Settlement Location Shapes the Integration of Forced Migrants: * Evidence from Post-War Germany
Total Page:16
File Type:pdf, Size:1020Kb
Settlement Location Shapes the Integration of Forced Migrants: * Evidence from Post-war Germany Sebastian T. Braun University of Bayreuth, CReAM, IZA, IfW, RWI Essen Nadja Dwenger University of Hohenheim, CEPR, CESifo Abstract Following one of the largest displacements in human history, almost eight million forced migrants arrived in West Germany after WWII. We study empirically how the settlement location of migrants affected their economic, social and political integration in West Germany. We first document large differences in integration outcomes across West German counties. We then show that high inflows of migrants and a large agrarian base hampered integration. Religious differences between migrants and natives had no effect on economic integration. Yet, they decreased intermarriage rates and strengthened anti-migrant parties. Based on our estimates, we simulate the regional distribution of migrants that maximizes their labor force participation. Intra-German migration in the 1950s brought the actual distribution closer to its optimum. This work is licensed under a Creative Commons \Attribution- NonCommercial-NoDerivatives 4.0 International" license. Keywords: Forced migration; Regional Integration; Post-War Germany JEL Classification: N34; J15; J61 Declarations of interest: none Published version: Sebastian Till Braun, Nadja Dwenger (2020). Settlement location shapes the integra- tion of forced migrants: Evidence from post-war Germany. Explorations in Economic History, Volume 77, 101330. https://doi.org/10.1016/j.eeh.2020.101330 *The research in this paper was funded by Deutsche Forschungsgemeinschaft (project number 267108021, \Die volkswirtschaftlichen Effekte der Vertriebenen und ihre Integration in Westdeutschland, 1945-70"). An earlier version of this paper has been circulated under the title `The Local Environment Shapes Refugee Integration: Evidence from Post-war Germany'. All remaining errors are our own. University of Bayreuth, Centre for Research and Analysis of Migration, IZA { Institute of Labor Economics, Kiel Institute for the World Economy and RWI Research Network. Address: University of Bayreuth, Faculty of Law, Business and Economics, 95440 Bayreuth, Germany. Email: [email protected]. Tel.: +49-921-55 6256. University of Hohenheim, CEPR and CESifo. Address: University of Hohenheim, Lehrstuhl fur¨ Finanzwis- senschaft (520D), Schloß-Mittelhof (Ost), 70593 Stuttgart, Germany. Email: [email protected]. Tel.: +49-711-459 22990. 1 Introduction At the end of 2018, UNHCR, the UN Refugee Agency, counted more than 70 million forced migrants, including 25.9 million refugees and 41.3 million internally displaced people (IDPs) (UNHCR 2019).1 This is the highest level recorded by UNHCR in its 70 years of existence. The dramatic increase in the number of forced migrants has sparked heated debates about the ability of receiving regions to integrate forced migrants into their economies and societies. As a response to rising migrant numbers, many Western countries have introduced dispersal policies or re-location bans to prevent immigrant concentration in large cities (Brell et al. 2020, Fasani et al. 2018), and the European Union continues to debate mechanisms for allocating forced migrants between member countries (Hatton 2015, Fern´andez-Huertas Moraga and Rapoport 2015). One key motivation for assigning forced migrants to resettlement location is to foster their integration (Bansak et al. 2018, Brell et al. 2020). However, many important questions on the effect of resettlement location on integration outcomes still await conclusive answers. Does it matter, for instance, whether forced migrants are re-settled to agrarian or industrialized regions, to culturally close or distant regions, in small or large numbers? This paper addresses these questions in the context of one of the largest forced population movements in history, the displacement of Germans from East and Central Europe during and after World War II. At the time, authorities could not account for local integration prospects when settling forced migrants, due to the size and pace of the inflow and the chaotic post-war circumstances. Furthermore, strict moving restrictions prevented migrants from sorting into specific regions after their arrival. These features of our historical setting allow us to abstract from many confounding factors that usually complicate empirical analyses of the effect of resettlement location on integration. Between 1944 and 1950, eight million displaced Germans, so-called expellees (Heimatver- triebene), arrived in West Germany. The majority of them (55 percent or 4.4 million) were IDPs who came from the eastern provinces that Germany ceded after World War II. The remaining ones had lived in East and Central Europe before the war, mostly in the Czech Sudetenland or in the eastern territories that Germany had lost after World War I. The mass inflow of expellees dramatically increased the West German population. By 1950, every sixth West German resi- dent was an expellee. The integration of the economically impoverished newcomers into the West German economy and society is widely seen as one of the key challenges that the war-ridden country faced after 1945 (Connor 2007, Grosser 2006). This paper uses newly digitized administrative data at the county level to provide a com- prehensive empirical assessment of the regional conditions that influenced the economic, social, and political integration of expellees. As we document in the first part of the paper, integration outcomes differed strongly across West German counties. These differences were not hidden from contemporary observers (Reichling 1958, Pfeil 1958). Our empirical analysis in the second part of the paper tests the three key hypotheses that have been put forward by historians and social sci- entists to explain the differences in integration outcomes (Connor 2007, Pfeil 1958, Schulze 2002). First, we assess whether high population shares of expellees hampered integration. Second, we check whether agrarian regions were less successful in integrating expellees than industrialized 1The key difference between refugees and IDPs is that the latter never cross an international border to find safety while the former do (UNHCR 2019). 1 regions. Third, we test whether religious differences between expellees and non-expellees ham- pered integration. We measure integration by a broad set of economic, social and political outcomes, which include expellees' labor force participation, inter-marriage rates between ex- pellees and native West Germans, and electoral support for expellee and anti-expellee parties. We find that regional conditions were indeed key for integration outcomes. Three features of our setting are crucial for the empirical analysis. First, integration outcomes varied widely across the 526 West German counties in our data. The expellee labor-force-to- population rate in 1950, for instance, ranged from 31.6% to 59.0%. Second, expellees encoun- tered very different local socioeconomic conditions. West German counties were not only very different in their sectoral employment structure and predominant Christian confession. They also differed strongly in the population share of expellees, which ranged from 1.8% to 44.1% in 1950. Third, our historical setting arguably creates quasi-exogenous variation in the regional distribution of expellees. In particular, the initial distribution was driven by the geographic proximity to expellees' origin regions, not by integration prospects (Connor 2007, Muller¨ and Simon 1959, Nellner 1959). At the end of World War II, Germans from Eastern Europe fled from the approaching Red Army to nearby regions in West Germany. Expellees thus gathered in the eastern parts of West Germany that were most accessible to them. While expellees did not choose their initial location based on integration prospects, the local housing supply did influence their resettlement location in West Germany. Authorities in West Germany, overwhelmed by the size and pace of the inflow, were unable to distribute expellees according to their religious affiliation or local job prospects. Instead, their prime concern was to provide expellees with a roof over their heads (Nellner 1959). Yet, most of the housing stock laid in ruins after the war. One key challenge for our empirical analysis is thus that war destruction might have driven both the regional distribution of expellees and their integration outcomes. We ease this concern by controlling for various indicators of war destruction. We then show that conditional on covariates, regional expellee shares do not correlate with the pre-war economic situation. Likewise, the skill level of expellees did not differ systematically across regions. The occupying powers imposed severe restrictions on intra-German migration until 1949. The initial regional distribution of expellees thus persisted for several years after the war. Our empirical analysis uses an instrumental variable strategy to address remaining concerns that expellees relocated endogenously within Germany after their initial placement. The instruments isolate the variation in regional expellee shares and religious distance that is due to the initial placement of expellees, not their later movements. Our key results underline the importance of the socioeconomic conditions at the resettlement location for expellee integration. First, the regional expellee share had strong negative effects on the economic, social and political integration of expellees in 1950, i.e.,