Social Capital: The Problem of Measurement

by Francis Fukuyama

Since the fall of the Berlin Wall, there has been an extraordinary amount of attention paid to the interrelated issues of social capital, civil society, trust and social norms as central issues for contemporary democracies. The propensity for civil society was said to be an essential condition for the transition to stable democracy in Eastern Europe, and the decline in social capital in the United States is said to be a major problem for American democracy today.

Social capital can be defined simply as the existence of a certain set of informal values or norms shared among members of a group that permit cooperation among them. The sharing of values and norms does not in itself produce social capital because the values may be the wrong ones: The norms that produce social capital must substantively include virtues like truth-telling, the meeting of obligations and reciprocity.

It is clear that the norms that produce social capital are partible, that is, they can be shared among limited groups of people and not with others in the same society. While social capital exists in all societies, it can distributed in very different ways. Families are obviously important sources of social capital everywhere. But family structure differs from one society to another and the strength of family bonds differs, not simply from family ties in other societies, but relative to other types of social ties. In some cases, there appears to be something of an inverse relationship between the bonds of trust and reciprocity within kinship groups and between kin and non-kin; while one is very strong, the other is very weak. What made the Reformation important for Weber was not so much that it encouraged honesty, reciprocity and thrift among individuals, but that these virtues were for the first time widely practiced outside of the family.

Robert Putnam has argued that the quality of governance in the different regions of Italy is correlated with social capital, and that social capital has been in decline in the United States since the 1960s. I do not want to engage in a prolonged discussion of the so-called Putnam debate in this context, except to use it to illustrate some of the difficulties involved in measuring social capital. Putnam has been very ingenious in coming up with a wide variety of statistical measures of social capital, both in Making Democracy Work and in "Bowling Alone." These include information on groups and group membership, such as that coming from the General Social Survey (GSS), survey research on values (such as World Values Survey) concerning perceptions of honesty and trust, and measures of political participation such as voter registration and newspaper readership. Putnam has collected time-series and cross-sectional data on groups from sports clubs and choral societies to trade unions and political parties.

Much of the debate over Putnam's research concerns the empirical validity of his basic finding that American social capital has been declining over the past two generations. Numerous scholars have earlier pointed to contradictory data showing that groups and group membership have actually been increasing over the time period covered by Putnam, or else that the available data simply do not capture the reality of group life in a complex society like that of the United States. By definition, a newly formed group will be less institutionalized than a more established one, and hence less likely to keep good statistical information on itself, or to be observed by third parties collecting statistics on group memberships. There are a large number of informal social networks for which no data exist at all. On the other hand, Putnam excludes from his measures social capital in families, for which abundant evidence of decline exist.

There are three further measurement problems. First, social capital has an important qualitative dimension. While a bowling league or a garden club might be, as Tocqueville suggests, schools for cooperation and public-spiritedness, they are obviously very different institutions from the US Marine Corps or the Mormon Church, in terms of the kinds of collective action they foster. A full account of social capital needs to take account of the degree of cohesive action of which a group is capable.

The second problem has to do with what one might call the positive externalities of group membership. While all groups require some degree of social capital to operate, some build bonds of trust and hence social capital outside of their own memberships. As Weber indicated, Puritanism mandated honesty not simply towards other members of one's religious community, but towards all human beings. On the other hand, norms of reciprocity can be shared among only a small subset of a group's members. While what Putnam terms a "membership group" like the American Association of Retired People has a huge number of members, there is no reason to think that any two given members will trust one another or achieve coordinated action just because they have paid their yearly dues to the same organization.

The final problem concerns negative externalities. Some groups actively promote intolerance, hatred and even violence toward non-members. While the Klu Klux Klan, Nation of Islam, and Michigan Militia possess social capital, a society made up of such groups would not be particularly appealing, and might even cease to be a democracy. Not only do such groups have problems cooperating with each other, the bonds of community uniting them are likely to make them less adaptive by sealing them off from influences in the surrounding environment.

It should be clear that coming up with a believable number expressing the stock of social capital for a large and complex society like the United States is next to impossible. We have empirical data, of varying reliability, on only a certain subset of the groups that actually exist, and no consensus means of judging their qualitative differences. It is clear that much more work needs to be done on the measurement of social capital, which is the starting point for understanding how to increase our stock of it.

Further Reading

Francis Fukuyama. 1995. Trust, the Social Virtues and Creation of Prosperity. The Free Press Paperbacks, New York.

---