This thesis focuses on three different empirical questions related to the economic incentives to the formation of identity. The first chapter examines the dynamics of occupational segregation of self-identified homosexual workers, specifically accounting for the fact that part of their observed distribution may come from selective disclosure of sexual identity. We present a simple labour supply model where individuals choose both an occupation and their revealed sexual identity. Using confidential data from the UK, we show that selfreported homosexuals are concentrated in opposite gender typical occupations. As tolerance increases, a greater fraction of homosexuals reports their homosexual identity, particularly in marginal homosexual occupations, hence occupational segregation falls. The finding suggests that part of the observed segregation of homosexuals in opposite sex occupations is due to selective disclosure rather tastes or comparative advantage. The second chapter uses an original dataset covering the universe of local elections in England spanning over 40 years to investigate whether the electoral success of women and ethnic minorities leads to increases in these groups' representation as political candidates in subsequent elections. Using a regression discontinuity approach, we find that both groups enjoy a personal incumbency advantage. One direct consequence is an increase in the fraction of women and ethnic minority candidates contesting a seat previously held by someone from the same group. In the case of women, this increase is also driven by an inflow of new women candidates. The third chapter focuses on the impact of television on religious identity. We use detailed survey data on individuals' self-reported religious sentiment, behaviours and attitudes from Indonesia. We use the variation in signal reception due to geographic topography at the sub-district level to estimate the causal effect of media exposure. Individuals exposed to a higher number of television channels are less likely to report being religious and following religious practices. Furthermore, they also display lower interfaith hostility. At the village level, higher exposure to television increases the supply of religiously forbidden activities. However, higher exposure to television seems to have no effect on political preferences.