I’m Ready to Be Automated
How AI can shift supply and demand—perhaps with benefits for everyone.
I’m Ready to Be AutomatedSebastien Thibault
The US government in March instituted the Coronavirus Aid, Relief, and Economic Security (CARES) Act in response to the COVID-19 pandemic. Among other provisions, the CARES Act temporarily, until July 31, delivered a $600 weekly payment to households that had suffered job losses. This provision was an emergency measure designed to replace 100 percent of the average national salary when combined with mean state unemployment benefits.
It ended up more than replacing salary for most workers, finds research by University of Chicago’s Peter Ganong and Chicago Booth’s Pascal Noel and Joseph S. Vavra. Three-quarters of US workers received more from unemployment insurance (UI) under the CARES provision than they did from wages alone when they were working—and the payments proved most helpful for people with the lowest incomes, facing the biggest job losses.
With lump-sum payments, such as those sent under the CARES Act for administrative reasons, it is difficult to replace income for most workers without also having some receive more than their precrisis wages, the researchers note. While their paper doesn’t prescribe an optimal, modified policy, they agree that the government should continue some form of expanded UI, Vavra says.
To understand how CARES supplementary payments were distributed across the national income spectrum, the researchers looked at micro data from households across the United States. They used census data (from the most recent Current Population Survey Annual Social and Economic Supplement) to determine prior earnings in unemployed households, then simulated a worker’s quarterly earnings history and applied the UI benefit formula by state to translate her earnings into current unemployment benefits.
Under the CARES Act, the median replacement rate—the percentage of salary replaced by UI—was 145 percent. It was more than 200 percent for workers in the bottom 20 percent of the US income spectrum, and more than 300 percent for workers in the bottom 10 percent. This compares with a typical pre-CARES rate of 40–50 percent of lost income, which had been the average state UI rate.
The figures fall slightly, to 69 percent of workers and a 134 percent replacement rate, when the researchers take benefits and taxes into account.
The payments reversed some group-level income patterns that would otherwise have arisen in this crisis. Take retail workers and teachers, for example. Unemployment rose more for retail workers than for teachers, which in normal times would lead income to decline more for the former than the latter. But because of the CARES payment, income for retail workers rose, “both in absolute terms and relative to the median teacher,” the researchers write.
It’s important to understand these patterns and see other data in this light, especially as more projects explore how the CARES payments affected spending and labor supply, write Ganong, Noel, and Vavra.
The COVID-19 crisis has been unprecedented in scope, and there was an immediate need in the spring for workers to remain at home during shelter-in-place restrictions. The CARES payments provided the most benefit to the lowest-income workers, who were employed in sectors such as food service, according to the research.
Since the CARES provision expired, critics have argued that continuing the benefits at the same levels would hurt the economy by keeping many workers on unemployment when they otherwise might find a job that would pay less. That’s a suggestion that research is exploring but has not at this point confirmed, write Ganong, Noel, and Vavra, who note that others in academia and policy are exploring how best to balance such trade-offs.
Peter Ganong, Pascal Noel, and Joseph S. Vavra, “US Unemployment Insurance Replacement Rates during the Pandemic,” Working paper, August 2020.
How AI can shift supply and demand—perhaps with benefits for everyone.
I’m Ready to Be AutomatedA Q&A with Chicago Booth’s Constantine Yannelis on policies to address the student-loan crisis.
The Case for Pausing, Not Canceling, Student DebtTurkey’s experience suggests ‘place-based policies’ can have unintended effects.
Does Targeting Aid Help Poorer Regions?Your Privacy
We want to demonstrate our commitment to your privacy. Please review Chicago Booth's privacy notice, which provides information explaining how and why we collect particular information when you visit our website.