Have Enhanced Unemployment Benefits Discouraged Work?

The Issue:

The CARES Act greatly increased the generosity of unemployment insurance benefits by adding a fixed $600 federal weekly supplement to state benefits and expanding eligibility for benefits. As a result, many of the unemployed received benefits that surpassed their prior weekly wages until the benefit expired at the end of July. The $600 benefit became a point of contention in negotiations over additional coronavirus aid. Concerns have been voiced that the generosity of these benefits has kept people from reentering the labor force, slowing the pace of economic recovery. But, recent research provides a more nuanced picture.

Even with the additional $600 in unemployment benefits, there are more job applications for every new job vacancy posted since COVID.

The Facts:

• Unemployment Insurance typically replaces less than 100 percent of the income workers lose. There is evidence that, in normal times, an increase in unemployment benefits tends to increase the length of time that people remain unemployed and also tends to decrease the effort made by people in searching for a new job. But, the generosity of unemployment benefits is not the only factor that impacts the intensity with which workers search for jobs or the speed with which they become reemployed — especially during a pandemic. Issues such as the availability of childcare or health concerns can also hamper a return to work. Moreover, the impact on the economy from disincentivizing job seeking through generous unemployment benefits also depends on the balance between available jobs and the number of workers who are unemployed.
• In response to the unprecedented economic shock triggered by the coronavirus pandemic, the CARES Act boosted unemployment benefits substantially. Up to two thirds of unemployed workers were eligible to receive more in benefits than they earned in their prior jobs, according to one estimate. COVID-19 had a swift and dramatic impact on the U.S. economy, with many business closures and unemployment skyrocketing to 14.7% in April 2020. Since then, new unemployment claims have remained in excess of 1.2 million each week (the previous high was 695,000 in October 1982) and the unemployment rate remained above 10% in July 2020. In response, lawmakers passed the Coronavirus Aid, Relief and Economic Stimulus (CARES) Act on March 27, 2020. Among other measures, the CARES Act increased the level of unemployment benefits by adding a $600 weekly federal supplement to state benefits and expanded eligibility to workers who would not normally be covered by unemployment insurance, such as gig workers. The $600 amount was chosen to raise unemployment benefits to replace 100% of lost wages for the average U.S. worker — since technical issues in state unemployment insurance systems did not make it possible to rapidly tailor benefits to replace each worker's prior earnings.
• The extent to which the boosted unemployment benefits replaced prior earnings (or went beyond and provided supplemental income) varied greatly from worker to worker depending on their prior wages and their state of residence. The fixed $600 payment represented a proportionally greater boost for lower-wage workers and for workers in states with lower pre-COVID benefit replacement rates. (The replacement rate is the ratio of the amount received in benefits by an unemployed worker to the wages received while working.) Benefits are determined and administered at the state level. Workers typically receive UI benefits that equal half of their pre-unemployment wages up to a maximum level of benefits, which differs across states ranging from the $200s per week for low-benefit states to the $800s for high-benefit states (see here). With the additional $600, an unemployed worker who normally received $600 per week would see their benefit double (100% increase), while a worker who normally received $800 a week would see their benefit increase by only 75%. 
• To understand what impact these boosted benefits have had on employment, we first look at the evolution of job vacancies and job applications since the onset of the pandemic in mid-March. While both fell markedly after March 16, the drop in vacancies was triple that of applications. As a result of the collapse in job openings, we find there has been an increase in the number of job applications per open position since the onset of the pandemic. Daphne Skandalis, Daniel Zhao and I analyzed data from Glassdoor, one of the world’s largest jobs and career sites. The data covered the period December 30, 2019 to June 21, 2020. The total job openings on this site are about four-fifths of all job openings reported by the Bureau of Labor Statistic Job Openings and Labor Turnover Survey (see here). We looked at the changes in the numbers of postings for job vacancies and in the number of job applications. We find a decline of job applications of 21 percent from the week of March 2-8 to the week of March 16-22 (note that there was a peak in Google searches for “coronavirus” around March 16), but no significant subsequent additional decline after the CARES Act passed on March 27. Postings for new job vacancies dropped by 64 percent since mid-March, triple the decline in job applications; this was due to a 20 percent decline immediately after the crisis started and a further 50 percent decline after the CARES Act.
• If the $600 federal unemployment boost has the effect of reducing workers' motivation to seek work, you would expect to see a bigger disincentive effect among those who experienced a greater increase in the unemployment benefit replacement rate. Indeed, when we separate workers into groups based on differences in the benefit replacement rate increase caused by the CARES Act, we find that higher increases in replacement rates are associated with lower applications and lower applications-per-vacancy which is consistent with the enhanced benefit reducing applications. We calculate average earnings in each occupation category in each state and use this information to create four groups based on the increase in the replacement rate, each with the same number of job applications. The ratio of job applications to job vacancies increased for all groups — an indication of a weaker job market overall. But there are significant differences across the replacement rate groups. The group with the highest replacement rate increase (the 4th quartile) had a significantly lower ratio of applications per opening (see chart). It is important to note however that this does not prove that the enhanced benefits caused a decrease in applications. Indeed, this group of lower-wage workers also had a slightly lower level of applications in the pre-COVID period. Overall, even for the group with the highest replacement rate increase, the applications-to-vacancy ratio was higher after the CARES Act than in the pre-COVID period. This would indicate that, given the overall weakness of the labor market, employers have many more potential applicants for each job opening in all groups and thus the $600 benefit increase is not likely to have had an impact on aggregate employment during this period.
• Other recent studies show similar results concerning the lack of an effect of the increase in unemployment benefits through the $600 benefit on overall employment.  A group of researchers at Yale University studied hourly workers employed by small businesses and find no decline in employment in groups that saw a greater increase in the unemployment benefit replacement rate after the CARES Act. They also found workers who faced larger expansions in these benefits returned to their jobs at similar rates as others (see here). Another recent research paper shows that the majority of the decline in jobs for hourly workers occurred between March 14th and 28th, and this was driven by low-wage service jobs, particularly in the retail and leisure and hospitality sectors. More generous UI benefits were associated with milder declines and faster recoveries (see here). And a study using a different design strategy and different data source (the Census Household Pulse Survey) also finds no clear indication that high replacement rates had an impact on employment through late July. Our findings help explain the absence of an association between more generous unemployment insurance and employment: when there are too many applicants per job, a relatively small reduction in the number of applications has no material effect on the likelihood of that job being filled.
• Restaurants and bars provide a particularly interesting window into job dynamics during this period. Lockdowns directly affected these establishments since they were a potential locus of the spread of the disease. Workers in restaurants and bars receive relatively low wages, and therefore they saw a high unemployment insurance replacement rate with the $600 weekly supplement. We conducted a separate analysis of this sector for these reasons and also because this sector was a focus of other studies (see here and here). We redefined replacement quartiles for this sector alone and, because of lower average wages, the replacement rates and their increase after the CARES Act were higher than for the broader economy. We found that the decline in vacancies was roughly similar to the decline in applications in this sector so the number of applications per vacancy was roughly the same after the CARES Act as it was before. Unlike the broader sample, there was no evidence that, in this sector, quartiles with the higher replacement rates had fewer applications per job after the CARES Act. Therefore, in contrast with our results for the whole economy, there is no evidence that employers in this sector found it harder to attract applicants due to higher replacement rates even though replacement rates were higher in this sector than in the broader economy.