Essay 1: This paper studies tipping behaviour in the residential mobility of the native population inSweden between 1990 and 2007. Using regression discontinuity methods, we find that thegrowth in native population in a neighbourhood discontinuously drops once aneighbourhood’s immigrant share exceeds the identified tipping point. In the 1990s the dropcan be attributed both to increased out-migration of natives (native flight) and to thedecreased in-migration of natives (native avoidance) while native flight appears to be drivingthe segregation pattern between the years 2000 and 2007. Further, we find native migrationfrom neighbourhoods that have tipped is selective, in the sense that natives with a high levelof educational attainment are the most likely to move from such neighbourhood. We concludethat the native residential mobility has contributed to increased ethnic segregation but it alsoappears to have increased socio-economic segregation in Sweden between 1990 and 2007. Essay 2: This paper focuses on second-generation immigrants and analyses the short- and long-termeffects of immigrant and ethnic group concentration in childhood neighbourhood on earnings,unemployment, reliance on income support and educational attainment. The results show thata high immigrant concentration in a childhood neighbourhood is negatively associated witheconomic outcomes of both second-generation immigrants and natives. Ethnic groupconcentration seems to work in the opposite direction, improving economic outcomes forsecond-generation immigrants. Furthermore, the results highlight the importance of includingtime dynamics in any analysis of the effect of childhood neighbourhood ethnic compositionon economic outcomes.

This thesis consists of four empirical essays on the topics of ethnic segregation, gender economics, and self-employment.    Essay I investigates how the residential mobility of Sweden's native population contributes to ethnic segregation, by applying regression discontinuity methods. The results show that the growth in the native population in a neighbourhood discontinuously drops as the share of non-European immigrants exceeds the tipping point. Tipping is driven by the departure of natives and their avoidance of tipped neighbourhoods. Tipping behaviour is selective in the sense that highly educated and high earning natives are more likely to leave neighbourhoods that have tipped.    Essay II studies the relationship between the childhood neighbourhood's ethnic composition and economic outcomes in adulthood for second-generation immigrant sand natives. The results reveal that a high concentration of immigrants in aneighbourhood is associated with a lower probability of second-generation immigrants continuing to higher education. Natives' earnings and educational attainment are negatively correlated with, and the probability of social assistance and unemployment are positively associated with a high immigrant concentration. Among non-Nordic second-generation immigrants, reliance on social assistance and unemployment are negatively correlated with the share of co-ethnics and positively associated with the proportion of other ethnic groups.    Essay III explores the role of social norms and attitudes about gender for labour market outcomes of immigrant men and women in Sweden. The results show that immigrants originating from countries with large gender disparities in labour force participation also have large gender gaps in labour force participation within their immigrant group on the Swedish labour market. In contrast, source country gender differences in earnings are not correlated with gender gaps in earnings within immigrant groups in Sweden. In addition, gender gaps in labour force participation among immigrants assimilate towards the corresponding gap among natives as time inSweden increases.    Essay IV empirically tests the Jack-of-all-trades theory, which states that individuals who are more balanced in their abilities are more suitable for self-employment. Using Swedish Military Enlistment data, a measure of balance in endowed abilities is constructed and this balance measure is, in relation to previous research, less likely tobe endogenous. The results support the Jack-of-all-trades theory, in the sense that propensity for being or becoming self-employed is greater for individuals with abalanced set of abilities. In addition, earnings from self-employment tend to be higher among individuals with a balanced set of skills.