How do global crises affect privileged migrants? Return migration of German emigrants one year into the Covid-19 pandemic

ABSTRACT

This article discusses how global crises affect the return of international migrants. It develops a theoretical model and tests this contribution empirically taking Covid-19 as an instance of a global crisis. Existing studies have treated migration and mobility mostly as phenomena of virus dispersion. The reverse impact of Covid-19 on migration has attracted less attention. This article draws on a unique probability-sample of German emigrants to investigate actual return migration during the first year of Covid-19. These panel survey data were collected immediately before the pandemic onset in winter 2019 and one year later in winter 2020. We enhance these data with publicly available country-level data on Covid-19 incidence, excess mortality and Covid-19-induced democratic violations. Our findings from multi-level regression models suggest a moderate impact of country-specific pandemic indicators on return migration behaviour. Moreover, democratic violations in countries of residence and aggravating economic situations of individual households increase the likelihood of return. We conclude with a discussion of what can be learnt from the case of German return migrants about crisis migration in general.

KEYWORDS:

• Crisis migration
• Covid-19
• high skilled migration
• return migration
• origin country perspective

Acknowledgements

Thanks to Jonas A. Weik and Antonia Lang for their assistance with preparing the country-level data. We are also grateful for comments from Victoria Finn, Wagner Faria de Oliveira, participants of the seminar ‘Labor market Outcomes of Migrants’ at University of Frankfurt, and two anonymous reviewers. Earlier versions of this paper have been presented at the German Sociological Association's Section ‘Migration and Ethnic Minorities’ meeting at University of Göttingen (remote), the IMISCOE annual conference in Luxembourg (remote), and the NCCR Graduate Conference at University of Neuchatel (remote).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Article 12(4) of the International Covenant on Civil and Political Rights states that ‘[n]o one shall be arbitrarily deprived of the right to enter his own country’.

2 These statistics update the procedure by Ette and Erlinghagen (2021) for 2019/20, netting the raw statistics of ethnic Germans and de-registrations ex officio.

3 According to Henley and Partners, an agency that assists citizenship by investment, the German passport ranked 3rd in 2022 in a global ranking and yielded visa free travel to 190 destinations.

4 Additionally, 87 participants who lived abroad in wave 3 and moved to another country in 2020 have been excluded. These onward migrants have been excluded for conceptual reasons because their migration decisions follow different rationalities than return migration decisions. Note that our main findings are robust when we adjust the sample to alternative reference periods (see Appendix, Table A3).

5 Note that the coefficients are slightly smaller when we exclude January (see Appendix, Table A3).

6 In an additional question, participants were also asked whether they have seriously considered moving back to Germany or moving on to another country. Additional analyses based on this alternative operationalisation yield identical results with respect to all substantial hypotheses. Because of lower item non-response, we give preference to information about intentions to stay.

7 All Covid-19 measures throughout this study are standardised to 100,000 population always based on the United Nations standard projection for 2020 (United Nations, Department of Economic and Social Affairs, Population Division 2019).

8 The model fit is estimated applying the fit_meologit_2lev package for Stata provided by Wolfgang Langer (https://langer.soziologie.uni-halle.de/stata/index.html).

9 Furthermore, replications based on two smaller samples excluding countries of residence with fewer than 20 (5) observations did not change the substantial results.

Additional information

Funding

This research has been supported by the German Research Foundation [345626236].