The US employment-population ratio—the share of the civilian population ages 16 and older that is currently working—has fallen from 62.8 percent in the fourth quarter of 2007 (the peak of the business cycle prior to the Great Recession) to 60.3 percent in the first quarter of 2018, a 2.5 percentage-point decline.1 While this decline is substantial, it is entirely the predictable consequence of the aging of the US population, as the movement of a larger fraction of the population into older age groups that are less likely to work has mechanically lowered the overall employment rate. However, when adjusted for age—that is, applying earlier employment rates for different age groups to the current, now older population—the employment rate is essentially back to its pre-recession level.
The roughly unchanged age-adjusted employment rate over the last decade reflects very disparate experiences for different age groups. Among those ages 55 and older, the age-adjusted employment rate rose by 1.7 percentage points over this period, while it fell by 2.3 percentage points for those under age 25 and by 0.4 percentage point for those ages 25 to 54. This is the same pattern as in the previous business cycle, though the pace of change has generally slowed, particularly for older and younger workers.
By itself, this analysis cannot answer the question of whether or by how much the US economy is operating below its potential, or, in other words, how much slack remains in the US economy today. The answer to this question also depends on other factors like the state of employment in the fourth quarter of 2007, the base period used for this analysis, and also trends in part-time employment. Ultimately, any complete assessment of US economic potential would need to examine not just labor market quantities but also trends in wages and prices.
Adjusted for Aging, the Employment Rate Is Back to Where It Was Prior to the Great Recession
Older Americans are much less likely to work than younger Americans: The employment rate for workers ages 25 to 54 averaged 79 percent in 2017, while the average for workers ages 60 to 64 was 55 percent. As a result, the aging of the population will tend to drive the employment rate down, even if employment rates for particular age groups (such as older men or younger women) remain constant. This represents a purely "demographic" effect that largely reflects child-bearing decisions made decades ago and which, as a result, does not reflect current economic conditions.
To determine how much the employment rate has declined purely as a result of the aging of the US population, we can ask a straightforward question: What would the overall employment rate be today if the population maintained its actual age structure, but the likelihood of employment for individuals at a particular age was the same today as it was in 2007? If this were the case, the employment rate in the first quarter of 2018 would be 60.2 percent, which is slightly lower than the actual employment rate of 60.3 percent in the first quarter of 2018. In other words, adjusting for the aging of the population, the employment rate rose slightly over that period as shown in figure 1 (for more detail on this calculation, see the technical note at the end of this blog post).
The Employment Rate Has Fallen for Younger Workers and Prime-Age Workers and Has Risen for Older Workers
The roughly flat age-adjusted employment rate from the fourth quarter of 2007 to the first quarter of 2018 shown in figure 1 represents the net effect of changes in participation rates since 2007 within particular age groups. Table 1 shows the effects by decomposing this age-adjusted increase into the portions attributable to six age-sex groups.
|Table 1 Contribution to non-aging component of decline in US employment-population ratio, 2007:Q4–2018:Q1
|Note: Totals may not sum due to rounding.
Source: Bureau of Labor Statistics, Current Population Survey; authors' calculations.
As table 1 shows, within-group declines in employment rates among younger workers (ages 16 to 24) and prime-age men (25 to 54) have been pulling down age-adjusted employment rates. Younger individuals have increasingly stayed in school and are also less likely than in the past to work while in school. Since prime-age men still represent almost a quarter of the working-age population, even relatively small declines in their employment rate can have a meaningful effect on the overall employment rate. At the same time, employment rates for older Americans have increased, in part due to increases in life expectancy, occupational safety, and educational attainment.
To put some longer-term perspective on these trends, figure 2 shows the age-adjusted changes in employment rates by age and gender since 1981. As the figure shows, the direction of the changes in employment rates in the current business cycle is similar to previous decades (with the exception of prime-age women, who saw increasing employment rates through 2000).
The magnitudes of the changes have evolved somewhat over time, as shown in table 2. Relative to the 2001–07 business cycle, the current cycle has featured slower employment declines for younger workers, slower employment gains for older workers, a similar downward drift in prime-age male employment, and a considerable improvement in prime-age female employment. Relative to earlier decades, male employment trends are roughly similar while trends for women have worsened, consistent with an overall decline in women's employment since around 2000.
|Table 2 Change in age-adjusted employment-population ratio by age
(percentage point change, annual rate)
|Source: Bureau of Labor Statistics, Current Population Survey; authors' calculations.|
Labor Force Participation Rates, Adjusted For Age, Are Slightly Below Their Pre-recession Rates
The employment-population ratio reflects a combination of two factors: the labor force participation rate (i.e., the share of the population working or looking for work) and the unemployment rate (i.e., the share of the labor force that is not working). On an age-adjusted basis, the slight improvement in the overall employment rate between the fourth quarter of 2007 and the first quarter of 2018 reflects a 0.6 percentage point reduction in the unemployment rate outweighing the 0.3 percentage point decline in the labor force participation rate.2
Specifically, the labor force participation rate fell over 3 percentage points from the fourth quarter of 2007 through the fourth quarter of 2013 and has been roughly stable since then. The underlying aging of the population, however, should have pushed the labor force participation rate down steadily over this period. As a result, the labor force participation rate is now very close to where it was at the end of 2007 after adjusting for aging—as shown in figure 3. The underlying age-gender components have evolved similarly to the ones described above for the employment rate.
Is the United States At or Above Full Employment?
This analysis should not be taken as a definitive comment on whether the US economy is operating above, at, or below full employment. Any such analysis would need to consider not just quantities, as discussed above, but also prices and wages, both of which provide evidence about the degree of slack in an economy.
Nevertheless, the analysis above does help frame the argument about where the United States is relative to full employment:
The case for the United States being above full employment would argue that age-adjusted employment rates are higher today than they were at the end of 2007, a period when the economy was slightly above full employment. Moreover, the unfavorable trends for prime-age workers are not unique to this expansion: The employment rate for prime-age men failed to recover to its pre-recession peak in all but three expansions in the postwar period. In fact, the decline in prime-age employment rates is smaller than it was in the last business cycle. Finally, the sustained high levels of unemployment, especially long-term unemployment, seen during the last recession would be expected to have persistent effects on employment rates, effectively lowering the employment rate corresponding to full employment. To the degree that the non-age component of the change in the employment rate is affected by these factors, it would mean the economy is currently at or above full employment.
The case for the United States being below full employment is that trends prior to the current expansion in fact reflected chronic underemployment; as a result, age-adjusted employment rates that are even higher than those in late 2007 do not indicate that the economy is operating above capacity. Moreover, proponents of this case could argue that the improvement in employment rates for older workers has been common across advanced economies—and may be a structural indicator of, for example, the decline in manual jobs—and thus should not be credited as an offset for the decline in employment rates for prime-age workers, which are a better proxy for labor market slack. Another argument for the continued existence of slack is that the fraction of the workforce employed part-time for economic reasons—those currently working part-time but who would prefer a full-time job—remains 0.2 percentage point above its value in the fourth quarter of 2007.
Time will help distinguish which of these two perspectives is correct. Still, what is unambiguous is that the United States has enjoyed a steady improvement in age-adjusted employment rates over the last seven years—but these improvements mask the lack of a full recovery in employment for prime-age men.
To calculate the effects on employment rates and labor force participation of the aging of the US population, counterfactual quarterly histories of these rates are constructed using data from the Current Population Survey conducted by the US Bureau of Labor Statistics. Employment and participation rates for age-sex groups are held fixed at their 2007 annual average values, while each age-sex group's population share evolves along its actual path since the fourth quarter of 2007.
Data for the following age-sex groups are used in constructing the counterfactual (since they are the most granular data published by the Bureau of Labor Statistics on a not seasonally adjusted basis):
To correct for slight differences resulting from the use of non-seasonally adjusted data, the within-group employment and participation rates are scaled such that the counterfactual rates are equal to the actual, seasonally-adjusted participation rate in the fourth quarter of 2007 for each age-sex group for which seasonally-adjusted data are available.
The age-adjusted change in the employment and participation rates represents both the effect of changes in these rates since 2007 within particular age groups and the interaction of these within group changes with changing demographics. The non-aging contribution for each of the six groups included in table 1 is calculated by determining the change in the group-specific employment rate that is not due to aging and multiplying it by the group's share of the working age population in the first quarter of 2018. The change in the age-adjusted employment rate for each group is the difference between the total change and the change due to aging for that group (which is calculated following the same approach used to determine the change in the overall employment rate due to aging).
1. Figures may not sum to totals due to rounding. The remainder of this blog post refers to the employment-population ratio as the employment rate.
2. The age-adjusted unemployment rate here is calculated using relative population shares of each demographic group, not the relative labor force shares.
The data underlying this analysis are available here [zip].