This article appears in the February 24, 2014, issue of National Review.
In his State of the Union address last month, President Obama called on Congress to extend federal emergency unemployment benefits that expired in December 2013. “I’m also convinced we can help Americans return to the work force faster by reforming unemployment insurance so that it’s more effective in today’s economy. But first, this Congress needs to restore the unemployment insurance you just let expire for 1.6 million people.” The president didn’t seem to recognize the internal contradiction in his speech: The extended insurance was surely a significant cause of the surge in long-term unemployment that has left us with a lingering unemployment problem.
Extended unemployment benefits lower workers’ incentives to search for jobs and to take jobs that may not be a perfect fit, and they may also lower firms’ incentives to hire new workers. It might seem intuitive that these incentives sometimes lead workers to delay their return to work, but many on the left dispute this. There is, however, a long line of literature demonstrating that unemployment insurance (UI) extends the duration of unemployment spells. A classic paper by Bruce Meyer and Lawrence Katz in 1990, for example, used data from a large number of U.S. households to show that a potential increase of one week in unemployment benefits lengthened a spell of unemployment by about a fifth of a week. Using their calculations, increasing unemployment benefits from 26 weeks to 99 weeks—the increase in UI benefits since 2007—would on average lengthen a person’s time in unemployment by several months. The problem with such an increase is that it can create a structural problem. A person who has been out of work for a little bit of time has a much easier time finding a new job than a person who must explain a lengthy period of unemployment.
While the Meyer and Katz result is well documented, it may not necessarily apply to job searchers in this particular recession. After the massive financial calamity, perhaps 99 weeks was an essential and equitable response to the absence of job offers for most everyone and did little to extend stays on the unemployment rolls.
The academic literature may well resolve this in the fullness of time, but international data can shed some light on the debate now. Some countries extended benefits greatly, some did not. The key question is, Did those with longer benefits see a bigger increase in long-term unemployment? The nearby chart explores the relationship between UI duration and the change in the proportion of unemployed workers who were unemployed for a year or longer between 2007 and 2012. The sample is for all OECD countries, with Australia and New Zealand omitted for lack of data, Belgium because its UI benefits are unlimited in duration, and Greece and Ireland because their employment situations were extreme outliers in the sample. There is a clear positive link. Countries with longer-lasting unemployment benefits saw a significantly larger increase in the share of long-term unemployment. The correlation be tween the two measures is highly statistically significant.
SOURCE: OECD. NOTE: A SIMILAR PATTERN WAS APPARENTWHEN USING MONTHS OF UI BENEFITS AVAILABLE IN 2007.
The data suggest that if the U.S. had not increased the duration of UI benefits to 99 weeks, but had instead left them at 26 weeks, the percentage of the unemployed stuck in long-term unemployment would be 8 percent instead of 37.7 percent, which amounts to about 3 million fewer longterm unemployed workers. Now, some of this effect might have resulted from workers’ dropping out of the labor force. But other workers might have avoided the fate of long-term unemployment by being spurred to take jobs more quickly or to search more vigorously in the early months of unemployment, knowing they could not rely on UI indefinitely.
Going forward, finding policies to help the long-term unemployed should be a high priority for all politicians. Extending benefits further would only make the problem worse.