How will Brexit affect immigration?

Brexit and the employment of foreigners in Great Britain: from model country to bulwark?

Many old member states viewed the introduction of the free movement of workers in the course of the eastward expansion of the EU as problematic from the outset. They feared that Western European labor markets would be inundated with cheap labor from Central and Eastern Europe. For this reason, transitional rules were agreed in 2001 that allowed national labor markets to be sealed off for a period of up to seven years. While Germany made full use of this waiting period, Great Britain opened its labor market as early as May 1, 2004 without any major restrictions. But not even twelve years later, the British vote to leave the EU is seen in a direct connection with labor migration from Central and Eastern Europe. Great Britain no longer appears as a beneficiary of the free movement of workers, but as a victim of its liberal opening policy. Could it be that the ambivalent relationship between the British and the EU has suffered the decisive rift through the exemplary opening of the British labor market and that Britain will become a bulwark in the wake of Brexit?

British welcoming culture

The 2001 Gothenburg Compromise contained transitional arrangements that allowed Member States to postpone the opening of their labor markets to workers from the candidate countries in three phases over a total of seven years. Only Great Britain, Ireland and Sweden waived a transition phase and opened their labor markets completely on May 1, 2004. Many EU states followed these liberal pioneers only two to three years later. Germany and Austria, however, were the only old members to fully exhaust the transitional regulations

In the following years, Great Britain developed into a beneficiary of the free movement of workers: as early as 2005, more than 15% of workers from the later twelve accession countries (E12) who found employment in the EU15 countries had their jobs in Great Britain (see Figure 1).

illustration 1
Employment of workers from the E12 acceding countries1 in the EU15 countries2

1 E12 candidate countries: Bulgaria, Estonia, Latvia, Lithuania, Malta, Poland, Romania, Slovakia, Slovenia, the Czech Republic, Hungary and Cyprus. 2 EU15 countries: AT = Austria, BE = Belgium, DE = Germany, DK = Denmark, ES = Spain, FI = Finland, FR = France, GB = Great Britain, GR = Greece, IE = Ireland, IT = Italy, LU = Luxembourg, NL = Netherlands, PT = Portugal, SE = Sweden.

Source: Eurostat: Database: Labor Market: Employment by sex, age and citizenship (1000) [lfsa_egan], http://ec.europa.eu/eurostat/data/database?node_code=lfsa_egan (15.5.2018); own representation and calculations.

Only Spain, Germany and Italy each had more employees from the acceding countries. But in 2017, Great Britain led this ranking with a share of 31%, well ahead of Germany. It is noteworthy that from 2005 to 2017 the number of E12 employees in the EU15 countries tripled, but in Great Britain even increased by 7.5 times. During this period, the number of employees from the accession countries did not fall, even in the crisis years or when the economy slowed - the increase only slowed down (see Figure 2).

Figure 2
Employment of workers from the E12 acceding countries in the EU15 and in the UK

Source: Eurostat: Database: Labor Market: Employment by sex, age and citizenship (1000) [lfsa_egan], http://ec.europa.eu/eurostat/data/database?node_code=lfsa_egan (15.5.2018); own representation and calculations.

However, with the accession of Bulgaria and Romania to the EU on January 1, 2007, a change in policy began in Great Britain: Like other EU15 countries, the country took full advantage of the seven-year transition phase until the end of 2013 to ensure free access for Bulgarian and Romanian workers prevent UK labor market. These restrictions are understandable since the global economic and financial crisis and the subsequent downturn in the labor market fell during this transition period (see Figure 3). The British demand for foreign workers corresponded very closely with economic development, so that when the British labor market recovered in 2013, immigration also increased significantly again. In contrast to previous years, however, the predominant workforce was no longer from Poland, but from Bulgaria and Romania - Romania alone accounted for almost half of all immigrant workers from the accession countries up to 2017. Demand for the Romanian and Bulgarian workers was in particular from the construction industry, the transport industry, as well as from the hotel and catering trade.2 Only recently have fewer workers from the E12 countries been registered in Great Britain.

Figure 3
EU labor immigration to the UK and UK unemployment1

PL = Poland, BG = Bulgaria, RO = Romania.

1 Left scale: Annual immigration of employees from EU countries based on registration with the British social security system (“NINO registration”); Right scale: Harmonized unemployment rate in% of the labor force.

Sources: Eurostat; Database: Labor Market: Unemployment - LFS adjusted series (une), unemployment by sex and age - annual average [une_rt_a], http://ec.europa.eu/eurostat/data/database?node_code=une_rt_a (May 17, 2018) ; Department for Work and Pensions: NINO Registrations To Adult Overseas Nationals Entering The UK; own representation and calculations.

The openness of the British labor market to foreign workers is not only benefiting workers from the EU accession countries, which account for around 38%. In addition, there are currently 27% employees from the EU15 countries and 35% from non-EU countries (see Figure 4). In the observation period since 2005, the share of foreign workers in total British employment has almost doubled, with the growth being driven in particular by the E12 countries. Because of this development, the UK has had a higher proportion of foreign workers than the EU average since 2008.

Figure 4
Employment of foreign workers in the UK 1

1 Left scale: employees by country of origin; Right scale: share of foreign workers in total British employment.

Source: Eurostat: Database: Labor Market: Employment by sex, age and citizenship (1000) [lfsa_egan], http://ec.europa.eu/eurostat/data/database?node_code=lfsa_egan (15.5.2018); own representation and calculations.

Even after the Brexit vote, Great Britain remained attractive to foreigners: In the last year of observation up to September 2017, there was still an increase in EU citizens on balance, even if this was smaller than in previous years due to a lower number of jobseekers; on the other hand, the net immigration of non-EU citizens to Great Britain continued to increase.3 So there can be no question of a bulwark against foreigners.

Free movement of workers as the cause of Brexit?

The British voter vote for Brexit is often directly associated with labor migration to Great Britain in the course of the EU free movement of workers.4 This presumed connection can be found in single regressions at constituency level within the framework of the available data from the Office for National Statistics and the Department for Work and Work Pensions ”, however, did not confirm.5 In a comparison of the socio-economic indicators, the variance of the“ Leave ”share of the vote is best explained by the level of education of the voters, as an overview shows: the higher the level of education of the resident population, the lower the share of the“ Leave "voices. This correlation is particularly strong for the group with high and highest qualifications (corrected R2 = 0.64) - even if this group is expanded to include medium qualifications, a strong correlation remains (corrected R2 = 0.52).

In the case of a further group of indicators, a connection with the proportion of “leave” votes can also be seen, but the explanatory content is significantly less. These include:

  • the level of the average income in the constituencies, which is negatively correlated with the “leave” share of the vote (corrected R2 = 0.25);
  • the proportion of natives without a migration background, which is positively correlated (corrected R2 = 0.28);
  • the proportion of people over 60 in the resident population that is positively correlated but contributes even less to the explanation (corrected R2 = 0.18).

Relevant labor market indicators have no significant explanatory value: The unemployment rate, the change in the unemployment rate in the observation period from 2004 to 2015 and the share of unemployment benefit recipients cannot explain the variance in the “leave” share of the vote. The latter also applies to the majority of the indicators used to record the employment of foreign workers in the constituencies. Indicators that show the registration of foreign workers (National Insurance number, NINo) from the Central and Eastern European EU accession countries (E10) or their changes in the constituencies do not play a role - this also applies if the subgroups "Accession Countries 2004 "(E8) and" Bulgaria and Romania "(E2) 6

A multiple regression analysis largely confirms these relationships from the single regressions (see Table 1). A model approach without variables that capture registered foreign workers can already explain the variance of the “leave” share of the vote well (corrected R2 = 0.74) (model 1). If the probability of error is less than 1%, the variable “Proportion of highly qualified” has the greatest influence on the “Leave” share of the vote, which decreases with increasing proportion of highly qualified. In the same way, the variable “Proportion of over 60-year-olds” has the second largest influence - the “Leave” proportion of the vote is higher, the more older people live in an electoral district. The variable “proportion of citizens without a migration background” is also highly statistically significant in this model, but it has a comparatively small positive influence. Statistically less secure is the variable “income level” with an even smaller influence on the “leave” share of the vote. All labor market indicators and the proportion of unskilled workers are statistically insignificant. 7

If this model is expanded to include variables that record the registration of foreign workers (NINo registrations) in the constituencies, the explanatory content increases only slightly (corrected R2 = 0.75) (model 2). The additional variables are highly significant, but their influence on the proportion of “leave” votes is only minor. The coefficient of the inventory variable has a negative sign, while the coefficient of the change variable has a positive sign. This means that the increase in foreign workers leads to a higher “leave” share, while a higher stock of foreign workers is associated with a lower “leave” share. If, instead, only the stock and the change in NINo registrations of workers from the E10 accession countries are taken into account, this finding remains largely unchanged (model 3). Although the coefficients of the E10 variables both have a positive sign, they are just as small - only the explanatory content is slightly higher overall (corrected R2 = 0.76).

The regression analyzes at constituency level, for example, suggest that the level of education of the electorate had the greatest power to explain the “leave” share of the vote in the Brexit referendum. The higher the level of education in a constituency, the lower the “leave” share. These constituencies seem to be aware of the advantages of UK membership in the EU. There can be a variety of reasons for this: an information advantage, lack of fear of loss, a cosmopolitan attitude or own economic advantages would be conceivable. In the case of the low-skilled, on the other hand, these advantages do not seem to be relevant or to correspond to personal experience. The stronger support for the “leave” by older voters, who also see the dangers of EU membership rather than the advantages, is likely to go in the same direction. The actual economic situation of the voters, on the other hand, had no direct influence on voting behavior. The number of foreign workers also did not play a major role, only the increase in foreign workers as a result of the EU free movement of workers had a weak influence in favor of the “leave”.

Classification of the examination results

The present research results and their interpretation are in line with a number of alternative analyzes that also try to empirically explain the outcome of the Brexit referendum. This includes in particular the work of Goodwin and Heath: 8 At the counting district level, these show that there has been a polarization of the British electorate, especially along educational boundaries. Less educated and less qualified people tended to vote in favor of “leave” regardless of their other political ties. The voters of the anti-EU UKIP and supporters of the Euro-skeptical wing of the Conservatives evidently were joined by many Labor voters. The authors' analysis makes it clear that age differences were less decisive for the voting result than differences in education - less educated younger people were more likely to vote for “leave”, while educated older people were more likely to vote for staying in the EU. In general, Goodwin and Heath conclude that groups at risk of poverty - low-wage earners, pensioners, low-skilled unemployed - support Britain's exit from the EU out of fear of a change in economic and social structures and cultural upheaval. These groups differ significantly from the predominantly cosmopolitan British “elites” who also benefit personally from the EU and globalization. On the other hand, at Goodwin and Heath, too, the proportion of EU migrants had at most a minor influence on the voting result. Only the strong increase in labor migration from the accession countries in Central and Eastern Europe within a relatively short period of time has evidently given an additional boost to EU skepticism beyond the borders of the political camps.

Becker, Fetzer and Novy, who carry out an econometric analysis at the district level, find similar explanations for the “Leave” share of the vote.9 They also see the reasons for the voting result in the fact that larger sections of the population increasingly felt economically and socially marginalized. In their analysis they show, on the one hand, that the "leave" share increases significantly as the level of education or qualification falls. On the other hand, the “leave” share increases with the unemployment rate and the share of people over 60. The authors also confirm this picture with a number of other indicators: Those who benefit little from wage increases, who do not have adequate health services available, who suffer particularly from government austerity measures and who are not a London commuter, tended to vote in favor of Brexit.

Table 1
Explanations for the Brexit vote on June 23, 2016: regression analysis at constituency levela
Dependent variable: "Leave votes" at constituency level in%
Model 1Model 2Model 3
Explanatory variablescoefficientStandard errorcoefficientStandard errorcoefficientStandard error
Higher qualifications (NVQ4 +) in% of the resident population (ages 16 to 64) 2015b-0,770,04***-0,710,04***-0,730,04***
Unqualified persons as% of the resident population (ages 16 to 64) 20150,030,09-0,020,090,060,09
Over 60 year olds as% of the resident population in 20150,320,08***0,400,08***0,350,08***
Indigenous population groups as% of the resident population in 20110,080,02***0,020,030,130,03***
Average gross annual income in% of the national average in 20150,040,02**0,030,02*0,050,02***
Unemployment rate 2015-0,050,130,000,13-0,060,13
Recipients of unemployment benefits in% of residents (ages 16 to 64) 2014/2015-0,430,37-0,260,36-0,450,37
NINo registrations per 1000 voters 2014/2015 c-0,030,01***
Change in NINo registrations from 2004 to 2015 in%0,020,00***
NINo registrations of E10 employees per 1000 voters (2014/2015)0,050,02***
Change in NINo registrations of E10 employees from 2004 to 2015 in%0,000,00***
constant64,613,4963,924,3054,324,28
Number of observations508508490
Corrected R20,740,750,76

a Ordinary Least Squares Analysis (OLS Analysis); Selection of variables according to Akaike and the Bayesian information criterion; Significance levels: *** 1%, ** 5%, * 10% probability of error. b Qualification levels according to the five-level “National Vocational Qualifications” (NVQ), here: residents with an educational level of level 4 and higher. c For NINo registrations, see Figure 3. d E10 = E12 acceding countries excluding Malta and Cyprus.

Sources: Department for Work and Pensions (DWP): NINo Registrations to Adult Overseas Nationals Entering the UK-Westminster Parliamentary Constituencies (Northern Ireland excluded) and Nationality by Quarter and Year of Registration; Office for National Statistics (ONS): Nomis - Official Labor Market Statsitics by Westminster Constituencies; own calculations and compilation.

The authors make it clear that, against this background, the immigration of foreign workers could not be decisive for the voting result: the “leave” share of the vote is positively correlated with the increase in migrants from the Central and Eastern European accession countries, but only weakly statistically confirmed. The proportion of migrants from the EU15 and from non-European countries of origin is even negatively correlated with the “leave” votes. The explanation given is that the low-skilled migrants from the E12 countries could lower wages and make rents for simple housing more expensive, which would put local low-skilled workers at risk. The E12 migrants appear to reinforce existing problems in already disadvantaged social milieus. In cosmopolitan milieus with higher incomes, where the residents see themselves as beneficiaries of “Europeanization”, however, labor migration is seen as beneficial and the “leave” share is correspondingly lower.10

Perception and reality

The empirical analyzes of the Brexit referendum leave the impression that it was not economic facts that were decisive for the “leave” share of the vote, but rather vague fears of loss. This is consistent with the fact that the actual income development in Great Britain after the EU's eastward enlargement hardly gives any indications of income losses. At best, the migrant workers had a slight dampening effect on the wage growth of the unskilled and unskilled labor.11 Nevertheless, the median disposable income of the bottom fifth of the British population rose by 15% in the ten years after the economic and financial crisis alone, while the median income of the top fifth rose only increased by 0.5%. Correspondingly, inequality in the distribution of income in Great Britain decreased during this period, as shown by a falling Gini coefficient as a measure of the inequality in income distribution

Pensioner households also experienced an absolute and relative income improvement before the 2000s: their incomes have grown much faster than those of non-pensioner households in most years since the late 1970s. The unequal distribution of income also decreased for both household types. Therefore, fears of loss among the elderly can hardly be attributed to systematic loss of income or a growing income spread. 13

In addition, studies show that since the 2000s, low-income and pensioner households have on average not had to accept any loss of purchasing power as a result of increasing general living costs. A notable exception is the affordability of home ownership, which has risen disproportionately over the past five years, especially in the growth regions of London and south-east and east England. The chances of low-wage earners to acquire affordable housing there have fallen significantly, but this is more of an urbanization phenomenon. 14

Finally, the recipient structure does not justify any fear of loss when it comes to social benefits, as there is no displacement of indigenous people in need by migrants. At the beginning of 2017, just 7.4% of the recipients were non-British, while EU citizens from Central and Eastern Europe accounted for 1.1%. At the beginning of the decade, these proportions were 6% or less than 0.5%, but this development cannot be viewed as dramatic - especially since the absolute number has changed little for non-British recipients and even at the current edge was declining.15 There was no massive immigration into the social security systems as a result of the free movement of workers.

Lessons from Brexit

Against the background of the present analysis, Great Britain can be seen as a major beneficiary of the EU free movement of workers up to the current edge. It was able to attract the manpower required for its prosperity development on the European labor market and thus compensate for its own shortages across all qualification levels. The immigration of workers from the EU accession countries proved to be cyclical, despite immigration, British unemployment fell and incomes and purchasing power rose for all types of households, while there was no significant immigration into the British welfare systems.16 The needs of the British The economy of foreign workers will not change as a result of Brexit. Therefore, even a future national control of labor migration should not turn Great Britain into a bulwark.

The empirical analysis of the voting results also does not support the hypothesis that the free movement of workers in the EU was an important factor for the majority of the “leave” camp. The explanatory content of indicators that record the stock and growth of foreign workers in the constituencies is extremely small. Personal experience with foreign workers was not decisive for the “Leave” share of the vote, on the contrary. The only exception is the rapid increase in labor migration from Central and Eastern Europe. The high level of immigration within a very short period of time may have increased diffuse fears of foreign infiltration and the fear of loss of control - even if Great Britain has traditionally been a country of immigration due to its colonial heritage.

Instead of the free movement of workers, an increasing rejection of British EU integration by voters with a lower level of education - regardless of age and political orientation - was decisive. This group of voters strengthened the camp of traditional EU skeptics, which had not been able to hold a majority until then. The majority in favor of Brexit can be seen as a rejection of an “elite project” by those who tend to expect disadvantages from European integration. Regardless of their actual situation, fear of loss seems to prevail among these voters: For them, the “Europeanization” of Great Britain threatens social security as well as national identity and the cultural values ​​associated with it. The EU is obviously not perceived as a protection against future dangers, but rather counted among the causes of the threat. From this point of view, the influx of EU workers, for example, is perceived as a foretaste of a “flooding” of the British labor market in the course of future rounds of enlargement. The simultaneous deepening of the EU through the communitarisation of further policy areas is understood as a progressive loss of national control, which makes it more likely that social acquisitions will melt away. In contrast to the cosmopolitan educated elite, the “Leave” supporters obviously have doubts about their own assertiveness in a European society. The result of the referendum therefore does not reflect a temporary mood, but reveals a deep rift through British society.

Since such social tensions are by no means unique to Great Britain, European politics cannot simply go back to business. However, it should also resist the temptation to make an example of Great Britain and thereby cause further damage to European cooperation. Rather, Brexit should be the occasion for a fundamental discussion: Wouldn't it be advisable to forego new rounds of enlargement until the economic gap within the EU has been further reduced and the possible enlargement candidates have visibly caught up? And what speaks against a restriction of the EU to core competencies according to a strict subsidiarity principle, whereby it would be left to the individual member states to decide on their integration steps in their own sovereignty? Promote populism and only do further damage to the European project.

  • 1 Cf. in detail HH Glismann, K. Schrader: Boundless Freedom on Europe's Labor Markets - Exclusion in Germany ?, in: Wirtschaftsdienst, 91st year (2011), no. 5, pp. 315-317, https: // archiv .wirtschaftsdienst.eu / year / 2011/5 / unlimited-freedom-on-europe-labor markets / (2.8.2018).
  • 2 See Office for National Statistics: Living abroad: dynamics of migration between the UK and the EU2, London 2017.
  • 3 See Office for National Statistics: Migration Statistics Quarterly Report: February 2018, London 2018.
  • 4 Cf. BBC: Eight reasons Leave won the UK’s referendum on the EU, June 24, 2016, https://www.bbc.com/news/uk-politics-eu-referendum-36574526 (March 14, 2018).
  • 5 data sources: Department for Work and Pensions (DWP): NINO Registrations to Adult Overseas Nationals Entering the UK-Westminster Parliamentary Constituencies (Northern Ireland excluded) and Nationality by Quarter and Year of Registration; Office for National Statistics (ONS): Nomis - Official labor Market Statsitics by Westminster Constituencies; own calculations and compilation.
  • 6 For the NINo statistics, see Department for Work and Pensions: Statistics on National Insurance Number Allocations to Adult Overseas Nationals, Assessment Report, No. 321, London 2017.
  • 7 If the variable “proportion of highly qualified” is omitted in this model and only the variable “proportion of non-qualified” explains the influence of the level of education on the proportion of “leave”, this variable becomes highly significant; its coefficient becomes relatively high with a positive sign Value on; the "income" gains significance and explanatory power with a negative sign; "Age" and "British nationality" also gain in explanatory power - however, the explanatory content of the model decreases overall (corrected R2 = 0.55). The key messages of model 1 therefore remain unchanged.
  • 8 Cf. M. Goodwin, O. Heath: The 2016 Referendum, Brexit and the Left Behind: An Aggregate-Level Analysis of the Result, in: The Political Quarterly, 87th year (2016), no.3, pp. 323-332; This: Brexit vote explained: poverty, low skills and lack of opportunities, https://www.jrf.org.uk/report/brexit-vote-explained-poverty-low-skills-and-lack-opportunities (28.2. 2018).
  • 9 Cf. S. O. Becker, T. Fetzer, D. Novy: Who voted for Brexit? A Comprehensive District-Level Analysis, CEP Discussion Paper, No. 1480, London 2017; and this: What voting patterns are the Brexit referendum based on ?, in: Wirtschaftsdienst, 98th Jg. (2018), no. 13, pp. 41-45; in addition S. O. Becker, T. Fetzer: Does Migration Cause Extreme Voting ?, http://www.ceistorvergata.it/public/CEIS/file/seminari/2017/Becker(1).pdf (28.2.2018).
  • 10 An analysis by the BBC of the reasons for the victory of the “Leave” camp confirms the picture of a divided British society: the political and economic elites who experience the EU as beneficial are opposed to the low-skilled, the elderly and euro-skeptics who are one fear social decline and the loss of national identity, see BBC, op. a. O. Whyman and Petrescu also show in a migration analysis that although the British economy benefits from the immigration of highly skilled workers, the low skilled are exposed to wage pressure, albeit slight, from EU migrant workers. See P. B. Whyman, A. I. Petrescu: The Economics of Brexit, London 2017, pp. 185-221.
  • 11 See S. Nickell, J. Saleheen: The impact of immigration on occupational wages: evidence from Britain, Bank of England, Staff Working Paper, No. 574, London 2015.
  • 12 Office for National Statistics: Household disposable income and inequality in the UK: financial year ending 2017, London 2018, pp. 7-12.
  • 13 See ibid, pp. 13-17.
  • 14 See Office for National Statistics: Household Costs Indices, UK: preliminary estimates 2008 to 2017, London 2018, pp. 13-25; same: Housing affordability in England and Wales: 2017, London 2018.
  • 15 Department for Work and Pensions: Statistics: Nationality at point of NINo registration of DWP working age benefit recipients: data to Feb 2017, https://www.gov.uk/government/statistics/nationality-at-point-of-nino -registration-of-dwp-working-age-benefit-recipients-data-to-feb-2017 (8.8.2018).
  • 16 Cf. also H. H. Glismann, K. Schrader, op. a. Cit., Pp. 320-322.
  • 17 Cf. for example J. Stehn: The core problem of the EU, Kiel Policy Brief, No. 106, Kiel 2017.

Title: Brexit and Labor Migration in the United Kingdom: Transition of a Model Country into a Bulwark?

Abstract: In the course of the EU’s Eastern enlargement, the United Kingdom was one of the few EU15 countries which took advantage of the opportunities provided by the freedom of movement for workers from the beginning. Since then, the United Kingdom is prospering and still relies heavily on migrant workers. As the empirical analysis shows, labor migration from the accession countries was not a determining factor in the outcome of the Brexit vote. It was rather the less educated voters who perceive a threat to their social status and their national identity as the EU’s grows. As a result, a policy discussion on varying EU policies is advisable.

JEL Classification: D72, F22, R23