What happens when you misuse migration statistics

Listen to this article


Newsnight brings more support for this telling chart about immigration:

Britons want immigration reduced, though they are not as rabidly concerned about it as conservative tabloid newspapers would have us believe. At the same time, Britons do not trust the government to meet the Tories’ target for reducing net migration. Little wonder. The whole debate is marred by exaggerations and broken commitments that engender more cynicism. This makes politicians keener to appear tough … and to seize anything that looks like supporting evidence.

The BBC’s Chris Cook says that Downing Street is “suppressing” a report by civil servants that undermines the headline finding from a previous report, by the independent Migration Advisory Council in January 2012. This earlier report was used by Home Secretary Theresa May to argue that: “every 100 non-European working age immigrants were associated with 23 fewer British-born people in work”.

The government’s trouble is depressingly inevitable. The findings of the MAC paper were under-analysed and over-stated. This was not the fault of the economists who wrote the report; the MAC is clear about the study’s limitations and how it should be used. But rather than recognise caveats, the Home Office neglected them.

Here is the table presenting the study’s main findings:

The fourth column suggests a “displacement” rate of 23 native workers for every new 100 non-EU migrants over the period 1995-2010. Hence the quote from Ms May.

But if you read the MAC report, it is clear that the authors are keen for this finding to remain in context. Here are just six of important qualifications they used.

1. It is only one paper in an extensive literature: “Most studies estimating the impact of migrants on employment and unemployment in the UK find little or no association between migrant flows and changes to employment or unemployment.”

2. It shows correlation not causation: “Our findings should therefore be considered as estimating the association between migration and the native employment rate rather than the impact of migration on the native employment rate.”

3. The results are only relevant for a downturn, not when the economy is back to “normal”: “Nevertheless, our analysis shows a statistically significant impact of migration on native employment in an economic downturn but not in an upturn”.

4. The headline displacement effect may be temporary: “Other results in the Annex to this report tentatively suggest that it is only recent migrants (those who have been in the UK for under five years) that are associated with possible displacement. Those migrants here for over five years are not shown to be associated with any displacement of British workers”. (And if data from the year 2010 were removed then the displacement effect all but disappears, Jonathan Portes suggests.)

5. Remember outflows: “We find tentative evidence that any reduction in native employment associated with migrant inflows is equal to an increase in native employment associated with equivalent migrant outflows”.

6. The paper found no evidence of migration reducing wages: “It may be assumed that migration has no impact on average wages, but that migration increases wages at the top of the wage distribution and lowers wages at the bottom of the wage distribution.”

The “suppressed” report was a “comprehensive literature review”, according to the BBC’s Nick Robinson. If this is the case, no one in government should have been surprised that it undermined the MAC’s report headline result because the MAC report itself includes a comprehensive literature review! As per point 1 above, this review notes how the new MAC results diverge from the established evidence base.

Ultimately, when it comes to an issue like immigration, we put too much emphasis on reports with the government imprimatur. There has been extensive research on the topic and the state has no monopoly on the relevant data. Sadly, the main reason why we should pay special attention to government research in this field is because it is used by politicians to mislead a rightly sceptical electorate.

Copyright The Financial Times Limited 2017. All rights reserved. You may share using our article tools. Please don't copy articles from FT.com and redistribute by email or post to the web.