Professional Documents
Culture Documents
Not to be quoted
Abstract
This paper examines patterns of popular trust in political and public institutions in
Nepal. Thereafter, the paper examines to what extent such trust is linked to, on the
one hand, Nepalese citizens' social and political identities, and on the other hand, to
citizens' perceptions of institutional performance. Our findings demonstrate that trust
varies extensively among different public institutions. Trust is high in a number of
professional institutions such as schools and hospitals. It is also quite high in local
government institutions such as the Village/Town and District Development
Committees. Trust in the parliament and the government is much lower. Second,
identity based trust such as demographic and social characteristic of respondents, their
caste, religious, and political identities have less explanatory power on the level of
citizens trust in political and public institutions compared to performance based
variables. Trust in political and public institutions primarily depend upon how citizens
assess the performance of such institutions. Hence, patterns of institutional trust
depend very much on how respondents evaluate the current macro political situation
of Nepal, whether recent political changes are judged to be in the right direction.
1. INTRODUCTION
The purpose of this paper is to examine patterns of popular trust in public institutions
in Nepal. Trust in public institutions implies that citizens extend positive expectations
to the members of such institutions and assume that they follow procedures that will
produce beneficial outcomes for themselves and for society at large. When citizens
trust institutions they may believe in the normative idea of an institution, i.e. that a
democratically elected parliament will be the best way to govern a society, or that an
independent court system will secure a fair and neutral interpretation and application
of the laws adopted by the parliament. Furthermore, trust in public institutions
presupposes that people have confidence in the mechanisms established to sanction
the behaviour of office holders, so that when the latter deviate from what is prescribed
they will be held accountable.
In what follows we explore to what extent such trust is linked to, on the one hand,
Nepalese citizens' perceptions of institutional performance, and on the other hand
citizens' social and political identities. The study is built on the assumption that the
more trust citizens have in public institutions and the process of governance, the
closer is the relationship between the state and society. Such close proximity may
enhance partnership between government and civil society, foster democratic
practices, and facilitate better provision of public services (Askvik and Bak, 2005;
Boukaert et al. 2005: 460; Putnam, 1993; Evans, 1996; Fukuyama, 1995). In fact trust
in public institutions can be interpreted as a basic requirement for the proper workings
of a democratic political regime: it tends to promote popular support and reduce
resistance against the regime.
Why is trust a relevant issue in Nepal? After the recent political turmoil in Nepal, the
country has now opted for a democratic constitution and is undergoing reforms in
order to reduce the influence of monarchism and elitism in the government. The (new)
Nepal is promising to be more democratic on the basis of more inclusiveness and
representativeness of all sections of the community. In order to achieve this, trust is a
key variable to reach consensus about the type of government and the process of
governance in Nepal as well as for enhancing collective action. Trust lowers costs of
transactions, reduces risks and uncertainties in choosing an option, and makes the
actions of individuals and organizations more predictable. Without people confidence
and trust in government institutions, the realization of all the promises may be a far
cry and the system of governance in the country may again struggle with the problem
of legitimacy.
Sztompka (1999:41-45), on the other hand, argues that people may have different
targets of trust, and social groups, organisations, and institutions may all stand out as
objects of social trust because they encompass human activity and human beings.
When Sztompka trusts Lufthansa and is prepared to fly with them, the meaning of
which is that he trusts "their pilots, the cabin crew, the ground personnel, technicians,
controllers, supervisors". In order to trust them he does not have to engage with all of
them in face to face interaction. He will have some notions of Lufthansa and its
employees based on various sources: certain limited personal experiences, references
from friends, commercials, newspaper reports, TV etc. According to Sztompka the
difference between interpersonal trust and trust directed towards other social objects
is not a fundamental one in so far as also the latter imply some references to human
actions. As suggested by Fukuyama (1995; 1999), the radius of trust may cover
different objects in different societies: from the strong, specific trust most people
extend to members of their own family, their friends and immediate neighbours to the
3
In a parallel fashion we assume that people may extend trust to organisations and
institutions. Standing out as combined structures of rules, roles and human actors that
generate activities, people may trust or distrust such entities depending upon how they
perceive them, and how they assess their actions. The kinds of institutions we focus
upon in this paper include the key public and political institutions of Nepal: i.e. the
parliament, the central government, the civil service, the judiciary, and the police.
When we ask people in Nepal to what extent they trust these institutions we
presuppose that they have an image of each institution as a combination of people,
positions, procedures, and processes. Such images may be more or less stable and
distinct, and they may be more or less anchored on valid information. In order to
construct their pictures of government institutions Nepalese citizens do not need to
know in detail how various institutions are designed. Yet their images of the
parliament or the civil service imply that respondents develop some expectations
towards the institutions in question, in particular how they will act, and whether their
actions will be beneficial to different social groups. A recent study from Bangladesh
demonstrated that poor people may have quite elaborate conceptions of the history
and structure of the national government, and their sources of information may be
mass media, exposure to election campaigns, and own experiences of approaching
civil servants (Ali and Hossain 2006).
When citizens indicate that they trust certain institutions we interpret that to mean that
they find these institutions trustworthy: i.e. based on the relevant institutional images
they have constructed, and in accordance with whatever criteria people adopt to
decide if an institution can be labelled trustworthy. As suggested below such criteria
may manifest themselves as generic, normative expectations:
Below we distinguish between two main forms of institutional trust: trust emanating
from social identity groups and trust judgements based on institutional performance.
Trust based on social identity refers to a relationship where trust primarily is extended
to members of a particular social identity group. We assume that people tend to trust
each other when they perceive themselves to be bearers of a common identity, be it
extended family, social class, ethnicity, religion, geography etc. Such identity based
trust may be extended to public and political institutions when the institutions in
question are seen as representing the interests and values of certain identity groups, in
Nepal, for instance, defined by social position, caste, or regional and religious
affiliation. We ask to what extent are social identities affecting Nepalese citizens
trust in public institutions.
Performance based trust, on the other hand, refers to trust anchored in how citizens'
assess the current policy achievements of public institutions (Van de Walle and
Bouckaert, 2003). The main idea is that when citizens are satisfied with the output of
relevant institutions, they will tend to trust and support those institutions. Thus from a
performance perspective we ask to what extent is institutional trust in Nepal linked to
performance evaluations. Institutional performance may be evaluated from different
perspectives. Political assessments may focus upon how democratic principles are
developed and implemented, whether human and political rights are respected, if
elections are free and fair etc. Economic evaluations are concerned with how
government institutions are contributing to economic growth and development, and
whether they promote economic well-being. From their study of political support in
post-communist regimes in Central and Eastern Europe and the former Soviet Union
Mishler and Rose (2002) concluded that assessments of political performance had a
greater impact than economic performance. However, as Pharr et al. (2000) have
observed, it is important to keep in mind that citizens' performance evaluations reflect
subjective perceptions rather than objective measurements, and as such they are
dependent upon access to information through mass media and other sources.
Institutional trust is taken as dependent variable here because of the argument that it is
democratic capital (modern political institutions, electoral processes, social and
economic structural reforms, and accountability) and not social capital as argued by
Putnam (1993) that fosters democratic transition in post-colonial societies (Mitra,
2008: 557). According to the main argument of social capital cultural attributes such
as trust, civic associationism, shared norms, and social network in society trickle up to
the political institutional level and makes democracy work. In contrast, in societies
such as Nepal where caste, religions and linguistic groups have remained unchanged,
it is the political structure rather than the social structure that is the main driving force
of social change. The same is argued by studies on Scandinavian countries that the
performance and the organization of democratic and bureaucratic institutions generate
trust (Kumlin and Rothstein 2005: 343; Rothstein, 2004; Rothstein and Stolle, 2001).
Flanked by India in the South and the Tibetan region of China in the North, Nepal is a
landlocked country. The issue of trust in public and political institutions is very
relevant in the context of Nepal because of varied and sometimes volatile political
developments (Whelpton, 2005). The country in its present form was established as a
kingdom more than two hundred years ago. Since then the country has experienced
the rule of a dozen kings, the Ranas (1846-1951), party less Panchayat system (19601991), and at present a recently installed elected government through the constituent
assembly election in 2008. Nepal has now become a federal democratic republic by
abolishing monarchism, a secular state from a previously and only Hindu Kingdom,
and now a coalition government comprising different political parties and interests.
However, the Maoists who vehemently opposed monarchism won the majority of
seats to this Assembly. 2 Prior to this constituent assembly which was sworn in 2008,
Nepal experienced decade long Maoist armed insurgency in the period 1996-2006 and
mass movement for the restoration of democracy to put an end to the feudalistic
power
structure
and
the
hegemony
of
the
monarchy
(http://www.nepalelectionportal.org/EN/political-development/concise-history.php,
accessed 27th November 2008).
Communist Party of Nepal (Maoists) won 229 of 601 seats to Constituent Assembly election held in
April, 2008 (http://en.wikipedia.org/wiki/Elections_in_Nepal, accessed on 29th January, 2009).
3
C R Uprety and Tek Nath Dhakal (1992): Inculcation of Political Culture: Urgent Need for Nepal. A
Conference paper presented at the First SAARC Political Science Conference on Democracy in South
Asia: Problems and Prospects held on June 19-20, Kathmandu: Political Science Association Nepal.
people sovereign. However, less than six years after the restoration of multi-party
system, the Communist Party of Nepal-Maoist started an armed insurgency in 1996
claiming
for
more
justice,
fairness,
and
inclusion
in
governance
(http://www.nepalelectionportal.org/EN/political-development/political-history.php,
accessed 27th November, 2008).
In the wake of escalation of Maoist insurgency which almost destabilized the state a
tug of war between political parties followed. The lack of democratic culture among
the leaders of different political parties has seriously disrupted the functioning of
political institutions. In addition the Palace massacre in 2001 opened the door for
King Gyanendra to take control over state power in 2005. However, his absolute rule
was short lived as people under the banner of different political parties staged massive
demonstrations for the restoration of democracy. Meanwhile, political parties formed
an alliance with the Maoist rebels that ended the decade long violence and insurgency.
As a result of peoples movement, the king, eventually, relinquished power and an
interim government took over in 2006 whose major task was to hold election to the
Constituent Assembly. Election to the Constituent Assembly was held in 2008 with a
mandate to formulate a new constitution for a so called new Nepal. However, in the
absence of political culture of democracy coupled with diverse regional based
interests, the country faces an uphill task when the issues of participation of different
ethnic groups and inclusions of hitherto neglected castes such as Dalits4 in the process
of governance are concerned. As the process of governance taking shape at a slow
pace, many interested parties view this as an opportunity to put forward their demands
and be part of the political bandwagon to power. As a result, the restoration of
democracy is now hanging on numerous tug-of wars and horse-trading between
different political parties.
Coupled with political divisions and cleavages, the small Himalayan state it is argued,
is divided into castes, religions, and ethnic groups rather than economic classes
(Berreman, 1979, Sharma, 1977). Hindu customs, values, and traditions are deeply
ingrained in society and are nurtured in the various institutions of state as well. Caste
4
In the Indian caste system, Dalit means oppressed. A Dalit, often called an untouchable, or an outcast,
is a person who does not have any varnas (color) (Berg, 2007: 5). They are also known as Scheduled
Caste (Wikipedia)
If socio-economic and ethnic identities affect the level of trust in political and public
institutions then Nepal may have a long way to go to establish legitimate governance
acceptable to all. These ethnic and religious, especially caste cleavages are permanent
structures that often are reasons for exclusion of the lower caste from participation
and unequal share in public policies. In a situation where trust is more within ones
own caste, class, and ethnicity, then civic associationism and solidarity across
different groups may become difficult to achieve, which may in turn affect firm
footing of democratic norms.
On the other hand, if trust in institutions is contingent on how policies reflect peoples
needs and demands, then political stability and proper functioning of democratic
institutions are necessary to enhance trust.
4. METHODOLOGY
The study is based on a country wide door to door questionnaire survey of Nepal
where 1836 (originally the intention was to survey 20000 households) households
participated. 5 In order to achieve a representative sample size from all regions in
Nepal, we used the ecological and development regions which are generally used to
divide Nepal geographically from east to west and from north to the south. In
addition, rural-urban classification was also used to represent both urban and rural
areas. Households were randomly selected
Geographically Nepal is divided into three ecological and five development zones.
Ecological zones divide Nepal into three east-west zones. These are Terai (plain land
in the south bordering India), Mountain (high mountainous region in the north
bordering China), and in between is the Hill region where the capital Kathmandu is
located. Population wise, Terai and Hill zones are densely populated areas, while the
northern mountainous region is the least and thinly populated region. In addition, the
country is divided into five north-south administrative development zones. These are
Eastern, Central, Western, Mid Western, and Far Western regions.
10
11
Development
Regions
Districts
Ecological Regions
Mountain
RURAL
WDR
M/FWDR
CDR
URBAN
EDR
Total
CDR
WDR
Gorkha
Mustang
Kalikot
Darchula
Kailali
Banke
Udaypur
Mahottari
Ilam
Kathmandu
Lalitpur
Bhaktapur
Dolakha
Kaski
Kapilavastu
Dang
Morang
M/FWDR
EDR
5 Administrative Deve17 Districts
lopment
Regions
Hill
100
Total no. of
respondents
Terai
100
100
100
100
45
46
71
103
175
100
100
100
197
99
100
100
200
100
100
100
675
1836
45
46
71
103
175
100
100
100
197
99
100
100
200
191
970
According to the population census of 2001, the rural population of Nepal is around
75%. However, in our survey the proportion of urban and rural populations are almost
the same which means urban areas are more represented and given more weight
because of the higher level of education, political activism and awareness. After
selecting districts, households were selected both from VDCs and Municipalities
(districts are further divided into Village Development Committees which are rural
local governments and municipalities which are urban local governments).
Respondents from each household were selected on the basis of age (those who are 18
and above) and gender in order to obtain a sample that is representative of the
population. Every fifth household was selected and administered questionnaire. Yet,
in the case of the mountain districts such as in Kalikot, Mustang and Darchula every
second household was selected because of sparse population scattered across a large
12
geographical area. This method was followed to obtain a sample that is representative
of Nepals ethnic and religious diversity.
Female
Male
Master degree or
higher
Graduate
Education Secondary
Primary
Literate
Illiterate
61 and above
Age
46-60
Groups
31-45
18-30
Hindu
Buddhist
Muslim
Religion
Kirat
Christian
Others
Gender
Respondents
(our sample)
%
45
55
Population
Census Data
of Nepal
2001*
%
50
50
5
23
42
5
12
12
5
19
38
38
87
8
3
.7
.7
.4
54
46
8
11
23
58
81
10
4
4
.4
.4
Ca. 23
1836
Total
million
* Population Census Data 2001, Central Bureau of Statistics,
Kathmandu, Nepal (http://www.cbs.gov.np/national_report_2001.php,
accessed 20th November, 2008)
The above table describes demographic characteristics of our sample and compared
with the Population Census Data of 2001 carried out by the Central Bureau of
Statistics of Nepal. Compared to the Population Census Data which includes all age
categories in the population, our sample collected data from only those respondents
who have filled 18 years of age and above. Also given the length and complexity of
our questionnaire which included mapping opinions on diverse economic and social
13
The sample was reduced to 1836 because of political violence in some Terai districts
and inaccessibility to certain Mountain districts. The survey was carried out in the
period February-April, 2008. Prior to the nation-wide survey, a pilot survey was
carried out in October 2007 to test the questionnaire.
As noted above the dependent variable of our analysis is trust in public and political
institutions. Table 1 below presents the overall distribution of responses to the main
question of how much confidence people have in various institutions in Nepal.
Altogether 18 institutions were listed. The numbers (%) suggest that, for instance, for
political parties 36% had no confidence at all, 38% had not very much confidence,
22% had quite a lot of confidence, while 5% had a great deal of confidence.
When we add the two last categories of responses the sum indicate that 27% of the
respondents have "positive" confidence in the political parties. We use this sum as an
indicator of the overall amount of trust in each institution. In the table institutions
have been ordered from lowest trust at the top (political parties) to highest at the
bottom (school, college).
We observe that there are great variations among the institutions in terms of how
many people who trust them. The most popular institutions are schools and colleges,
university, hospital, media. Also local institutions such as village/town development
committees, chief district administration offices and district development committees
attract a lot of trust. Less popular are central national institution such as the
judiciary/courts, the civil service, the army, and the police. Among civil society
organisations, trade unions are more popular than NGOs. Central government and the
parliament attract also less trust than most institutions. The exception is the king and
political parties which appear at the very bottom of the institutional trust ranking,
14
indicating that most Nepalese citizens do not have very positive attitudes toward these
institutions.
36
Quite a lot
confidence
(3)
22
A great deal
confidence
(4)
Sum of quite
a lot+a great
deal
27
King
45
25
21
29
Central Government
18
37
40
45
Parliament
15
39
41
46
CIAA
20
31
36
13
49
The Police
17
31
44
53
NGOs
12
32
46
10
56
Army
12
29
44
14
58
Civil Service
10
31
51
59
Judiciary/Courts
10
23
52
14
66
District Development
Committee
25
62
70
Trade Unions
23
51
20
71
Chief District
Administration Office
23
63
72
Village/Town Development
committee
19
63
13
76
Media
15
55
27
82
Hospital
12
46
39
85
University
50
39
89
School / College
51
40
91
Question: I am going to name a number of organizations and institutions. For each one, could you tell
me how much confidence you have in them: is it a great deal of confidence, quite a lot of confidence, not
very much confidence or none at all?
More generally, differences among the institutions suggest that some institutions are
less contested than others. Among the least contested are institutions in education,
health, and media and local government institutions. All of these attract a lot of trust.
However, in this paper we want to focus on confidence in the key political institutions
of Nepal and we have concentrated on the following, i.e. the parliament, the central
government, the civil service, the judiciary, and the police. These are less trusted than
some of the professional and local institutions included in the above table, yet
probably more important and critical, we believe, for the emerging new political
regime of Nepal. In the data analysis we assume that trust in these key governance
institutions may be interpreted as having a strong common component which provides
the basis for constructing an index of institutional trust. This common component
15
refers to the whole set of key institutions that are critical for development of the new
Nepal.
Originally we considered to also include the king and the army among the key
national institutions, but an examination of the correlation matrix of institutional trust
(see Appendix table A1) and an exploratory principal components analysis (presented
in Appendix table A2: the two-factor model), suggested otherwise. If these two
institutions were included, a two-factor solution would be more appropriate for
explaining variations in the data. Most probably, because they are seen as belonging
to the old political regime that has been toppled, trust and distrust in these two
particular institutions is not part of the common institutional trust pattern we seek to
identify. To some extent, the data indicate, that may also be the case with the police
and the judiciary, but since they score high on both dimensions, we infer that they are
also perceived as part of the new political system. Hence we have decided to include
them in our index despite their association with the king and the army. Furthermore,
in support of such a decision, when we include the five said institutions the principal
component analysis produces a one-factor model which explains more than 50 per
cent of the total variation (Appendix table A2: one-factor model). The high factor
scores of the police and the judiciary definitely suggest that they are influenced by the
common factor.
We interpret the above to support our assumption that we can use the responses to the
five institutions in question as indicators of a generalised, common trust attitude
toward central governance institutions in the emerging new political regime of Nepal
at the time of the data collection. On the basis of the above analysis we have created
an index of institutional trust in the core institution (see Appendix table A2). The
index will be used as a measure of our dependent variable. The index varies between
0 and 15, where 0 means that a respondent have no confidence at all in any of the five
mentioned institutions, and 15 means that a respondent have a great deal of
confidence in all of the five institutions. Of the total distribution 2.6 per cent of
respondents have a minimum score of 0 and 4 per cent have a maximum score of 15
(Mean = 7.23).
16
(i)
and social position. We have examined to what extent do such variables as age,
gender, living area, region, formal education employment sector, and social class
affect trust in public institutions. When we bring all of these variables together in one
model it appears from table 4 that such a model only explains 5.7 per cent of the total
variation in the institutional trust index. Most of the individual variables in the model
do not have a significant impact. Yet the two variables indicating living area suggest
that people living in rural (as opposed to urban) areas and the population of the Terai
region are more inclined to trust public institutions.
Table 4: Regressions of identity variables on trust in public institutions.
Dependent variable=trust index. Method=enter. Standardized beta coefficients.
Age: high
Gender: female
Area: rural
Region: Terai
Formal education
Civil servant
Soc.class
(Brahmin)
Chettri
Baysha
Sudra
(Hindu)
Buddhist
Muslim
Religious
Nationalism
Political interest
Congress
UML
Maoist
Demography,
education,
social position
,035
,034
,206***
,111**
-,009
,046
-,034
Caste and
religion
Political
identities
Combined
model
,006
,075
,167***
,084*
-,030
,047
-,006
-,006
-,051
-,059
--,043
-,007
,090**
--,058
-,063
-,080*
-,010
-,096*
,164***
,095**
,080**
,075*
,006
,121**
,120**
,085*
,076
,025
17
Explained
variance (R2)
.057
.011
.047
.096
n=926
n=1155
n=1208
n=713
*p<.05 **p<.01 ***p<.000
NB for caste is Brahmin, and for religion is Hindu excluded from the analyses, because they
are used as reference categories. Also for religion, a small non-uniform group (1.8%) of
Christians, Kirats and others is removed from the analyses. Multicollinearity among
independent variables should not be a problem since the highest Pearson's r=0.31.
(ii)
trust encompasses caste and religion. We have examined to what extent Chettris,
Bayshas and Sudras deviate significantly from the other castes in terms of trust scores,
or whether Buddhists or Muslims differ from Hindus. Also here the amount of
explained variance is very limited: i.e. 1.1 percent. The data of table 4 suggest that
there are no significant differences among different castes and among the various
main religions. How respondents view public institutions is not very much affected by
which specific caste or religion they belong to. The only exception is the more generic
religiosity of respondents: whether they see themselves as a religious person or not.
The data suggest that people who, regardless of any specific belief system, see
themselves as religious, are more inclined to trust the public institutions than, for
instance, atheists or persons without a strong religious belief. Yet the correlation is
very small, and the overall finding is that the impact of caste and religion is very
limited.
(iii)
The third group of variables that may be linked to identity is what we have
18
that political identities may affect institutional trust patterns in certain ways, and more
so than caste and religion.
(iv)
regression equation. This model explains almost 10 per cent of the total variation
which is clearly more than any of the other models. From table 4 it appears that most
of the coefficients stand out as statistically significant even when we, in this way,
expand the number of independent variables. Thus, following from demographic
identity characteristics, if respondents are living in a rural area, and/or if they are from
the Terai region they are more inclined to trust the public institutions. Resulting from
religious identity characteristics, in the combined model there is also an indication
that respondents viewing themselves as religious are more trusting, while those who
classify themselves as Sudra are slightly less trusting of public institutions. And from
the political identity characteristics we can infer that taking pride in being a Nepalese
citizen and/or expressing interest in politics contribute to institutional trust in a
positive way. So does also political support of the Congress party.
In general, however, we may conclude from this part of our analysis that the impact of
identity variables on institutional trust is limited. Although we do find some
correlations the overall impression is that neither of these is very strong.
19
see how far different explanatory models can take us when we use regression
techniques.
(i)
(ii)
The second model focuses on how respondents view the current and previous
political system of Nepal and how they assess its development toward full democracy.
From table 5 it appears that positive evaluations of the political system, both of today
and yesterday, clearly are linked to respondents trust judgements. All the coefficients
are statistically significant, and in particular one coefficient suggests a strong
relationship between institutional trust and to what extent the people now holding
government positions are handling the country's affairs in a satisfactory way. For all
the five variables taken together the amount of explained variance of 26 per cent
demonstrates that assessments of political system performance and democratic
development make up an important component of institutional trust evaluations.
(iii)
A third model taps the impact of popular attitudes toward politicians. To what
degree are politicians perceived as competent, knowing what they are doing? And are
they, in fact, doing what is right most of the time? Although the survey questionnaire
contains a number of other questions about politicians the above two questions seem
to be those who most clearly link attitudes toward politicians to institutional trust
judgements. For both variables the coefficients reveal significant statistics. From table
5 it appears that the amount of explained variance is about seven per cent, thus
indicating that popular attitudes toward politicians behaviour are relevant for the
overall interpretation of how institutional trust evaluations vary.
20
Satisfied with
life as a whole
Satisfied with
life 5 yrs ago
Satisfied - now
Financial Situation
How well is political
system today
How well was
political system 5
yrs after
Democracy
developing
People at national
office handling
country
This country is run
for benefit of all
people
Most politicians are
competent & know
what they are doing
Politicians do what
is right most of time
Personal
wellbeing
,084*
Democratic
development
Views on
politicians
Views on
civil
servants
Policy
evaluations
,047
-,018
-,014
,085*
,037
,076*
,042
,104***
,094**
,140***
,106**
,309***
,228***
,145***
,086**
,126***
,007
,207***
,087**
,260***
,114****
-,146***
,146***
-,084**
,030
,278***
,117***
-,102***
,019
Explained variance
.017
.263
.070
.151
.108
(R2)
N
1202
1054
1263
1234
1338
*p<.05 **p<.01 ***p<.000
Multicollinearity among independent variables should not be a problem since the highest
Pearson's r=-0.43 and the lowest tolerance statistic is 0.60.
(iv)
Combined
model
.309
884
institutional trust judgements refers to popular views on civil servants. From a number
of questions about civil servants we have identified three variables that seem to
influence our dependent variable more than others: i.e. if civil servants are perceived
to be prompt and efficient, if they are experienced as not very helpful, and/or if they
are seen as reliable. The statistic results demonstrate that all of these variables are
linked to institutional trust evaluations. In particular if respondents consider civil
servants as to be prompt and efficient will they also tend to express trust in public
institutions? The explanatory power of this model is about 15 per cent suggesting that
21
performance evaluations of civil servants are more important than similar appraisals
of politicians.
(v)
The fifth model we introduce in table 5 departs from how respondents more
(vi)
In the final model we have combined all the performance variables in one
regression analysis. The combined model explains almost 31 per cent of the variation
which is hardly five per cent more than the democratic development model, but
otherwise clearly more than any of the other models. Most of the original beta
coefficients are reduced in the combined model and some are not statistically
significant any more. Thus the variables measuring personal well-being become
insignificant in the combined model. So do also variables such as: satisfaction with
the present political system, politicians competence, civil servants reliability, and the
policy performance index. Satisfaction with people now in government office stands
out as the most important single variable of the combined regression model. Four out
of seven significant coefficients originate from the democratic development,
suggesting that assessments of democratic performance probably are the most
important component of trust in public institutions in Nepal.
22
8. CONCLUDING REMARKS
The above analyses have addressed the issue of popular trust in public institutions in
Nepal at a point in time when the political regime is changing and efforts are made to
strengthen democracy. Starting from an assumption that political trust is critical for
the survival of any democratic regime we wanted to examine to what extent Nepalese
citizens trust public and political institutions, and how such trust judgements are
affected by identity and performance based evaluations.
Our findings demonstrate firstly that trust varies extensively among different public
institutions. Trust is high in a number of professional institutions such as schools and
hospitals. It is also quite high in local government institutions such as the
Village/Town and District Development Committees. However, in a number of other
national institutions it is not so high, although when 66 per cent indicate that they trust
the courts this do suggest quite extensive support for the judiciary, and almost
likewise for the civil service. Trust in the parliament and the government is lower, yet
we must admit that it is difficult to compare such survey result across countries since
we do not know whether the survey questions have similar meanings in different
cultures. Still for political parties and the monarchy trust is definitely low since less
than 1 out of 3 respondents indicate that they trust these institutions.
In a second step we have analysed how the institutional trust patterns differ among
different social and political groups. Asking whether various types of group based
identities affect trust patterns, we compared the explanatory power of three different
models: one model based on the demographic and social characteristic of respondents,
another model was linked to caste and religion as sources of identity, and a third
model departed from political identities. For all three models the explanatory power is
rather limited, and the combined model, which included all the relevant variables,
explained less than 10 per cent of the total variance in institutional trust. Still, some
selected variables, such as for instance living area and sense of citizenship, reveal
certain identity based relationships of statistical significance. In general, we conclude
that identities do not appear as prominent factors for explaining variations of
institutional trust in Nepal.
23
The weak relationship between identity variables and institutional trust is to some
extent surprising. Other observers has emphasised that the political institutions of
Nepal until recently have been excluding significant socio-cultural groups from
democratic participation (e.g. Lawoti 2007). Hence, we could have expected more
divisions among advantaged and disadvantaged groups in terms of institutional
support. This seems not to be the case. Our findings probably represent good news for
Nepal, suggesting that trust is not dependent on ethnic, religious and social identities.
On the other hand, the greater explanatory power of the performance model on trust
indicates that better performance of political institutions enhances citizens trust in
these institutions. Positive interaction and experience with public institutions builds
confidence among people in these organizations which may in turn foster democratic
institutions to develop in Nepal.
24
REFERENCES
Ali, Tariq and Naomi Hossain
2006 "Popular Expectations of Government: Findings from Three Areas in
Bangladesh", PRCPB Working Paper No. 13, Programme for Research on Chronic
Poverty in Bangladesh (PRCPB) Bangladesh Institute of Development Studies
(BIDS), Dhaka
Bak, Nelleke and Steinar Askvik,
2005 Introduction in Steinar Askvik and Nelleke Bak (ed.), Trust in Public
Institutions in South Africa, pp. 1-23, Ashgate, England
Berg, Dag Erik,
2007 Sovereignties, the World Conference against Racism 2001and the Formation of
a Dalit Human Rights Campaign Questions de recherch/Research in Question No 20.
(Paris: Centre d'tudes et de Recherches
Internationales (CERI),
Berreman, G.
1979 Caste and Other Inequities: Essays on Inequality; (Meerut: Folklore Institute).
Bouckaert, Geert, Per Lgreid, Steven Van de Walle
2005, Introduction, Public Performance and Management review, (Volume 28,
number 4, June, 2005), pp. 460-464.
Evans, Peter ,
1996 Introduction: Development Strategies Across the Public-Private Divide, in Peter
Evans (ed.) State-Society Synergy: Government and Social Capital in Development,
(University of California, Berkeley, 1996).
Fukuyama, F.
1995 Trust: The Social Virtues and the Creation of Prosperity, Free Press, New
York.
Fukuyama, F.
1999 The Great Disruption Human Nature and the Reconstitution of Social Order
The Free Press, New York
Hardin, R.
1999. Do we want trust in government? In Warren, M.E. (ed), Democracy and
Trust. Cambridge: Cambridge University Press, 22-41.
Jamil, Ishtiaq, Rameshwor Dangal
2009, The State of Bureaucratic Representativeness and Administrative Culture in
Nepal, forthcoming in Contemporary South Asia, June.
25
Kumlin, Staffan, and Bo Rothstein. 2005. Making and Breaking Social Capital:
The Impact of Welfare-State Institutions. Comparative Political Studies Vol. 38 (No.
4):339-365.
Lawoti, M.
2007 'Political Exclusion and the Lack of Democratisation: Cross-National
Evaluation of Nepali Institutions using a Majoritarian-Consensus Framework'
Commonwealth & Comparative Politics Vol. 45, No. 1, 5777, February 2007
Levi, M. and Stoker, L.
2000 Political Trust and Trustworthiness, Annual Review of Political Science, Vol.
3, pp. 475-508.
Luhmann, N.
1979, Trust and Power. New York: Wiley
Miller, A.H. and O. Listhaug, 1990, Political parties and confidence in government: a
comparison of Norway, Sweden and the United States, British Journal of Political
Science 20:357-86
Mishler, W. and Rose R., 2002, Learning and re-learning regime support: The
dynamics of post-communist regimes, European Journal of Political Research, 41, 536.
Mitra, Subrata K.
2008 When Area Meets Theory: Dominant, Dissent, and Democracy in India,
International Political Science Review, Vol. 29, No. 5, pp. 557-578.
Offe, C.
1999 How can we trust our fellow citizens? In Warren, M.E. (ed), Democracy and
Trust. Cambridge: Cambridge University Press, 42-87.
Pharr, S.J., R.D. Putnam, and R.J. Dalton
2000 "A Quarter-Century of Declining Confidence" Journal of Democracy 11.2
pp.5-25
Population Census Data
2001, Central Bureau of Statistics, Kathmandu, Nepal
(http://www.cbs.gov.np/national_report_2001.php, accessed 20th November, 2008)
Putnam, R.
1993 Making Democracy Work: Civic Traditions in Modern Italy, Princeton
University Press, Princeton, New Jersey, 1993.
Rothstein, B. (2004), 'Social Trust and Honesty in Government: A Causal Mechanism
Approach' in J. Kornai, B. Rothstein and S. Rose-Ackerman (eds.), Creating Social
Trust in Post-Socialist Transitions, Palgrave Macmillan, New York.
Rothstein, Bo, and Dietlind Stolle. 2001. Social Capital and Street level Bureaucracy:
26
27
APPENDIX
Table A1:Correlation matrix of trust in key governance institutions of Nepal
Parliament
Parliament
Central
Government
Civil
Service
Judiciary/Courts
The
Police
King
1,000
Central
Government
Civil Service
,572
1,000
,402
,451
1,000
Judiciary/Courts
,299
,353
,382
1,000
The Police
,243
,295
,331
,419
1,000
(,060)
(,086)
,120
,150
,245
1,000
,148
,188
,255
,317
,483
,379
King
Army
Unrotated
One- factor model including five
institutions
Single factor
Parliament
,808
-,016
,725
Central Government
,826
,059
,775
Civil Service
,700
,220
,737
Judiciary/Courts
,539
,418
,683
The Police
,367
,664
,628
King
-,075
,726
Army
,147
,809
Eigenvalue
2,277
1,849
2,531
Percentage of variance
32,5
26,4
50.6
10
11
12
13
14
15
Total
2,6
1,3
2,8
4,7
5,7
9,8
11,6
11,4
14,1
11,0
13,8
6,3
3,0
1,3
,2
,4
100
M=7.23. Std.Dev.=2.92
28
Std.
Deviation
Minimum
Maximum
Mean
1821
1,00
4,00
1,9023
,87411
1827
,00
1,00
,4444
,49704
1836
,00
1,00
,5120
,49999
1836
,00
1,00
,3676
,48230
1828
,00
1,00
,7588
,42796
1277
,00
1,00
,1989
,39933
1721
1,00
5,00
3,2098
,98927
1566
,00
1,00
,3895
,48780
1566
,00
1,00
,2656
,44182
1566
,00
1,00
,2452
,43035
1566
,00
1,00
,0996
,29958
1831
,00
1,00
,8706
,33578
1831
,00
1,00
,0814
,27349
1831
,00
1,00
,0300
,17074
1779
,00
1,00
,4109
,49214
1808
3,50
,661
1633
2,31
1,000
1757
,00
1,00
,2043
,40332
1757
,00
1,00
,1429
,35003
1757
,00
1,00
,0962
,29493
875
Mean
Std.
Deviation
Minimum
Maximum
1739
10
6,64
2,287
1699
10
6,23
2,456
1773
10
6,25
2,447
29
1694
10
3,63
2,513
1481
10
6,71
2,766
1664
2,19
,772
1634
2,10
,745
1650
2,33
,895
1615
2,04
,867
1711
2,06
,715
1643
2,28
,767
1692
2,49
,833
1673
2,10
,769
1825
1,00
4,95
3,1574
,55397
1819
1,00
5,13
3,7481
,73961
1001
30