Migration statistics, continued. How many migrants are still here?

Last week’s ONS publication on migration statistics threw some light on the debate, although they generated even more heat – and frankly some flat out lies, as I explained here. However, this Wednesday, we’ll get a different set of numbers; this will focus not on how many migrants are entering or leaving the country, short or long-term but how many actually live and work here – which is after all most of what matters.

The ONS Labour Market Statistics, published every three months, set out the headline figures for employment, unemployment and so on, based on the Labour Force Survey (LFS), a large survey designed to be (broadly) representative of UK households. Every three months, including this one, the ONS publishes a more detailed analysis of the data, looking specifically at “employment by nationality and country of birth“.  It is these statistics that have been repeatedly misunderstood and misrepresented, by everyone from the Daily Express to the Prime Minister, to make the entirely false claim that “all new jobs went to immigrants” (or some variant of this).

The most recent statistics, published in February, showed that there were 2.1 million people born elsewhere in the EU in employment in the last quarter of 2015 (out of a total of 2.6 million EU-born of working age counted by the LFS; with an employment rate of about 80%, the EU-born are significantly more likely to be in work than the UK-born).  This represented an increase of about 200,000 year-on-year, while employment for the UK-born grew about 250,000 (these figures are not seasonally adjusted, so year on year figures are more useful).

But are the headline totals accurate?  Perhaps the most interesting single figure in last week’s ONS release was this one buried right at the end of Annex 3 -finally, at least a snapshot of the data I’ve long been seeking, and that the government has long been refusing to provide, on “active” National Insurance numbers:

“HMRC also shared with us analysis of those who had arrived, or registered for a NINo in the 4 years, to 2013-14 and were subject to income tax NICs and/or received HMRC benefits at some point in 2013/14 – the latest year available. They found 1.0 million such individuals were from the EEA and 0.4 million were from non-EEA countries.”

We should in principle be able to compare this number to the LFS, with some adjustments, since the LFS also asks immigrants how long they’ve been in the country. And being subject to income tax, NICs or receiving HMRC benefits (mostly tax credits) ought to be a pretty good proxy for being in employment. But the LFS data tells a completely different story to the HMRC analysis.  DWP’s analysis of LFS data, in March 2013, estimated that there were 390,000 EEA national adults of working age resident in the UK who had arrived in the UK within the previous 4 years. Looking at 2014, Migration Observatory came up with a similar number (307,000, but excluding Bulgaria and Romania).  And both those numbers actually include people who will not be counted in HMRC’s 1.0 million (those claiming only DWP benefits and some non-working spouses, for example).

Now it is true that the HMRC data runs over a whole year – and so will include a significant number of short-term migrants and others who will leave in the course of a year – while the LFS data is a snapshot of people who are long-term residents at any one time. But even on stretching assumptions I don’t think this could reduce HMRC’s 1.0 million figure by more than perhaps 300-350,000 (while short-term migrants are a significant proportion of total inflows, maybe about half in 2014, they are a much smaller proportion of the total stock of migrants, even relatively recent migrants, in the country at any one time).

Another way of looking at this discrepancy is to note that HMRC also estimated that these same migrants paid more than £3 billion in income tax and National Insurance during 2013-14. If the LFS were correct this would suggest recent EEA migrants paid considerably more – on an annualised basis, so allowing for the fact short-term migrants are here short term –  than their proportionate share of direct taxes, which is not consistent with what we know about the occupational distribution of recent European migrants.

So what should we conclude from this?

First, it seems implausible the LFS numbers are correct. I think there clearly are more recent EU migrants present and active in the UK labour market than suggested by the official statistics.  By how much, we don’t know. Nor do we know why.  But, on the basis of what we know now, this seems to be a bigger discrepancy, and one that is more difficult to reconcile, than that found in the migration statistics themselves.  Note that in itself this doesn’t mean the overall numbers are wrong – perhaps new migrants are misclassifying themselves as having been here longer – but it is cause for concern.

Second, as I’ve already noted, this puts to bed the Prime Minister’s dodgy claim, back in November, that “around 40 percent of all recent European Economic Area migrants are supported by the UK benefits system”.  This was based on an estimate that about 130,000 recent EEA national adults were claiming benefits in March 2013; given the number recorded as being “active” at some point in 2013-14 reached a million, his 40 percent looks like a gross exaggeration.

Third, this doesn’t in itself change the broad assessment of the labour market impacts of EU migration, summarised by my LSE colleagues here. Although much of the research they discuss is based on LFS data, by no means all is – in particular, as they note, recent analysis by me using National Insurance registration data confirms the broad picture of no significant impact on employment, and small if any impacts on wages.  But it does perhaps mean we need to pay more attention to the impacts of short-term migration – which the LFS, by construction, cannot tell us.

Finally, though, it does raise some broader issues. Unlike the immigration statistics – which are used for only limited purposes – the LFS is used for everything from the headline unemployment statistics to a wide variety of academic research on labour market issues, as well as being an international standard. If it’s wrong or simply giving us an increasingly incomplete picture, we (and ONS) have a problem.

I have submitted a number of further FOI requests to HMRC to try to shed some more light on this; so has my indefatigable colleague in these endeavours, Michael O’Connor, who points out that there is a whole wealth of detail in their data, not yet published, which could throw further light on all this. Let’s see if we get a response. And I’ve asked the UK Statistics Authority for a view.