An analysis of the identity formation of second generation immigrants using city-level data

This paper examines the process of identity formation from an economic perspective using data on second generation immigrants in Western European countries. Relying on the identity formation mechanism suggested by Bisin, Patacchini, Verdier and Zenou (‘Bend it like Beckham’: Identity, Socialization, and Assimilation, 2006), identity formation is modelled using an intergenerational model, where the identity of children is determined through three processes: parental identity investments, social interaction and an element of identity choice. While building on the theoretical part of their paper, a different empirical approach is taken here. We use data on the children’s (rather than the parents’) generation to examine the model’s main outcome of interest, the children’s identity. Language in which parents chose to raise their children is used as a measure of investment, as there is a large psychological literature on the link between language and identity and this may be less affected by non-linearity and measurement issues than homogamy (the measure they used). A cross-country comparative perspective is used to assess the robustness of this identity formation mechanism. City level variation is used to examine the role of local responses to immigration. We hope that this contributes to the literature on identity (so far limited in economics) by examining a formal theoretical model of identity formation, providing insights which may later be used to overcome endogeneity problems in regressions attempting to measure the effect of identity on economic outcomes such as education or employment.