Receive free Employment updates
We’ll send you a myFT Daily Digest email rounding up the latest Employment news every morning.
This article is an on-site version of Martin Sandbu’s Free Lunch newsletter. Sign up here to get the newsletter sent straight to your inbox every Thursday
Greetings to Free Lunch readers whether you are on holiday or not. I will be away next week, but my colleague Claire Jones will keep you supplied with your global economy newsletter. I’ll be back in August.
For two years or more, policymakers have been complaining that labour markets are “hot”, “tight” and “strong”. This, they said, has led to problems, starting with “shortages” of workers in the early post-pandemic recovery (although I argued at the time that those were blessings in disguise) followed by inflationary pressures over the past year and a half. The main quantitative measures through which all these economic experts assess hotness, tightness and strength have been job vacancy ratios (high) and unemployment (low).
But these are not the only measures one could think of. Another obvious gauge of labour market strength is the employment rate. In both the US and UK, this has languished below or at pre-pandemic levels (depending on how you adjust for population). Far from hot or tight, in other words.
Not so in the EU, where employment and activity rates have broken old records. But there, too, the labour market pressure looks weak when measured by hours. In the eurozone, average hours worked have never recovered to pre-pandemic levels. (As for the UK, it has stagnated on both metrics.)
My colleague Delphine Strauss wrote a nice piece recently that highlighted the contrast between the US and the eurozone. As her chart shows (reproduced below), hours and jobs have risen together in the former, while hours have lagged behind jobs in the latter. Take the sharp rise in the total number of jobs in the US with a grain of salt, though, since it partly reflects population growth. The more informative employment and labour force participation rates (as a share of population) have not performed nearly as impressively, and less impressively than in the EU.
The similarity is that in both economic regions, average hours worked have been stagnating. In the eurozone, they peaked below pre-pandemic levels in the spring of 2022 and have been declining since then (although the most recent data now shows they rose in the first quarter of this year). In the US, they peaked a year earlier, somewhat above pre-pandemic levels, to which they have since slipped back.
So here is a question: why, at a time of supposedly excessive demand for labour, are people working fewer hours than they used to? Or from the point of view of companies, why are they not “sweating the intensive margin” — jargon for making existing employees work longer hours? And why have average hours been falling just when employers are being forced to raise wages — so we are told as an explanation of inflation — to get enough workers to meet the demand they face?
This matters enormously for the popular view that high price growth on both sides of the Atlantic is at least in part a reflection of excessive demand pressures, and that central banks must therefore crack down on job and income growth to bring inflation back down. You can only hold that view if you believe labour markets are indeed too tight. But a sustained decline in average hours looks decidedly like labour market slack.
Average hours are, in other words, a data point that should make us question whether we are not facing excess aggregate demand at all. It raises the possibility that we don’t even have enough demand to make people work all the hours they might. How could you have slack in the economy and yet such high inflationary pressures, you ask? But it’s not so hard to imagine. Go back to any earlier recession and suppose energy prices had then increased tenfold, as they did up to summer 2022. Of course, you would have had a bad patch of stagflation.
That, of course, is not the mainstream view. As I don’t subscribe to it, however, I am not well placed to defend it. But there are some obvious attempts one could make. A not very good one is to dismiss falling average hours worked because other signs, such as vacancy ratios and high wage growth, clearly show tightness. But high vacancy ratios could reflect fast job reallocation after the pandemic. Nominal wages, meanwhile, are rising less than prices, and falling real wages are surely no sign of tapped-out labour supply.
A better retort would be to say that lifestyle and health changes after the pandemic mean people want to work fewer hours — ie, a permanent fall in the labour supply. That same argument is used to explain why employment rates can’t rise further in the US and the UK even if they are below historic highs (even adjusted for ageing). But that has clearly not stopped employment from breaking new records in the eurozone, so surely one can’t explain shorter hours there either. And in the US, average hours seem to have fallen across the board, also in sectors with fast-expanding employment, such as transportation and warehousing (see chart).
If average hours are falling in many sectors, it complicates another explanation that is more plausible on the face of it. This would be that job growth has disproportionately happened in sectors where average hours are low for long-term, unlikely-to-change reasons. There could be some of this in the eurozone: Strauss points out that one-third of its job growth has been in the public sector, more than its share of total employment. (There are big differences — virtually all new jobs in Spain and Germany are public sector ones, while virtually all those in France and Italy are in the private sector.) Since hours are shorter in the public sector, this could explain the initial shift down of average hours. But since the initial pandemic shock, the periods when average hours stagnated do not line up well with when public sector job growth was ahead of the pack. Over 2022, public sector jobs grew no faster and often more slowly than employment overall — and the decline in hours that year happened across virtually all big sectors.
No doubt we can attribute some of the fall in average hours to each of these explanations. But there still seems to be something left over to explain. And that something looks very much like labour market slack.
