For the first time, UK long-term international migration statistics no longer rely on survey data but are, instead, produced using administrative data. While this is a landmark moment and gives us an improved picture of past migration patterns, using admin data can present fresh challenges, as Mary Gregory explains.
Since the pandemic – when the International Passenger Survey (IPS) was paused – the ONS has increased its use of administrative data, such as visa data and tax information, to improve estimates of UK long-term international migration (LTIM) – people changing the country they live in for one year or more. This enables us to produce estimates based on observed behaviours rather than relying on what people told us about their future intentions while travelling through a port or airport.
For people outside the European Union (EU), we’ve been using Home Office Borders and Immigration (HOBI) data since 2021. Over time, we’ve been learning more about the HOBI data and now have a better understanding of non-EU+ migrants than ever before.
More recently, our focus has been on improving how we estimate EU+ and British migration. As of today, we’ve aligned our data source and methods for EU+ estimates with non-EU+ estimates, while we’ve moved our estimates of British nationals away from the IPS to Department of Work and Pension (DWP) tax and benefit records. This is the same system we previously used to estimate EU+ migration. We have taken learnings from that as we have developed our methods for British nationals.
Non-EU+ migration has been the main driver of international immigration in recent years, with people coming to the UK to work and study (as well as for humanitarian reasons such as Ukraine resettlement) driving the record levels of net-migration in 2022 and 2023.
While our non-EU+ methods are increasingly mature, our use of admin data means that as our understanding improves or we receive new data, we will see updates to our estimates, as my blog last November set out. In our latest estimates we have updated our non-EU+ migration estimates to reflect improvements by the Home Office to the data matching process within the HOBI data they provide to us. The maturity of our methods for non-EU+ migration means we expect fewer changes in future. However, we will always make quality improvements as our data and understanding mature further. When we do make changes, we will be transparent about these and their impact.
Our new methods for EU+ and British Nationals, alongside the updated data for non-EU+ migration, mean our recent net migration estimates show a higher and slightly earlier peak of 944,000 in the year ending March 2023, compared with our previously published estimate of 906,000 in the year ending June 2023. This is followed by a sharper fall since the peak. Our latest estimate for year ending December 2024 is now 345,000 compared with our provisional estimate of 431,000 published in May this year.
While overall trends are similar to those previously published, beneath the surface, these improvements paint a different picture of EU+ and British migration in recent times.
Understanding British migration
Understanding the long-term international migration patterns of British nationals has been, and always will be, challenging due to the sheer number of British people crossing the border on a daily basis. Very few of these will be migrants and British people don’t need a visa to travel to the UK so we cannot use HOBI data to estimate migration of British nationals.
Not only was the IPS stretched beyond its intended purpose, Census 2021 showed it was underestimating British emigration. For these reasons, we have been exploring many different data sources to improve our estimates.
Our ambition is to move to using travel information to measure British migration if it is proven to be of sufficient quality. In the meantime, we are moving to use of the Registration and Population Interaction Database (RAPID), maintained by DWP. This offers good coverage of British nationals, but it is not intended as a data source for migration. We therefore have to make a range of assumptions based on people’s activity to determine whether or not someone is counted as a long-term migrant. We will continue to monitor and refine these assumptions.
Based on these new data and methods, it is clear the IPS continued to underestimate British emigration since 2021, and it also underestimated immigration. From year ending December 2021 to year ending December 2024, immigration of British nationals has been revised upwards by a total of 306,000 and emigration by 650,000. Cumulatively, this means compared with our previous estimates, 344,000 more British nationals left the UK than arrived between year ending December 2021 and year ending December 2024.
We know we have more work to do to unpick the migration of British nationals. We have data on age and sex, but RAPID doesn’t give us a reason for migration or tell us where British people are going to. We will need improved data sources to truly understand what is going on with British migration and we are continuing to explore other data sources which have the potential to support this.
EU+ migration
Since 2021 we’ve been using RAPID to estimate EU+ migration. As I’ve just explained, RAPID is designed to track individuals’ interactions with public services over time. It does not directly record migration events but infers them based on the presence or absence of interactions with government systems.
Recent policy changes, particularly the end of freedom of movement following Brexit and the implementation of the EU Settlement Scheme, have significantly impacted migration patterns and data collection practices. HOBI has become increasingly relevant for tracking EU+ nationals post-2021 [as they need settled status or a visa]. We’ve therefore taken this opportunity to move estimates of EU+ migration to HOBI. This brings greater consistency with our non-EU+ estimates, is a more timely data source and enables us to understand more about EU+ migration patterns.
Our new estimate of EU+ net migration over the period is higher than our previous published estimate. The biggest difference is at the end of 2021 and early 2022, where HOBI shows around 100,000 fewer EU+ nationals emigrating. In the most recent period (year ending December 2024) our net migration estimate is -69,000 compared with the previously published estimate of -96,000.
Looking ahead
Today’s changes mark a significant moment in our journey to move away from the IPS to estimate long-term international migration. This has not been without challenge.
Before the pandemic, migration was relatively stable. The pause of the IPS at the start of the pandemic accelerated our move to new methods. However, this came during a period when patterns and behaviours have been shifting considerably, influenced by a new immigration system, policy changes and a series of world events.
During this period, our international migration estimates have been badged as official statistics in development as we’ve worked to better understand new data sources. This continuous development is an important part of our commitment to providing the best possible estimates and we will continue to make improvements as we get access to better data.
For instance, we are working with the Home Office to investigate inconsistencies in some visa data, mainly affecting family emigration. This does not impact the overall migration patterns.
We will also continue to make revisions to our provisional estimates. To support timely statistics we initially produce provisional estimates informed by assumptions based on past behaviours. This means we can provide an estimate of net-migration before the individuals have been in the country for a year (the definition of a long-term international migrant). As our assumptions improve the scale of these revisions should continue to reduce.
Next week we publish our latest long term international migration statistics for the year ending June 2025. Again, these will be provisional and subject to revision, but they offer the best and most timely insights into current migration trends.
Mary Gregory – director of population statistics