Schoolboy errors in UCL immigration report

“Schoolboy errors” in UCL report claiming fiscal benefit to immigration

• Calculations discredited by fellow UCL professor of statistics
• Questions over prominence given by BBC to flawed report on immigration

A study claiming to show that EEA immigrants have made a net contribution to the UK’s public finances is fatally flawed, an in-depth statistical analysis published by Civitas shows.

The findings of Christian Dustmann and Tommaso Frattini of University College London’s Centre for Research and Analysis of Migration (CReAM) were reported prominently by the BBC last month and hailed as proof that immigration has a positive economic impact on Britain.

But an independent analysis of their reasoning and methodology demonstrates that they significantly overstated their case.

Mervyn Stone, Emeritus Professor of Statistics at UCL, and Civitas statistician Nigel Williams spell out in two complementary papers a series of flaws in Dustmann and Frattini’s findings:

• Their claim is based on large assumptions about contributions and receipts, particularly in respect of transport and housing;
• The actual difference in net contributions is a relatively small sum, which would vanish under a different treatment of a few assumptions, such as capital investment in social housing;
• They have used methods that the suppliers of the data specifically warned against;
• Their modelling of the receipt of benefits is based on an inadequate understanding of the structure of the original survey.

Dustmann and Frattini claimed to have calculated “precise” estimates for annual expenditure on immigrants and the revenues those immigrants have generated in comparison to native born workers.

Professor Stone writes: “If any honest statistician had made the same painstaking but assumption-based calculations, the last word he/she would have used to describe the estimates is ‘precise’ – unless exhaustion had affected judgement.

“Most of the underlying crude assumptions that the all-embracing approach has been obliged to make have not been subject to sensitivity tests that have might been made if the study had not been so obviously driven to make the case it claims to have made.”

Mr Williams says “the level of precision claimed is beyond what the data can reasonably support”.

“Superficially, understanding the give and take between native and immigrant populations looks like a helpful contribution to the debate.

“A closer look at the methods used in ‘The Fiscal Effects of Immigration to the UK’ by Christian Dustmann and Tommaso Frattini shows how many pitfalls are inherent in the attempt.

“The economic argument about the contribution of migration is more complex than an individual profit and loss account.”

Prof Stone and Mr Williams point out that Professor Dustmann was one of the academics behind the pre-accession projections of immigration from Poland and other Eastern European nations in 2004 which were proved entirely wrong by events.

That report, ‘The Impact of EU Enlargement on Migration Flows’ (2003), predicted an annual net inflow of between 5,000 and 13,000 individuals. In the event, more than a million have arrived in almost a decade.

Prof Stone was one of the first to raise the alarm over their findings, in another Civitas paper, ‘Prediction of future migration flows to the UK and Germany’ (2003).

Civitas director David Green said:

“The UCL study’s claims were held up by many as evidence that immigration from Romania and Bulgaria is unquestionably a good thing.

“But, as our forensic analysis by two respected statisticians shows, the study’s methodology was flawed and its findings were highly questionable. The authors made schoolboy errors in their handling of statistics which fell short of the professional standards recognised by statisticians.

“It is regrettable that even academics from an august institution like UCL can only see what they want to see.

“Yet again the BBC has failed in its duty to provide the public with impartial information. It has the resources to double-check conclusions based on statistical studies but instead seized on the conclusions to bolster its pre-existing prejudices.

“It was given headline prominence for 24 hours when many newspapers did not report it, no doubt because their subject specialists could see the flaws in the study.”

Notes

‘Plain Assumptions and Unexplained Wizardry Called in Aid Of “The Fiscal Effects Of Immigration To The UK”‘ by Professor Mervyn Stone can be accessed here and ‘Responding to “the Fiscal Effects of Immigration to the UK”‘ by Nigel Williams can be accessed here.

Mervyn Stone is an emeritus professor of statistics at University College London where he has been a teacher, researcher and statistical adviser since 1968. He has been editor of the methodology journal of the Royal Statistical Society and a member of its executive committee. His academic work has appeared in more than 100 papers.