Is social capital good for your health? A European perspective

Presenter: Lorenzo Rocco, University of Padova

Abstract

Using data from the European Social Survey for 14 European countries, supplemented by regional level data coming from the EUROSTAT REGIO database, we study whether individual and/or community level social capital (SC) positively affect individual health. Intuitively SC is beneficial to health because it favours 1) cooperation among individuals in terms of safety nets and reciprocal support in case of illness and 2) social interaction and circulation of health-relevant information. Both health and SC measures are affected by error-in-variables and the causal relationship between SC and health is likely to be circular. We address these empirical issues by setting a structural model and instrumenting SC indicators. Reported individual SC is assumed to be a function of the true (and unobserved) individual SC and of reported community SC up to a zero-mean error. Moreover reported community SC is defined to be equal to the true (and unobserved) community SC plus a zero-mean error. By substituting these relationships into the main equation defined in terms of true SC indicators, we obtain an estimable equation where reported individual health depends on reported individual SC, reported community SC and the interaction between the two (plus other controls). Given measurement errors and reverse causality, all SC indicators are endogenous: an IV estimator is therefore adopted and the chosen instruments are both individual and local characteristics likely to affect SC stocks without directly impacting on health. The large variance of the overall error term due to measurement errors requires the adoption of strong instruments in order to get reliable estimates in finite sample. Our dataset is rich enough in information to allow us to find a strong causal relationship going from social capital to individual health. Also we obtain that community social capital does not affect health when individual social capital is controlled for, thus contributing to the long lasting debate on the relative merits of individual and community SC.