Albeit vague, the notion of “social capital” has a nice ring to it, evoking neighborhoods where people don’t have to lock their front doors, care about each other, and participate in a vibrant community life. As de Tocqueville put it,

Americans make associations to give entertainments, to found seminaries, to build inns, to construct churches, to diffuse books, to send missionaries to the antipodes; in this manner they found hospitals, prisons, and schools. If it is proposed to inculcate some truth or to foster some feeling by the encouragement of a great example, they form a society.

Understood as “the networks, norms, and social trust that facilitate coordination and cooperation for mutual benefit”—and measured through associational activity or self-reported levels of trust—social capital is intuitively seen as a good thing—reducing transaction costs and making it easier to solve collective action problems.

But is social capital always desirable? Closer networks and higher trust levels among criminals facilitate the operation of criminal groups, for example. And, in contrast with Tocqueville’s idea of strong, tightly knit communities as the basis for a limited government, higher levels of social capital can go hand in hand with a lack of political competition and weak constraints on the government.

Higher levels of social capital sometimes facilitate the rise of authoritarian movements. Satyanath, Voigtlaender, and Voth show how dense networks of civic associations were related to the rise of the Nazi Party in Germany. In the face of weak and often dysfunctional institutions of the Weimar Republic, personal, face-to-face networks fostered the spread of the radical new movement. Incidentally, in Prussia, which had stronger democratic institutions than the rest of Germany, the link between the levels of social capital and the spread of membership in Nazi Party was much weaker than elsewhere—suggesting that it is the quality of formal institutions determines whether social capital plays a beneficial role or not.

Social capital can be captured by a predatory political elite. In a new paper, Acemoglu, Reed, and Robinson look at the effects of political competition on governance in Sierra Leone. Unsurprisingly, areas with a smaller number of ruling families display worse development outcomes because the political elites are less constrained by competition. At the same time, these areas display higher levels of social capital:

[C]hiefs that face fewer constraints build social capital as a way to control and monitor society. This mechanism may also induce people to invest in patron-client relations with powerful chiefs, thus giving them a vested interest in the institution. Hence, if in surveys people say that they respect the authority of elders and those in power, this is not a reaction of the fact that chiefs are effective at delivering public goods and services or represent the interests of their villagers. Rather, rural people appear to be locked into relationships of dependence with traditional elites.

In such situations, political elites leverage the existing networks to exercise control over their subjects. In turn, the citizens are incentivized to play along, as the lack of well-functioning political competition and constraints on the rulers makes it worthwhile to invest into the relations of patronage with people up in the hierarchy, leading to further entrenchment of the ruling elite.

Hence, whether cooperation induced by more social capital is desirable or not ultimately depends on the institutional context. Now, add another element to the discussion: homogeneity. It is commonly argued that homogeneity—ethnic, cultural, religious—facilitates the creation of social capital, whereas heterogeneity reduces trust levels. This is the basis of Paul Collier’s call for tighter immigration controls—with more immigrants, it becomes more difficult to sustain the trust levels existing in Western societies.

That argument presupposes that all of the social capital in West is of the benign variety. But what if social capital helps sustain large, intrusive governments? For all the admiration one can have for the high-social-capital societies of Scandinavia, it is plausible that these countries would be better off without sizable welfare states that would not have been able to operate without the high degree of civic-mindedness among the population.

While Western societies look nothing like Sierra Leone or Weimar Germany, it is at the very least conceivable that some parts of our existing stock of social capital foster undesirable forms of cooperation—political rent seeking or wasteful government redistribution. So maybe it is time to start trusting your neighbors a little bit less.