|Institution:||University of Helsinki|
|Keywords:||Kansantaloustiede: Yleinen linja|
|Full text PDF:||http://hdl.handle.net/10138/153568|
Starting from the beginning of 1990s, Finland has experienced rapid growth rates in population fractions of non-native residents, with almost half of all immigrants eventually settling in Helsinki metropolitan area. This thesis studies immigrant settlement patterns in the capital region, with the purpose of reliably quantifying and documenting dynamics of ethnic residential segregation. Formation of “ethnic enclaves” is a widely-debated issue which might have ambiguous impact on immigrants’ economic and social outcomes. Robust inference on segregation dynamics is a prerequisite for furthering understanding of the issue. However, qualitative inference on segregation dynamics is often hindered by the fact that different spatial unit sizes and/or immigrant population fractions generate random segregation of varying magnitudes. This thesis overcomes the problem by employing index of systematic segregation which is expressed as a fraction of maximum excess dissimilarity (net of random) that could possibly occur. Index of dissimilarity (Duncan index) is chosen as a baseline measure for calculating systematic index, ensuring comparability of my research to the existing body of knowledge on the phenomenon. While index of dissimilarity reports only marginal growth in segregation, systematic measure reports almost two-fold increase in ethnic residential segregation. Employing systematic index allows reliable comparison of segregation across various immigrant groups and localities. Thus, I find that immigrants from Balkan and African countries are more segregated than other foreign-born individuals. Comparison of systematic indices across largest Finnish cities reveals that Turku has been substantially more segregated than its counterparts starting from year 1995. Eventually, I isolate the measure of segregation along ethnic lines from sorting along other dimensions, such as income. I find no evidence that increase in residential segregation is driven by widening income gap between native and immigrant population. Finally, all the findings are presented compactly in two web applications, allowing flexible controls over the choice of locality of interest, immigrant group, statistics and map-types.