Professional Documents
Culture Documents
Organizational innovation, its persistence, its relationship with technological innovation, and their influence on firm performance remain under-researched. These issues are investigated using an integrated firm-level data set obtained from two recent waves of the Norwegian Community Innovation Survey (CIS 3 and 4) and firms financial accounts. A Heckman two-step estimation, to ensure against potential sample selection bias, was used to demonstrate that between 1999 and 2004 a number of Norwegian firms were persistent in organizational innovation, and this persistence raised the (positive) effects of organizational innovation on their performance. The results also reveal that many firms undertook, and benefited from, different types of organizational innovation, and such benefits were increased by the combinative effect of organizational and technological innovation. The study also found that older and larger firms are more inclined to attempt organizational change, while smaller firms are more able to benefit. JEL classification: L25, O21, O39.
1. Introduction
Recent decades have seen a remarkable increase in scholarly attention devoted to innovation (Fagerberg, 2004; Fagerberg and Verspagen, 2009; Fagerberg and Sapprasert, 2010). Most research has focused on understanding product and process innovation (Damanpour et al., 2009). By contrast, research on organizational innovation has been limited (Schmidt and Rammer, 2005; Damanpour et al., 2009) because of a lack of empirical data, established definitions and measurement constructs (Lam, 2004).
*Koson Sapprasert, Centre for Technology, Innovation and Culture, University of Oslo, Postbox 1108, Blindern, N-0317 Oslo, Norway. e-mail: koson.sapprasert@tik.uio.no **Koson Sapprasert, School of Entrepreneurship and Management, Bangkok University, 10110, Bangkok, Thailand. e-mail: koson.s@bu.ac.th
y
Tommy Hyvarde Clausen, Nordland Research Institute, 8049 Bod, Norway. e-mail: tommy.clausen@nforsk.no z Main author for correspondence.
The Author 2012. Published by Oxford University Press on behalf of Associazione ICC. All rights reserved.
1284
This potentially constrains our understanding of the impact of (different types of) innovation on economic growth and performance at the firm and country levels, particularly as technological and non-technological change are complementary and co-evolve (Nelson, 1991; Freeman, 1995). Organizational innovation can enable and even enhance the effect of technological innovation on firm performance (Chandler, 1962; Lam, 2004). As Wischnevsky et al. (2011: 133) argue, administrative processes are key in enabling organizations to launch new products and services and to manage change.1 By taking advantage of a new data set on organizational innovation, this article aims to improve our understanding of organizational innovation and its performance effects by analyzing: (i) the influence of prior organizational innovation on current organizational innovation, (ii) the influence of prior organizational innovation on the performance effects of current organizational innovation, and (iii) the combined effect of technological and organizational innovation on firm performance. In doing so, the article moves from previous qualitative studies to undertake a quantitative study that contributes to the literature in four ways. First, the article uses a definition and measurement construct of organizational innovation that has the potential to standardize the different definitions and measurement constructs used in previous research (see, Wolfe, 1994; Lam, 2004). Second, the article extends prior cross-sectional studies (e.g. Schmidt and Rammer, 2005) by adopting a more longitudinal approach as suggested by Damanpour et al. (2009). Third, the article includes a preliminary analysis of the persistence of organizational innovation and its effects on performance extending prior works on technological innovation (Geroski et al., 1997; Peters, 2009; Clausen et al., 2012) and prior studies that have used industry-specific samples (e.g. Wischnevsky et al., 2011). Fourth, the article examines the combined effects of technological and organizational innovation on firm performance (Damanpour et al., 2009). The remainder of the article is organized as follows. Section 2 explores how organizational innovation has been understood in the literature, and how it can be measured in applied empirical work. Section 3 provides the theoretical background and hypotheses. Section 4 presents data and methods. Section 5 discusses descriptive statistics and empirical findings from the econometric analysis, while Section 6 provides a discussion and conclusion.
The authors defined change in administrative processes as change in the social system, related mainly to the organizations administrative domain. This includes shifts in structure, policies, reward systems, labor relations, control systems, and coordinating mechanisms of the organization, which affect the knowledge and procedures used in performing the work of management.
1285
1286
management/organization studies as an adoption of any novelty in an organization (see, for example, Evan, 1966; Daft, 1978; Teece, 1980; Kimberly and Evanisko, 1981; Damanpour, 1987, 1991).2 More specifically, Edquist et al. (2001) distinguish between technical and organizational process innovation. However, our definition has the advantage of being increasingly standardized.
The term administrative innovation is used as opposed to technical innovation in this line of research.
1287
Recent studies have started to address these issues. Boer and During (2003), for example, compare and contrast findings from three empirical studies on product, process, and organizational innovation (measured as the implementation of Total Quality Management), in different industries in the Netherlands, UK, and Belgium to show that the differences between product, process, and organizational innovation processes are surprisingly few (103). Similarly, a large scale quantitative study of German manufacturing and service sectors by Schmidt and Rammer (2005) found that technological and non-technological innovation were often linked to each other, and had similar determinants, suggesting that the decision to innovate was driven by similar factors. They also found that firms that combined product and process innovations with marketing and organizational innovations performed better in terms of innovation sales and process innovation driven cost reductions (conditional upon the adoption of both organizational and marketing innovations). A combination of product with organizational innovation had a significant and positive effect on firm profitability, although the firms that reported the highest effects on profit margins only introduced technological innovations. Overall the study by Schmidt and Rammer (2005) suggests that firms tend to adopt both technological and non-technological innovation and their determinants are roughly the same, but the performance implications of adopting different types of innovation are more complex. Some of this complexity may be due to the cross-sectional research design as the introduction of one innovation at a single point in time may not matter for firm performance as much as the innovative history of the firm (Roberts and Amit, 2003; Damanpour et al., 2009), underscoring the need to study firms over time (Damanpour et al., 2009). Adopting a longitudinal approach, Wischnevsky et al. (2011) found in a study of bank holding companies that changes in products are followed by changes in both technological and administrative processes and that the three types of change (change in products, change in technological process, and change in administrative process) exhibit momentum, as firms were more likely to implement changes similar to those recently undertaken. This is consistent with recent studies on the persistence of technological innovations (for example, Flaig and Stadler, 1994; Crepon and Duguet, 1997; Peters, 2009)3 which explore a firms probability to innovate over time (Clausen et al., 2012). Based more or less implicitly on a linear view of innovation, innovation persistence can be seen as a result of sunk costs (Sutton, 1991). This does not necessarily contradict the evolutionary view (Nelson and Winter, 1982; Dosi, 1988) as a firm can persist in innovating by learning and collecting knowledge
3
To the authors knowledge, the present study is probably one of the first research attempts which, in part, looks at the topic of persistence of innovation in an organizational aspect. See, for instance, Raymond et al. (2006) and Clausen et al. (2012) for detailed discussions of research on the topic of persistence of technological innovation.
1288
that furthers its innovative capability. Because of the cumulative nature of learning itself (Rosenberg, 1976), a firm might be able to extend and use this capability to develop new products or processes (Raymond et al., 2006), as well as to improve its organizational routines, at decreasing marginal costs (Amburgey et al., 1993). As Amburgey and Miner (1992) and Kelly and Amburgey (1991) argue, organizational change may then become a self-reinforcing process. We therefore put forth the following hypothesis:
H1: Past attempts at organizational innovation increase the probability of (new) attempts at organizational innovation
However, research has also demonstrated that factors which may enable innovation at some stages in the innovation process may act as a deterrent at other stages (Wolfe, 1994) as changes in organizational routines can disrupt performance (Hannan and Freeman, 1984). Persistent organizational innovation may therefore be disadvantageous and decrease a firms performance, particularly for firms which change frequently and do not fix the problems which arise from disruption (Amburgey et al., 1993). An alternative perspective understands innovative persistence as a process of creative accumulation (Schumpeter, 1942) whereby knowledge obtained from past innovation(s) supports new innovations. Firms learn (to change) by changing, as in conformity with learning by doing (see, for example, Arrow, 1962; Nelson and Winter, 1982; Dosi, 1988). This also means that having changed, firms may be more able to routinize change (Kelly and Amburgey, 1991) by developing a modification routine (Nelson and Winter, 1982; Aldrich, 1999) and adapt to their changing environments. Hence, persistent organizational innovators may be more capable of effectively reorganizing repeatedly, and may benefit more from doing so. Malerba and Orsenigo (1999), for example, show that firms which persistently innovate possess an advantage in consistently improving their performance. Much of this research has focused on firms persistent ability to develop technological innovations and its performance effects. Our knowledge about the performance effects of organizational innovation, including the role of prior organizational innovation, remains scarce (Damanpour et al., 2009; Wischnevsky et al., 2011). Prior, mainly cross-sectional, research has suggested that technological and nontechnological innovation share many of the same determinants and driving forces (Boer and During, 2003; Schmidt and Rammer, 2005). Based on the literature on persistence of technological innovation and its performance effects we thus put forth the following hypothesis:
H2: Persistent organizational innovation increases the effects of (current) organizational innovation on firm performance
As was argued earlier, technological and organizational innovation have complementary effects on firm performance (Chandler, 1962; Nelson, 1991; Sapprasert, 2007; see
1289
Bruland and Mowery, 2004 for historical evidence on the steam engine). This joint contribution still seems to be important as firms reorganize their businesses to exploit the introduction and diffusion of Information and Communication Technology (ICT) (Brynjolfsson and Hitt, 2000, 2003; Bresnahan et al., 2002; Brynjolfsson et al., 2002). Because information processing and transfer can be significantly improved by ICT, it allows more decentralization and task delegation (Brynjolfsson and Mendelson, 1993). Quantitative research on the combinative effects of different types of innovations on firm performance remains scarce (Damanpour et al., 2009). In one of the few studies conducted, Damanpour et al. (2009) analyzed organizational performance from adopting compositions of innovations (service, technological process, and administrative process) over time among public service organizations. Their findings showed that divergence from the industry norm in the adoption of innovation types could be beneficial to organizational performance. Extending this line of research within the context of for-profit firms in the manufacturing and service sector, we put forth the following hypothesis in relation to the combinative effect of technological and organizational innovation on firm performance:
H3: Technological and Organizational innovation have a complementary effect on firm performance
1290
The three sources were then combined, and the resulting data set contains around 1700 respondent firms in the manufacturing, service, and other industries (see Table 1). Since this number of firms refers to an overlap of 430% of firms from the three sources, the data set seems to be reasonably representative. In order to examine the determinants and effects of organizational innovation on the basis of this integrated data set, the following two-step model was constructed: ORG PASTORG PASTPERF HAMPi SIZE AGE IND 1
ORG Dummy for the attempt at organizational innovation (2002 2004) EFORG Factor score for six types of effects of organizational innovation 2005; see more description below PASTORG Dummy for the past attempt at organizational change (1999 2001)
1291
PASTPERF Past performance in terms of profitability growth (1999 2001) HAMPi Hampering factors (2002 2004; see more description below) INCOMP Dummy for the joint contribution of technological and organizational innovation (2002 2004; see explanation below SIZE Firm size in terms of employment (LogEmp) and turnover (LogTurn) AGE Firm age (LogAge) IND Dummy for industrial classifications (NACE)
Downloaded from http://icc.oxfordjournals.org/ at University of New South Wales on April 4, 2013
Because only those firms which reported to the CIS 4 that they had undertaken organizational innovation between 2002 and 2004 were allowed to answer the question about its effects, i.e. since only organizational innovators are included in (2), it is important to inspect for the potential of sample selection bias when analyzing this data. Thus, two-step estimate, which can indicate the existence/significance of this bias, is employed (see for example, Heckmans (1979); Zucker et al., 1998; Hall, 2002; Catozzella and Vivarelli, 2007).4 Based on this estimate, the selection equation explains whether, and the extent to which, the independent variables included in Stage 1 affect firms decisions to undertake organizational innovation (ORG), while the outcome equation examines the influence of the independent variables included in Stage 2 on the outcome of such an undertaking (EFORG). The variables of interest in this Heckman two-step procedure are organizational innovation (ORG), its effects (EFORG), past/persistent organizational change (PASTORG), past performance (PASTPERF), hampering factors (HAMP), the complementarity of organizational and technological innovation (INCOMP), firm size (SIZE), firm age (AGE), and industry dummies (IND). The measure of organizational innovation (ORG), employed as a dependent variable in the selection equation (Stage 1), is obtained from the answers to the question in the CIS4 which asks whether or not, between 2002 and 2004, the firm introduced organizational innovation, defined as being a new or significant change in the firms structure or management methods seeking to improve the firms use of knowledge, quality of goods or services, or workflow efficiency. The three types of organizational innovation concerned in the survey are: (i) a new or significantly improved knowledge management system implemented to better use or exchange information, knowledge, and skills within the firm (ORGSYS); (ii) a major change to the organization of work within the firm, such as change in the management structure or the integration of different departments or activities (ORGSTR); and (iii) a new or significant change in the firms relationships with other firms or public institutions, such as through
4
Since the Heckman results show no sign of selection bias, the OLS (Ordinary Least Square) estimation is also used in the second stage experiment. Three types of organizational innovation (ORGSYS, ORGSTR, and ORGREL) are added, in order to examine their potentially differential impacts. See below.
1292
alliances, partnerships, outsourcing or sub-contracting (ORGREL). It is useful to have details of the contents of change within the firm, as well as linkages between the firm and external actors.5 Based on the three measures, a dependent variable ORG for Stage 1 (Probit) is constructed.6 ORG equals one if the firm has a positive answer for at least one of the three types of organizational innovation, and zero otherwise. The variable used to assess the impact of these three types of organizational innovation is based on the next question in CIS4, which inquired (in 2005) about the effects of such innovation.7 As mentioned earlier, only the firms which carried out organizational innovation, i.e. for which ORG 1, shall respond to the question about its effects. This question asks the firm to rate (from 0 to 3) the importance of six types of effects: (i) reduced response time to customer needs; (ii) improved quality of goods or services; (iii) reduced costs per unit output; (iv) improved employee satisfaction and/or reduced employee turnover; (v) increased enterprise capacity; and (vi) higher enterprise profitability. This information is deemed suitable for use in investigating the effects of organizational change, as it seems to meet the two criteria suggested by Barnett and Carroll (1995), i.e. it captures the effects at the firm level and is broadly applicable (for example, not specific to one or only a few industries or business categories). A factor analysis was conducted for the six measures (see Table A1). One factor was retained from this, and the factor score for each firm is used as a dependent variable (EFORG) in the outcome equation, which examines how the effects of organizational innovation are influenced by the predictors included in Stage 2. Several explanatory variables are employed in the selection and outcome equation. It should be noted that some, but not all,8 of them are taken into account in both stages. These include PASTORG, used to determine the influence of prior organizational change (between 1999 and 2001) on the probability of another attempt at organizational change by the firm between 2002 and 2004 (ORG) in Stage 1 (testing
5
See Barnett and Carroll (1995) for a good discussion on the process and content of organizational change. ORG is applied because this Heckman estimation can have only one dependent variable in a binary format (0 or 1) in the selection equation (Stage 1). This means that such a variable (ORG in this case) cannot be a measure of the scale of organizational innovation and, thus, does not (to a great extent) explain its heterogeneity. It is important to emphasize that, although the information on organizational innovation and its effects both come from the CIS4 (20022004) which may seem to provide somewhat little time for the effects to be realized and thus have a causality problem, the question on the effects of organizational innovation was designed to be rather explicit by asking the respondent firms to evaluate in 2005 the effects of organizational innovation introduced between 2002 and 2004. The Norwegian CIS4 questionnaire was sent out 6 months after the year of reference (2004). This is because of a requirement associated with this regression technique (Heckman, 1976, 1979).
1293
H1). As explained earlier, since only the organizational innovators between 2002 and 2004 (ORG 1) are included Stage 2, PASTORG is used also in the outcome equation to assess the extent to which the combined prior and current efforts at organizational change (between 1999 and 2001 and between 2002 and 2004, i.e. persistence of change) increased the effects of organizational innovation felt in 2005, EFORG (testing H2). In other words, this variable, employed in both equations, helps to answer two questions: to what extent were the sampled firms persistent in organizational innovation? And to what extent did those who were benefit more from being so? PASTORG, constructed on the basis of the CIS3 data, has a value equal to one if the firm has introduced change between 1999 and 2001 in at least one of the following types related to reorganization: corporate strategies, management techniques, and organizational structures. The age and size of a firm are also taken into account in both equations as important control variables, since older firms are not necessarily larger than younger firms, and vice versa (Penrose, 1959).9 Based on the information from the financial accounts, the explanatory variables for firm age and size are created and included in both Stages 1 and 2. Firm age (LogAge) is calculated as the log value of the time period between the year the firm was established and 2001 (the last year before entering the period of main interest, i.e. 20022004). Firm size is measured on the basis of information about the number of employees (LogEmp) and the firms total turnover (LogTurn) in 2001.10 Also, industrial classification dummies (IND), constructed from the CIS3 information, are employed in both stages to control for the influence of industry heterogeneity on the firms propensity to innovate, as well as on its effects. IND equals one if the firm belongs to the respective industry (classification based on the standard NACE code), and zero otherwise. PASTPERF and HAMP are included in the selection equation (Stage 1). PASTPERF, measured based on the financial accounts data as firm growth in profitability (profit per employee) between 1999 and 2001, captures a recent change in the firms economic performance since performance variation usually induces the firm to change (Cyert and March, 1963; Greve, 2003). HAMP represents three types of obstacles to organizational change perceived by the sampled firms between 2002 and 2004. These include high innovation costs (HCOST), a lack of funds (HFUND), and a lack of qualified personnel (HPER), which are often regarded as factors which
See Table A2 for a simple correlation test between firm age and size (in terms of both total turnover and number of employees).
10 9
Having both of these proxies is advantageous since they possibly explain the size of the firm in different dimensions. That is, while LogEmp is deemed to relate more to the scale of human resource, and may thus better depict a degree of complexity/hierarchy of the firms structure, LogTurn represents the size of the firm in terms of financial capacity. A simple correlation test conducted shows that turnover does not necessarily very strongly correlate with the number of employees (see Table A2).
1294
affect innovation in the literature (see for example, Kline and Rosenberg, 1986; Galia and Legros, 2004). Using information from the CIS4, the three proxies are constructed from the firms rating (from 0 to 3) of the importance of these three impediments to innovation.11 Finally, since all the firms included in Stage 2 were organizational innovators between 2002 and 2004 (firms with ORG 1), a dummy for technological innovation in terms of new or significantly improved product(s) or process(es) (INCOMP) between 2002 and 2004 is simply used to measure the joint contribution of technological and organizational innovation in Stage 2 (testing H3), i.e. INCOMP is equivalent to the result of multiplying itself by ORG (which always equals one in this stage). This variable, applied to examine their combinative effect on firm performance (EFORG), is extracted from the CIS4 data on technological innovation, and equals one if the firm introduced at least one product or process innovation between 2002 and 2004. Table A2 provides a correlation matrix for the explanatory variables employed, with no indication of a multicollinearity problem.
5. Analysis
The descriptive statistics in Table 1 demonstrate that more than one-third of the firms in the sample are organizational innovators (having introduced at least one type of organizational innovation between 2002 and 2004). In terms of the descriptive picture of heterogeneity of organizational innovation (the three measures of organizational innovation obtained from the CIS4), change in the firms structure (ORGSTR) is the most common, followed by change in the firms knowledge management systems (ORGSYS) and change in the firms external relations (ORGREL), respectively, regardless of the firms age, size, and sector. The results from Table 1 also show that only a small share of firms undertook all of the changes considered. Table 2 contains the descriptive statistics of other variables in the data set. The results demonstrate that 450% of the firms had carried out organizational change between 1999 and 2001, and many of these had made another attempt at organizational change between 2002 and 2004 (supporting H1). Contrary to, for example Geroski et al. (1997) and Cefis and Orsenigo (2001), who found a rather low persistence of technological innovation using patent information, almost one-quarter of the sampled Norwegian firms were persistent in organizational innovation between 1999 and 2004. However, the present study finds that technological innovation (product/process) was more common than organizational innovation within the
11
These three variables were selected on the basis of their relevance to organizational innovation (those related only to technological innovation were excluded, for example, a lack of information on technology and an uncertain demand for innovative goods and services), their significance during models tests, and their uniqueness reported in the results of the factor analysis (not reported here; available upon request).
1295
sample between 2002 and 2004 (47 percent of the firms reported undertaking technological innovation, compared with the 35 percent which adopted organizational innovation). When comparing across sectors, it can be seen that a greater share of manufacturing firms engaged in technological innovation, while a greater share of service firms were active in organizational innovation between 2002 and 2004, which is consistent with previous findings.12 The results of the econometric analysis are displayed in Table 3. First, considering the lower part of the first two columns (Model I with LogEmp and Model II with LogTurn), the Heckman Stage 1 (with ORG as a dependent variable) results provide some evidence of persistence of organizational innovation in line with the descriptive statistics in Table 2 and recent studies, such as Crepon and Duguet (1997) and Peters (2009) in the context of technological innovation. Prior organizational change between 1999 and 2001 influences the probability of another attempt at organizational
12
As usually argued in the literature on service innovation (for example, Evangelista, 2000; Miles, 2004; Sapprasert, 2007), non-technological and intangible characteristics of services are very significant and particularly linked to organizational change.
1296
innovation by firms between 2002 and 2004 (ORG), which supports H1. This can be seen from the significant positive coefficients of PASTORG (Past Organizational Change) in Models I and II (0.832 and 0.794, respectively, both significant at the 5% level). Further, the results in Table 3 shed light on how the effects of organizational innovation (EFORG) can be explained by several determinants. Since there is no clear evidence of selection bias (insignificant Mills ratios in both Heckman Models I and II), the results of both the Heckman outcome equation (Stage 2the upper part of the results for Models I and II) and OLS (Ordinary Least Square) estimations (Models III and IV in the last two columns), which are quite comparable, are reported and discussed. First, the results of the Heckman outcome equation (coefficients of 0.129 and 0.132, both significant at the 10% level in Models I and II, respectively)13 indicate the existence of a positive relationship between persistence of organizational innovation (PASTORG) and firm performance (EFORG). This supports H2 and prior research such as that undertaken by Malerba and Orsenigo (1999), suggesting that innovation persistency is conducive to the consistent improvement of firm performance within the context of technological innovation. Next, the results of all models in Table 3 confirm H3 in terms of the combinative effect. The coefficients of INCOMP, measuring the complementarity and joint effect of organizational and technological innovation, are positive and statistically significant at the 10% level in Model I (coefficient of 0.146) and at the 5% level in Models II, III, and IV (coefficients of 0.154, 0.159, and 0.169, respectively), supporting the claim that this combined presence is associated with improved firm performance (Chandler, 1962; Nelson, 1991; Damanpour et al., 2009). With regard to the influence of important control variables, most notably size and age, our results suggest that larger and older firms are more inclined to make an attempt at organizational innovation, while, in terms of outcomes, smaller firms are more able to benefit from such an attempt. In addition, the OLS results demonstrate that all three types of organizational innovation are strongly associated with improved firm performance.14 The Norwegian firms would seem to have benefited substantially from changes in firms structure (ORGSTR), and to a lesser extent, from changes in knowledge management systems (ORGSYS) and changes in external relationships (ORGREL).15
13 14
Nonetheless, the same signs are found in the OLS estimations (Models III and IV).
The results (not reported here; available upon request) of a detailed analysis of different effects (six types of effects as dependent variables, one at a time) of these three types of change also go along similar lines as the evidence discussed here using factor score (EFORG) as a dependent variable. This finding somewhat conflicts with the basic view of organizational ecologists, that change in an organizations structural core, which naturally impinges on, or even disrupts, some of its existing major routines (i.e. reduces reliability and accountability), hinders its performance.
15
Constant Persistent organizational change (PASTORG) Complementarity (INCOMP) Firm size Number of employees (LogEmp) Total turnover (LogTurn) Firm age (LogAge) Industry dummies (IND) Organizational innovation (in OLS only) ORGSYS ORGSTR ORGREL Selection equationHeckman Stage 1 (dependent variable ORG) Past organizational change (PASTORG) Profitability growth (PASTPERF) Hampering factors (HAMP) High innovation costs (HCOST) Lack of funds (HFUND) Lack of qualified personnel (HPER) Firm size Number of employees (LogEmp) Total turnover (LogTurn) Firm age (LogAge) Industry dummies (IND) Mills ratio Wald-test R2 Number of observations Uncensored
1297
1298
6. Concluding discussion
Although the importance of organizational innovation has been widely recognized, it has been subject to limited research (Lam, 2004; Damanpour et al., 2009), even though historical research has demonstrated its importance for technological and economic change (Chandler, 1962; Bruland and Mowery, 2004). In this article, we have undertaken an initial attempt to improve our understanding of organizational innovation and its effect on firm performance, paying particular attention to its interrelationship with technological innovation and the impact of persistence. Three hypotheses were formulated and empirically tested using two waves of CIS data for Norway and a Heckman regression that tests for selection bias. Extending recent research on determinants of organizational innovation (e.g. Schmidt and Rammer, 2005) and the persistence of technological innovation (e.g. Peters, 2009; Clausen et al., 2012), our article has shown that past organizational innovation is a positive and significant predictor of current organizational innovation (H1) and that the effects from undertaking current organizational innovation are enhanced by prior experience with organizational innovation (H2). The influence of diversity of organizational change on firm performance has also been partially assessed, and the evidence shows that the three types of change considered influence firm performance to different degrees. We also find, in line with the argument that the complementarity of organizational and technological innovation is part-and-parcel of economic change (Nelson, 1991), that the combined effect of undertaking both types of innovation on firm performance is positive and significant (H3). Put differently, firms can better reap the rewards of reorganization by jointly reorganizing with technological innovation. Most papers have shortcomings however, and ours in no exception. Since the Norwegian CIS4 was conducted around the middle of 2005, there was only a short time for the respondent organizational innovators to realize the effects of organizational innovation introduced between 2002 and 2004. Therefore, the analysis could only show how the firms benefited from organizational change in the near term. This limitation relates to the cross-sectional nature of data from the CIS, which may also lead to a simultaneity problem in some cases, because certain variables (which refer to the same, or an overlapping, time period) included in an estimate may be jointly determined. Furthermore, the relationships between some of the variables included in the analysis in the present study may have been influenced by common method bias, because they were extracted from the CIS questions which used similar scale format and/or anchors.16 This bias may have been the case, since these questions
16
Strong correlations between such variables may have been, in part, due to this reason. Criscuolo et al. (2007) explain that, in order to attempt to avoid this bias, the CIS questionnaire was designed to incorporate a mixture of Likert scales and questions which required responses in a binary (yes/ no) or numerical format (absolute numbers, percentages), so that the respondents needed to answer the questions in different parts in different ways. For example, the variables used to measure
1299
were answered based, in part, on the (same) respondents (subjective) evaluation. The reliance on the respondents subjective knowledge or perception may also have led to subjective indicators in the estimate, such as in the case of the CIS questions about obstacles to innovation (Clausen, 2008). Furthermore, it can be argued that the data on organizational innovation made available by the CIS4 is not very detailed. The CIS4 provides only three measures with no scaling of the magnitude of organizational innovation, and, as discussed earlier, these measures are at the firm level (but not plant- or project-level). Therefore, the heterogeneity of organizational innovation within and among firms could not be taken into account in great detail in this study. Moreover, there may be other interesting organizational issues to be investigated on the basis of the CIS data (arguably, the most detailed large-scale survey data currently available for innovation research). For example, it is possible to look further into the differential and complementary effects of different types of organizational innovation (such as by means of a multivariate analysis), or of different combinations of technological and organizational innovation. The relationship between knowledge or skilled workers and organizational change also remains to be explored.17 These are examples of important future research topics which, nonetheless, go beyond the scope of this study. Another issue owing to the scope of the study is that we have not had the opportunity to analyze in detail the role of the industrial environment for organizational innovation and its effects. Although we have controlled for industry fixed effects in our analysis, the influence of these on organizational innovation and its effects needs to be better understood. For example, our results show that firms in some industries are more inclined to report an organizational innovation when compared to firms in other industries but that no industry fixed effect is significant when attempting to explain the effects of organizational innovation. In this article, we have focused on the firm level, but an analysis of how and why organizational innovation differs across industries would add to the existing research within innovation studies on the inter-industrial heterogeneity of (technological) innovation (e.g. Pavitt, 1984; Marsili and Verspagen, 2002). Lack of emphasis on this issue is a weakness of the present article but suggests an interesting venue for further research.
organizational innovation and its effects in this analysis were extracted from two (consecutive) question sets which were associated with yes/no and Likert-scale items. As described earlier, the variables for (the three types of) organizational innovation are binary, while the variables for (the six types of) its effects have a scale of 03.
17
For instance, Leiponen (2000, 2005) empirically analyzes the relationship between firms innovation and their employees skills/competencies, and suggests that this relationship is complementary. However, her analyses concern innovation in a rather technological sense, e.g. R&D and product/process innovation.
1300
Lastly, our longitudinal approach consists of data about firms at only two points in time (i.e. CIS 3 and CIS 4), which has then been analyzed within a cross-sectional framework. Results from cross-sectional analyses mainly show associations between variables, and there is always the possibility that associations between variables are partly driven by omitted variables. Although we test and control for the presence of selection bias in our analysis (in addition to the fact that some variables are measured using the CIS 3 data, while organizational innovation and its effects are measured using the CIS 4 data), we may not claim causality in this article. Despite this shortcoming, we believe that our results are still interesting because there is a lack of empirical studies on the under-researched but important topic of organizational innovation. We acknowledge however that our article is only a small step in this direction and that more research on this topic needs to be undertaken, especially on the longer-term performance effects of organizational innovation.
Acknowledgements
The authors are grateful to Jan Fagerberg, David Mowery, Bart Verspagen, Fulvio Castellacci, Glenn Carroll, Ed Steinmueller, Paul Nightingale, and Martin Srholec for their advice on this work. Many helpful comments received from anonymous referees at Industrial and Corporate Change are also greatly acknowledged.
References
Aldrich, H. (1999), Organisations Evolving. SAGE: London. Amburgey, T. L., D. Kelly and W. P. Barnett (1993), Resetting the Clock: The dynamics of organisational change and failure, Administrative Science Quarterly, 38(1), 5173. Amburgey, T. L. and A. S. Miner (1992), Strategic momentum. The effects of repetitive, positional, and contextual momentum on merger activity, Strategic Management Journal, 13, 335348. Arrow, K. (1962), The Economic implications of learning by doing, The Review of Economic Studies, 29(3), 155173. Barnett, W. and G. Carroll (1995), Modeling internal organizational change, Annual Review of Sociology, 21, 217236. Becheikh, N., R. Landry and N. Amara (2006), Lessons from innovation empirical studies in the manufacturing sector: A systematic review of the literature from 19932003, Technovation, 26, 644664. Boer, H. and W. E. During (2001), Innovation, what innovation?: A comparison between product, process and organizational innovation, International Journal of Technology Management, 22, 83107.
1301
Bresnahan, T. F., E. Brynjolfsson and L. M. Hitt (2002), Information technology, workplace organization, and the demand for skilled labor: Firmlevel evidence, Quarterly Journal of Economics, 117, 339376. Bruland, K. and D. Mowery (2004), Innovation through time, in J. Fagerberg, D. Mowery and R. Nelson (eds), The Oxford Handbook of Innovation. Oxford University Press: Oxford. Brynjolfsson, E. and L. Hitt (2000), Beyond computation: Information technology, organizational transformation and business performance, Journal of Economic Perspectives, 14(4), 2348. Brynjolfsson, E. and L. Hitt (2003), Computing productivity: Firm-level evidence, The Review of Economics and Statistics, 85(4), 793808. Brynjolfsson, E., L. Hitt and S. Yang (2002), Intangible assets: Computers and organizational capital, Brookings Papers on Economic Activity, 1, 137198. Brynjolfsson, E. and H. Mendelson (1993), Information systems and the organization of modern enterprise, Journal of Organizational Computing, 3, 245255. Catozzella, A. and M. Vivarelli (2007), The catalyzing role of in-house R&D in fostering the complementarity of innovative inputs, IZA Discussion Paper No. 3126. Cefis, E. and L. Orsenigo (2001), The persistence of innovative activities: A cross-countries and cross-sectors comparative analysis, Research Policy, 30, 11391158. Chandler, A. (1962), Strategy and Structure. The MIT Press: Cambridge, MA. Clausen, T. (2008), Search pathways to innovation, TIK Working Papers on Innovation Studies, Centre for Technology Innovation and Culture. University of Oslo, Norway. Clausen, T., M. Pohjola, K. Sapprasert and B. Verspagen (2012), Innovation strategies as a source of persistent innovation, Industrial & Corporate Change, 21(3), 553585. Crepon, B. and E. Duguet (1997), Estimating the innovation function from patent numbers: GMM and Count panel data, Journal of Applied Econometrics, 12(3), 243263. Criscuolo, P., N. Nicolau and A. Salter (2007), The elixir or burden of youth?: Exploring differences among start-ups and established firms in innovation behaviour, Paper presented at the DRUID summer conference 2007, Copenhagen, Denmark, 1820 June. Cyert, R. M. and J. G. March (1963), A Behavioral Theory of the Firm. Prentice Hall: Upper Saddle River, NJ. Daft, R. L. (1978), A dual-core model of organizational innovation, Academy of Management Journal, 21(2), 193210. Damanpour, F. (1987), The adoption of technological, administrative and ancillary innovations: Impact of organizational factors, Journal of Management, 13(4), 675688. Damanpour, F. (1991), Organizational innovation: A meta-analysis of effects of determinants and moderators, Academy of Management Journal, 34(3), 555590. Damanpour, F., R. M. Walker and C. N. Avellaneda (2009), Combinative effects of innovation types and organizational performance: A longitudinal study of service organizations, Journal of Management Studies, 46(4), 650675.
Downloaded from http://icc.oxfordjournals.org/ at University of New South Wales on April 4, 2013
1302
Dosi, G. (1988), Sources, procedures, and microeconomic effects of innovation, Journal of Economic Literature, 26(3), 11201171. Dosi, G. and R. Nelson (1994), An introduction to evolutionary theories in economics, Journal of Evolutionary Economics, 4, 15372. Edquist, C., L. Hommen and M. McKelvey (2001), Innovation and Employment: Process versus Product Innovation. Edward Elgar: Cheltenham. Evan, W. M. (1966), Organisational lag, Human Organization, 25(1), 5153. Evangelista, R. (2000), Sectoral patterns of technological in services, Economics of Innovation and New Technologies, 9, 183221. Fagerberg, J. (2004), Innovation: A guide to the literature, in J. Fagerberg, D. Mowery and R. Nelson (eds), The Oxford Handbook of Innovation. Oxford University Press: Oxford. Fagerberg, J. and K. Sapprasert (2010), Innovation: Exploring the knowledge base, TIK Working Papers on Innovation Studies. Centre for Technology Innovation and Culture: University of Oslo, Norway. Fagerberg, J. and B. Verspagen (2009), Innovation studies: The emerging structure of a new scientific field, Research Policy, 38(2), 218233. Flaig, G. and M. Stadler (1994), Success breeds success: The dynamics of the innovation process, Empirical Economics, 19, 5568. Freeman, C. (1995), The national system of innovation in historical perspective, Cambridge Journal of Economics, 19, 524. Galia, F. and D. Legros (2004), Complementarities between obstacles to innovation: Evidence from France, Research Policy, 33, 11851199. Geroski, P. A., J. van Reenen and C. F. Walters (1997), How persistently do firms innovate? Research Policy, 26(1), 3348. Greve, H. R. (2003), Organisational Learning From Performance Feedback: A Behavioral Perspective on Innovation and Change. Cambridge University Press: Cambridge. Hall, B. H. (2002), Notes on Sample Selection Models, Mimeo, available online at http:// emlab.berkeley.edu/users/bhhall/sampsel.pdf, Accessed on 29 June 2007. Hannan, M. T. and J. H. Freeman (1984), Structural inertia and organizational change, American Sociological Review, 49(2), 14964. Hauser, J. R., G. J. Tellis and A. Griffin (2006), Research on innovation: A review and agenda for marketing science, Marketing Science, 25, 687717. Heckman, J. (1976), The common structure of statistical models of truncation, sample selection, and limited dependent variables and a simple estimator for such models, Annals of Economic and Social Measurement, 5, 475492. Heckman, J. (1979), Sample selection bias as a specification error, Econometrica, 47(1), 153161. Kelly, D. and T. L. Amburgey (1991), Organizational inertia and momentum: A dynamic model of strategic change, Academy of Management Journal, 34(3), 591612.
1303
Kimberly, J. R. and M. J. Evanisko (1981), Organizational innovation: The influence of individual, organizational, and contextual factors on hospital adoption of technological and administrative innovation, Academy of Management Journal, 24(4), 689713. Kline, S. J. and N. Rosenberg (1986), An overview of innovation, in R. Landau and N. Rosenberg (eds), The Positive Sum Strategy: Harnessing Technology for Economic Growth. National Academy Press: Washington, DC. Lam, A. (2004), Organizational Innovation, in J. Fagerberg, D. Mowery and R. Nelson (eds), The Oxford Handbook of Innovation. Oxford University Press: Oxford, See also working paper version: http://mpraub.uni-muenchen.de/11539/1/MPRA_paper_11539.pdf.
Downloaded from http://icc.oxfordjournals.org/ at University of New South Wales on April 4, 2013
Leiponen, A. (2000), Competencies, innovation and profitability of firms, Economics of Innovation and New Technologies, 9, 124. Leiponen, A. (2005), Skills and innovation, International Journal of Industrial Organization, 23, 303323. Malerba, F. and L. Orsenigo (1999), Technological entry, exit and survival: an empirical analysis of patent data, Research Policy, 28(6), 643660. Marsili, O. and B. Verspagen (2002), Technology and the dynamics of industrial structures: an empirical mapping of Dutch manufacturing, Industrial and Corporate Change, 11, 791815. Martin, B. R. (2008), The evolution of science policy and innovation studies, TIK Working Papers on Innovation Studies. Centre for Technology Innovation and Culture: University of Oslo, Norway. Miles, I. (2004), Innovation in services, in J. Fagerberg, D. Mowery and R. Nelson (eds), The Oxford Handbook of Innovation. Oxford University Press: Oxford. Nelson, R. R. (1991), Why do firms differ, and how does it matter? Strategic Management Journal, 12, 6174. Nelson, R. R. and S. G. Winter (1982), An Evolutionary Theory of Economic Change. Harvard University Press: Cambridge, MA. OECD (2005), Oslo Manual: Guidelines for Collecting and Interpreting Innovation Data. OECD: Paris. Pavitt, K. (1984), Sectoral patterns of technical change: towards a taxonomy and a theory, Research Policy, 13(6), 343373. Penrose, E. T. (1959), The Theory of the Growth of the Firm. Wiley: New York, NY. Peters, B. (2009), Persistence of innovation: Stylized facts and panel data evidence, Journal of Technology Transfer, 34(2), 226243. Pettigrew, A. M. and E. M. Fenton (2000), The Innovating Organization. SAGE: London. Raymond, W., P. Mohnen, F. Palm and S. S. van der Loeff (2006), Persistence of innovation in Dutch manufacturing: Is it spurious? CESifo Working Paper No. 1681. Roberts, P. W. and R. Amit (2003), The dynamics of innovative activity and competitive advantage: the case of Australian retail banking, 1981 to 1995, Organization Science, 14, 10722.
1304
Romanelli, E. and M. L. Tushman (1994), Organizational transformation as punctuated equilibrium: An empirical test, Academy of Management Journal, 37(5), 11411166. Rosenberg, N. (1976), Perspectives on Technology. Cambridge University Press: Cambridge. Sapprasert, K. (2007), The impact of ICT on the growth of the service industries, TIK Working Papers on Innovation Studies, Centre for Technology Innovation and Culture, University of Oslo, Norway. Sapprasert, K. (2009), Acknowledging organisational innovation, in W. Ostreng (ed.), Confluence: Interdisciplinary Communications 2007/2008. Centre for Advanced Study at the Norwegian Academy of Science and Letters: Norway. Schmidt, T. and C. Rammer (2007), Non-technological and Technological Innovation: r Europa ische Strange bedfellows? ZEW Discussion Papers 07-052. ZEW - Zentrum fu Wirtschaftsforschung/Center for European Economic Research. Schumpeter, J. (1942), Capitalism, Socialism and Democracy. Routledge: London. Sorensen, J. B. and T. E. Stuart (2000), Age, obsolescence, and organizational innovation, Administrative Science Quarterly, 45(1), 81112. Smith, K. (2004), Measuring Innovation, in J. Fagerberg, D. Mowery and R. Nelson (eds), The Oxford Handbook of Innovation. Oxford University Press: Oxford. Sutton, J. (1991), Sunk Costs and Market Structure. The MIT Press: Cambridge, MA. Teece, D. J. (1980), The diffusion of an administrative innovation, Management Science, 26(5), 464470. van der Panne, G., C. van Beers and A. Kleinknecht (2003), Success and failure of innovation: a literature review, International Journal of Innovation Management, 7(3), 130. ndez (2011), Influence of environmental Wischnevsky, J.D., F. Damanpour and F. A. Me factors and prior changes on the organizational adoption of changes in products and in technological and administrative processes, British Journal of Management, 22, 132149. Wolfe, R.A. (1994), Organizational innovation: Review, critique and suggested research directions, Journal of Management Studies, 31(3), 405431. Zucker, L. G., M. R. Darby and J. Armstrong (1998), Geographically localized knowledge: Spillovers or markets? Economic Inquiry, 36, 6586.
1305
Appendix
Table A1 Principal components analysis for the effects of organizational innovation
Effects of organizational innovation Factor loadings EFORG Reduced response time to customer needs Improved quality of goods or services Reduced costs per unit output Improved employee satisfaction and/or reduced employee turnover Increased enterprises capacity Higher enterprises profitability 0.639 0.699
One factor with eigenvalue 41 detected, which explains 47% of total variance.
Table A2 Correlation matrix for the explanatory variables employed in the model
Age Emp Turn PASTORG PASTPERF HCOST HFUND HPER INCOMP ORGSYS ORGSTR
Age Emp Turn PASTORG HCOST HFUND HPER INCOMP ORGSYS ORGSTR ORGREL
1.000 0.118 0.050 0.006 0.086 0.096 0.052 0.030 0.001 0.009 0.013 1.000 0.595 0.115 1.000 0.051 1.000 1.000 0.016 0.034 0.002 0.002 0.024 0.014 0.043 1.000 0.762 0.556 0.387 0.126 0.186 0.177 1.000 0.555 0.355 0.142 0.200 0.163 1.000 0.330 1.000 0.138 0.250 0.181 0.228 0.125 0.142 1.000 0.434 0.257 1.000 0.400
PASTPERF 0.102 0.008 0.050 0.004 0.001 0.011 0.149 0.022 0.006 0.135 0.054 0.084 0.132 0.160 0.144 0.036 0.122 0.039 0.230 0.080 0.129 0.063 0.181 0.020 0.123
Age, Emp (Number of employees), Turn (Total Turnover) and PASTORG (Past/Persistent Organizational Change) are included in Heckman-Stage 1 and 2 and OLS estimation. INCOMP (Complementarity) is included in Heckman-Stage 2 and OLS estimation. PASTPERF (Productivity Growth), HCOST (High Innovation Costs), HFUND (Lack of Funds), and HPER (Lack of Qualified Personnel) are included in Heckman-Stage 1. ORGSYS, ORGSTR, and ORGREL are included in OLS estimation.