By John Robertson and Ellyn Terry
The labor force participation rate (LFPR) is an important ingredient in projecting employment growth and the unemployment rate. However, predicting the LFPR has proven difficult. For example, in 2011 the Congressional Budget Office (CBO) projected that the LFPR in 2015 would be about 64.3 percent. In reality, the LFPR turned out to be 62.6 percent. Based on the CBO projection, the economy would have needed to create about 4 million more jobs to reach the 2015 unemployment rate of 5.3 percent.
Why is the LFPR so hard to predict? Leaving aside the challenge of projecting the size of the population, movements in LFPR primarily reflect shifts in the age distribution of the population as well as a number of behavioral factors. Although the aging trends are largely baked in, the behavioral factors vary over time. According to our estimates, about half of the 3.4 percentage-point decline in the LFPR between 2007 and 2015 is the result of the aging of the population, while behavioral factors account for the rest.
The complication is that the specific behaviors can change. The following chart shows a decomposition of the change in LFPR from 2007 to 2011 and from 2011 to 2015. Though the aging of the population contributed about the same amount to the decline in LFPR in both periods, the contributions from other factors have varied a lot. (We delve into the changes in the factors following the chart.)
Aging: The single largest factor contributing to the decline in the overall LFPR has been the rising share of older Americans in the population. In 2007, about one in five Americans were over 60 years old. In 2015, almost one in four were over 60. Moreover, this demographic force will continue to suppress the overall LFPR as the share of older Americans increases further in coming years. (For an in-depth discussion of the economic implications of an aging population—including changes in the labor market—please read the Atlanta Fed's 2015 annual report.)
Later retirement: One countervailing factor to an aging population has been the rising LFPR of older individuals. The retirement rate of those over 60 declined between 2007 and 2011 by a similar amount as it had before the recession. However, the trend toward later retirement has slowed considerably in recent years. The reason for this slowing is a puzzle and has important implications for the future course of overall LFPR.
Schooling among the young: The enrollment in educational programs by American youth has been generally rising over several decades, and this trend has put downward pressure on the overall LFPR. During the 2007–11 period, the share of 16- to 24-year-olds who do not want a job because they were in school or college accelerated relative to the prerecession trend. However, enrollment rates have since flattened out. The slowing may reflect enrollment rates catching up to the longer-term trend or may be a result of changes in the opportunity cost of education.
Not in the labor force but want a job: The share of the population saying they want a job but are not classified as unemployed by the U.S. Bureau of Labor Statistics definition is countercyclical—it tends to go up during bad times and down during good times. The relative size of this group increased between 2007 and 2011 and has since retraced about half of that increase as the economy has strengthened. We expect that this category will continue to shrink some more as the economy continues to expand.
Health: The share of individuals who do not want a job because they were too ill or disabled to work has increased over time. The relative size of this group increased between 2007 and 2011. Since 2011, the rate of increase has slowed, and it actually declined in 2015. It is not clear what drove the larger increase during the 2007–11 period, but there is some literature linking weak labor market conditions to poor health outcomes.
Prime-age reasons for not wanting a job (other than health): During the recession, the share of prime-age (ages 25 to 54) women not wanting a job because of household or family responsibilities decreased. One explanation is that some women entered the labor force to help make ends meet. At the same time, there was an offsetting effect from a rise in educational enrollment. Since the recession, nonparticipation because of household or family responsibilities has returned to near prerecession levels, and educational enrollment has leveled off.
To summarize, we find that relative to the 2007–11 period there has been a:
- flattening in the rate of retirement by older individuals,
- flattening in the rate of educational program enrollment by younger individuals,
- declining share of the population saying they want a job but not officially counted as unemployed,
- smaller drag from nonparticipation because of health, and
- larger drag for reasons other than health among prime-age individuals.
Where will LFPR be by the end of 2016? What about five years from now?
During the first three months of 2016, there has been an increase in the overall LFPR. This was largely the result of a decline in the share of prime-age people citing health reasons for nonparticipation, with some contribution from a decline in the share who want a job but are not "unemployed."
Though these boosts to participation may offset the effect of an aging population in the short term, most forecasts have the LFPR declining over the next several years. How much participation will actually decline depends on the answers to several difficult questions, such as: Will older individuals push retirement out even farther? Will school enrollment rates rise more rapidly again? Will the health status of the population improve? The difficulty of answering these questions helps explain why making accurate labor force projections is challenging.