Both the experience from previous economic crises and first indications on labour market and social outcomes during the current pandemic suggest that the COVID‑19 crisis is likely to have a disproportionate impact on immigrants and their children. This policy brief provides first evidence on how the pandemic has affected immigrants and their children in terms of health, jobs, education, language training and other integration measures, and public opinion, and describes host countries’ policy responses. It complements a previous brief on the impact of the pandemic on migration management.
Due to a range of vulnerabilities such as higher incidence of poverty, overcrowded housing conditions, and high concentration in jobs where physical distancing is difficult, immigrants are at a much higher risk of COVID‑19 infection than the native-born. Studies in a number of OECD countries found an infection risk that is at least twice as high as that of the native-born.
COVID-related mortality rates for immigrants could also be significant, exceeding those of the native-born population.
Immigrants are potentially in a more vulnerable position in the labour market due to their generally less stable employment conditions and lower seniority on the job. Studies also suggest that discrimination strongly increases in times of a slack labour market, while networks of contacts – of which migrants have fewer – become more relevant for finding a job.
The negative impact on immigrants’ labour market outcomes is increased still further by the fact that they are strongly overrepresented in those sectors most affected by the pandemic to date. For example, in the particularly hard-hit hospitality industry, a quarter of employees in the EU are foreign-born, twice their share in overall employment.
It is still early to gauge the labour market effects of the pandemic – especially in European OECD countries, where job retention schemes have cushioned the immediate impact of the lockdowns. That notwithstanding, the available evidence on the initial impact shows a disproportionately negative toll on immigrants in the vast majority of countries for which data are currently available, especially in the Southern European countries, Ireland, Norway, Sweden and the United States.
The school closures and distance learning measures put in place to slow the spread of COVID‑19 put children of immigrants at a disadvantage, in several ways. Their parents tend to have fewer resources than native-born parents to help them in their homework, and 40% of native-born children of immigrants do not speak the host-country language at home. Such children are also less likely than students with native-born parents to have access to a computer and an internet connection at home or to a quiet place for study.
The pandemic gave a push for remote language learning for adults as well. A number of countries introduced innovative new schemes. In Germany, for example, online tutorials were set up to compensate for the temporary closure of immigrant integration courses. However, such online learning has proved difficult for low-educated immigrants, especially at early stages of language learning, leading to delays in both language learning and broader social integration.
In light of growing unemployment and the role of international travel in the initial spread of the pandemic, there is a risk of a backlash in public opinion against immigrants. A number of communication campaigns have aimed at addressing this issue, with a particular focus on tackling misinformation regarding the role of immigrants in the spread of the virus.
The COVID‑19 pandemic comes at a crucial time for international migration. Just prior to the crisis, record-high inflows were recorded in a number of countries, and populations of immigrants and native-born children of immigrants have grown virtually everywhere. They now account for one in five people across the OECD. Among the many recent arrivals, refugees account for a relatively high share in some countries and this group faces particular vulnerabilities and has specific needs.
At the same time, integration had improved on many fronts prior to the pandemic. In all OECD countries except Turkey and Colombia, immigrants were more successful in finding and keeping jobs over the past five years, although in most countries the gap with native-born did not close a lot. Likewise, attitudes towards immigrants had seen improvements in most countries over the past decade. Educational outcomes of children of immigrants had improved since 2005, with higher educational attainment levels and outcomes – in both absolute and relative terms – narrowing the previous gap between themselves and peers with native-born parents. This progress is now put at risk because of the pandemic. The COVID‑19 crisis is affecting the health, education and jobs of already vulnerable immigrants and their children, as well as their broader social integration.
This policy brief sets out to assess the likely impact of the pandemic in each of these areas, along with first evidence on outcomes and policy measures taken by countries to attenuate negative impact. It concludes with an outlook on the longer-term consequences of COVID‑19 on immigrants, along with policy recommendations.
Immigrants are younger on average than the native-born population, so theoretically less susceptible to develop serious health effects from COVID‑19. In European OECD countries, for example, 8% of the foreign-born are over 75 years old, against 12% of their native peers. However, it is a well-established fact that groups of population with socio-economic disadvantages are more likely to have bad health conditions and suffer more from chronic diseases, which can increase the risk of co-morbidity in the context of COVID‑19. Immigrants in most OECD countries are overrepresented among these vulnerable groups.
Around 30% of immigrants live in relative poverty in the OECD, compared with 20% of the native-born. They are also more likely to live in sub-standard accommodations (23% against 19%) and twice as likely in overcrowded dwellings (17% against 8%). Poor housing conditions increase the likeliness of infection, especially given that immigrants are more likely to live in extended cohabiting families.
Alongside poor housing conditions, immigrants are also more likely to live in higher density buildings and neighbourhoods, which makes the respect of social distancing more difficult. This is particularly the case for certain groups who live together in collective housing, such as asylum seekers. As an example, University of Bielefeld found that, compared with other forms of housing, collective housing for asylum seekers and refugees increased the risk of COVID‑19 transmission in case of a first positive diagnostic by 17%. Immigrants are also a group with a higher use of public transportation , which has become a risk in times of pandemic.
In addition, immigrants are concentrated in essential occupations that cannot be undertaken from home. They are less likely to be able to telework: in three‑quarters of OECD countries, the share of immigrants able to telework is at least 5 percentage points below that of their native counterparts (Figure 1).
For example, immigrants account for more than half of all domestic services workers in Southern European countries, Israel and Canada . Not only do immigrants have to go to work, in some specific sectors they also have to deal with difficult and unsafe working conditions with respect to the COVID‑19 transmission. In Germany, for instance, a coronavirus outbreak in a slaughterhouse spread to more than 1 500 employees, the vast majority of whom were EU migrants, which triggered a local lockdown.
A number of immigrant-specific factors push some specific groups at higher risk of COVID‑19 transmission. Immigrants in an irregular situation, for example, may be less inclined to go for a test or to hospitals. Migrants, notably temporary migrants, often tend to have less extensive coverage by social security systems, including health. What is more, the lack of host-country language proficiency for some immigrants may hamper access to information on COVID‑19. Although many countries provide access to information on health care services and good practices, usually accessible in several languages through multiple information channels , reaching the most vulnerable groups remain challenging. Immigrants also often have a smaller support network to rely upon in case of a COVID‑19 infection.
For some groups such as refugees, the poor conditions in origin and transit countries, as well as health issues related to their (forced) migration, also enhance the risk of COVID‑19 comorbidities. A report by the WHO regional office for Europe that summarised the available evidence found, for example, a higher risk of ischaemic heart disease and stroke among the refugee and migrant population in Europe. What is more, refugees and migrants in Europe have a higher incidence, prevalence and mortality rate for diabetes than the host population, with higher rates among women than men, though this depends on the country of origin. At the same time, migrants tend to have a lower risk for most neoplasms.
Assessing to what extent COVID‑19 has affected the immigrant population is challenging. Basic statistics on the number and characteristics of COVID‑19 cases are registered by national health systems, built on official statistics from many sources, such as on testing and hospitalisations. What is more, not many countries ask for immigrant status in these registered data, neither with respect to country of birth nor nationality. In some countries where this information is not available, such as the United Kingdom, New Zealand and the United States, ethnic origin or race are asked, using national categories, which partly cover the immigrant population alongside native-born with ethnic minority background. In a number of countries, none of the above is registered.
The few data available by origin usually show a significant over-representation of immigrants in the incidence of COVID‑19. Among the confirmed cases in Norway, 31% are foreign-born (mainly from countries with a high share of humanitarian emigrants), almost twice as much as their share among the population. The trend during the peak of the pandemic (13 March to 7 May) was the same in Sweden, with 32% of cases being immigrants (against 19% in the population) as well as in Denmark, where immigrants from lower-income countries and their native-born children account for 18% of the infected, twice as many as their share in the Danish population. According to the Estudo Instituto Saude Publica in Portugal, 24% of COVID‑19 infections in Lisbon are immigrants (mainly from Africa); while foreign-born represent around 11% of the population of the Capital-City Metropolitan area. Outside Europe, immigrants are also disproportionately affected by COVID‑19. For instance, in Ontario, Canada, permanent immigrants accounted for 43.5% of cases, while they account for 25% of its population.
Some countries, however, did not register a higher incidence. In Italy, for example, which has been disproportionately hit by the first wave of the pandemic, data from the Istituto Superiore di Sanità showed that only 5% of cases concerned foreigners, only about half of their share in the population.
While these conflicting findings partly reflect differences in demographic structure and the different ways in which the pandemic hit across countries, they also challenge the reliability of the number of registered COVID‑19 cases by origin to fully grasp the impact of the pandemic on the immigrant and minority population. There are actually many caveats to consider when using the overall number of cases. First, it does not take into account the different demographic composition of both groups, which is an important bias, especially in countries where migration is a more recent phenomenon and immigrants are younger on average. In Italy, for example, a significant part of the foreign population consists of young migrants and native-born children of immigrants, who are less likely to show COVID‑19 symptoms. What is more, the number of confirmed cases is driven by the national testing strategy. Limited testing capacities in some countries in the early phases of the pandemic hampered large-scale population testing.
In addition to the broad overall testing capacity, the number of confirmed cases by origin is also impacted by the ability of each country to reach the most vulnerable groups, among them immigrants and their specific issues (e.g. lack of host-country language proficiency, legal status). Rates of testing were lower for immigrants compared with native-born in Ontario for instance.
Other measures are thus needed to better grasp the impact of COVID‑19 on the immigrant population. A priori, the number of deaths from COVID‑19 by origin tends to be a better measurement of the impact of COVID‑19 on this population, although the country of origin is not necessarily mentioned on those certificates or the information is not readily available. In Germany, for example, while nationality is in principle registered in the death registers, this data is currently not transmitted for the compilation of the national mortality statistics. To be reliable, COVID‑19 death statistics must be based on death certificates that include at least the cause of death and other key sociodemographic variables, especially age, which dramatically influences mortality for the disease.
A number of OECD countries do not have data on mortality for immigrants, but for ethnic minorities. This notably concerns the United Kingdom and the United States. The composition of the two groups differs but is partly overlapping. In the United Kingdom, for example, among the persons who self-identify as Black Asian and Minority Ethnic, 69% are immigrants. In the United States, among the Hispanic ethnic minority, almost half (46%) are foreign-born and a further 24% are native-born with immigrant parents. Among the Asian ethnic minority, the respective figures are even 73% and 20%, respectively.
The available data show that ethnic minorities have paid a higher price than the majority population concerning COVID‑19. In the United Kingdom, one third of patients critically ill were registered as BAME (Black Asian and Minority Ethnic) as of 7 September 2020 , while they made up only 22% of the British population at the time of the last census (2011). Taking into account the size, gender and age structure of the population, data from March to mid-May by the Office for National Statistics (ONS) show that the age-adjusted mortality rate from COVID‑19 was 0.26% for Black males, 0.19% for Bangladeshi/Pakistani, 0.16% for Indians, and 0.09% for Whites. Age-adjusted mortality rates were even 3.3 times higher for Black men compared to their white peers; the corresponding figure was 2.4 for black women. Taking into account further factors, such as living conditions and socio-economic variables, the risk of COVID‑19 death remained twice as large for black men and 1.4 for black women as well as 1.5 times greater for South Asian men, compared with their white peers. Studies found that the greater risk of severe COVID‑19 diseases and COVID‑19 deaths in BAME populations can only be partly explained by higher risk factors for mortality (such as cardio metabolic and vitamin D status), or by other patient care or socio-economic characteristics (such as social deprivation and frailty, see Apea et al. and Raisi-Estabragh et al. ).
In the United States, the Center for Disease Control and Prevention (CDC) estimates death statistics standardised by age and accounting for differences in the geographic distribution. 41% of all COVID‑19 adjusted deaths involved Hispanic people. The expected percentage based on their population share – again adjusted for age and geographical distribution – was 33%.
Higher mortality from COVID‑19 compared with their white peers affected all groups among the black population but only the working age population among Hispanics. Minorities in the United States are not only affected by higher COVID‑19 mortality overall, this also holds after controlling for pre-existing health conditions and other demographic and socio-economic factors. Bertocchi and Dimico estimated that the probability of dying remains higher for ethnic minorities in so-called “lower-graded” (run-down) neighbourhoods.
A lack of death certificate data disaggregated by origin and other socio-demographic characteristics does not necessarily imply that assessing the impact of the COVID‑19 on the immigrant population is impossible. France, the Netherlands and Sweden, for example, which have recent mortality data by origin but not by cause of death, observed an uneven impact of mortality by country of birth (Figure 2).
In France, between March and April 2020, excess mortality – the difference in mortality compared with the same period in 2019 – among the foreign-born was twice that of native-born . The immigrant groups which were hit the worst by the excess mortality were from North Africa (+54% deaths compared to the same period in 2019), sub-Saharan Africa (+114%) and Asia (+91%), compared with 22% excess mortality for the native-born. Higher excess mortality for immigrants was even observed among the youngest cohorts. Immigrants’ excess mortality remained twice to four times as much as that for the native-born, even after taking into account that foreign-born are more likely to live in densely populated areas that were more affected by the pandemic.
In Sweden, the share of foreign-born among deaths, which varied between 12 and 14% over the 2015‑19 period, reached 16% in March-April 2020. The number of deaths among persons aged 40 years and more who were born in countries from which many refugees have migrated to Sweden in the last decades (Syria, Iraq and Somalia) was 220% higher in March-May 2020 compared to the average in 2016‑19. In contrast, the respective increase during these three months was only 18% for those born in Sweden, EU or North America, despite an older age composition.
In the Netherlands, death statistics by parental place of birth for March and April 2020 show that deaths were 47% higher than usual for immigrants from lower-income countries and their immediate children, 49% higher for immigrants from high-income countries and their descendants, and 38% higher for native-born with Dutch parents.
In summary, immigrants have paid a higher toll with respect to the incidence of COVID‑19, with higher infection risk and higher mortality, despite having a younger age on average. In the longer term, further research is needed to estimate the long-term effect on the overall health of immigrants, including on their mental health. To address these questions, OECD countries must also try to address the lack of data on COVID‑19 incidence and outcomes.
Most OECD countries offer access to testing and to emergency health care for migrants, including those in an irregular situation, if they contracted COVID‑19. Free access to necessary treatment related to COVID‑19 irrespective of status is, for example, possible in Belgium, Chile, Finland, France, Germany, Hungary, Israel, Luxembourg, Mexico, Portugal, Spain and Switzerland. Some countries restrict this access to certain health services or make it subject to conditions. Portugal temporarily regularised migrants in an irregular situation to ensure full access to the health care system. Similarly, Spain suspended the obligation to have valid documents in order to continue receiving aid covering basic needs. In Greece, access is available for minors, but only in case of emergency for adult immigrants. In the Czech Republic, migrants in an irregular situation might have to reimburse their treatment later. In some countries, such as Australia and Canada, regional authorities handled health care access matters, resulting in free full access across Canada and free access in some states in Australia.
A number of countries also launched specific information campaigns for migrants (Box 2; see also OECD ).
In the course of the COVID‑19 pandemic, governments needed to adapt their communication on migrant integration to set of additional objectives. In order to limit the spread of the virus, governments needed to provide immigrants with timely and accurate information on the pandemic itself, public health measures as well as access to medical services.
In many OECD countries, governments make use of digital communication channels to communicate with migrants in the current crisis context – combining the advantages of both dedicated online websites as well as communication via social media channels (see OECD for a comprehensive overview). Newly established multilingual websites allow providing one place for both migrants and host societies to access reliable and up-to-date information related to COVID‑19. They often include innovative visual tools such as the mapping of regional confinement measures and safety regulations, including COVID‑19 testing. Additionally, social media play a central role in advertising these online platforms and their content, as they allow directly targeting specific groups of migrants in their mother languages and responding to their questions. Social media have also proven to be the most important communication channel to track and respond to circulating misinformation.
In Germany, the Federal Government has implemented a digital campaign combining both web and social media platforms in multiple languages in order to communicate on COVID‑19. In the scope of the current crisis, it has been providing information related to COVID‑19 in all relevant languages on both its own website as well as the dedicated online platform of “Handbook Germany”. This information has been further promoted and circulated through a Facebook page with over 200.000 followers for migrants living in Germany, presenting information in the form of articles and videos in the relevant languages. The platform actively answers questions by users and responds to comments containing misleading information on the topics concerned.
Similarly, Canada’s Immigration, Refugees and Citizenship service adapted the content provided on its website and social media channels to the COVID‑19 crisis. In addition to resources of the country’s Public Health Agency translated in multiple languages, the service provides information on changes in travel restrictions and application processes for each specific group of migrants, such as workers, students and asylum claimants. The website also hosts a live chat to answer individual questions. This allows users to easily access the information relevant to their personal situation.
In addition to online communication, intermediaries such as local authorities and NGOs play an important role in increasing the outreach of information provided by national governments. Local intermediaries are able to directly reach out to immigrants living in their local communities, including those who are particularly isolated and hard to reach through online communication channels. For example, in collaboration with national and local authorities in the United Kingdom, the NGO Migration Yorkshire developed a Migrant Information Hub for the Yorkshire and Humberside region. This platform is dedicated to informing migrants about the COVID‑19 pandemic, access to public services and the status of their residency rights. Furthermore, the Migrant Hub provides multilingual videos and organises free workshops for migrants on topics such as “How to access health services in the UK” to inform about how to access local services.
Partnerships between governments and NGOs to manage the immediate translation of communication content facilitate the implementation of multilingual communication campaigns related to COVID‑19. An example of such a partnership is the collaboration of public authorities in Denmark with the Danish Refugee Council (DRC) in order to enable them to establish a website and phone hotline dedicated to answer questions related to COVID‑19 in 25 languages. Likewise, in Norway, the government has set aside 20 million NOK for NGOs to spread information on COVID‑19 to migrant communities.
Immigrants face a number of particular vulnerabilities in the current labour market situation. First, immigrants tend to be overrepresented among employees with temporary contracts, especially in European and Asian OECD countries. Immigrant also tend to have lower seniority and are overrepresented in cyclical sectors (see OECD for a discussion). Permit conditions may also enhance their vulnerability in terms of a job loss, if they do not allow for change of employers or professions. Even where this is possible, the perceived additional administrative burden or uncertainty may prevent employers from considering migrants in a context of a relative abundance of workers.
More generally, employers tend to be more selective during slack labour market conditions, and characteristics such as language difficulties, which may hamper productivity, are used to screen out immigrant applicants. Even where such factors are lacking, a number of studies have shown that employers tend to selectively omit immigrants during an economic downturn . Migrants have also fewer networks, and the importance of such networks tends to increase when labour market conditions worsen .
While the above-mentioned factors have been observed in a number of downturns such as the global financial and economic crisis of 2008, there are also several factors that specifically enhance migrants’ labour market vulnerability in the current situation. First, the outbreak of the COVID‑19 pandemic occurred following record-high inflows of immigrants to OECD countries, among which many refugees. New arrivals tend to be particularly hard hit during the crisis, with lasting negative impact on their long-term employment prospects. This holds true especially for those who did not yet managed to find employment, which is still the case for many refugees who arrived during the 2015/16 refugee crisis. In Sweden, an analysis of the impact of the recession of the early 1990s, which followed a period of high refugee inflows, showed a disproportionate negative impact of the recession on recent arrivals. This effect was strong and lasting – with much lower chances of being employed even 10 years after the crisis; see also Aydemir .
Second, the pandemic particularly affected employment in a number of services sectors where immigrants are largely overrepresented. This is particularly the case for hospitality and security and cleaning services, which have been hard hit during the lockdown (Table 1). For example in the EU, immigrants account for a full quarter of employment in the hospitality industry, twice their share in overall employment. In Austria, Finland, Germany, Ireland, Luxembourg, Norway, Switzerland and Sweden, they even account for at least 40% of employment in this sector (Figure 3). The likely impact on migrants is further heightened by the fact that hospitality is the sector where new arrivals are most strongly overrepresented when compared with long-term settled migrants .
Security and cleaning services
Total (all sectors)
Source: OECD Secretariat calculations based on labour force surveys.
Not only migrant employees, but also the self-employed tend to find themselves among the most vulnerable in the current situation. Across the OECD, migrants tend to be overrepresented among the self-employed, albeit only slightly . Their businesses also tend to be smaller and to have a lower capital stock. What is more, as is also the case among workers, many immigrant businesses find themselves in the hard-hit hospitality sector. Indeed, in Canada, Denmark, Germany, Luxembourg, Norway and Sweden more than 40% of all self-employed in this sector are immigrants (Figure 3).
While most OECD countries are gradually re-opening their economy after the often strict confinement phase at the peak of the COVID‑19 crisis, the impact on the labour market is likely to deepen significantly going forward. Changes in countries’ unemployment rates since the onset of the COVID‑19 crisis have varied widely, reflecting fundamental differences in policy responses but also the complexity of collecting and comparing labour market statistics in times of a pandemic (see for a comprehensive discussion and overview).
Given this, and the uncertainty over the situation in the near future, it is still early to assess the impact on immigrants, but the available evidence clearly suggests a disproportionate impact in most countries. Across the OECD, employment rates were higher among foreign- than native-born in almost two‑thirds of countries. Immigrants were particularly affected in Southern European countries, Ireland and Austria, where employment rates decreased by at least 4 percentage points, at least twice as much as for their native counterparts (Figure 4).
Among the exceptions with respect to the declining employment rates for immigrants are the United Kingdom, the Netherlands, Finland and the Czech Republic, where no negative impact was observed thus far on the aggregate.
The overall decline in employment rates – for both immigrants and native-born – tended to be more marked outside of Europe. At the same time, there was a much stronger volatility in these countries throughout the pandemic thus far.
The effect on employment may be associated with higher unemployment, higher inactivity or both. Indeed, it is conceivable that migrants who lost their jobs were discouraged by the difficult labour market situation and that those who were looking for a job before the pandemic stopped searching or were simply not available for a job (because of the lockdown or because of family responsibilities due to schools closing). This removed them from the labour force according to the ILO definition. In countries where large job retention schemes have been put place and where temporary employment is less frequent, most of the effect may result in a rise in inactivity. The latter may even be accompanied by a decrease in unemployment. In other cases, increases in the unemployment rate may absorb most of the effect.
In the United States, while overall unemployment reached at its peak 14.4% in April, it dropped to 8.5% in August 2020. Prior to the pandemic, the unemployment rate for immigrants was lower than that of the native-born, but it is now 2 percentage points higher than the unemployment rate for native-born. Part of the reason for the stronger overall unemployment reaction in the United States, and also Canada, is because retention schemes are not prevalent like in many European countries . To which degree immigrants are covered by these schemes is unfortunately not known. Nonetheless, the initial labour market figures suggest that they seem to have a less stabilising effect for immigrants.
That being said, most OECD countries experienced increases in the unemployment rate for both native-born and immigrants, with a much larger increase for the latter. Figure 5 provides an overview of the percentage point increase in unemployment rates comparing the most recent data from 2020 with the corresponding data from the same period of 2019. Among the countries for which data is available, in Norway, the United States, Canada, Sweden and Spain, immigrants experienced an increase of more than 4 percentage points in the unemployment rate. Among native-born, this was only the case in Canada and the United States.
The disproportionately high incidence on immigrants is especially notable in Sweden, where 58% among the newly unemployed since August 2019 are immigrants. In Austria, Norway and Germany, more than a third of the newly unemployed are immigrants or foreign nationals.
In Greece, the changes in unemployment rates for immigrants and native-born even go in different directions. While the unemployment rate for immigrants increased by more than 2 percentage points, the unemployment rate declined slightly for the native-born.
The situation is also continuously evolving. This is evident for example in Denmark, where figures on the changes in unemployment since the beginning of March 2020 show that initially, there was a stronger increase among native-born. However, over time, the situation changed and since August, a higher incidence on immigrants has been observed.
In a context of declining activity, increases in unemployment must be interpreted with caution, as they can reflect the higher weight of the unemployed under lower employment figures. While for most countries, increases in unemployment were associated with declines in employment rates, there are a few notable exceptions. These include Belgium, France and Italy, where declines in employment were not mirrored in higher unemployment but rather by higher inactivity rates.
These developments may partly reflect shortcomings in using labour force survey data in the early days of the pandemic – indeed, administrative data show a more consistent picture. In Belgium, for example, administrative data relating to temporary unemployment claims by nationality show an over-representation of workers of foreign nationality in temporary unemployment claims for the entire period March-August 2020. The over-representation increased from +5.9 percentage points in March to +8.7 percentage points in August. Likewise, in the figures on final payments to people who were temporarily unemployed for at least one day (available until July), people of foreign nationality are over-represented by 7.5 percentage points compared to their share among employees.
The impact on specific groups has thus far not been uniform. For example, the increase in unemployment rates for EU nationals in Germany and in Austria was about twice as high as the increase in unemployment rates for nationals, but still lower than for other migrant groups. In Spain, however, the unemployment rate for EU citizens jumped from 14.8% to 20.5% – an even higher increase than for non-EU foreigners. In Norway, immigrant women from Central and Eastern Europe were the hardest-hit group thus far – with an increase in unemployment by more than 10 percentage points. As for women, data from Germany also suggest a disproportionate impact
Likewise, data on the initial impact on refugees show divergent trends. In Germany and Austria, the unemployment rate for nationals from the main origin countries of refugees was above 30% in 2019 and increased rapidly by more than 5 percentage points in 2020. This increase is more than three times higher than the unemployment increases for nationals in Austria and Germany. In both countries, the unemployment rate for refugees now exceeds 40%. In Sweden, in contrast, first data from the public employment service on the labour market status of persons 90 days after the finalisation of the introduction programme for refugees and their family members do not show higher incidences of unemployment or non-employment. This somewhat surprising result, which stands in contrast with the observed higher incidence of unemployment as a whole, may be attributable to the fact that the Swedish authorities prolonged the duration of the various wage subsidy schemes by an additional year. This is a measure that disproportionately benefits recent refugees.
Longitudinal data with disaggregation by country of origin for the United Kingdom show a higher job loss of BAME immigrants during the lockdown – while this was not observed for immigrants who self-declared to be of “white” origin. Interestingly, while the former group had a higher probability to lose their job, they appeared to be slightly less likely to be furloughed. This means that BAME immigrants were more likely to have lost their jobs permanently instead of being granted leave of absence.
Data on the impact on native-born descendants of immigrants is currently only available for Norway. In that county, the increase in the unemployment rate among descendants was, at 7.5 percentage points, even slightly higher than for immigrants themselves; indeed considering their lower initial levels of unemployment (3.4% in Q2‑2019), the relative impact was even stronger.
To date, there is little evidence on the impact by education level, but the scarce evidence suggests that the crisis predominantly affect lesser-skilled immigrants. In Italy, for example, foreign nationals without educational credentials experienced a decline in the employment rate of more than 10 percentage points – twice the decline of those with at least an upper secondary degree. Likewise, the observed increase in unemployment in the United States was twice as large among immigrants with at most a high-school degree than for their peers with higher degrees. In addition, differences by education level were less pronounced for the native-born.
Clearly, the outlook for immigrants’ labour market integration will largely depend on the evolving overall economic conditions, including with respect to the impact on specific sectors – and here the outlook is not good, considering immigrants’ strong overrepresentation in those services sectors most likely to be hardest-hit. First indicators on vacancies already suggest that the negative impact on immigrants is likely to magnify disproportionately. In the Netherlands, for example, where no negative impact on the employment or unemployment of immigrants has yet been observed, the Central Planning Bureau calculated the probability of being in a profession that observed a drop in vacancies of more than 20% combined with the probability of having a temporary contract. Compared with those with native-born parents, immigrants from high-income OECD countries and their descendants are 21% more like to be in this group, and other immigrants and their descendants are even 34% more likely to find themselves in this group.
a. Facilitating stay in case of unemployment and reduced income
Migrant workers who lose their jobs often struggle to comply with the conditions of their residency permits. In response, many countries have extended permits or removed obligations to leave, to prevent legally staying migrants from falling into an irregular situation. Spain, Greece, the Czech Republic and Germany, for example, did not withdraw permits for migrants who lost their job during the pandemic. Other countries including France, Slovenia, Estonia, Italy, Ireland, Poland, and Portugal automatically extended or renewed permits for a certain period, in some cases until after the end of the state of emergency, in other cases until a pre-defined calendar date. In many OECD countries, changes introduced allowed for overstay on a temporary visa, without any negative consequences for future visa applications;. In countries where the validity of certain work permits is conditioned on reaching a certain level of income, for example in Australia, Austria and the United Kingdom, additional flexibility was introduced. A number of countries, including the Netherlands and Ireland, introduced policies to prolong the post-graduation stay of international students seeking employment .
b. Extending coverage of support measures
A number of OECD countries modified their access to, and the duration of, unemployment benefits, with some changes likely to benefit migrants in particular as these often have less stable contracts and a lower contribution history. Belgium, for example, included the COVID‑19 pandemic within its “temporary unemployment due to force majeure” regime. As a result, workers do not need to prove a sufficient number of days as an employed worker in order to receive support. Similarly, in Spain, the minimum duration of work required to qualify for unemployment benefits, 360 working days during the last 6 years, was suspended. In Ireland, any person who lost their job due to the pandemic can receive COVID‑19 unemployment benefits. France modified the regulations for partial unemployment to employees affected by the lockdown. Sweden extended the period of subsidised jobs by one year; a measure that particularly benefits migrants – especially recent arrivals – as these are a main target group.
Many OECD countries also introduced support measures for employers and businesses. These schemes tend to be general and hence open to foreigners as well. One example is the emergency law of March 2020 in Italy, introducing a wide set of measures to address the economic impact of the crisis. In the Netherlands, foreigners with a start-up visa or self-employed residence permit can request the ‘Temporary Bridging Measure for Self-employed Professionals’ allowing for income support and loans for business capital. Sweden prolonged financial support to newly started businesses, among which businesses operated by immigrants are overrepresented.
c. Extending work rights
Work permits are often restricted to a specific sector or employer. In light of the pandemic, some countries have eased these restrictions during the COVID‑19 crisis . In the Czech Republic, for instance, migrant workers who have lost their job could receive authorisation to change employer as well as sector. In Finland, foreign workers with valid residence permit can change employer or field of employment until the end of October 2020. New Zealand has allowed temporary foreign workers in essential services to adjust their work hours and perform different roles for their current employer or their current role in a different workplace.
Policies introduced also extended working right of other migrant groups, including students and asylum seekers. In Belgium for example, until the end of June, asylum seekers hosted by the employer were allowed to work immediately, instead of a previous waiting period of four months. Spain allowed young third-country nationals aged 18‑21, including students and asylum seekers, to work immediately until the end of September 2020. Ireland, France and Belgium for instance, allowed international students to work more hours, while Canada and Australia removed a previous cap on the maximum number of hours international students may work for those employed in essential services.
d. Easing foreign credential recognition and other measures
Many OECD countries have adopted policies to ensure faster labour market access of foreign health professionals already in the country . Italy as well as several provinces in Canada and several states in the United States enabled a temporary licencing of doctors with foreign medical degrees. Chile and Spain facilitated recruitment in the national health services, while other countries including Belgium, Germany, Ireland and Luxembourg expedited applications for the recognition of foreign qualifications of health professionals. France allowed the employment of foreign-trained health workers in non-medical occupations in the health sector.
In Italy, individuals in an irregular situation who can prove their presence in the country prior to 8 March, and are working in the agricultural or domestic care sector are now eligible for a new regularisation pathway. By the beginning of July, over 30 000 requests were submitted, the vast majority from domestic workers.
Apart from these administrative changes, few countries have provided specific targeted support measures. The Norwegian government proposed a package of measures of NOK 456 million targeted towards immigrants, aimed at providing more assistance and increasing competence to facilitate swifter participation and inclusion into the labour market.
Finally, Germany has implemented a number of specific support measures for migrant entrepreneurs, including an online platform with information on available crisis support measures in five languages and a network of dedicated caseworkers to support ailing migrant business in all federal states, including with respect to the filing of applications for state aid.
The progressively comprehensive closure of schools across OECD countries has made online learning opportunities critical for education at all levels. Although schools are better equipped with digital tools than ever before, access to digital learning opportunities is still not equal: children of immigrants tend to be less equipped to face this new transition. In most OECD countries with significant shares of children of immigrants (Figure 6), students with immigrant parents are less likely than students with native-born parents at the age of 15 to have access to a computer and an internet connection at home. However, in spite of the gaps, in all countries, the overwhelming majority of students with immigrant parents do have access to a computer and to an internet connection.
However, children with immigrant parents tend to be overrepresented among those with a low economic, social and cultural status compared to those with native-born parents (Figure 7). In Denmark, Slovenia, Iceland and Greece, more than half of children of immigrants aged 15 were among the bottom quarter of the economic, social and cultural status index. On average across the OECD, students with immigrant parents are twice as likely to be in the bottom quarter of the PISA index of economic, social and cultural status. Prior OECD work has shown that students from such disadvantaged backgrounds are less likely to have access to a quiet place to study than their more advantaged peers and to be proficient in using technology for schoolwork – already before the pandemic . Indeed, students with immigrant parents are consistently less likely to have access to a quiet place to study, with the exception of foreign-born students in New Zealand, but again the differences are not large and in all OECD countries, at least three out of four children of immigrants report to have access to a quiet place to study at home.
Online learning reinforces the role of families in supporting students’ learning. Students with immigrant parents tend have parents with, on average, lower education and poorer socio-economic resources, and who may have a lower understanding of the education system. Students with immigrant parents thus tend to face additional difficulties compared to their peers with native-born parents.
In addition to the consequences of disadvantaged backgrounds on migrant students’ ability to make the most out of online learning and the preparedness of their school systems to this transformation, other factors affect the capacity of children with immigrant parents to succeed in this new way of learning. The lack of fluency in the language spoken in the host country can exacerbate difficulties with home schooling. Language barriers are more challenging when instruction is online, and in particular as it reinforces the potential lack of parental support. In the context of online learning, and without daily contacts with peers and few interactions with teachers, it can be more difficult for students with immigrant parents to overcome these language barriers and learn the host country language. It is particularly challenging in some countries, where the proportion of students who do not speak the host country language is high. On average across OECD countries in 2018, close to half (48%) of 15‑year‑old students with immigrant parents did not speak the language of the PISA assessment at home. This proportion was particularly large in Austria, Finland, Germany, Iceland, Luxembourg and Slovenia (Figure 8).
While the available comparative data on student disadvantage is based on PISA data and thus refers to students at the age of 15, the negative impact on children of immigrants is likely to be higher for smaller children. Smaller children likely depend more on parental support, and the impact of not having many chances to practice the host-country language during several months may impact particularly at younger ages (see also OECD ).
The OECD-Harvard Graduate School of Education Survey estimated the impact of school closures on education continuity to be at least two months of instruction for half of the primary and secondary school students . During this time, a number distance-learning solutions such as online classrooms and broadcasts, as well as computer-assisted learning were implemented to bridge the gap between schools and learners, but the impact on education outcomes remains uncertain, especially for children of immigrants. In terms of material resources, however, there is little evidence of a disproportionate impact. Already prior to the pandemic, on average across OECD countries, there was almost one computer for educational purposes available at school for every 15‑year‑old student and in most countries, the availability of computers in schools tends to be more equitable than at home. In fact, in many OECD countries, the computer-student ratio is actually greater in disadvantaged schools than in advantaged schools. In reaction to the pandemic, many countries have distributed computers to all students in need, which should further lower a negative impact on disadvantaged groups such as children of immigrants.
Likewise, first evidence on the impact on the suspension of final exams on children of immigrants shows divergent influences. Research from the Netherlands, a country with tracking in school, suggests that because there was no central exam at the end of primary school, 14 000 children in that country are likely to have an “advice” for the level of the secondary school that was lower than it would have been had there been a central exam. Children with migrant parents are expected to be overrepresented among this group. In contrast, for those at the end of secondary school, comparison with data from previous years suggests that children with migrant parents are more likely than children with native-born parents to have graduated due to the fact that there was no central exam .
In any case, while the potential learning loss may be largely temporary, other elements associated with the absence of traditional schooling, such as lower educational aspirations or the disengagement from the school system, can have a long-term impact on students’ outcomes. This so-called “hysteresis” effect in education requires specific attention .
Indeed, in a number of OECD countries such as France and Norway, there have been reports of school disengagement by children of immigrants following the pandemic. Preventing negative long-term impact on migrant students’ outcomes requires to make sure that students are not discouraged by their difficulties and stay engaged in learning, and to provide innovative opportunities for interaction in the host-country language beyond teaching hours. This period of pandemic might otherwise widen the gap between students with immigrant parents and native-born students, at a time where many countries had just managed to start to close the gap.
For many migrants, in particular recent arrivals, the near-global lockdown caused by COVID‑19 meant a disruption to language courses they were receiving. Most countries were forced to end in-person integration courses as restrictions were imposed, though many have recently started again with in-person courses. While some countries had online language course options available prior to the pandemic, such programmes had not been scaled to reach every eligible migrant. Several countries halted their integration programmes altogether. Across the OECD, volunteer organisations with language and integration missions also paused their operations.
A break in language learning places some migrants at a significant disadvantage. The first five years after arrival is a critical time span, as migrants need to acquaint themselves with public institutions and available services. Discontinuity of language learning can lead to loss of progress in language courses, often resulting in demotivation. Social isolation from host country natives who could help with informal learning is also a concern. Immigrants who speak the host-country language have greater contacts with native speakers than immigrants with little or no command of the host-country language, in part because they are more likely to be employed in a context that requires speaking to natives. Migrants with a low initial level in the host country language are thus particularly affected by programme shutdowns, including in terms of employability and broader social integration.
In spring 2020, countries took a variety of actions to mitigate these issues. Sweden extended eligibility for integration programmes by 12 months. Denmark postponed obligatory examinations and made it easier to progress to new modules without a test. Some countries tackled the isolation aspect of confinement, setting up online language mentorship programmes and sport tutorials with native speakers. The United Kingdom launched a call for evidence seeking tools to improve social connection for isolated groups. As the lockdown often implied financial hardship for course providers as well, Germany put in place a specific programme to offer them financial support.
The crisis stimulated rapid innovation in the digital and distance-learning sector. Recognising that deferring classes altogether likely results in learning losses and potential dropouts, many countries turned to novel digital tools or ramped up existing offerings (see Box 3). Still, for some subgroups of migrants, such as recent arrivals with little or no-host country language knowledge, the transition to distance learning poses particular challenges that are not present for more advanced courses. For example, while initial language training courses in Germany – as in other countries – virtually came to a halt during the lockdown period, a third of the more advanced vocational language courses smoothly transitioned online. While some migrants have responded positively to moving coursework online, others have abandoned study. An assessment of Australia’s online courses during the lockdown suggested that the move online worked well for those with caring responsibilities but not for asylum-seekers, refugees and other groups with very low literacy. In total, 60% of participants enjoyed the online offers, 40% welcomed the increased flexibility, and 80% of students said they improved their skills during this period. However, 70% stated that the skills would have improved even more in a classroom situation.
During lockdown, Austria’s Österreichischer Integrationsfonds provided free online language courses for levels A1‑B1 of the Common European Reference for Languages (CEFR), and 20 000 eligible migrants participated. Given this popularity, the Ministry is considering maintaining blended and online courses going forward.
Germany invested EUR 40 million and approved nearly 7 000 online classrooms to avoid disruption of courses due to suspension of government services. It also promoted use of its learning portal (vhs-Lernportal), which allows students to link with teachers who can correct any errors with color-coded markers. The portal is both desktop and smartphone compatible. Many migrants reported enjoying the online offerings during confinement, but because the courses could only reach a limited number of migrants, the Federal Office for Migration and Refugees determined classes taken during this period would be a “bonus” that would not count against the migrant’s language learning entitlement. Thirty-eight percent of eligible migrants transitioned to the online course, which was offered free of charge during this period. Job-specific language training could be maintained at 1/3 capacity, while integration courses persisted at lower rates.
To ensure continuity of integration during confinement, France provided 15‑24 hours of distance learning per week to those migrants who had already begun French courses under their integration contract. Because of the need for internet access and basic technological and French literacy, distance training was targeted toward 100‑hour courses (for migrants closest to A1 level during initial placement) with groups of 6‑10 participants, and 200‑hour courses with groups of 3‑5 participants. Out-of-class work took place via app, learning wall, and email. Individual support and follow-up on written exercises was provided. The programme was only introduced on a small scale to test the approach. Based on the lessons taken from the experiment, France plans to integrate e-learning modalities into their general course offerings in the future.
Some countries were able to rely on existing tools. For example, Portugal’s Plataforma de Português Online allows learners to progress through two modules. Users are invited to register and identify a native language, and instructions and materials are made available in that language. Throughout the lessons, tools are explained through both text and audio, and users are assigned a tutor. Estonia offers e-courses designed for both English and Russian speakers. Migrants are supported by an Estonian teacher with whom they can exchange messages via email.
It is important that immigrants not be penalised for lack of technology or digital literacy by missing out on coursework. In Finland, which had a robust and long-standing distance learning programme already in place (a main provider reported a 3% dropout rate in 2019, with 52% of migrants who completed the 200‑day course reaching the target of level B1.1 or above), migrants are commonly provided with devices at the start of a course. During confinement, a postal correspondence option for course materials and homework assignments was introduced for those who could not navigate such tools. Germany provided online tutorials as a supplement rather than subtracting them from the normal entitlement. Such a policy may provide migrants with low digital literacy greater incentive to resume in-person lessons when possible. In Belgium, migrants who do not have access to the necessary equipment to follow the online course received devices from the responsible agencies.
It seems that system in which the organisation of language training and other introduction activities is largely decentralised have faced more difficulties in adapting to distance learning. For example, a report by the Norwegian Research Institute FAFO on how municipalities were able to adapt their introduction activities – chief among which language learning – in the pandemic showed that one in two municipalities faced difficulties in adapting to the new situation.
Moving forward, countries could benefit from the experience of COVID‑19 by identifying areas for improvement so they can add flexibility to their programmes and better harness the potential of online courses in the future. At the same time, it became evident that there is currently little alternative to classroom learning for the most vulnerable groups. In this context, it is essential to understand how best to reach those migrants who are likely to encounter difficulties and to tackle these difficulties.
With COVID‑19 monopolising attention and public debates worldwide, it appears that the salience of immigration issues has decreased in public opinion. However, immigration is at the core of two key and contrasting issues related to COVID‑19. On the one hand, the health care sector relies significantly on immigrant labour. Compounded with the fact that many immigrants work in other essential jobs (agriculture, shop clerks, delivery services), this has put the economic and social contribution of immigrants in OECD countries in the spotlight. At the same time, as cross-border movements have been a key factor of the initial spread of the virus, there have been occurrences of immigrants being scapegoated for the spread of the virus.
The net impact on public opinion regarding immigration of these two opposite narratives is unclear. However, the experience of the Great Recession and of the downturn following the global financial crisis provide insights on public opinion on migration with respect to economic conditions. It suggests that increasing unemployment may be associated with worsening opinion about immigration, especially among the low skilled, implying a tendency for polarisation;;. As unemployment increases and public finances come under pressure, it may thus be expected that public opinion regarding immigration and immigrants becomes more negative.
A number of communication campaigns have aimed at preventing a backlash in public opinion against immigrants in the course of the pandemic, with a particular focus on tackling misinformation regarding the role of migrants in the spread of the virus . In this context, the United Nations have issued a “guidance note” on addressing and countering hate speech related to COVID‑19, with a series of recommendations for governments and media, including social media. The International Organisation for Migration (IOM) implemented a large and decentralised social media campaign in co-operation with Mexico. The campaign specifically targeted populations in local communities with migrant shelters, safe houses or temporary camps for migrants in order to prevent discrimination based on the assumption that migrants accelerate the spread of COVID‑19. A number of local communities, such as Barcelona and New York, also launched specific communication campaigns to address misinformation and discrimination against migrants.
While most information campaigns regarding COVID‑19 aimed at public opinion have been implemented by international organisations and local authorities, a number of countries have also implemented specific measures. For example, the German anti-discrimination agency observed a rise of racist and anti-Semitic discrimination in light of the pandemic – as measured by the increase in complaints to the agency. In reaction, it has launched a dedicated campaign to raise awareness about the rise and provides information on how victims can obtain help. In Finland, the government launched a large, nation-wide campaign on tackling misconceptions about COVID‑19 through social media influencers.
Finally, some countries have publicly recognised the contribution of immigrants in the pandemic. For example, in France, accelerated naturalisation procedures are foreseen for immigrants who were working in the frontline during the lockdown.
As this policy brief has shown, all indicators suggest that immigrants and their children have been disproportionately affected by the initial phase of pandemic and are likely to continue to be highly affected, both with respect to health and integration outcomes. This has followed a period of high migration inflows and better integration outcomes for immigrants across a range of indicators. As the crisis continues and its negative impact, notably in terms of employment, further unfolds, especially for the many recent arrivals, it is urgent to prevent the pandemic from turning into an integration crisis. To this end, a number of policy options could be considered:
Monitor the situation of immigrants and their children regarding health, employment, and education to better identify the issues and the appropriate policy responses.
Ensure access of immigrants to testing and treatment for COVID‑19, in order to alleviate the disproportionate impact on migrants’ health. While many countries have taken efforts to remove such barriers, notably regarding testing, in others reach out needs to be stepped up and remaining financial or legal access barriers removed, especially in countries where testing is still limited.
Ensure that housing and employment conditions for migrants respect sanitary norms in order to avoid spreading the virus. This is particularly the case for some groups of migrants in collective housing (asylum seekers, seasonal workers) and those working in specific difficult work environments (immigrants working in meat processing) face particular health vulnerabilities due to these conditions.
Raise awareness of discrimination and reinforce anti-discrimination measures, as the labour market impact of the pandemic risks increasing discrimination against immigrants and their native-born children.
Make sure that economic and employment support measures do reach migrants, including both workers and self-employed. This implies not only formal access but also information and reach-out to these groups.
Make sure that the contributions of migrants are not forgotten but rewarded.
Provide more flexibility in introduction measures such as language training, particularly where participation or certain outcomes are obligatory.
Address the specific difficulties of children of immigrants in distance learning. This needs to be factored in with respect to decisions about school closure and distance learning, and appropriate support measures taken.
Provide for possibilities for immigrants to interact with native-born outside of regular integration measures. In this context, the online tutorials provided by Germany during the lockdown of integration course are an interesting initiative.
Finally, it is now more important than ever to bear in mind that integration is an investment. As countries implement fiscal stimulus programmes, some of the resources should be used to address key structural problems of the integration of migrants. Lack of investment today will lead to long-term negative integration prospects on migrants, especially on the many recent arrivals, and these will affect not only on migrants themselves, but also across generations on their native-born children. Given the growing population shares of immigrants and their children, the spectre of widening gaps and unequal opportunities is ultimately also a threat to social cohesion – unless appropriate action is taken.
 Anger, S. et al. (2020), Befunde der IAB-Forschung zur Corona-Krise – Zwischenbilanz une Ausblick, IAB, Nuremberg.
 Apea, V. et al. (2020), Ethnicity and outcomes in patients hospitalised with COVID-19 infection in East London: an observational cohort study, Cold Spring Harbor Laboratory, http://dx.doi.org/10.1101/2020.06.10.20127621.
 Aydemir, A. (2003), “Effects of business cycles on the labour market assimilation”, Statistics Canada Research Paper 203.
 Baert, S. et al. (2015), “Is There Less Discrimination in Occupations Where Recruitment Is Difficult?”, ILR Review, Vol. 68/3, pp. 467-500, http://dx.doi.org/10.1177/0019793915570873.
 Basso, G. et al. (2020), “The new hazardous jobs and worker reallocation”, OECD Social, Employment and Migration Working Papers, No. 247, OECD Publishing, Paris, https://dx.doi.org/10.1787/400cf397-en.
 Behtoui, A. (2004), “Unequal Opportunities for Young People with Immigrant Backgrounds in the Swedish Labour Market”, Labour, Vol. 18/4, pp. 633-660, http://dx.doi.org/10.1111/j.1121-7081.2004.00281.x.
 Bertocchi, G. and A. Dimico (2020), “Covid-19, Race, and Redlining”, IZA Discussion Paper, Vol. 13467, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3648807 (accessed on 22 September 2020).
 Brun, S. and P. Simon (2020), “Inégalités ethno-raciales et coronavirus”, De Facto 19.
 EMN/OECD (2020), “EU and OECD Member States responses to managing residence permits and migrant unemployment during the COVID-19 pandemic”, EMN Inform 1.
 EMN/OECD (2020), “Impact of COVID-19 on International Students in EU and OECD Member States”, EMN Inform 2, http://www.oecd.org/coronavirus/policy-responses/managing-international-migration-under-covid-19-6e914d57/.
 Fafo (2020), ““We have kept the wheels in motion” The municipalities’ integration work with refugees during the outbreak of the coronavirus”, Fafo-rapport 2020:16.
 Finseraas, H., A. Pedersen and A. Bay (2014), “When the Going Gets Tough: The Differential Impact of National Unemployment on the Perceived Threats of Immigration”, Political Studies, Vol. 64/1, pp. 60-73, http://dx.doi.org/10.1111/1467-9248.12162.
 Guttmann A, F. et al. (2020), COVID-19 in Immigrants, Refugees and Other Newcomers in Ontario: Characteristics of Those Tested and Those Confirmed Positive, as of June 13, 2020, ICES, http://www.ices.on.ca.
 Hansson, E. et al. (2020), “Överdödlighet våren 2020 utifrån födelseland”.
 Hatton, T. (2016), “Immigration, public opinion and the recession in Europe”, Economic Policy, Vol. 31/86, pp. 205-246, http://dx.doi.org/10.1093/epolic/eiw004.
 Hu, Y. (2020), “Intersecting ethnic and native–migrant inequalities in the economic impact of the COVID-19 pandemic in the UK”, Research in Social Stratification and Mobility, Vol. 68, p. 100528, http://dx.doi.org/10.1016/j.rssm.2020.100528.
 ICNARC (2020), ICNARC report on COVID-19 in critical care.
 Kunst, A. et al. (2020), Oversterfte tijdens de eerste zes weken van de corona-epidemie, CBS.
 Mcginnity, F. and G. Kingston (2017), “An Irish Welcome? Changing Irish Attitudes to Immigrants and Immigration: The Role of Recession and Immigration”, The Economic and social review, Vol. 48/3, pp. 281-304, https://www.researchgate.net/publication/319670153_An_Irish_Welcome_Changing_Irish_Attitudes_to_Immigrants_and_Immigration_The_Role_of_Recession_and_Immigration (accessed on 2 October 2020).
 OECD (2020), “Contribution of migrant doctors and nurses to tackling COVID-19 crisis in OECD countries”, OECD Policy Responses to Coronavirus (COVID-19), OECD Publishing, Paris, https://www.oecd.org/coronavirus/policy-responses/contribution-of-migrant-doctors-and-nurses-to-tackling-covid-19-crisis-in-oecd-countries-2f7bace2/.
 OECD (2020), “Education and COVID-19: Focusing on the long-term impact of school closures”, OECD Policy Responses to Coronavirus (COVID-19), OECD Publishing, Paris, https://www.oecd.org/coronavirus/policy-responses/education-and-covid-19-focusing-on-the-long-term-impact-of-school-closures-2cea926e/.
 OECD (2020), International Migration Outlook 2020, OECD Publishing, Paris, https://dx.doi.org/10.1787/ec98f531-en.
 OECD (2020), Learning remotely when schools close: How well are students and schools prepared? Insights from PISA, OECD, Paris, https://read.oecd-ilibrary.org/view/?ref=127_127063-iiwm328658&title=Learning-remotely-when-schools-close (accessed on 2 October 2020).
 OECD (2020), Managing international migration under COVID-19, OECD, Paris, https://read.oecd-ilibrary.org/view/?ref=134_134314-9shbokosu5&title=Managing-international-migration-under-COVID-19 (accessed on 2 October 2020).
 OECD (2020), OECD Employment Outlook 2020: Worker Security and the COVID-19 Crisis, OECD Publishing, Paris, https://dx.doi.org/10.1787/1686c758-en.
 OECD (2020), “Performance and academic resilience amongst students with an immigrant background”, in PISA 2018 Results (Volume II): Where All Students Can Succeed, OECD Publishing, Paris, https://dx.doi.org/10.1787/263bde74-en.
 OECD (2008), International Migration Outlook 2008, OECD Publishing, Paris, https://dx.doi.org/10.1787/migr_outlook-2008-en.
 OECD (2007), Jobs for Immigrants (Vol. 1): Labour Market Integration in Australia, Denmark, Germany and Sweden, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264033603-en.
 OECD (forthcoming), Communicating on migration and integration in the context of the Covid-19 crisis.
 OECD (forthcoming), Making Integration Work: Youth with Migrant Parents, OECD Publishing, Paris.
 OECD/European Union (2018), Settling In 2018: Indicators of Immigrant Integration, OECD Publishing, Paris/European Union, Brussels, https://dx.doi.org/10.1787/9789264307216-en.
 Papon, S. and I. Robert-Bobée (2020), “Une hausse des décès deux fois plus forte pour les personnes nées à l’étranger que pour celles nées en France en mars-avril 2020”, Insee Focus, No. 198, INSEE, https://www.insee.fr/fr/statistiques/4627049 (accessed on 17 September 2020).
 Raisi-Estabragh, Z. et al. (2020), “Greater risk of severe COVID-19 in Black, Asian and Minority Ethnic populations is not explained by cardiometabolic, socioeconomic or behavioural factors, or by 25(OH)-vitamin D status: study of 1326 cases from the UK Biobank”, Journal of Public Health, Vol. 42/3, pp. 451-460, http://dx.doi.org/10.1093/pubmed/fdaa095.
 Swart, L., D. Visser and M. Zumbuehl (2020), Effect schrappen centraal examen zonder aanvullende maatregelen, CPB.
 Swart, L. et al. (2020), Schrappen eindtoets groep 8 kan ongelijkheid vergroten, CPB.
 University of Bielefeld (2020), SARS‐CoV‐2 in Aufnahmeeinrichtungen und Gemeinschaftsunterkünften für Geflüchtete, https://pub.uni-bielefeld.de/download/2943665/2943668/FactSheet_PHNetwork-Covid19_Aufnahmeeinrichtungen_v1_inkl_ANNEX.pdf (accessed on 25 September 2020).
 van den Berge, W., S. Rabaté and L. Swart (2020), Crisis op de arbeidsmarkt: wie zitten in de gevarenzone?, CPB.
 White, C. and V. Nafilyan (2020), Coronavirus (COVID-19) related deaths by ethnic group, England and Wales - 2 March 2020 to 15 May 2020, ONS, https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/articles/coronaviruscovid19relateddeathsbyethnicgroupenglandandwales/2march2020to15may2020 (accessed on 22 September 2020).
 WHO Bureau for Europe (2018), Report on the health of refugees and migrants in the WHO European Region: no public health without refugee and migrant health, WHO, Geneva, https://www.euro.who.int/en/publications/abstracts/report-on-the-health-of-refugees-and-migrants-in-the-who-european-region-no-public-health-without-refugee-and-migrant-health-2018 (accessed on 2 October 2020).
This brief uses the words “migrants”, “immigrants” and “foreign-born” synonymously. Unless mentioned otherwise, it includes all persons born abroad, regardless of their migration category, legal status, or nationality. Likewise, unless mentioned otherwise, native-born include all persons born in the country, regardless of the country of birth of their parents or of the ethnic minority to which they may belong. Children of immigrants, in contrast, includes all persons with foreign-born parents. It thus includes children who are born in the country but have immigrant parents.
The PISA index of economic, social and cultural status (ESCS) was created using student reports on parental occupation, the highest level of parental education, and an index of home possessions related to family wealth, home educational resources and possessions related to “classical” culture in the family home.