Volume 52, September/October 2017, Number 5 · pp. 257-263
Integration of Refugees: Lessons from Bosnians in Five EU Countries
Mikkel Barslund, Centre for European Policy Studies, Brussels, Belgium.
Matthias Busse, Centre for European Policy Studies, Brussels, Belgium.
Karolien Lenaerts, Centre for European Policy Studies, Brussels, Belgium.
Lars Ludolph, Centre for European Policy Studies, Brussels, Belgium.
Vilde Renman, Permanent Representation of Sweden to the EU, Brussels, Belgium.
In 2015 Europe experienced the largest influx of refugees since the breakup of Yugoslavia in the early 1990s. While arrivals were down in 2016, the security situation in the Middle East and the instability in North Africa and elsewhere mean that Europe will play host to refugees for the foreseeable future.
Apart from the considerable humanitarian issues at stake, this compels European countries to facilitate the labour market integration of newcomers. The reasons are manifold. It would allow refugees to live a decent life and might accelerate their integration into the host society. At the same time, it would ease the strain on public finances and might improve the often negative public sentiment towards foreign nationals.
In this study, we follow up on Bosnian refugees displaced during the Bosnian war to investigate if it is possible to draw lessons for the current wave of refugees. Specifically, we looked at the experience in five countries: Austria, Denmark, Germany, the Netherlands and Sweden. Looking closely at Bosnian refugees enables us to see past immediate integration outcomes and take a longer-term view. Another consideration is that there is significant overlap among countries affected by the two refugee crises. The same five countries all saw a large absolute and relative influx of refugees in both 2015 and between 1992 and 1995, when the bulk of Bosnians arrived in Western Europe (Table 1).
Table 1 (back to the text)
Overview of registered refugees from Bosnia and Herzegovina, 1992-95
Registered refugees from Bosnia and Herzegovina
Refugees from Bosnia and Herzegovina as a share of the host country's population (%)
Source: M. Barslund et al.: Labour Market Integration of Refugees: A comparative survey of Bosnians in five EU countries, CEPS, 2016.
After a brief historical introduction, we review the reception and labour market conditions Bosnian refugees faced in the five host countries upon arrival. We then turn to the labour market outcomes and the educational attainment of Bosnians and their descendants, followed by a brief conclusion.
Historical background: Reception and labour market conditions in the five host countries
The Bosnian war – fought among the different ethnic groups in the current state of Bosnia and Herzegovina – took place from April 1992 until the Dayton Agreement in December 1995. Following the breakup of Yugoslavia, Bosnian Serbs rejected a referendum vote in favour of the country's secession from what was left of Yugoslavia. Subsequent ethnic cleansing against Muslim Bosniaks and Catholic Croats quickly led to a full-scale war. Zwierzchowski and Tabeau estimate the total number of fatalities and missing people at a minimum of 89,186, or two per cent of Bosnia and Herzegovina's 1991 population.Bosniaks suffered the most, with casualties and missing persons estimated at 57,992, equating to 3.1% of their overall population. About 1 million people were displaced internally, and about 1.2 million fled the country as war refugees.
Approximately half of the displaced Bosnians fled to Serbia, Montenegro and Croatia. The vast majority of the other half, consisting largely of Muslim Bosniaks, fled to Western European countries.
Initially, all Western European countries granted temporary protection to Bosnian refugees at the time of their arrival. For many host countries, it was the only way of dealing with the large influx of refugees without amending or overburdening their asylum systems. However, the UN Refugee Agency had deeper concerns. The organisation was interested in pushing the issue of burden-sharing of refugees across Europe.In fact, discussions at the time bear some resemblance to those taking place among EU members today. Temporary protection left the door open to involve (at a later stage) those Western countries that had not initially received many refugees from former Yugoslavia. This strategy turned out to be largely unsuccessful. A further concern, given the ethnic cleansing motive for the war, was the potential signal effect to warring parties of granting permanent residency immediately. It was feared that permanent residency could be seen as tacitly accepting that no return would be possible after the war.
However, after the initial period of temporary protection, the five host countries differed vastly in the type of residency granted to Bosnians, their access to national labour markets and education, integration measures offered to them, and the financial support they received. These differences are laid out in Table 2.
Table 2 (back to the text)
Overview of the institutional and legal framework in the five host countries during the Bosnian refugee crisis
This article is part of the Forum Integration of Immigrants in European Labour Markets.
Type of residency granted
Access to labour market and
Source: For a full list of sources, see M. Barslund et al.: Labour Market Integration of Refugees: A comparative survey of Bosnians in five EU countries, CEPS, 2016.
The table shows the level of hospitality extended to Bosnian with lighter colours indicating more favourable starting conditions. For example, Germany never converted the temporary status of refugees into permanent residency, the refugees had only very limited access to German labour markets and very few integration measures were set up to support them. At the other end of the scale, Sweden quickly granted most Bosnians permanent residency, allowed unrestricted access to the labour market and the Swedish education system, and offered integration measures such as language courses to refugees. The other countries fell in between those two approaches, with differences in the reception conditions reflecting the varying levels of political willingness to host refugees.
Three years after the outbreak of the war, a peace agreement, the "General Framework Agreement for Peace in Bosnia and Herzegovina" (commonly referred to as the Dayton Agreement), was signed in December 1995. The return of displaced Bosnians featured prominently in the Dayton Agreement, and all five countries in this study adopted various measures to facilitate return. Despite the Dayton Agreement, the ethnic aspect of the war meant that return was often complicated. Even after the peace agreement, many internally displaced Bosnians could not, or were not willing to, return to their former homes in so-called minority areas where they feared persecution. Economic conditions for the returnees were also difficult, with the official unemployment rate around 40% for most of the post-war period.
As a consequence, by 2005 only Germany had repatriated the vast majority of Bosnian refugees, while others had moved on to different countries (Figure 1). That Bosnians left Germany – and none of the other four countries – in large numbers is likely due to the aforementioned fact that permanent residency was rarely given in Germany.
Figure 1 (back to the text)
Annual number of returnees to Bosnia and Herzegovina
Note: The left-hand scale applies to Austria, Denmark, the Netherlands and Sweden; the right-hand scale applies to Germany.
Source: Own elaboration based on UNHCR data.
Thus, ten years after the Dayton Agreement, less than ten per cent of the initial refugee population remained in Germany. In the four other countries, the vast majority of Bosnians had settled in the countries permanently (Table 3).
At the time the Bosnian war was unfolding, the economic environment was unfavourable and quickly deteriorating in all host countries, with the possible exception of Austria. A global recession in the aftermath of the 1987 "Black Monday" stock market crash characterised much of the late 1980s and early 1990s, and Sweden in particular was going through a major economic crisis. Unemployment rates were on the rise in all five countries (Figure 2).
Figure 2 (back to the text)
Unemployment rates, 1990-99
Source: Own elaboration based on the IMF World Economic Outlook.
This observation is important for two reasons. First, economic variables that are subject to high short-term volatility, such as productivity growth or unemployment, are not likely to be decisive in refugees' choice of destination.However, these variables (if unfavourable, as was the case during the Bosnian war) do aggravate the labour market opportunities of refugees, an issue often exacerbated by priority systems. Fast integration into labour markets is often seen as crucial for social integration into societies. Weak labour market conditions, and the associated low probability of employment, may also contribute to path dependencies regarding the medium- to long-term employment prospects of refugees.
Second, adverse labour market conditions may also impact the host countries' hospitality towards refugees, especially among workers within low-skilled sectors. This sentiment appears common despite recent evidence from the UK, Austria and Denmark showing that an inflow of low-skilled refugees rarely leads to a displacement of local workers.It would therefore appear that Bosnian refugees did not face very favourable conditions for either short- or long-term integration.
Table 3 (back to the text)
Outflows of Bosnians from their country of reception and stock of Bosnian refugees in 2005
Moved to different country of reception
Repatriation to Bosnia and Herzegovina (1996-2005)
Number of (former) Bosnian refugees remaining in 2005
Source: Elaboration based on M. Valenta, S.P. Ramet: The Bosnian Diaspora, Integration in Transnational Communities, Burlington 2011, Ashgate.
Labour market and integration outcomes of Bosnian refugees
We now turn to the contemporary integration outcomes of Bosnian refugees in the four host countries of interest, excluding Germany due to the small percentage of Bosnians who remained in Germany after the end of the war. Of course, taking such a long-term perspective in a country comparative study is not without its problems. There is a lack of harmonised data to draw from, and national statistical databases and labour force surveys often do not disaggregate data on former Yugoslav nationals, nor between Bosnian war refugees and those coming later via family reunification or as economic migrants. The latter group of immigrants often differs significantly with respect to their success in labour markets and within education systems. Aggregating the groups, in particular when the composition can only be estimated roughly, complicates analyses.
An early insight thus emerges from this work: Internationally, there is a need for much better long-term tracking of immigrants in general and refugees in particular in order to analyse their experiences and draw conclusions for policy.
Despite these caveats, there is still considerable value in trying to extract information about integration experiences from various national data sources and comparing outcomes with those of the population of the host countries. Apart from national sources, we also utilise the European Labour Force Survey (EULFS).
Educational attainments of Bosnians at the time of arrival
It is interesting to note that the educational attainment of Bosnians differed among countries of arrival. The level was generally slightly higher than that of other refugees arriving to Europe at the same time but still significantly lower than the average educational attainments in the host countries (Figure 3).
Figure 3 (back to the text)
Educational attainment of Bosnians in various host countries compared to the host country's population
Source: For a full list of sources, see M. Barslund et al.: Labour Market Integration of Refugees: A comparative survey of Bosnians in five EU countries, CEPS, 2016.
Austria in particular, but also Denmark, received a disproportionally high share of low-educated Bosnian refugees compared to the average education level of the host country. For Austria, which had relatively favourable labour market conditions, this seems not to have affected labour market integration outcomes, as we document below. In Denmark, where the unemployment rate in 1992 stood at nine per cent, the relatively large number of low-skilled refugees may have played a role in defining initial labour market outcomes.
Labour market outcomes
We first turn to labour market outcomes shortly after the end of the Bosnian war. While the aggregating of all refugees from the region as "former Yugoslavian" does make it difficult to identify Bosnian refugees per se, it is possible to make some approximation of the true measure. The timeline of the wars in former Yugoslavia and comparisons of the total number of Yugoslav refugees with those from Bosnia suggest that immigrants from the former Yugoslavia who entered host countries from 1993 onwards are a reasonable proxy measure for Bosnian war refugees. Figure 4 shows employment rates in the different receiving countries of said group in 1998, along with those of the native-born population of the host country.
Figure 4 (back to the text)
Employment rate of former Yugoslav nationals in 1998 in various host countries
Sources: Eurostat; and J.D. Angrist, A.D. Kugler: Protective or Counter-Productive? Labour Market Institutions and the Effect of Immigration on EU Natives, in: The Economic Journal, Vol. 113, No. 488, 2003, pp. 302-331.
The employment rates show a very positive picture of the labour market outcomes experienced by former Yugoslav nationals in Austria, where this number had already reached 64% for recent immigrants in 1998. In all other countries of our sample, these employment rates remained subdued, between 26% (the Netherlands) and 33% (Denmark).
As shown in Figure 5, the gap in employment rates between Bosnian refugees and the native-born population for 25-54 year olds had been closed by 2008 in Austria and Sweden. In the Netherlands, the employment rate for Bosnians still trailed that of natives. However, it had vastly improved relative to 1998. For 2014, data is only available for Sweden and Austria, but the overall picture remains.
Data from national statistical sources for Denmark is again much less positive. The labour force participation of Bosnians in 2013 was around 20 percentage points below that of Danish nationals, approximately the same gap as for the reported employment rate in 2015.
It is worth noting that the data indicates that the gap in employment rates between men and women is larger for Bosnians than for natives. This is also found for other immigrant groups.The employment rate difference between men and women in the Netherlands is just above ten percentage points for nationals, whereas the difference is almost 30 percentage points for Bosnians, equating to a supplemental gender gap of close to 20 percentage points. Similar gaps are found in 2014 for Sweden (more than five percentage points) and Austria (around ten percentage points). While the extent differs, gender inequality in labour market integration is an issue among Bosnian refugees.
Figure 5 (back to the text)
Employment rate of Bosnian refugees in 2008 and 2014 in various host countries, 25-54 year olds
Note: Bosnian refugees in the EULFS are defined as a) migrating for humanitarian reasons, b) with the region of origin being Europe, outside of the EU28 countries, and c) having arrived in the early to mid-1990s. Cross-checking with UNHCR refugee macro statistics shows that the vast majority of migrants identified are Bosnians. The above leaves us with an overall sample of 597 observations for the three countries in 2008 and 2014.
Source: Own elaboration on EULFS data.
The larger variation in the employment rates of Bosnian refugees compared to the native-born population is likely a result of their relatively small sample size in the EULFS. However, Figure 5 suggests that in Austria and Sweden, the labour market performance of Bosnian refugees is very close to that of the the native-born population. Note that the employment rates could, however, be somewhat overestimated by two factors. First, a positive self-selection of those Bosnians staying in the country who found employment is likely. Second, the age composition of Bosnian refugees compared to the host countries' population is favourable. Many refugees who fled from the war in Bosnia were in their prime working age at the time of the studies. The first concern is clearly valid, but our results nevertheless hold for those who stayed in the country. The latter is the reason we restrict the sample to 25-54 year olds.
Educational outcomes of second generation
Although the evidence on labour market integration is mixed for Denmark and the Netherlands, a more positive picture emerges for young Bosnians' educational attainment in Denmark (Figure 6). While not all are second generation in the strict sense of being born in the host country by immigrant parents, those who were born outside and travelled with their parents as refugees were very young at the time. They therefore constitute a very close proxy for the educational performance of second-generation Bosnians.
Figure 6 (back to the text)
Share of 20-24 year olds pursuing further education in Denmark, by gender, 2012-13
Source: Ankestyrelsen: Bosniske krigsflygtninge fra medio 90'erne: Fakta om integration, 2014, available at https://ast.dk/filer/tal-og-undersogelser/integrationsomradet/bosnieranalyse_endelig.pdf.
For both male and female migrants, Bosnians are more likely to pursue tertiary education than ethnic Danes and non-Western migrants in Denmark. The situation is similar in the Netherlands, where over 40% of children to Bosnian refugees obtained a university or vocational university degree, a number above the national average.According to de Boom et al., 54.9% of the former Yugoslavs aged 15-24 were enrolled in full-time education in 2007, which is on par with Dutch nationals. Furthermore, Hessels reported that Yugoslav children in the Netherlands were outperforming their parents in terms of educational attainment, and most of them were enrolled in higher education. We interpret this as an indication that the initial disadvantageous labour market positions may be overcome and might not necessarily lead to unfavourable education outcomes in the second generation.
In this study, we reviewed the integration experience of Bosnian refugees displaced during the Bosnian war in the early 1990s. We compare reception conditions among the five Western European countries that played host to the majority of refugees at the time and track their labour market experience as well as, to the extent possible, education outcomes of second generation Bosnians.
While the overall integration experience has been a success story, a number of interesting, more nuanced findings emerge that may inform policymakers. In Austria, where both labour market conditions and the integration support extended towards Bosnians were relatively favourable, employment rates quickly reached those of the native-born population. In the Netherlands and Sweden, it took a little more than a decade for Bosnian refugees to catch up to the native-born population. Unfavourable initial labour market conditions may have inhibited policy efforts towards integration but did not have a long-term negative effect. In Denmark, Bosnian refugees faced difficult labour market conditions and received relatively little support upon arrival. This may have led to the observably worse employment outcomes that continue to prevail.
Despite differences in the educational attainments that Bosnians held when arriving in the host countries, we do not find any association between the level of education and labour market outcomes, neither in the short nor the long term.
Lastly, it needs to be stated again that Germany never intended to host Bosnian refugees permanently, provided little integration support and returned most refugees as soon as the Dayton Agreement of 1995 marked the end of the Bosnian war.
The recent evidence we find of young and second-generation Bosnian refugees' educational outcomes is promising. In both Denmark and the Netherlands, the share of this young group enrolled in tertiary education exceeds that of the native-born populations. By this metric, the integration experience has been a success.
We end by reiterating the call to policy makers of the need for better longitudinal data on integration outcomes. This is necessary not only to follow these outcomes in a descriptive way, but more importantly to be able to analyse the effects of interventions to improve integration. For third-country immigrants as a group, labour market integration is lacking in most European countries, and going forward, improving labour force participation and employment rates is of paramount importance.
* This article is based on research funded by the Mercator Foundation in the context of the MEDAM Project (Mercator Dialogue on Asylum and Migration). The authors thank the participants at the "Migration and the European Welfare States" conference in Malmö, 19-21 October 2016, Matthias Lücke, Tobias Stöhr and Alessandra Venturini for helpful suggestions and comments. Part of this article builds on M. Barslund, M. Busse, K. Lenaerts, L. Ludolph, V. Renman: Labour Market Integration of Refugees: A comparative survey of Bosnians in five EU countries, CEPS, 2016.
1J. Zwierzchowski, E. Tabeau: The 1992-95 War in Bosnia and Herzegovina: Census-based Multiple System Estimation of Casualities' Undercounting. Conference Paper for the International Research Workshop on 'The Global Costs of Conflict', The Households in Conflict Network (HiCN) and The German Institute for Economic Research (DIW Berlin), 1-2 February 2010.
2 M. Valenta, Z. Strabac: The Dynamics of Bosnian Refugee Migrations in the 1990s, Current Migration Trends and Future Prospects, in: Refugee Survey Quarterly, Vol. 32, No. 3, 2013, pp. 1-22.
3 R. Black, K. Koser (eds.): The End of the Refugee Cycle: Refugee Repatriation and Reconstruction, Oxford 1999, Berghahn Books.
4 E. Neumayer: Asylum Destination Choice – What Makes Some West European Countries More Attractive Than Others?, in: European Union Politics, Vol. 5, No. 2, 2004, pp. 155-180; W.H. Moore, S.M. Shellman: Whither Will They Go? A Global Study of Refugees' Destinations, 1965-1995, in: International Studies Quarterly, No. 51, 2007, pp. 811-834.
5 A priority system is an employment protection measure that usually puts temporary residence holders last. For example, in Austria, when Bosnian refugees held the legal status of aliens, they could only take up positions that could not be filled by Austrians, guest workers, recognised refugees under the 1951 Refugee Convention, labour migrants with social welfare credits or second generation aliens with at least five years of schooling in Austria. See J. van Selm-Thorburn: Refugee Protection in Europe: Lessons of the Yugoslav Crisis, Hague, Boston, London 1998, The Martinus Nijhoff Publishers.
6 C. Dustmann, T. Frattini, I. Preston: The Effect of Immigration along the Distribution of Wages, in: Review of Economic Studies, Vol. 80, No. 1, 2013, pp. 145-173; J. Bock-Schappelwein, P. Huber: Auswirkungen einer Erleichterung des Arbeitsmarktzuganges für Asylsuchende in Österreich, Österreichisches Institut für Wirtschaftsforschung, WIFO Project No. 9714, 2015; M. Foged, G. Peri: Immigrants' Effect on Native Workers: An Analysis on Longitudinal Data, IZA Discussion Paper No. 8961, Institute of Labor Economics, 2015.
7 M. Barslund, M. Busse, K. Lenaerts, L. Ludolph, V. Renman: Labour Market Integration of Refugees: A comparative survey of Bosnians in five EU countries, CEPS, 2016; Statistics Denmark: Statistical Yearbook 2017.
8 M. Barslund, A. Di Bartolomeo, L. Ludolph: Gender Inequality and Integration of Non-EU Migrants in the EU, CEPS Policy Insights No. 2017/06, February 2017.
9 M. Bolwijn, G. De Mooij: Ex-Joegoslaven zijn een succesverhaal, De Volkskrant, 24 September 2015.
10 J. de Boom, A. Weltevrede, S. Rezai, G. Engbersen: Oost-Europeanen in Nederland. Een verkenning van de maatschappelijke positie van migranten uit Oost-Europa en voormalig Joegoslavië, Rotterdam 2008, Risbo, Erasmus Universiteit Rotterdam.
11 T. Hessels: Voormalig Joegoslaven in Nederland. Bevolingstrends, 1e kwartaal 2005, Centraal Bureau voor de Statistiek, 2005.