(Bloomberg) — Follow @Brexit, sign up to our Brexit Bulletin, and tell us your Brexit story.

Immigration is a divisive topic in the U.K., with the surge in the number of foreign nationals coming to Britain since the mid-2000s widely regarded as the driving force behind the vote to leave the European Union.

For almost a decade, the Conservative government has had a target to limit net migration — immigration minus emigration — to less than 100,000 a year. It has never come close to meeting it. Last year, the figure stood at almost 260,000.

But research published Wednesday cast doubt on the data used to gauge the flows of people in and out of the U.K.

According to the Office for National Statistics, official estimates of net migration from other EU countries have been too low. Using adjusted figures, it was around 16% higher than the published estimate in the year to March 2016, with the error relating to migrants from countries that joined the bloc in 2004, including Poland.

the same time, net migration from outside the EU has been overestimated — by more than 8% in 2018 — because more foreign students left than first thought.

The problem centers on the International Passenger Survey, which questions people at the main U.K. ports. The survey began in 1961 and its approximately 800,000 interviews a year are used to estimate the number of people arriving or departing for at least a year — the cut-off point for inclusion in the official count.

Statisticians have been criticized for relying on this IPS when other sources such as Annual Population Survey paint a diverging picture of migration over time.

The ONS said it had decided to downgrade the classification of its migration figures to “experimental.” New data for the year through March are due to be published Thursday.

To contact the reporter on this story: Andrew Atkinson in London at a.atkinson@bloomberg.net

To contact the editors responsible for this story: Fergal O’Brien at fobrien@bloomberg.net, Lucy Meakin, Mark Williams

©2019 Bloomberg L.P.