Thread by @M_Sumption

Important report today for @ONS provides revised figures for EU and non-EU net migration since 2009. Here's my take on what the key points are. ons.gov.uk/peoplepopulati… @MigObs

/THREAD/

It has been clear for some time that something wasn't quite right in the net migration stats. Non-EU net migration looked higher than expected and EU net migration looked lower, when you compare it to other data sources, such as the Annual Population Survey (APS) - see chart

There are various reasons this could be (e.g. different definitions used in different sources, migrants' uncertainty about whether they're migrating long-term) and @ONS has been working to understand exactly what's going on. The biggest step in that work comes in today's report.

They've drawn on some new data sources e.g. DWP data to get a better picture. And they concluded that the previous estimates were systematically underestimating EU net migration and overestimating non-EU net migration.

Exactly how much the official figures have underestimated EU net migration is not yet known. But this initial set of adjusted figures increase the estimate of EU net migration by 16% or 29,000 for the year ending March 2016, the most recent period for which adjustment possible.

The underestimate of EU net migration was mostly due to immigration from EU8 countries such as Poland. @ONS think this is probably because these migrants are often particularly uncertain about how long they're coming for, so might not be counted as 'long-term' migrants.

ONS has also concluded estimates of non-EU net migration have been too high, due to undercount of emigration of non-EU students. Adjusted figures non-EU net migration are 13% (or 25,000) lower than previous estimates in the year ending March 2016, and 8% or 18,000 lower for 2018.

What does this mean for overall net migration? Not that much, actually. The two errors broadly cancel out. Based on what we know from other data sources, the headline net migration figure seems to have been about right.

The adjustments @ONS has made today are important and give us a better picture of what's really going on, but it's not the end of the story. There's still an unexplained gap between IPS and other data sources, and @ONS is planning further work to get to the bottom of this.

In the meantime, they have decided to reclassify the migration stats as "Experimental Statistics" a label designed to communicate that the stats are still being developed, are subject to significant uncertainty and may change over time. So watch this space!