On the Empirics of Social Capital

This paper critically examines the way in which empirical evidence is developed in support of a role for social capital in socioeconomic outcomes. Three leading studies of social capital are reviewed and in each case argued to suffer from various identification problems. A general set of conditions under which social capital effects may be identified in linear models is given. Careful attention to these conditions combined with greater openmindedness as to what constitutes appropriate evidence seems the best route to improving empirical studies of social capital.