Print
Share
Birth of Immigration Fiction: Watch a UK Minister Create an Economic Myth about Migration from Developing Countries
February 7, 2012
Some myths leave us to wonder who dreamed them up. Other myths we can observe as they are born. Last week a UK minister created an economic myth about immigration to his country, and it’s useful to watch how and why it arose.
First the fact: In January, an expert research team called the Migration Advisory Committee (MAC) released a rigorous review of the economic effects of immigration to the UK. The MAC found no association between UK unemployment and immigration to the UK from outside the EU, over the whole period 1975-2010. But when they separated out the last 15 years of that period, 1995-2010, the MAC found that there was a positive association between unemployment and immigration from developing countries in particular. During those years, for every 100 working-age non-EU immigrants who arrived in a particular UK region in a particular year, 23 fewer UK natives were employed in that region in that year.
But—as the MAC authors know—the fact that two things happen at once does not mean that one caused the other. If you get sick on the same day you purchase a toy, it does not mean that the toy caused you to get sick. Both the sickness and the toy purchase could have been caused by something else: visiting your sick nephew.
Likewise, no association between unemployment and immigration shows that immigration caused a single job loss for a UK native. For example, old businesses and industries are constantly declining and new ones arising. UK native employees in the old businesses might not be willing or able to work in the new ones, while many immigrants are willing and able—especially in times of rapid technological change, such as the past 15 years. A pattern like that could explain a positive association between UK native unemployment and immigration, by place and year, even if the true effect of immigration were to reduce native unemployment—for example, by raising the return to investment in new businesses.
The MAC researchers understand this well. They write,
“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” (p. 62).
Sadly the UK Minister of State for Immigration, Damian Green, saw the MAC report and badly misread it. He certainly has a responsibility to read every page of a landmark report so critical to his job. But he appears ignorant of the authors’ definitive statement above. Last week, in an important speech on the economic effects of immigration, he said:
“Do immigrants displace British workers in the labour market? The MAC research showed that in certain circumstances there can be displacement of British-born workers by non-EEA [European Economic Area] migrants, up to a level of 23 displaced for every 100 additional working age non-EEA migrants… This analysis gives us the basis for a more intelligent debate. It supports a more selective approach to non-EU migration. The old assumption was that as immigration adds to GDP—national output—it is economically a good thing, and that therefore logically the more immigration the better, whatever the social consequences.”
I’m sorry, Minister, but the report you cite simply does not say, does not find, and does not suggest that non-European immigrants “displaced” UK workers. It does not even find that such displacement happens “in certain circumstances” or “up to a level of” 23 per 100.
In fact, the report’s authors go out of their way to state that their evidence must not be interpreted as clear evidence of any displacement, any impact of migration on native unemployment. They do this because the authors know, and they unequivocally state, that the association they find is also consistent with several different economic phenomena that are not caused by immigration. They also know that non-EU immigrants are much less likely to be close substitutes for UK workers than immigrants from the EU; European workers tend to be similar to each other. This too suggests that the finding of zero association between UK native unemployment and EU immigration, and the positive association between UK native unemployment and non-EU immigration, both represent something other than ‘impacts’ of immigration on unemployment.
If that wasn’t clear enough, the MAC flatly concludes that the available data do not allow for an accurate accounting of the economic costs and benefits of immigration for UK natives:
“On balance it is clear that, on the basis of current data and knowledge, any attempt to calculate the NPV [Net Present Value] of migration policies will be subject to considerable uncertainty and likely biases” (p. 98).
But the minister’s myth propagates anyway, with help from a docile press. The BBC article on the minister’s speech, for example, simply quotes the minister’s false interpretation of the MAC report, without qualification. The article does not bother to interview any of the MAC report’s authors, who could clarify what they did or did not say. The BBC article does bother to interview anti-immigration activist Sir Andrew Green, who (shocker!) shares the minister’s sadly fictional interpretation of the MAC report.
So we’re likely stuck with this unfortunate, mutated thought, to be quoted over and over again by people who have various reasons to wish there were fewer foreigners around them. For years you’ll hear that a commission of top experts proved that immigrants massively displace UK native workers. Every time it’s said, it will be false.
What does the best economic research show? As I’ve discussed in a peer-reviewed article in a journal of the American Economic Association, barriers to migration from developing countries are far and away the most impoverishing obstacle to the global economy. Even slightly greater labor mobility out of developing countries would add trillions of dollars to the world economy, and most of those gains happen in countries of destination like the UK.
No economic effect of immigration on nonimmigrants comes close to meaningfully offsetting those gains, including in the UK (a good starting point, as highlighted by the MAC, is the work of Christian Dustmann at University College London). That clear finding of the best research is not called into question by the MAC report, and its authors know that. But for politicians and journalists to know that, they would have to read the whole report. And who has the time?
Update: To focus this post I did not address the question of whether or not the MAC report clearly establishes even the association between non-EU immigration and UK unemployment. Ian Preston of University College London insightfully critiques this aspect of the MAC report in this post from the Centre for Research and Analysis of Migration.
Possibly Related Posts
- Do the Gains from International Migration “Go to the Immigrants”?
- Celebrating America’s Long-Term Strength, Not Its Impending Death, on International Migrants Day
12 Responses to “Birth of Immigration Fiction: Watch a UK Minister Create an Economic Myth about Migration from Developing Countries”
Post a Comment
We value frank and constructive exchanges and encourage you to use your real name in your comments.
February 8th, 2012 at 1:40 pm
Thanks for the kind words about my comment.
The MAC report was, as you say, careful to disavow any causal interpretation to the associations reported though a few mentions of “displacement” might have been enough to encourage others to report the results less cautiously. The report was published on the same day as a paper from NIESR looking at the same question but using different administratively-sourced data. That paper found opposite effects to MAC at regional level and no effects at a finer spatial disaggregation, giving grounds for skepticism about the existence of any real persistent employment effects. Both of these were preceded in the same week by a publication by MigrationWatch, asking simply whether it could be a coincidence that immigrant numbers and the rise in youth unemployment were of such a similar magnitude. It is revealing to see which of these three got most press coverage.
Another myth took wing the following week. Damian Green, the minister whom you mention, coauthored an article in the Telegraph with fellow minister, Chris Grayling, revealing, without any of the careful contextualisation needed for interpretation, figures on numbers of immigrants claiming welfare benefits. As many have pointed out, these figures were compatible with immigrants collectively being net contributors to public funds and actually showed they were substantially less likely to be welfare-dependent than the UK-born. Nonetheless another myth was set in motion. Intervention from the UK Statistics Authority will do little to limit the myth’s spread.
The truth is that what we know about economic effects of immigration is known with little certainty but suggests, if anything, that it can be beneficial both to the immigrants and the country that receives them, albeit that some attention should be paid to addressing short term distributional impacts. Political debate is woeful at accommodating that sort of uncertainty particularly when there are people who want evidence to validate already strongly held views.
February 8th, 2012 at 2:42 pm
Thanks for reading the MAC report with such attention. As you say, the associations we found were mostly statistically insignificant and the results were sensitive to specification in some cases. The recent changes in the Points Based System anyway suggest future migrants will be more likely to have shortage skills and be complements to the existing workforce. My personal view is that there is much more to take further from the research the MAC commissioned and published on other impacts of migration, and it would be sensible to find ways of pinpointing any negative effects that could be identified – for example, is there congestion in the use of specific public services? It is unfortunate that the debate tends to focus only on searching for employment and wage effects when most research so far suggests these are small with non-obvious causal relationships. It may be that voter concerns relate to other impacts, but we do not have the data or research to know.
February 8th, 2012 at 4:46 pm
Diane:
Disentangling causal relationships with respect to public attitude formation is, of course, even more treacherous than dealing with other questions in this area, but we do have data sources. We have just published work (an earlier ungated version is here) on analysis of the European Social Survey that is at least strongly suggestive that, while economic concerns do matter, they are not as important as the other impacts you mention.
February 8th, 2012 at 7:52 pm
I’m very grateful to Ian Preston and Diane Coyle, two of the top experts on this subject, for their thoughtful comments. Coyle is one of the co-authors of the MAC report.
February 9th, 2012 at 5:48 am
“Docile media”!? Since when!? I think you’ll find that what causes UK politicians to spout this kind of stuff is how it will play in the highly un-contextualised tabloid press.
February 9th, 2012 at 8:39 am
Sad though this is to say, I am not greatly surprised at this outcome. Much as in science we are taught to leave any particular biases we may have at the door when analysing data, we also need to pay due attention as to for whom the report is intended. Many politicians are ideologues first and foremost (ditto some ‘journalists’). Therefore however many caveats you surround your conclusions with they are likely to cherry pick and distort. I would reluctantly recommend that any conclusions which seem to susceptible to idealogically-inspired misinterpretation should either be deliberately buried or phrased in such a way as to discourage their use. I struggle to think of an apposite example, but choices such as a double negative, active v passive voice, presenting findings under headings such as “No Causal Factors Identified” etc could all help. The blame for this episode lies with Damian Greene, but as scientists we can learn from this outcome too.
February 9th, 2012 at 10:59 am
MJ: I think your comment is critically valuable, thank you.
The world of private sector health has something analogous to what you suggest. In the U.S., pharmaceutical companies making claims for their products must have those claims approved by the Food & Drug Administration (FDA), or they must prominently label those claims as not having been evaluated by the FDA. In practice that means that the statements are being labeled as “causality demonstrated” or “causality not demonstrated”.
Stepping outside of regulatory analogies, related informal conventions have arisen in other areas, such as the near-universal use of the word “allegedly” in journalism. That wasn’t the product of legislation, as far as I know, but a convention that arose in the journalism profession in response to a lot of messy mistakes over the years, some resulting in lawsuits. But in the case of economic research, many journalists (no matter how smart and talented) simply might not have the training to assess whether a given technical analysis can or cannot be described as demonstrating causation.
That brings it squarely back to us researchers, as you suggest MJ. Especially those of us writing for a policy audience simply must embed qualifiers about causality into all of the takeaway sentences, particularly in the abstract and introduction—right there in the same sentence, to raise the cost of decontextualized cherry-picking. For example, “We find a correlation, some or all of which may reflect other economic forces driving both outcomes, between immigration and unemployment.”
Alternatively, I like your idea of a conventional phrase (playing a role like that of ‘ceteris paribus’, all else equal) that adds the appropriate caveat without too much clumsy wording. Focus-groups might be required to find a phrasing that would be clear to any educated person regardless of background.
I don’t believe that all this is a failing of the MAC authors; it is a generalized phenomenon, and certainly I myself have fallen short of this ideal at times. But it is time for a big change in the culture of what’s acceptable. And in this particular case, Minister Green either should have the expertise to distinguish correlation and causation transparently, or if he does not, he has the responsibility to gather around him a staff that does.
February 9th, 2012 at 11:22 am
Graham Brown: I believe that the BBC story had a responsibility, when reporting a research finding, to at least try to talk to the researchers in question. If they had, they would understand that the minister, willingly or not, misrepresented the research finding. Instead, they chose to interview an anti-immigration activist with little interest in research. ‘Docile’ is one of the nicer adjectives I could’ve chosen.
Too often, reporters covering a politically controversial issue feel the need to cover ‘both sides’, as if everything were a matter of mass opinion and no facts exist. In this case a clear fact exists: what the MAC report did and did not say. Reporters have a responsibility to get the facts.
February 14th, 2012 at 2:57 am
The reason why the Court of Appeal decided that the requirement was not binding was that it was not imposed by the Immigration Rules but by the Points Based System Policy Guidance which is not subject to Parliamentary scrutiny in the same way that the Immigration Rules are.
February 14th, 2012 at 4:04 pm
Michael: Good catch on this one. It is always fascinating to see how these stories get propagated in real time. Often harder to discover (and correct) in retrospect.
The comments suggest to me that in addition to properly writing and qualifying research findings, we need to call out those who mislead others on the basis of research – as you’ve done. But in the latter case, we have to raise the volume. I’d suggest a title like: “British Minister Misleads the Public and BBC Colludes.” Maybe some tabloids would pick that up and reduce the chances of propagating the myth!
February 15th, 2012 at 5:53 pm
The whole supposed debunking of the associations of immigration impacting native jobless rates are logical economics. Of course an influx of cheap labor will displace natives as employers dump natives for cheaper immigrant labor. The dissection of the details muddies the point that the opostion view suppoorting no impact of labor on natives from immigrants is a liberal plot to support the naturalization of 10 million illegal immigrants in the US. Naturally the liberal left want to show that immigrants help instead off hurt an economy to throw a cover over Obamas need to pull in more uneducated immigrants to tank the US economy and offer some form as socialism as the only way to salvage the problem. What amazes me is how anyone with any amount of wealth can undermine capitalism so brazenly.
February 17th, 2012 at 12:12 pm
If asking ministers to accurately report the content of scientific reports constitutes a “liberal plot”, then I am definitely part of the nefarious socialist conspiracy!
I’m grateful to “Real Truth”, above, for providing a great example of the incoherent idiocy behind much of the anti-immigrant movement. It’s hard to imagine a more pro-capitalist policy than relaxing tight government restrictions on labor mobility, but people who have other reasons to hate immigrants can magically convince themselves that relaxing massive government regulation is “socialist” and “anti-capitalist”—for reasons only accessible to the voices in their heads.
Sadly for purveyors of fear like “Real Truth”, the economic evidence contradicts his/her statement that “of course” immigrants do economic harm, just as the scientific evidence contradicts the view that “of course” orange juice cures influenza. I summarize the economic research literature on the economic benefits of migration in this peer-reviewed paper.