"One in ten new roles created in a year went to people born in Romania and Bulgaria", Daily Mail, 20 February 2014
"Romanians and Bulgarians have taken one in 10 new jobs in the past year as the number working in Britain has risen by 40 per cent." Daily Telegraph, 20 February 2014
Employment trends which consider nationality or country of birth cannot tell us anything about 'new jobs' or whether one group has gained more new jobs than another.
Guidance published in every release of the figures says as much.
Compared to last year, there are 42,000 more Romanian- and Bulgarian-born workers working in the UK. Overall, employment went up by 425,000. So these workers account for about 10% of the increase in employment.
But, the number of people in work is not the same as the number of jobs that exist. One person can have more than one job: that's why there are 30 million people in employment but as many as 32.5 million workforce jobs. So these figures can't show the number of new jobs.
Using a proportion can cause problems as well. Between the start of 2011 and 2012, total employment rose by 44,000 while employment for Romanian- and Bulgarian-born workers fell by 13,000. Using the logic of the papers today, this would mean that -30% of new jobs went to these workers, which doesn't make a lot of sense.
It's not a new problem: a few years ago when the mistakes were being made in the press time and again we asked the Office for National Statistics to introduce guidance on interpreting the numbers with every release.
Since this happened, until recently we saw far fewer misunderstandings. Now that Romanian and Bulgarian migration is hot on the agenda, now is as important a time as any to stop these mistakes becoming routine once again.