The Job Market is Tight Again

Wednesday, June 22, 2022

Graph titled "Unemployed-to-openings ratio". The lines climb upward during the grey banded area to a ratio of roughly 6 workers for every opening, then decline to just below 1 until a sharp and narrow spike early in 2020. Finally, the line declines back below 1 again in mid 2021.

Data from FRED, Federal Reserve of St. Louis

In August 2020, I wrote about the perhaps obvious fact that the economic issue of the early pandemic was a job shortage. This problem has mostly disappeared—there are now more openings than unemployed people and there have been for months.

The larger problem was that it wasn’t safe to work or patronize businesses. Unfortunately, that problem hasn’t disappeared.

Workplace risks from COVID-19 have improved in some ways: there’s no longer a shortage of masks; vaccines that, when boosted, reduce long-term symptoms and greatly reduce hospitalization and death are available; and doctors know a lot more about treating severe cases.

But we have to balance that against the areas where safety has stagnated or worsened: The latest variants are extremely contagious, masks are no longer required or widely used, few workplaces have improved filtration, and emergency sick leave requirements have lapsed. Meanwhile, not that many people are working from home—about 7.4 percent in the latest jobs report. Because of this mixed bag, it’s difficult to ascertain the level of risk. Deaths are lower than this time last year. Hospitalizations are higher in June and about the same year to date as last year. Cases are much higher, even if you don’t adjust for the fact that many tests are done at home and not reported. These figures suggest that most people are at lower risk of the worst outcome but may be at a similar risk of hospitalization and a higher risk from long COVID1.

year January February March April May June 2 All
2020 0 1 4,303 58,836 41,239 14,094 118,473
2021 94,603 62,422 33,829 22,172 16,596 6,222 235,844
2022 55,736 61,978 33,644 12,351 8,309 5,866 177,884
All 156,856 134,593 75,798 95,589 71,484 27,219 561,539

Deaths (Source: COVID Act Now)

year January February March April May June 2 All
2020 N/A N/A N/A N/A N/A N/A N/A
2021 3,225,525 1,599,981 1,077,048 1,130,948 817,893 335,640 8,187,035
2022 4,212,441 1,881,606 551,048 348,881 690,019 632,761 8,316,756
All 7,437,966 3,481,608 1,628,096 1,479,829 1,507,912 790,485 16,325,896

Hospital Admissions (Source: COVID Act Now)

Having so many openings is generally to workers’ benefit. Most basically, workers have more bargaining power since their labor is in demand. Workers may have raised their standards and explored a wider range of options. Seeing so much upheaval may have more workers reconsidering their current position, as Arindrajit Dube hypothesized; previous research (full) has found that workers don’t always accurately estimate their opportunities.3 The improved market might provide its own protection. Workers shouldn’t be forced to leave their job in order to avoid a high risk of COVID infection, but at least that option exists.

Not all of this shortage is for good reasons. Some workers died due to COVID or other preventable causes. Another troubling reason may be due to workers with Long COVID. The Brookings Institute estimates that 1.1 million people left the workforce entirely due to Long COVID, and 0.5 million have reduced hours.

The unemployment rate is actually not quite at its historic low. U3 is at 3.6, whereas the historical low over the same time period is 3.5.4 The reason the ratio is low is that there are so many jobs. (To put it in more mathy terms, both the numerator and the denominator are relatively big.)

To reiterate from a previous post,

U-4 adds workers who’ve given up looking but still want jobs, and U-5 adds still more workers who want to be part of the labor force but can’t for reasons other than being discouraged (maybe they’re taking a class).

Given the sheer number of people who lost jobs during the early stages of the pandemic, it’s a little surprising U-4 and U-5 aren’t higher. Part of it is a lot of jobs have come back recently. It could also be that some people gave up on work altogether, perhaps by retiring early.

In addition to the persistent risk of COVID, there are also storm clouds on the horizon, as Jeanna Smialek and Ben Casselman write for the New York Times:

Pay gains have been falling behind inflation for months. Credit card balances, which fell early in the pandemic, are rising toward a record high. Subprime borrowers — those with weak credit scores — are increasingly falling behind on payments on car loans in particular, credit bureau data show. Measures of hunger are rising, even with unemployment still low and the overall economy still strong.

In short, we have a surging economy marred by the unnecessary risks forced upon workers by an indifferent administration and an ever-exploitative owning class.

Correction Friday, August 5, 2022

A previous version lacked attribution of the data. I’ve added data source attributions and alt text to the graph and data source attributions to tables.

Further Reading


  1. The assumption being that more cases entails higher risk. Vaccines, prior infections, and a different variant could, in theory, lower the per-infection risk enough to make up for the increase in cases. 

  2. Through June 21. 

  3. It is not really on topic for this post or this research paper, but I will quickly point out that employers deliberately seek to keep workers ignorant to their options, by discouraging workers from discussing pay, among other things. 

  4. The all-time low is 2.5, which the U.S. reached in February and March of 1953. While the measure’s definition hasn’t changed to my knowledge, changing social dynamics may not make this directly comparable. In 1953, roughly 80 percent of adult men were part of the workforce, whereas only about 1/3 of adult women were. Now, just under 70 percent of men and just over half of women are. (The Bureau of Labor Statistics did not and still does not collect data on nonbinary workers as a separate category, although the bureau is studying the possibility.)

    In 1953, it also was legal to discriminate on the basis of sex and race in hiring, with the exception of federal government and its contractors.