Depression-Level Job Losses in April


Payroll employment is expected to plunge by more than 34 million jobs in April as the toll of national lockdowns to combat the spread of the virus has mounted. Initial unemployment insurance (UI) claims surged by more than 26 million during the five weeks covered by the survey for the month of April, but those losses understate the actual losses in payrolls for the month. Many who qualified for unemployment insurance were unable to get their applications processed. A survey by the Equitable Policy Institute suggests that somewhere between nine and 14 million people could not get access to UI after they were thrown off payrolls as a result of COVID-19 during the five-week period. Another eight million undocumented workers are not even eligible to apply for UI. A good portion of them also suffered job losses, many in the leisure and hospitality industry.

The BLS announced that it also adjusted the birth and death rates of firms to more accurately capture the loss of jobs due to COVID-19 during April. The birth rate of firms fell dramatically, a detail which will help to better capture the loss in employment in small business employment.

This is at the same time that hiring screeched to a virtual halt. Posting for new jobs on all the major job websites plummeted between 30% and 40% in March and April. Some sites saw a falloff in posting as early as late February when the tech giants canceled offsite meetings and major conferences to stem contagion. What’s even more stunning is the breadth of losses. They spanned virtually every industry and income category. One of the greatest surprises to many during this health crisis has been the surge in layoffs in the health care industry. The fear of going to dental and doctors offices and contracting COVID-19 prior to national lockdowns in March had already triggered the largest ever drop in health care jobs in a single month.

Our estimate of 34 million job losses includes the 26 million job losses captured by UI, plus an additional eight million workers who either were unable to process their application for UI or were undocumented workers who lost jobs. Undocumented workers recieve pay from their employers so those payrolls drop when they are laid off. Total job losses could exceed 40 million given the miss in UI claims and the ranks of undocumented workers, but response rates will be low. The BLS reported a fairly large, 9% shortfall in the response rate by firms for the survey in March. Downward revisions to the March data are expected to be subtracted as they tracked down some of the companies that missed the initial survey.

Wages and hours worked were hit hard across a broad spectrum of industries as firms scrambled to preserve cash. The blow to wages is much more immediate and abrupt than we have seen in the past and could take a toll on average hourly earnings in a way we have never witnessed. Firms usually lay off workers before they scale back compensation schemes; both have happened simultaneously during this crisis.

The best we can come up with in terms of a translation to average hourly earnings is an unprecedented 1% drop in average hourly earnings between March and April. That would lower year-over-year gains in wages to 1.9%, more than one full percentage point below the annual gains in March. The blow to hours worked will be even greater with more than 20% of the workforce with no hours worked at all during April.

Separately, the unemployment rate, which is derived from the household survey, is expected to rise to 17% in April. This is well below the Great Depression levels of 20-25% implied by the job losses above. The problem is how we classify different kinds of workers for the household survey, which distorts our reading of the headline (U-3) unemployment rate. Mike Horrigan, the president of the W.E. Upjohn Institute and former associate commissioner of the BLS penned an article laying out the measurement problems that could arise in the April data. He identified three areas of concern:

  1. The number of people reporting they were absent from work but not unemployed in March was unusually high. It was so high that the data crunchers at BLS suspected the question was misinterpreted; the BLS estimated that the unemployment rate could be as much as 5.4% in March, a full percentage point higher than was actually reported, if the question had been answered correctly. A more accurate classifications of workers absent from work should show up as more unemployed in April.
  2. The number of people who expect to be called back to work may have fallen quite dramatically between March and April, even if well-intending employers said layoffs were temporary. That, coupled with a surge in the number of people who were not actively searching for work in the month that passed (because workplaces were closed) could dramatically reduce the ranks of those classified as unemployed.
  3. Finally, the labor market could shrink, as those who might have rejoined it, such as parents returning from paternity leave and retirees - or students entering the labor market for the first time decided to sit it April out.


We are also worried about a precipitous drop in those participating in the labor market. Parents (mostly women) who are now taking care of children home from school cannot as easily search for work. Indeed, shuttered schools are one of the biggest concerns for employers hoping to get their workers back when the economy reopens. Low-wage workers have a particularly hard time covering daycare costs. Social distancing is another hurdle as it could preclude even the wealthiest of households from hiring help to watch children at home instead of at camp or school over the summer.

Indeed, this is a time when it will be useful to focus more on the U-6 measure of unemployment, which was used frequently during the Great Recession. It includes both workers forced to scale back hours from full-time to part-time work and workers not classified as unemployed because they are on permanent instead of temporary layoff and/or have not looked for a job in the past month. That figure is expected to move well above 20% to a record high in April. It could even exceed the 24.9% Depression-era peak in the U-3 measure of unemployment hit in 1933.

Finally, it is important to look at the gap between the unemployment rate for blacks and whites. Minorities with fewer economic advantages, facing overt discrimination, tend to be hit much harder by a recession than their white counterparts. In fact, the gap in unemployment between blacks and whites already widened in March, when layoffs surged. Look for it to widen further in April.

Media Contact
Karen Nye
T +1 312 602 8973

Other Inquiries
Na Tasha Lowe
T +1 312 754 7368