British jobs for foreign workers
What puts more pressure on British jobs - migrant workers coming from the European Union or from outside the EU? This matters, of course, because workers from beyond the EU are affected by the immigration points system and potentially by Conservative proposals for a cap. Workers from inside are free to come and go.
I think I may have a few clues.
Last week I quoted official figures showing that, in 2008, 67,000 non-EU workers came to the UK and 99,000 EU workers. On the face of it, this suggests the effect on jobs is greater from inside rather than outside the European Union.
But the stats only tell us how many workers came into the UK - not how many others left. It is the overall figure, the net number, which matters. After all, if fewer EU workers arrived than departed and more non-EU workers came than left, the picture would be completely different.
The net statistics I quoted last week are not as useful as they might be because they measure people's reasons for migrating rather than accurately tracking the arrival and departure of foreign workers.
The hunt is on, therefore, for a better way of identifying trends. With the help of a friendly chap at the Office for National Statistics, I was guided to a table within the Labour Force Survey , "Employment levels by country of birth and nationality".
The figures show the number of UK and non-UK workers in the labour force and, within that, the number of EU workers with jobs in Britain. It is possible, therefore, to identify UK, EU and non-EU workers and to see how the picture is changing.
The latest data show that, comparing the last quarter of 2009 with the same period in 2008, there are 76,000 fewer non-EU workers and 13,000 fewer EU workers employed in the UK. The EU change is not large enough to be statistically significant - in terms of a trend, the picture is stable.
What does this mean for British workers - after all, fewer of them are employed too? Well, the proportion of all jobs occupied by UK citizens has risen - very slightly. At the end of 2008, 91.9% of jobs were held by Brits. At the end of 2009, it was 92.1%.
Interestingly, while the proportion of jobs held by non-EU workers (the ones affected by the immigration points system) has fallen a tad - from 4.6% to 4.4% of British jobs, EU workers' share of the jobs has gone up slightly - from 3.4% to 3.5%. (The figures don't add up exactly because of rounding.)
Even though the number of EU workers with British jobs has fallen, their share of the UK jobs market has risen a little as the market contracts. Again, this effect may not be statistically significant.
But this is all a far cry from the story in 2002, before the big increases in immigration from central Europe and beyond kicked in. Back then, 95.1% of British jobs were held by British workers. Around 3% were held by non-EU citizens and 1.9% by EU migrants.
The year everything really changed was 2007. The economy was expanding and about 615,000 new jobs were created. Who got them? Roughly 200,000 went to non-EU workers; 275,000 were taken by EU migrants; the remainder, about 140,000, went to Brits.
Looking at 2009 again, one can see that all three groups lost jobs. UK workers lost the most jobs - 335,000 - but things look rather different when you analyse it in percentage terms.
Among all UK workers, 1.3% lost their jobs in 2009. Among EU workers, 1.2% lost their jobs, a very similar figure. But among non-EU workers, 6% were made redundant in that year.
So what can we conclude in terms of where the pressure on British jobs comes from most? I think the figures suggest that the effect of non-EU jobs is declining, but EU migration's effect is broadly stable.