The Proposition Sounds Plausible,
But the Data Don’t Support It
As life expectancy increases and the retirement income system contracts, households face an enormous challenge in ensuring a secure retirement. Working longer is the best way to increase retirement incomes. But some suggest that more work by older persons reduces the job opportunities for younger persons.
This contention, known as the “lump of labor” theory, is widely accepted in many European countries and has provided an economic rationale for early retirement programs. However, economists in the United States generally reject this theory, arguing that the labor market is dynamic and the economy can adapt to labor force changes. Nevertheless, “crowding out” has received increased media attention in the wake of the Great Recession and, if generally accepted, could impede the trend towards working longer.
My colleague April Wu and I just completed an investigation of whether any empirical support exists for the lump-of-labor theory. The existing literature on the relationship between the labor force participation of younger and older individuals is relatively small. A series of papers examined whether employment of older individuals crowds out employment of younger individuals in 12 countries, including the United States. Based on individual country and cross-country analyses, none of these international studies finds evidence that increasing the labor force participation of older persons reduces the job opportunities of younger persons.
Skeptics could argue that the international study did not fully investigate the issue. The authors were constrained to methods and data that could be applied to all 12 countries for ease of comparison. Further, they measured the impact of older workers’ labor force participation only on employment of younger workers, ignoring any potential impact on hours worked or wages. Moreover, the period they examined was before the Great Recession and, whatever the likelihood of crowding out in a growing economy, the dynamics might be very different in a stagnant one.
Our study uses data from the nation’s largest annual labor market survey, the Current Population Survey, and exploits the variation across states in labor force activity of both the old and young for the period 1977-2011. The sample, which consists of state averages for individuals aged 20 to 64 in the survey year, is divided into three age groups: 20-24 (the “young”), 25-54 (the “prime-aged”), and 55-64 (the “old”). The variables of interest include labor force participation, employment and unemployment, hours worked, and wage rates.
The basic equation is estimated separately with three different dependent variables to capture various measures of youth labor force activity: the unemployment rate, the employment rate, and the number of hours worked per week. If crowding out were occurring, an increase in the employment of older persons would increase youth unemployment, decrease employment, and reduce hours worked. However, the coefficients all show the opposite effects.
The question is how robust are these results? Would slicing and dicing the data in different ways produce different answers? We found that the patterns are consistent for both men and women and for groups with different levels of education. We found that the results held when we controlled for differences among states. And we found that the effects of older worker employment on other segments of the labor market during the Great Recession do not differ from those during typical business cycles. In short, we beat the horse to death.
Convincing employers and policymakers that the lump-of-labor theory does not hold is extremely important, given the state of the U.S. retirement system and the need for people to work longer in order to have a secure retirement. Employers already have reservations about older workers, so adding the false argument that retaining older workers hurts younger ones could impede the ability of older workers to remain in the labor force. Therefore, we need to put the lump-of-labor theory to rest. The theory may sound plausible, but the data do not support it.