Recent EU migration to the UK: can we trust government numbers?
The Prime Minister’s claim that “around 40 percent of all recent European Economic Area migrants are supported by the UK benefits system” has been widely criticised (see my brief explanation here) and the UK Statistics Authority has made clear that it is very displeased that Number 10 chose to use unpublished (and inconsistent) statistic (which DWP’s post-hoc publication doesn’t even come close to validating).
But, leaving the abuse of the statistical process aside, the reason the Prime Minister came up with a number that is much higher than most researchers think credible is interesting – and potentially very revealing. In order to arrive at the “around 40 percent”, Number 10 took the number of “recent EEA migrants supported by the UK benefit system” from the DWP and HMRC computer systems that record tax credit and benefit claims and also PAYE and NI contributions. These are linked to National Insurance numbers, which in turn record nationality at time of issue, when they were issued, and when the person to whom they were issued arrived in the UK. Leaving definitional issues aside (the government appears to be counting at least some UK-born children) this seems a sensible place to start.
But in order to estimate the actual number of “recent” EEA migrants - the denominator in the “40 per cent” calculation - Number 10 used the Labour Force Survey (LFS) – an entirely different and survey-based source. According to the LFS, there are about 525,000 recently (in the last 4 years) arrived EEA nationals. Why not use one source for both numbers? Well, if we use the LFS for both numbers, we know what happens, because external researchers have done it. We get a much lower figure than the 40 percent.
But what about if we use the administrative data from DWP and HMRC – which, in principle at least, are far more comprehensive? How many recent EEA migrants are there in the country? Is it similar to the 525,000 number? We don’t know, because the government won’t tell us. But the government does know this number, or at least is in a position to know this number. Precisely the same computer systems which it used to calculate the number claiming benefits will also tell us how many are paying income tax or National Insurance; together with those claiming benefits, this should cover the substantial majority of those who are indeed still here. And we know the government has the capacity to perform this analysis, because it has denied various related Freedom of Information requests (mostly from the indefatigable Michael O'Connor) on grounds of cost or (spuriously) confidentiality; but has admitted that it does indeed hold this information.
And what would that tell us? Well, in the last 4 years, nearly 2 million EEA nationals registered for National Insurance numbers in the UK. Now that number is not directly comparable to the 525,000 for a number of reasons, most importantly that we don’t know how many of the 2 million are still here. Does this mean the LFS is substantially underestimating the number of EU migrants? Again, we don’t know. But a possible clue is found by comparing National Insurance number registrations with the official immigration statistics (from another survey, the International Passenger Survey). This chart shows, for the EU 15 (the EU pre-2004) and for the EU2 (Bulgaria and Romania) the divergence between the two series – which was always there but has grown considerably in the last few years.
So what would happen if the government revealed how many EEA migrants are, according to its own computer systems, actually active in the UK labour market, one way or another? It seems highly probable that we would learn two things. First, that the Prime Minister’s figure of “around 40 percent” of recent EEA migrants claiming benefits of some kind is considerably exaggerated. Second, and perhaps even more interesting, that there are actually considerably more such recent migrants than the official immigration or labour market statistics actually suggest.
Neither piece of information, of course, would be particularly helpful to the government at the present time. It would tell us both that EU migration is an even bigger issue than the official statistics suggest - and that the government's preferred solution is even less relevant than we already thought. But that is not a good reason for concealing the facts.