Following two decades of rapid growth, the EU citizen population in the UK has declined in the early 2020s. This briefing outlines what we know about mobility from the EU to the UK and how the EU-origin population in the UK has changed over time.
-
Key Points
- Excluding Irish citizens, there were around 4.1 million EU citizens estimated to be living in the UK at the time of the 2021/22 Census. Including Irish passport holders, the total rose to 5.3 million. Some EU citizens may have been outside the UK at the time of the Census, which was conducted during the pandemic.
More… - EU Settlement Scheme statistics overstate the number of EU citizens currently in the UK – it is likely that many of those who applied to the scheme no longer live in the country. By the end of 2024, around 5.7 million EU nationals applied to the EUSS.
More… - Two main periods of high EU migration to the UK took place in the mid-2000s and early 2010s. EU migration collapsed during the Covid-19 pandemic and remained negative in subsequent years, due to the more restrictive post-Brexit immigration regime.
More… - Some of the main official migration statistics have a history of understating EU migration, although the much more accurate Census data has brought some relief for the time being. Other data sources should be treated with caution, such as the experimental immigration and net migration estimates.
More…
- Excluding Irish citizens, there were around 4.1 million EU citizens estimated to be living in the UK at the time of the 2021/22 Census. Including Irish passport holders, the total rose to 5.3 million. Some EU citizens may have been outside the UK at the time of the Census, which was conducted during the pandemic.
-
Understanding the Evidence
Which Statistics are Available? The UK government publishes several different statistics on migrants in the UK. ... Click to read more.They can be broadly divided into two main types:
Population data estimate the number of EU citizens or EU-born migrants who are living in the UK at a given point in time. The main sources of data on the population are the decennial Census and annual Office for National Statistics (ONS) estimates of the population by nationality and country of birth. The government holds data on tax and benefits records by EU citizens, in a database known as RAPID, but only uses it to publish immigration and emigration figures, not population data.
Migration (immigration and emigration) data show movement to and from the UK. ONS estimates follow the standard UN definition of people arriving or leaving for at least one year, and thus exclude short-term migrants.
In theory, one would expect these two sources of data to say the same thing. That is, if net migration of EU citizens was 100,000 in a given period, the EU migrant population should increase by a similar amount (with a small difference due to births and deaths). However, sources of data on population and migration have traditionally been largely independent and come with different types of error. This means that the population and migration data may be inconsistent, requiring us to consider several different data sources to get an overall picture of migration patterns. Table 2 at the bottom of this document lays out the different available data source and their limitations.
There are also indirect sources of data on EU migration or the number of EU migrants living in the UK. These include data on the number of National Insurance Number (NINo) registrations, and the number of employees on payroll in the UK who were EU citizens at the time they registered for a NINo. The payroll figures provide high-quality data on the number of employees but do not include the self-employed and economically inactive.
How many EU citizens live in the UK?
The latest Census took place in March 2021 in England, Wales and Northern Ireland, and in March 2022 in Scotland. We combine these data to obtain ‘Census 2021/22’ estimates.
An estimated 5.3 million people living in the UK at the time of the 2021/2 Censuses held EU passports. This included 634,000 people who were dual nationals, holding both a UK and an EU passport.
Totals also include almost 460,000 Northern Ireland residents who were born in Northern Ireland and do not hold UK passports. Citizenship is more complex in Northern Ireland, because people born there usually have the right to hold either or both Irish and British citizenship. This means that most people holding EU passports in Northern Ireland were born there and never migrated from an EU country.
Because Irish citizens continue to enjoy free movement-like rights in the UK under the Common Travel Area, they are often excluded entirely from statistics on EU-origin residents. Excluding Irish citizens, the number of EU citizens falls to around 4.1 million, of which around 322,000 were dual UK-EU nationals.
If we analyse the data by country of birth rather than citizenship, there were around 4 million EU-born migrants, or 3.6 million excluding the Irish-born.
Table 1
Although more recent population data are not available, it is likely that the size of the EU population in the UK has declined since the Census. ONS estimates show that the net migration of EU nationals to the UK has been negative since 2021, at an average of around -85,000 per year between July 2021 and June 2024. These estimates are discussed further below. The EU citizen population figure will also be affected slightly by births and deaths of EU citizens in the UK, but not very much.[1]
An interactive map showing the EU-born and EU-passport holding population by local authority in England, Wales and Northern Ireland can be found here.
Why were there many more EUSS applicants than EU residents in the UK?
The high number of EU citizens who applied to the EUSS prompted speculation that ONS had dramatically underestimated the number of EU citizens. By the end of 2024, around 6.34 million people had applied to EUSS. This included 5.7 million EU citizens, 64,000 EEA/Swiss citizens, and 510,000 non-EEA family members.
The 5.7 million EU citizen applicants by 31 December 2024 was 1.9 million higher than the roughly 3.8 million EU citizens living in the UK at the time of the 2021/22 who would have had to apply to the EUSS (dual UK-EU nationals and Irish citizens were not required to obtain status). As explained above, the latter figure is likely to have declined further after the 2021/22 Census, as more EU nationals left the UK than arrived in the country.
Applicant figures are lower than the 8.35 million applications made to the EUSS until the end of 2024, because 1.63 million people had applied more than once, including many people who initially secured pre-settled status and applied again to secure settled status.
However, EUSS figures are not a good way to assess the current size of the EU population. The Census was not perfect (as discussed further below), but it is much better than any other data source available.
What explains the gap between the 5.7 million EU nationals who applied for EUSS and the just under 4 million non-Irish non-British EU citizens living in the UK at the time of the Census?[2] An undercount of EU/non-EU dual passport holders (described further below) could account for a share of the gap, but the largest factor is likely to be emigration. Some EU citizens who applied to the EUSS no longer live in the UK. Anyone who lived in the UK by 31 December 2020, no matter how briefly, was potentially eligible to apply to the EUSS once the scheme opened fully in March 2019. (Although the original application deadline was 30 June 2021, people can still make a late application to the EUSS if they provide reasonable grounds for having missed the deadline, such as serious illness or having been in an abusive relationship; in 2024, around 97,000 late applications were made.)
Data from the ONS suggest that gross emigration of long-term EU citizen residents could have been over 200,000 per year in recent years, and these figures do not include short-term migrants. The estimated number of EU citizens making short-term visits to England or Wales of at least a month but less than a year averaged 370,000 annually in the most recent two years of data before this data series was discontinued (years ending June 2018 and 2019). Therefore, it is plausible that more than one million EU citizens with EUSS status might no longer be in the UK.
How has the EU population living in the UK changed over time?
Before enlargement in the 2000s, EU migration to the UK was relatively low. At the time of the 2001 Census, there were an estimated 1.4 million people born in the EU living in the UK (excluding Estonia, Latvia, Lithuania, Bulgaria, Slovakia, and Slovenia, for which data were unavailable; see Figure 3).
During the 21st Century, there were two main periods of elevated EU migration: the mid-2000s, immediately after the 2004 EU enlargement; and the early 2010s, following the financial crisis and the end of transitional controls on citizens of Bulgaria and Romania.
The total EU-born population in the UK grew substantially in the 2000s and 2010s. In England and Wales, where more complete data by country were available, growth was slightly larger in the 2010s (+1.20m) compared to the 2000s (+1.08m).
Figure 1
The available data suggest that EU immigration fell sharply after 2016. There are several possible explanations, and it is difficult to disentangle them to say which was most important. They include the fall in the pound’s value after the EU referendum result in June 2016, the easing of unemployment in Southern European Member States, and a perception that the UK was no longer a welcoming destination due to Brexit. Lower immigration and slightly increased emigration means that net migration of EU citizens declined after 2016 and became negative after 2021, once free movement ended.
Figure 2
Note that in recent years, the ONS has moved to a new method for producing statistics on the movement of EU citizens. The figures draw on administrative tax and benefits records and are very much a work in progress. They have undergone several revisions and could well see further changes as the methodologies are refined.
How accurate are the UK’s migration statistics?
Census
The Census is the best available data source for measuring the population as of March 2021. In England and Wales, the Census achieved an estimated 97% coverage, and the estimates were then adjusted to compensate for those who did not respond.[3]
The Census is not perfect. It only collected information on one non-UK or Irish passport. As a result, it is likely that some dual-national EU and non-EU citizens will have entered their non-EU nationality despite being in the UK under free movement rules. Nonetheless, it is encouraging that over 550,000 non-EU-born people reported holding an EU passport, which suggests that undercounting of dual citizens has not been a major problem.
The timing of the Census was unusual, in the middle of the COVID-19 pandemic. As a result, some additional people who considered the UK to be their main home at the time of the Census are likely to have been living temporarily overseas due to the pandemic. Many of these people would also be excluded from the population statistics for the same reason, because they were not part of the population at that point in time. The scale of temporary stays overseas during the pandemic is not currently possible to measure.
Population statistics
Between Censuses, the main source of data on the population by nationality or country of birth used to come from an ONS survey. Unfortunately, the survey ran into difficulties during the pandemic. ONS suspended publication of its regular data tables on the population by nationality and country of birth, due to concerns that changing response rates had made the data less reliable. The ONS plan is currently to reintroduce this publication when a redesigned new labour market survey is fully operational (relying more heavily on an ‘online first’ data collection mode, in which most people are expected to respond online and those who do not will receive telephone or in-person follow-ups).
It remains unclear whether and to what extent the new survey will resolve the problems the old survey faced. ONS does not appear to have short-term plans to adjust the current survey figures to produce more reliable estimates in the meantime. For the moment, therefore, it remains unclear when any new population estimates will arrive to update the Census.
When the new population figures do arrive, they may well continue to undercount the migrant population to some extent. Census data indicate that official population data sources have underestimated the EU population in the past. For example, Census data indicated that there were around 3.9 million EU passport holders in England and Wales in 2021, or 3.5 million excluding Irish citizens (excluding UK-EU dual nationals). Official population statistics for the year ending June 2021 — roughly the same period — suggested that these figures were 3.1m and 2.8m respectively. This suggests that by the year ending June 2021, population statistics were undercounting EU citizens by roughly 20% and the EU-born by roughly 13% — or a bit more than this if meaningful numbers of EU/non-EU dual citizens were not captured as EU citizens in the Census.
Table 2
Immigration, emigration and net migration
Concerns about the accuracy of UK statistics first gained prominence in 2013 after the 2011 Census results indicated that official immigration statistics had systematically underestimated immigration from new EU Member States after 2004. Despite various adjustments, concerns that EU immigration and net migration were being underestimated persisted throughout the 2010s. In theory, the move to administrative records helps to resolve the problems the earlier estimates faced, which resulted primarily from the methodology and design of the survey that produced the figures (the International Passenger Survey).
Immigration and emigration statistics are now based on administrative records. However, currently these records do not tell the ONS when EU citizens enter or leave the country. Instead, the ONS is assuming that people who appear in or leave the tax and benefits system for long enough are immigrating or emigrating. This is a substantial limitation in the data.
In summary, UK migration statistics are in a period of transition. Many of the data sources analysts previously relied on have become less accurate or have been suspended. However, certain data sources are nonetheless reasonably reliable. Table 3 provides a summary overview of the main data sources, what they measure and their reliability.
Table 3
Footnotes
[1] The England and Wales Census data recorded 429,000 EU passport holders born in the UK, of whom 69% were under the age of 16 and 25% were under the age of 5. As a back-of-the-envelope calculation (ignoring future emigration and naturalisation of UK-born children), the 108,000 UK-born EU citizen children under the age of 5 suggest that births added roughly 20,000 to the EU citizen population per year. ONS figures suggest that the number of deaths of EU citizens each year also averages just over 20,000.
[2] Note, the EUSS figures include Irish as they cannot be excluded from the published individual applicant data, but the numbers applying for EUSS are sufficiently small that they do not make much difference to the overall calculation.
[3] The Census coverage adjustment effectively ‘adds’ extra respondents to the Census population. The characteristics of the people added to the data match those of other people in the areas where non-responders lived. This will help to correct for undercounting of migrants even if migrants had higher rates of non-response in the Census. It is not possible to know to what extent an undercount may remain after this correction, although for most nationality groups it is likely to be small.
