For reasons I have explained before, it is vital to know whether the mass unemployment we now have in Europe after the crisis is mostly structural or mostly cyclical in nature.
For the USA two recent blog posts (from Paul Krugman and Mark Thoma) show quite convincingly that unemployment over the pond is overwhelmingly cyclical, that is faster demand and output growth is needed to bring unemployment down, rather than labour market institutional reform in order to push, sorry, help people into jobs.
Krugman’s simple method of comparing the extent to which unemployment has risen by industry of previous employment is intuitive, and the finding that the increase is broadly the same across US sectors a powerful argument against those using anecdotal arguments about the difficulty of getting former construction workers to do, say, old-age care.
What about Europe? Well, I am sorry to say the answer is: we don’t really know. In fact all the macro signs (low core inflation, falling wage inflation) point in the obvious direction, namely that such a rapid and substantial run-up in unemployment can only be linked to the collapse of demand. Still, looking forward matters such as the sectoral distribution of the rise in unemployment are critical, and we would like to have European-wide comparable information on such issues.
But we don’t. (Or at least I can’t find it.)
Eurostat has a series, based on the quarterly labour force survey, on ‘previous occupations of the unemployed’. But it uses rather unhelpful broad categories like ‘professionals’ and ‘elementary occupations’, and, much worse, more than half – I kid you not – of the total is in the completely unhelpful category ‘unknown’. It is, in short, a fat lot of good.
We are left with employment data, which is available by economic sector (NACE). This is not satisfactory for the matter at hand because of the secular growth of the labour force and longer-term sectoral shifts. But ok: needs must… What do we see? Overall between the third quarter of 2008 and the same quarter of 2010, total employment in the EU (almost exactly 200m at the start) fell by just under 1.5%. (I am looking at adult workers 25+ as the relevant group for structural (un)employment.) Taking all those sectors which started out with at least 10m workers and subtract this average rate of change from the sectoral change, we get a measure of the relative ups and dows of employment since the crisis in the sectors that really count for employment (and unemployment).
There are two big losers: manufacturing and construction, with employment losses over the period of just over and just under 8% respectively more than the average. You can see this from a slightly different angle in the figure, which shows the absolute employment change for the 8 largest sectors.
Something over 3 million manufacturing jobs have been lost and 1.5 million construction jobs in net terms. That is not good, but it can hardly be considered an insuperable structural problem in an economy with 200 million jobs and over 22 million unemployed. Meanwhile 13 out of 21 NACE sectors (admittedly mostly smaller ones) experienced net employment growth over the two year period of the crisis. Some of these additional jobs will have been former manufacturing and construction workers. It’s that famously sclerotic European labour market at work again. Manufacturing employment, at least, has already begun to pick up. Once again it seems that the real problem is the sluggish pace (and fragility) of the recovery.
Anyway, my point here is not so much to argue the for the ‘cyclists’ over the ‘structuralists’ , but simply to point out that we need better labour market and particularly unemployment data at EU level. Any study along the lines of the American ones cited can at present only be done at national level in Europe. Such studies are valuable – and if you are aware of one for your country, please let me know – but raise comparability and of course aggregation issues. After all, we do have a (nascent) EU labour market and limited but increasing cross-border labour migration. We need timely and detailed data if we are to analyse labour market developments in Europe and make the right policy choices.