Unemployment Insurance Generosity and Aggregate Employment
During the Great Recession, unemployment rose substantially, triggering unprecedented increases in the duration of unemployment insurance (UI) benefits. Thanks to the Extended Benefits (EB) program created in 1970, which automatically adds up to 13 weeks of benefits when a state’s unemployment exceeds certain thresholds, and the Emergency Unemployment Compensation (EUC) program created during the onset of the Great Recession, under which up to 60 additional weeks of benefits were made available in high-unemployment states, the maximum duration of UI benefits for unemployed workers rose from 26 weeks to as much as 99 weeks. Benefit extensions may have two offsetting effects on employment. On the one hand, longer benefit eligibility may mean that workers take longer to find a job, causing employment to be lower. The effects of benefit duration on labor supply have been widely studied. On the other hand, putting more money in workers’ pockets could raise aggregate demand, leading employment to be higher. Less is known about this effect. Boone, Dube, Goodman, and Kaplan seek to estimate the overall effect of UI extensions on aggregate employment during the Great Recession.
An important fact about the increases in the duration of UI benefits during and after the Great Recession is that the benefit extension timeline differed considerably across neighboring states. This allowed Kaplan and his collaborators to compare areas that were otherwise similar but experienced extensions at different times. Comparisons at the state level to learn about the effects of changes in aggregate benefit payments are potentially problematic. This is because those relationships may be affected by reverse causation—changes in aggregate benefit payments associated with changes in maximum benefit duration could be driving employment, but it is also possible that changes in employment due to other factors affect unemployment and, through the formulas contained in the EB and EUC legislation, drive changes in aggregate benefit payments. Kaplan and his collaborators address this reverse causality using a border county pair strategy. The intuition behind this strategy is that basing comparisons on neighboring counties, especially neighboring counties in which employment trends had been similar prior to the introduction of differential UI benefits at the start of the Great Recession, should mitigate any reverse causation concerns. They find no evidence that UI benefit extensions substantially affected aggregate employment. This is consistent with a small negative effect on labor supply together with a modest positive effect on aggregate demand in the local economy.
A frequently-expressed concern about extending the duration of unemployment benefits when unemployment rises is that, by depressing labor supply, such extensions could have the counter-productive effect of reducing employment. These findings suggest that such concerns may be unwarranted.