U.S. job matching holds up, keeping a soft landing in sight

As the Federal Reserve tightens monetary policy to slow inflation, will U.S. employers reduce job vacancies while avoiding a large increase in unemployment? The outcome could mean the difference between an ideal “soft landing” for the U.S. economy and a painful recession.

Many Fed officials still see a path to a soft landing. “Has it narrowed? Yes. Is it still possible? Yes,” said Fed Chair Jerome Powell following the November 2022 meeting of the Federal Open Market Committee. The “array of data in the labor market is highly unusual, and to many economists, there is a path.” Fed Governor Chris Waller and Board of Governors economist Andrew Figura found that “under most assumptions” the U.S. job vacancy rate could decline to typical, pre-COVID levels with an increase in the unemployment rate of less than 1 percent—an outcome “consistent with a soft landing.”

Economists Olivier Blanchard, Alex Domash, and Larry Summers attracted the spotlight earlier this year with their pessimistic assessment. Calling it “Bad News for the Fed from the Beveridge Space” (a reference to the Beveridge curve at the heart of the economic debate—Figure 1), they declare the soft landing to be wishful thinking and warn policymakers to brace for “substantial unemployment pain.”

Their economic model identifies the alleged culprit: Many job seekers, they say, are fundamentally mismatched to the jobs available. If so, the number of job openings required to sustain a given level of unemployment is much higher. The economists find a “sustained decline in matching efficiency” in the U.S. labor market, down 20 percent from before the pandemic.

Such a fundamental change in the labor market would imply that the Beveridge curve has “shifted out” into a new space—up and to the right—where a fall in vacancies must be accompanied by a sharp increase in unemployment. A bigger mismatch between job hunters and employers also implies that the natural rate of unemployment is higher than we thought. According to the historical record, this observation would suggest that the unemployment rate will need to rise substantially to return to a level consistent with lower inflation.

The three economists were less definitive on why openings and workers would have fallen so suddenly and steeply out of alignment with one another. In their policy brief, they suggest that attitudes toward telework and a generally greater restlessness (technical term: “reallocation”) among workers might play a role.¹

Running the numbers: Job matching looks just fine

Minneapolis Fed Senior Research Economist Simon Mongey brings a practical notion to this debate: If job-matching efficiency is at the center of the “hard landing” case, let’s just measure it directly: How effectively are firms hiring?

Mongey had established how to do this in a 2018 paper in the American Economic Review with Alessandro Gavazza and Giovanni Violante (and in a subsequent paper with Violante). In building a model to study how macroeconomic shocks affect business recruiting, they developed a formula to measure job matching efficiency via the essential inputs of hiring, job separations, job vacancies, employment, and unemployment.²

Applying the latest U.S. economic data, Mongey generates a clear answer to Blanchard, Domash, and Summers: We are not mismatched after all. After a steep, initial drop, job-matching efficiency in the U.S. has bounced back close to its pre-COVID level (Figure 2).

This is a much more positive picture than what Mongey and his co-authors measured after the Great Recession of 2007–2009 (Figure 3).³ More than four years from the end of that recession, matching efficiency remained stubbornly depressed at a level 20 percent worse than before—a clear sign of a structural shift.

By contrast, the rebound after the initial pandemic shock happened rapidly and has been more complete. While many variables remain in play, Mongey’s updated calculation of job-matching efficiency appears to clear away one dark cloud over the path to the soft landing.

Which Beveridge curve?

Mongey says the gloomy conclusion of Blanchard, Domash, and Summers arises from their assumption that the rate of hiring remains fixed no matter how tight the labor market gets. But in this unprecedented moment, when so many job postings are chasing such a tiny pool of unemployed workers, the rate of hires-per-vacancy (the job-filling rate) will naturally become very low—exactly what we would expect under these tight circumstances. This does not mean there is anything structurally altered in the ability of the labor market to match job seekers and open positions.

Whereas Blanchard, Domash, and Summers suggest the Beveridge curve has shifted out, Mongey believes the evidence points to an economy on a very steep extension of the traditional curve (Figure 4). From this point, vacancies can decline substantially without a large increase in unemployment. Mongey says that employers closing many of those excess open positions—thereby slowing upward pressure on wages—would be unlikely on its own to generate a large increase in unemployment.

Supporting evidence: Spending, vacancies, and quits

Recent data are consistent with this, showing falling job openings while the unemployment rate remains low. Sectors like retail and restaurants, where vacancies have been highest, are seeing spending growth cool and have been eliminating job openings since the start of the year (Figure 5), without widespread layoffs.

Mongey also notes that job vacancies and the rate of job quitting have maintained a consistent, positive relationship for many years. This relationship held up through the pandemic, as surging quits pushed up job openings. But quits have been gradually returning to pre-pandemic levels during 2022. If this trend continues, we would expect vacancies to follow.

Both sides would agree we are in uncharted territory. Blanchard, Domash, and Summers point out that there are no comforting examples of this type of soft landing in the modern economic data. They document that since the 1950s, job vacancies have never come down without a significant rise in unemployment.

Mongey counters that our present moment is unique. “There is no historical example of the ratio of vacancies to unemployment being so high,” he said. “But there is little evidence that the Beveridge curve has shifted given that match efficiency is at pre-pandemic levels.”

The story of the worker-employer mismatch offered a tempting explanation for our unusual, unbalanced labor market. However, Mongey says, its authors failed to account for how that same tight labor market would affect their readings, and the insight falls apart on further scrutiny. Deciphering the true forces driving our economy back from its unprecedented, pandemic journey remains the ongoing challenge of Fed economists and policymakers.

Endnotes

1 Another notable recent paper to weigh in on this debate came from economists Laurence Ball, Daniel Leigh, and Prachi Mishra. They do not offer a firm conclusion, but state that the path to a soft landing depends heavily on the path of the Beveridge curve and is “reasonable only under quite optimistic assumptions” about vacancies, unemployment, and inflation expectations.

2 The authors describe the formula in the online appendix to their AER article.

3 For additional figures and a technical breakdown of Mongey’s economic reasoning, see his recent Twitter thread.