"White flight": evidence-based debate or headline-based evidence?
In our continued discussion of his book "The British Dream" in the London Review of Books, David Goodhart argues that what he describes as a "small number of factual errors", and I describe as his "slapdash approach to evidence and analysis", do not have "any bearing on the book's main arguments."
I thought I would elaborate here on one specific argument David makes where it is quite clear that his failure to do even the most basic homework does have a considerable bearing on an important argument, and one that goes well beyond David's deeply flawed book. This is the topic of "white flight" - a US phrase David seems determined to import into the UK debate.
Writing in the Financial Times, David said:
For it is important to understand that the proportion of white British Londoners fell so dramatically – from 60 per cent in 2001 to 44.9 per cent in 2011 – not only because of high levels of immigration but also thanks to a mass exodus of white Britons. Over the decade between the 2001 and 2011 censuses, the number of white British Londoners fell by more than 600,000 (17 per cent). That is about three times the fall over the previous census period, 1991 to 2001.
David didn’t stop with his FT article; he repeated these claims in a Prospect article, and in a Guardian video, where he described the "extraordinary movement" of white British people out of London. And his efforts to popularise the "white flight" factoid have, not surprisingly, had the desired effect in some quarters. The Mail tells us "How the rise of white flight is creating a segregated UK" while the Telegraph describes: "a continuing pattern of “white flight” from areas where indigenous Britons find themselves surrounded by new minority communities."
There is no doubt that London is undergoing rapid demographic change, and that this is a legitimate subject for both academic research and policy debate. But, as I have already pointed out, as yet there is actually little or no evidence of a "mass exodus of white Britons". In fact, the ONS data we have so far suggests that the number of people leaving London each year for the rest of the country (as opposed to moving abroad) has remained rather stable, at rather more than 200,000 or so a year, for the last 20 years. Meanwhile, somewhat smaller, but still very large, numbers of mostly white Britons continue to move to London from elsewhere in the UK. Indeed, the ONS even talks about a "general trend of a reduction in the net outflow of migrants for the London region [to other parts of the UK] over the last ten years."
But David goes from unsupported rhetoric to simple error when he asserts, as fact, that, comparing the 1990s with the 2000s, census data show that the net fall in the number of white British Londoners tripled. There is just one problem with this "fact". The 1991 census did not have a category for "white British". It had a category for white. Anyone familiar with – or who’d actually bothered to look at - the census data would know that.
So what do the 1991, 2001, and 2011 censuses actually tell us, and how does David come up with his numbers? Here's a table:
London population ('000s)
White White British White (not British)
5103 4288 815
4887 3669 1218
So, the white British population did indeed fall by more than 600,000 between 2001 and 2011. But the total white population only fell by about 200,000; on a simple like-for-like comparison, that is almost exactly the same as the fall between 1991 and 2001. The raw census data simply don't tell us what happened to the white British population between 1991 and 2001, because in 1991 there was no such category. David is comparing the fall in the white population (in the 1990s) with the fall in the white British population (in the 2000s), even though these two variables are quite different, both conceptually and in the data. David's figures would be correct if and only if the white (non-British) population hadn't changed at all between 1991 and 2001, which on the face of it is implausible.
What's David's response to this point? In his articles and in emails to me, he has cited only one source: Eric Kaufmann, Professor of Politics at Birkbeck College, who is working on an ESRC research project on this topic, and undoubtedly knows the data and evidence well; see his article here. But Professor Kaufmann responded publicly to David's most recent twitter debate with me:
Will know firm numbers soon. Data I have seen suggest white Other less likely to leave than white Brits...But there isn't "white flight". The pattern caused by "avoidance" destination of white Brits v others.
Professor Kaufmann is examining the longitudinal data from the Census; this will allow him to estimate the ethnicity breakdown in 1991 in a comparable way to the two later censuses. But, as he makes clear, he hasn't got, let alone published, any results yet. It is possible that when he does that they will support David's assertions at least in part. My expectation however is that while they will show some increase in the rate at which the "white British" population of London is decreasing, it will not be nearly as dramatic as David's tripling. But we don’t know yet. Equally importantly, Professor Kaufmann also directly contradicts David's "white flight/white exodus" explanation for the decrease in London's white British population; as he says, the factors at play are much more complex and explicit ethnic preferences are likely to be only a small part.
So where does that leave David? He has repeatedly asserted the existence of a phenomenon on the basis of statistics which are "not even wrong" [yet], since we don't and can't know the correct numbers, let alone the causal links. When he's challenged, he cites a recognised authority – who then contradicts both his numbers and his interpretation. And when he's caught, he dismisses this as "sniping in the footnotes", although his claims are clearly central to his argument.
So, given what we know now, his talk of an “extraordinary mass exodus” of white Britons from London is not a sensible contribution to an important policy discussion; instead, it distorts the debate and appeals to the worst elements of the media, who then further misrepresent the facts. If he does in fact have any respect for the evidence, and wants to make a constructive contribution to the debate, he will publicly correct the record. As always, I am happy of course to publish his response here.