Year of Publication


Document Type



Arts and Sciences


Political Science

First Advisor

Matthew Gabel

Second Advisor

Mark Peffley


While the virtues of social capital in democracies are widely recognized, previous studies have repeatedly shown that social capital is in short supply in heterogeneous communities with ethnic minorities. Against the view that levels of social capital are culturally predetermined, I argue that it is possible to generate social capital by carefully formulating political institutions. Drawing from theories of institutional management of ethnic conflict and theories of institutional learning, I construct an integrated theory of social capital which hypothesizes that citizens learn to trust one another based on their experiences with political institutions during an extended period of democratic rule. To test this integrated model of social capital, I use a probit analysis to examine how democratic longevity in different institutional settings (e.g., majoritarian vs. consensus) influences social capital. To overcome the endogeneity problem that exists between social capital and democratic longevity, I adopt an instrumental variables approach, drawing on theories in international relations. My analysis of World Values Survey data yields three main conclusions concerning the institutional arrangements that foster social capital. First, I find that democratic longevity fosters higher levels of trust in countries with consensus institutions containing powersharing arrangements through cabinets, executive-legislative balances, party systems, and electoral systemspresumably because cooperation among different groups enhances social capital. Second, a longer period of democratic rule in highly federal institutions undermines trust, as the devolution of powers through territorial units is thought to fragment the political system and society. Finally, consistent with the theoretical expectations, I find that these two conclusions hold only among ethnic minorities. Among ethnic majorities, the effect of democratic longevity disappears once we purge the endogenous component (i.e., the effect of social capital on democratic longevity), using an instrumental variables approach. Case studies of the Baltic States, the Canadian province of Quebec, and Malaysia corroborated the findings from the statistical analyses. By uncovering a mechanism through which social capital can be generated in multiethnic states, this study makes an important contribution to the literature.