17 July, 2017
1. Should the UK remain subject to free movement rules after Brexit as a member of the European Economic Area (EEA), we can expect net EU migration to continue at about 125,000 a year into the medium term. The mitigating effect of uncertainty surrounding Brexit, the risk of further devaluation of sterling and the possibility of a period of slower UK economic growth are likely to be offset by the continuation of strong incentives for immigration from the EU. These include a much higher minimum wage in the UK relative to the countries of Eastern Europe, continuing high youth unemployment in southern Europe and the pull effect of more than three million people from other parts of the EU who are already in the UK.
2. With the second round of talks on the UK’s departure from the EU set to take place this week, a significant body of opinion continues to argue that negotiators should seek a ‘soft’ Brexit in which the UK joins the EEA and remains subject to the free movement of people. This paper examines the prospects for EU migration in the medium term should this occur.
3. Figure 1 below shows that net migration from the EU rose from 18,000 in 1997 to 133,000 in 2016. The solid part of the line reflects the official ONS data on EU migration and the dotted part of the line reflects data that we have adjusted to account for the fact that migration from some parts of the EU (the Eastern European - EU8 – countries which joined the bloc in 2004) was undercounted during the years 2004-2008 by 50,000 a year. The undercount was revealed by the decennial Census in 2011.
Figure 1: EU and Non-EU Net Migration since 1997, ONS.
4. Figure 2 below breaks down recent EU migration by region. Net migration from the EU14 remained below 50,000 a year until 2012 after which it rose sharply in the wake of the Eurozone debt crisis. It peaked at 84,000 in both the years ending June 2015 and June 2016.
5. Official figures show that net migration from the EU8 rose sharply following the accession of those states in 2004 and peaked at 87,000 in 2007. However, as already noted, the 2011 Census identified a significant undercount in the Long-Term International Migration (LTIM) figures which the ONS attributed to EU8 during the period 2004-2008. The undercount assigned by the ONS to EU8 migration amounted to 250,000, meaning that EU8 migration was in fact 50,000 a year higher over this period than recorded in the LTIM data. Therefore actual net migration from the EU8 was 137,000 at its peak in 2007. Figure 2 below shows net migration from the EU8 with the adjusted figures (to reflect the undercount) shown by a dotted line.
6. The rise in net migration from Eastern Europe has likely been driven in large part by the fact that wages available in the UK are considerably higher than in countries of origin. For example the minimum wage in the UK, between 2010 and 2015, was over three times higher than in Poland. However, although net migration from the EU8 averaged around 43,000 a year between 2010 and 2015, it fell to 5,000 in 2016.
7. Net migration from Romania and Bulgaria (the so-called EU2 states which joined the EU in 2007) increased sharply from 2013 onwards, more than doubling from 21,000 that year to 54,000 in 2016. This increase was likely augmented by the ending of UK transitional controls on migration from those countries in January 2014.
Figure 2: Net migration from the EU by region, 1997-2016, ONS.
8. According to the Labour Force Survey, the workforce of EU nationals in the UK now numbers 2.32 million and increased by 171,000 over the course of the most recent year (Quarter 1 2016-Quarter 1 2017). Looking at it from the standpoint of country of birth, the number of EU-born people working in the UK increased by 148,000 to 2.36 million in the same period.
Figure 3: Change in UK workforce by EU region 2007 and 2017, ONS.
9. This contrasts with Department of Work and Pensions data on National Insurance Number (NINo) registrations by nationals of different EU regions. Figure 4 below shows that annual registrations by EU14 and EU2 nationals have overtaken those by EU8 nationals.
Figure 4: NINo registrations to adult overseas nationals by EU region, DWP.
10. Figure 5 below shows that the bulk in the rise in NINo registrations by EU14 nationals has been accounted for by those from Italy, Spain, Portugal and (to a lesser extent) France. The rise since 2009/10 has likely largely been driven by recent high unemployment, and in particular, by high youth unemployment in those countries. Particularly notable has been the rise in NINos granted to Italian nationals since 2013/14, at a time when grants to all other EU14 nationalities (apart from registrations by Greek nationals) witnessed a decline.
Figure 5: NINo registrations to adult overseas nationals from selected countries, DWP.
11. Figure 6 below shows that NINo registrations by Polish nationals have declined to levels similar to those seen in 2009/10.
Figure 6: NINo registrations to adult overseas nationals from selected countries, DWP.
12. In contrast, Figure 7 below shows that the rise in NINo registrations by EU2 nationals has been largely accounted for by an increase in registrations by Romanian nationals - from 4,000 in 2006/7 to 182,000 in 2016/17.
Figure 7: NINo registrations to adult overseas nationals from selected countries, DWP.
13. High projected levels of youth unemployment in southern Europe and ongoing disparities between UK wages and those in Eastern Europe suggest that net migration from the EU is likely to continue at high levels into the medium term. We analyse the various factors that could affect flows below.
14. Some commentators have attributed the recent fall in EU immigration, and especially the reduction in net flows from the EU8, to political, economic and social uncertainty in the wake of the EU referendum. However, the government’s offer of ‘settled status’ for the vast majority of EU migrants already in the UK, along with the greater clarity that is likely to emerge as the Brexit negotiations progress, should help to reduce any understandable uncertainty. However, as the Migration Observatory has noted, despite the fall in EU8 net migration, inflows from the EU14 and EU2 have not significantly declined. Indeed, the ongoing high level of NINO registrations to EU nationals (as indicated in Figure 4 above) suggests that the UK has not become notably less attractive for EU migrants from both Western and Eastern Europe.
15. The OECD has said the pound’s drop since the Brexit vote had pushed up inflation, denting household income growth and household spending. Oxford University’s Migration Observatory has argued that the devaluation of sterling could make the UK less attractive to foreign-born workers who send remittances to associates in other countries. However, it adds that devaluation might also influence some foreign-born workers to decide to stay longer in the UK to accumulate higher savings.
16. The OBR have forecast UK economic growth of 2% in 2017, falling to 1.6-1.9% in the period 2018-2020 and then rising again to 2% in 2021. This contrasts with the World Bank, OECD, the United Nations and European Commission which have all predicted lower growth over the next three or four years. For example, in its global economic outlook released in June 2017, the OECD said predicted a slowdown in GDP growth to 1.6% in 2017 and to just 1% in 2018. In contrast, the IMF projects that UK GDP growth will only fall slightly to just over 1.9% by 2022.
17. The UK unemployment rate for the period from March to May 2017 of 4.5% (or 1.49 million people) was the lowest since 1975. The OBR expects UK unemployment to rise slightly to 5.1% by 2021. The OBR notes that: “The 0.7 million rise in employment over the forecast is therefore attributable to additional population growth.” High levels of net migration from the EU are the probable result of a continued large number of job opportunities.
18. The ONS recently reported that in the period May to March 2017 average weekly earnings for employees in Great Britain in nominal terms grew by around 2%, which is lower than nominal pay growth prior to the economic crisis. However, continuing large differences between wages in Eastern Europe and those in the UK are expected to ensure that there remain strong incentives for migrants to move to the UK, even in the face of the recent devaluation of sterling. The UK’s 2017 minimum wage (1,397 Euros) remains three times the level of Poland’s (453 Euros), while Romania’s (275 Euros) remains a fifth that of the UK’s. Figure 8 below shows wage differences in 2017. The planned increase in the National Living Wage to £9 per hour by 2020 is only likely increase the attraction of the UK to potential migrants. Moreover, the fact that lower rates will apply to those under the age of 25 may make the prospect of hiring younger people from abroad even more attractive to UK employers than it is at present.
Figure 8: Minimum wage in 2017 (Euros per month) in selected states, Eurostat.
19. As figure 9 below shows, the OECD forecasts that there will be no convergence of GDP per capita in the UK and Poland before 2035. The OECD have also projected that there will be no convergence of UK and East European wages in the next 20 years.
Figure 9: Projections for UK and Polish GDP per capita (dollars), OECD.
20. The pull factor resulting from kinship and growing existing communities of EU nationals in the UK may act as a draw for immigration. The EU-born population has already grown from just under 1.8 million to more than in 3 million by 2015. The 2015 figure includes 1.5 million people who were born in the EU14, 1.3 million who were born in the EU8 and around 290,000 who were born in the EU2. As figure 10 below shows, the most notable rise has been in the Polish-born population, which has tripled in size over the decade from 265,000 to 831,000. Also notable is the rise in the Romanian-born community over the period – it increased nearly 13-fold from 17,000 to 220,000 over the period. Note that the German-born population in the UK is inflated by the number of UK citizens born in Germany, often to parents stationed there on military duty.
Figure 10: EU-born population in the UK by selected country, ONS.
21. Unemployment in the Euro area (9.3% in April 2017) has declined since its high of more than 12% in 2013. However, the average Eurozone unemployment rate remains more than double than that of the UK. As figure 11 below shows, unemployment across southern European countries is expected to fall by 2020, although it is projected to remain above 10% in Greece and Spain. In Poland unemployment is expected to rise slightly, while Romania’s jobless rate is projected to remain relatively constant.
Figure 11: Projected unemployment rates in selected EU countries 2017 to 2020, Trading economics.
22. Figure 12 below shows that youth unemployment is projected to remain at or above 30% in Greece, Spain and Italy by 2020, while Romania’s youth unemployment rate is expected to reach more than 20%. However, Poland’s rate of youth unemployment is projected to fall below the UK’s.
Figure 12: Projected youth unemployment rates in selected countries until 2017, Trading economics.
23. Romania’s GDP grew at an annual rate of 5.6% in the first quarter of 2017, the highest in the EU. However, figure 13 below shows that the annual growth rates of Poland, Spain and Romania are projected to fall by 2020. Conversely, growth rates are projected to increase in Greece.
Figure 13: Projected annual GDP growth rates in selected countries until 2020, Trading Economics.
24. The continuing effect of the Eurozone debt crisis means that previous alternative destinations in southern Europe remain less attractive to migrants from Eastern Europe than previously. The Romanian population in Italy reportedly almost doubled between 2006 and 2007. This trend has slowed in recent years as Italy has suffered its longest post-war recession. Similarly, what have been traditionally large inflows of Polish citizens into Germany may be mitigated as Germany attempts to integrate hundreds of thousands of migrants from the Middle East and Africa. On the other side of the coin, however, Germany is projected to have healthier GDP growth and relatively low unemployment by 2020 so this could help to draw larger numbers of Polish citizens across the border into Germany.
25. Declining populations of younger people in Eastern and Southern European states may have a mitigating effect on EU migration to the UK in the longer term. For example, the population in Romania of those between the ages of 20 and 34, who have a high propensity to migrate, is projected to decline between 2015 and 2037 from 4.2 million in 2017 to around 3.2 million in 2035 (see figure 14 below).
Figure 14. Population of Romania aged between 20 and 39, OECD.
26. Political turbulence in Eastern Europe could act as a push factor for potential migrants to other parts of Europe. In February 2017, as a result of political disputes over corruption, Romania experienced its biggest anti-government demonstrations since the fall of communist leader Nicolae Ceausescu in 1989. A crisis following the 2016 presidential election in Bulgaria, meanwhile, seemed to have been stabilised by the time of the victory of the centre-right party GERB, headed by Prime Minister Boyko Borisov, during the parliamentary elections of March 2017. Poland’s ruling Law and Justice Party has been accused of pursuing a polarising agenda, a factor which may drive opponents of the government to seek residence elsewhere in the EU.
27. The migration crisis reached a peak in October 2015 with sea arrivals to Greece and Italy reaching an average of more than 7,000 per day. The migration crisis reached a peak in October 2015 with sea arrivals to Greece and Italy reaching an average of nearly 7,500 each day. Flows into Greece were somewhat stemmed after EU and Turkey implemented a plan to halt illegal migration across the eastern part of the Mediterranean in March 2016. However, during 2016 a total of more than 360,000 migrants arrived in Europe by sea. By 4th July 2017 more than 100,000 migrants had arrived in Europe by sea since the start of the year. Moreover, of 1.1 million first-instance decisions on asylum applications in EU member states in 2016, around 61% (or just over 670,000) resulted in positive outcomes (including grants of refugee or subsidiary protection status, or an authorisation to stay for humanitarian reasons). Those who are granted asylum will be able to gain citizenship after qualifying periods of between two and twelve years. The majority of migrants have gone to Germany where those granted protection can apply for citizenship after six years. Such migrants will acquire the right to freedom to move anywhere in the EU after gaining citizenship. As a result, this could account for a significant source of secondary movement of migrants to the UK in the period from 2017 to 2030 and beyond.
28. Looking ahead, the attractiveness of the UK for potential migrants from the EU could be reduced by uncertainty stemming from the Brexit process, by the risk of a further devaluation of sterling and by the possibility of a period of lower growth. On the other hand, there will continue to be a large wage disparity between the UK and Eastern Europe, while the substantial diaspora of EU nationals already in the UK will continue to act as a pull factor.
29. As regards Europe, the conditions in home countries will vary substantially over the medium term. The EU14 can be expected to have moderate growth in GDP but to retain quite high levels of unemployment, especially youth unemployment. Accordingly, we can expect that net migration from the EU14 will be somewhat lower than its 2016 level of 73,000, and to perhaps fall to around 50,000 a year.
30. As for the EU8, there has clearly been a sharp reduction in net migration in the last year which may reflect improving conditions in Poland, a reduced exchange rate in the UK and possibly uncertainty arising from Brexit. Nevertheless, the large Polish diaspora will be a significant factor and we would expect net migration to recover from its 2016 level of 5,000, to about 25,000 a year.
31. Romania and Bulgaria are a different proposition. Economic and political conditions in those countries are poor. It is therefore likely that there will continue to be a significant EU2 inflow. The number of National Insurance Number registrations, particularly by Romanians, remains high (registrations by Romanian nationals stood at 180,000 in 2015/16 while registrations by Bulgarians stood at 40,000). We expect net migration from these two countries to continue at 50,000 a year. In total, this gives an estimate of EU net migration of the order of 125,000 a year.
32. Note that this estimate does not include the potential secondary flows of migrants who have been granted humanitarian protection or refugee status after arriving in Europe.
33. Current levels of non-EU net migration stand at 175,000. Unless further restrictions are introduced that number is not likely to fall significantly. Subtracting net British emigration of 50,000 a year and adding our estimate for EU migration, total net migration would run at around 250,000 a year. We note that the ONS population projection high migration scenario is 265,000 a year. It is therefore likely that net migration in the medium term will remain around this level rather than fall in line with their principal projection, which has a migration assumption of 185,000 a year. This in turn means that the UK population is likely to increase by nearly six million over the next decade and by just over ten million over the next twenty years. In effect, a failure to control EU immigration would result in the ONS high migration scenario under which the population rises at half a million each year; 75% of that increase would be the result of the direct and indirect effects of immigration.