This paper explores whether the concept of social capital as popularized by Robert Putnam is a good social science concept. Taking Gerring’s work on concept evaluation as the starting point, the paper first presents a set of criteria for conceptual ‘goodness’ and discusses how social capital performs on these criteria. It is argued that social capital eventually may be a good concept if it can be shown empirically to be a unidimensional concept. An empirical section therefore explores the validity of the unidimensionality assumption and rejects it in four separate tests at both the individual and aggregate level. We conclude that even if social capital has been a remarkably productive idea, it is not a good concept as most popular conceptualizations define social capital as several distinct phenomena or as phenomena that already have been conceptualized under other labels.
Social Indicators Research, 2013, Vol 114, Issue 3, p. 1225-1242