How Do Network Size, Voluntary Association, and Trust Affect Civic Engagement? Evidence from the Asian Barometer Survey

doi:10.4236/aasoci.2012.24033

Paper Menu >>

Journal Menu >>

Advances in Applied Sociology

2012. Vol.2, No.4, 253-259

Published Online December 2012 in SciRes (http://www.SciRP.org/journal/aasoci) http://dx.doi.org/10.4236/aasoci.2012.24033

How Do Network Size, Voluntary Association, and Trust Affect

Civic Engagement? Evidence from the Asian Barometer Survey

Harris H. Kim

Department of Sociology, Ewha Womans University, Seoul, South Korea

Email: [email protected]

Received August 2nd, 2012; revised September 5th, 2012; accepted September 17th, 2012

This study examines the role of social capital in shaping the individual likelihood of “civic engagement”

defined specifically as informal and formal political participation. Based on a subset of the Asian Ba-

rometer Survey of Democracy, Governance and Development (2006), a representative cross-national

dataset, it examines how and to what extent network size, voluntary association, generalized trust, and

particularized trust differentially influence the political behaviors of the survey participants in Korea. The

dependent variable is measured in terms of first, discussing political topics in an informal social context

and, second, getting together with others in order to raise a political issue or sign a petition. Four inde-

pendent variables are measured: 1) the size of egocentric network; 2) the membership in voluntary or-

ganizations and formal groups; 3) the degree of trust placed in generalized others (i.e., strangers); and 4)

the extent to which survey respondents place their trust in particularized others (those with whom one has

a personal relationship). Quantitative analyses show that, ceteris paribus, network size and voluntary as-

sociation have a strong causal impact on both outcome variables. Generalized trust is found to be a

non-significant factor, however, while particularized trust has a contingent effect. Along with the inter-

pretation of statistical results, their broad theoretical implications are also discussed.

Keywords: Civic Engagement; Network Size; Social Trust; Voluntary Organization; Asia

Introduction

Social Capital as Generalized Trust & Voluntary

Association

What are the social structural conditions under which people

make political decisions and take political actions? Like any

other type of social action, individual political behavior, both

formal and informal types of civic participation, does not take

place in a social vacuum. Whether it’s going to the voting booth,

participating in a protest, or simply engaging in a casual politi-

cal discussion, people’s behaviors are shaped by a number of

relevant situational factors, i.e., those in which individual deci-

sion-making and actions are socially embedded (cf. Granovetter,

1985). When it comes to this particular empirical question,

researchers have made much use of the concept social capital.

According to Putnam’s (1993) much-cited work, social capital-

broadly cast in terms of interpersonal networks, shared organ-

izational membership, mutual trust, and norms of cooperation

and reciprocity, serves to increase political participation and

can improve democratic governance and accountability. There

is a voluminous and growing literature that shows a causal link

between social capital and various political outcomes (see e.g.,

Braithwaite & Levi, 1998; Dekker & Uslaner, 2001; Jamal &

Noorudin, 2010; Mishler & Rose, 2005; Paxton, 2002; Roth-

stein & Uslaner, 2005; Uslaner, 2002; Newton & Zimerli, 2011;

Zmerli & Newton, 2008).

The vast majority of the extant scholarship focuses on the

role of generalized trust and participation in voluntary associa-

tion in facilitating democratic values, support and involvement

on the part of individuals (Marsh, 2005; Tavits, 2006; Zmerli &

Newton, 2008). Concerning the causal relationship between the

propensity to trust others and the likelihood of civic engage-

ment, Rothstein and Uslaner (2005: p. 41) write that “at the

individual level, people who believe that in general most other

people in their society can be trusted are also more inclined to

have a positive view of their democratic institutions, to partici-

pate more in politics, and to be more active in civic organiza-

tions.” Researchers have found a similar causal link connecting

voluntary organizational membership and a certain set of values

and orientations that are conducive to individual democratic

responsibility (Paxton, 2002; Stolle, 2001; Van Egmond et al.,

1998). As Putnam (1993) explains, “associations instill in their

members habits of cooperation, solidarity, and public-spirited-

ness” and that “participation in civic organizations inculcates

skills of cooperation as well as a sense of shared responsibility

for collective endeavors” (90). Using the World Values Survey

data across 43 countries, Inglehart (1997) provides empirical

support for this view: that the stability of democracy among

different nations is indeed highly correlated with the individual

membership in voluntary associations as well as the aggregate

levels of social trust. In a case study of Sweden, Teorell (2003)

offers further evidence by showing that “organizational in-

volvement provides bridging social capital by connecting the

individual to a wide range of people” (49).

Network Size as Social Capital: A Neglected Variable

The preceding discussion serves to highlight two major

strands of research in the extant literature: that is, how general-

ized trust and voluntary association can lead to greater political

interest and participation of individual citizens in democratic

societies. While fully recognizing the past research to be em-

H. H. KIM

pirically and theoretically important, the current study proposes

to examine two additional factors that have been relatively

overlooked, namely network size and particularized trust. That

the network variable, though causally relevant, has not been

investigated fully in the literature is surprising since social

capital is typically defined as trust, norms of reciprocity, and

social networks (Putnam, 1993; see also Rothstein & Stolle,

2008). Part of the reason for this selective scholarly attention

lies with the limited availability of data on interpersonal net-

works and political behavior.

This study has a twofold purpose. Given the relative scarcity

of research findings related to the causal role of social networks,

one of its main objectives is to fill this empirical gap in the

literature. Network size, though understudied, is indeed a sig-

nificant factor driving individual political interest and behavior.

According to the empirical investigation carried out by Lake

and Heckfeldt (1998), along with political interaction frequency

and network expertise, the size of one’s interpersonal network

has a powerful influence on the person’s political actions, while

controlling for a host of background factors. Theoretically,

larger networks are associated with multiple interpersonal

sources that can provide access to novel and timely information

(Burt, 1992). They can also serve as a communication bridge

(cf. Granovetter, 1973). In the case of political participation,

being situated in large networks opens up more opportunities to

be in touch with others who may be politically interested or

active (Huckfeldt et al., 1995).

Particularized vs. Generalized Trust

Along with network size, the concept of particularized trust

has also failed to attract sufficient scholarly attention. A con-

ceptual distinction is drawn between “generalized” and “par-

ticularized” trust in the literature. Simply put, the former is

about trust in unknown others or strangers. It is related to a

situation where “a community shares a set of moral values in

such a way as to create regular expectations of regular and

honest behavior” (Fukuyama, 1995: p. 153) among people who

do not know each other personally. This is very different from

particularized trust which is based on specific face-to-face in-

teractions between concrete individuals (see Bahry et al., 2005;

Hardin’s, 2002; Yamagishi & Yamagishi, 1994). The vast ma-

jority of the past studies attempting to link social capital and

political participation have relied on the concept of generalized,

not particularized, trust. A study by Uslaner and Conley (2003)

is one important exception, which focuses on the latter concept.

They categorize people into two groups: generalized trusters

and particularized trusters. According to their definition, the

former group tends to engage in broad social interactions and

helps to build general social capital that is conducive to the

proper workings of democracy. The latter group, on the other

hand, is prone to remain disconnected from the mainstream

society. The empirical analysis carried out by Uslaner and

Conley (2003) is based on a sample of Asian immigrants in the

U.S. Their findings reveal that particularized trusters, compared

with their counterparts, are indeed less likely to engage in

mainstream American politics.

As Newton and Zmerli (2011) show, however, the relation-

ship between generalized trust and particularized trust may be

more nuanced than as described above. The authors challenge

the conventional notion that particularized trust is “either

harmful or of little importance in modern democracies” (2011:

p. 169) and that only generalized trust is instrumental in creat-

ing social capital that promotes democratic processes. They

contend that depending on the situation the two forms of trust

may be mutually exclusive or supportive. In their analysis of

the World Values Survey (1995-2007), the authors also demon-

strate that both trust types are positively associated with peo-

ple’s attitude toward (level of confidence in) political institu-

tions across 22 countries. In other words, contrary to the impli-

cations of the findings in Uslaner and Conley (2003), they re-

port that those who score higher on the particularized trust scale

are more likely to hold a positive view of their political leaders

and institutions. Based on the above literature review, a fol-

lowing set of hypotheses can be formulated concerning the

concepts of generalized trust, particularized trust, voluntary

associational memberships, and network size, as each relates to

the probability of individual political participation.

Hypothesis 1: People who are more willing to trust general-

ized others are more likely to participate in political activities,

ceteris paribus.

Hypothesis 2: People who are more active in voluntary asso-

ciations are more likely to participate in political activities,

ceteris paribus.

Hypothesis 3: People who have larger social networks are

more likely to participate in political activities, ceteris paribus.

Hypothesis 4a: People who have greater particularized trust

are more likely to participate in political activities, ceteris

paribus.

Hypothesis 4b: People who have greater particularized trust

are less likely to participate in political activities, ceteris pari-

bus.

Hypothesis 4 has two opposing versions since, given the

contradictory views in the literature, it is debatable whether

particularized trust promotes or hinders the likelihood of infor-

mal and formal political engagement. The remainder of this

paper is devoted to evaluating the validity of these hypotheses

through empirical testing. The next section describes the data,

the variables, and the methods used for the quantitative analysis,

followed by the interpretations of the findings and the broad

implications of this study.

Data, Model & Variables

Data analyzed in this paper were collected by the Asian Ba-

rometer Project, which was co-directed by Professors Fu Hu

and Yun-han Chu and received major funding support from

Taiwan’s Ministry of Education, Academia Sinica and National

Taiwan University. This is the second wave of the Asian Ba-

rometer Survey (ABS) collected in 2006 covering major politi-

cal systems in Asia. This study focuses specifically on the Ko-

rean dataset. There have only been a limited number of nation-

ally representative studies on Korea, which previously pro-

duced conflicting results based on older data (see Kim, 2005;

Lee, 2008). The dataset available from the ABS provides more

recent information that can be used to shed novel light on the

role of social capital in individual political participation in Ko-

rea. The survey was conducted during the month of September

(2006) by Gallup Korea. The survey population consisted of all

citizens aged 19 and older residing in the territory of South

Korea at the time it was conducted. The survey, using a

multi-stage probability sampling method, was done through

face-to-face personal interviewing, which lasted on average 60

minutes. A total of 1212 interviews were completed, with the

254

H. H. KIM

response rate of 44.9%.

Dependent Variable

The dependent variable for this study is “political participa-

tion”. A number of questions were asked in the survey to tap

into the respondents’ levels of political interest and behavior. In

this study, the focus will be on one informal and one formal

type of political participation. The exact wording for these two

questions is as follows:

Q.1 When you get together with your family members or

friends, how often do you discuss political matters? (Frequently,

Occasionally, Never).

Q.2 Here is a list of actions that people sometimes take as

citizens. Tell me whether you, personally, have never, once, or

more than once done any of these things during the past three

years. For example, “Got together with others to raise an issue

or sign a petition”.

The answers for the first question was coded on a 3-point

scale (“Frequently” = 3; “Occasionally” = 2; “Never” = 1) to

create the first dependent variable (DISC_POL). From the sec-

ond question, a variable called SIGN_PETN is constructed by

assigning the value of “0” if the answer is never and “1” other-

wise. In the sample, 31% stated that they never talk about poli-

tics with their friends and families; about 59% said that they do

so occasionally, while 10% claimed to do so frequently. When

it comes to getting together with others to raise an issue or sign

a petition, about 12% gave an affirmative answer.

Independent Var iables

A number of independent variables are created to test the va-

lidity of the hypotheses stated above. The standard question in

the literature used to measure generalized trust, as found in the

American General Social Survey (GSS) or the World Values

Survey (WVS), is typically phrased as: General speaking,

would you say that “most people can be trusted” or that “you

must be very careful in dealing with people”? The same word-

ing was used in the ABS. In the sample, about 33% of the Ko-

rean respondents agreed with the question, a figure that is rela-

tively lower in comparison with the earlier studies of Europe

and North America (see Nannestad, 2008). A variable named

GEN_TRUST, coded “1” if answered “yes” to the above state-

ment, is used to test the validity of Hypothesis 1. Hypothesis 2

deals with the level of participation in voluntary groups or or-

ganizations. Under the subheading of “Social Capital,” the ABS

asks subjects to name up to “3 most important organizations or

formal groups you belong to”. Based on the individual re-

sponses, a variable called VOL_ASSOC is constructed.

As discussed above, the variables for generalized trust and

voluntary associations are standard measures frequently used in

the literature. The two additional variables examined in this

study are network size and particularized trust. One of the new

questions asked in the ABS is the following, which was not

available in the first wave of the survey conducted in 2003: “On

average, about how many people do you have contact with in a

typical week day? Please include only people you know.” This

question is designed to directly tap into the size of the interper-

sonal network in which the individual subjects are embedded. It

resembles the question in the GSS, which asks for the number

of confidants with whom one discusses important matters. The

range of answers varies from 1 to 5, with 2.7 being the average

network size for the Korean sample. This mean value is higher

than that for the Americans, which decreased over time, from

2.94 in 1985 to 2.08 in 2004 (McPherson, Smith-Lovin, &

Brashears, 2006). For this study, a variable called NET_SIZE is

created to evaluate Hypothesis 3.

Lastly, in order to test the validity of Hypothesis 4a and 4b,

particularized trust is measured. In addition to generalized trust,

the ABS inquires about the extent to which the respondents

trust three groups of people whom they know personally,

namely “your relatives”, “your neighbors”, and “others you

interact with”. Based on the answers given, three separate vari-

ables are put together using a 4-point scale (“A great deal of

trust” = 1; “Quite a lot of trust” = 3; “Not very much trust” = 2;

“None at all” = 1), namely TRUST_REL, TRUST_NEIGHB,

TRUST_OTHERS. The mean value for trusting one’s relatives

is 3.14, that for trusting one’s neighbors is 2.8, and for trusting

others it is 2.4. According to an earlier study done by Gibson

(2001), 59.6% of the Russians feel that people they know per-

sonally can be “trustful”. In the Korean sample, 59.8% said that

they put “quite a lot of trust” in their relatives and 64.4% ex-

pressed the same feeling concerning their neighbors, while

40.3% gave the same answer for the third trust category (“oth-

ers you interact with”).

Control Variables

A number of relevant control variables are included in the

analysis. According to Newton and Zimerli (2011: p. 183), “a

reading of the already voluminous literature on trust suggests a

fairly short and consistent list of individual variables associated

with it”. In this study, such factors are controlled for—includ-

ing income, education, gender, marital status, age, and the level

of subjective interest in politics. The age variable (AGE) is a

continuous variable with the mean value of 42.7; the income

variable (INCOME) is measured as an ordinal one whose scale

is divided into quintiles; the level of education (EDUC) is

gauged on a 10-point scale (“No formal education” = 1;

“Post-graduate degree” = 10); the gender variable (FEMALE)

is coded 1 if “female”; the marital status variable (MARRIED)

is also dichotomously coded (“married” = 1); and the variable

that gauges the degree of interest in politics (INTEREST) is

coded using a 4-point scale (“Not at all interested” = 1; “Very

interested” = 4).

The descriptive information concerning all the variables is

summarized in Table 1. Table 2 contains the bivariate correla-

tions among the variables. The quantitative results are pre-

sented in two separate tables (Tables 3 and 4) since there are

two different outcome variables under investigation: discussing

political matters (DISC_POL) and signing a petition (SIGN_

PETN). Since the dependent variables are all categorically dis-

tributed, nominal logistic regression models are estimated. The

following section describes the statistical findings and their

interpretations.

Results and Discussion

For each of the two regression tables, two models are pre-

sented. Model 1 contains only the independent variables used to

test the hypotheses. Model 2 introduces the control variables,

offering a more conservative test of their validity. Table 3 re-

ports the findings from regressing DISC_POL on the selected

independent and control variables. According to Model 1, coef

H. H. KIM

256

Table 1.

Descriptive statistics.

Min Max Mean Std. Dev.

INCOME 1 5 3.06 1.22

EDUC 1 10 6.79 1.98

MARRIED 0 1 .72 .44

FEMALE 0 1 .50 .50

AGE 19 80 42.65 14.01

INTEREST 1 4 2.38 .80

NET_SIZE 1 5 2.71 1.17

VOL_ASSOC 0 3 .39 .79

GEN_TRUST 0 1 .32 .46

TRUST_REL 1 4 3.14 .62

TRUST_NEIGHB 1 4 2.80 .60

TRUST_OTHERS 1 4 2.44 .66

DISC_POL 1 3 1.77 .59

SIGN_PETN 0 1 .11 .32

ficients for two of the six variables are found to be statistically

significant (p < .001), namely NET_SIZE and VOL_ASSOC.

In other words, individuals with larger social networks are more

likely to participate in informal political discussions with their

friends and relatives. Also, those with more active organiza-

tional life or formal group affiliation are more prone to engage

in such informal political behavior, a finding that complements

the earlier research by Lee (2008) on Korea. Contrary to the

hypotheses stated earlier, however, neither generalized nor

particularized forms of trust are found to exert any significant

causal influence.

Model 2 tests the robustness of the coefficients in Model 1

by including the control variables. Among them, three have a

significant effect on the outcome variable. Specifically, indi-

viduals with higher educational attainment (p < .001), those

who are older (p < .05), and people who profess to have greater

interest in politics (p < .001) are more likely to converse with

their friends and relatives about political issues. More impor-

tantly, the coefficients for network size and associational

membership remain their significance at .05 and .01 level, re-

spectively, thus offering strong empirical evidence that con-

firms Hypothesis 2 (“People who are more active in voluntary

associations are more likely to participate in political activities,

ceteris paribus”) and Hypothesis 3 (“People who have larger

social networks are more likely to participate in political ac-

tivities, ceteris pa ribus”).

Table 2.

Correlation matrix.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

1. INCOME 1

2. EDUC .400** 1

3. MARRIED .020 −.114** 1

4. FEMALE −.016 −.229** .066* 1

5. AGE −.300** −.611**. .438** .050 1

6. INTEREST .061* .123** .074* −.215** .065*1

7. GEN_TRUST .021 .052 .045 −.048 .024 .095** 1

8. TRUST_REL −.007 .090** .010 −.061* .003 .093** .168** 1

9. TRUST_NEIGHB −.011 −.018 .092** −.037 .122** .100** .289** .499** 1

10. TRUST_OTHERS .023 .070* .014 −.034 −.033 .105** .335** .320** .501** 1

11. DISC_POL .121** .157** .115** −.177** .059*.435** −.005.040 .025 .005 1

12. SIGN_PETN .061* .138** .019 –.049 −.087** .098** .031 .048 −.015 −.021 .093** 1

13. CONTACT .008 .047 .064* −.107** .008 .082**.052 .022 .036 −.041 .108** .146** 1

14. NET_SIZE .143** .103** −.029 −.143** −.060*.105**.059*.049 .032 .060*.104** .087** .121** 1

15. VOL_ASSOC .092** .108** .137** −.190** .083** .170** .086** .050 .095** .045 .198** .128** .183** .121**1

Note: *<.05; **<.01; ***<.001.

H. H. KIM

Table 3.

Logistic coefficients from regressing DISC_POL on selected vari-

ables.

Model 1 Model 2

Estimate S.E. Estimate S.E.

INCOME .106# .062

EDUC .172*** .048

MARRIED .290 .178

FEMALE −.259# .146

AGE .017* .007

INTEREST 1.211*** .101

NET_SIZE .189*** .054 .122* .060

VOL_ASSOC .521*** .082 .287** .089

GEN_TRUST −.163 .140 −.253 .155

TRUST_REL .095 .114 −.032 .128

TRUST_NEIGHB .015 .135 .051 .150

TRUST_OTHERS −.074 .110 −.313# .122

INTERCEPT1 4.090*** .707

INTERCEPT2 8.015*** .757

−2LL 861.853 1463.182

N 1069 983

Note: # < 0.1; * < .05; ** <.01; *** <.001 (two-tailed tests).

Table 4.

Logistic coefficients from regressing SIGN_PETN on selected vari-

ables.

Model 1 Model 2

Estimate S.E. Estimate S.E.

INCOME –.012 .093

EDUC .161* .079

MARRIED .341 .281

FEMALE .207 .221

AGE –.015 .011

INTEREST .352** .136

NET_SIZE .222** .083 .229* .088

VOL_ASSOC .398*** .100 .309** .109

GEN_TRUST .290 .217 .273 .226

TRUST_REL .442* .174 .389* .187

TRUST_NEIGHB –.196 .207 –.125 .221

TRUST_OTHERS –.323# .174 –.450* .181

INTERCEPT1 3.024*** 4.396***

–2LL 417.631 668.580

N 1069 981

Note: # < 0.1; * < .05; ** < .01; *** < .001 (two-tailed tests).

Table 4 contains regression results from examining the sec-

ond outcome variable (SIGN_PETN), which has to do with a

more formal way of political engagement, i.e., getting together

with other citizens in order to raise an issue or sign a petition.

Looking at Model 2, the full model with the control variables

included, a new set of findings that is different from the one in

Table 1 emerges. As was the case with Table 1, people sur-

rounded by a larger number of interpersonal contacts (p < .05)

and those with greater organizational and group ties (p < .01)

exhibit a higher propensity toward political behavior, while

controlling for other covariates. In addition, two new variables

are found to be significant (at p < .05), namely TRUST_REL

and TRUST_OTHERS. Interestingly, however, the causal

flows are in opposite directions. As the negative sign for

TRUST_OTHERS suggests, individuals who shows greater

trust in those whom they know personally are less likely to

engage in formal political behavior, in accordance with Hy-

pothesis 4b. This finding thus lends support to the view pro-

posed by Uslaner and Conley (2003), as mentioned above. To

the contrary, the positive coefficient sign for TRUST_REL

offers a completely different picture, i.e., that people who score

higher on particularized trust are more prone to get involved

politically, which confirms Hypothesis 4a. This regression re-

sult challenges Uslaner and Conley (2003) and instead supports

Newton and Zmerli (2011), thereby adding further to the con-

troversy surrounding the relationship between particularized

trust and individual political behavior.

In sum, the quantitative analyses provide consistent and rela-

tively strong empirical support for Hypothesis 2 and Hypothe-

sis 3, as illustrated by the fact that the coefficients for network

size and organizational or group affiliation are robust across all

the models. Contrary to Hypothesis 1, however, generalized

trust is not related to political participation. In fact, the coeffi-

cients for this variable (GEN_TRUST) are insignificant in all of

the estimated models, a finding that deviates from much of the

previous research that confirms the function of generalized trust

in facilitating individual democratic values, responsibilities and

activities. The two contradictory hypotheses (Hypothesis 4a &

Hypothesis 4b) concerning the causal link between particular-

ized trust and individual political behavior receive partial sup-

port. As shown by the mixed statistical results, particularized

trust is not causally related in any way to informal political

participation, as measured by the frequency of holding political

discussions in a social setting with friends and relatives. None

of the coefficients for the three measures (TRUST_REL,

TRUST_NEIGHB, TRUST_OTHERS) reach the conventional

significance level. But when it comes to being mobilized to

sign a petition or to raise a political issue, trusting in relatives

and trusting in others do have significant, albeit opposite, ef-

fects, though trusting in neighbors is found to be causally in-

significant.

Conclusion

The purpose of this study is to examine the role of social

capital in predicting the level of individual civic engagement, at

both informal and formal levels. It tests four distinct arguments

related to the causal influences of generalized trust, particular-

ized trust, social network, and voluntary associational mem-

bership. Data analysis produces results that are expected as well

as surprising. First, when it comes to the role of voluntary asso-

ciation, measured in terms of membership in an organization or

H. H. KIM

a formal group, this study finds evidence that bolsters one of

the oldest views in the social capital theory—that healthy asso-

ciational life fosters citizen interests and activities in domestic

politics—an argument that dates back to Tocqueville’s (1969

[1865]) original observation about the workings of democracy

in America. Second, generalized trust seems to have no impact

on the individual probability of getting politically involved.

This finding is actually not very surprising in light of what

some scholars have pointed out in the past: namely that “there

are patchy and weak associations between social and political

trust” (Newton, 2001: p. 202) and that “trust has small if any

independent effect on support for the current regime” (Mishler

& Rose, 2005: p. 14). In fact, despite the voluminous literature,

there is still an ongoing controversy concerning the exact link-

age between social/generalized trust and political/institutional

trust, as well as various dimensions of democratic political

engagement (see, e.g., Delhey & Newton, 2003; Mishler &

Rose, 2005; Newton & Norris, 2000; Rothstein, 2002). The

main issue stems from disagreements concerning the definition

and the measurement of this frequently used, yet thorny, con-

cept. Broadly speaking, there are two opposing conceptions of

generalized trust found in the literature: one that is character-

ized by strategic rationality, on the one hand, and one that is

norm-driven, on the other (Nannestad, 2008). Many studies do

not make an explicit differentiation between them but conflate

the two approaches, which calls for future research that better

theorizes about and gauges generalized trust in understanding

its multifaceted causal role.

Third, network size as a relatively understudied concept is

found to be of major significance. In his investigation of Russia,

Gibson (2001) writes that “weak social networks” are critical

for the building of civil society in transitional societies such as

Russia. By “weak,” he means networks that are open and thus

cut across multiple social groups (cf. Granovetter, 1973). This

particular characteristic of network is also related to size (see

Burt, 1992). To the extent that a network is large, there is a

greater possibility of cross-cutting that would enable actors to

be connected to groups that transcend their narrow circles of

contacts based on kinship or other characteristics of similarity.

The argument is that networks that are large, disconnected, or

weak create bridging, rather than bonding, social capital (Put-

nam, 2000: p. 22) that encourages individual political aware-

ness, interest and engagement, which is ultimately beneficial

for the entire society. One shortcoming of this research is that it

relies on network size as a proxy variable and, due to data un-

availability, does not deal with the direct measures of network

openness or density. More nuanced network information is in

order to probe into the complex relationship between interper-

sonal networks and individual political engagement.

Lastly, the quantitative results in this study further add to the

debate concerning the value of particularized trust in promoting

political behavior. Does particularized trust deter people from

engaging in mainstream politics, as Uslaner and Conley (2003)

contend? Or does it serve as a foundation on which generalized

trust and political confidence in government institutions could

be built, as Newton and Zmerli (2011) insist? Perhaps the right

answer is that the role of particularized trust is contingent, that

is, “different types of social networks ··· lead to dissimilar

types of civic engagement” (Uslaner & Conely, 2003: p. 355).

It is the task of researchers to figure out the conditions under

which social networks of one kind result in one form of politi-

cal behavior, while another kind leads to a different form. So-

cial capital theory has provided a great deal of analytical mile-

age in explaining, for example, why people support different

types of political regimes, what democratic values they espouse,

when they are likely to participate in collective action, or how

they decide to trust government institutions. Regretfully, as is

the case with other popular theoretical innovations, the concept

of social capital has run into many conceptual, definitional, and

methodological problems. Greater attention is thus required in

correctly framing future inquiries about the political conse-

quences of social capital so as to maximize its heuristic value

and minimize the pitfalls associated with it.

REFERENCES

Bahry, D., Kosolapov, M., Kozyreva, P., & Wilson, R. (2005). Ethnic-

ity and trust. American Political S cience Review, 99, 521-532.

doi:10.1017/S0003055405051853

Braithwaite, V., & Levi, M. (2002). Trust and governance. New York:

Russell Sage Foundation.

Burt, R. (1992). Structural holes: The social structure of competition.

Cambridge: Harvard University Press.

Dekker, P., & Uslaner, E. (2001). Social capital and participation in

everyday life. London: Routledge.

Delhey, J., & Newton, K. (2003). Who trusts: The origins of social trust

in seven societies. European Societies, 5, 93-137.

doi:10.1080/1461669032000072256

Fukuyama, F. (1995). Trust: The social virtues and creation of pros-

perity. New York: Basic Books.

Gibson, J. (2001). Social networks, civil society, and prospects for

consolidating Russia’s democratic transition. American Journal of

Political Science, 45, 51-68. doi:10.2307/2669359

Granovetter, M. (1973). The strength of weak ties. American Journal of

Sociology, 78, 1360-1380. doi:10.1086/225469

Hardin, R. (2002). Trust and trustworthiness. New York: Russell Sage.

Ikeda, K., & Richey, S. (2005). Japanese network capital: The impact

of social network on Japanese political participation. Political Be-

havior, 27, 239-260. doi:10.1007/s11109-005-5512-0

Inglehart, R. (1997). Modernization and postmodernization: Cultural,

economic, and political changes in 43 societies. Princeton, NJ:

Princeton University Press.

Jamal, A., & Nooruddin, I. (2010). The democratic utility of trust: A

cross-national analysis. Journal of Politics, 71, 45-59.

doi:10.1017/S0022381609990466

Kim, J. Y. (2005). Bowling together isn’t a cure-all: The relationship

between social capital and political trust in South Korea. Interna-

tional Political Science Review, 26, 193-213.

doi:10.1177/0192512105050381

Lake, R., & Huckfeldt, R. (1998). Social networks, social capital, and

political participation. Political Psychology, 19, 567-584.

doi:10.1111/0162-895X.00118

Lee, J. (2008). Path toward democracy in South Korea: Social capital &

democracy embedded in the citizens. Asian Survey, 48, 580-602.

doi:10.1525/as.2008.48.4.580

Marsh, J. (2005). Social capital and democracy in a new democracy.

The Sociological Quarterly, 46, 593-615.

doi:10.1111/j.1533-8525.2005.00028.x

Mishler,W., & Rose, R. (2005). What are the political consequences of

trust? A test of cultural and institutional theories in Russia. Com-

parative Political Studies, 38, 1050-1078.

doi:10.1177/0010414005278419

Nannestead, P. (2008). What have we learned about generalized trust, if

anything? American Review of Political Scienc e , 11, 413-436.

doi:10.1146/annurev.polisci.11.060606.135412

Newton, K. (2001). Trust, social capital, civil society, and democracy.

International Political Science Review, 22, 201-214.

doi:10.1177/0192512101222004

Newton, K., & Norris, P. (2000). Confidence in public institutions:

Faith, culture, or performance? In S. Pharr, & R. Putnam (Eds.), Dis-

258

H. H. KIM

affected democracies: What’s troubling the trilateral countries?

Princeton, NJ: Princeton University Press.

Newton, K., & Zimerli, S. (2011). Three forms of trust and their asso-

ciation. European Political Sc i e n c e Review, 3 , 169-200.

doi:10.1017/S1755773910000330

Paxton, P. (2002). Social capital and democracy: An interdependent

relationship. American Sociological Review, 67, 254-277.

doi:10.2307/3088895

Putnam, R. (1993). Making democracy work: Civic traditions in mod-

ern Italy. Princeton, NJ: Princeton University Press.

Putnam, R. (2000). Bowling alone: The collapse and revival of Ameri-

can community. New York: Simon and Schuster.

Rothstein, B. (2002). Social capital in the social democratic state. In R.

Putnam (Ed.), Democracies in flux. The evolution of social capital in

contemporary society. Oxford: Oxford University Press.

Rothstein, B., & Uslaner, E. (2005). All for all: Equality, corruption,

and social trust. World Politics, 58, 41-72.

doi:10.1353/wp.2006.0022

Rothstein, B., & Stolle, D. (2008). The state and social capital. Com-

parative Politics, 40, 441-459.

doi:10.5129/001041508X12911362383354

Stolle, D. (2001). Clubs and congregations: The benefits of joining an

association. In K. Cook (Ed.), Trust in society. New York: Russell

Sage Foundation.

Tavits, M. (2006). Making democracy work more? Exploring the link-

age between social capital and government performance. Political

Science Quarterly, 59, 211-225.

De Tocqueville, A. (1969 [1865]). Democracy in America. New York:

Harper Perennial.