Statisticians are questioning the accuracy of the official data that underpins UK government policymaking after the latest migration figures were significantly out of line with analysts’ expectations.
Last week, experts were left puzzled when the Office for National Statistics reported that net migration reached 606,000 in 2022. Though that figure marked a record for the UK, it also represented a relative levelling-off from the previous two quarters — a far cry from the leap analysts had projected.
While upward revisions to previous estimates contributed to the gap between expectations and the official figures, experts point to another factor: the ONS’s decision to reform the way it calculates migration data.
“The ONS is starting to transform the way these [migration] statistics are estimated and it’s very transparent about that, but is quite complicated and technical,” said Alan Manning, a professor at the London School of Economics and former chair of the Migration Advisory Committee, which advises the government.
“Has the headline figure just gone up because they’re changing the methodology or because of something underlying?”
The disparity between the official figures and analysts’ projections has sparked debate among experts, with some questioning whether the UK’s new methodology — which underpins government policy — is fit for purpose.
Net migration is calculated by estimating how many people migrate to the UK for at least a year and subtracting the number of long-term residents who emigrate overseas.
The ONS previously based its migration estimates on a questionnaire conducted in airports such as Heathrow and Manchester called the International Passenger Survey. However, it ignored important gateways in other parts of the country, such as Luton and Leeds.
The department has long admitted the approach was flawed. In 2019, the data was stripped of the “national statistic” badge of approval after the survey was found to have systematically underestimated EU and overestimated non-EU net migration.
Since April 2021, the ONS has produced an “experimental” migration data set using administrative data, such as visa and tax records, alongside statistical modelling and other surveys.
Madeleine Sumption, director of the Migration Observatory at Oxford university, pointed to potential limitations of the new approach.
“Administrative data sources have a lot of benefits, but they’re also less transparent,” she said. “There are lots of fiddly technical issues that affect the estimates in ways that are not always easy to predict.”
Uncertainty over the accuracy of migration data presents a challenge for policymakers and statisticians, who rely on it to produce population estimates in between censuses.
Simon Briscoe, a former specialist adviser on statistics for the House of Commons’ public administration select committee covering migration, said the “rough and ready” figures used under the new data-collection system are “all but meaningless”.
“To have sensible policies for filling gaps in the labour markets and making sure there’s enough staff in the NHS we need accurate figures and we haven’t got them.”
The Home Office declined to comment.
The ONS relies on a patchwork of data to estimate the headline migration statistic. Calculating non-EU numbers is easiest because people in this group require visas, whereas many EU citizens have settled status and so can travel without one.
However, the ONS must still adjust for the fact that not everyone issued with a visa should be counted in the statistics. Some visa-holders stay for less than a year, while others do not come at all.
Measuring emigration from the UK is also challenging because people may leave before their visas expire.
Recent changes to the methodology show how counting changes can substantially affect the headline number.
Ukrainian refugees, for example, were previously considered long-term migrants. But 39,000 people were removed from the latest estimates based on data that suggested some stayed in the UK for less than a year.
Measuring asylum seekers is also challenging as this group is typically not reflected in Home Office border systems data. They were included in the migration statistics for the first time this year, with figures based on the number of recorded claims.
The ONS recognised its methods were “clearly experimental and in development” but insisted it had confidence that the statistics were “the best estimates possible from the available data”.
It added that they were underpinned by many sources, including the 2021 census, using methods developed in consultation with experts in the field.
“These new methods are based on actual travel patterns, rather than the stated intentions of survey respondents. While this provides a more accurate picture of migration overall, it takes time to get a complete picture of people’s behaviour.”