Happy 2023 and welcome back to Free Lunch. I hope all readers had a restful break.In my last column before the break, I warned central banks against seeing fast wage growth as necessarily presenting an inflationary danger that calls for tighter monetary policy to restrain jobs and income growth. It could instead reflect a more competitive labour market — more competitive for workers, that is. If more workers than before are shifting from worse-paid to better-paid jobs, then wage acceleration is a welcome indicator of an equally welcome reallocation of labour towards more productive activities. (After all, the employers to whom workers are switching could only pay those higher wages if productivity justifies it.)I could only refer to it in passing in the column, so here I want to give more credit to the excellent recent research suggesting that this is precisely what is going on, at least in the US. Last month David Autor of MIT presented the findings that he, together with Arindrajit Dube and Annie McGrew, have gleaned from US population survey data — you can watch his presentation for yourself here. I want to highlight four of the most telling graphs from the slide deck.First, wage growth has been much stronger for the lowest-paid since the start of the pandemic, sharply reversing decades of rising wage inequality:
This recent wage compression is broad-based: it has taken place between occupations, between the young and the old, between those with less and more education, and to the advantage of minorities. Second, even though inflation is high, the lower-paid have still seen real wage growth:(This is true even for the shorter period of just the most recent 12 months.) Third, people are shifting between jobs much faster than before the pandemic:
And job mobility has increased in particular among young workers with little formal education; ie those people most likely to have formerly been stuck in bad, poorly paid jobs.Fourth, by far the biggest acceleration in wage growth is among those who switch jobs rather than those who stay in place:
Note that the chart shows two separate things: that wage growth is always higher for job switchers, and that this advantage over job stayers has roughly doubled in size in the current strong labour market. This should make us rethink the standard story we are told about a dangerously “tight” labour market. For one, common indicators of overheating may not be saying what we think they are. In particular, high vacancy rates may not be the sign that excessive demand puts upward pressure on prices, but rather reflect more footloose workers (especially low-wage ones). After all, the more workers move, the more often you would expect employers to look for new staff. So we should expect a higher vacancy rate for any given state of aggregate demand. (In fact, alternative measures of vacancies suggest the US labour market is less “tight” than it appears on the conventional yardstick.)More fundamentally, if greater job mobility makes for higher productivity — as Autor and his colleagues say it theoretically should — then the current labour dynamics should be expanding the economy’s capacity to produce. That would be a force for lower, not higher prices — and thus a reason for central banks to relax rather than tighten monetary policy.This speculation, however, runs into the fact that, so far, a boost to productivity is hard to spot in the numbers (unlike early in the pandemic). As a recent New York Times story shows, many companies find that higher staff churn temporarily lowers productivity because more time needs to be put into training.But the key word here is “temporary”. Look at output per hour worked in the US in the chart below: it fell in the first two quarters of 2022. But that fall came after a spike in the early pandemic that lasted for more than a year. (Productivity edged up again in the third quarter of 2022 on a whole private-sector basis but slipped further for non-financial corporations.) So examine how productivity behaved over the whole pandemic period, including both shutdown and recovery. Taking the past three years of available data, from the third quarter of 2019 to the third quarter of 2022, non-farm business output per hour worked grew by 1.6 per cent annually (1.3 for the non-financial corporate sector). That was about the same productivity growth rate as in the preceding three years, and faster than the average rate in the preceding 12 (a period encompassing the previous big crisis). So productivity remains at or above the pre-pandemic trend. Given all the disruptions over the past three years, that is a strong record.I caught up with Dube, one of the researchers, to hear more. (Free Lunch has previously featured his work on minimum wages and the US’s pandemic-era supplementary unemployment payments.) He said their interpretation of the data was indeed that workers were moving from lower- to higher-productivity jobs, but he wondered if we should expect it to show in aggregate productivity data amid “all the background noise” of shutdowns and reopenings. He suggested there could also be “growing pains” related to hiring and training: “in the interim, new workers may not be as productive in the short run”. So we should watch how the productivity data evolve. But there is at least cause for optimism. And — in my own view at least — corresponding room for caution on central bank tightening. Dube pointed out that the “usual story about how a wage-price spiral may take hold is that inflation expectations change and workers negotiate higher wages”. But job stayers, he says, have “not [had] unusually high wage growth. It’s all driven by job switchers.” That, says Dube, “limits the scope of inflationary pressures” from the wage increases actually observed. To reiterate, these findings are only for the US economy. While most of Europe also shows historically high job vacancy rates, I haven’t found timely data on job-to-job moves to see if that rate has gone up too (Free Lunch readers, do send me any pointers). So even if this benign view of wage growth is correct for the US, it is not so clear for Europe. Dube points out that stronger minimum wage laws mean Europe has fewer of the low-paid jobs that drive his team’s findings in the US. And another “reason it may have happened more in the US is because we pursued what ironically seemed at the time like a worse way to help” — namely letting people lose their jobs and pay unemployment benefits rather than protecting employment relationships with furlough payments.Other readablesOver the Christmas break, I noticed a number of sometimes surprising pieces that in various ways reflect the biggest economic and political issues of the year that just finished. Start with the wonderfully quirky way in which Cinderella reflects protectionist industrial policy: Charles Perrault, who wrote down the fairytale of the girl with the glass slippers, was also in charge of outfitting the Palace of Versailles — including its Hall of Mirrors — and for setting up a national glassworks, which ensured that at time of economic patriotism (today we might say “reshoring”) the Sun King’s most spectacular ballroom was furnished with domestically sourced products.Meanwhile, global carmakers are quietly cutting ties with China.My colleague Jemima Kelly, who always saw the crypto bubble for what it was, writes on what the year in crypto taught us.Tales from the coalface, or rather the factory floor: how European manufacturers are coping with high energy prices, and how a chocolatier is using robots to manage labour shortages.China’s about-turn in Covid-19 policy may have an unexpected cause: how the zero-Covid approach exacerbated inequality.History-loving Vladimir Putin somehow never mentions Nicholas I, the dead tsar he most resembles.Numbers newsThe IMF has warned that a third of the global economy will suffer recession this year.The “moron premium” that sent UK borrowing costs soaring after the “mini” Budget in September has largely disappeared from gilt yields — but not from mortgage rates, Chris Giles finds.German, French and Spanish inflation slow more than expected. Who would have thought? More
