The intergenerational cost of randomly allocating refugees

In terms of 2015/16 refugee migration from Syria to Europe and North America, Dustman et al. (2016) and Hatton (2018) call for the renewal of the European refugee resettlement system. However, what should be an optimal allocation process remains an open question. Strict allocations to high-unemployment zones in Germany, for example, negatively affect the economic and social integration of refugees (Aksoy et al. 2020). Fasani etc. (2016) argue that free residential mobility can encourage labor market integration. Importantly, existing research has investigated the integration of refugees themselves. Whether and how the background and local conditions affect not only the first generation but also the children of the refugees is still unknown.

The ability to communicate is especially important for mothers. Communication can be facilitated by peers who provide important information on childbirth or by proficiency in the local language (Lazear 1999). Language skills are considered to be one of the most important components of human capital for immigrants, which is crucial for their successful participation in host society. Communication power is associated with labor market integration (Auer 2018), electoral participation (Houle 2019), and social capital (Cheung and Philimore 2014). In the context of forced displacement, language training is especially important for newly arrived refugees. Due to the perceived lack of preparation and often the absence of bonds and social networks, most refugees have to learn the language of their destination country (Foged et al. 2021).

In a recent study (Auer and Kunz 2021), we exploit a feature of Switzerland’s asylum policy: newly arrived refugees are randomly assigned across the state and, consequently, across three distinct language regions (German-, French- and Italian-speaking regions). As a result of this external distribution, refugees – who may be from countries with large French-speaking or Italian-speaking populations – are incidentally assigned to a familiar or unfamiliar language environment (see Figure 1). The additional fact that allocations are mandatory – as refugees are not allowed to move to other areas (with very few exceptions) – allows us to follow their integration trajectory over time.

Figure 1 Significant French- and Italian-speaking populations in the Swiss-speaking region and around the world

We study the well-being of children born to refugee mothers from the same country and the allocation in familiar vs. unfamiliar language environments with otherwise comparative characteristics. For example, a pregnant refugee from C ডিte d’Ivoire (French-open) who has been randomly assigned to Geneva in the French-speaking region of Switzerland may be able to obtain and follow information on where to get pregnancy check-ups. The doctor’s advice was compared to that of his pregnant co-conspirator, who was randomly assigned to the German-speaking region of Zurich. In addition, he may benefit from other aspects of his local surroundings; Being able to read the ingredients of food packaging can, for example, facilitate positive health behaviors.

Our data includes all child births in the country, including administrative accounts and detailed health information for all refugees who came to Switzerland between 2010 and 2017. We find no evidence of structural differences, election imbalances, or differentiated fertility preferences among refugees whose language matches their assigned territory and who do not. In other words, randomly assigned to a familiar language environment does not increase one’s chances of giving birth, nor does it change the time of conception.

Original search

The primary result of our interest is the birth weight of the baby. Birth weight predicts educational achievement, income, credit default, and health in later life, among many other results. Comparing refugees from the same source across the destination shows a significantly consistent pattern of their baby’s birth weight (Figure 2). The gray line shows the birth weight of the children of refugees without language correspondence in Switzerland (such as the Persian-speaking refugees in Afghanistan). The average birth weight is almost flat, with little advantage for those who are the most economically prosperous in the German-speaking region. In contrast, blue lines (light: any French exposure; dark: French as an official language) show large birth weight gain, both bilingual but predominantly French-speaking and exclusively French-speaking in Switzerland.

Finally, the green line shows that refugees from large Italian-speaking countries (Libya and Somalia) benefit greatly from being allocated to Italian-speaking areas (a small group consisting of only two sending countries and a region of Switzerland; therefore, the results of these groups Should not be over-explained).

Figure 2 The birth weight of the children of the refugees by the expression of the original language of the parents and the main language of the assigned birth zone

This analysis shows that the benefits of being able to communicate are great. We find that, on average, children of mothers who were accidentally assigned to a familiar language environment weighed 72 grams more than children of co-mothers who arrived in Switzerland at the same time but were assigned to an unfamiliar language region. Compared to the average birth weight of about 3,200 grams of our refugee population, this is an increase of 2.2% by birth weight. (In the paper, we show that these findings contain different regression specifications that differentiate between local environment, country of origin, and time of arrival, accounting for well-known risk factors for child health).

The effect is felt not only in the birth weight distribution average, but also in the lower tail, where weight changes can significantly change the well-being of the child and later life: Clinical low birth weight indicator (weight <2,500 g) decreased by 2.9 percentage points (6.98% from average). These effects are substantial compared to targeted interventions; For example, they are two to three times more effective than attending a $ 3,867 annual tax credit for low-educated black mothers in the United States (Hoynes et al. 2015) or a nutritious food program for low-income mothers in the United States (Rosin-Slater 2013).

In addition, we find evidence that a more extensive local network is an alternative to the convenience of being placed in a familiar language environment, mainly when this network includes refugee mothers who are relatively well acquainted with the newborn. In other words, the mother’s reliance on communication skills decreases when conscious co-nationals are present. While these networks are less likely to affect the quality of communication with physicians or the language coherence with physicians and healthcare professionals, they appear to be important for sharing knowledge about health care and health-related behaviors.

We find no evidence that the difference in potential earnings drives the effects we observe through employment or further extended residency (language) assimilation in Switzerland. Importantly, we have already observed the high birth weight of pregnant mothers in Switzerland, indicating that assimilation cannot explain our results. Conversely, we do not notice any difference in gestational age, which is often associated with external shocks during pregnancy (such as losing a mate or experiencing a natural disaster). Furthermore, the positive impact of a familiar language environment on a child’s health is greatest when mothers are born in a country where medical personnel are not usually present for delivery – that is, when they are less likely to be familiar with state-of-the-art healthcare services, such as regular check-ups. Up

The impact of our results

At the immediate policy level, our results highlight the importance of early policy intervention in addressing the systematic difficulties of vulnerable groups. Measures include language training, comprehensive interpretation services for newly arrived refugees and language-adequate information campaigns to raise awareness of available welfare and health services.

On a more general level, we echo the arguments against the current form of ‘spontaneous’ allocation of asylum seekers and in favor of a more comprehensive rehabilitation program (Hatton 2016). In a large survey, Bansak et al. (2016) found that European citizens support fair allocations to incoming refugee countries. From a policy standpoint, recent efforts to improve the allocation of refugees have been multiplied. Bonsak et al. (2018) and Acharya and others. (2019) suggest an algorithmic approach to maximize the potential earnings of refugees; Delacretaz et al. (2016, 2019) and Jones and Teytelboym (2018), on the other hand, favor a choice-based matching algorithm that respects the preferences of refugees. None of these approaches, so far, have taken the next generation into account. Our findings, however, suggest that increasing children’s well-being significantly affects how their parents are allocated across countries and territories and thus should be part of the discussion of fair and optimal allocation approaches.


Acharya, A, K Bansak and J Hainmueller (2019), “Matching Refugees for Country Position Based on Preference and Outcome”, eScholarship, University of California.

Aksoy, CG, P Poutvaara and F Schikora (2020), “Near First Time: Local Conditions and the Integration of Refugees”,, 11 December.

Auer, D (2018), “Language Roulette – Impact of Random Placement on Refugee Labor Market Integration”, Journal of Ethnic and Migration Studies 44 (3): 341–62.

Auer, D, and JS Kunz (2021), “Communication Barriers and Child Health: The Intergenerational Impact of Random Allocation of Refugees Across the Language Region”, No. 867 GLO Discussion Paper.

Bansak, K, J Ferwerda, J Hainmueller, A Dillon, D Hangartner, D Lawrence and J Weinstein (2018), “Improving Refugee Integration through Data-Driven Algorithmic Assignments”, Science 359 (6373): 325-29.

Bansak, K, J Hainmueller and D Hangartner (2016), “How economic, humanitarian and religious concerns shape European attitudes towards asylum seekers”, Science 354 (6309): 217-22.

Cheung, SY, and J Philimore (2014), “Refugee, Social Capital, and Labor Market Integration in the UK”, Sociology 48 (3): 518-36.

Delacrétaz, D, SD Kominers and A Teytelboym (2016), “Refugee Resettlement”, Oxford University of Oxford Department of Economics Working Paper.

Delacrétaz, D, SD Kominers and A Teytelboym (2019), “Matching Mechanisms for Refugee Rehabilitation”, No. 2019-078.

Dustmann, C, F Fasani, T Frattini, L Minale and U Schӧnberg (2016), “On the Economics and Politics of Refugee Migration”,, 18 October.

Fasani, F. T. Frattini and El Minalle (2018) “(The Struggle) Refugee Integration in the Labor Market: Evidence from Europe”,, 9 April.

Foged, M, L Hasager and G Peri (2021), “Language Training and the Success of Refugees”,, 20 March.

Hangartner, D, and L. Schmid (2020), “Migration, Language and Employment”, unpublished manuscript, ETH Zurich and the University of Lucerne.

Hatton, T (2016), “Migration Crisis and Refugee Policy in Europe”,, 23 May.

Houle, C (2019), “Religion, language, race and ethnic voting”, Electoral Studies 61: 1-11.

Hoynes, H, D Miller and D Simon (2015), “Income, Earned Income Tax Credit, and Child Health”, American Economic Journal: Economic Policy 7 (1): 172–211.

Jones, W, and A Teytelboym (2018), “Local Refugee Match: Aligning Refugee Choices with Local Capabilities and Priorities”, Journal of Refugee Studies 31 (2): 152-78.

Lazear, EP (1999), “Culture and Language”, Journal of Political Economy 107 (S6): S95 – S126.

Rossin-Slater, M (2013), “WIC around you: new evidence on the impact of geographic access to clinics”, Journal of Public Economics 102: 51-69.

Leave a Reply

Your email address will not be published.