Health Behaviors and Risk Equilibria

The paper models the individual's decision to adopt a health risky behavior, smoking or binge drinking, as the result of maximizing expected utility, in which the presence of social capital reduces the risky behavior. The model extends Folland (2006) by incorporating health production. The model is tested against data from the National Longitudinal Survey of Youth 79 as merged with the DDB marketing database and covering 1976-1998. The tests include prediction equations for the presence of smoking or binge drinking as well as predictions of the decision to quit. The empirical work overall focuses on the effort to identify the social capital effect and support the claim that the relation of social capital to health is causal.

Psychosocial Resources and Health Inequalities in France

This study addresses the issue of psychosocial determinants of health, in addition to the material and biological ones, in order to explain social inequalities in health in France. Whereas the impact of psychosocial resources on health is most often described from a "contextual" perspective, considering that a higher level of resources in the area improves health on average for individuals living in the area, psychosocial resources are here assessed at the individual level to take into account the actual or perceived access of individuals to the level of resources.

We use a general population survey, representative of employed individuals aged 16 to 64 living in France in 2004, to study the association between health status and the subjective perception of social capital, social support, sense of control and self-esteem, controlling for standard socio-demographic factors (occupation, income, education, age and gender). Several health outcomes are considered: self-assessed health, long term activity limitations, main chronic conditions, obesity, tobacco consumption, and chronic excessive alcohol consumption.

We find empirical support for the link between the subjective perception of psychosocial resources and health or health-related behaviours. More specifically health status is positively associated with access to social capital, emotional support and sense of control at work. Since access to psychosocial resources is not equally distributed in the population, these findings suggest that psychosocial factors can partly explain social inequalities in health in France.

The Role of Community Social Capital in Reducing the Prevalence of Serious Mental Illness

We show that lagged community social capital is strongly and inversely related to the prevalence of serious mental illness and also exhibits strong diminishing returns. We estimate a mental health production function, which controls for sex, age, race/ethnicity, marital status, education, and individual social capital, as well as unobserved area-level heterogeneity using three year (1999-2001) of U.S. data on 48,222 adults. The presence of mental health is measured using a valid and reliable indicator of serious mental illness: the Kessler K6. Our measure of community social capital is a new validated ecological measure of community social capital: the Petris Social Capital Index (PSCI).

Social Capital and Health in Eight CIS Countries

This paper starts from an empirical assessment of different dimensions of social capital in the transition countries of Central and Eastern Europe (CEE) and the Commonwealth of Independent States (CIS). On the whole, the level of social capital is lower in CEE-CIS countries compared to other countries in Europe and beyond, and some social capital indicators have deteriorated in recent years. Given the alleged importance of social capital for health we use a unique data source to carefully investigate the impact of social capital on individual self-reported health for a sample of eight countries from the Commonwealth of Independent States (Armenia, Belarus, Georgia, Kazakhstan, Kyrgyzstan, Moldova, Russia, Ukraine). We rely on three indicators for social capital - individual degree of trust, participation in local organisations, social support - and employ alternative procedures to consistently estimate the impact of social capital on health. Memberships in organisations, social support or trustful behaviour are choice variables implying that social capital indicators are by definition endogenously determined and depend on individual specificities. We tackle this endogeneity problem using instrumental variable estimates. The wealth of the data set allows us to distinguish the social capital impact from other community effects (such as health care supply) that are simultaneously correlated with health and measures of social capital. Our results show that, in the pooled sample of all eight countries, the individual degree of trust and the indicator for social support positively affect self-reported health, this being true with classical profit estimates as well as when we address the endogeneity issue through IV methods. We finally offer an interpretation of the differential effect that social capital has across the different countries.

Social Capital and Health in England

Objectives

A large body of evidence shows important health variations in England, both in relation to individual socio-economic characteristics, particularly social position and education, and in relation to area-level characteristics. Recently, reports have also shown a relationship between health and aspects of social cohesion and social capital. This study aimed to use the most detailed source of data on social capital and health in England to investigate the role of aspects of social capital acting at the individual and the community levels on various dimensions of health, and possible interactions between these and other determinants of health. The expectation is that social capital variables reflect the strength of social networks and circulation of information, as well as identification with models of good citizenship that may increase the adoption of healthy lifestyles and behaviours. Trust is generally seen as a factor that reduces transaction costs, which is also likely to have an indirect effect on health.

Data Sources

The main source of data used in the study is the Health Survey for England (HSE) a cross-sectional household-based survey of the health status, risk factors and health-related behaviour of the population in England, including information on aspects of social capital at the individual level. Information on area characteristics was linked at the Health Authority and at the Ward levels.

Methods

Secondary analysis of data from five HSE waves, 2000 to 2004, including over 70 thousand adults. The following outcome measure were used: (a) self-reported health (five point scale); (b) limiting longstanding illness (presence or absence of a limiting condition in at least one of 13 longstanding disease groups); (c) psycho-social health (General Health Questionnaire). Individual level covariates included equivalised household income, education, age, gender, ethnicity, participation in membership organisations, trust, social support. Community level variables included area-level deprivation, distribution of deprivation, and ethnic residenial segregation from the decennial census (2001); indicators of community social capital (CSC) from the Annual Business Inquiry (ABI). Both fixed effects and random effects (multilevel) models were used, broadly drawing on a Grossman production function approach. The models allowed for possibly decreasing marginal returns to individual covariates. The possibility of a u-shaped relationship between CSC and health outcomes was tested using a quadratic term. The possibility of interactions between individual participation in membership organisations and CSC measures was also tested for. All analyses were conducted using Stata 9.0.

Principal Findings

All of the individual social capital variables are strongly correlated with health outcomes. The correlation is slightly stronger for generalised trust. Of the CSC variables, mean participation in membership organisations was most consistently associated with good health outcomes. A positive and significant interaction term between household income and area-level participation in membership organisations (p=0.045 for self-reported health; p<0.001 for limiting long-standing illness and psycho-social health) appears to indicate that the positive effect of CSC on health is greater when income is lower, with a clear gradient from the lower to the upper income quintiles. The possibility of a u-shaped relationship between CSC and health outcomes did not appear to be supported. Similarly, a possible interaction between CSC and ISC (individual participation in membership organisations) appeared to be non significant. The Health Authority (average population of 500,000) proved an inadequate geographical level for the analysis of community-level effects in multilevel models, showing poor intra-class correlation.

Social Capital and Health in Indonesia

This paper empirically examines the role of community social capital in the individual's health production function. We focus on health measures relating to physical as well as mental health. In addition to exploring the relationship between social capital and health, we test for interrelationships between social and human capital in the production of health. Data come from more than 10,000 adults surveyed in the Indonesian Family Life Surveys of 1993 and 1997. We identify a robust positive empirical association between community-level social capital and good health. We find weak evidence for an interrelationship between human and social capital and mental health.