Who’s left?

There’s been plenty of anecdote about a Brexodus of EU nationals from the UK, what with all the uncertainty and hostility generated by the referendum and subsequent developments. And business responses to the government White Paper on a new immigration system seem universally alarmed at the prospect of proposals that might choke off flows into lower skilled (and/or lower-paid) jobs. So here are a few data-based observations of labour market changes over the past two years, largely from the Office of National Statistics Labour Force Survey.

Observations are made more difficult now by the public datasets having less detailed information than previously, with individual country of birth details no longer available for most countries. But usefully we can still distinguish Poland from the rest of the EU8 countries that joined the EU in 2004 and the EU2 (Romania and Bulgaria) who joined more recently. The EU14 can’t be very usefully broken down as only Ireland can be separately distinguished.

Looking simply at ‘Western Europeans’ and ‘Eastern Europeans’, the numbers working in the UK do appear to have been differently affected by the referendum. Western Europeans don’t appear to have been affected at all. Of course, a flattish series doesn’t mean that no one has left, but it does mean that for anyone stopping work in the UK, someone else has started. Caveats always apply re Labour Force Survey sample size and confidence intervals, but at this level of aggregation they aren’t material. On the other hand, the number of Eastern European workers climbed by a not inconsiderable 200,000 from early 2016 to a peak in Jul-Sep 2017 before gently giving up much of that gain over subsequent quarters.

When looking at the mix of Eastern European workers, the picture is a bit different. On these data, the number of Polish workers peaked just after the referendum and fell from then until reaching a stable level at the beginning of 2018. But this fall was more than made up for by an increasing number of Romanians and Bulgarians which growth tapered off during 2017 but ticked up again in 2018. Other Eastern Europeans haven’t varied by anywhere near as much. So compared to the beginning of 2016, while overall both Western and Eastern European numbers are essentially unchanged, a loss of Poles has been balanced by a gain of Romanians and Bulgarians. So far so anodyne. But if one group is simply replacing another, what’s all the bleating from business about?

I’ve looked at this from a number of angles, and from the perspective of occupational groupings the picture is quite interesting. Compared with the beginning of 2016, EU employment has fallen in Skilled Trades and in Elementary Occupations. It would have fallen in Process, Plant and Machine Operatives and by much more than it did in Elementary Occupations (in both cases driven mainly by falling numbers of A8 workers) if it had not been for increasing numbers of Romanians and Bulgarians. On the other hand, employment in the two highest-skilled occupational groups increased by nearly 100,000.

The picture changes a bit if we start instead at the point of peak Polish employment just after the referendum and also disaggregate Poles from other A8 workers. From their peak, the number of Polish workers has dropped by some 160k, or 22%. This really is quite substantial. The number of workers from other A8 countries dropped by less than 30,000 or 8%. The losses are again heavily concentrated in three occupational groups, and although made up for to some extent by Romanians and Bulgarians in the lowest-skilled sectors, the Skilled Trades Occupations saw nothing but reductions. Now these tens of thousands aren’t necessarily anything to get excited about when there are millions in each occupational group. However, over this period there was a reduction in the number of people working in Skilled Trades Occupations of just over 1% which is entirely accounted for by there being fewer Eastern Europeans working in the sector.

For context, we can look at how these changes fit into overall change in employment in the UK. Over the past two and a half years all growth has been concentrated in (essentially) higher skilled work and ‘office jobs’. Employment has fallen in lower skilled and ‘hands-on’ occupations. In Skilled Trades Occupations and Process, Plant and Machine Operatives, the entire drop is attributable to fewer EU workers.

So, has there been a big Brexodus? Well that depends on how full or empty you like your glass. If you like it half-full, then at the whole economy level, the number of EU workers hasn’t changed much. However, if you prefer it half-empty, there are clear signs that more people have been leaving the UK from some sectors than have been replaced in these sectors by other new EU arrivals (as it seems unlikely that individuals will have been moving from elementary occupations to being Managers and Directors). The implication is that while higher-skilled jobs in the UK continue to be attractive to people from around the EU, lower-skilled ones have become less so. This seems to be particularly true for people from Poland. Of course, this isn’t necessarily anything to do with a newly hostile or unwelcoming environment. If that were so, one would expect all groups to change similarly as the ‘unwelcoming’ would likely be as ill-disposed to Lithuanians, Hungarians et al as to Poles.

Instead it seems rather more likely to reflect changes in the Polish economy than anything else. Indeed, the change in the number of Polish workers in the UK from just before the referendum rather neatly tracks (inverted of course) the changing rate of GDP growth in Poland.

And not to forget either the strengthening of the Zloty against the pound.

Obviously the pace of growth in the Polish economy is not as such anything at all to do with Brexit (although the currency movements are). The fact that prospects in Poland for Poles might be booming may well be making life more difficult for some UK employers in specific sectors but that seems to be no more than a function that the employer offer just isn’t as attractive as it once was. UK employers have been very happy to take advantage of differentials in job availability and wage levels between the UK and some other EU countries, but those differentials were bound to narrow sooner or later whether the UK stayed (or stays) in the EU. I’m afraid my rather harsh judgement is that anyone who’s been caught out hasn’t been forward-planning enough.

Edited: Removed a sentence left over from a earlier draft looking at a different period so didn’t match the chart to which it referred!