New Research from UMass Amherst Economist Shows Unemployment Insurance Extension During Great Recession Did Not Destroy Jobs

Arindrajit Dube
Arindrajit Dube

AMHERST, Mass. – A new study released by a team of economists including Arindrajit Dube, associate professor of economics at the University of Massachusetts Amherst, reveals that the large extension of unemployment insurance (U.I.) benefits during the Great Recession had neither statistically significant nor economically meaningful effect on employment, positive or negative.

During and after the Great Recession, the federal U.I. program was expanded significantly as unemployed workers became eligible to receive up to 99 weeks of U.I. payments, compared to the usual 26 weeks available before the crisis. While few would disagree that these benefits helped laid-off workers weather the hardship of job loss, some economists have theorized that the increased generosity of U.I. benefits might have reduced employment and slowed economic recovery. Some economists also say there is a possibility that providing funds to workers helped boost consumption and profits, bolstering employment.

The new paper, “Unemployment Generosity and Aggregate Employment,” published on Dube’s website, is co-authored by Christopher Boone, assistant professor in the School of Hotel Administration at Cornell University, Lucas Goodman, Ph.D. student at the University of Maryland and Ethan Kaplan, associate professor of economics at the University of Maryland.

“Our baseline results suggest that extending benefits by 73 weeks increased employment by 0.2 workers per 100 working age residents, a negligible amount,” Dube explained. “While our employment estimates are not statistically distinguishable from zero, they do rule out moderate-sized negative employment effects of the U.I. extensions of 0.5 or more workers per 100 working age residents.”

Dube and his coauthors say the nature of the U.I. extensions during the Great Recession—on which the federal government spent nearly $50 billion each year from 2009 to 2013—makes studying this question very difficult. In order to target more generous benefits to those areas in greatest need, the benefit extension programs were structured to grant longer benefit lengths to states with higher unemployment rates.

“So it would be difficult to infer the impact of U.I. on employment simply by comparing states with more generous benefits to those with less,” Kaplan said. “Some states are provided more generous benefits precisely because their labor markets were doing worse rather than because unemployment insurance provisions lowered employment.”

To avoid this reverse causality, the authors examined 1,161 county pairs that straddle the border between two states, such as Allegany County, Md., and Bedford County, Pa. Within each pair, counties share a similar geography and economic environment, but may have access to different lengths of benefits, largely due to the economic situations in the rest of the state.

The researchers then compared changes in county employment to changes in the maximum length of benefits within each pair over the course of the Great Recession. They also focused on two policy changes: one that expanded benefits in November 2008, and one that eliminated all extended benefits at the end of 2013. They found that these changes were due to shifts in national policy rather than individual state economic performance.

“The changes in U.I. generosity between neighboring states due to these national policy changes are particularly suitable for causal analysis,” Boone said.

The results of this new paper contrast with two recent papers by economists Marcus Hagedorn, Iourii Manovskii and Kurt Mitman, who were joined for one paper by economist Fatih Karahan.

“Like us, they employ a strategy comparing county pairs which straddle state borders, but they find a substantial negative effect of U.I. on employment. In our new paper, we provide a detailed accounting of why our result is different from theirs,” Goodman said.

The differences are largely accounted for by two factors. First, Dube and his coauthors say that they use better measured employment data from the Quarterly Census of Employment and Wages, instead of imputed unemployment data from the Local Area Unemployment Statistics produced by the Bureau of Labor Statistics. Second, they say the methodology used in this new paper uses empirical strategies that “provide a more transparent analysis of the underlying data and are less reliant upon model-based assumptions.”

“The overall small aggregate effect on employment is consistent with a small number of people waiting longer to take jobs along with a moderate boost to economic activity from increased recipient spending,” Dube explained. “This should reassure policymakers facing a future recession that providing relief to job-losers by extending U.I. will not have substantial unintended effects upon economic activity.”