![Sample our Built Environment journals, sign in here to start your access, Latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5926/TOC-Subject-Sample-2021-Built-Environment.jpg"})
Abstract
This article contributes to research on the homeownership gap between immigrants and native-born residents in Western countries, extending earlier research using longitudinal data and studying a country with a short history of immigration. Discrete-time survival analysis and statistical decomposition are applied to compare the duration of entry to homeownership between non-Western immigrants and native-born residents moving to the Helsinki Metropolitan Area in Finland, using individual-level register-based data from 1990 to 2008. The results show considerable differences between groups in the speed of entry to homeownership. The majority of these differences can be explained by observed differences in economic and demographic characteristics. Therefore, differences in economic integration are an important explanation for the homeownership gaps. However, for some groups, considerable gaps remain, requiring additional explanations. From a methodological viewpoint, the results indicate that in cross-sectional analyses, the significance of economic resources as an explanation for the homeownership gaps may be underestimated.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 The socioeconomic resources can themselves be strongly influenced for instance by cultural factors, stratification processes in the labor market, welfare state arrangements, and integration policies, but the focus here is on the housing market processes.
2 In this study, the term ‘migrant’ is used to refer to all persons moving to the HMA, including both Finnish-born and foreign-born persons and both internal and international migrants. Therefore, this term does not imply an immigrant origin.
3 Each individual’s first move to the HMA between the years 1991 and 2005 was selected. The data include only those migrants who were not in a child’s position in a family at the end of the year of their migration to the HMA. A sensitivity analysis using 25–49-year-old migrants was conducted, and all conclusions stayed the same as with the 18–49-year-old migrants.
4 In the remaining sample, survival analyses indicate that within 10 follow-up years, 10 per cent of those Finnish-born migrants who do not enter homeownership or leave the Uusimaa region to another part of Finland either out-migrate or die. Among immigrants, this proportion varies between 13 and 35 per cent (Russia 13 per cent, Estonia 19 per cent, Other East Europe 21 per cent, North Africa and West Asia 25 per cent, Sub-Saharan Africa 35 per cent, and Other Asia 32 per cent).
5 Because some ethnically Finnish persons are born in other countries, the native speakers of Finnish, Swedish, and Sami languages were removed from the remaining immigrant groups (with the exception of Finnish speakers born in Russia and Estonia, see Overview of Immigrants’ Housing Position …).
6 Educational level is often used as a proxy of expected long-term income (cf. Alba & Logan, Citation1992; DeSilva & Elmelech, Citation2012). It is not used in this article, as education is not measured reliably for immigrants in the Finnish registers. Data on wealth were not available either.
7 Models with random effects at the individual level were also tested, but statistically significant random effects were identified only among the Finnish-born group and these models yielded almost exactly the same coefficients and significance levels as the models reported here.
8 When the decomposition was done separately by sex (with 10-year follow-ups), it turned out that a higher share of the gap can be explained in the case of females (49–71 per cent among females and 36–61 per cent among males), with both demographic and economic characteristics explaining more. The only exception was the ‘Other Asia’ group, with no clear difference between the sexes. The order of the groups is approximately the same for both sexes. In addition to differences in the explained share by sex, the explained share was found to be smaller (29–57 per cent) in a shorter five-year follow-up, i.e. during the initial years in the HMA.
9 In an alternative specification with absolute instead of logarithm-transformed income values, a 1000 EUR increase in household income from the previous year had an average marginal effect of 0.19 percentage-points among the Finnish-born group and between 0.03 (Sub-Saharan Africa) and 0.13 (Other East Europe) percentage-points among immigrants.
10 The longitudinal models were the same as assessed above, and the cross-sectional models used data for a single year, either 2000 or 2005, with purely cross-sectional versions of the explanatory variables and the housing tenure. The sample for the cross-sectional design was constructed similarly to the longitudinal sample, except that those entering homeownership were not removed from the data. Survey weighting was applied when running pooled models to take the stratified sampling design into account.
11 The large difference in the average income level also partly explains why entry to homeownership is faster among the Finnish-born group, even if the relative income growth is faster among the immigrants: despite faster relative income growth, immigrants still lag behind in the income level.
12 In the sample consisting of those who stayed in Finland for at least two years after the year of arrival, 19 per cent left (or died) during the first 10 follow-up years. Pungas et al. (Citation2012) found that 24 per cent of the Estonians in Finland have intentions to return, with ‘income migrants’ working in jobs below their educational qualifications especially prone to having such intentions.
13 Thirty-five per cent of the Sub-Saharan group and 36 per cent of Somalis left Finland—or died—during the first 10 follow-up years.