political decisions makers with
respect to the social and political impacts of public policies.
Proposition 3. Firms with greater B2B marketing capabilities tend to
have stronger political capabilities in utilizing information provision
related to policies affecting social or political aspects of their business
than firms with less expertise in B2B marketing.
Constituency Building
Building constituency is another popular and important political tactic that
firms can employ to gain political influence (Walker, 2014). Constituency
building requires firms to call upon various stakeholders, such as employ-
ees, suppliers, industry associations, community leaders, and individual
shareholders, to influence public policy decision makers (e.g., Aplin &
Hegarty, 1980; Baysinger, Keim, & Zeithaml, 1985). To effectively obtain
the support of an informed and motivated constituency, firms must have
the political capabilities to identify appropriate target groups that may be
most affected by the specific public policies that the firm supports or
opposes, to communicate with members of the constituency, to motivate
them to support the firms positions, and to educate constituents regarding
how to increase their direct contact with elected officials (Baysinger et al.,
1985; Keim, 1981).5
These political capabilities share many common traits with a firms
ability to market to individual consumers (B2C). Firms that focus on B2C
marketing have extensive experience in communicating with individual
consumers through advertising and have accumulated significant knowl-
edge about how to motivate consumers to take certain actions, both of
which are necessary for gaining popular support. Although constituency
building entails mobilizing a wider range of stakeholders than just the
firms customers customers are sometimes, but not always deemed an
effective interest group to mobilize in the political process it is the under-

lying capabilities of psychological motivation and persuasion that are
developed through interacting with customers in the market, rather than
the direct contacts with a specific customer base, that facilitate a firms
ability to effectively execute constituency building.
First, successful B2C marketers have extensive experience in identifying
the right customers, and this ability may help firms identify the most effec-
tive constituents. Not all interest groups are ideal candidates to help a firm
exert political influence on political decision makers (Keim, 1981), and it
would be costly and less rewarding for the firm to target the wrong stake-
holders who have limited political incentives and capacities. Therefore, the
ability to identify the right constituency greatly enhances the effectiveness
of a firms political efforts in constituency building.
Moreover, firms with capabilities in B2C marketing not only have exten-
sive experience in communicating with individual consumers through var-
ious channels, such as advertising, but also excel at making emotional
appeals to influence the decisions and attitudes of a vast and diverse audi-
ence. Experienced B2C marketers deeply understand the psychology of
consumer purchasing decisions and how to use a variety of tactics (e.g.,
play on emotions, build a bond between the consumer and the product) to
influence customers behavior. Firms abilities in this regard facilitate the
key step in constituency building: appealing to the members of the constitu-
ency and educating them regarding how they may be affected by a public
policy and why their support for the firms stance is important. In particu-
lar, an emphasis on understanding and influencing the psychology of mem-
bers of the constituency enhances the effectiveness of constituency building,
as many of the constituents are also individuals, such as employees and
individual shareholders (Baysinger et al., 1985).
Furthermore, B2C marketing capabilities may also help firms to educate
and mobilize constituents to take actions that exert political pressure on
elected officials. Firms with stronger marketing capabilities have accumu-
lated extensive knowledge about how to motivate consumers to take certain
actions. This experience may translate into a firms ability to mobilize a
political stronghold consisting of effective interest groups and stakeholders
and to thus obtain voter support, which is highly valued by politicians.
Finally, extensive experience and capabilities in B2C marketing activities
and market campaigns may shape the firms understanding of possible
solutions to current political issues and may induce a firm to consider
constituency building to be an effective political tactic. For example, the
internet-based high-tech firms, such as Facebook, Wikipedia, and Craiglist,
are particularly successful at motivating and utilizing grass-root support, in

terms of winning the battle for public opinion, using their websites and
communication platforms to launch campaigns, urging consumers to lobby
their representatives over the proposals.6 This preference over motivating
grass-root support seems consistent with these firms strong customer
orientation and adroit maneuver of the social media in their businesses
(e.g., Walker, 2014).
We do not suggest that firms mainly motivate and use their own custo-
mers in constituency building, but that firms with strong capabilities in
B2C marketing know how to influence individuals and that they can use
these capabilities to mount effective constituency-building campaigns that
are not simply focused on their own customer base. It is the underlying
capability of knowing how to use psychology to influence individual beha-
vior and motivate actions (usually purchase decisions but also political
actions) that results from having strong B2C marketing capabilities. This is
consistent with the observations in public affairs consultancy that firms of
well-known household brands are the most active in conducting grass-root
campaigns whereas firms that supply business buyers and are less known to
the vast individual customers are rarely observed to employ the constitu-
ency building tactics (Walker, 2014). Therefore, we propose the following
proposition.
Proposition 4. Firms with greater B2C marketing capabilities tend to
have stronger political capabilities in utilizing constituency building to
influence political decision making than firms with less expertise in
B2C marketing.
SUSTAINABILITY OF POLITICAL ADVANTAGE

Internal Alignment
In addition to creating advantages for firms through political strategies,

researchers are taking the next step to examine the sustainability of these
advantages. The concept of sustainability of political advantages refers to
whether the political advantages that are gained by a firm are long lasting
or only temporary. A key factor in the political market that could diminish
the political advantages that a firm has gained is competition (or retalia-
tion) from other firms (Capron & Chatain, 2008). Although this research
area is still in its infancy, we build on one of the foundational articles in
this area, Capron and Chatain (2008), who examine the situations in which
the advantages gained by the focal firms attacking competitors in the
factor market and the political markets may be less sustainable. In this
section, we examine how the fit between a firms market capabilities and its
political capabilities influences the sustainability of advantages gained
through the use of political tactics.
Much has been written about the sustainability of competitive advantage
within markets (e.g., Ghemawat, 2009). In the political arena, the sustain-
ability of advantages gained from using particular political tactics can be
addressed by considering how difficult it is to develop the focal firms poli-
tical capabilities. If political capabilities are easy to develop, then we would
expect the advantages gained from them to be short lived. However, if dif-
ferences exist in political capabilities between various firms, then we predict
that there would be significant differences in competitive advantage across
firms and potentially more enduring competitive advantages among firms
with greater political capabilities. The issue then becomes determining
what contributes to the development of greater political capabilities that
are difficult to imitate.
We propose that the alignment between a firms market capabilities and
the political tactics for which it develops expertise is an important element
in the sustainability of political advantage. Specifically, we argue that when
a firm develops political capabilities that are aligned with its market cap-
abilities (i.e., the two kinds of capabilities are built on the same knowledge
base or underlying skill set), the political advantages gained through these
capabilities will be more sustainable because it will be more difficult for
rivals to imitate the focal firms political capabilities.
Imitation is a key driver that threatens the sustainability of the gains
arising from a firms political strategies. Capron and Chatain (2008) argue
that when a firms actions in the political market are easy to imitate,
the economic rents gained through these actions are less sustainable.
Because we have stressed that the effective utilization of two key political
tactics (i.e., information provision and constituency building) critically
depends on a firms relevant political capabilities, we argue that the diffi-
culty of imitating such political capabilities lies at the heart of whether
imitation is likely to occur. For example, for a firm that has successfully
used constituency building, a rival firm that lacks the capability to mobilize
consumers en masse poses a relatively weak threat to the sustainability of
any advantage that the first firm gained through political means.
Firms that try to build political capabilities can do so by increasing their
experience in using particular political tactics; however, their rate of
learning is likely to be slower if their market capabilities do not benefit

from using that particular tactic, than if their market capabilities can facili-
tate building political capabilities. That is, if the firms market capabilities
require the same types of underlying skills that are required to successfully
employ the political tactic being used to achieve a particular outcome, then
the firms-related experience in developing the market capabilities is likely
to lead to the development of greater political capabilities compared to
firms that seek to develop the same political capabilities without the com-
plementary market capabilities.
When market capabilities and political capabilities are integrated
(i.e., closely support each other), this complementary relationship leads to
the development of greater political capabilities. In turn, when a firm has
greater political capabilities, it is much more difficult for rivals to achieve
different, and possibly conflicting, political goals because the level of politi-
cal capabilities will play a key role in determining the policy outcome.
Proposition 5. When a firm uses a political tactic that draws on the same
underlying skills as its market capabilities (i.e., the internal alignment
between market capabilities and political capabilities), the resulting
advantage gained through political means is more sustainable than if the
political tactic and the firms market capabilities are not aligned.
External Alignment
Research in competitive strategy argues that the fit between external envir-
onmental factors, such as industry structures, and a firms strategy and
strategic positioning influences the sustainability of the firms competitive
advantage (e.g., Ghemawat, 1991; Porter, 1980). Following this line of
thought, the fit between the political capabilities that a firm develops in
using particular political tactics and the relevance of those tactics to the
policies that the firm seeks to influence will enhance the sustainability of
the advantages gained from using the political tactic. As with internal
alignment between market capabilities and political capabilities, external
alignment between the firms political capabilities and the tactics best suited
to achieve the firms political goals enhances sustainability.
In some cases, there are policy outcomes that firms could pursue equally
effectively by using different political tactics; in such a situation, there is no
real external alignment because any tactic, if used effectively, may be suc-
cessful at achieving the desired policy outcome. However, not all political
tactics are equally effective tools for addressing a certain political issue.
For example, when facing widespread denunciation of their practices before
and during the financial crisis, many major banks sought to seal their infor-
mation by lobbying behind closed doors, holding private meetings with
politicians to try to sway strengthened regulations because the banks
unpopularity would have made it impossible, if not detrimental, for them
to use constituency building to garner public and widespread voter support
for their political goals.7 When the political issues are difficult to be commu-
nicated to the public or framed in a way to appeal to the public, corpora-
tions tend to conduct quiet politics behind closed doors with policy
makers without involving constituents (Culpepper, 2010).
Institutional contexts may also render certain policy outcomes to be
more effectively achieved through one type of political tactics than others.
Considering the value of internal alignment in two distinct situations in
which one political tactic may be more likely to achieve a desired policy
outcome will better illustrate our point in this regard. Consider the case in
which the focal firms internal alignment leads to a strong capability in con-
stituency building that has helped it achieve a particular advantage via poli-
tical means. In Country A, by virtue of the issue and the motives of policy
makers, constituency building is the tactic that is best suited for influencing
policy outcomes, and neither information provision nor financial contribu-
tion is likely to have as great of an impact on policy related to this issue
(such environmental conditions are often specific to certain issues). In this
situation, firms seeking to diminish the political advantage gained by
the focal firm must develop constituency-building capabilities; however,
the focal firm is far ahead as a result of the close links between its market
capabilities and the skills that are required for constituency building.
In Country B, by contrast, the policy makers are less concerned with the
views of their constituents (which is often the case in less democratic
regimes), and it is more likely that political advantages would be gained
through information provision. In this case, a focal firm with strong capabil-
ities in using constituency building will have a more difficult time achieving
its desired policy outcomes because it must either use an unfamiliar tactic or
use a tactic that is less likely to have a strong impact on policy.
Indeed, the institutional environments of different types of modern democ-
racies shape how businesses interact with governments. For example, Hillman
and Keim (1995) suggest that the political institutions in the parliamentary
systems of countries such as Britain and Germany make information lobby-
ing more desirable than those in the presidential-congressional systems of
countries such as the United States where constituency building is a more

effective option.
Thus, we argue that when firms use tactics that are aligned with the
needs of the policy environment and the specific policy outcome that they
seek, they are more likely to gain a sustainable advantage from their politi-
cal strategy.
Proposition 6. When a firm uses the political tactic that is best suited to
bring about the needs of the desired policy outcome (i.e., external align-
ment between the political tactic used and the needs of the political
environment), the resulting advantage is more sustainable.
Consistency of Internal and External Alignments
Propositions 5 and 6 concern the benefits of internal and external align-

ment in achieving sustained advantage from political strategy. A key issue
arises, however, when these two factors push a firm in different directions.
A firm is in a difficult position if it has capabilities that lend themselves
to the use of a particular tactic but the policy outcome that they seek in the
political arena requires a different tactic. For example, firms with strong
B2B marketing skills might be able to convince relevant politicians that a
particular policy is a good idea, but if it is in a highly visible area where
individuals (voters) have strong views about the policy, then constituency
building may be necessary to convince the politicians that supporting the
policy that the firm desires will not adversely affect their chances for
re-election. Likewise, capabilities in constituency building for policies with
little pubic visibility or impact might not be as useful in influencing policy
makers as the persuasion tactics used by firms with capabilities developed
through extensive experience with B2B marketing. When there is a misa-
lignment between the needs of the policy environment and the firms inter-
nal capabilities in using particular tactics (based on their market
capabilities), then the firm is less likely to build a sustainable advantage
from political strategy.
However, the internal alignment between a firms market capabilities
and its political capabilities is particularly valuable for creating sustained
advantage when it fits the needs of the political environment. If the political
tactic that is employed by the focal firm is clearly the most effective
mechanism by which to influence a particular outcome, then the firms
political advantage will be more sustainable than if the policy outcome

could be effectively influenced by firms wielding a wide array of politi-
cal tactics.
Proposition 7. When the political tactic that is most suitable to bring
about the desired policy outcome (i.e., external alignment) is also one
that is aligned with the firms internal capabilities (i.e., internal align-
ment), the advantaged gained from the firms political strategy will be
more sustainable.
DISCUSSION
Understanding how to more effectively manage political strategies to facili-
tate business activities in a particular business environment is an important
part of an overall business strategy. A firms business strategy can rarely be
isolated from its non-market environment and non-market activities, which
are composed of social, legal, and political structures (Baron, 1997; Ring
et al., 2005). A firms strategies and actions in non-market domains shape
market competition and influence the returns to many market activities. In
this paper, we have focused on a firms political environment and how it
builds capabilities in using different political tactics to achieve its political
goals. To gain and sustain a competitive advantage, firms need to consider
how their market capabilities influence their political capabilities in using
particular tactics and thus how their market capabilities affect the effective-
ness of their political strategy. Therefore, integrating market strategies with
the political strategies is important for firms to establish an overall compe-
titive advantage (Baron, 1995a, 1995b, 1997, 2001).
Although a firms political activities actions targeting governments
and other players in the political system that firms use to advance their pri-
vate ends have been found to considerably affect the firms overall eco-
nomic performance (Hillman et al., 1999; Peng & Luo, 2000; Li & Zhang,
2007), we are among the first to explore the complementarity between its
market activities and these political activities. If firms make any effort to
integrate their political activities and their market activities, then their mar-
ket-oriented capabilities may influence both their incentives for pursuing
political strategies and their capabilities in implementing political strategies.
When studying political strategies and political capabilities in isolation, we
may miss important complementarities between a firms market capabilities
and its political capabilities that can influence the success and sustainability
of its political endeavors.
Can Firms Entirely Replace Political Capabilities with Outsourcing?
One may wonder why firms need to develop their own political capabilities
rather than outsourcing political activities altogether. For example, firms in
many industries use professional lobbyists who have both extensive experi-
ence and knowledge in the relevant industry and strong political clout to
persuade regulators and politicians to adopt favorable positions on a broad
range of legislative and regulatory issues8 (for an example of outsourcing
lobbying by universities, see de Figueiredo & Silverman, 2006), or hire pro-
fessional public relations firms to engage the mass grass-root population.
The topic of the make or buy decisions in building political capabilities
is quite nascent and has not received much attention in the literature (with
the exception of de Figueiredo & Tiller, 2001), but we provide some argu-
ments about why we believe that in most cases, firms still need to develop
certain political capabilities even though they have the option to contract
with external talents to conduct political activities.
First, recent research has debunked the notion that money buys political
victory (for a review, see Walker & Rea, 2014), such that firms can rarely
expect to achieve success from entirely outsourcing political activities with-
out any corporate involvement. A key reason is that, when using external
talents to conduct political activities, firms still need to integrate the exter-
nal political tools with the deep knowledge of their own business operations
for which the political activities intend to create value (Baron, 2012). To
enhance the overall firm value, political activities should fit in the larger
picture of the overall firm strategy (King & Walker, 2014), and the firm
managers are in a better position than external professionals to know
where and how political activities should be connected with other key ele-
ments of the firms operations, because external professionals specialize in
utilizing political tools and tend to have less comprehensive and less in-
depth knowledge of the firms overall operations. Therefore, the firm still
needs to contribute a certain amount of internal resources and knowledge
to facilitate the political activities that are led or helped by external profes-
sionals. For example, there is more to the capability of information provi-
sion than simply hiring the right lobbyists. Firms must work with their
lobbyists to provide the right types of information and steer a dynamic pro-
cess toward the desired outcome, which still requires that the firm possess
abilities similar to those discussed above. Therefore, many political activ-

ities conducted by external lobbyists can be made more effective if firms are
involved to work with the lobbyists, rather than adopting a hands-off
approach by entirely outsourcing these activities.
Moreover, firms need to possess a reasonable amount of political kno-
whow in order to gain the awareness of politically sensitive issues or poten-
tial policy influence, without which firms will be slow to realize and react
when new needs to engage in politics and to undertake political strategies
arise (Baron, 2012). This is possibly why firms that are generally active in
pursuing political activities almost always have their in-house public or
government affairs offices (Meznar & Nigh, 1995).
Third, unlike the cases of economic production where consumers typi-
cally care only about the end products but not the process of producing
the products, many key audiences of political activities such as policy
makers and wary constituents are mindful of the process or the manners
in which political activities are conducted. Firms involvement becomes a
signal of greater accountability, genuineness, and reliability, whereas some
may view lobbyists as simply utilizing political connections to engage in
rent-seeking with policy makers, and public relations professionals simply
as skillful with PR stunts and tricks. For example, although many
corporations hire professional grass-root lobbying firms to engage in con-
stituency building, the corporations themselves are often closely involved
rather than adopting an arms-length transitional approach (e.g., see
Walker, 2014).
Finally, if the firms completely outsource their political activities and
develop little political knowledge or political capabilities of their own, then
they face the risks of becoming over-dependent on the lobbyists or PR pro-
fessionals, in which case external lobbying or PR professionals can hold up
the firm and extract most of the rents gained by political activities. Firms
are also reluctant to outsource when lobbying involve proprietary informa-
tion (de Figueiredo & Tiller, 2001).
Therefore, we believe that political capabilities remain highly relevant to
firms even though some talents of conducting political activities can be
hired from the market. Many observations are consistent with this conclu-
sion. Firms increasingly strengthen their in-house lobbying operations or
internal government affairs divisions, such as by hiring political activity
professionals as employees, while continuing to utilize external profes-
sionals (Baron, 2012). Moreover, political activities of firms are usually
controlled by the senior management rather than a functional office such
as the human resource office (Walker & Rea, 2014).
Of course, there exists variation among firms. For example, the consid-
eration of make or buy may be moderated by other factors, such as
the frequency of political activities; the above analysis may be more rele-
vant for firms that need to engage in political activities reasonably often
rather than in a one-off situation. The overall topic of outsourcing political
activities calls more extensive future investigation.
How Far Can Capabilities Travel within the Firm?
One may question how far capabilities can travel inside a firm. We build,
to a degree, on the insight of the literature into corporate diversification
about the degree of relatedness in that more closely related industries gen-
erally result in more successful diversification attempts because the capabil-
ities and resources are more applicable (e.g., Silverman, 1999). We are
looking at the underlying capabilities required, in a general sense, in mar-
keting and technology capabilities, and B2B versus B2C customer focus,
and seeing how those might be best deployed to the purpose of political
influence. The relatedness of skills is the foundation for our statements
about the applicability of capabilities that appear more relevant to market-
ing versus technological, and B2B versus B2C strategies to different types
of political influence tactics.
We argue that the context-specific aspects of market and political arenas
should not stop knowledge transfer of relevant underlying capabilities that
can be applied in both areas. Just as Philip Morris bought 7UP and Miller
Beer to utilize its advertising capabilities built for the cigarette business,
firms can apply capabilities built for the market to political influence. Just
as capabilities can cross industry boundaries, we argue that they can cross
market to political spheres of influence as well.
Meanwhile, we acknowledge that the mechanisms through which cap-
abilities travel within a firm from marketing or R&D divisions to the
divisions that handle political matters may be shaped by certain organiza-
tional factors. First, the structure of the organization may play a role in
facilitating or obstructing internal transfer of capabilities. In some lobbying
cases, personnel from functional areas such as the R&D or marketing may
be consulted when political affairs staff prepare for their arguments, such
as being called upon to help with developing technical papers. The ease for
the talents and capabilities to travel between divisions thus the extent to
which the firm could draw on the capabilities developed in market activities
to assist the implementation of political tactics would be critically shaped
by the organizational structure and the relationships amongst different

divisions in the firm. Second, corporate culture also plays an important
role. As discussed earlier, anecdotes suggest that many high-tech firms in
the Silicon Valley seem very proud and avid in motivating grass-root sup-
port through social media, which is possibly bolstered by their successful
experience of directly interacting with mass individual customers, but they
tend to treat the idea of working with lobbyists with aversion or even dis-
tain in some cases. Therefore, a firms corporate culture developed in busi-
ness activities can shape how its staff approaches political affairs.
Theoretical Contributions
The theoretical insights developed in this paper advance the research on

integrated strategy, that is, how non-market strategies firm actions in
the context of social, legal, and political environments should be integrated
with market strategies to maximize overall returns (Baron, 1995a, 1995b,
1997, 2001), a perspective that Hillman, Keim, and Schuler (2004, p. 852)
consider to be the most promising for studying corporate political activ-
ities. Although researchers in related areas have shown increasing interest in
integrated strategy, such research is still in its infancy. Existing studies report
that weaker firms in the market are more active than stronger firms in pursu-
ing political strategies to insulate themselves from market competition
(e.g., Lenway et al., 1996; Morck et al., 2001). However, a firms market cap-
abilities do not always crowd out a firms political activities. We propose that
a firms market capabilities may complement its political activities by increas-
ing its political capabilities for utilizing certain political tactics. Our arguments
are focused on the impact of a firms market capabilities on political activ-
ities. To be more effective in the political arena, firms must utilize political
tactics such as information provision and constituency building; we examined
how the political capabilities that are needed to effectively utilize each politi-
cal tactic can be accumulated through the firms capabilities in marketing or
R&D and in its customer focus (B2B or B2C). These capabilities enhance the
skills that are required to effectively utilize a particular political tactic.
Prior research has explored how firms formulate their political strategies
and how external and internal conditions influence the deployment of
various corporate political tactics (e.g., Hillman & Hitt, 1999; Oliver &
Holzinger, 2008; Shaffer & Hillman, 2000). Although many firms are
assumed to have political strategies, how a firms ability to utilize specific
political tactics is influenced by the firms market capabilities has received

little attention in the literature. We examine how a firms ability to use such
tactics is built and how it influences the sustainability of a firms advan-
tages gained through political means. Knowing when political tactics are
used is important, but when determining the ability of political action to
achieve economic rents (i.e., some advantages for the firm), it is also impor-
tant to know when such attempts are more likely to be successful because
of the firms capabilities in using particular tactics, and to understand the
needs of the political environment.
We also contribute to the emerging body of research on firms political
capabilities. Prior research has examined the general nature of firms politi-
cal capabilities, the influence of prior political activities on multinational
firms political capabilities, and the home countrys sociopolitical influence
on the development of firms political capabilities (e.g., Garcia-Canal &
Guillen, 2008; Henisz & Delios, 2002; Holburn & Zelner, 2010). This analy-
sis constitutes a new addition to the emerging body of literature on firms
political capabilities, which has focused primarily on how prior experience
with political activities in both home countries and host countries increases
the likelihood of success of future political activities. In this paper, we are
able to significantly increase our theoretical understanding of political cap-
abilities, by maintaining a sharp focus on the link between market capabil-
ities and political capabilities. Political capabilities should not be viewed in
isolation; although some firms may isolate political activity from all market
activities, more experienced firms recognize that if the skills that underlie
their market capabilities can help them to compete more effectively in the
political arena, they should integrate their market and political capabilities,
at least to some degree. At a minimum, firms should draw on their stock of
applicable knowledge and skills to provide a strong foundation for their
political capabilities.
In addition, firms are more likely to sustain their success with political
strategies if they develop the capabilities to use political tactics that build
on their existing capabilities than if they develop political capabilities that
are completely disconnected from their market capabilities. Just as related
diversification is more attractive than unrelated diversification, firms will
have more success in developing political capabilities in areas that are
related to their market capabilities than they would in areas in which the
firm has no relevant skills (i.e., a political capability unrelated to a firms
market capabilities such that the two capabilities do not draw on any kind
of common knowledge or skill set).
The sustainability of political advantage is an important but nascent

concept and an important research area with many implications for corpo-
rate political strategy and political tactics (e.g., Capron & Chatain, 2008).
This paper adds new insights into this area by discussing how market
capabilities require underlying skills that contribute to particular political
tactics; that is, internal alignment increases the sustainability of the advan-
tage that the firm gains from its political efforts by making imitating the
political capability more difficult. Moreover, external alignment with the
institutional environment also enhances the sustainability of a firms politi-
cal advantage. We highlight the need for both internal alignment (between
the firms market capabilities and political tactic) and external alignment
(between the political tactic utilized and the needs of the political environ-
ment). Political advantage is most sustainable when the tactic that is best
suited to the policy environment is also the tactic that is aligned with the
firm market capabilities. When there is a mismatch between what a firm is
capable of doing and what is most suited to the environment, the firm faces
an unenviable choice. It can use a political tactic with which it is relatively
unfamiliar and that does not draw closely upon its market capabilities, or it
can apply what it knows how to do in an environment in which such a tac-
tic is unlikely to influence policy outcomes. Most firms in this situation are
likely to seek to develop capabilities in using the most relevant political tac-
tic, but such skills are not general and can take time to develop, especially
in the context of most political environments with highly specific rules
(formal and informal) by which firms must abide.
Limitations and Future Research
The paper has several limitations and may open new avenues for future
research. First of all, this paper develops theories that have clear implica-
tions for testable hypotheses, so systematic empirical tests of the theories
are needed. In addition, while we have collected anecdotal evidence about
the outcome of political activities of firms, we have obtained fewer direct
observations on the mechanisms how firms deploy resources such as
human capital internally to allow capabilities developed in market opera-
tions be used in conducting political activities. It is typically very difficult
to collect fine-grained data on firms internal operations at a systematic
level across a large number of firms, so we consider in-depth case analysis
to be appropriate in the circumstances.
Second, we have examined two general types of political tactics: infor-

mation provision and constituency building, but there are other ways to
categorize political tactics based on different dimensions. For example,
firms may decide to be leaders or followers or to stay inactive with respect
to political issues (Lenway & Rehbein, 1991), or they may choose to take
political actions collectively or individually (Hillman & Hitt, 1999).
According to Hillman and Hitt (1999), firms choose to engage in political
activities at two levels the individual level and the collective level, and at
each level firms may choose to deploy generic political tactics (information
provision, constituency building, and financial contribution). The existing
literature on political capabilities, including this paper, has placed greater
focus on individual-based political strategies. The knowledge gained about
the political capabilities required by individual-based political strategies is
important because individual-based political capabilities are highly preva-
lent and common political activities are not first and foremost collective-
based.9 For example, individual political activities are preferred in pluralist
countries (Hillman & Hitt, 1999) and institutional contexts where the state
has greater redistributive power (Jia, 2014); they are also preferred by firms
of certain characteristics such as owning greater financial resources
(Hillman & Hitt, 1999) and seeking to bend the rules instead of chan-
ging the rules (Harstad & Svensson, 2011). One may question if the
knowledge developed in this paper is relevant to collective political strate-
gies. We consider this paper relevant because to successfully implement
collective political activities, firms also need to possess certain political
capabilities firms cannot entirely free ride on other firms in the political
coalition. Nevertheless, the political capabilities critical to collective politi-
cal strategies may comprise of different elements, such as an ability to man-
age collective action and align the interest of all firms that participate in the
political coalition. For example, it is possible that a strong ability to man-
age alliances could transcend to help the firm gain collaborative skills to
manage collective political activities, which is an interesting topic to investi-
gate in the future.
Third, some basic assumptions underlying the theory developed here are
largely based on the institutional environment in developed countries, such
as the United States and Europe. The institutional environment in emer-
ging economies, however, may have dramatically different characteristics.
For example, political economy scholars have extensively discussed the
governments role in the economy as a grabbing hand in many countries
where the market-supporting institutions are underdeveloped, that is, in
countries where governments have the power to expropriate firms without
legal basis because checks and balances on government behavior are inef-
fective (e.g., Frye & Shleifer, 1997). Therefore, firms in such environments
may pursue additional political goals that are not common in the U.S. In
addition, the political tactics that are common in the United States and
Europe may not be equally prevalent in other parts of the world. For
example, because China has no structured lobbying system, formal infor-
mation lobbying, which is important for firms in North America, is not yet
deemed an important political strategy in that country (e.g., Kennedy,
2005).10 Chinese laws also prohibit campaign contributions (e.g., Li, Meng, &
Zhang, 2006). It would be interesting for future studies to expand the scope
of the political goals that firms intend to achieve and the range of political
tactics available to gain a more comprehensive understanding of integrated
strategy in the international setting.
Finally, we examined the sustainability of the gains of using political stra-
tegies from the perspective of aligning firm market capabilities and political
capabilities and described how this relationship is moderated by the require-
ments of the policy environment. The sustainability of the competitive
advantage gained in a non-market setting is a subject of critical importance
in the development of theories on corporate political strategies and requires
additional examination that is beyond the scope of this paper. For example,
political rivalry may also diminish the benefits that a firm gains from politi-
cal ties with politicians who have lost their political power (e.g., Fisman,
2001; Siegel, 2007).
CONCLUSION
We propose that this research represents an important extension of the lit-
erature on corporate political activities and the relationship between politi-
cal and market strategies. The literature on corporate strategy puts market
capabilities on center stage and does so appropriately, as the market is
the primary battleground for competition among firms. We propose, how-
ever, that the role of political strategy and its importance in shaping the
rules of the game under which markets operate will require more theoreti-
cal and empirical attention if we are to better understand the critical inter-
play between markets and political strategy, because market capabilities
influence a firms ability to more effectively implement political tactics. The
interplay between market capabilities and political capabilities is critical;
although further research is necessary, this paper constitutes an important

step toward addressing this topic.
NOTES
1. Financial Times (2012b). Also see, for example, The Bloomberg Business
Week (2012).
2. See, for example, Los Angeles Times (2011a; 2011b) and The Wall Street
Journal (2010).
3. Firms, in some cases, use a combination of tactics to achieve their political
goals instead of only one (Keim & Zeithaml, 1986). Here, we examine each indivi-
dual political tactic as a first step in understanding how a firms market capabilities
affect the effectiveness of political tactics.
4. To further elaborate this example, Boeing has a commercial aviation business
in which they sell to other corporations while Northrop Grumman only operates in
the defense industry (multiple segments of it but all in defense) and sells almost
exclusively to the U.S. government (with some very small sales to select foreign
governments). Thus, Boeing has capabilities in B2B marketing to support the sale
of their aircraft and maintenance services to airlines and air freight companies
around the globe, while Northrop Grumman lacks these capabilities because their
marketing and sales efforts are exclusively focused on the (US) government. Thus
we think this represents a strong example of two firms that operate in the same
industry but the different scopes of the firms indicate different capabilities in B2B
marketing because only one firm has a B2B business.
5. We acknowledge that we focus on mobilizing individuals. The political
capabilities required by mobilizing non-government organizations (NGOs) can be
quite different.
6. For example, over 10,000 internet sites successfully coordinated a black out
day to use the social media power to protest the antipiracy legislation lobbied by
traditional media companies. As a result, thousands of voters contacted their repre-
sentatives, disrupting phone and email systems in Congress. The groups have
seized on a cautiously worded White House statement Saturday offering partial
support for their cause and quickly joined a brewing blackout plan, outflanking the
traditional media industry, according to The Wall Street Journal (2012). Also see,
for example, Financial Times (2012a).
7. The Wall Street Journal (2011).
8. Bloomberg Business Week (2011).
9. Recent research has questioned the notion that the unity of collective action
of businesses necessarily generates more political power (for a review, see Walker &
Rea, 2014). For example, research has shown that business political power in the
United States becomes increasingly fragmented (Drutman, 2010; Martin, 1999;
Mizruchi, 2013; Vogel, 1996) and specific business interests are increasingly cited in
policy debates (Baumgartner, Berry, Hojnacki, Leech, & Kimball, 2009).
10. Formal information lobbying is becoming increasingly more accessible
and important to firms in China because it is being progressively facilitated by
institutional reform initiatives, including the recent Legislation Law, which attempts
to endorse a more open and consultative legislative process to encourage wider citi-
zen participation in the legislative process (Paler, 2005).
ACKNOWLEDGMENTS
We would like to thank Edward Walker, David Primo, and the participants
of the 2015 Annual Conference of the International Society for New
Institutional Economics (ISNIE) for helpful comments.
REFERENCES
Aplin, J. C., & Hegarty, W. H. (1980). Political influence: Strategies employed by organiza-
tions to impact legislation in business and economic matters. Academy of Management
Journal, 23(3), 438450.
Barney, J. B. (1986). Organizational culture: Can it be a source of sustained competitive
advantage? Academy of Management Review, 11(3), 656665.
Barney, J. B. (1991). Firm resources and sustained competitive advantage. Journal of
Management, 17(1), 99120.
Baron, D. P. (1995a). Integrated strategy: Market and nonmarket components. California
Management Review, 37, 4765.
Baron, D. P. (1995b). The nonmarket strategy system. Sloan Management Review,
37(1), 7385.
Baron, D. P. (1997). Integrated strategy and international trade disputes: The Kodak-Fujifilm
case. Journal of Economics & Management Strategy, 6(2), 291346.
Journal of Economics & Management Strategy, 10(1), 745.
Baron, D. P. (2012). Business and the environment (7th ed.). Englewood Cliffs, NJ:
Prentice Hall.
Baumgartner, F., Berry, J., Hojnacki, M., Leech, B., & Kimball, D. (2009). Lobbying and
policy change: Who wins, who loses, and why. Chicago, IL: University of Chicago Press.
Baysinger, B. (1984). Domain maintenance as an objective of business political activity: An
expanded typology. Academy of Management Review, 9(2), 248258.
Baysinger, B. D., Keim, G. D., & Zeithaml, C. P. (1985). An empirical evaluation of the poten-
tial for corporate constituency programs with shareholders. Academy of Management
Journal, 28(1), 180200.
Bloomberg Business Week. (2011). The power brokers: Fifteen regulators, lawmakers, and
lobbyists shaping the torrent of regulations. Bloomberg Business Week, January 2431.
Bloomberg Business Week. (2012). Startups cant ignore Washington. The Bloomberg Business
Week, April 12.
Bonardi, J.-P. (2008). The internal limits to firms nonmarket activities. European Management
Review, 5, 165174.
Bonardi, J.-P., Holburn, G. L. F., & Vanden Bergh, R. G. (2006). Nonmarket strategy perfor-
mance: Evidence from U.S. electric utilities. Academy of Management Journal,
49(6), 12091228.
Bonardi, J.-P., & Keim, G. D. (2005). Corporate political strategies for widely salient issues.
Capron, L., & Chatain, O. (2008). Competitors resource-oriented strategies: Acting on compe-
titors resources through interventions in factor markets and political markets.
Culpepper, P. D. (2010). Quiet politics and business power: Corporate control in Europe and
Japan. Cambridge: Cambridge University Press.
Cyert, R. M., & March, J. G. (1963). A behavior theory of the firm. Englewood Cliffs,
NJ: Prentice Hall.
de Figueiredo, J. M., & Silverman, B. S. (2006). Academic earmarks and the returns to lobby-
ing. Journal of Law and Economics, 49(2), 597625.
de Figueiredo, J. M., & Tiller, E. H. (2001). The structure and conduct of corporate lobbying:
How firms lobby the federal communications commission. Journal of Economics &
Management Strategy, 10(1), 91122.
Delios, A., & Henisz, W. J. (2003). Political hazards, experience, and sequential entry strate-
gies: The international expansion of Japanese firms, 19801998. Strategic Management
Journal, 24(11), 11531164.
Drutman, L. (2010). The business of America is lobbying: The expansion of corporate political
activity and the future of American pluralism. Berkeley, CA: University of California.
Financial Times. (2012a). Lobbying campaign scuttles US piracy bills. Financial Times,
January 20.
Financial Times. (2012b). Silicon Valley learns fast in game of lobbying. Financial Times,
January 18.
Fisman, R. (2001). Estimating the value of political connections. The American Economic
Review, 91(4), 10951102.
Fremeth, A. R., & Shaver, J. M. (2014). Strategic rationale for responding to extra-jurisdictional
regulation: Evidence from firm adoption of renewable power in the US. Strategic
Frye, T., & Shleifer, A. (1997). The invisible hand and the grabbing hand. The American
Economic Review, 87(2), 354358.
Frynas, E. G., Mellahi, K., & Pigman, G. A. (2006). First mover advantages in international
business and firm-specific political resources. Strategic Management Journal,
27(4), 321345.
Garcia-Canal, E., & Guillen, M. F. (2008). Risk and the strategy of foreign location choice in
regulated industries. Strategic Management Journal, 29(10), 10971115.
Ghemawat, P. (1991). Commitment: The dynamic of strategy. New York, NY: Free Press.
Ghemawat, P. (2009). Strategy and the business landscape (3rd ed.). Upper Saddle River, NJ:
Pearson Prentice Hall.
Grier, K. B., Munger, M. C., & Roberts, B. E. (1994). The determinants of industry political
activity, 19781986. The American Political Science Review, 88(4), 911926.
Harstad, B., & Svensson, J. (2011). Bribes, lobbying, and development. American Political
Science Review, 105(1), 4663.
Henisz, W. J., & Delios, A. (2002). Learning about the institutional environment. In P. Ingram &
B. Silverman (Eds.), New institutionalism in strategic management (Advances in Strategic
Management) (Vol. 19, 339372). Amsterdam: Elsevier.
Henisz, W. J., & Zelner, B. (2012). Strategy and competition in the market and non-market
arena. Academy of Management Perspectives, 26(3), 4051.
Hillman, A. J., & Hitt, M. A. (1999). Corporate political strategy formulation: A model of
approach, participation, and strategy decisions. Academy of Management Review,
24(4), 825842.
Hillman, A. J., & Keim, G. D. (1995). International variation in the business-government
interface: Institutional and organizational considerations. Academy of Management
Review, 20(1), 193214.
research agenda. Journal of Management, 30(6), 837857.
Hillman, A. J., Zardkoohi, A., & Bierman, L. (1999). Corporate political strategies and firm
performance: Indications of firm-specific benefits from personal service in the U.S. gov-
ernment. Strategic Management Journal, 20(1), 6781.
Holburn, G. L. F., & Vanden Bergh, R. G. (2002). Policy and process: A game-theoretic
framework for the design of non-market strategy. Advances in Strategic Management,
19, 3366.
Holburn, G. L. F., & Vanden Bergh, R. G. (2008). Making friends in hostile environments:
Political strategy in regulated industries. Academy of Management Review,
33(2), 521540.
Holburn, G. L. F., & Zelner, B. A. (2010). Policy risk, political capabilities and international
investment strategy: Evidence from the global electric power industry. Strategic
Management Journal, 31(12), 12901316.
Jia, N. (2014). Are collective and private political actions substitutes or complements?
Empirical evidence from Chinas private sector. Strategic Management Journal,
35(2), 292315.
Keim, G. D. (1981). Foundations of a political strategy for business. California Management
Review, 23(3), 4148.
Keim, G. D., & Zeithaml, C. P. (1986). Corporate political strategy and legislative decision
making: A review and contingency approach. Academy of Management Review,
11(4), 828843.
Kennedy, S. (2005). The business lobbying in China. Cambridge, MA: Harvard
University Press.
King, B. G., & Walker, E. T. (2014). Winning hearts and minds: Field theory and the three
dimensions of strategy. Strategic Organization, 12(2), 134141.
Kotler, P., & Keller, K. (2008). Marketing management (13th ed.). Upper Saddle River, NJ:
Prentice Hall.
Leiblein, M. J., & Miller, D. J. (2003). An empirical examination of transaction- and firm-level
influences on the vertical boundaries of firms. Strategic Management Journal,
24, 839859.
Lenway, S. A., Morck, R., & Yeung, B. (1996). Rent seeking, protectionism and innovation in
the American steel industry. The Economic Journal, 106(435), 410421.
Lenway, S. A., & Rehbein, K. (1991). Leaders, followers, and free riders: An empirical test
of variation in corporate political involvement. Academy of Management Journal,
34(4), 893905.
Levitt, B., & March, J. G. (1998). Organizational learning. Annual Review of Sociology,
14, 319338.
Li, H., Meng, L., & Zhang, J. (2006). Why do entrepreneurs enter politics? Evidence from
China. Economic Inquiry, 44(3), 559578.
Li, H., & Zhang, Y. (2007). The role of managers political networking and functional experi-
ence in new venture performance: Evidence from Chinas transition economy. Strategic
Li, Y., Peng, M., & Macaulay, C. (2013). Market-political ambidexterity during institutional
transitions. Strategic Organization, 11(2), 205213.
Los Angeles Times. (2011a). Although Apple is a tech goliath, it has the mind-set of a David in
D.C. power circles. Los Angeles Times, March 29.
Los Angeles Times. (2011b). Facebook, already ubiquitous in Washington, aims to beef up its
lobbying power. Los Angeles Times, February 3.
Martin, C. (1999). Stuck in neutral: Business and the politics of human capital investment policy.
Princeton, NJ: Princeton University Press.
Meznar, M., & Nigh, D. (1995). Buffer or bridge? Environmental and organizational determi-
nants of public affairs activities in American firms. Academy of Management Journal,
38, 975996.
Mizruchi, M. (2013). The fracturing of the American corporate elite. Cambridge, MA: Harvard
University Press.
Morck, R., Sepanski, J., & Yeung, B. (2001). Habitual and occasional lobbyers in the U.S.
steel industry: An EM algorithm pooling approach. Economic Inquiry, 39(3), 365378.
Oliver, C., & Holzinger, I. (2008). The effectiveness of strategic political management: A
dynamic capabilities framework. Academy of Management Review, 33(2), 496520.
Paler, L. (2005). Chinas legislation law and the making of a more orderly and representative
legislative system. The China Quarterly, 182, 301318.
Peng, M. W., & Luo, Y. (2000). Managerial ties and firm performance in a transition
economy: The nature of a micro-macro link. Academy of Management Journal,
43(3), 486501.
Peteraf, M. A. (1993). The cornerstones of competitive advantage: A resource-based view.
Strategic Management Journal, 14(3), 179191.
Porter, M. E. (1980). Competitive strategy. New York, NY: Free Press.
Ring, P. S., Bigley, G. A., DAunno, T., & Khanna, T. (2005). Perspectives on how govern-
ments matter. Academy of Management Review, 30(2), 308320.
Roberts, R. W., Dwyer, P. D., & Sweeney, J. T. (2003). Political strategies used by the US
public accounting profession during auditor liability reform: The case of the private
securities litigation reform act of 1995. Journal of Accounting and Public Policy,
22, 433457.
Rucker, D. D., & Petty, R. E. (2004). Emotion specificity and consumer behavior: Anger,
sadness, and preference for activity. Motivation and Emotion, 28(1), 321.
Schuler, D. A. (1996). Corporate political strategy and foreign competition: The case of the
steel industry. Academy of Management Journal, 39(3), 720737.
Schuler, D. A., Rehbein, K., & Cramer, R. D. (2002). Pursuing strategic advantage through
political means: A multivariate approach. Academy of Management Journal,
45(4), 659672.
Scott, D. (2010). The new rules of marketing and PR: How to use social media, blogs, news
releases, online video, and viral marketing to reach buyers directly. Hoboken, NJ: Wiley.
Shaffer, B., & Hillman, A. J. (2000). The development of business-government strategies by
diversified firms. Strategic Management Journal, 21(2), 175190.
Siegel, J. (2007). Contingent political capital and international alliances: Evidence from South
Korea. Administrative Science Quarterly, 52, 621666.
Silverman, B. S. (1999). Technological resources and the direction of corporate diversification:

Toward an integration of the resource-based view and transaction cost economics.
Management Science, 45(8), 11091124.
The Wall Street Journal. (2010). Lobbying war over net heats up. The Wall Street Journal,
December 10.
The Wall Street Journal. (2011). Banks return with a goal: Pushing back. The Wall Street
Journal, January 26.
The Wall Street Journal. (2012). New, old media battle over net rules. The Wall Street Journal,
January 18.
Vogel, D. (1996). Kindred strangers: The uneasy relationship between politics and business in
America. Princeton, NJ: Princeton University Press.
Walker, E. T. (2014). Grassroots for hire: Public affairs consultants in American democracy.
Cambridge: Cambridge University Press.
Walker, E. T., & Rea, C. M. (2014). The political mobilization of firms and industries. Annual
Review of Sociology, 40, 281304.
Wernerfelt, B. (1984). A resource-based view of the firm. Strategic Management Journal,
5(2), 171180.
HOW PATENT STRATEGY
AFFECTS THE TIMING AND
METHOD OF PATENT
LITIGATION RESOLUTION
Deepak Somaya
ABSTRACT
Patent litigation consists of non-market actions that firms undertake to

access intellectual property rights defined by prior legislation and
enforced by the courts. Thus, patent litigation provides an interesting
context in which to explore aspects of firms non-market strategies. In
contrast with prior non-market strategy research that has largely focused on
how political institutions define the rules of the game for market competition,
non-market actions within patent litigation primarily seek to access and
apply these broad policies to specific situations, products, or assets that
matter to the firm. Furthermore, because such non-market actions are
directly influenced by the firms market strategies, they represent a promising
area for research on integrated (market and non-market) strategies as well.
The goal of this paper is to explain how generic patent strategies that
firms use to support their competitive advantage in the product-market
influence non-market outcomes related to the timing of patent litigation

ISSN: 0742-3322/doi:10.1108/S0742-332220160000034014
471
472 DEEPAK SOMAYA
resolution. In contrast with prior research that has studied settlement in

patent litigation essentially as a one-shot bargaining game, this paper
seeks to explain litigation resolution as an outcome of the competing
mechanisms of settlement and adjudication that operate continually
during litigation. Using a large sample of patent litigations in research
medicines and computers, I model the timing of patent litigation resolution
in a proportional hazards framework, wherein settlement and adjudication
are competing risks. The evidence found is consistent with the proposition
that the speed with which patent litigation is resolved by either settlement
or adjudication reflects the use of proprietary, defensive, and leveraging
patent strategies by firms. These findings also help to explain unexpected
and anomalous findings regarding the settlement of patent litigation
reported in prior research.
Keywords: Patent strategy; litigation; suit settlement; adjudication;
duration models
The availability and strategic use of barriers to imitation are central to

sustaining firm competitive advantage, and preventing the erosion of profits
from firm innovation (Barney, 1991; Lippman & Rumelt, 1982; Peteraf,
1993; Rumelt, 1984). Intellectual property rights, and patents in particular,
have long been recognized as a significant type of imitation barrier
(Mahoney & Pandian, 1992; Rumelt, 1984), and research has shed substan-
tial light on the strategic role played by patents in several important ways
(see Somaya, 2012 for a review). In particular, three generic patent
strategies proprietary, defensive, and leveraging have been highlighted
as having a proximate impact on firm competitive advantage and are
central motivations for firms acquisition and use of patents (Blind, Edler,
Frietsch, & Schmoch, 2006; Cohen, Nelson, & Walsh, 2000; Duguet &
Kabla, 1998; Somaya, 2012). Consistent with the definition of firm strategy
in other domains (Rumelt, 2011), each generic patent strategy consists of a
distinct theory or logic of how a firm may use patents to compete and
succeed in the marketplace. In turn, prior research has shown that these
generic strategies affect key outcomes such as patent value (Reitzig, 2004),
use of oppositions and reexamination (Blind, Cremers, & Mueller, 2009;
Clarkson & Toh, 2010), and decisions to undertake and settle litigation
(Polidoro & Toh, 2011; Somaya, 2003).
Timing and Method of Patent Litigation Resolution 473
Broadly, patent strategy entails actions by firms in three domains of patent

activity the obtaining of patent rights, the licensing of these rights, and last
but not least, the enforcement of patents (Somaya, 2012). In contrast with
research on non-market strategies that has largely focused on how political
institutions define the rules of the game for market competition (De
Figueiredo, 2009), non-market actions within these domains primarily seek
to access and apply the legal rules relating to intellectual property that have
been created by prior legislation, agency regulations, and case law. For exam-
ple, actions relating to obtaining patent rights largely occur at the patent
office, where firms must decide which technologies to expend resources on
and what legal provisions (e.g., provisional applications, reexaminations, and
continuations) must be availed of in each case. While patent licensing entails
private contracting between parties, it relies on and is shaped by provisions of
contract law, as well as aspects of patent and antitrust law. Similarly, patent
litigation actions seek to access enforcement rights enshrined under patent
law, and shaped by the institutional features of courts. How firms seek com-
petitive advantage or performance advantages from patents in the market
environment their patent strategy in turn has implications for actions
they undertake to access legal rules and rights available in each of these
domains in the non-market environment for patents.
In the area of patent enforcement, which is the focus of the current
paper, prior research has sought to understand a number of important
questions inter alia, why companies decide to file patent suits (Lanjouw &
Schankerman, 2001; Nerkar, Paruchuri, & Khaire, 2007; Polidoro & Toh,
2011), how they use predatory litigation through preliminary injunctions
(Lanjouw & Lerner, 2001), their choice of litigation forum (Somaya &
McDaniel, 2012), and the strategic factors that lead to non-settlement in
patent litigation (Somaya, 2003). While these studies inform us substantially
about several aspects of patent enforcement, all of them suffer from one
significant drawback they essentially take a one shot view of litigation,
which tells us very little about choice and timing of suit resolution mechan-
isms within litigation and how patent strategy influences them. Specifically,
litigation can be resolved by the parties reaching a settlement or by an
adjudicatory action taken by the court, and these alternatives act continually
as competing risks for resolution of the dispute (Hughes & Savoca, 1999;
Spier, 1992). During patent litigation, the parties have the option to simulta-
neously pursue adjudication in the court and parallel negotiations to settle
the dispute, and they can strategically choose to delay or speed up either
one or both of these mechanisms. In turn, these resolution mechanisms, and
the speed with which they are reached, directly influence and help to explain
settlement outcomes in litigation (Spier, 1992).
474 DEEPAK SOMAYA
Accordingly, in this paper, I draw on and extend asymmetric stakes and

asymmetric information theories of settlement in litigation (Priest & Klein,
1984) to propose and test hypotheses regarding the timing of patent-suit
resolution. To explain the impacts of proprietary and defensive patent
strategies, I develop a novel extension of asymmetric stakes theory for the
timing of patent-suit settlement and adjudication. I also apply extant
asymmetric information theory regarding the timing of dispute resolution
(e.g., Ma & Manove, 1993; Spier, 1992), which relies on a signaling logic,
to explain the impacts of a leveraging patent strategy on the timing of
patent litigation resolution. Thus, from a non-market strategy perspective,
the current paper focuses on the firms strategic choices in the non-market
arena specifically whether to speed up or delay settlement and adjudica-
tion in litigation and seeks to explain how they are influenced by different
market (patent) strategies. Empirically, I test these propositions by modeling
the hazard of patent litigation termination in a competing risks framework,
where settlement and adjudication are the two competing risks. The empiri-
cal tests are conducted on a sample of patent suits from the U.S. Federal
District Courts in two highly innovative industries computers and
research medicines.
The empirical findings are largely (but not entirely) consistent with the
theoretical hypotheses developed in this paper. Proprietary strategies are
associated with delays in settlement during patent litigation but with
faster adjudication; however, statistical support for the former relation-
ship (with delays in settlement) is weak. Defensive patent strategies are
associated with both faster settlement and faster adjudication in patent
litigation. Leveraging strategies are only associated with faster adjudica-
tion. These findings provide a deeper understanding of the mechanisms at
work in patent litigation when compared with prior work that focuses
only on settlement outcomes (Lanjouw & Schankerman, 2001; Somaya,
2003). For example, proprietary strategies can be understood to decrease
settlement (Somaya, 2003) largely because adjudication is speeded up,
rather than settlement being delayed. Similarly, prior research found that
there was no association between defensive strategies and settlement in
patent litigation (Somaya, 2003), but this finding can now be understood
to result from both settlement and adjudication being simultaneously
speeded up. Additionally, the empirical findings shed light on the impacts
of institutional differences between courts in their handling of patent
litigation (notably, so-called rocket dockets) and on inter-industry
differences in patent strategy (e.g., Cohen et al., 2000; Reitzig, 2004;
Somaya, 2003).
LITERATURE REVIEW AND HYPOTHESES

In recent management research, a new stream of work on the patent strate-
gies of firms has emerged from the broader inquiry into the economics of
patents and innovation. This research has its roots in the comparative
study of appropriability mechanisms (Levin, Klevorick, Nelson, & Winter,
1987; Teece, 1986), and the motivations cited by firms for obtaining patents
(Cohen et al., 2000; Duguet & Kabla, 1998). Other work in this tradition
has examined the role of patents in strategic deterrence (Lerner, 1995), in
safeguarding know-how leakage in licensing (Arora, 1995), and in defensive
strategies against patent enforcement by other firms (Grindley & Teece,
1997; Hall & Ziedonis, 2001).
When a company finds that its patent is being infringed it can choose to
either ignore it (e.g., if the patent is unimportant or enforcement is too
costly), or respond with other non-patent actions (such as competitive
market moves or attempts to use other appropriability mechanisms like the
firms complementary assets), or it can seek to enforce its property right.
Typically such patent enforcement begins with private negotiations, which
may result in a settlement on mutually agreed terms (such as an end to the
infringement, a license, and/or payments). If these negotiations fail, or are
expected to fail, one of the parties may take the dispute to court. In studies
focused specifically on patent litigation, research has found that more valuable
patents are more likely to be litigated (Lanjouw & Schankerman, 2001), and
that the types of knowledge building entailed in the patent moderate this
relationship (Nerkar et al., 2007). Research on patent litigation has also
examined the predatory use of preliminary injunctions against financially
distressed firms (Lanjouw & Lerner, 2001), and examined how generic patent
strategies influence decisions to settle patent litigation (Somaya, 2003).
Studying patent litigation outcomes in the manner of these studies,
however, has the drawback that litigation itself is treated as a one-shot
game, which essentially overlooks the competing risks between different
litigation resolution mechanisms by which litigation outcomes actually
come about. As Fig. 1 illustrates, such a competing risks approach views
adjudication by the court and settlement negotiations between the parties
as parallel mechanisms, either of which can lead to the resolution of
patent litigation. In turn, the impact of patent strategies on litigation
outcomes for example, on litigation settlement can result from their
influence on adjudication, or settlement negotiations, or both. In order to
fully understand how patent strategies operate in the context of patent liti-
gation, we need to examine their impacts on these competing mechanisms
476 DEEPAK SOMAYA
Fig. 1. Comparison of a One-Shot Outcomes View and a Competing Risk

Mechanisms View of Patent Litigation.
and how they in turn affect litigation outcomes. Further, as described

below, such an investigation may also provide a stronger test for theories
of patent strategy than would be feasible by examining one-shot litigation
outcomes alone.
The timing and method of litigation resolution have been modeled in the
law-and-economics tradition by Spier (1992), using an information-revealing
bargaining game set in the shadow of impending trial costs. The primary
prediction of Spiers model is a U-shaped distribution of settlements over
time, where some settlements occur early due to low information asymme-
tries and the expected costs of delay and litigation costs, and most remaining
settlements occur late but just before the trial deadline when costs are
expected to escalate rapidly (on the court steps, in one colorful descrip-
tion). However, this analysis has some drawbacks when applied to patent
litigation. To begin with, the base model uses a fixed timing for trial, which
acts as a deadline for the bargaining parties. It cannot, therefore, inform us
about the ways in which the parties may seek to influence the timing of adju-
dication. Moreover, when the timing of adjudication is endogenized into the
model, it loses the ability to make predictions about the timing of settlement
(Spier, 1992).
Spiers model is essentially a dynamic extension of a model by Bebchuk
(1984), which is one of several studies in law and economics that explain
non-settlement in litigation as resulting from information asymmetries
between the parties (Png, 1983; Schweizer, 1989; Spier & Spulber, 1993).
However, these asymmetric information theories are merely one explanation
for non-settlement of litigation (for a review, see Cooter & Rubinfeld,
1989), and an alternative theory based on asymmetric stakes between the
parties (Priest & Klein, 1984) has been empirically shown to play more sig-
nificant role in the resolution of patent (and also other types of intellectual
property) litigation (Siegelman & Waldfogel, 1996; Waldfogel, 1995).
Asymmetric stakes imply that one party has more to gain from winning the
suit than the other party loses, and this leads to non-settlement because the
resulting asymmetry in transfers from adjudication shrinks (or eliminates)
the bargaining surplus available to the parties in settlement (Priest &
Klein, 1984).1
In theoretical models of patent litigation, asymmetric stakes in patent
litigation generally arise from the diminution of joint profits as a result of
post-settlement competition between the patentee and its rival (Lanjouw &
Lerner, 1998; Meurer, 1989). Thus, the patentee gains more from (the
expected probability of) winning the suit and keeping its rival out of the
market than by settling the dispute, despite the savings in litigation costs.
In turn, the rival finds it difficult to compensate the patentee adequately for
giving up its exclusivity in the patent, given the reality of post-settlement
competition. However, these formal models of patent litigation only
employ lump-sum transfers as the negotiated outcome in settlement. With
more creative licensing arrangements in settlement, including running
royalties, it is perhaps conceivable that the patentee can be adequately
compensated even in a settlement outcome. Therefore, asymmetric stakes
in patent litigation must arise from difficulties in crafting and enforcing
such settlement contracts, which have been linked in prior work to the
impacts of patent strategies on the non-settlement of patent litigation
(Somaya, 2003).
However, prior research has not examined the implications of asymmetric
stakes for the timing of litigation resolution, whether through settlement or
adjudication. Ultimately, this question must be analyzed in the context of
the extremely high costs of continuing patent litigation, and the substantial
478 DEEPAK SOMAYA
back loading of these costs. According to one estimate, typical legal costs
for patent litigation were in the range of $1.03.0 million for each side in
1997, with over 50% being incurred in late-stage court hearings and trial
(AIPLA, 1997). In addition to pecuniary costs, patent litigation causes
considerable organizational dislocation, absorbing the time and energy of
managers, star scientists, and company lawyers in preparing for litigation,
and participating in discovery, depositions, and court hearings. Ongoing
patent litigation also creates substantial uncertainty in the marketplace,
which has an indirect but equally important economic impact.
The upshot of these ongoing and impending costs is that the parties
have strong incentives to settle the dispute. Resolution by settlement avoids
the costs of continuing with litigation, as well as the costs of trial (or
another form of adjudication), which becomes increasingly likely as litiga-
tion continues. While it is uncontroversial that the parties in litigation
determine the timing of settlement, it is important to recognize that they
can also exert an influence on the timing of adjudication, either by delaying
progress of litigation into its later stages (e.g., by seeking and taking more
time for litigation procedures such as filing motions, conducting deposi-
tions and discovery proceedings, and collecting and presenting evidence) or
by seeking earlier-stage court decisions (e.g., through summary motions).
Resolution by adjudication invariably involves costs, but these costs are
lower for earlier-stage adjudication than a full-blown trial.2 By seeking
such early adjudication, the parties forego at least some cost savings from
settling and avoiding adjudication altogether, and must therefore have con-
cluded that the prospects for settlement are poor relative to the advantages
gained from adjudication.
As Fig. 1 illustrates, pressures to terminate the litigation by either
settlement or adjudication operate continuously from the time a lawsuit is
filed, even as the litigation progresses from its early stages towards trial.
Therefore, as one moves forward through time and litigation-stage, fewer
and fewer disputes will remain in litigation, having been previously
disposed of by either settlement or adjudication. In practice, delaying
tactics in both settlement and adjudication can be constrained by the
discretionary powers of the supervising judges, who may set strict time-tables
for the conduct of litigation. In some jurisdictions, such as the so-called
rocket-docket courts where judges have taken a particularly strong stand,
the parties may be forced to progress relatively quickly with both litigation
and settlement-related actions. Thus, the timing and method of litigation
resolution may also entail an institutional dimension, which is incorporated
in the empirical specification of this paper.
Generic Patent Strategies and the Timing of Patent Litigation Resolution
This main goal of this paper is to demonstrate how generic patent strategies
influence the timing of patent litigation resolution, and how this differs
when the litigation is resolved by settlement or by adjudication. As noted
earlier, patent strategies can be understood as key underlying logics or
theories about how a firm plans to derive competitive advantage from its
patents. I examine three types of generic strategies pursued by firms with
their patents namely, proprietary strategies, defensive strategies, and
leveraging strategies. Each of these strategies can be understood to reflect
choices regarding patents in a particular technology (or product) area,
which in turn affects the firms decision making around a key patent that
may be involved in litigation. I link these patent strategies to the timing of
litigation resolution through asymmetric stakes reasoning, except in the
case of predatory litigation, which is based on a signaling rationale within
an asymmetric information view of litigation.
Proprietary Patent Strategy

Firms rely on patents to different extents for the protection of their
commercial interests. Some patents are more important because they are
built on, both technologically and commercially, to create valuable firm
assets. The patents then function as isolating mechanisms for the rents
emanating from these assets that is, they raise barriers to imitation of the
assets by other firms (Lippman & Rumelt, 1982; Rumelt, 1984; Somaya,
2003). Such patents may protect rents from co-specialized organizational
assets (Teece, 1986), follow-on innovations, or associated investments in
commercial development (Kitch, 1977). Moreover, they may be central to
strategic positions built by firms (Teece, Pisano, & Shuen, 1997), which
provide dynamic advantages in subsequent rounds of competition
(Stefanadis, 1997). Since the barriers to imitation created by these patents
are important for the competitive advantage of the firm, I describe them as
having high strategic stakes (Somaya, 2003), which also leads firms to
employ a proprietary strategy with these patents (Somaya, 2012).
When patents have high strategic stakes, it becomes difficult to craft
litigation settlement agreements which the non-patentee would be willing
to accept that also address the patentees interest in retaining the strategic
advantages available from such patents. Due to the substantial technological
and commercial uncertainties that are typical with patents, potential future
advantages arising from patent-based exclusivity are generally difficult to
predict and therefore specify in a licensing contract. For example, new market
480 DEEPAK SOMAYA
segments may emerge or new applications may be uncovered that are not
adequately covered by the contract. Or, new complementary resources
that the licensee develops and brings to bear may erode the patentees
rents in completely unanticipated ways. While a higher guaranteed rate of
compensation from the non-patentee may offset some of these concerns,
the non-patentee would be unwilling to enter into such an arrangement
due to concerns about the negative and uncertain impacts on its own
profitability. Along with these uncertainties, there are also substantial pro-
blems with monitoring and enforcing complex and amorphous licensing
contracts that seek to account for them. The higher the strategic stakes
associated with a litigated patent, the more acute these contracting pro-
blems become, and therefore, the more difficult and time-consuming will
it be to craft a suitable settlement to address these concerns.
Patentees are likely to make relatively unattractive settlement offers if
adequate contractual safeguards are not present, which in turn would be
rejected by their rivals in litigation. Therefore, high strategic stakes which
increase the asymmetric stakes in litigation are likely to be associated with
delays in settlement, since these contractual impediments are difficult and
time-consuming for the parties to resolve. By comparison, early-stage adjudi-
cation, which is less costly than a full trial, and also avoids the costs asso-
ciated with further litigation steps, protracted negotiations, and continuing
uncertainty, may be an attractive alternative. Moreover, the costs of ongoing
uncertainty and infringement are likely to be more significant when patentees
are relying on their patents to provide commercial exclusivity in the market-
place; in other words, when strategic stakes are high. Thus, the proclivity
toward seeking early adjudication may also be high with high-stake patents
(i.e., with a proprietary patent strategy).
H1a. The greater the strategic stakes of the patentee in the litigated
patent, the lower the hazard rate of patent litigation resolution by
private settlement.
H1b. The greater the strategic stakes of the patentee in the litigated
patent, the higher the hazard rate of patent litigation resolution by court
adjudication.
Defensive Patent Strategy

In multi-invention (or systems) product industries, such as the computer
industry, patents pose a somewhat different challenge to firms. Multi-
invention products are typically made using numerous patented inventions,
which are often owned by many different firms (Somaya, Teece, &
Wakeman, 2011). Therefore, firms in these industries are almost always
infringing each others patents, and must devise ways to operate without facing
the brunt of full-blown litigation, and potential exclusion from the market.
Because of this constant threat, firms that operate in multi-invention industries
must devise defensive patent strategies to avoid costly hold-up, which occurs
when their own prior investments and commercialization plans are put at risk
by patent enforcement threats from rival firms. Since the mid-1980s, large
portfolios of so-called defensive patents have been widely used to obtain access
to others patents in these industries.3 In semiconductors, for example, firms
have patented more intensively to build up their defensive portfolios in
response to a heightened threat of patent litigation (Hall & Ziedonis, 2001).
Strong defensive patent portfolios enable firms to counter attempted
patent enforcement with credible threats to enforce their own patents in
retaliation. Such retaliatory litigation can exclude both firms from operating
in the relevant market, resulting in a situation of mutual hold-up. The high
cost of mutual hold-up for both parties induces reciprocal access to patents,
and thus constitutes the underlying logic behind defensive patenting. In
some instances, this implicit mutual threat has resulted in tacit truces, with
neither side suing the other (Von Hippel, 1988), whereas in others it has led to
the conclusion of formal cross-licenses to ensure commercial freedom
(Grindley & Teece, 1997).
In the context of patent litigation, the non-patentees ability to threaten
(and follow through on) a mutual hold-up defensive strategy makes it very
costly for both sides to continue with litigation. These costs are manifested
both in increased commercial uncertainty and in the risks of potential mar-
ket exclusion for both firms. Given these high costs, the parties experience a
greater urgency to settle patent litigation quickly. Moreover, these high costs
are also relevant for the speed with which the parties seek adjudication.
When the parties in patent litigation find themselves in a situation of mutual
hold-up, adjudication may also be pursued more actively: either because
settlement negotiations are at an impasse, or because a favorable quick deci-
sion can work to a firms advantage in obtaining attractive terms during
post-adjudication cross-licensing. In either case, mutual hold-up raises the
time discount rate for continuing with litigation, and thus creates pressures
for both parties to conclude adjudication and settlement quickly.
The arguments about mutual hold-up apply primarily to patent litiga-
tion in multi-invention product industries, such as the computer industry in
the current study. Although computer patents are often no longer cutting
edge by the time they issue, ensuring commercial freedom is of critical
482 DEEPAK SOMAYA
importance for firms in the industry. Patents are highly valued as tools to
negotiate licenses with other firms and to counter patent suits brought
by them (Blind et al., 2006; Cohen et al., 2000). While innovation in
biotechnology-based research inputs has taken on more of a multi-invention
character over time (Somaya et al., 2011; Walsh, Arora, & Cohen, 2003), the
time frame of this study (19831993) predates that technological shift.
H2a. The greater the ability of rivals to effect defensive patent strategies
based on mutual hold-up the higher the hazard rate of patent litigation
resolution by private settlement.
H2b. The greater the ability of rivals to effect defensive patent strategies
based on mutual hold-up the higher the hazard rate of patent litigation
resolution by court adjudication.
Leveraging Patent Strategy

The essence of leveraging patent strategies is that patentees can use the
bargaining power conferred by the exclusionary rights granted by patents
to appropriate rents and/or gain other strategic benefits from putative
infringers. In patent litigation, there are clearly ongoing costs in continuing
with litigation, and it is likely that some parties such as smaller, poorly
financed firms face higher costs than others, reflecting their relative
bargaining strength. In turn, leveraging strategies that are employed by
firms with greater bargaining strength can have an influence on the timing
of litigation resolution through their effect on bargaining tactics between
the parties. Formal models in the bargaining literature have found that
delays in making or responding to offers can be used as a strategic device
to signal relative bargaining strength (Admati & Perry, 1987; Cramton,
1992). Thus, firms can obtain better settlement offers in settlement negotia-
tions by using delaying tactics that are more costly for their rivals than
themselves. When the court adjudication timeline is treated as exogenous,
formal bargaining models predict that parties with bargaining strength will
delay settlement initially in order to force settlement later, under deadline
pressure, with offers that favor themselves (Ma & Manove, 1993). At the
outset, therefore, bargaining models appear to suggest that when patentees
have greater bargaining strength relative to their rivals in patent litigation,
which reflects a greater likelihood of employing leveraging strategies, the
timing of settlement is likely to be delayed.
Turning our attention to the timing of adjudication, however, Ma and
Manove (1993) also show that when settlement deadlines are endogenous,
firms with bargaining strength seek earlier deadlines so as to put their rivals
under deadline pressure. Consistent with this prediction, prior research has
found that (early-stage) preliminary injunctions, which create early dead-
line pressures, are more likely to be used in patent litigation with financially
distressed firms (Lanjouw & Lerner, 2001) against whom patentee firms
have a bargaining advantage. Thus, we would expect that the use of a lever-
aging strategy, which is more likely when a patentee has a bargaining
advantage relative to its rival in litigation, it will likely speed up the timing
of adjudication.
H3a. The greater the relative bargaining strength of the patentee in
patent litigation, the lower the hazard rate of patent litigation resolution
by private settlement.
H3b. The greater the relative bargaining strength of the patentee in patent
litigation, the higher the hazard rate of patent litigation resolution by
court adjudication.
DATA AND METHODS

I use data on patent suits in computers and research medicines filed between
1983 and 1993 in the U.S. federal district courts. Data from this time period
are particularly attractive for studying patent litigation because of the rela-
tive policy stability during this period, as well as the relative stability of the
nature of innovation in the two industries.4 The litigation data are compiled
from two sources the Federal Judicial Center (FJC)5 and the U.S. Patent
and Trademark Office (USPTO). Of these two sources, only the USPTO
dataset has the patent number(s) of the patents being litigated, but it consists
of only a sample of patent suits (comprising 5058% of suits filed in each
year). The FJC data has more complete coverage (nearly the universe of
patent suits) and more complete information about the dates and outcomes
of the suits. I take advantage of the respective strengths of the two data
sources by matching patent suits between them, using a number of different
common fields with a success rate of over 95% (of the USPTO records).
Data on patent characteristics, and the patent portfolios of firms from the
Case WesternNBER database were also collected (see Hall, Jaffe, &
Trajtenberg, 2001).6
I narrowed my sample to patent suits in two highly innovative industries
in the economy where patent strategy is likely to be a significant success
484 DEEPAK SOMAYA
factor for firms computers and research medicines (pharmaceuticals plus

medical biotechnology). Both R&D expenditures and patenting in the two
industries sum up to around 1520% of these activities in the U.S. economy
as a whole, highlighting the importance of these industries to overall innova-
tion in the economy. The primary reason for using a two-industry sample is
to allow for differences in patent strategy between industries, as anticipated
by the hypotheses about mutual hold-up strategies and by prior survey
research about inter-industry differences in patent use (e.g., Blind et al.,
2006; Cohen et al., 2000; Duguet & Kabla, 1998; Somaya, 2003). The
two-industry sample was constructed by employing a set of representative
patent classes and subclasses for each industry, which in turn were identified
through patents owned by pure-play firms in the respective industries (based
on within-industry sales as reported in Compustat).7 Since this approach
requires information about the litigated patents (gathered by identifying the
patent numbers and then linking out to other data), the focal suits in the
sample were identified by using the USPTO dataset.
Each litigated patent thus identified was manually checked to ensure that
it belonged to one of the two industries being studied. All suits in the sample
were consolidated into litigation units by combining multiple suits that
were sometimes filed between the same parties over the same patent around
the same time (in different courts), and all further analyses were conducted
at the litigation unit level. This unit-of-observation also reflects the practical
reality that these multiple suits were eventually consolidated (in every case)
into a single proceeding at one venue. After accounting for missing data in
some observations, 607 litigation units were identified, of which 369 were in
computers and 238 in research medicines. Consistent with prior research
(Lanjouw & Schankerman, 2001; Somaya, 2003), when multiple patents
are involved in litigation, the first patent listed in the litigation was used for
analysis, and the number of patents in the litigation was coded for use as a
control variable.
Dependent Variable
Since I am interested in examining the speed with which patent litigation is

terminated, I use a duration or hazard model, where the dependent vari-
able is the length of time or spell for which the observation (litigation)
stays in the same state. The likelihood that the litigation exits this state in
any given time interval is the hazard rate. Formally, a hazard rate (t) is
defined as the probability that an observation will end in a time interval t,
given that it has survived till time t, as t0. Thus, the hazard rate can
be expressed in terms of the density function f(t) and cumulative density
functions F(t) of duration as follows:
t f t = 1 Ft 1
I measure the length of spells in litigation in days by using the filing and
closing dates of suits.
The independent variables used in the analyses are developed by translating

each of the three generic strategies into one or more distinct measures, each
of which indicates conditions under which these strategies are more (or
less) likely to be employed. A list of all the variables used in this paper, and
their definitions, are included in Table 1.
Proprietary Strategy: Patentee Stakes and Patent Age

The use of proprietary strategies with patents entailing high strategic
stakes is measured by using two characteristics of the litigated patent: self-
citations (by the patentee) to the patent and the age of the patent. Self-citations
reflect the extent to which the patentee has built on its own technology, and
thus measure the related technological and commercial assets that may be
isolated by the litigated patent. Although self-citation has been widely used in
research to indicate internal spillovers and knowledge transfer within the firm,
it is used here in a manner consistent with another common usage to proxy
the firms resource investments in the domain of the patent and the firms
competitive strength or relative position (e.g., McMillan & Hamilton, 2000;
Somaya, 2003). I label this self-citation variable Patentee Stakes, because it
reflects the extent to which the patentee has built a strategic stake in the
litigated patent. Patentee Stakes is measured as the ratio of self-citations to the
overall level of citations to the patent. The intent is to orthogonalize this
variable relative to overall patent citations, which is included separately as a
control variable (see below). A second proxy for proprietary strategy is Patent
Age. Although Patent Age is not itself a strategic choice, it is likely to affect the
strategic choices of patentees because as patents reach the end of their life the
advantages from (and need for) a proprietary strategy decline. By contrast,
when Patent Age is low and patents are farther away from expiration, it is
harder to contract over the potential benefits yielded by its strategic isolation
486 DEEPAK SOMAYA
Table 1. List of Variables Used and Their Hypothesized Impact on the

Hazard Rate of Patent Litigation Resolution (by Settlement/by
Adjudication).
Proprietary strategy (strategic stakes) variables:
Patentee stakes Ratio of patentee self-citations of the litigated patent to all
citations (/+)
Patent age Age in years from patent issue date to the date when suit is filed (+/)
Defensive strategy (mutual hold-up) variables:
Portfolio citations Citations from the patentees patents to the non-patentees patents divided
by the number of patentees patents (for the 8 years prior to suit) (+/+)
Counter-suit A dummy variable (= 1 if a counter-suit is simultaneously being pursued
by the non-patentee, using its own patent) (+/+)
Leveraging strategy (bargaining strength) variables:
Size ratio The ratio of the number of patents in the non-patentees patent portfolio
to that in the patentees portfolio (/+)
Control variables:
Citations The number of citations to the patent (estimated for 17 years from issue)
Firm size The number of patents in the patentees portfolio
Court congestion The ratio of the duration of an average patent suit that went to trial in a
given district to the average suit that went to trial in all districts
Issue lag The lag in years from application date to issue date for the litigated patent
Same-party suits The number of patent suits involving either litigant in a 10-year window
centered around the litigation event
Same-patent suits The total number of suits filed by the patentee over the same patent
Individual A dummy variable (=1 if the patent is being litigated by an individual)
Foreign A dummy variable (=1 if the patentee is based in a foreign country)
Time Time trend variable with 1/1/1983 = 0, and each year = 1 unit, based on
suit filing date
of rent-yielding assets far into the future, which (following our theoretical
logic) would delay settlement and encourage early-stage adjudication.
Defensive Strategy: Portfolio Citations and Counter-Suit

A defensive strategy based on mutual hold-up is underpinned by the ability
of non-patentees to use their own patents against rivals in litigation.
However, simply having a large patent portfolio may not be an effective
defense against all patent enforcers (Somaya, 2012). For example, some
potential opponents may be niche specialists (Hall & Ziedonis, 2001) who
commercialize their narrow technology through licensing or component
markets (Linden & Somaya, 2003). Such firms are not engaged in building
entire integrated systems products in-house, and have no need for access to
broad portfolios of patents. By contrast, if the patentees own patents cite
the non-patentees patents extensively, this indicates that the latters patents
can potentially hold up the patentee as well and function in a defensive
role. I measure these Portfolio Citations as the number of citations per
patentee patent to patents in the non-patentees patent portfolio (using
only patents obtained in the 8 years immediately before litigation to ensure
relevance to the patentees current technologies).
When faced with litigation in multi-invention contexts, firms with defensive
portfolios can retaliate with counter-suits by using their own patents (Somaya,
2003). These counter-suits constitute the actual execution of a mutual hold-up
threat by non-patentees as a part of their defensive strategy, which I measure
using a dummy variable that indicates if a Counter-Suit is being litigated
simultaneously with any focal patent litigation. This variable is time-varying
within litigation so that it measures the impact of a counter-suit only after
one is filed.8 Both Counter-Suit and Portfolio Citations are only relevant for
defensive purposes in a multi-invention (or systems) industry such as the
computer industry, as discussed in the development of hypotheses earlier.
It comes as little surprise then that Counter-Suit equals zero for all research
medicine litigation in the data. Furthermore, Portfolio Citations affects the
timing of suit resolution (at statistically significant levels) only for the computer
litigations in the sample (as discussed below in the Results section).
Leveraging Strategy: Size Ratio

A leveraging patent strategy is associated in this study with relative bargaining
strength, which is measured using the ratio of the firms patent portfolios as a
proxy. It is essentially impossible to get reliable accounting data for many
firms in patent litigations within my sample (such as small, private, or foreign
firms), which necessitates the use of some other source of data to estimate
relative bargaining strength. Therefore, I measure relative bargaining strength
by the Size Ratio of the non-patentees patent portfolio to the patentees
patent portfolio. In addition to being easily measured by using public patent
data, the sizes of rival firms patent portfolios are also likely to more
accurately capture (relative to accounting measures of firm size) the relative
resources that the firms have deployed in the patent arena.
Control Variables
In addition to the key independent variables described above, I also use a

number of control variables. For those control variables that are highly
skewed the variables are logged, which also results in better model fit
488 DEEPAK SOMAYA
(measured by the Bayesian Information Criterion). The overall level of

patent citations may be correlated with the timing of litigation resolution
because it is a well-recognized measure of patents technological impor-
tance and overall value (Hall, Jaffe, & Trajtenberg, 2005; Jaffe, Fogarty, &
Banks, 1998; Trajtenberg, 1990a, 1990b); therefore I control for (log)
Citations. I also introduce controls for (logged) Firm Size (measured using
patents), and for the (logged) delay in the patent office between application
and issue of the litigated patent (Issue Lag) (Somaya, 2003). I control for
differences in the administrative speed with which different federal court
districts deal with patent litigation by using a variable labeled Court
Congestion, which measures the mean duration of patent suits till trial in
any given district, divided by the mean of the variable across all districts.
For example, the value of Court Congestion for the Eastern District of
Virginia, which was known to be a rocket docket during the time period
of this sample, was 0.36 i.e., this district took only 36% of the time taken
by an average district for patent suits that went to trial. I also control for
the (log) number of patent suits involving the same parties (over a 10-year
window), or involving the same patent as the focal litigation, labeled Same-
Party Suits and Same-Patent Suits, respectively. These variables account for
the possibility that the speed of litigation resolution in any individual dispute
may be influenced by interactions with other related patent disputes, including
through reputation effects (Siegelman & Waldfogel, 1996; Waldfogel, 1995).
Finally, I control for a linear time trend (Time) to capture changes in litigation
practices over time, and for cases in which the patentee was an Individual or a
Foreign entity, which have been shown to affect settlement in patent litigation
(Somaya, 2003). I expect Citations, Court Congestion, and Same-Party Suits
to decrease the hazard rate of settlement, and for Court Congestion to decrease
the hazard rate of adjudication. Same-Patent Suits are likely to increase the
hazard rate of both settlement and adjudication in a manner analogous to
the defensive strategy variables.
A number of other controls were also used in unreported estimations,
which did not significantly improve the performance of the empirical model.
These variables include alternative measures of patent value (number of
claims, patent classes, and backward citations), as well as controls for unique
features of the litigation (number of defendants, number of patents, and a
dummy variable for suits filed by the non-patentee). Because patent citation
data are only available for a limited number of years, Citations and Patentee
Stakes may be biased due to truncation of the citation series. Therefore,
I estimate the lifetime (17 year) citations to each litigated patent by dividing
the truncated citations measured by the average fraction of lifetime citations
that occur within an industry for the same duration that the patent is at risk
of citation (see Hall et al., 2005). For example, if computer patents accumulate
70% of their citations over the first 10 years on average, and a specific
computer patents citation series is truncated after 10 years from grant,
then I divide the measured number of citations for this patent by 0.7 to
obtain an unbiased measure of its lifetime citations. I follow prior research
in using patent citations as a proxy for underlying attributes of a patent,
which are known at the time of litigation even if the citations themselves
have not yet occurred (Hall et al., 2005).
Empirical Model
Statistically, I model the timing of patent litigation resolution using a

hazard model with competing risks, where either settlement or adjudication
can terminate the litigation so long as the other has not already done so
(Hughes & Savoca, 1999). One common approach to hazard models is to
specify a functional form for time-dependence of the hazard rate; for example,
a constant hazard rate, which would imply that litigation durations are
distributed exponentially. Despite some parametric flexibility, this approach
inherently entails specific assumptions about the time-dependence of the
hazard function, which may be too restrictive for its use in modeling (patent)
litigation. In particular, I expect the courts adjudication schedule (time limits
set by the judges) to impose constraints on the parties ability to further delay
dispute resolution after a patent suit has been running for some time. Prior
research on the duration of litigation supports these expectations, which is
manifested in findings of increasing time-dependence of the hazard rate (Fenn &
Rickman, 1999; Fournier & Zuehlke, 1996; Hughes & Savoca, 1999).
While these prior studies have adopted modified hazard functions to deal
partially with this problem, I follow Kessler (1996) in using the proportional
hazards model and the associated estimation methods pioneered by Cox
(Cox, 1972; Lancaster, 1990; Petersen, 1995). This model decomposes the
hazard function into two parts as follows:
t exp 0 X0 t 2
where the time-varying part of the function, 0(t), is not specified in

advance, and is thus non-parametric and estimated from the data, which
provides the desired flexibility for the hazard rate to change (and in parti-
cular increase) over time. The model is proportional in the sense that the
490 DEEPAK SOMAYA
hazard rates of two observations have the same ratio over time if their
independent variables do not change. One reason proportional hazards
models are attractive is that violation of the proportionality condition is not
a major misspecification concern (Allison, 1984). The parametric component
of the model is estimated by Coxs partial likelihood method, which then
enables estimation of the baseline hazard function. I stratify the model by
industry, which estimates separate baseline hazards for each industry, and
compute robust standard errors to address the potential inconsistency of
standard errors arising from model misspecification.
RESULTS
Descriptive sample statistics for the sample used in this study are reported
in Table 2. Of the 607 litigations in the sample, 417 end in settlement, while
180 are resolved by court adjudication. Ten litigations are censored with no
further information about their outcome. One variable, Counter-Suit, can
be time-varying over the course of litigation (whereas all other variables are
time-invariant), which results in 17 litigations being broken up into two
Table 2. Descriptive Statistics of Key Variables.

Variables No. Mean Std. Min. Max. Predicted Sign
Obs. Dev.
Settlement Adjudication
Suit duration (days) 607 517 483 7 2,805

(1) Patentee stakes 607 0.31 0.29 0.00 1.00 +
(2) Patent age 607 6.61 5.39 0.00 21.31 +
(3) Portfolio citations 607 0.02 0.11 0.00 1.47 + +
(4) Counter-suit 607 0.05 0.22 0.00 1.00 + +
(5) Size ratio 607 2.81 2.77 0.10 10.04 +
(6) Log citations 607 3.42 1.04 0.00 7.04
(7) Log firm size 607 4.10 2.88 0.00 9.17
(8) Court congestion 607 1.05 0.25 0.36 1.79
(9) Log issue lag 607 1.19 0.33 0.40 2.94
(10) Log same- 607 3.14 1.02 0.00 5.42
party suits
(11) Log same- 607 0.83 1.12 0.00 3.66
patent suits
(12) Individual 607 0.08 0.28 0.00 1.00
(13) Foreign 607 0.13 0.34 0.00 1.00
(14) Time 607 6.65 3.07 0.22 12.48
spells and results in a total number of 624 observation spells. The mean
and median of litigation duration spells are 517 and 372 days, respectively.
Correlations between variables in the sample suggest few problems with
multi-collinearity (Table 3).
Parameter estimates from the Cox hazard rate models for private settle-
ment and court adjudication are reported in Tables 4 and 5, respectively.
Because the propositions advanced in this paper predict how each patent
strategy impacts both litigation resolution mechanisms settlement and
adjudication the results from Tables 4 and 5 are discussed together, which
also facilitates comparison between the findings. For ease of interpretation,
hazard ratios and the corresponding standard errors are reported, instead of
coefficient estimates.9 Hazard ratios are simply the exponentiated coefficients
and have an intuitive multiplicative interpretation. Hazard ratios that
are less than (greater than) the value 1.0 decrease (increase) the overall
hazard rate, and thus correspond to variables that slow down (speed up) the
resolution of litigation.
The results are broadly consistent with the hypothesized impact of
proprietary patent strategies as measured by Patentee Stakes and Patent
Age on the timing of patent litigation adjudication (H1b), but not settle-
ment (H1a). The coefficient of Patentee Stakes has the correct hypothesized
sign and is significant at the 5% level of significance in the adjudication
models, but it is only significant at the 10% level in the settlement models
(with the right sign). The coefficient of Patent Age also has the right sign,
but it is only statistically significant (at the 1% level) in the model for the
timing of adjudication. Thus, the results indicate that Patentee Stakes and
Patent Age are significantly associated with speeding up and slowing down
adjudication, respectively. Overall, proprietary strategies appear to impact
the timing of adjudication much more strongly than settlement, which sug-
gests that prior findings about the higher propensity for non-settlement in
patent litigation with proprietary strategies (Lanjouw & Schankerman,
2001; Somaya, 2003) may come about because of firms seeking quicker
adjudication rather than by slowing down settlement.
Hypotheses 2a and 2b suggest that defensive patent strategies, as
measured by Portfolio Citations and the filing of Counter-Suits, will be
associated with higher hazard rates of both settlement and adjudication.
Although both defensive strategy variables have the hypothesized signs and
are significant at the 10% level or higher in both sets of models, each vari-
able is only significant at conventional statistical levels in one set of models.
Specifically, Portfolio Citations has a statistically significant (at 1%) asso-
ciation with the hazard rate of settlement (H2a), whereas Counter-Suit is
492
Table 3. Correlation Table.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14)
(1) 1.00
(2) 0.02 1.00
(3) 0.10 0.08 1.00
(4) 0.01 0.05 0.16 1.00
(5) 0.25 0.22 0.14 0.11 1.00
(6) 0.35 0.30 0.09 0.03 0.06 1.00
(7) 0.39 0.32 0.07 0.08 0.59 0.01 1.00
(8) 0.09 0.02 0.10 0.03 0.02 0.07 0.06 1.00
(9) 0.04 0.13 0.05 0.04 0.15 0.08 0.02 0.06 1.00
(10) 0.23 0.32 0.05 0.12 0.22 0.04 0.52 0.06 0.04 1.00
(11) 0.13 0.31 0.06 0.12 0.44 0.11 0.25 0.04 0.10 0.37 1.00
(12) 0.30 0.22 0.06 0.07 0.25 0.27 0.40 0.14 0.15 0.09 0.05 1.00
(13) 0.01 0.01 0.04 0.02 0.07 0.16 0.13 0.09 0.00 0.00 0.10 0.08 1.00
(14) 0.05 0.06 0.07 0.11 0.08 0.18 0.05 0.03 0.11 0.03 0.19 0.04 0.09 1.00
DEEPAK SOMAYA
Table 4. Cox Proportional Hazard Model Estimates for the Resolution

of Patent Litigation by Private Settlement.
Variables Model 1 Model 2 Model 3 Model 4 Model 5
Hazard ratio Hazard ratio Hazard ratio Hazard ratio Hazard ratio
(Std. err.) (Std. err.) (Std. err.) (Std. err.) (Std. err.)
Patentee stakes 0.701* 0.667*

(0.150) (0.142)
Patent age 1.006 1.008
(0.013) (0.013)
Portfolio citations 3.445*** 3.914***
(0.920) (1.081)
Counter-suit 1.444* 1.511*
(0.321) (0.341)
Size ratio 1.038 1.053**
(0.026) (0.028)
Log citations 0.858*** 0.894* 0.847*** 0.853*** 0.879**
(0.045) (0.052) (0.044) (0.045) (0.053)
Log firm size 1.000 1.007 1.003 0.981 0.985
(0.022) (0.024) (0.023) (0.025) (0.027)
Court congestion 0.552*** 0.537*** 0.546*** 0.565** 0.542***
(0.124) (0.124) (0.119) (0.128) (0.123)
Log issue lag 0.977 0.969 0.996 0.996 1.015
(0.152) (0.151) (0.155) (0.156) (0.160)
Log same- 0.868** 0.868** 0.848*** 0.887** 0.872**
party suits (0.048) (0.048) (0.047) (0.051) (0.051)
Log same- 1.192*** 1.190*** 1.224*** 1.166*** 1.186***
patent suits (0.066) (0.069) (0.069) (0.067) (0.072)
Individual 0.902 0.843 0.949 0.906 0.880
(0.192) (0.195) (0.204) (0.193) (0.208)
Foreign 1.045 1.060 1.044 1.019 1.027
(0.171) (0.171) (0.170) (0.170) (0.168)
Time trend 1.081*** 1.081*** 1.081*** 1.080*** 1.079***
(0.018) (0.018) (0.018) (0.019) (0.018)
No. of subjects 607 607 607 607 607
No. of obs. (spells) 624 624 624 624 624
No. of failures 417 417 417 417 417
Log likelihood 2003.0 2001.3 1998.1 2002.0 1994.2
*, **, and *** denote .10, .05, and .01 levels of significance, respectively.
Baseline hazards are stratified by industry and standard errors are robust.
significantly associated (at 5%) with the hazard rate of adjudication (H2b).
These variables are measured for the sample as a whole in the reported
results; however, their effects come about due to their impacts in Computer
patent litigation. Specifically, the Counter-Suit dummy variable is non-zero
494 DEEPAK SOMAYA
Table 5. Cox Proportional Hazard Model Estimates for the Resolution

of Patent Litigation by Court Adjudication.
Variables Model 1 Model 2 Model 3 Model 4 Model 5
Hazard ratio Hazard ratio Hazard ratio Hazard ratio Hazard ratio
(Std. err.) (Std. err.) (Std. err.) (Std. err.) (Std. err.)
Patentee stakes 1.971** 2.104**

(0.637) (0.677)
Patent age 0.936*** 0.937***
(0.020) (0.020)
Portfolio citations 3.849* 4.802*
(3.129) (3.787)
Counter-suit 2.352** 2.518**
(0.832) (0.913)
Size ratio 1.177*** 1.199***
(0.047) (0.049)
Log citations 1.097 0.945 1.078 1.071 0.903
(0.094) (0.093) (0.091) (0.088) (0.088)
Log firm size 0.982 1.004 0.986 0.887*** 0.901**
(0.032) (0.038) (0.031) (0.036) (0.042)
Court congestion 0.543* 0.586 0.554* 0.596 0.632
(0.183) (0.202) (0.186) (0.205) (0.222)
Log issue lag 0.599* 0.657 0.609* 0.709 0.776
(0.160) (0.180) (0.166) (0.178) (0.204)
Log same- 1.070 1.083 1.037 1.189* 1.171
party suits (0.100) (0.106) (0.098) (0.115) (0.120)
Log same- 1.617*** 1.767*** 1.672*** 1.384*** 1.521***
patent suits (0.123) (0.135) (0.129) (0.111) (0.125)
Individual 0.950 1.555 1.039 0.897 1.604
(0.342) (0.595) (0.379) (0.307) (0.590)
Foreign 1.055 0.952 1.049 0.948 0.817
(0.243) (0.232) (0.239) (0.226) (0.209)
Time trend 1.109*** 1.109*** 1.106*** 1.099*** 1.098***
(0.031) (0.031) (0.031) (0.030) (0.031)
No. of subjects 607 607 607 607 607
No. of obs. (spells) 624 624 624 624 624
No. of failures 180 180 180 180 180
Log likelihood 789.53 781.50 786.05 780.67 768.25
*, **, and *** denote .10, .05, and .01 levels of significance, respectively.
Baseline hazards are stratified by industry and standard errors are robust.
only for Computer litigation observations, and the results are very similar
if the Portfolio Citations variable is only coded and employed for
Computer (but not Research Medicine) observations.10 Analyses within the
Computer sub-sample also yielded similar results to those reported here.
Overall, these findings provide a nuanced understanding of how the defen-

sive strategy of mutual hold-up operates in the context of patent litigation,
and help to explain the lack of empirical support for the relationship
between defensive strategies and settlement of patent litigation in prior
work (Somaya, 2003). Specifically, because mutual hold-up is associated
with increases in the speed with which both settlement and adjudication are
sought, its net impact on settlement outcomes can be ambiguous.
The findings regarding the association between leveraging strategies and
the timing of patent litigation resolution are also mixed. The coefficient of
Size Ratio in the settlement models has the wrong sign (indicating faster
resolution by settlement), and is statistically significant in the full model (at
the 5% level). Thus, the evidence appears to contradict the bargaining
strength signaling hypothesis (H3a). By contrast, the coefficients of size
ratio in the adjudication models are the right sign and are statistically sig-
nificant at the 1% level, which is consistent with the hypothesis that firms
with relative bargaining strength might bring to bear the deadline pressures
of speedier adjudication as part of a patent leveraging strategy.
The results also indicate some interesting statistically significant associa-
tions between the hazards of litigation resolution and some of the control
variables. Same-Patent Suits are positively and significantly associated with
the hazard rates of both settlement and adjudication, indicating that when
patents are enforced against multiple parties, it broadly acts as a spur to
speedy resolution.11 By contrast, settlement (but not adjudication) is slower
in the presence of other Same-Party Suits, indicating that these firms may be
seeking reputations for being tough in patent litigation, as suggested by prior
research (Siegelman & Waldfogel, 1996; Waldfogel, 1995). The Court
Congestion variable has a statistically significant impact on the timing of
settlement, but not necessarily adjudication. Citations to the litigated patent
decrease the hazard rate for settlement (indicating a reluctance to settle
litigation over these important patents), but has no significant impact on the
speed of adjudication. In addition, the Time Trend variable indicates that
the hazard rate of litigation resolution through both settlement and adjudi-
cation increased over the time period of this study, which is consistent with
the long-term decrease in patent litigation durations in the United States
(Somaya, 2004). Finally, the estimated baseline cumulative hazard functions
for both settlement and adjudication (not reported) show an increasing
time-dependence, as predicted by the deadline effect of court time-tables in
theoretical models of settlement timing (Ma & Manove, 1993; Spier, 1992)
and found in prior empirical work (Fenn & Rickman, 1999; Fournier &
Zuehlke, 1996; Hughes & Savoca, 1999).
496 DEEPAK SOMAYA
DISCUSSION
As patents become increasingly important to companies competing in the
knowledge-based economy, patent strategy has become a matter of great
relevance to top management (Rivette & Kline, 2000). Consequently, there
is significant value in developing approaches to patents that are strategic,
managerial, and holistic. In this spirit, the current paper has investigated
the implications of three generic patent strategies for the timing of adjudi-
cation and settlement in patent litigation. By employing an extension of
asymmetric stakes theory from law and economics, and drawing on prior
work regarding strategic signaling and deadline setting in the bargaining lit-
erature, hypotheses were developed about the timing of both settlement
and adjudication. The empirical findings are consistent (at statistically sig-
nificant levels) with some predictions of the theoretical hypotheses, suggest-
ing that proprietary, defensive, and leverage strategies may be important
factors in how firms conduct their patent litigation activities.
The specific phenomenon studied in this paper points to a larger set of
questions within non-market strategy that have been largely ignored and
might benefit from further research. Specifically, the extant non-market
strategy literature has largely been interested in understanding how firms
influence political institutions to shape the rules of the game under which
they compete in the market (De Figueiredo, 2009). However, these rules
may define the firms non-market environment only at a broad abstract
level, and further actions may be needed to access and apply these rules to
specific situations, products, or assets (including technological assets) that
matter to the firm. For example, while legislation and case law determine
the broad rules of the game in patent policy, firms must decide when and
to what extent they wish to avail of these policy provisions for specific
inventions. In other words, firms must determine which non-market actions
to undertake for a specific business situation by choosing among alterna-
tives within the broad policy umbrella available to them. This paper has
focused on explaining a particular type of non-market action, namely the
speed and method by which firms resolve patent litigation, which is shown
to be influenced by their associated market strategy. Similar types of
non-market actions are likely to be widespread in other areas of policy as
well, and future research may seek to further investigate the drivers and
consequences of such actions. Inter alia, such a research program will likely
involve substantial consideration of the market strategies of firms, and thus
represent a new direction for integrated strategy research as well.
While the empirical evidence uncovered in this study is consistent with

some of the proposed hypotheses, it also has some limitations. The data
employed in the analyses was restricted to two industries and a specific time
period, which may limit the validity of the findings in other contexts. The
analyses are also limited to the sample of litigation reported to the USPTO,
and there may be systematic differences between this sample and the larger
population of patent lawsuits (despite prior research indicating that the
USPTO sample is not biased in its measurable attributes Lanjouw &
Schankerman, 2001). Further, only some patent disputes end up in litigation,
and it would therefore be appropriate to interpret the findings as being
applicable specifically to patent disputes in litigation (rather than all patent
disputes). Last but not least, the empirical evidence provided in this paper
consists essentially of associations between variables, from which it is
difficult to make strong causal claims. Thus, future research may build on
the current study by exploiting exogenous shocks or variations for
example, changes in patent law or the composition of court judges may
exogenously make it easier or more difficult to pursue certain generic patent
strategies that help to better identify causal relationships between firm
strategies and the timing of litigation resolution.
Subject to these caveats, the results of this study illustrate the value of trying
to understand (patent) litigation outcomes as resulting jointly from settlement
and adjudication mechanisms. As research on litigation gains ground in
strategy, this approach promises to provide a deeper understanding of
litigation than studies that focus only on one-shot litigation outcomes. This
paper introduced an informal extension of asymmetric stakes theory (Priest &
Klein, 1984) to make predictions about the timing of both settlement and adju-
dication, which has been difficult to do in formal models based on asymmetric
information theory alone (Spier, 1992). Furthermore, while the empirical
literature on the timing of litigation resolution has used proportional hazards
(Kessler, 1996) and competing risks (Hughes & Savoca, 1999) models
separately, this paper demonstrates the advantages of combining the two
approaches to develop insights into the drivers of litigation outcomes. Last but
not least, the empirical findings suggest a more significant role for the hazard
of adjudication (which is more strongly associated with the main independent
variables than the hazard of settlement) in determining settlement outcomes in
litigation. In contrast with conventional law-and-economics theory regarding
non-settlement (Bebchuk, 1984; Priest & Klein, 1984; Spier & Spulber, 1993),
which emphasizes the difficulties of reaching settlement, these results suggest
that more research is needed to understand the factors that drive the ability
498 DEEPAK SOMAYA
and willingness of disputing parties to influence the timing of adjudication

in litigation.
The current paper also makes a number of contributions to the research
literature on patent strategy. Prior work on the non-market arena in patent
strategy is quite limited, and this study points to the significant opportu-
nities that remain to be tapped. While the emphasis here has been on firms
decision making regarding their choices in litigation, which are in turn
influenced by their (market) patent strategies, future research may seek to
develop a deeper understanding of firms non-market capabilities in the
patent arena (e.g., Somaya & McDaniel, 2012). For example, while this
paper has made the case for specific optimal tactics in patent litigation, it
is likely that firms non-market capabilities both generically in patent
litigation, as well as in specific litigation venues (courts) may influence
the effectiveness with which these tactics are pursued. Additionally, future
research may also investigate the impact of patent litigation tactics on firm
performance and long run competitive advantage. The logic advanced in
this paper would suggest that litigation tactics which are well aligned with
the firms generic patent strategies will have more positive long run benefits
for the firm. For example, if a firm quickly settles patent litigation related
to a high strategic stakes patent, for which the firm should ideally pursue a
proprietary patent strategy, then this type of action may undermine the
firms long run competitive advantage within the technological domain
of the patent. Future research may leverage stock market reactions to
litigation resolution events to assess such long run performance impacts of
well-aligned (versus poorly aligned) patent litigation tactics.
Within the context of this papers research, its empirical results help to
explain and amplify a number of prior findings by unpacking the competing
resolution mechanisms of adjudication and settlement that drive settlement
outcomes in patent litigation. Specifically, in prior work, the use of proprietary
strategies as measured by Patentee Stakes and Patent Age has been found
to be associated with the non-settlement of patent litigation (Lanjouw &
Schankerman, 2001; Somaya, 2003). The current research shows that this
increased propensity to not settle patent litigation when firms are pursu-
ing a proprietary strategy stems primarily from a higher hazard rate of
adjudication rather than delays in settlement negotiations between the
parties. This insight is particularly important given the predominant
focus on barriers to settlement as the primary mechanism in canonical
law-and-economics models of patent litigation (Lanjouw & Lerner, 1998;
Lanjouw & Schankerman, 2001).
Similarly, the papers findings also shed light on how defensive strategies
affect settlement outcomes, which have been found to have no significant
empirical relationship in prior research (Somaya, 2003). The results

reviewed here show that the variables measuring defensive strategies are
associated with a higher hazard rate of litigation resolution by both settle-
ment and adjudication (albeit at a 10% level of significance in some cases).
These competing settlement and adjudication mechanisms appear to work
against each other in affecting settlement outcomes. Although the settle-
ment of litigation is speeded up by defensive strategies, faster adjudication
can also interpose itself and lead to non-settlement outcomes. Ultimately,
the current research shows that even if defensive strategies based on mutual
hold-up have no measurable impact on the likelihood of settlement in
patent litigation (Somaya, 2003), they are not irrelevant and can manifest
themselves in the speed with which patent litigation is resolved.
This papers empirical findings also illuminate the use of patent enforce-
ment as a component of patent leveraging strategies by firms that have rela-
tively stronger bargaining positions versus rivals in the marketplace. While
patent leveraging strategies often play out in the context of licensing nego-
tiations, the option to pressure potential licensees through litigation actions
is clearly an important threat point under which these negotiations occur
(Lanjouw & Lerner, 2001). The results reviewed above suggest that such
leveraging strategies are more strongly associated with the strategic use of
court procedures to speed up adjudication (Lanjouw & Lerner, 2001; Ma &
Manove, 1993), and not with delaying tactics to signal strength to rivals
and induce compromises from them (Admati & Perry, 1987; Cramton,
1992). Again because relative bargaining strength is associated with higher
hazard rates of both adjudication and settlement, its net effect on settle-
ment outcomes is ambiguous.
Enforcement of intellectual property through the courts is a critical aspect
of the value and strategic use of intellectual property by firms. Patent
enforcement sets the context conditions under which other aspects of patent
strategy, including the obtaining and licensing of patents, are pursued
(Somaya, 2012). Thus, although this paper has investigated the implications
of patent strategy for a specific question within patent enforcement, its
findings shed light into the strategic management of patents by firms more
generally and its implications for firms actions in the non-market arena.
NOTES
1. Formally, the bargaining surplus would be the litigation costs C + c, which

would be incurred by the plaintiff and defendant, respectively. With asymmetric
stakes, the plaintiff may expect a transfer wT from the suit, which is larger than the
500 DEEPAK SOMAYA
expected transfer away from the defendant wt. Here, w is the probability of a
plaintiff win, and T and t are the transfers to the plaintiff and away from the
defendant, respectively, when the plaintiff wins. As the asymmetry in the transfers w
(T t) increases, it threatens to overwhelm the bargaining surplus (C + c), which in
turn reduces incentives for the parties to settle the suit.
2. The tradeoff is that early adjudication may involve greater risk, since it does
not include all the evidence available at trial. In addition, it may sometimes fail to
deliver a clear decision that results in resolution of the suit.
3. Patent pools were used in the past to accomplish the same ends
(Bittlingmayer, 1988; Merges, 1996). Patent pools have existed in many industries
including automobiles, aircrafts, synthetic rubber, sewing machines, bathtubs, door
parts, seeded-raisin, and coaster brakes, but their role has since been diminished by
increased antitrust scrutiny.
4. On the policy front, this time period does not straddle the creation of the
Federal Circuit for patent-suit appeals (in October 1982), which was a very significant
institutional change in patent litigation. Moreover, it avoids straddling the mid-1990s
when software patents became rapidly institutionalized (Graham & Mowery, 2003),
the TRIPS (Trade-Related Intellectual Property Rights) agreement was enacted into
law (1995) and Markman hearings were introduced into patent litigation proceedings
(1996) (Markman v. Westview Instruments, Inc., 517 U.S. 370). On the technology
front, not only did software patents begin to play an important role in the computer
industry but the boundaries of the industry became increasing difficult to define as
software patenting proliferated into other industries and eventually led to the filing of
large numbers of business method patents. Moreover, biotechnology innovation
began to take an increasingly multi-invention character with the advances into
genomics and other related technologies (Somaya et al., 2011), so that the sharp
distinction between research medicine and computer innovation (as non-systemic and
systemic, respectively) was no longer valid.
5. Federal Judicial Center. Federal Court Cases Integrated Database (various
versions, 19701997), Ann Arbor, MI: Inter-university Consortium for Political
and Social Research (distributor).
6. In addition to the match of roughly 3,000 manufacturing firms to their
subsidiaries in 1989 from this database, I track all name variations for firms in my
dataset that are not listed there.
7. Some of the technologies included in the research medicines sample were
pharmaceuticals, drug delivery systems, assays, and dental innovations (but not
veterinarian drugs, nutritional foods, cosmetic products, implants, and prosthetic
devices), and the computers sample included data storage, computer systems, I/O
devices, and networking technologies (but not electronic displays and clocks, video
and audio recording, and telecommunications).
8. For all other variables, I treat the measures as capturing invariant essential
properties of the litigated patent, and as a result these variables are not time-varying
in the empirical analyses. For example, although citations accrue to the litigated
patent over time, they are used here as measures of the patents importance or
strategic stakes (in the case of self-citations), which are fixed attributes of the patent
and do not change every time a citation occurs. Therefore, the actual citations (or
self-citations) to a patent simply proxies its expected lifetime citations, which is a
measure of its importance (or strategic stakes) (Hall, Jaffe, & Trajtenberg, 2005).
9. The hazard ratio standard errors are estimated using the delta method (based
on the Taylor series expansion), but the p-values for assessing significance are based
on the coefficient estimates. The distribution of the z-statistic for the coefficient is
well characterized and symmetric, making it easy to conduct two-tailed tests, but
this is not the case for the hazard ratios. However, since the coefficients are maxi-
mum likelihood estimates, so are the hazard ratios.
10. Table 2 suggests that Portfolio Citations and Counter-Suit have very skew
distributions. In large part, this skewness is due to the fact that the value of these
variables for most observations is zero. Even though Portfolio Citations is a contin-
uous variable, its effects are similar to that of a dummy variable, and the results are
largely identical if the variable is simply replaced by a dummy variable (=1 when
its value is higher than a certain threshold, say 0.1). To check for the possibility that
the results may be driven by outliers, I dropped the top 5% of observations
(for Portfolio Citations) within this set and found very similar results (although the
statistical significance level on the settlement model fell to 5%).
11. I investigated three possible causal factors providing a signal for other
contemporaneous litigation, the effect of prior settlements, or the effect of prior
court decisions (under the legal principle of res judicata, where courts adopt prior
decisions on issues that have been adjudicated) as the mechanism through which
this variable operates, but the results were inconclusive, possibly because only crude
measures were available for these causal factors.
ACKNOWLEDGMENTS
This research would not have been possible without David Mowery and
Bronwyn Hall, who provided valuable early advice and access to data.
Martin Dresner, Phil Evers, Brent Goldfarb, Curt Grimm, Josh Newberg,
James Prieger, David Teece, Robert Windle, Rosemarie Ziedonis, and
participants at the Academy of Management Conference provided valuable
feedback on earlier drafts. Detailed comments and suggestions from
Editor Rick Vanden Bergh and two anonymous referees for this volume
are gratefully acknowledged.
REFERENCES
Admati, A. R., & Perry, M. (1987). Strategic delay in bargaining. Review of Economic Studies,
54, 345364.
AIPLA. (1997). Report of economic survey. Washington, DC: American Intellectual Property
Law Association.
Allison, P. D. (1984). Regression for longitudinal event data. Beverley Hills, CA: Sage.
Arora, A. (1995). Licensing tacit knowledge: Intellectual property rights and the market for
know-how. Economics of Innovation and New Technology, 4, 4159.
502 DEEPAK SOMAYA
Barney, J. (1991). Firm resources and sustained competitive advantage. Journal of

Management, 17(1), 99120.
Bebchuk, L. A. (1984). Litigation and settlement under imperfect information. Rand Journal of
Economics, 15(3), 404415.
Bittlingmayer, G. (1988). Property rights, progress, and the aircraft patent agreement. Journal
of Law & Economics, 31(1), 227248.
Blind, K., Cremers, K., & Mueller, E. (2009). The influence of strategic patenting on companies
patent portfolios. Research Policy, 38, 428436.
Blind, K., Edler, J., Frietsch, R., & Schmoch, U. (2006). Motives to patent: Empirical evidence
from Germany. Research Policy, 35, 655672.
Clarkson, G., & Toh, P. K. (2010). Keep out signs: The role of deterrence in the competition
for resources. Strategic Management Journal, 31, 12021225.
Cohen, W. M., Nelson, R. R., & Walsh, J. (2000). Protecting their intellectual assets:
Appropriability conditions and why U.S. manufacturing firms patent (or Not). NBER
Working Paper No. 7552.
Cooter, R. D., & Rubinfeld, D. L. (1989). Economic analysis of legal disputes and their resolution.
Journal of Economic Literature, 27(3), 10671097.
Cox, D. R. (1972). Regression models and life tables. Journal of the Royal Statistical Society,
Series B, 34, 187202.
Cramton, P. C. (1992). Strategic delay in bargaining with two-sided uncertainty. Review of
Economic Studies, 59, 205225.
De Figueiredo, J. M. (2009). Integrated political strategy. Advances in Strategic Management,
26, 459486.
Duguet, E., & Kabla, I. (1998). Appropriation strategy and the motivations to use the patent
system: An econometric analysis at the firm level in French manufacturing. Annales
dEconomie et de Statistique, 49, 289327.
Fenn, P., & Rickman, N. (1999). Delay and settlement in litigation. The Economic Journal,
109, 476491.
Fournier, G. M., & Zuehlke, T. W. (1996). The timing of out-of-court settlements. Rand
Graham, S. J. H., & Mowery, D. C. (2003). Intellectual property protection in the U.S. software
industry. In W. M. Cohen & S. Merrill (Eds.), Patents in the knowledge-based economy.
Washington, DC: National Academies Press.
Grindley, P. C., & Teece, D. J. (1997). Managing intellectual capital: Licensing and cross-licensing
in semiconductors and electronics. California Management Review, 39(2), 841.
Hall, B. H., Jaffe, A. B., & Trajtenberg, M. (2001). The NBER patent citation data file:
Lessons, insights and methodological tools. NBER Working Paper No. 8498.
Hall, B. H., Jaffe, A. B., & Trajtenberg, M. (2005). Market value and patent citations. RAND
Hall, B. H., & Ziedonis, R. M. (2001). The patent paradox revisited: An empirical study of
patenting in the semiconductor industry, 1979-1995. RAND Journal of Economics,
32(1), 101128.
Hughes, J. W., & Savoca, E. (1999). Accounting for censoring in duration data: An application
to estimating the effect of legal reforms on the duration of medical malpractice disputes.
Journal of Applied Statistics, 26(2), 219228.
Jaffe, A. B., Fogarty, M. S., & Banks, B. A. (1998). Evidence from patents and patent citations
on the impact of Nasa and other federal labs on commercial innovation. Journal of
Industrial Economics, 46(2), 183205.
Kessler, D. (1996). Institutional causes of delay in the settlement of legal disputes. Journal of
Law, Economics and Organization, 12(2), 432460.
Kitch, E. W. (1977). The nature and function of the patent system. Journal of Law and
Lancaster, T. (1990). The econometric analysis of transition data. New York, NY: Cambridge.
Lanjouw, J. O., & Lerner, J. (1998). The enforcement of intellectual property rights: A survey
of the empirical literature. Annales dEconomie et de Statistique, (49/50), 223246.
Lanjouw, J. O., & Lerner, J. (2001). Tilting the table? The use of preliminary injunctions.
Journal of Law and Economics, 44(2), 573603.
Lanjouw, J. O., & Schankerman, M. (2001). Characteristics of patent litigation: A window on
competition. RAND Journal of Economics, 32(1), 129151.
Lerner, J. (1995). Patenting in the shadow of competitors. Journal of Law & Economics,
38(2), 463495.
Levin, R. C., Klevorick, A. K., Nelson, R. R., & Winter, S. G. (1987). Appropriating the
returns from industrial research and development. Brookings Papers on Economic
Activity, 3, 783820.
Linden, G., & Somaya, D. (2003). System-on-a-chip integration in the semiconductor
industry: Industry structure and firm strategies. Industrial and Corporate Change,
12(3), 545576.
Lippman, S. A., & Rumelt, R. P. (1982). Uncertain imitability: An analysis of interfirm
differences in efficiency under competition. Bell Journal of Economics, 13(2), 418438.
Ma, C. T. A., & Manove, M. (1993). Bargaining with deadlines and imperfect player control.
Econometrica, 61(6), 13131339.
Mahoney, J. T., & Pandian, J. R. (1992). The resource-based view within the conversation of
strategic management. Strategic Management Journal, 13(5), 363380.
McMillan, G. S., & Hamilton, R. D. (2000). Using bibliometrics to measure firm knowledge:
An analysis of the US pharmaceutical industry. Strategic Management Journal,
12(4), 465475.
Merges, R. P. (1996). Contracting into liability rules: Intellectual property rights and collective
rights organizations. California Law Review, 84(5), 12931393.
Meurer, M. J. (1989). The settlement of patent litigation. Rand Journal of Economics,
20(1), 7791.
Nerkar, A., Paruchuri, S., & Khaire, M. (2007). Business method patents as real options:
Value and disclosure as drivers of litigation. In J. J. Reuer & T. W. Tong (Eds.), Real
options theory. Advances in strategic management (Vol. 24, pp. 247274). Bingley, UK:
Emerald Group Publishing Limited.
Peteraf, M. A. (1993). The cornerstones of competitive advantage: A resource-based view.
Petersen, T. (1995). Analysis of event histories. In G. Arminger, C. C. Clogg, & M. E. Sobel
(Eds.), Handbook of statistical modeling for the social and behavioral sciences.
New York, NY: Plenum.
Png, I. P. L. (1983). Strategic behavior in suit, settlement, and trial. Bell Journal of Economics,
14(2), 539550.
Polidoro, F., & Toh, P. K. (2011). Letting rivals come close or warding them off? The effects
of substitution threat on imitation deterrence. Academy of Management Journal,
54(2), 369392.
Priest, G. L., & Klein, B. (1984). The selection of disputes for litigation. Journal of Legal
Studies, 13, 155.
504 DEEPAK SOMAYA
Reitzig, M. (2004). The private values of thickets and fences: Towards an updated picture of
the use of patents across industries. Economics of Innovation and New Technology,
13(5), 457476.
Rivette, K. G., & Kline, D. (2000). Discovering new value in intellectual property. Harvard
Business Review, (JanuaryFebruary), 212.
Rumelt, R. P. (1984). Towards a strategic theory of the firm. In R. B. Lamb (Ed.), Competitive
strategic management. Englewood Cliffs, NJ: Prentice Hall.
Rumelt, R. P. (2011). Good strategy bad strategy: The difference and why it matters. New York,
NY: Crown Business Publishing.
Schweizer, U. (1989). Litigation and settlement under two-sided incomplete information.
Review of Economic Studies, 56, 163178.
Siegelman, P., & Waldfogel, J. (1996). Towards a taxonomy of disputes: New evidence through
the prism of the priest/Klein model. Yale University (mimeo).
Somaya, D. (2003). Strategic determinants of decisions not to settle patent litigation. Strategic
Somaya, D. (2004). Firm strategies and trends in patent litigation in the United States. In
G. Libecap (Ed.), Advances in the study of entrepreneurship, innovation and economic
growth. Greenwich, CT: JAI Press.
Somaya, D. (2012). Patent strategy and management: An integrative review and research
agenda. Journal of Management, 38(4), 10841114.
Somaya, D., & McDaniel, C. (2012). Tribunal specialization and institutional targeting in
patent enforcement. Organization Science, 23(3), 869887.
Somaya, D., Teece, D. J., & Wakeman, S. (2011). Innovating in multi-invention contexts:
Mapping solutions to technological and intellectual property complexity. California
Spier, K., & Spulber, D. (1993). Suit settlement and trial: A theoretical analysis under alternative
methods for the allocation of legal costs. Journal of Legal Studies, 11, 5581.
Spier, K. E. (1992). The dynamics of pretrial negotiation. The Review of Economic Studies,
59(1), 93108.
Stefanadis, C. (1997). Downstream vertical foreclosure and upstream innovation. Journal of
Industrial Economics, 45(4), 445456.
Teece, D. J. (1986). Profiting from technological innovation: Implications for integration,
collaboration, licensing and public policy. Research Policy, 15(6), 285305.
Teece, D. J., Pisano, G., & Shuen, A. (1997). Dynamic capabilities and strategic management.
Trajtenberg, M. (1990a). Economic analysis of product innovation. Cambridge, MA: Harvard
University Press.
Trajtenberg, M. (1990b). A penny for your quotes: Patent citations and the value of innovations.
Rand Journal of Economics, 21(1), 172187.
Von Hippel, E. (1988). The sources of innovation. New York, NY: Oxford University Press.
Waldfogel, J. (1995). The selection hypothesis and the relationship between trial and plaintiff
victory. Journal of Political Economy, 103(2), 229260.
Walsh, J. P., Arora, A., & Cohen, W. M. (2003). Effects of research tool patents and licensing
on biomedical innovation. In W. M. Cohen & S. A. Merrill (Eds.), Patents in the
knowledge-based economy. Washington, DC: National Academies Press.
ABOUT THE AUTHORS
Jose Miguel Abito is Assistant Professor of Business Economics and Public

Policy at the Wharton School, University of Pennsylvania. He specializes
in Industrial Organization, Regulation and Environmental Economics.
He received his PhD in Economics from Northwestern University in
2013. Prior to his doctoral studies, he attended the Toulouse School of
Economics in France and the National University of Singapore.
David P. Baron is David S. and Ann M. Barlow Professor of Political
Economy and Strategy (Emeritus) in the Graduate School of Business of
Stanford University. He conducts research in economics, political science,
and strategy and the business environment. He has been at Stanford
since 1981.
David Besanko is IBM Professor of Regulation & Competitive Practices at
the Kellogg School of Management at Northwestern University. Professor
Besankos research covers topics relating to industry dynamics, competitive
strategy, industrial organization, the theory of the firm, and economics of
regulation. He has received grants from the National Science Foundation
and from the Citicorp Behavioral Science Research Council to support this
research. In 2015, he received the Aspen Institutes Faculty Pioneer award.
Jean-Philippe Bonardi is Professor of Strategic Management and currently
the Dean of HEC Lausanne (the Faculty of Business and Economics of the
University of Lausanne, Switzerland). He earned his PhD and a Master
from HEC School of Management, as well as a Master in Economics from
the University of Aix-en-Provence, and was previously an Associate
Professor at the Richard Ivey School of Business of the University of
Western Ontario. He also held visiting appointments at the University of
California Berkeley, Tulane University, and the University of New South
Wales. His research focuses on the relationships between corporate strategy
and public policies, both in the context of public and private politics.
Dominik Breitinger is working for the World Business Council for
Sustainable Development (WBCSD), where he currently serves as Director
for Finance & Capital Markets. He holds a Bachelor and Masters Degree
505
506 ABOUT THE AUTHORS
from WHU School of Management as well as a PhD from HEC Lausanne

(the faculty of business and economics from the University of Lausanne).
Dominik also spent several years with Accenture Strategy Consulting where
he specialized in Merger and Acquisitions and Sustainability.
John M. de Figueiredo is Edward and Ellen Marie Schwarzman Professor
of Law at the Duke Law School and Professor of Strategy and Economics
at the Fuqua School of Business. He is also Research Associate at the
National Bureau of Economic Research. de Figueiredo studies competitive
strategy, corporate political and legal strategy, and the management of
innovation. His research in non-market strategy explores how firms use
political influence, regulatory lobbying, and strategic litigation to enhance
competitive performance.
Rui J. P. de Figueiredo, Jr. is Associate Professor in the Haas School
of Business and Department of Political Science at the University of
California at Berkeley. His primary research interests are in political
economy with a focus on political institutions and organization. He has
published in a range of political science, economics and management
journals including the American Political Science Review, American Journal
of Political Science, Journal of Law and Economics, Journal of Public
Economics, Journal of Law, Economics and Organization, and Management
Science, among others.
Daniel Diermeier is the Dean of the Harris School of Public Policy and
Emmett Dedmon Professor of Public Administration at the University of
Chicago. His teaching and research focuses on formal political theory, poli-
tical institutions, the interaction of business and politics, text analytics,
public perception, as well as crisis and reputation management. Professor
Diermeier is a fellow of the American Academy of Arts and Sciences, the
Guggenheim Foundation, and the Canadian Institute of Advanced
Research (CIFAR). Prior to joining the University of Chicago Professor
Diermeier has taught at the Graduate School of Business at Stanford
University and the Kellogg School of Management, Northwestern
University, most recently as IBM Professor of Regulation and Practice in
the Department of Managerial Economics and Decision Sciences (MEDS),
and Director of the Ford Motor Company Center of Global Citizenship.
Geoff Edwards is Vice President with the European and Asia-Pacific
Competition Practices of Charles River Associates, based in London and
Sydney. His primary field of expertise is competition and regulatory eco-
nomics (with a focus on the telecommunications industry). His publications
About the Authors 507
include many articles on the economics of competition law and policy as

well as papers on the political economy of telecommunications regulation
that have been published in the Journal of Regulatory Economics and the
Journal of Economics and Management Strategy.
Adam Fremeth is Assistant Professor in the Business, Economics, and
Public Policy Group at the Ivey Business School, University of Western
Ontario. His work focuses on the intersection of firm strategy and public
policy making. He is primarily concerned with how firms both shape and
respond to public policy, specifically in the electric utility and oil and gas
sectors. His research has been published in leading journals, including
American Economics Journal: Applied Economics, California Management
Review, Strategic Management Journal, and the Journal of Law, Economics,
and Organization.
Benjamin A. T. Graham is Assistant Professor of International Relations at
the University of Southern California. His core research is on political risk
and foreign investment in emerging markets, focusing particularly on the
role of diaspora investors and on the ongoing evolution of international
property rights. He also studies federalism, powersharing, and unrecog-
nized states. Professor Graham serves as a Special Sworn Researcher at the
U.S. Bureau of Economic Analysis. He received his Ph.D. in political
science from the University of California, San Diego.
Hillary Greene is Professor of Law at the University of Connecticut School
of Law and a Visiting Scholar at Harvard Law School. Professor Greenes
research and teaching interests focus upon antitrust/competition policy,
patent law, and First Amendment Law. From 2007 to 2012 she served
as Inaugural Director of the University of Connecticut Law Schools
Intellectual Property and Entrepreneurship Law Clinic. Professor Greene is
the Chairperson of the American Association of Law Schools Section on
Antitrust and Economic Regulation. She is a graduate of the Yale Law
School and of Yale College where she earned her bachelors degree in
Economics and Political Science. Prior to entering academia, Greene
worked as a litigation associate at Cahill Gordon & Reindel and served as
Acting Deputy Assistant General Counsel for Policy Studies and as Project
Director for Intellectual Property at the Federal Trade Commissions
Office of the General Counsel where she was a primary organizer of the
FTC/DOJ hearings which resulted in the FTC Report, To Promote
Innovation: The Proper Balance of Competition and Patent Law and
Policy (2003).
Nan Jia is Assistant Professor of Strategic Management at the Marshall

School of Business, University of Southern California. She received her
PhD from the Rotman School of Management, University of Toronto. Her
research interests include corporate political strategy, business-governance
relationships, and corporate governance in international business. Her
research has been published in the Management Science, Strategic
Management Journal, Organizational Science, and the Journal of Politics.
She serves on the editorial board of Strategic Management Journal and
Journal of International Business Studies.
Noel P. Johnston is Postdoctoral Fellow at Oxford Universitys Blavatnik
School of Government, and Research Fellow of Nuffield College. His
research covers economics, political science, and strategy, with specific
focuses on trade policy and the structure of compliance with international
property rights in foreign investment markets. He received his Ph.D. in
political science from Washington University in St. Louis.
Allison F. Kingsley is Assistant Professor at the University of Vermonts
Grossman School of Business. Her research focuses on investor strategies
to manage political risk in emerging markets. She earned her Ph.D. in poli-
tical science from Columbia University and her M.S.L. from Yale Law
School, and worked on Wall Street for banks and buy-side funds for
ten years.
Michael Lenox is Samuel L. Slover Professor of Business at the University
of Virginias Darden School of Business where he teaches the core MBA
strategy course. He also serves as Associate Dean and Academic Director
of Dardens Batten Institute for Entrepreneurship and Innovation. He
found and served as Inaugural President of the Multiple-University
Alliance for Research on Corporate Sustainability. Prior to joining Darden
in 2008, Professor Lenox was Professor at Duke Universitys Fuqua School
of Business where he served as Area Coordinator for Fuquas Strategy
Area and Faculty Director and Founder of Dukes Corporate Sustainability
Initiative. He received his PhD in Technology Management and Policy from
the Massachusetts Institute of Technology in 1999 and the degrees of
Bachelor and Master of Science in Systems Engineering from the University
of Virginia. Professor Lenox has served as Assistant Professor at New York
Universitys Stern School of Business and as Visiting Professor at Stanford
University, Harvard University, Oxford University, and IMD. Professor
Lenoxs research has appeared in over 25 refereed academic publications and
has been cited in a number of media outlets including the New York Times,
the Financial Times, and the Economist. In 2009, he was recognized as a

Faculty Pioneer by the Aspen Institute and as the top strategy professor
under 40 by the Strategic Management Society. In 2011, he was named one
of the top 40 business professors under 40 by Poets & Quants. Professor
Lenoxs primary expertise is in the domain of technology strategy and policy.
He is broadly interested in the role of innovation and entrepreneurship
for economic growth and firm competitive success. In particular, he
explores the business strategy and public policy drivers of the direction of
innovative activity. Professor Lenox also has a long-standing interest in
the interface between business strategy and public policy as it relates
to the natural environment. Recent work explores firm strategies and
non-traditional public policies that have the potential to drive green
innovation and entrepreneurship.
Thomas P. Lyon holds the Dow Chair of Sustainable Science, Technology
and Commerce at the University of Michigan, with appointments in both
the Ross School of Business and the School of Natural Resources and
Environment. From 2006 to 2011 he served as Director of the Universitys
Erb Institute for Global Sustainable Enterprise. Professor Lyon is a
leader in using economic analysis to understand corporate environmental
strategy and how it is shaped by emerging government regulations, non-
governmental organizations, and consumer demands. His book with John
W. Maxwell, Corporate Environmentalism and Public Policy, published by
Cambridge University Press, is the first rigorous economic analysis of this
increasingly important topic. Professor Lyon earned his bachelors degree
at Princeton University and his doctorate at Stanford University. His
current research focuses on corporate environmental information disclo-
sure, greenwash, and voluntary programs for environmental improvement.
Professor Lyon has been Visiting Professor at Stanford University, the
University of Chicago, the University of Bonn, University of California at
Santa Barbara, and the University of Paris. He was Fulbright Scholar at
the Scuola SantAnna in Pisa, Italy, during 1997. He spent the academic
year 2002/2003 as a Gilbert White Fellow at Resources for the Future in
Washington, DC, and 2003/2004 as Visiting Economist in the Antitrust
Division of the U.S. Department of Justice. Professor Lyon serves on the
editorial boards of the Journal of Economics and Management Strategy and
the Journal of Regulatory Economics. His teaching experience includes
energy economics and policy, environmental governance, non-market strat-
egy, regulation, managerial economics, business and government, game
theory, business strategy, and the management of innovation.
John W. Maxwell holds the W. George Pinnell Chair in Business

Economics at the Kelley School of Business, Indiana University. In addi-
tion to his principal appointment at Kelley, Professor Maxwell is Research
Professor at the Richard Ivey School of Business at the University of
Western Ontario. He is Fellow of The Royal Society for the
Encouragement of the Arts, Manufactures and Commerce (FRSA) and
also Research Fellow at CCISSR, Peking University and ZEI, University
of Bonn. He is Founding Board Member of the Alliance for Research in
Corporate Sustainability and has also served as Chairman of the Foreign
Scholars Advisory Committee to the Department of Environment,
Resource and Development Economics at the Peking (Beijing) University.
Professor Maxwells areas of research interest concern the interface
between industry, government, non-governmental organizations and our
natural environment. He has published numerous scholarly and practi-
tioner articles and edited volumes in this area including his book with
Thomas P. Lyon, Corporate Environmentalism and Public Policy, which
was published by Cambridge University Press in 2004. Professor Maxwell
has been Visiting Scholar at University College, London, Bonn University
and Peking University. Professor Maxwell has taught courses on the
Global Environment of Business, Corporate Non-Market Strategy,
Managerial Economics, and Regulation & Public Policy to undergraduate,
MBA and PhD students.
Kyle Mayer is Professor of Strategic Management at the Marshall School
of Business, University of Southern California. He studies how firms gov-
ern relationships with other firms, with particular attention to the contract
and its role in establishing a framework for the relationship. His research
has been published in Organizational Science, Academy of Management
Journal, Management Science, and Journal of Law, Economics, and
Organization. He serves on the editorial board of Academy of Management
Journal, Organization Science, Academy of Management Review, and
Strategic Management Journal.
Dylan Minor is Assistant Professor of Managerial Economics and Decision
Sciences at the Kellogg School of Management and Visiting Assistant
Professor of Strategy at Harvard Business School. His research focuses on
the nexus between organizations and social issues. He obtained his PhD
and MS from the Haas School of Business, UC Berkeley.
Felix Oberholzer-Gee is Andreas Andresen Professor of Business
Administration in the Strategy Unit at Harvard Business School. A member
of the faculty since 2003, Professor Oberholzer-Gee received his Masters

degree, summa cum laude, and his PhD in Economics from the University
of Zurich. His first faculty position was at the Wharton School, University
of Pennsylvania. He currently teaches competitive strategy in executive
education programs such as the Program for Leadership Development, the
Senior Executive Program for China, and in a program for media executives
titled Effective Strategies for Media Companies. His course Strategies
Beyond the Market is a popular elective class for second-year MBA stu-
dents. Professor Oberholzer-Gee won numerous awards for excellence in
teaching, including the Harvard Business School Class of 2006 Faculty
Teaching Award for best teacher in the core curriculum, and the 2002 Helen
Kardon Moss Anvil Award for best teacher in the Wharton MBA program.
Prior to his academic career, Professor Oberholzer-Gee served as managing
director of Symo Electronics, a Swiss-based process control company.
Professor Oberholzer-Gees research and consulting are centered on compe-
titive strategy, international competition, and non-market strategy, a
branch of strategic management that studies how companies best work with
government and non-governmental groups. In recent work, he studied how
entertainment companies can successfully manage the digital transition.
Dating back to a study abroad program as an undergraduate, Professor
Oberholzer-Gee has a long-standing interest in the Chinese economy and
Chinese companies. In recent academic work, he compared the financial
performance of Chinese companies with the performance of multinationals
operating in China. In a related study, he explored how and why Chinese
companies diversify their activities. Professor Oberholzer-Gees academic
work has been published in the very best, peer-reviewed journals of his
profession, including the American Economic Review, Journal of Political
Economy, Journal of Financial Economics, and Journal of Law & Economics.
His work has been profiled by media outlets around the world, including
ABC Nightly News, Financial Times, Guardian, Le Figaro, Neue Zurcher
Zeitung, New York Times, Singapore Straits Times, Suddeutsche Zeitung,
Wall Street Journal, and Washington Post.
Brian Kelleher Richter is Assistant Professor in the Business, Government,
and Society Department at the University of Texas at Austins McCombs
School of Business. He is an interdisciplinary scholar who has published
in leading journals in the management, economics, and political science
disciplines. His research focuses on understanding firms interactions with
governments and other non-market, societal and legal, actors and
institutions.
Brandon Schaufele is Assistant Professor in the Business, Economics and

Public Policy Group at the Ivey Business School, University of Western
Ontario. His research concentrates on the relationships between firms and
governments as well as the implications of energy and environmental
policy-making. Among other places, he has published papers in the American
Economics Journal: Applied Economics and the Journal of Environmental
Economics and Management.
Kenneth W. Shotts is David S. and Ann M. Barlow Professor of Political
Economy at the Stanford Graduate School of Business. He received his BA
in Political Science from Stanford in 1993 and his PhD from the GSB
in 1999. In addition to his time at the Stanford, he has taught at
Northwestern and the University of Michigan, and has been a visiting
scholar at Princetons Woodrow Wilson School of Public and International
Affairs, the Hoover Institution, and the Center for Advanced Study in the
Behavioral Sciences. In his research, Shotts uses game theory to analyze
how elections and political institutions influence policy choices made by
government officials. He has published articles on presidential leadership,
racial redistricting, term limits, and the politics of regulatory enforce-
ment. He is currently doing research on several topics, including political
risk, industry-level self-regulation, electoral accountability, and policy
entrepreneurship.
Deepak Somaya is Associate Professor of Strategy and Entrepreneurship
and the Steven and Christy King Faculty Fellow at the College of Business,
University of Illinois at Urbana-Champaign. He also holds courtesy
appointments in the College of Law and the Institute for Genomic Biology
at the University of Illinois. In his research, he studies how companies stra-
tegize about and derive competitive advantage from their intangible assets,
particularly their human capital, relational assets, and intellectual property.
His research has been published in over 25 journal articles, book chapters,
and conference proceedings, and has received numerous awards including a
best dissertation award, several conference best paper awards, and the 2012
California Management Review Best Article Award. He currently serves on
the editorial boards of the Journal of Management, Organization Science,
Strategic Entrepreneurship Journal, and Strategic Management Journal, and
is Program Chair for the Strategic Human Capital interest group of the
Strategic Management Society.
Richard G. Vanden Bergh is Associate Professor in the School of Business
Administration, University of Vermont. He graduated from the Haas
School of Business, University of California at Berkeley, where he com-

pleted his PhD in Business and Public Policy and earned his MBA. He
graduated with a BA in Economics from Swarthmore College. Professor
Vanden Berghs areas of research include: firm strategy in the political
environment, the design of political and regulatory institutions, and the
effects of political institutions on business investment. For the past several
years he has been exploring issues in the energy sector including studies of
how merger and acquisition activity affects firm strategy in political and
regulatory arenas. Professor Vanden Berghs research has been published
in many academic journals including the Academy of Management Journal,
the Academy of Management Review, the Journal of Law & Economics, the
Journal of Law, Economics & Organization, and the Strategic Management
Journal. Professor Vanden Bergh currently teaches courses on competitive
strategy and political strategy. He has won awards for excellence in teach-
ing, including the Teacher of the Year Award for the School of Business
Administration. Prior to joining the University of Vermont in 2000,
Professor Vanden Bergh worked in corporate banking for Wells Fargo in
Los Angeles and in investment banking for Stone & Youngberg in
San Francisco.
Craig Volden is Professor of Public Policy and Politics, with appointments
in the Frank Batten School of Leadership and Public Policy and the
Woodrow Wilson Department of Politics. He studies legislative politics and
the interaction among political institutions, with a focus on what policy
choices arise from legislative-executive relations and from American feder-
alism. His previous works include Legislative Effectiveness in the United
States Congress: The Lawmakers, co-authored with Alan E. Wiseman
(Cambridge University Press), and Revolving Gridlock: Politics and Policy
From Jimmy Carter to George W. Bush (Westview Press), co-authored
with David Brady. He has published numerous articles in journals
such as American Political Science Review, American Journal of Political
Science, Journal of Politics, Legislative Studies Quarterly, Journal of Law,
Economics & Organization, and Publius: The Journal of Federalism.
Alan E. Wiseman is Associate Professor of Political Science and Law
(by courtesy) at Vanderbilt University. He has research and teaching
interests in American political institutions and positive political economy,
focusing on legislative and electoral politics, regulation, bureaucratic
politics, and business-government relations. His previous works include
Legislative Effectiveness in the United States Congress: The Lawmakers,
co-authored with Craig Volden (Cambridge University Press), and The
Internet Economy: Access, Taxes, and Market Structure (Brookings

Institution Press). He has published numerous articles in journals such
as the American Political Science Review, American Journal of Political
Science, the Journal of Politics, Legislative Studies Quarterly, Business &
Politics, and the Journal of Theoretical Politics.
Dennis A. Yao is Lawrence E. Fouraker Professor of Business
Administration and Head of the Strategy Unit at Harvard Business School.
He joined the HBS faculty in 2004 after having been at the Wharton
School, University of Pennsylvania. From 1991 to 1994 he served as
Commissioner, U.S. Federal Trade Commission where he and his four col-
leagues had decision responsibility for antitrust and consumer protection
matters in both prosecutorial and judicial roles. Professor Yaos research
involves the application of microeconomics and game theory to incentive
and information problems affecting firms. His recent work has focused on
the implications of mental models for competitive interaction and on
understanding contracting and knowledge flow issues in markets for ideas
and inventions. Earlier he worked on problems involving procurement and
on technology-forcing regulation. Professor Yao has also written exten-
sively in the areas of antitrust policy and non-market strategies.

Advances in Strategic Management Strategy Beyond Markets PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Advances in Strategic Management Strategy Beyond Markets PDF

Uploaded by

Copyright:

Available Formats

STRATEGY BEYOND MARKETS

RICHARD G. VANDEN BERGH

United Kingdom North America Japan

First edition 2016

Copyright r 2016 Emerald Group Publishing Limited

Reprints and permissions service

No part of this book may be reproduced, stored in a retrieval system, transmitted in

British Library Cataloguing in Publication Data

Certificate Number 1985

STRATEGY BEYOND MARKETS: A STEP BACK

POLITICAL RISK AS A HOLD-UP PROBLEM:

INCORPORATING LEGISLATIVE EFFECTIVENESS

A UNIFIED MODEL OF POLITICAL RISK

MOTIVATIONS FOR CORPORATE

THE MARKET FOR LEGISLATIVE INFLUENCE

CORPORATE REPUTATIONAL DYNAMICS, PRIVATE

SELF-REGULATION AND REGULATORY

PRIVATE POLITICS DAILY: WHAT MAKES FIRMS

NAVIGATING NATURAL MONOPOLIES: MARKET

THE ORGANIZATION OF NONMARKET STRATEGY

COMPLEMENTARITY IN FIRMS MARKET AND

HOW PATENT STRATEGY AFFECTS THE

ABOUT THE AUTHORS 505

Jose Miguel Abito Department of Business Economics and

Benjamin A. T. School of International Relations,

Brian Kelleher Richter McCombs School of Business, University

Traditional strategy scholarship is fundamentally about understanding the

In response to these pressures, firms design strategies to address the

Whether challenged through public policy or private pressure, managers

traditionally less regulated industries such as entertainment and retail.5 It is

journals that publish strategic management scholarship (e.g., Management

PUBLIC POLITICS ARTICLES

PRIVATE POLITICS ARTICLES

to rely on confrontation versus criticism and serves as a nice complement

INTEGRATED POLITICAL STRATEGY ARTICLES

of stable ratings standards, uncertainty about the quality of ratings, and

1. See CEOs on strategy and social issues, The McKinsey Quarterly,

This paper provides a perspective on the field of nonmarket strategy.

Strategy Beyond Markets

Market strategy focuses on the sources of competitive advantage and

Identifying the effects of these exogenous factors on strategy choice and

A STYLIZED (U.S.) HISTORY OF THE DEMAND

1970s that focused on the environment, occupational health and safety,

Nonmarket strategies can be thought of as playing one or more of six roles.

Applying the Approach.4 Fifth, nonmarket strategy is used to attract

Levels of Nonmarket Strategy Research

Research on nonmarket and integrated strategy can be thought of at three

The Political Economy Approach

A number of frameworks have been proposed for understanding the non-

potentially affected by the proposed legislation. As Olson (1965) argued,

The Firm Level

The challenge for a strategy field, either market or nonmarket, is to identify

the opportunity to reevaluate its strategy and adapt it to the situation in

in hesitation in some states. In some states such as Maryland and Texas

APPLYING THE APPROACH

Some entrepreneurial firms are able to operate without a nonmarket strat-

Kalanik decided to charge ahead and entered the underserved San

licenses or commercial insurance as taxi companies were required to carry

Private Politics: Rainforest Action Network and Citigroup

The Rainforest Action Network (RAN), a radical environmental NGO

An essential factor of production for the infrastructure projects was pro-

Integrated Strategy and Private and Public Politics: The Fast

The Fast Food Campaign (FFC) illustrates the approach to integrated