What to watch on jobs day: Slow closing of the massive jobs deficit

On Friday, the Bureau of Labor Statistics (BLS) will release its latest jobs report with a snapshot of the labor market in mid-September. By now, the pandemic recession has caused immense damage to the health and economic well-being of millions of people for over six months. The economic pain easily extends to over 33 million people in the economy today, and that doesn’t include those who had lost their jobs and regained employment but got behind on their bills or those who lost loved ones and providers to illness.

Now that the economic losses have lingered on for this long, it may be time to reevaluate the metrics we use to determine the extent of the jobs deficit in the labor market today. The employment losses in March and April totaled 22.2 million, while the economy gained 10.6 million jobs between May and August. By that measure, the labor market was still down around 11.5 million jobs in August.

However, the more appropriate counterfactual would be to compare jobs today to how many would have been created if the labor market hadn’t tanked in the spring. So, what’s the appropriate counterfactual: average monthly job growth in the three months prior to the recession (216,000), six months prior (217,000), or twelve months prior (194,000)? All are defensible, but let’s go with the lowest value. At 194,000 jobs per month, the labor market would’ve added another 1,164,000 jobs over the last six months. That would give us a jobs deficit of 12.7 million (the 11.5 million fewer jobs we have than we had in February, plus the 1.2 million jobs we would have added over that period if the recession hadn’t occurred).

Not only is the jobs deficit enormous, the rate at which job gains have occurred over the last few months has also slowed dramatically (4.8 million in June, 1.7 million in July, and 1.4 million in August). If this slowdown in progress continues—as consensus reports and other analysts suggest it will—we are in for a long haul of a recovery, even more so if policymakers fail to pass additional stimulus. Without additional aid to state and local governments as well as reinstituting the $600 weekly boost to unemployment insurance (UI), it is likely that job growth will be over 10 million jobs lower through the next year (5.3 million for state and local aid plus 5.1 million for UI extensions).

Because of the economic crises, states and localities are facing huge revenue shortfalls, which could be relieved with more federal aid. What we know from the last recession is that states that preserved or grew their public-sector workforce fared better, with fewer job losses overall, fewer private-sector job cuts, less growth in unemployment, and faster job growth. Without significant federal investment, it will be impossible for state and local governments to withstand the expected shortfall in revenues from the current economic disaster and return to their pre-pandemic employment levels.

On the local level, school districts are struggling to provide appropriate education for students within the complexities of the pandemic. Even prior to the pandemic, public K–12 education employment never reached its level prior to the Great Recession. As of the latest data, public K–12 education employment is still 441,000 below where it was a year ago, and 767,000 below where it would have to be to keep up with growth in student enrollment since 2008. Unfortunately, without adequate funding, the necessary relief, recovery, and rebuilding cannot take place and students across the country will continue to suffer. On Friday, we will get the latest numbers on public-sector employment, and I will be paying close attention to public K–12 jobs as the school year got up and running in September.

Another key measure to watch in the upcoming jobs report is the duration of unemployment. Now that the recession is over six months old, it is likely that we will see a significant uptick in long-term unemployment (which is defined as an unemployment duration of 27 weeks or more). In fact, over the last several months, we have seen workers with longer and longer spells of unemployment. The figure below illustrates the number of unemployed workers by duration of unemployment: less than five weeks, 5–14 weeks, 15–26 weeks, and 27 or more weeks.

When the job losses spiked in April, the vast majority of the unemployed had been unemployed less than five weeks—not surprising since millions of workers were laid off in a single month. As the recession has dragged on, the share of the unemployed with longer unemployment durations has ballooned. By May, the vast majority had been unemployed for over five weeks, and by August, most had been employed for over 15 weeks. Now that we are seven months out—and six months since the worst of the job losses—I suspect more and more unemployed workers will show up with at least 27 weeks of unemployment, a dire circumstance for those workers and their families, particularly since the $600 enhanced unemployment benefit is long gone with little hope for resurgence.

Figure A

Workers are experiencing longer and longer periods of unemployment: Number of workers at each level of unemployment duration, by month, February 2020–September 2020

Date Less than 5 weeks 5–14 weeks 15–26 Weeks 27 weeks and over
Feb 2,013 1,803 825 1,102
Mar 3,542 1,794 808 1,164
Apr 14,283 7,004 833 939
May 3,875 14,814 1,078 1,164
Jun 2,838 11,496 1,903 1,391
Jul 3,202 5,169 6,484 1,501
Aug 2,281 3,134 6,517 1,624
Sep 2,552 2,732 4,918 2,405


ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Data are seasonally adjusted.

Source: EPI analysis of Bureau of Labor Statistics Current Population Survey public data series.

Copy the code below to embed this chart on your website.

As bad as these numbers are, they understate the economic pain. These counts of the unemployed do not take into account the millions of workers who have left the labor force or were misclassified as employed but not at work or had their hours cut. Only those counted as unemployed can be classified by duration.

When we think about the unemployed and the recovery, we must always consider for whom are the jobs returning. For this, I’m going to turn to the share of the population with a job—the employment-to-population ratio—which helps solve some of the measurement issues with the unemployment rate, though does nothing to solve any issues with the nonresponse rates so prevalent during this recession, which skew the data toward a rosier picture of the labor market.

So far, the recovery has been more successful for white workers than Black, Hispanic, or Asian workers. The share of the white population with a job in August was 4.0 percentage points below where it was in February. On the other hand, the shares of the Black, Hispanic, and Asian populations with a job are still 6.9, 6.7, and 5.7 percentage points lower, respectively, in August than they were in February.

As policymakers—particularly Senate Republicans—continue to drag their feet on relief, it is important for us to continue to track workers’ labor market experiences by race as well as by gender, education, age, and wage level to measure the avoidable, yet immense, pain inflicted on workers today. It is a simple fact that the labor market damage would be significantly lessened by vital public health investments and economic relief for today’s workforce as well as state and local governments. Policymakers must act if there’s any hope for a swifter and more broad-based recovery.