Structuring Social Capital: Social, Cultural and Institutional Dimensions
Total Page:16
File Type:pdf, Size:1020Kb
Structuring Social Capital: Social, Cultural and Institutional Dimensions Nan Lin Yang-chih Fu Chih-jou Chen Academia Sinica This is a draft; do not quote without permission 5/ 24/2008 Paper presented at the International Conference on Social Capital, May 28, 2008 at the Academia Sinica, Taiwan. Data used in this paper were drawn from the thematic research project “Social Capital: Its Origins and Consequences", sponsored by Academia Sinica, Taiwan, through its Research Center for Humanities and Social Sciences, and the Institute of Sociology. The principal investigator of the project is Nan Lin. We acknowledge the assistance of Siyin Lee in the data analysis. Lin, Fu and Chen 1 Abstract Using the attribute data (gender, strength of ties and relationships) in the position generator from three societies (US, China and Taiwan) we examine how the societies are differentiated by how these attributes differentially describe access to social capital (embedded resources). We consider possible differential patterns relative to social (friendships), cultural (nuclear, extended and pseudo-kinships), and institutional (school, work, clan ties) dimensions. Results show that different patterns can be discerned between the US and China samples. For the US respondents, friends emerge as an important source in accessing better social capital. For the Chinese respondents, kin ties and institutional ties are more important. Taiwan seems to fall in-between with the continued but weakened significance of kin ties, the cont inued significance of institutional ties, and the increasing significance of friends. We suggest that these patterns are more consistent with an industrialization explanation than a cultural or political-economic regime explanation. The data also shed light on the nature of guanxi in Chinese societies and on the prevalence of the social principle (homophily) transcending developmental or economic processes. Lin, Fu and Chen 2 Structuring Social Capital: Social, Cultural and Institutional Dimensions For the past two decades, the position generator has provided a standard measurement instrument for social capital (Lin & Dumin, 1986; Lin et al., 2001; see the entire volume of Social Capital: An International Research Program, 2008). The principal idea is to ascertain an individual’s access to various valued resources through social ties. Typically, occupations are assumed to carry significant resources (capital) in most societies, so the instrument may consist of a list of sampled occupations. The instrument, asking an individual to indicate whether she/he knows of others having sampled occupations, may yield indications of the richness (access to high-status occupation) as well as diversity (number of sampled occupations accessed) of social capital for the person – access to capital through social ties. Thus the position generator is a theory-based instrument for social capital. As an estimation of social capital, the instrument has yielded reliable and valid empirical evidence of returns on social capital in life chances (e.g., market competition and social solidarity (Lin, 2005)). Aggregation of such indexes and returns afford comparisons across social groups (gender, race\ethnicity, etc.) as well as across societies. The instrument can also be adapted for other types of valued resources (van der Gaag et al., 2005; van der Gaag et al., 2008) and for groups/organizations (Lin, 2006; Son & Lin, 2008). One additional feature of the instrument is that it may also incorporate measurement of particular characteristics of the occupant of each accessed occupation (e.g., gender and race\ethnicity) as well as the relations (e.g., strength of the tie and the role relationship) between the occupant and the focal individual. These attributes can help estimate sources of inequality in social capital. For example, one may pose the question: whether certain types of social ties (e.g., men or whites) are more likely to accessing rich or more diverse resources? This question is posed based on two Lin, Fu and Chen 3 considerations: (1) the homophily principle favors individuals of certain characteristics in accessing social ties with similar characteristics and (2) individuals and groups of certain characteristics tend to occupy higher-status or accessing more diverse occupations in most societies. With information about attributes of tie characteristics (e.g., whether the occupant accessed is man or woman, white, black or Latino), it becomes possible to assess whether access through men rather than women and white ties rather than minority members, for e xample, tends to reach rich and/or diverse positions – better social capital. Multiple attributes incorporated in the position generator afford simultaneous examination of multiple characteristics of occupants and social relations as they are associated with rich or more diverse social capital. It then becomes possible to estimate the relative significance of attributes (e.g., is having a male tie more important than having a white tie to gain better social capital?). The objective of this paper is to ascertain the relative significance of attributes for accessing richer and/or more diverse occupations as sampled in the position generator instruments in a study of three societies – US, Taiwan and China. Since the data come from three national surveys, we can further seek to understand if different structures of social capital exist across societies. In the remainder of the paper, we will introduce the position generator instrument that secures attributes in regard to certain characteristics of the accessed occupants. We then introduce some conceptual considerations about how the associations of these attributes and the accessed social capital may be expected to vary across the three societies. A brief introduction to the surveys and the da ta will be followed by analyses. The paper will conclude with some reflections of the conceptual issues. THE INSTRUMENT - THE POSITION GENERATOR AND THE TIE ATTRIBUTES – Lin, Fu and Chen 4 A typical position generator poses a question, “I am going to ask some general questions about jobs some people you know may have. These people include your relatives, friends and acquaintances (acquaintances are people who people who know each other by face or name). If there are several people you know who have that kind of job, please tell me the one that occurs to you first.” Then, a list of sampled occupations is provided and the respondent is asked to indicate if she/he knows someone in that occupation (“Is there anyone you know who is a nurse?”). The sampling of the occupations should be from a list of occupations listed in a rank order of “values” – for example, status or prestige scale scores. They could be either sampled from equal-intervals along the scale, or some random sampling process. Not all initially sampled occupations should be used, as some of them may be obscured and unfamiliar to most people. Preferences should be given to occupations at or close to sampled occupations in rank which have substantive prevalence in the labor force (e.g., the size of occupants) thus are more familiar to the population. Typically, the sample should consist of at least 10 or 15 occupations to provide reasonable variance in the data. A further refinement of the sample may be to add some occupations oriented toward certain social groups in the population (e.g., female-dominant occupations). In our surveys, we sampled 22 occupations. An important feature of the position generator, as in contrast to the name position, is that it does not evoke any particular topics or issues on which the ties are based, thus avoiding the problem of a lack of knowledge of the population of topics or issues from which a probability sampling may be drawn. Also, the position generator differs from the name generator in that it focuses on positions in a social hierarchy rather than individual actors in the hierarchy. Thus the position generator estimates the extent to which an actor can access different positions in the hierarchy. In so far as such positions are reasonable estimates of valued resources in the hierarchy, the Lin, Fu and Chen 5 position generator is a reasonable instrument to estimate accessed valued resources – social capital. Thus the primary aim of the instrument is to estimate access to capital in the hierarchy and its validity depends on (1) the reasonable representation of capital captured by the positions (Do the positions reflect differential valued resources?) , and (2) a reasonable representation of the positions sampled (Do the sample of positions reflect the range of such positions?). In contrast, the name generator elicits other actors, regardless of their positions in the social hierarchy. It is well suited to study the ego-networks linking to other actors. Estimation of their relative vertical position and range is helpful to describe the clustering of hierarchical positions among stronger ties for ego, but it does not inform us as to whether ego can indeed access other positions in the hierarchy. Still another feature of the position generator is the less likelihood of eliciting strong ties, as compared to the name generator, with its limited number of names elicited. However, it does not completely overcome the tendency to elicit stronger ties. An actor may know multiple occupants at a sampled position. Typically, as in our surveys, we elicit the first occupant that comes to mind. This strategy has the advantage of efficiency (quick response) in a survey, but it also tends to draw stronger rather than weaker ties , as we assume stronger ties come first to mind. 1 Thus, the position generator does not eliminate the tendency to draw out occupants of stronger ties for a sampled position. Inclusion of a reasonable size of sampled occupations may increase the likelihood of the inclusion of some weaker ties, but it does not eliminate the general tendency of stronger-tie inclusions. Nevertheless, we should remind ourselves that attributes examined for the occupants also may be biased toward descriptions of stronger ties.