The effort to extend unemployment insurance is often portrayed as an unambiguously good idea. The basic premise, as I understand it, is that unemployment is entirely about demand and supply is of no consequence. There are no jobs, and thus the extension of unemployment insurance is a humanitarian measure and that’s it. But it’s not clear that this is always true, as Casey Mulligan argues:

“Traditionally, many economists have been leery of prolonged unemployment benefits because they can reduce the incentive to seek work. But that should not be a concern now because jobs remain so scarce, said Lawrence Katz, a labor economist at Harvard.” as quoted by the New York Times. (See also here for the same claim)

What is the basis for this claim? I’m not sure, but Professor Shimer kindly pointed me to an article about unemployment duration in Pittsburgh 1980-85. Unemployment rates got quite high in Pittsburgh in those days, reaching 16 percent at one point, and staying over 10 percent for two and a half years. The chart below shows some of the results. It graphs weeks from unemployment benefit exhaustion against the fraction of unemployment people either finding a new job or being recalled to a previous job in that week. “Exhaustion” refers to the time when benefits cease being paid to the unemployed person, regardless of whether they have found a job.

The striking chart suggests that exhaustion of unemployment benefits proved a sharp and effective spur to job-seekers.

Almost no one started working during the 2-3 weeks prior to the exhaustion of their unemployment benefits (weeks “-3″ and “-2″ in the chart). Miraculously, more than one quarter started work a week later (19% started a new job, 10% returned to a previous job). Economists agree that a huge reason for this behavior is that people are more willing to remain unemployed when unemployment itself generates a paycheck. (The job they take may not be great, but the data show that often there is a job to take).

Mulligan then linked to an earlier post for the NYT on what he calls the “summer jobs surge.”

Suffice it to say, not everyone embraces Mulligan’s worldview. He consistently maintains that we’ve pursued a long series of employment-reducing policies under President Bush and President Obama. By way of explanation, this is how Mulligan characterizes the impact of mortgage modification programs:

So when I say that “mortgage modification discourages people from earning too much” that does not only mean quitting one’s job to “game” the mortgage modification formula, but also failing to increase one’s income or take an new job when the opportunities arise.

One thing I find interesting is that many of the middle-class professionals I know think in precisely these terms when it comes to freelance work — given taxes, is the gig valuable enough to merit taking time away from family and other priorities? — yet consider it deeply offensive to suggest that people who face straitened economic circumstances see the world in roughly the same way.

I don’t actually have a settled view on the question of extending unemployment insurance. The jobs tax credit proposal seems worse, insofar as most of it will go to employers who already intended to hire. At least in theory, an extended period on unemployment insurance could allow one to upgrade skills and seek more remunerative employment opportunities down the road, as Raj Chetty’s research suggests. Tim Harford summarized Chetty’s work in Slate late in 2008.

But new research from Raj Chetty, a young Berkeley economist, suggests that moral hazard may not be why more generous benefits seem to lead to more unemployment. Chetty realized that unemployment benefits do not merely pay people to stay out of work; they also protect them from having to rush into an unsuitable job. It is nothing to celebrate if unemployed engineers cannot afford to spend three months finding a job for which they are qualified but are forced to work as real estate agents to put food on the table. A longer gap between jobs is sometimes preferable.

We don’t have any randomized experiments to test the hypothesis, Harford notes, but we do have a kind of natural experiment.

Chetty and two colleagues looked at the system in Austria, where severance pay is due to anyone employed for more than three years. By looking at—for example—a factory closure in which lots of staff are fired simultaneously, they could treat severance pay almost as a randomized experiment. Those lucky enough to get severance pay spent more time looking for a new job, despite the fact that severance pay provides no direct incentive to stay out of work.

In The Natural Survival of Work, a wonderful book that I very highly recommend, the authors offer a comparative analysis of unemployment insurance systems. Bruce Bergman of the Bureau of Labor Statistics wrote a useful guide to the book.

Looking for a job “ensures the reallocation of the labor force toward the most efficient jobs, and thus constitutes an essential source of growth.” In practice, however, unemployment insurance and employment services have varying degrees of effectiveness. In 2005, the average duration of unemployment in the United States was 18 weeks, while in France, it was 15 months. In trying to get behind what works well and what does not, the authors describe the history of French trade unions, which were created to facilitate job placement and information sharing. They conclude that a credible system of checking job-search activity is imperative. Citing examples from Switzerland and the Netherlands, the authors assert that public employment services, though, must go beyond simply checking to provide real assistance to jobseekers.

As we extend unemployment insurance, it’s worth wondering whether or not we’re doing it in a coherent, thoughtful, cost-effective way that will truly aid jobseekers, or if we’re doing the politically expedient thing. That said, it’s not obvious to me that Senator Bunning’s approach is the best way to devise a more coherent unemployment insurance system. But as always, I’d be interested in hearing your thoughts on this.