Explaining Attitudes toward Immigration: The Role of Regional Context and Individual Predispositions

Abstract

Existing research makes competing predictions and yields contradictory findings about the relationships between natives’ exposure to immigrants and their attitudes toward immigration. Engaging this disjuncture, this article argues that individual predispositions moderate the impact of exposure to immigrants on negative attitudes toward immigrants. Negative attitudes toward immigration are more likely among individuals who are most sensitive to such threats. Because country-level studies are generally unable to appropriately measure the immigration context in which individuals form their attitudes, this article uses a newly collected dataset on regional immigration patterns in Austria, Germany, and Switzerland to test the argument. The data show that increasing and visible diversity is associated with negative attitudes toward immigrants, but only among natives on the political right. This finding improves the understanding of attitudes toward immigrants and immigration and has implications for the study of attitudes toward other policies and for immigration policy itself.

Acknowledgements

We are grateful to Michael Burch, Duncan Lawrence, Anand Menon, and two anonymous reviewers for helpful comments. For research assistance, we thank Amira Jadoon and Kerri Anne Watson. A previous version of this paper was presented at the 2012 meeting of the Midwest Political Science Association.

Disclosure Statement

No potential conflict of interest was reported by the authors.

Supplemental Data

All replication data and code necessary to reproduce the analyses in this article are available at http://thedata.harvard.edu/dvn/dv/jkarreth. An online appendix is posted at http://www.jkarreth.net.

Notes

1. As far as regional-level data on immigrants and the respective regional identifiers for survey respondents are available, a replication of our study in other countries in future research would be a valuable test of the utility of a MSSD in this context.

2. See the website of the Swiss Federal Statistical Office at http://www.bfs.admin.ch/bfs/portal/de/index/regionen/11/geo/analyse_regionen/02a.html (accessed 15 November 2014).

3. In the empirical models below, we also performed robustness checks where we accounted for additional sources of heterogeneity between these regions by controlling for population size, geographic area, and (separately) population density. None of these variables exhibited a consistent relationship with the variables of interest.

4. Due to idiosyncrasies in the Swiss population data, we created an artificial region Zürich–Northwestern Switzerland for 2002; in the years 2004, 2006, and 2008, Zürich and Northwestern Switzerland are in the sample as separate regions. Austrian responses are not included in the 2008 data file of the ESS; hence respondents from Austrian Bundesländer are part of our sample for the years 2002, 2004, and 2006.

5. Due to data availability, these 11 countries are Austria, Belgium, Germany, Denmark, Finland, United Kingdom, Greece, Italy, Netherlands, Norway, and Sweden.

6. Data available at http://www.migrationinformation.org/datahub/countrydata/data.cfm (accessed 13 October 2013).

7. The figures were created with immigration data we collected across the three countries and from the European Social Survey, both of which are described in detail below.

8. Eighty per cent of respondents fall into this category. This is the most specific piece of information about respondents’ origin we can extract from the ESS. While it is theoretically possible that some respondents have an immigrant background despite both of their parents being born in their current country of residence, the proportion of such individuals should be so small that it would not bias our findings.

9. These results are shown in Tables A15, A16, A17, and A18 in the online appendix. Missing information on some regions for these years reduces the sample size in these models. Unfortunately, the lack of data on historical immigrant presence before 1991 for all countries in the sample prevents us from using immigrant levels from previous years for additional robustness checks.

10. Instead of using relative changes, we control for the baseline percentage of immigrants in 1991. Holding the baseline level constant allows us to make inferences about appropriately comparable changes in the presence of migrants.

11. Additional robustness checks with split samples instead of interactions return the same results (see Tables A20–A23 and Figures A3–4 in the online appendix).