Study: Extending Unemployment Benefits Increased Unemployment by More than 3 Percentage Points

There has been much debate recently over the merits of extending unemployment benefits — last week, the Senate reached cloture on a bill to do so for three months, and will vote on the bill tonight. Even just among conservatives, the debate has been robust: AEI economist Michael Strain has argued, on the Corner, for extending the benefits, while others, such as Senator Rand Paul, have argued against such an extension.

But the debate hasn’t focused on the academic literature on the subject – including an important, surprising new paper written by four well-respected economists — Marcus Hagedorn of the University of Oslo, Fatih Karahan of the University of Pennsylvania, and Iourii Manovskii and Kurt Mitman of the New York Fed, distributed by the National Bureau of Economic Research (NBER). The study’s finding: “Most of the persistent increase in unemployment during the Great Recession can be accounted for by the unprecedented extensions of unemployment benefit eligibility.”

In the aftermath of the financial crisis, the federal government has dramatically expanded the length of the unemployment insurance program from the typical 26 weeks, the period states more or less have the funds to offer, to as long as 99 weeks. That program has helped millions of people get by, but the NBER study suggests it’s also kept the labor market much much weaker than it would have been otherwise — in fact, they calculate, unemployment is 3.6 percentage points higher than it would be if benefits hadn’t been extended.

There are essentially two factors affecting whether the unemployed find jobs: the intensity of their search (supply of labor) and the quantity of jobs available (the demand for it). Unemployment benefits affect both. Conservatives often use the first variable to explain their opposition to unemployment benefits: Unemployed workers will be more selective in choosing a job than they would be without said benefits. Liberals like to caricature this as “the notion that giving people who lose their jobs unemployment benefits makes them lazy and unlikely to look for work” (h/t to Tim Carney for this quote), but, of course, incentives matter, and people do turn down jobs if it means losing benefits.

But unemployment benefits don’t just affect supply, they also affect demand. Why? The benefits allow workers to demand higher wages than they otherwise would, which drives up their wages. When prices go up, demand goes down – meaning fewer jobs. In this case, maybe millions fewer.

In the authors’ words, “Extending unemployment benefits exerts an upward pressure on the equilibrium wage,” which makes it less profitable to hire additional workers, unnecessary job vacancies, and more unemployment.

Past studies have analyzed the effects of unemployment benefits on the intensity of job searches, the supply side, but this study’s results on the demand side are astounding: “During the Great Recession, unemployment benefits have been on average at 82.5 weeks for approximately 16 quarters. . . . Translating this to rates, we would predict a rise in unemployment from 5% to 8.6%.”

In technical terms: Cowabunga!

By continuously extending unemployment benefits, the authors calculate, the federal government has increased unemployment by 3.6 percentage points. As explained, this is because extensions of unemployment benefits discourage workers from accepting lower-wage jobs and cause employers to create fewer jobs.

One would assume wonky reporters would have written extensively about this study, since it deals with such a hot topic. It’s not as the debate itself hasn’t gotten plenty of coverage: A simple Google search reveals dozens of articles and blog posts about the effects of extending unemployment benefits on unemployment rates. Yet I’ve only come across one post that deals seriously with the NBER study, from Danny Vinik of Business Insider — and even then, he makes some important mistakes.

He deems the study’s results “hard to believe,” because they imply “that high unemployment is almost entirely the result of unemployment benefits, not a lack of demand.” But our economy has now been growing for years, including fairly robust growth in recent months — aggregate demand is, in fact, increasing. Yet we’ve seen, as you’ve probably repeatedly heard, more or less a “jobless recovery.” Unemployment rates usually lag economic growth, since employers tend to think long-term, but extending unemployment benefits has made the rates lag even more, or in the authors’ lingo, “remain persistently high.”

Vinik thinks he’s found another mistake when he writes that the authors’ “model also presumes that extended unemployment benefits would drive up wages of incumbent workers by increasing their bargaining power.” But it doesn’t: Their model assumes that unemployment benefits drives up the wages demanded by non-incumbent (or unemployed) workers and the wages offered for job vacancies, an uncontroversial assumption. This contracts the demand for labor — of course incumbent employees can’t use unemployment benefits to demand higher wages, since they can’t get benefits if they quit their jobs.

He does raise one important concern: He criticizes the data source employed by the authors, county-level BLS data on jobs that are extrapolations, rather than actual employment numbers. I don’t have much to add to this point, except to note that researchers often are faced with incomplete and imperfect data, and have to do their best with their available resources. Hopefully, the authors will respond to this critique.

Despite his errors, Vinik deserves credit for at least engaging with a study that counters his own preferred policy agenda. To hear President Obama, members of Congress, and most liberal policy journalists tell it, extending unemployment support offers multifarious benefits, and no downsides at all — we’ve long known that’s not true, but the NBER paper offers a new, and potentially substantial, problem with the idea.