By Scott Blinder, University of Oxford
The immigration debate isn't a great place to look for rational or factual arguments at the best of times, and new rules concerning Romania and Bulgaria have spurred a new round of evidence-free speculation. As of January 1, people from both countries will be able to live and work in Britain and other EU nations as and when they please.
Given memories of large-scale - and vastly underestimated - migration from Eastern Europe to Britain after EU enlargement in 2004, many members of the British public are understandably anxious about the potential for a repeat. The government and much of the research community have been reticent to make predictions - also for understandable reasons. But this has left an opportunity for anyone willing to claim that they can forecast the future.
Last week saw a particularly unfortunate example. A report by the Democracy Institute, an American libertarian think-tank, predicted that 385,000 people will migrate from Romania and Bulgaria to the UK over the next five years.
This prediction was uncritically featured or cited in an array of newspapers including the The Times, The Daily Telegraph, the Daily Mail and the Daily Express. But the report is deeply flawed and should be dismissed as not credible until its authors can prove otherwise, as it gives no information about the methods by which the prediction was generated. For the British press to continue using the prediction, giving credibility to the dubious report, would be both misleading to the public and damaging to public debate.
To be clear, the problem with the report is not that it predicts a high number. The issue is the lack of transparency about its methodology. The prediction comes from a self-described "proprietary econometric migration model" - in other words, a secret method. This means that there is no way of knowing where the number came from or of scrutinising the approach taken to generate it.
Secrecy in method is contrary to the very idea of research, and devalues attempts to use scientific or social scientific inquiry to inform public debate and public policy. Transparency is a necessity, as it allows other researchers, as well as policy-makers, civil society groups and interested members of the public, to evaluate the quality and credibility of any piece of legitimate research.
Until the authors of the Democracy Institute report show how they have arrived at their conclusions, their "estimate" has as much value for public debate as a number pulled out of a hat, and should be treated as much. Indeed, for all we know, their "proprietary econometric migration model" might involve a large hat with numbers in it. The "proprietary" nature of it means that we will never know.
The most commonly cited prediction of the number of migrants that may come to the UK from Romania and Bulgaria is the central estimate of 50,000 per year suggested by Migration Watch.
I have previously argued that it is not possible to accurately predict Romanian and Bulgarian migration to Britain post-2014 given the present state of knowledge about what determines migration flows, and that it is not very useful to try. I therefore do not have a view on whether Migration Watch, Democracy Institute, or Romanian and Bulgarian officials with lower predictions are most likely to be proved correct.
However, in contrast with the Democracy Institute, the Migration Watch report has the important merit of transparency about its methods. Their published paper outlined the assumptions they made to come to their figure, and other people and groups have been able to critically evaluate their work. The report also is frank about being the result of the authors' considered judgements rather than a forecasting model.
The Democracy Institute, however, makes no such effort at transparency, and instead uses the language of science and expertise ("econometric migration model") to disguise a complete lack of public evidence for the number they have injected into Britain's public discussion.
This is a report that presents the British people with a number that cannot be scrutinised or tested. Until the authors show their work, 385,000 migrants per year represents a guess masquerading as a social scientific estimate. A press and political class concerned with evidence-based debate should avoid presenting it as anything more substantial than that.
Basing policy-making and public debate on evidence is a worthy goal, and the immigration debate in Britain is often highlighted as in particular need of an increase in the ratio of evidence to assertion. But for this to work, it is important to distinguish more and less credible contributions, and to put no stock in those who claim the language of research but present nothing of the kind. Citing a number and claiming to have model is not a sound basis for intervening in Britain's immigration debate.
Scott Blinder does not work for, consult to, own shares in or receive funding from any company or organisation that would benefit from this article, and has no relevant affiliations.
Follow The Conversation UK on Twitter: www.twitter.com/conversationuk