A jobs survey full of good news

For as long as I’ve been paying attention to economic news, pundits and investors have waited anxiously for the monthly report by the Bureau of Labor Statistics on the employment situation. That’s still true, and there was some important news in Friday’s report. More on that later.

But another report from the bureau, which came out Tuesday, was a real eye-opener. It was, in particular, the best news about inflation we’ve seen in a long time — even though it never mentioned inflation.

JOLTS — the Job Offerings and Labor Turnover Survey — tracks, um, job offerings and labor turnover. That is, it asks employers how many unfilled positions they have, how many workers have quit or been fired and so on. It may seem obvious that this information is useful, but these days, many economists believe that it’s even more so.

Some background: Standard macroeconomics relies a lot on an updated version of the Phillips curve. The original version of that theory asserted (based on historical evidence) that there was a downward-sloping relationship between unemployment and inflation: The higher the unemployment rate, the lower the inflation rate, all else being equal. Since the 1970s, almost everyone has assumed that expectations of inflation also play a big role. If the public expects a lot of inflation, as it did at the end of the 1970s, this will push up actual inflation for any given rate of unemployment.

So what explains the recent surge in inflation? The mystery is that while unemployment is low, it’s roughly the same as it was on the eve of the pandemic, yet inflation is much higher. This is true even if you focus on “core” inflation, which excludes volatile food and energy prices.

Are we having a ’70s-type problem, with inflation driven by self-fulfilling expectations? Well, we have a lot of direct evidence on expected inflation, from both surveys and financial markets — and it isn’t especially high. Everyone seems to expect that the Federal Reserve will get inflation down, and fairly soon. So expectations aren’t the story.

What many economists have been suggesting, instead, is that the unemployment rate is an inadequate measure of how hot the economy is running. And some have argued that the best measure is the so-called Beveridge curve: the ratio of vacancies — unfilled job openings — to the number of unemployed workers. This is the basis of an influential recent paper by Laurence Ball, Daniel Leigh and Pankaj Mishra that offers a pessimistic take on inflation based on the fact that the vacancy-to-unemployment ratio is very high, even though unemployment isn’t all that low.

Another recent paper, by Olivier Blanchard, Alex Domash and Larry Summers, showed that as of early summer, the relationship between unemployment and vacancies had greatly deteriorated, which they argued implied that the unemployment rate consistent with stable inflation is now well above its current level — roughly 5% versus its current 3.5%.

But then came the latest JOLTS report, which showed a large drop in job offerings in August, even though unemployment didn’t rise significantly. Until August, vacancies were much higher than pre-pandemic experience suggested they should be for any given rate of unemployment.

Until very recently, it looked as if restoring the pre-pandemic vacancy ratio would indeed require something like 5% unemployment. It’s only one month’s data, but August suggests that the trade-offs may be improving as the economy recovers from COVID disruptions. A high unemployment rate may not be necessary after all.

I can’t resist pointing out that Blanchard et al declared, back in July, that “it is highly unlikely that the [required] decrease in the vacancy rate can be achieved without a substantial increase in the unemployment rate.” But that’s exactly what seems to be happening. In fact, more than 40% of the apparent excess in vacancies has vanished over the past few months, with no significant rise in unemployment.

Again, most of this is just one month’s data, and monthly data can be noisy. But the notion that labor markets are looking better got some further reinforcement from Friday’s employment report. Unemployment is still low. But wage increases have been slowing. Over the past three months, average wages have risen at an annual rate of 4.4%, compared with 3.1% in the last three months of 2019. That suggests some excess underlying inflation, but not all that much.

So we got a JOLTS of good news. And yes, it adds to the case that the Fed is behind the curve on its fight against inflation.