Settlement municipality and integration among adult refugees in Norway
Refugees who are granted a residence permit in Norway are usually settled by public authorities in one of Norway’s more than 400 municipalities to participate in introductory courses and integrate into the Norwegian society. These municipalities vary substantially on a range of factors that might impact later integration, such as their population size and centrality, the number and proportion of immigrants, labor market conditions and the income and educational level of those who already live there. A relevant question is, therefore, whether the characteristics of the municipality a refugee is assigned to, causally impacts his or her later integration. To provide a general answer to this question it is essential that there are no systematic differences (that cannot be rigorously controlled for) in who is settled in what type of municipality. Thus far little systematic evidence exists about how The Directorate of Integration and Diversity (IMDi), the authority responsible for the settlement process, assigns refugees to municipalities, and whether this policy can form the basis of a causal research design.
The report starts with an overview of the relevant international literature, including an overview of various European settlement policies. Specific attention is given to Scandinavian policies and studies. The Norwegian settlement policy assigns refugees to municipalities based on a dialogue between IMDi (and previously The Norwegian Directorate of Immigration, UDI) and municipality-level authorities. To provide an overview of this process, as well the ways in which various guidelines of assignment have been executed in practice, we rely on both reports and other official documents as well as interviews with relevant employees at IMDi. This overview illustrates that several individual- and municipality-level factors influence the settlement process, whereas a swift assignment of refugees to municipalities long has been an overarching goal.
Next, we rely on data from Statistics Norway and UDI to empirically explore how the individual characteristics of refugees correlate with the characteristics of the municipality in which they are settled. We study all refugees who were settled through the above-mentioned settlement policy between 2002 and 2012 and include both quota refugees and asylum seekers. The results confirm the literature review and interviews and show that there exist systematic differences between refugees who are assigned to different municipalities (i.e., the settlement process is endogenous). We do for instance find that highly educated refugees, as well as women, to a greater extent are placed in more central, populous municipalities with high educational and income levels among the population, while families with children as well as quota refugees to a greater extent are placed in less central and less populous municipalities with lower educational and income levels. It seems that the settlement of quota refugees is more random than the settlement of asylum seekers, and that larger refugee inflows and a greater focus on a swift settlement is correlated with fewer systematic differences between individuals.
Given that it is not completely random who is placed where, it is methodologically challenging to use the Norwegian settlement policy as the starting point of general causal analyses of the effect of municipality characteristics on integrational outcomes. Descriptive analyses show that correlations exist between several municipality variables (including centrality, population size and unemployment rate) and several integration outcomes (including employment, education and the propensity to relocate to another municipality) but based on the current analyses we cannot determine whether these patterns are driven by selection on unobservable characteristics or not. We conclude by briefly discussing the potential for future causal analyses into these questions.