US job openings fell more than expected in June, but hires were mostly unchanged and quits remained extraordinarily elevated, underscoring the intractability of inflationary labor market distortions in the world’s largest economy.
Openings dropped more than 600,000 to 10.7 million on the last business day of June, key data out Tuesday showed. Economists expected a decline half that size.
Hires fell to 6.37 million from 6.5 million in May, leaving the gap near the widest on record, even as it also counted as the narrowest since November (figure below).
As Goldman wrote last week, “most of the reduction in the jobs-workers gap needed to slow wage growth to a pace consistent with the Fed’s inflation goal will have to come from a reduction in labor demand rather than an increase in supply.”
Tuesday’s data reflected what I’d tentatively describe as evidence of that dynamic beginning to unfold. We’re now getting close to a point beyond which I’d expect to see job openings move consistently lower, in line with myriad reports of corporate hiring freezes and layoffs.
The MoM drop in openings was the largest on record excluding the collapse associated with the onset of the pandemic (figure on the left, below).
Over three months, openings fell 1.16 million (figure on the right, above).
I’ve repeatedly suggested that a big drop in headline JOLTS is likely at some point over the next several months. A few days ago, I wrote that this week’s figures “will mark a turning point, even if this probably isn’t the ‘big one.'” The numbers appear to be fairly consistent with that assessment.
Of course, the situation is nothing like “normal.” Quits were 4.2 million in June, and the rate stuck at 2.8% (figure below). Churn is rampant, and why wouldn’t it be? Wage growth for job switchers is two percentage points higher than it is for so-called “stayers,” according to the Atlanta Fed’s wage tracker. Layoffs remained near record lows in June.
Quits remained very elevated in leisure and hospitality, where at least 800,000 workers have voluntarily left a job every month this year. In fact, the last time monthly quits in leisure and hospitality were below 800,000 was in June of 2021. Those figures are due almost entirely to ongoing tumult in accommodation and food services.
Tuesday’s figures should be considered in the context of an increasingly contentious debate between, on one hand, Larry Summers and Olivier Blanchard, and on the other, Christopher Waller. Late last week, Waller attempted to refute a July 13 paper by Summers and Blanchard who used a Beveridge curve analysis to suggest the Fed’s contention that job vacancies can be decreased without increasing unemployment is far-fetched.
On Monday, Summers and Blanchard refuted Waller’s refutation. “We looked at [Waller’s] note with interest, in the hope of being educated on a more optimistic view of the American economy’s soft landing prospects,” Summers and Blanchard wrote, in a short post dripping with sarcasm. “Unfortunately, our judgment is that it contains misleading conclusions, errors and factual mistakes.” Only in economics are “factual” matters subjected to “judgment.”
In any event, the bottom line from the JOLTS figures is that labor demand is softening. And quite rapidly. Just not rapidly enough. Labor demand needs to effectively collapse in order to short circuit the wage-price spiral evident in second quarter ECI data out last week.
7 thoughts on “US Job Openings Just Plunged. But Nowhere Near Enough”
I finally roused myself to go to the DOL website and check: the definition. JOLTS “INCLUDES: Full-time and part-time employees Permanent, short-term, and seasonal employees”
I could not easily find a breakdown between the full and part time positions cited in the monthly data.
Yes, part time and seasonal jobs are jobs. But could Dr. Waller survive on part-time wages?
It would be illuminating to see a breakdown of job openings by FT/PT, wage level, skill level, etc.
When I was in high school/college, I (and most of my friends) had seasonal jobs.
In the last decade, I noticed that only a small percentage of my kids’ friends had summer jobs. (My kids all had jobs). It would not be the worst thing if seasonal jobs get filled by high school and college aged people again. Great work experience.
The JOLTS data has always short term and part time openings. Unless you can make the case that the elevated JOLTS numbers are entirely due to an increase in those jobs, I believe H’s point remains valid. I expect that ratio between full time permanent openings and those less desirable openings has remained pretty constant.
Do not worry, the job market is going to roll over soon. Housing is in the early innings of a downturn and manufacturing is too. It may take a couple of quarters but once it switches it is going to switch fast. Most leading indicators are much weaker. It is only a matter of time…..
Employers have hoarded workers like other inputs. That can change in a hurry.
Need a little more time to sell my house on the mainland — Heis?
I think JOLTS openings can fall rapidly now. I’ve been looking at the “excess openings”, which is the extent to which each month’s job openings exceeds the prior month’s job openings less the month’s net hires (net hires = hires less total separations).
In 2021 and early 2022 excess openings was very positive. Employers were adding more and more new job openings, even while they were already adding new employees faster than they were losing existing employees (net hires have been significantly positive, above +0.5MM/mo.)
In 1Q22 cumulative net hires since the pandemic started went positive, meaning employers had added back all the employees they lost, and further new job openings and net hires would be for growth and expansion. With small businesses not in an expansionary mood, and large companies now pulling back, I think openings are likely to decline pretty steeply. Yes, quits get all the attention and have been very elevated, but total separations have not been much above-trend, because fires (layoffs, discharges, etc) have been very depressed.
And . . . “excess openings” just went negative.