Skip to content

A Very Steep Recession Curve


by Nancy Kimelman April 8, 2020

 This note highlights the most important features of what will go down — for hopefully a very long time — as a record-shattering report on the employment situation in America. As the number of jobs fell by 20.5 million in the month of April, the unemployment rate spiked to 14.7%. For reference, two months ago the unemployment rate was 3.5% and the economy added a healthy 230,000 jobs on the month. This swift reversal of fortunes is what happens when an economy is abruptly shut down. I will talk about the depth of our misery, to be sure, but the key now is look at the report for clues as to what happens next.

Before I get to the numbers, one should understand the way professional economists and financial market participants interpret the figures. Economists don’t sit around waiting for the monthly report to tell us how many jobs have been created or lost during the prior month, or what the unemployment rate was. We forecast the key figures in advance using all the data available to us. These forecasts then become the standard against which the real numbers are evaluated. Even with a horrifically bad employment report like we got this morning, the financial markets barely responded. This is not because the markets are heartless! It’s because the actual numbers, if they are close to expectations, don’t provide any “new” information and thus traders and investors have little new information to trade on. It’s evidence of rational expectations and efficient markets. For the record, the consensus forecast saw a 15% unemployment rate and a 21.8 million decline in nonfarm payrolls. (Yup, sometimes economists get it right.)

The monthly employment report from the Bureau of Labor Statistics (BLS) releases the results of two separate surveys. One is a survey of households. The other is a survey of establishments, or employers (and yes, there are adjustments to count the self-employed). While most people, including those in Washington, tend to focus on the results from the household survey, financial markets tend to look first to the establishment survey. This time round, both surveys have important information about our locked-down economy that may (note the emphasis) help us divine what the future holds as the economy emerges from the lockdown.

The headline from the household survey is the 14.7% of people who were counted as unemployed in April. That means that 14.7% of the labor force – people available to and looking for work – did not work. To be counted as unemployed an individual must not have worked and have looked for work. Now, in the midst of the pandemic, many of those laid off did not look for work because of the difficulty of doing so or because they believed their layoff to be temporary. Under normal circumstances, those people would not be in the labor force and not counted as unemployed (or employed). Because of the uncertainty of the current furloughs and layoffs, however, the BLS wanted to identify the temporarily laid off people as unemployed. They are counted in that 14.7% figure.

As I said, the headline unemployment figure of 14.7% marks the highest unemployment rate on record (the unemployment rate was estimated to have been 25% during the Great Depression but this data series doesn’t go back this far). The 14.7% rate also represents a deterioration (an increase) of 10.3 percentage points in the unemployment rate, the biggest one-month change in history. This number is called U3. The BLS also reports U6, which includes as unemployed the workers who were discouraged by the lack of jobs and didn’t look for one or others who simply stopped looking during the survey week. In April, U6 was 22.8% – another record-shattering result.

One of the other measures from this same survey that warrants attention this month in particular is the employment rate, which is the correlate to the unemployment rate. Only this month, with the labor market in such flux, many have argued that it may give a clearer picture. The rate of employment, which is the number of employed in the population, dropped 8.7 percentage points to 51.3. It represents a 22.4-million-person decline in the number of employed persons. To my mind, it’s corroboration, not new news. 

Stop reading for a moment and take all this in: Almost one-fourth of eligible workers in this country did not work in April. And only about half of all people aged 25-54 were working. The magnitude of these figures is astonishing and indeed indicative of depression-era conditions in the labor market.

The BLS report also provides demographic breakdowns of the unemployed. The differences are consistent with long-standing labor market trends (which doesn’t make them any more palatable, I should add). Here’s a summary:

“The rate was 13.0 percent for adult men, 15.5 percent for adult women, 31.9 percent for teenagers, 14.2 percent for Whites, 16.7 percent for Blacks, 14.5 percent for Asians, and 18.9 percent for Hispanics. The rates for all of these groups, with the exception of Blacks, represent record highs for their respective series.”

If there’s not much good news in the household survey, there is a bit of light in the results of the establishment survey. Of course, the figures show a devastating contraction of business. But, the industries where the jobs were lost along with what happened to hours worked and wages is not all doom and gloom.

The payroll figures (literally, how many people were on somebody’s payroll and getting paid), fell by 20.5 million in April. On top of that, the BLS has revised lower its estimates of February and March payrolls by an additional 214,000. The numbers here are telling a very similar story to those in the household survey (doesn’t always work out like that).

The beauty of the establishment survey is that it describes exactly where the jobs are – or aren’t. Thinking about this report over the past few weeks, I thought that the worst news would be far-reaching job losses in industries which we didn’t think had been severely affected. I don’t see that in this report. Don’t get me wrong: the job losses were widespread; they just weren’t surprisingly large in surprising places.

As expected the leisure and hospitality industries took the biggest hit. Employment in leisure and hospitality plummeted by 7.7 million people, or 47%. Almost three-fourths of that was in restaurants and bars, although layoffs were heavy in arts, entertainment, and recreation industries as well as in hotels. Job losses were significant in retail trade (all but food stores, pharmacies and Wal-Mart/Target/Home Depot-type stores were closed in the majority of states). And in government services – think teachers and state university closures – and personal services. (About the latter: how is it that the folks I see on TV look like they’ve had a haircut?) And in transportation and construction and real estate. The list of impacted industries goes on with few bright spots (big box stores added workers as did Amazon). Still, all this horrifically bad data were pretty much as expected.

[One side note: the mining sector also retrenched in April, due in part to reduced demand. However, don’t lose sight of the fact that a major battle was underway in April with Russia trying its best to break Saudi Arabia’s (and OPEC’s) control over the oil market and put to waste the U.S. oil and natural gas industries in the process. Russia’s strategy has been working.]

And finally a bit of good news. The average workweek on private payrolls edged up slightly. Those who kept their jobs worked a bit longer, which indicates there was work to do. That feels good. There was also an increase, a substantial one, in average hourly earnings. Average hourly earnings for all employees on private payrolls increased by $1.34 to $30.01. Average hourly earnings of private-sector production and nonsupervisory employees increased by $1.04 to $25.12 in April. Unfortunately, this is not really good news. It represents the reality of the job layoffs during the pandemic: workers who had to be present to work, typically lower paying jobs, were laid off. Those workers who could work from home, typically higher paying jobs, got laid off less frequently. The rise in hourly earnings is simply a reflection of who got laid off; it does not mean that people got raises (although some workers, especially in retail establishments, have been getting the equivalent of combat pay for working during these trying times).

So where do we go from here? That’s a bigger question than I can address here. Let me just summarize what this report has told us:

  1. As the economy reopens, those sectors of the economy that were the hardest hit are the same ones we want open first. That will work in our favor. Will the job losses be reversed good part in May and June? Probably not. But the numbers of unemployed will shrink. Maybe 10% unemployment in May?
  2. Hopes for a V-shaped recovery have pretty much flown out of the window. The rosy scenario now is for a Nike swish. Cute.
  3. There’s still a need to help both families and businesses stay afloat even as the re-opening process gets underway. Even 10% unemployment is very, very serious. Many economists don’t think the unemployment rate will get back to 5 or 6% for a few years. Until a vaccine can reduce the numbers of infected, or a miracle drug can reduce the deadly consequences or this disease, there will be no quick fix to the economy. And any resurgence in infection rates that requires shutting the economy back down will derail any economic recovery in short order.

The entire BLS report can be found at

And I can be reached at: