Eight Criticisms of Social Capital

Criticism 6: Social capital is difficult enough to define, but it is impossible to measure.

Following from the difficulty of defining and mapping social capital’s network contours is another pressing problem: that of measurement and other empirical indications related to the amount, or the importance of recent changes, in the type and quantity of social capital. Neither social capital nor its effects can be accurately measured in comparable ways, with some very influential research based exclusively on oversimplified measures and misleading comparisons (see Maraffi1994; Morlino 1995). Other attempts require reconceptualising social capital as assets or resources embedded within networks or to the value of a position within such networks, though these surrogates are based upon valuations of wealth, power, and status or access to bridging facilities (see Lin 1999: 37-38) and, as such, are factors that are arbitrarily valued or difficult to determine or grossly oversimplified. The most cited measures and statistics are also those with the most narrow conception of the concept, i.e. those cited by Robert Putnam, in which large United States data sets, in particular the General Social Survey and other longitudinal, national surveys, are analysed in detail to detect changes in a number of key measurements to account for the decay of social capital and a decline in engagement in American society (Putnam 2000). Rather than giving clear indications of such change, the research data is, on closer inspection, inconclusive. In an attempt to reconstruct Putnam’s findings using the same measures as Putnam, though on a smaller scale (trust, voting, church attendance, organisation membership, meeting with neighbours and friends, philanthropic giving) Claude Fischer shows that the relationships between the variables Putnam associates with social capital is not supported by the data:

The controls do little to improve the picture of quite modest associations for items meant to all indicate one thing. This is, of course, only a quick look, not an in-depth analysis. But if the “social capital” concept presumes a tight interconnection among its various elements, that does not immediately appear in the data (Fischer 2005: 165)

Methodological limitations also compound the problem. Foley and Edwards (1999) found that various attempts to quantify attitudes, norms, and social trust at a national level, yield no information as to which groups possess usable social capital. Steven Durlauf, examines the most cited “benchmarked” empirical research, and concludes that “the use of observational data to identify substantive forms of social capital is unlikely to be successful” (Durlauf 2002: F477). Attempting to develop a unit of analysis small enough to capture the effects of social capital in facilitating the dissemination of resources to groups or individuals, has limited data collection to a reliance on methods such as questionnaires, which provide data with which it is difficult to make the distinctions between outcomes of social capital with forms or indicators of its presence. Durlauf considers that social capital can be researched but is cannot permit the type of analysis with the clarity and precision that the advocates in the field claim:

The empirical social capital literature seems to be particularly plagued by vague definition of concepts, poorly measured data, absence of appropriate exchangeability conditions, and lack of information necessary to make identification claims plausible. These problems are especially important for social capital contexts as social capital arguments depend on underlying psychological and sociological relations that are difficult to quantify, let alone measure (Durlauf 2002: F475)

These are not the only likely limitations, as there are also problems associated with capturing the impact of period, cohort and lifecycle factors, identifying “nested” interaction factors and community context variables, in addition to problems of scale when evaluating aggregates. Each of these factors requires a different methodological approach. Economic experiments are one potential area that Durlauf considers might give some limited data, though he concludes that when examining the difficulties of the objective: “the use of observational data to identify substantive forms of social capital is unlikely to be successful” (Durlauf 2002: F477). Examining the literature of the past six years provides no counter-example with which to challenge this conclusion.