Refugee Immigrants in Sweden - educational investments and labour market integration

University dissertation from Department of Economics

Abstract: This thesis analyses educational investments and labour market attachment for refugee immigrants who received a permanent visa to Sweden during 1987-1991. It consists of five self-contained chapters. Chapter 2 presents a brief overview of the historical process behind immigration to Sweden from the post-war period until today, give an orientation on the development of Swedish immigration policy during this period, and analyses the composition of the immigrant population in a number of different dimensions. Chapter 3 provides a descriptive picture of the refugee immigrant cohorts included in our database, FLYDATA, and introduce areas where in-depth research on these cohorts will be undertaken. Chapter 4, the first of the empirical studies, analyses the determinants of investments in formal education for these immigrants. The study is conducted on the immigrants' first five years in Sweden. Two different categories of Swedish formal education, secondary and primary education, are considered. The number of semesters spent in each education category are analysed by count data models. According to our results there are four factors that are more important than others in determining the amount invested in the two categories of Swedish formal education. These are age at migration, the level of pre-immigration education, the year during which the immigrant received his/her residence permit, and to some extent also the country of origin. Chapter 5 analyses the determinants of the duration until the first transition into an investment in post-migration education. Research into immigrant labour market performance has shown that the degree of economic adjustment depends in part on the extent of investments in destination specific human capital by the immigrants. A neglected issue in the literature is the timing of these investments. The longer it takes until the immigrant decides to invest in destination specific human capital, the longer the labour market integration process is delayed. To analyse this issue we have used a non-proportional hazard rate model with a discrete non-parametric specification of the unobservable distribution. We find, among other things, that pre-immigration characteristics, for instance the level of foreign education, the probability of return migration, measured as the type of visa category, and the time of arrival to Sweden all are important determinants of the transition into education. Further, many of the parameter estimates are non-proportional, indicating different integration processes, for instance as a consequence of differences between individuals in their ability to find information on the returns to education in the new country, in the short and in the long run. Finally Chapter 6, analyses the labour market status of the FLYDATA immigrants in 1995. The analysis focus on those who were 20 to 60 years old in 1995. The study in Chapter 6 presents an empirical analysis, using Logit regressions, of the labour market status in 1995, in terms of being employed vs. unemployed and being employed with high annual income vs. low annual income. We focus on differences between different nationalities (rather than a comparison with native Swedes) and in a decomposition analysis we investigate to what extent the differences are due to differences in observed characteristics between the immigrant groups. The results indicate that there is a strong polarisation within the refugee population as regards the success on the Swedish labour market. An alarmingly poor attachment to the Swedish labour market is found for some of the immigrant nationalities. We also find that a larger amount of formal schooling, as measured both by the level of pre-immigration education and the number of semesters of Swedish formal education, often has no statistically significant positive effects, and sometimes even a negative effect, on the probability of being employed. Further, we decompose the time in Sweden into the time spent employed and not employed. We find that these variables have large effects, for both sexes, on the probability of being employed and having a good attachment to the labour market in 1995 for almost all nationalities. Hence, the results stress the importance of having been able to be part of the labour force in Sweden rather than to have invested in a Swedish education.

  This dissertation MIGHT be available in PDF-format. Check this page to see if it is available for download.