Center on Wage and Employment Dynamics

In April 2010, a record 45.9 percent of the unemployed were counted by the Bureau of Labor Statistics as long-term unemployed—defined as durations of six months or more. Those who were unemployed for more than a year, whom I shall call the very-long-term unemployed, numbered 23 percent of all the unemployed. We know that long-term unemployment and verylong-term unemployment generate serious and long-lasting harm to millions of individuals and to the economy. The scarring effects upon the economic, mental and physical health of longterm unemployed workers and their families are well-documented (von Wachter 2010). We also know that many of the very long-term unemployed eventually leave the labor force permanently, and some of those end up on the disability rolls. Very-long-term unemployment consequently generates adverse effects upon the Treasury and upon the capacity of the economy to grow in the long-run.

In the past two years the proportion of the unemployed who have been out of work six months or longer has grown to levels not seen since the Great Depression (Chart 1). In the recession phase– from the fourth quarter of 2007 to the end of the second quarter of 2009, when GDP growth resumed—the unemployment rate rose very rapidly and to a level higher than had been forecast. Given the length of the recession phase—over 18 months—it is not too surprising the number of long-term unemployed grew at about the same rate as the number of all the unemployed (Chart 2).

In the recovery phase that began in April or May of 2009, what went up quickly did not come down quickly. Thus far the overall unemployment rate has remained within a narrow range —hovering during the past year first a bit below 10 percent, then rising to a bit above 10 percent, and more recently, remaining a bit below 10 percent. But in the same period the long-term unemployment rate has continued to skyrocket as rapidly as during the recession phase. By April 2010, the proportion of the unemployed with jobless durations of six months or more had reached 45.9 percent. By comparison, in the 1957-59 recession this proportion peaked at about 10 percent; in 1982-83, it peaked at about 26 percent.

The surprising and continuing growth of long-term and very-long-term unemployment well into the economic recovery and in the face of repeated extensions of emergency benefits, up to 99 weeks in some states, poses important questions for employment policy. Have these unprecedented benefit extensions themselves kept very-long-term unemployed workers from searching for and accepting job offers? Will the economic recovery eventually reduce the number of very-long-term unemployed without further interventions?

According to the best credible research studies, the answer to the first question is a very clear no. In previous recessions, when extensions of time limits on UI benefits were more modest, unemployed workers receiving UI benefits on average stayed out of work only one or two more weeks. Then, as now, well under half of the unemployed were receiving any UI benefits at all. In the current environment, with more than five job searchers for every job opening, the evidence suggests that such effects are even smaller (Valletta and Kuang 2010).

The adverse effects of extending benefit limits beyond 99 weeks are also likely to be small. Employers generally choose to hire new labor force entrants or unemployed workers with short unemployment spells over those with longer spells. The very-long-term-unemployed are thus the least likely to receive job offers. Moreover, the remaining financial assets of this group are especially meager (Chetty 2010). Consequently, they are the least likely group among the unemployed to refuse job offers they do receive simply because they can continue to collect unemployment insurance benefits.

The very-long-term unemployed will be the last group to benefit from an economic recovery. Will the economic recovery nonetheless be strong enough to eventually help the very long-term unemployed? The Council of Economic Advisors, the Congressional Budget Office and many economists of all political stripes have forecast a very slow decline in the unemployment rate and argued that the continuing high level of unemployment is best combated by additional government-led increases in aggregate demand (Romer 2010; Zandi 2010). Programs such as relief for the states and relief for teachers facing layoffs are indeed called for, and on a large scale. Absent such policies, unemployment will decline only very slowly. Consequently, the argument for immediately renewing emergency unemployment benefits as well as extending recipient limits beyond 99 weeks in higher unemployment states is especially compelling. So, I would add, is the case for direct job creation programs, such as those proposed in Rep. George Miller’s Jobs for America Act.

In the remainder of this testimony, I will address why unemployment rates, short-term and long-term, have risen substantially higher than public and private forecasters predicted. I will also propose needed policy changes that would reduce the ranks of the unemployed and that would improve the workings of our labor market. Most of my proposals do not add big-ticket items to the Federal budget, so their merits can be considered independently of the question of how large the federal deficit should be at this time.

Why did the overall unemployment rate increase so much more than was expected? I argue that employers’ weakening attachment to their workers, a development that began in the 1980s, has heightened the labor market’s response to economic downturns. Well-working labor markets must maintain a balance between flexibility and security. But the increase in labor market flexibility has undermined this balance, generating costs for employers and workers alike. Unfortunately, the UI system itself creates employer incentives that have reinforced the turn to greater flexibility. I suggest some reforms of the UI system that would reverse the incentives in the system and improve labor market behavior.

My discussion of the rise of long-term unemployment begins with a usual suspect: the disappearance of many jobs that may never return, in industries such as manufacturing, construction, and the finance, insurance and real estate sector. A usual argument is that unemployed workers are locked in to these industries because they do not have the skills to transfer to industries that are or will be growing. My reading of the research literature suggests that the evidence for this argument is mixed. A more compelling explanation is based upon the long duration of the recession period and the weakness of the recovery, which keeps feeding the ranks of the very-long-term unemployed. The large numbers of workers who already have been unemployed for one year or more are going to be joined by many, many others. There are programs already in place to help workers weather difficult transitions. I suggest changes in some of the existing programs so that they can help the very-long-term unemployed more effectively.