Abstract
Social capital is a concept that attempts to describe the quantity and quality of social interactions in a community. This study explores the relationship between individual measures of social capital and alternative measures of health status within the context of a large national survey of population health. Using data for 13 753 adult participants in the 2003 Health Survey for England, linear regression with weighted least‐squares estimation and Tobit regression with upper censoring were used to model the relationship between individual measures of social capital and EQ‐5D utility scores. In addition, logistic regression was used to model the relationship between individual measures of social capital and a dichotomous self‐reported health status variable. The study demonstrated that low stocks of social capital across the domains of trust and reciprocity, perceived social support and civic participation are significantly associated with poor measures of health status. The implications for health economists and, potentially, for policymakers are discussed. Copyright © 2007 John Wiley & Sons, Ltd.