[ABSTRACT] A Study on the Measurement and Determinant of Social Capital in Local Community

A Study on the Measurement and Determinant of Social Capital in Local Community 

- Focused on the Korean General Social Survey Data -

Submitted by Choung, Yun Sik 

Supervised by Professor Park, In Kwon

Department of Urban Administration

Graduate School Of the University of Seoul

  This study aims to propose a method of measuring social capital and to analyze its determinants using data from the Korean General Social Survey (KGSS). Based on literature review, I define social capital as intangible and potential capital that is accumulating in various social relations and interactions of people and/or organizations. It is composed of trust, respect of institutions and norms, and networks and participation. Each of these components of social capital is a product of localities as well as individuals. Not only individuals' attributes such as experience, education, propensity to trust but only localities' attributes such as urbanity, networking density, and anonymity determine the degree of social capital formation. 

  This study introduced a structural equation model (SEM) to measure social capital, and a hierarchical linear model (HLM) to analyze its determinants. The three conceptual components of social capital are derived using a confirmatory factor analysis of KGSS data for 2004, and their inter-relations are analyzed by a path analysis. Finally, the determinants of social capital are explored by a HLM with explanatory variables at both individual and community levels. 

  The results first show that the three conceptual components of social capital are latent variables, which explains nine measurement factors, at the higher level. Measurement factors are also sub-latent variables that determines twenty six sub-measures in KGSS. 

Secondly, the path analysis shows that there is directional inter-relations between the components of social capital: from networks to institutions and from institutions to trust. It means that the respect of institutions and norms intermediates between trust and networks. 

  Finally, the HLM results shows that three components of social capital are determined by both individuals' and localities' attributes. The degree of trust is observed higher in rural areas than in urban areas, and the degree of networking and participation is higher in urban areas more than rural areas. On the other hand, there is no systematic difference between rural and urban areas for the respect of institutions and norms. 

Keywords : social capital, determinants, Korean General Social Survey, structural equation model, confirmatory factor analysis, hierarchical linear model