With over a million new unemployment claims still being filed each week, job growth slowing, and millions of workers about to lose emergency jobless benefits created through the CARES Act in March, it is imperative that Congress enact another COVID-19 emergency relief package as quickly as possible. In addition to key provisions such as aid to state and local governments and extending the emergency benefits for unemployed workers, Congress should also extend innovative provisions that helped prevent workers from being laid off in the first place—specifically the CARES Act’s federal subsidies for work-sharing.

Work-sharing provides a way for many businesses to cope with a drop in consumer demand without having to lay off staff. Under work-sharing, workers’ hours are reduced and they receive partial unemployment insurance (UI) benefits to make up a portion of their lost wages. For example, if a business anticipated having to lay off five workers, they might instead cut in half the hours of 10 staff—achieving the same reduction in total work hours—and those 10 workers would all receive partial unemployment benefits from the state to make up some of their lost income.

By keeping workers on the job, work-sharing mitigates the impact of joblessness and reduces unemployment peaks in downturns. Now that coronavirus vaccines appear to be on the horizon, maintaining a connection between more workers and their employers for the first part of next year makes even more sense. As the pandemic is brought under control and regular consumer demand picks up, companies with work-sharing programs won’t have to go through a process of recruiting, hiring, and training new staff; they will be able to quickly ramp back up simply by restoring participants in work-sharing to full time.

Although work-sharing has not historically been widely used in the United States, Figure A shows that it is fairly common in other parts of the world. In Belgium, the country with the highest take-up rate, work-sharing participants equaled about 5.6% of employment in 2009, a huge cushioning of Belgian families and businesses from the impact of the Great Recession. For comparison, if the U.S. had the same take-up rate of work-sharing as Belgium, roughly 2 million jobs might have been preserved in 2009.i Figure A also shows that in 10 other countries, work sharing equaled 1% to 4% of employment in 2009—five to 20 times larger than usage in the U.S.

Because of its benefits for businesses and workers, the March CARES Act included a 100% federal subsidy for work-sharing in states that already have work-sharing programs (some 27 states) and a 50% subsidy for states to establish such programs. As a result of these subsidies, when customer demand declined and some businesses needed to operate below full capacity to allow employees to social distance, use of work-sharing picked up after March. In July, U.S. Department of Labor data showed the number of U.S. employers with approved work-sharing programs rose to 14,770, over six times the 2,405 employers who had programs in February. Employee work-share claims spiked in this time as well. As Figure B shows, the number of employee continuing claims filed for work-sharing UI rose from less than 70,000 in February to about 1.7 million in July.

The use of work-sharing varies by state—but, notably, in a nonpartisan way. Table 1 shows how substantial work-sharing UI is relative to regular UI. (Specifically, the table shows continuing claims for work-sharing UI as a share of all UI continuing claims.) The table is ordered from top to bottom by states’ 2020 shares, starting with Washington’s work-sharing continuing claims equal to almost one-tenth of all UI continuing claims. Of the eight states that have used work-sharing the most to reduce unemployment in 2020, half are “blue” states and half are “red” states. Two of the biggest “red” states—Texas and Florida—also use work-sharing and could capitalize on an extension to use it more.

Also enticing to the politically pragmatic, several states often considered “battleground states” have work-sharing programs but rank in the middle or near the bottom of the list for work-sharing UI claims as a share of all UI claims. Extending the CARES Act subsidy would allow states like Michigan, Ohio, Wisconsin, Minnesota, and Pennsylvania to expand work-sharing and would likely boost their manufacturing sectors, since manufacturers are often the biggest users of work-sharing programs. The incoming Biden administration could also give the Great Lakes states a multi-state peer learning grant to expand work-sharing across the region, including in manufacturing.

Another reason to extend work-sharing: It’s cheap. It would cost less than $250 million to include work-sharing in the recent bipartisan proposal to extend the CARES Act’s UI provisions.ii

In closing, work-sharing has already made an important contribution to reducing unemployment in the COVID-19 recession. Failing to extend the work-sharing subsidies could sever hundreds of thousands of additional workers from their employers—right when we may be within sight of an end to the pandemic. This would be a form of snatching defeat from the jaws of victory.

(Acknowledgment: the authors gratefully acknowledge the great work of Mary Madsen, first author of Keystone Research Center’s October work-sharing brief and now back at graduate school while working part-time at KRC.)

Steve Herzenberg is the Executive Director of the Keystone Research Center in Harrisburg, Pa. Claire Kovach is a Senior Research Analyst at the Keystone Research Center.

i. Average total nonfarm employment in the U.S. in 2009 was 131.29 million. If work-sharing as a share of employment rose from 0.2% to 5.6% (Belgium’s rate), that equals 7.09 million. However, because participants in work-sharing are still partially employed, the resulting increase in employment is not a one-to-one increase in the number of jobs. For the first 10 months of 2020, full-time equivalent (FTE) weeks of work-sharing was equal to an average of 27% of all work-sharing claims. Thus, 27% of 7.09 million would be 1.91 million full-time jobs.

ii. The math behind this estimate: In October, there were 198,543 full-time equivalent weeks of UI work-sharing claims, according to the US DOL’s ETA 5159 data on work-sharing. Assuming claims stay at this level—they will more likely decline—and that average unemployment benefits for an FTE week of work-sharing benefits equal $300 (which is the average cost of regular UI benefits); the cost for a 4-month extension of benefits would be $238.2 million.