- Comments and Discussion
The recovery from the 2001 recession pushed the unemployment rate down to a level that would have seemed low by the standards of the two decades that ended in 1995. By March 2006 the civilian unemployment rate was just 4.7 percent. GDP growth between 2001 and 2005 was moderate, and improvements in after-tax incomes fueled a substantial rise in personal consumption.
Two aspects of the recovery sparked concern, however. First, employment growth was exceptionally slow in comparison with that in all previous postwar recoveries. Fifty-eight months after the business cycle peak in March 2001, payroll employment was only 1.8 percent higher than it had been at the peak. In the average previous postwar expansion lasting at least fifty-eight months, payroll employment increased more than 9 percent. Even in the weakest previous postwar expansion, employment grew 6.8 percent.1
Second, five years into the expansion there was little evidence of recovery in the before-tax incomes of median- and low-income households. Real wages and compensation earned by people holding jobs either improved or declined only slightly, depending on the wage benchmark used. But because households on average had fewer earners or had breadwinners who earned paychecks for fewer weeks in a year, real household income for families with average or below-average income remained lower in 2005 than it had been at the peak of the last expansion. Even though gross income per person was higher in 2005 than it had been in 2000, a large fraction of the before- and after-tax income gains were captured by a small fraction of households. [End Page 135]
This paper by Stephanie Aaronson and her coauthors assesses one possible explanation for some of the peculiarities of the recent recovery. One reason that employment gains may have been modest in an environment of low and declining unemployment is that the labor force participation rate has been unusually depressed. The BLS estimates that between 2000 and 2005 the overall participation rate fell 1.0 percentage point. The rate among males 16 and older fell 1.5 percentage points, and the rate among females fell 0.7 percentage point. With declining participation rates, it is conceivable that the unemployment rate did not provide a reliable signal of job market distress. If potential job seekers dropped out the labor force instead of looking for jobs, both the unemployment rate and the labor force participation rate would have been lower than in previous economic expansions. For example, if all of the decline in labor force participants were added to the ranks of unemployed job seekers, the unemployment rate in 2005 would have been 6.5 percent instead of 5.1 percent. Of course, it is very unlikely that all of the drop in participation between 2000 and 2005 was traceable to job seekers' distress over their labor market prospects. But if dropping out of the labor force was exceptionally common after 2001, it provides a partial explanation for the combination of slow employment growth and a low unemployment rate.
Among the questions posed by the authors are these:
—. How much of the 2000-05 decline in labor force participation is explained by predictable demographic factors, such as the aging of the population?
—. How much is explained by normal cyclical factors?
—. How much can be explained by an examination of trend influences on age-specific participation rates, such as lower fertility, changing retirement incentives, and a bigger payoff to staying in school?
—. How much of the decline in participation remains unexplained?
The authors' summary of their analytical findings seems to me sound. Most of the drop in participation between 2000 and 2003 was due to normal cyclical factors, and the partial recovery of participation since 2003 brings it close to the trend line that one would expect given the shifting age composition of the population. The long-term trend in participation is downward, as the large baby-boom generation moves toward retirement and begins its withdrawal from the workforce. In early 2006 the overall participation rate was about where one would expect if age-specific participation rates had remained unchanged after...