‘Where are all these East Europeans flocking from?’ That was the question Gillian Duffy asked Gordon Brown in 2010. Brown would have got into trouble if he’d answered ‘Eastern Europe’, though probably not as much as he got into for describing Duffy as ‘just a sort of bigoted woman’. Her prejudices, or fears, are widely shared. Nearly 150,000 people have signed a petition against the lifting of labour restrictions on Romanians and Bulgarians next year. According to the petition, ‘there is currently an estimated 1.5 million people seeking work within the two countries.’ But how many of them might actually move to Britain?

The British government has declined to venture any figures. But according to a BBC survey published last month, only 1 per cent of Romanians between the ages of 18 and 60, and 4.2 per cent of Bulgarians of the same age, said they were looking for work in the UK.

Migration Watch UK, an ‘independent think tank’, swooped on the figures: ‘This is a stunning survey, which the BBC has rather desperately tried to play down. The percentages look small but when multiplied by the size of the work force they produce large numbers.’ According to Migration Watch, 150,000 Romanians and 200,000 Bulgarians could be said to be ‘actively considering’ looking for work in the UK after 2014. These estimates were reported by the Telegraph, which attributed them to the BBC. Nigel Farage also accepted them, and accused the BBC of ‘influencing the debate’ rather than ‘reporting the facts’. Further prime-time 'debate' ensued. But no one seems to have looked into how Migration Watch came by their numbers, or how accurate they are.

If you take the latest (2011) World Bank estimate of Romania’s population, 21.38 million, and their figure of 69.8 per cent for the working-age population, then divide that by 100, you get a number slightly lower than 150,000. Data from Romania’s National Office of Statistics for 2011 gives you roughly the same number. But there are two problems with this.

First, the BBC survey asked people between the ages of 18 and 60, not 15 and 64: a demographic of 12.63 million, by their count. Second, the data is now two years old, so further decline should be factored in. The BBC figures assumed a total population of 19 million, as estimated in a recent press release giving preliminary results of the 2011 census. This constitutes a drop of more than 10 per cent since 2002. In fact only 18.3 million were actually present in Romania, with 3.5 per cent ‘temporarily absent’. The population is also declining naturally, as the so-called decreţeii – the boom children ‘born by decree’ as a result of Ceauşescu’s pro-natal edicts of the 1960s and 1970s – reach middle age and have fewer children of their own. All this would probably leave 1 per cent of 18 to 60-year-olds in Romania in 2014 more in the region of 120,000.

I wrote twice to Migration Watch about their forecast without receiving a reply. I then wrote to their demographic consultant, David Coleman – the only member of their advisory council with relevant professional expertise – who confirmed he had had no involvement in the preparation of the figures, and did not know whether any demographers were involved. Like the Telegraph, he attributed the numbers to the BBC, not Migration Watch.

When I finally spoke to a representative of Migration Watch on the phone, he conceded that the population of Romania was in flux, and also that a survey of a younger generation – say, people from 18 to 40 – might be a more relevant indicator than one of people aged 15 to 64.

Predicting migration flows is not an easy game. But for a self-styled watcher’s estimate to be out by around 20 per cent purely on documentary or arithmetical grounds must raise questions. It is disheartening to see an organisation with no clear accredited expertise being accorded so much authority.