Developments in residential segregation in selected major and central municipalities after 2005
This analysis describes residential segregation among immigrants and residents with two immigrant parents in Norway and changes in segregation patterns during the period 2005-2017. The motivation for the study is the significant increase in the immigration to Norway during this period. While the number of immigrants and inhabitants born in Norway with two immigrant parents was approximately 365,000 in 2005, the corresponding figure in 2017 was more than twice that size (about 884,000 people). The annual increase was relatively moderate in the years 2000-2006 before we gained stronger growth in the years 2007-2015.
In spite of the fact that immigration to Norway has increased considerable during the last 15 years, residential segregation as measured by the index of dissimilarity (D index) has decreased, both from 2005-2011 and from 2011-2017 among immigrants and Norwegian-born persons with two immigrant parents. When we divide immigrants according to country of birth, we find that the decline is due to a decline in residential segregation among immigrants from country group 2, i.e., new EU countries in Eastern Europe, and a decline among immigrants from country group 3, i.e., the countries of Eastern Europe that are not members of the EU, Asia, Africa, Latin America and Oceania, except Australia and New Zealand. For the other immigrants, i.e., those from country group 1 (EU/EØS countries in Europe and Switzerland, USA, Canada, Australia and New Zealand), we find a slight increase in residential segregation from 2005 to 2017. Male immigrants from country group 1 and 2 show somewhat larger residential segregation than women, in all the three years 2005, 2011 and 2017. Over time, there is a decline in residential segregation among all age groups, except for the oldest (67 years and older). While for immigrants from new EU countries in Eastern Europe there is no systematic link between residential segregation and duration of residence in 2015, we find a clear pattern for immigrants from more remote countries (country group 3). For these immigrants, there is a positive correlation between duration of residence and residential segregation. With regard to the association with education, residential segregation has declined over time in all educational groups when we look at immigrants from country group 2 and 3. This finding also applies for immigrants with no educational information. In 2015 there is a negative association between residential segregation and length of education.
Residential segregation also varies with labor market status. In 2015, immigrants who are employed and immigrants who are neither employed nor undergoing education are among those with the lowest residential segregation. This finding applies to immigrants from all three country groups. Among immigrants from remote countries (country group 3), the unemployed person has the largest residential segregation in all three years we study. Unemployed immigrants from country group 2, on the other hand, do not differ by having particularly high residential segregation in 2015, but this group had a significant decline in residential segregation from 2005-2011, and to a lesser extent from 2011-2015.
We have also looked at the contribution from the 21 municipalities we have studied to the D index. Oslo stands out by having a particularly high contribution in all three years 2005, 2011 and 2017, but it decreases somewhat over time. Other municipalities with high contributions (in descending order in 2017) are Bergen, Trondheim, Drammen, Bærum, Fredrikstad and Tromsø, while Skedsmo, Kristiansand and Stavanger stand out with an increase in the contribution to the D-index throughout the investigation period.
Since the situation in Oslo is very important for the development of the D index, we have also conducted separate calculations for Oslo alone. Regarding the local living areas in Oslo, we find that for many areas with a particularly high residential segregation in 2005, residential segregation has been reduced in 2017. We also find examples of the opposite, i.e., that local areas with relatively high residential segregation in 2005 have gained even higher segregation in 2017, but these areas are relatively few. Another important finding is that a number of areas with a particularly low fraction of immigrants in 2005 have got relatively more immigrants in 2017.
Residential segregation is measured by the dissimilarity index. This index is a measure of the uniformity of the distribution of immigrant residents, or residents with two immigrant parents versus the rest of the population over different local living areas, which together constitute a larger area. The value of the index can be interpreted as the percentage of immigrants included in the calculations that should move to other local living areas if one wishes a distribution that corresponds to the distribution in the area under one. Instead of looking at the whole country, we have selected 21 large and central municipalities with relatively many immigrants. These municipalities are then divided into 380 local living areas, and these are the areas that form the basis for most of the calculations we have made.