Douglas Carswell

Immigration and dodgy data?

Over the past decade or so, immigrants have paid billions of pounds more in taxes than they have cost the public purse, according to a report by the Centre for Research and Analysis of Migration (CREAM). And, what's more, immigrants are 45 percent less likely to claim state benefits.

Yet more justification for all those "Britain needs more immigration" articles one finds in the Economist, perhaps?

Except on reading the CREAM report, I discovered errors, guesswork and oversight.

First, an elementary error. When calculating the fiscal contribution made by migrants, the report's authors try to work out how much tax migrants pay. In doing so, the report's authors seem to count the £10 Billion business rates paid to the Treasury each year as if business rates were a tax paid by every self employed person in the country. It isn't. Most of that £10 billion tax contribution comes from big companies.

Crediting every self employed person with a £2,500 tax contribution that they do not make is not only wrong. Because a much higher proportion of EU migrants are self employed, it seriously distorts any net assessment of the overall financial contribution they make.

Then there is the guesswork. When trying to work out how much corporation and capital tax migrants pay, the report appears to make assumptions that look to me like a guesstimate.

Then there is the oversight. The report tells us that migrants are 45 percent less likely to claim benefits. Really? If you only look at Labour Survey data it is possible to draw that conclusion. But not if you look at HMRC data, which suggests that migrants are 20 percent MORE likely to claim certain benefits.

Is it not a bit odd that a serious research unit, associated with University College London should produce a report that does not appear to understand who pays business rates? Not nearly as odd as the reaction I provoked by pointing this out.

Writing for the Independent on Sunday, David Blanchflower dismissed what I had to say as the "visceral response" of a "Right-wing Tory MP". He did not, however, get around to countering the points I had actually made.

It was enough for Blanchflower to simply assert that "Christian Dustmann is definitively the number one expert on migration in the UK". While Carswell, on the other hand, is a mere "amateur".

Indeed I am.

But isn't it all the more shocking that a mere "amateur" can expose flaws in a report produced by the "number one expert on migration"? No wonder UK immigration policy seems to be such a mess.

"Be fair, Carswell" I hear you say "it's jolly hard to work out how much migrants pay in tax".

I agree. Hideously complicated. So much so that I believe we cannot possibly claim to know for certain. Which is my point.

Responding to my critique, Christian Dustmann did not rebut the substantive points I raised. Instead he said that it was a "well-established" way of doing it. Maybe the fact it is so "well-established" helps explain why so many similar reports have tended to draw the same kind of conclusions? That does not prevent it from being wrong.

He went on to accept that the data on which his report was based was "not perfect". "Assumptions" he wrote "have to be made". Quite. So why make such big and bold announcements about the impact of immigration if the data is known to be so imperfect?

It was, you may recall, Professor Dustmann who once "predicted that opening UK borders to 10 new EU countries in 2004 would increase the population by 13,000 a year". Data can indeed be imperfect.

Back to all posts