Social Capital and Its Institutional Contingency

A Study of the United States, China and Taiwan

Nan Lin, Yang-chih Fu, Chih-jou Jay Chen

• Publisher: Routledge
• Year: 2013
• ISBN:
  9781135012328
• Edition: 1
• Topic: Política y relaciones internacionales
• Subtopic: Política

This volume is a collection of original studies based on one of the first research programs on comparative analysis of social capital. Data are drawn from national representative samples of the United States, China and Taiwan. The three societies selected for study allow the examination of how political-economic regimes (command versus market) and cultural factors (family centrality versus diverse social ties) affect the characteristics of social ties and social networks from which resources are accessed and mobilized.

Part I

Measuring Social Capital

2 Contact Status and Finding a Job

Validation and Extension

Nan Lin, Hang Young Lee and Dan Ao

Research on the use of social relations and resources embedded in social relations (social capital) in the labor market has had a long tradition in both economics and sociology (Granovetter 1995; Lin 1999b). A major focus has been the utility of contacts in job searches. The accumulated evidence over a period of four decades has confirmed two general findings. First, the mere use of contacts in the job search does not show any advantage in job attainment (e.g., occupational status or income; see Elliott 2000; Green, Tigges and Diaz 1999; Lin 1999b; S. Smith 2000). Second, among those who use contacts, contact status as a measure of social capital has consistently shown some advantage in obtaining better jobs, after controlling for education and other relevant demographic variables (Bian and Ang 1997; De Graaf and Flap 1988; Ensel 1979; Marsden and Hurlbert 1988; Lin, Ensel and Vaughn 1981; Requena 1991; Volker and Flap 1999; Wegener 1991). This second finding, however, has been challenged for its validity (nonspuriousness). Mouw (2003) suspected that contact status may be spurious in its effect in part because of occupational homophily (similarity of occupations between the job seeker and the contact), considered a social preference for friendship rather than social influence.

The present chapter reports a study clarifying and extending contact status research in several directions. First, it reassesses the validity of the effect of contact status by taking into account the correctly specified occupational homophily. In Mouw's study, the matching of occupations was based on the contact's and the job seeker's current occupations. Theoretically, the matching should be between the contact's occupation and the job seeker's previous occupation. The nonspuriousness of the effect of contact status should be evaluated when the correct occupational homophily is specified. Second, the chapter estimates the relative effects of contact status in the status attainment process for different gender and racial/ethnic groups. A significant body of literature suggests that females and minority group members are disadvantaged relative to males and whites in their social networks and social capital. The present study will systematically examine the effects of contact status for various gender and racial/ethnic groups. Third, the chapter will further explore the nature of homophily; a concept that we argue encompasses both social preference and social influence. An attempt is made to decompose these two components and ascertain the extent to which homophily may in fact reflect social influence, thus part of the contact effect as well.

If homophily, reflecting social influence, also affects status attainment, then the question becomes whether it supplements or substitutes the existing theory, postulating that reaching up to better contacts (with better resources or status) yields better returns. If it substitutes the reaching-up process, then the existing theory requires drastic reformulation. On the other hand, if it supplements the reaching-up process, the theory needs to take into account how help from lateral ties (ties with similar status) adds value in return and whether this additional benefit applies to certain social groups, especially those groups who may not benefit as much from the reaching-up process.

These research questions are examined with data drawn from two national representative sample surveys conducted in 2005 and 2008 in the United States. The surveys were specifically designed to investigate the production of and return to social capital, probably one of the first such national surveys. A portion of the data pertains to the use of contacts and contact statuses in the search for current (or last) jobs. The combination of the two identical data sets provides a sufficiently large sample size for the detailed analyses proposed in the preceding. We first explicate the significance of these issues.

ISSUES AND HYPOTHESES

Nonspuriousness of Contact Status effect

In his analysis of the validity of the effect of contact status, Mouw (2003, p. 869) suspected that effects found for contact status may be spurious because of the presence of a nonrandom variable—social preference in friendship. For the indicator of friendship, he constructed occupational homophily to identify those respondents who had the same occupation as their contacts. When occupational homophily was accounted for (deletions of cases), the data from the 1970 Detroit Area Study showed that the effect of contact's job prestige became insignificant (Models 2 and 3 in table 5, p. 883).

However, occupational homophily (in Model 3) was misspecified. Mouw eliminated the cases where contact's and respondent's current occupations were the same. The social capital theory clearly proposes that the “reaching-up” phenomenon refers to the process from ego's initial position to contact's. A comparison between contact's position and ego's current position may be useful for other purposes, such as for examining the extent to which contacts introduce or refer others to their own occupations (e.g., teachers, telemarketers and so on). However, it cannot reflect social influence (e.g., a respondent approaching the contact for help in the labor market) that must precede the current position of the respondent. For occupational homophily as a possible causal agent to current job status, the correct operationalization should thus be a comparison between the respondent's initial or prior occupation and contact's occupation.

The second operationalization, also employed by Mouw, constructed three dummy variables—contact status higher, the same or lower than the status of respondent's previous job (Model 4, table 5, in Mouw 2003). It showed no added effect when contact status was higher than ego's in predicting occupational prestige of ego's current job.¹ However, limitations and constraints of available data in the Detroit study made the analyses and results tentative at best. The Detroit study's 1970 data had only 219 qualified respondents for general analyses (Models 1, 2 and 4) and 154 for the analysis eliminating the “homophilous” cases (Model 1).

A proper test of the contact status effects should therefore correctly identify the homophilous cases and be based on a sufficiently large sample. We argue that once these correction and sample size issues are taken care of, the analysis will confirm the nonspurious effect of contact status.

H1: Contact status significantly affects attained status when other known factors, including homophily, are accounted for.

Gender and Racial/Ethnic Inequality in Social Capital

Inequality of social capital among social groups has been well documented (Campbell and Rosenfeld 1985; Drake 1965; McPherson and Smith-Lovin 1982; Marsden 1987, 1988; Martineau 1977; Moody 1983; G. Moore 1990; Munch, McPherson and Smith-Lovin 1997; Royster 2003; S. Smith 2005). Women tend to be located in structural positions that are inferior in resources. Gender segregation appears in occupations, jobs and careers (England 1992; Reskin and Bielby 2005; Tomaskovic-Devey 1993). That is, gendered occupations, jobs and careers also result in women being more likely to associate with other women than with men in organizations. For example, women's ties are more likely to be kin and neighbors compared with men's ties (G. Moore 1990; Marsden 1987, 1988). These ties are less heterogeneous and diverse in resources in comparison to nonkin ties, most of which are more work and organizationally related (Marsden 1987, 1988; G. Moore 1990). Likewise, women's social ties tend to be located in peripheral organizations that are smaller and more focused on domestic and community affairs (McPherson and Smith-Lovin 1982).

In formal organizations, men's ties tend to be other men, coworkers and others in a greater variety of occupations (Campbell 1988; Hanson and Pratt 1991; Green, Tigges and Browne 1995; Marx and Leicht 1992; Straits 1998; Beggs and Hurlbert 1997; Reskin 1993; S. Smith 2000). Even when women participate in organizations that include men, they are not well integrated into men's networks (Brass 1985; Roscigno, Garcia and Bobbitt-Zeher 2007). Further, since men hold key positions in most organizations, men's ties tend to be affiliated with core organizations or organizations that have significant economic resources or occupational resources (McPherson and Smith-Lovin 1982; Moody 1983). Even in voluntary organizations men tend to hold key positions. Thus men's ties also reflect their access to key occupants in voluntary organizations.

That women tend to associate with other women who themselves tend to be located in disadvantaged positions suggests a reproduction of disadvantage in resources through women's social capital. The exception to this principle is family ties, which tend to be gender heterogeneous (Marsden 1990; Hanson and Pratt 1991); that is, family members consist of both men and women and thus to some extent may overcome some network disadvantages for women. However, family ties also tend to be homogenous in resources. Thus, depending on the resources embedded in a family, family ties may or may not actually provide better access for women.²

It is not surprising that women have a decided disadvantage in their social networks and embedded resources compared to men. For example, the size of social networks appears to be larger for men than for women (Campbell and Rosenfeld 1985), but more importantly the types of social ties and their resources sharply differ for men and women. These theoretical and empirical developments lead us to propose the following hypot...

Table of contents

• Cover
• Half Title
• Routledge Advances in Sociology
• Title Page
• Copyright Page
• Table of Contents
• List of Figures
• List of Tables
• Acknowledgments
• Availability of Data Sets and Code Books
• Introduction
• Part I Measuring Social Capital
• Part II Endogeneity of Social Capital: Structural and Network Features
• Part III Accessing and Mobilizing Social Capital: Institutional, Networking and Organizational Factors
• Part IV Social Capital and Well-Being
• Appendix I: Social Capital USA 2004 Telephone Interview Questionnaire
• Appendix II: Sample Characters for U.S., Taiwan and China 2004–2005 Social Capital Surveys
• Bibliography
• Contributors
• Index