Do Immigrant Integration Policies Matter? A Three-Country Comparison among Turkish Immigrants

Bibliographic Data

ID6405973
AuthorsEvelyn Ersanilli (0000-0002-6286-1688, University of Oxford), Ruud Koopmans
Year2011
Volume34
Issue2
Pages208-234
Publication date2011-03-01
Peer ReviewedYes
Open AccessNo
TypeARTICLE
VenueWest European Politics (JOURNAL)
Journal identifiersISSN: 0140-2382 • E-ISSN: 1743-9655
PublisherInforma UK Limited (PUBLISHER • GB)
DOI10.1080/01402382.2011.546568
OpenAlexW1972980990
LanguageEN
Citations received99
References cited43

Various theoretical perspectives make strong, but often contradictory, claims about effects of immigrant integration policies on immigrants' retention of their ethnic cultures and their adoption of the host country's culture. However, there is very little empirical research investigating these competing claims. This article addresses this gap by investigating the merits of four prominent theoretical perspectives, which emphasise respectively the material costs and benefits of retention and adoption, acculturative stress, the permeability of ethnic boundaries, and reactive ethnicity. It uses original survey data on Turkish immigrants in Germany, France, and the Netherlands, countries with different integration regimes, and investigates identification, language proficiency and use, religious observance and interethnic social contacts. The results indicate that policies have only a modest effect on immigrants’ degree of adoption and retention. The findings clearly contradict the reactive ethnicity and acculturative stress hypotheses, and provide support for a combination of the material cost/benefit and boundary permeability perspectives. Notes 1. See http://www.integrationindex.eu, http://emilie.eliamep.gr/, http://www.imiscoe.org/natac/ (accessed 30 October 2009). 2. See http://www.integrationindex.eu/ (accessed 30 October 2009). 3. See the yearly reports ‘Multiculturele programmering’ of the public broadcasting organisations, e.g. http://pics.portal.omroep.nl/upnos/ZakoiolHC_RAP_MC2004_21.pdf (accessed 30 September 2009). 4. South-central Anatolia consists of the provinces Afyon, Aksaray, Karaman, Kayseri, Konya, Nevşehir, and Niğde. East-central Anatolia encompasses Adiyaman, Amasya, Elaziğ, Malatya, Tokat, Tunceli and Sivas. 5. For more information on the sampling procedures and response rates see Ersanilli (2010 Ersanilli, Evelyn. 2010. Comparing Integration. Host Culture Adoption and Ethnic Retention among Turkish Immigrants and their Descendants in France, Germany and the Netherlands, Amsterdam: Vrije Universiteit. [Google Scholar]). 6. For respondents without a partner or children, the scale was based on the average of the remaining items. 7. For male respondents without a partner, the scale excluded the headscarf question and was based on the average of the remaining items. 8. Since, except for religious observance that is based on z-scores, our dependent variables are categorical, we repeated the analyses with Ordered Logistic Regression. These analyses showed the same cross-national differences as OLS. The only exception is that the significance of the difference in interethnic social contacts between France and the Netherlands drops to the p < 0.10 level. 9. Age was not included as a control variable because of the consequent multicollinearity. However, we repeated all analyses with age and age2 instead of the generation dummies. This led to minimally different results. The most notable difference is the loss of significance of the level of proficiency in the host country language between the Netherlands and France. Full tables are available on request. 10. For the adoption variables we also ran analyses using the share of the total immigrant population instead of the Turkish immigrant population, following the reasoning that adoption might be positively related to the population share of host country ethnics rather than the share of non-Turks. For social contacts the difference between the Netherlands and France then decreases to the p < 0.10 level. For the other variables the results do not change (tables available on request). 11. For religious identification the model improvement after adding the country dummies is significant at p < 0.05, for religious observance at p < 0.001. 12. For host country identification the model improvement after adding country dummies is significant at p < 0.001, for host country language proficiency marginally significant at p < 0.10, for host country language use at p < 0.001, and for social contacts at p < 0.05. 13. Estimates are based on all people who are married or living in common law. Partners are estimated to be marriage migrants when they were born in Turkey and came to the host country after the respondent turned 18

Acculturation · Country of origin · Demographic economics · Economics · Ethnic group · Geography · Immigration · Political science · Sociology · Turkish · Diaspora, migration, transnational identity · Law · Migration and Labor Dynamics · Migration, Refugees, and Integration

Tools

Open DOISci-Hub