The Short Version: The Congress is locked in a protracted debate over “extenders” – including continuing unemployment insurance (UI) benefits to the long-term unemployed. In this debate, some have asked whether extending UI from its traditional 26 weeks up to as much as 99 weeks is increasing the unemployment rate. The answer is yes. Estimates attribute as little as 0.4 percentage points or as much as 1.8 percentage points of the 5.5 percentage point rise in unemployment rate since the onset of the Great Recession to extended benefits. The corollary is it is not always a good idea to extend benefits.

The Long Version: Unemployment benefits are good – they support standards of living and permit workers to find a new employer that is a good match – and bad – they reduce the intensity of job search, extend the duration of unemployment, and permit job skills to depreciate. Getting the generosity and length of UI benefits “just right” is hard – it is possible to give too much or give too lone.

Three recent research efforts examine the effect of UI benefits on the length of unemployment duration. The papers differ in their research methods and reach different bottom lines, but the research contributes to a fuller understanding of the costs and benefits underlying the Congressional debate over the “extenders bill” that includes continuing extended UI benefits.

1. The Labor Market in the Great Recession, by Elsby, Hobijn, and Sahin (forthcoming in the Brookings Panel on Economic Activity). The authors attribute from 0.7 percentage points to 1.8 percentage points of the rise in unemployment to extending benefits, but prefer estimates at the lower end of the range.

The paper provides an extensive overview of the labor market in the Great Recession, but authors highlight the “exit rate from unemployment” as the key factor in unemployment durations. Outflows fell during this recession, mostly due to poor matching of skill sets, depreciation of skills as unemployment duration lengthens, and statistical discrimination against the long-term unemployed.

In addition, using previously published estimates, the authors place significant importance on the extension of UI benefits. They estimate that one additional week of UI benefits will contribute to 0.08 to 0.2 additional weeks of unemployment – translating into their estimate of an increase of 0.7 to 1.8 percentage points.

2. What is behind the rise in long-term unemployment? by Aarsonson, Mazumder, and Schechter (Chicago Federal Reserve Bank) attributes 10 to 25 percent of the increase in duration to extension of UI benefits (based on previous published research). Meyer estimated that an additional week of benefits added 0.16 to 0.2 weeks of unemployment while Card and Levine estimated that an additional week of benefits added 0.1 week of unemployment. The authors used these estimates as their range of possible values. Using this range and their own simulation, the authors reported a 0.7 percentage point increase in unemployment that can be accredited to the extension of benefits.

3. Extended Unemployment and UI Benefits, by Rob Valletta and Katherine Kuang (San Francisco Federal Reserve Bank). The authors report the smallest effect the extension of UI benefits had on the unemployment rate: 0.4 percentage points.

The paper documents an unprecedented spike in the number of people who have been unemployed for six months or more. In March of 2009, 4.3 percent of the labor force had been unemployed for six months or more, compared to the previous high of 2.6 per

cent in 1983, which is noteworthy because the unemployment rate in 1982 was higher than 2009 at 10.8 percent.

It estimates the impact of extended UI by comparing the duration of unemployment for those eligible for extended UI to the duration for those who were not eligible when benefits were extended. The former saw an increase of 18.7 weeks, while the latter experienced an increase of 17.1 weeks. The difference of 1.6 weeks translates to the estimated 0.4 percentage point increase. The authors attribute the remainder to weak labor demand in the business cycle.

While clever, the authors are essentially extrapolating the experience of a single episode of extending UI – a weak research design that lends little confidence to the estimate.

Bottom line: It is well established that unemployment benefits increase the duration of unemployment – and thus raise the unemployment rate. The only question is how big the effect might be. The recent research places the impact in the ballpark range of 0.7 to 1.0 percentage points.