In Depth Analysis: CalculatedRisk Newsletter on Real Estate (Ad Free) Read it here.

Monday, September 07, 2020

A Pandemic Employment Model

by Calculated Risk on 9/07/2020 12:37:00 PM

It is possible to consider the impact of the pandemic on employment as being similar to a large number of simultaneous worker strikes, and a concurrent severe recession.

When a strike happens, a large number of jobs are immediately lost.  And then, when the strike is settled, the jobs are added back to the economy.

An example would be the August 1983 telecommunications strike when close to 700,000 went on strike.  This led to a sharp decline in employment in the August 1983 employment report.   The strike ended in late August 1983, and the approximately 700,000 jobs were added back in the September employment report.  A true "V" shaped employment recovery.

In March and April of 2020, close to 22.2 million jobs were lost.  A large number of these jobs were similar to a "strike", and the employees were able to return to work very quickly.

IMPORTANT: Of course, unlike a labor settlement, returning to work during a pandemic has serious health risks with over 30,000 Americans dying from COVID-19 in August alone.   And many more with serious health issues, see this story in the Eugene, Oregon Register-Guard: Marist alum Natalie Hakala went from collegiate runner to coronavirus long-hauler

Employment Recessions, Scariest Job ChartClick on graph for larger image.

This graph shows the job losses from the start of the employment recession, in percentage terms.

Some workers were able to return to the jobs fairly quickly. By August, about 10.6 million workers had returned to their jobs (almost half of the jobs lost in March and April).

This would be similar to a large number of simultaneous worker strikes, and then the workers returning to work when a settlement is reached.

However, these returning workers mask the concurrent severe recession.

Year-over-year change employmentThis graph shows permanent job losses as a percent of the pre-recession peak in employment through the August report.

This data is only available back to 1994, so there is only data for three recessions. Note that the permanent jobs losses data is from the CPS (Household survey), and the jobs data in the first graph is from the CES (Establishment survey).

In August, the number of permanent job losses increased sharply to 3.411 million from 2.877 million in July.  In addition, there have been a number of large company layoff announcements recently (some of these layoffs are happening as the CARES act corporate bailouts expire).    So the number of permanent job losers will probably increase sharply in the coming months.

Looking back at previous recessions, usually permanent job losers account for about 60% of total jobs lost in the early stages of the recession.   It was a much smaller percentage this time since many of the initial job losses were temporary (more like a strike).   My guess is about half the remaining jobs lost are already due to the severe recession - and the number is growing quickly.

Note: There is also some distortion from temporary decennial Census hiring too.  There were 288 thousand temporary Census workers in August (238 thousand hired in August alone).   This boosted employment, especially in August - and will subtract from employment in a few months.

Treating the current situation as a combination of simultaneous strikes, and a concurrent severe recession explains the fairly rapid employment increases, and also the decline in the unemployment rate.  However, this also suggests there are only a few million more jobs that can be gained in the short term - until we address the rapidly increasing severe recession.