Job Reallocation Shock - The New Post-COVID-19 Economy

A working paper entitled "COVID-19 is also a Reallocation Shock" by Jose Marie Barrero, Nick Bloom and Steven J. Davis looks at the future of employment in the post-COVID-19 era using data from the Survey of Business Uncertainty (SBU) from April as well as anecdotal evidence gleaned from media coverage of the economy.  From this dataset, the authors have created a forward-looking measure of expected sales and job reallocation across the United States over the coming year.  For the purposes of this posting, I will only examine the authors' analysis of the job situation in the United States.

As the authors note, the coronavirus pandemic is resulting in the most rapid reallocation of labor since the end of the Second World War, thanks to the government-ordered shuttering of the economy.  This has meant that corporations and governments have been forced to mobilize millions of laid-off workers into new work activities that are needed in the post-COVID-19 era.

If we want to look at a historical perspective on job reallocation we only need to look back two decades or less.  Thanks to the economy shifting from mom and pop and smaller, independent retailers to national chains like Walmart, Target, Home Depot and Best Buy as well as online retailers like Amazon, employees (and owners) at the smaller retailers were forced into new jobs as their employers shuttered their operations.  

Let's start by looking at some anecdotal evidence of hiring and job reallocation:

1.) Walmart - hiring 150,000 new employees in one month (effective April 18, 2020) and plans to hire 50,000 additional employees.

2.) Amazon - hired 100,000 new employees in recent weeks and aims to hire another 75,000 employees.

3.) Lowes - plans to hire 30,000 new employees this spring.

4.) Instacart - adding 300,000 shoppers to its payroll.

Some companies are partnering up with other companies to exploit the reallocative nature of the coronavirus economy, particularly since spending on travel, hotels, rental cars, ride sharing and other related industries is down between 75 percent to 95 percent on a year-over-year basis:

1.) Kroger - partner with Sodexo, Sysco and Marriott International to hire workers laid off from hospitality and food service industries.

2.) CVS Healthcare is recruiting 50,000 new staff by partnering with the Hilton hotel chain.

3.) Uber is listing jobs for its unemployed drivers at McDonald's, 7-Eleven and Amazon.

Now, let's look at how the authors used the Survey of Business Uncertainty.  The SBU polls senior executives on a monthly basis in all 50 states and in every major non-farm industry.  Survey questions examine the forecasts of these executives for their own firm over the next 12 months.  The authors used these firm-level employment forecasts in the SBU to calculate the following:

"gross expected job gains at firms expecting to grow over the next year plus gross expected job losses and firms expecting to shrink over the next year minus the absolute value of the net change obtained by summing over all of the forecasts"

The firm-level forecasts in this calculation are then activity weighted and divided by aggregate employment to obtain the expected excess job reallocation rate looking forward over the coming year.

In the April 2020 SBU, two special questions were asked:

1.) the coronavirus impact on own-company staffing from March 1, 2020

2.) the anticipated coronavirus impact on own-company staffing over the next four weeks.

Cumulatively, the response to this suggest that there were near-term layoffs equal to 12.8 percent of March 1, employment and new hires equal to 3.8 percent of March 1 employment.  This means that the COVID-19 shock caused the hiring of 3 new employees for the layoffs of every 10 employees.  If contractors and leased workers are included, the ratio is about 2.7 new employees for every 10 employees laid off.  In January 2020, the expected job excess reallocation rate was 1.54 percent (average 2.23 percent between September 2016 and January 2020); this rose to 5.39 percent in April.  Gross staffing reductions are equal to 14.9 percent of March 1 employment (thanks to layoffs and furloughs) and near-term net job contraction (job losses plus new hires) is equal to 11.9 percent of March 1, employment with 92 percent of this contraction occurring between March 1, and mid-April.   Expected employment growth over the next 12 months fell 2.2 percentage points from January to April 2020.

Here is a graphic showing the expected employment growth rate and the expected excess job reallocation rate:

In closing, let's quote from the conclusions of the authors:

"Many of the American jobs lost since early March will return in the coming weeks and months as the pandemic and lockdown recede….temporary layoffs and furloughs account for 77 percent of gross staffing reductions from March 1 to mid-May. Recall that 27.9 million Americans filed new claims for unemployment benefits in the six weeks ending on April 25. 23 percent of that number is 6.4 million permanently lost jobs. Of course, there remains tremendous uncertainty about the economic outlook. For many firms, cash-flow problems today will become insolvencies in the future, and “temporary” layoffs will become permanent. The longer it takes to bring the economy back on line, the larger the fraction of recent layoffs that will turn out to be permanent…."

Using a different calculation, the authors found the following:

"To get a sense for the fraction of layoffs that will lead to actual recalls, we turn to evidence from Katz and Meyer (1990), who analyze a sample of UI recipients in Missouri and Pennsylvania from 1979 to 1981. They find that 72 percent of UI recipients who initially expected to be recalled were actually recalled. In addition, 13 percent of ex ante “permanent” layoffs were, in fact, recalled….According to this calculation, 42 percent of the gross staffing reductions.…will result in permanent job losses. Applying the 42 percent figure to the 27.9 million new claims for unemployment benefits in the six weeks ending on April 25 yields 11.6 million permanently lost jobs. This number does not include future job losses caused by the COVID-19 shock."

There is no doubt that many of these shifts in employment will reverse once the pandemic recedes and governments roll back their lockdown orders, some aspects of the changes to employment will persist into the future, particularly job reallocation which may or may not be to the financial benefit of workers.  Over the next year and into the future, millions of currently unemployed Americans are about to find out just how costly the governments' responses to the COVID-19 pandemic have been.