Understanding economic and political integration has long been a central concern for economists. An important missing ingredient in the existing literature is the analysis of endogenously determined social identity. By “social identity” I refer to the fact that individuals often care deeply about the groups to which they belong. By “endogenously determined” I refer to the fact that individuals do not automatically identify with every group they belong to: whether or not an individual identifies with a given group depends on the characteristics of this group as well as on how close to this group the individual perceives herself. Empirical results obtained over the past decade allow us to integrate identity concerns into standard economic models. I propose to develop and test a theory of integration that does just that.Consider two states that may either be independent countries or form a union. The stability and desirability of unification may sometimes depend on the extent to which citizens identify with the union or with their states. But the profile of identities itself depends on the political-economic outcome under unification. The first step in developing the theory is to translate the evidence concerning behavior in groups into a concise statement of what it means to “identify” with a particular group and what factors shape identification decisions. The theory will then study the equilibrium outcomes of a political economy model of integration, where actions and identities are endogenously determined.The second part of the project will empirically examine the relation between social identities, individual characteristics, and European integration. To appropriately measure identification, I propose to employ experimental methods based on revealed preference conducted with a large and diverse sample of European citizens. This will be complemented by historical multi-country survey data on self-reported identity, political attitudes and behavior.