The turmoil of the COVID-19 crisis put a spotlight on the particularly high economic and social vulnerability of migrants and refugees in Europe. As many severely affected sectors in the economy were characterised by a large proportion of low-skilled and low-paid labour, immigrant workers were disproportionately exposed to the risk of unemployment and earning losses. Poorer working and housing conditions, over-representation in essential frontline jobs and fewer possibilities for remote work, as well as language barriers contributed to significantly higher health risks among migrants. Already existing educational disadvantages and insufficient skill-building opportunities were exacerbated by the interruption of active labour market policies, school closures and a lack of digital literacy (OECD, 2022). Emerging deficits in access to income protection and social security benefits, particularly evident among new immigrants and asylum seekers, forced several European governments to grant, at least temporarily, exceptional rights of access (OECD, 2021).
The challenges of economic and social inclusion of migrants
The shock of the pandemic has worn off, but immigrants remain particularly at risk under changing economic conditions. Due to their generally higher poverty rates, they are disproportionately affected by the real loss of purchasing power resulting from the current inflationary shock combined with the poorly targeted ad hoc transfers of many governments. In the coming years, the ongoing structural transformation towards a digital and green European economy will place higher skill demands on individual workers and accelerate labour market turnover. Since immigrants in many settings are systematically disadvantaged in both school and vocational training, they could suffer rather than benefit from this development, and fall even further behind non-immigrants economically and socially.
Against this backdrop, the population with a migration background in the EU (and in other major destination countries outside Europe) is an important target for social investment that should primarily focus on the development, maintenance and activation of human capital, and aim to promote individual employability and greater resilience in order to achieve economic and social integration (Bonoli, 2020; Hemerijck and Patuzzi, 2021). A systematic reason why agents who settle in a new country for economic reasons need such support is that they misjudge ex ante the expected net returns to settling in the destination country. As crossing borders is associated with uncertainty, immigrants may be exposed to unpredictable idiosyncratic or aggregate shocks after the move that leave lasting scars. Even if the actual net returns to moving thus become negative, they may not return to their country of origin because the migration costs originally invested have sunk in. The sunk costs issue may also come into play when migrants base their migration decision on incomplete information or are overly optimistic, for example, about the real barriers to labour market entry or the transferability of their human capital.
A second important reason is that receiving countries do not have perfect control over arrivals. Because of the free movement of workers in the European single market, it is possible that internal EU migrants who want to benefit from the different labour market developments and economic growth in the member states are on average more likely to have characteristics that hinder economic and social inclusion than natives. This may be the case even if they are positively self-selected in that they are determined to make a career at the destination. The same may apply to family migrants tied to economic migrants, and even more so to the admission of migrants for humanitarian reasons. It is evident that many of the asylum seekers currently reaching Europe do not possess the necessary skills to quickly catch up economically and socially with the receiving population.
In the migration context, the approach of supporting vulnerable people with precautionary measures to promote human capital and employability seems especially relevant, as they face particular risks of falling into a spiral of cumulative disadvantage. Research suggests that unfavourable labour market conditions at the beginning of arrival have a significant negative impact on the subsequent employment trajectory of migrants, as human capital depreciates, and on-the-job learning and the development of professional networks slow down (Ho and Turk-Ariss, 2018). The lasting impact of unemployment or inactivity at an early stage after arrival on future careers at the destination may be particularly pronounced for migrants with a refugee background (Marbach et al., 2018). Incomplete transferability or recognition of human capital from the country of origin, as well as the lack of host-country specific skills, especially language skills, result not only in immigrants having difficulties in pursuing skill-adequate careers. They also reduce opportunities for further education and training. This disadvantage can be long lasting and even affect the educational success of second-generation migrant children, unless mitigated by tracking arrangements in the education system or active integration measures (van de Werfhorst and Heath, 2019).
Positive spillover effects of social investment in migrants
State intervention in the form of social investment to promote immigrant economic and social integration can be justified by a number of potential positive externalities to the benefit of the host society or the incumbent population. The arguments for policies to improve the skills and employability of immigrants are similar to those for selective admission practices to increase the proportion of skilled immigrants.
First, given the increasing pressure on public budgets, especially social budgets, in light of demographic ageing, there are significant external benefits to improving immigrants’ skill levels as better-integrated immigrants make on average higher net payments into public coffers over their life course. Comparative simulation studies indicate that the net fiscal gains of EU member states resulting from labour migration systematically depend on the design of the tax-transfer system in the receiving country and on the characteristics of migrants (Belanger et al., 2020). At the level of the individual immigrant worker, net payments to the host state budget, at a given age, are strongly dependent on the employment rate and wage income earned. This reflects intra-generational redistribution, i.e. the transfer of resources from the relatively rich to the relatively poor via the tax-transfer system. It follows that the expected net tax payments of migrants correlate strongly with their level of education and the speed of their labour market integration.
For example, Bonin (2016), analysing the potential fiscal impact of 2015 humanitarian migration to Germany, estimates that upskilling 20% of the roughly one million refugees to match the fiscal power of workers with completed vocational training, in the long term, would generate €60 billion in additional revenue net of social transfers for the state. Accelerating refugee integration by one year would generate additional net revenue of about €10 billion in total. As long as the state had to spend less than these amounts for active integration measures to achieve these goals, it would be a profitable social investment, which in turn would render public finances more sustainable.
A second source of positive spillover effects of social investment in immigrants’ human capital and employability relates to labour shortages. These existed in many European economies already before the COVID-19 crisis. However, the associated disruption to the reallocation of labour has exacerbated hiring difficulties, contributing to slower economic recovery. In the coming years, the problems faced by employers in filling their vacancies with suitable specialist staff are likely to become even more severe. This is not only due to soaring replacement demand that comes with the retirement of the baby boomers, but also due to the strong dynamics of transformative structural changes that go hand in hand with digitalisation and decarbonisation. With search and matching frictions, a policy of upskilling migrants that increases employers’ chances to hire in tight labour market segments can remove barriers to growth. Furthermore, upskilling migrants can lead to an equilibrium with higher job creation rates and thus better labour market opportunities for non-migrants as well. To the extent that public social investment displaces employer initiatives to support and train migrants to enter shortage occupations, these effects naturally diminish.
A related mechanism by which social investment in immigrants can lead to higher GDP per capita is higher total factor productivity resulting from a boost in innovation or knowledge spillovers. However, the upskilling of initially lower-skilled migrants may affect total factor productivity, as there is less scope for migrants and non-migrants to specialise in tasks that require different skill sets, and non-migrants in simple jobs have less incentive to escape competition with migrants by switching to more complex, complementary jobs. Overall, the effect of upskilling less-skilled immigrants on total factor productivity is therefore theoretically ambiguous. A review of the evidence suggests that the impact of immigration on total factor productivity in receiving countries tends to be positive but rather small (Bonin et al., 2020). This observation presumably also applies to the productivity effects of social investments in the qualification of less-skilled immigrants, provided these do not exceed a plausible magnitude.
The availability of social investments aimed at improving economic and social inclusion provides a form of insurance and conserves private resources. These features could help countries that offer such investments to attract more immigrants, especially younger ones, who can benefit longer from the returns on investment. As a result, the ratio of employed persons to population would rise, and with it GDP per capita. As mentioned above, this effect is amplified if the additional immigrants also help to overcome labour shortages. A possible drawback could be that the skill composition of the newcomers deteriorates. If immigrants can count on active support in training and improving employability in the host country, they will need to bring less transferable human capital from abroad. Due to the insurance motive, immigrants might also be more negatively self-selected or display greater risk aversion. However, such effects could be controlled by setting minimum qualification requirements for economic migrants and strong activation components in the social investment measures used.
Finally, social investment in immigrants can have positive externalities for the host society by promoting pro-social outcomes such as social cohesion, trust and tolerance, or by limiting anti-social outcomes such as crime. Expected returns of this kind at the societal level are a common argument for public investment in education in general and for investment in the education of marginalised groups, such as less-skilled immigrants, in particular. One drawback is that such benefits are systemic and therefore especially difficult to measure. But the other potential social spillover effects of social investment in immigrants described above cannot be precisely quantified either ex ante or ex post. Nonetheless, given the large and diverse groups of asylum seekers that arrived in Europe in 2015-16, policymakers in key European destination countries intuitively adopted a strategy of social investment by allocating massive spending to active integration measures to avoid past mistakes that led to a cumulative disadvantage for less-skilled immigrants. Sweden, for example, spent more than 1% of GDP on refugee reception and upfront support in 2015, while Germany spent about 0.5% of GDP, and high levels of spending continued in the following years (Hemerijck and Patuzzi, 2021). The recent paradigm shift allows to study the impact of social investments targeting vulnerable immigrants from the outset.
Active labour market policies as a profitable social investment
In Germany, a comprehensive evaluation of all integration measures taken within the framework of active labour market policy is still ongoing. Interim estimates based on a control group approach indicate that for migrants with a refugee background who arrived from 2015 onwards, most of these measures lead to significantly better labour market integration and less welfare dependency within the first 40 months after the start of training (Table 1). An exception is the provision of public job opportunities, which are used by the employment agencies as a last measure for a small part of the hard-to-place persons and do not function as a stepping stone. Otherwise, the employment rates of participants develop more favourably compared to refugees who do not receive treatment. This is true for small-scale interventions such as activation and vocational orientation, which usually start when refugees get access to the labour market; for a range of training and (re)qualification measures of different scope and having different target groups; and for wage subsidies, which give employers incentives to hire immigrants despite initial productivity disadvantages and necessary additional in-company training. The differences brought about by these active labour market policy measures appear to be long lasting.
The resulting positive treatment effects are a necessary but not a sufficient condition for spending on active integration measures to be a profitable social investment. In fact, the cumulative average value of additional net public revenues (taxes and social contributions plus savings in social transfer expenditure) calculated over a period of only 40 months is still not always sufficient to cover the upfront costs of treatment. This concerns treatments where the long-term investment character is particularly pronounced because they focus on the development and adaptation of human capital. However, projecting the gains achieved over a further 20 months shows that all benefit-cost balances turn positive. Since immigrants on average have many more years to spend in the labour market, the ultimate expected fiscal returns of active integration measures in this environment appear very high.
Table 1
Estimated treatment effects and fiscal benefit-cost analysis of active labour market integration measures for refugees in Germany
| Estimated effects over 40 months from start of treatment | Forecast over 60 months in euros | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Integration effects | Fiscal effects in euros | ||||||||
| Months in employment |
Days on social transfers |
Taxes and contributions | Social transfers |
Average cost per treatment | Benefit-cost | Benefit- cost |
|||
| Activation and professional orientation | |||||||||
| with an employer | 9.2 | -68 | 2818 | -1683 | 635 | 3866 | 5105 | ||
| with a provider | 1.9 | -19 | 726 | -466 | 635 | 557 | 815 | ||
| Career choice and vocational training | |||||||||
| entry level qualification | 17.0 | 3 | 600 | 67 | 2331 | -1797 | 473 | ||
| assistance during training | 3.2 | 13 | 1149 | 324 | 2697 | -1872 | 2783 | ||
| Continuing professional education | 4.0 | -43 | 2571 | -1072 | 5738 | -2095 | 332 | ||
| Integration grants to employers | 13.2 | -232 | 9781 | -5903 | 4330 | 11354 | 17612 | ||
| Public work opportunities | -0.5 | 7 | 29 | 179 | 1683 | -1833 | -1736 | ||
Notes: Results refer to the population group refugees who arrived in Germany as of 1 January 2015 and started a treatment between August 2017 and September 2018.
Source: Own adaptation of results by Bonin et al. (2021).
Several factors that were important to success in the German case are also increasingly present in the (re)design of integration policies for humanitarian migrants in other European countries.
Early intervention. Active measures should reach migrants with a high likelihood of staying and characteristics that make economic and social integration significantly more difficult as early as possible. The experience in the initial phase after arrival in a new country can be crucial in determining whether migrants embark on a positive or negative integration path. From a dynamic perspective, postponing profitable social investments means that their potential benefits are not fully realised.
Accessibility. Of course, active labour market integration measures can only be successful if the targeted persons are actually allowed to work. Their more intensive use therefore demands and in fact often goes hand in hand with the removal of formal barriers to labour market access for migrants. Formal access to the arsenal of integration measures also plays a role. Immigrants are usually initially systematically excluded from the benefits of the contribution and insurance systems. This argues for the provision of active integration measures through welfare channels. Finally, formally eligible persons may be denied access due to positive or negative biases in the selection of programme participants (e.g. skimming) or even discrimination (e.g. application of gender stereotypes in programme design or selection). The latter could contribute, for example, to the underrepresentation of female refugees in active labour market policies, despite the fact that they would especially benefit from them, which has been observed in practice (Bonin et al., 2021).
Integrated services. As vulnerable immigrants are often disadvantaged in several areas at the same time, one success factor is to design combination treatments that address these areas in a coherent manner. Such combinations, like the provision of childcare services in conjunction with vocational training for women, may be necessary to encourage voluntary participation. But combining different elements, such as job-related language support in conjunction with in-service training, can also help to reinforce effectiveness. Combination treatments may also consist of planning logical sequences of individual measures. For example, basic language training would lay the groundwork for subsequent qualification measures.
Private engagement. The quality of progress in social and economic inclusion of immigrants is not determined by public authorities alone. It also depends on the commitment of businesses and civil society, which are essential parts of local integration ecosystems. Public integration measures that embed active private engagement therefore tend to achieve better results. Directly involving employers in active inclusion measures in real workplaces, e.g. by publicly assisted on-the-job vocational training, is proving to be a particularly profitable approach.
Sufficient resources. The implementation of profitable investment in integration measures asks for considerable resources at the start. These need to be stable enough to ensure continuity in well-functioning measures, pursue long-term strategies and allow gains from experience to unfold. It is also important to have high quality case workers who perform the key task of matching immigrants with the integration measures that are the most appropriate in the individual case. This requires experience and an excellent understanding of the specific needs within heterogeneous migrant target groups.
Conclusions
The principles set out above are also relevant when considering social investments in disadvantaged and vulnerable people with a migrant background who have lived in the host country for a long time or who may even have grown up there as descendants of immigrants – save for early intervention, of course. The latter is a massive issue, as the force of dynamic cumulative disadvantage is very difficult to revert once it is unfolded. Active public support may then primarily promote social stabilisation rather than economic well-being and is often unlikely to yield a positive return as social investment, at least in monetary terms.
Therefore, having to decide under budget constraints, the government may give preference to active inclusion policies that target newly arrived migrants in need of support. Many governments in Europe (if not being plain unwilling) are struggling to put this complex and challenging task into practice with another sharp rise in the number of asylum seekers in sight. Budgetary room for manoeuvre has shrunk, educational and social support capacities have been cut back, and parts of the administrations in charge suffer from labour shortages. This could lead to foregoing the substantial long-term social and economic returns that are possible if social investment in immigrant inclusion and integration is done well.
References
Belanger, A., M. Christl, A. Conte, J. Mazza and E. Narazani (2020), Projecting the net fiscal impact of immigration in the EU, Publications Office of the European Union.
Bonin, H. (2016), Gewinne der Integration: Berufliche Qualifikation und Integrationstempo entscheiden über die langfristigen fiskalischen Kosten der Aufnahme Geflüchteter, böll.brief – Teilhabegesellschaft, 1.
Bonin, H., B. Boockmann, T. Brändle, J. Bredtmann, M. Brussig, G. Demir, R. Kamb, H. Frings, A. Glemser, A. Haas, L. S. Höckel, S. Huber, A. Kirchmann, J. Kirsch, G. Klee, A. Krause-Pilatus, J. Kühn, P. Kugler, M. Kusche, A. Maier, U. Rinne, A. Rossen, T. Scheu, K. Schilling, E. Teichert, R. Wapler, K. Wolf and A. Zühlke (2021), Begleitevaluation der arbeitsmarktpolitischen Integrationsmaßnahmen für Geflüchtete, IZA Research Report, 123.
Bonin, H., A. Krause-Pilatus, U. Rinne and H. Brücker (2020), Economic Impact of Intra-EU Labour Mobility in Receiving and Sending Countries – A Brief Survey, IZA Research Report, 103.
Bonoli, G. (2020), Immigrant Integration and Social Investment, Journal of European Social Policy, 30(5), 616-627.
Hamerijck, A. and L. Patuzzi (2021), ‘Diversifying’ Social Investment: European Welfare States and Immigrant Integration in the Wake of the COVID-19 Crisis, Migration Policy Institute Europe.
Ho, G. and R. Turk-Ariss (2018), The Labor Market Integration of Migrants in Europe: New Evidence from Micro Data, IMF Working Paper, 18/232.
Marbach, M., J. Hainmueller and D. Hangartner (2018), The Long-Term Impact of Employment Bans on the Economic Integration of Refugees, Science Advances, 4(9).
OECD (2021), Labour Market Outcomes of Immigrants and Integration Policies in OECD Countries, in International Migration Outlook 2020, Chapter 2, OECD Publishing.
OECD (2022, 30 August), What has been the Impact of the COVID-19 Pandemic on Immigrants? An Update on Recent Evidence, OECD Policy Responses to Coronavirus (COVID-19), https://www.oecd.org/coronavirus/policy-responses/what-has-been-the-impact-of-the-covid-19-pandemic-on-immigrants-an-update-on-recent-evidence-65cfc31c/ (10 September 2023).
van de Werfhorst, H. G. and A. Heath (2019), Selectivity of Migration and the Educational Disadvantages of Second-Generation Immigrants in Ten Host Societies, European Journal of Population, 35(2), 347-378.
