US Labour Force Participation Rates

Richard Berner on the MS GEF:

Do Teens Mask Labor-Market Weakness?
June 22, 2007

By Richard Berner | New York

Here’s a puzzle: Despite slower US economic and job growth, the unemployment rate has failed to rise. Some think that teenagers are the culprits; they are dropping out of the labor force, especially lately. The teenage labor force fell by 352,000 in the first five months of 2007, resulting in a 240 basis point decline in the teen participation rate (the ratio of labor force to population) to a seasonally-adjusted record low of 41%. This 5-month dip in the participation rate was the largest since the 2001 recession, and accounted for half of the decline in the overall participation rate during that period. Other things equal, had those teens kept looking for work without finding jobs, the jobless rate would have been 4.7% in May, not 4.5%. Thus, the teen exodus appears to be holding down the unemployment rate and masking a weakening labor market. The deceleration in average hourly earnings from 4.3% in December to 3.8% (year over year) in May seems to confirm such a softening.


I’m suspicious of that line of reasoning. As I see it, recent teen behavior simply renews a 28-year downtrend in teen labor force participation, and while hourly wage gains may have peaked, other wage metrics show ongoing acceleration. Employment gains have slowed, and probably will continue to do so. Along with some rise in participation, that will likely boost the jobless rate, and wage pressures will level off. But most labor-market indicators so far do not suggest underlying weakness in either the job market or wages.

Both demand and supply matter for the balance in labor markets. Regarding demand, I think employment growth will decelerate in coming months, but the slowing probably will be gradual. That’s because companies’ hiring discipline has left some pent-up demand for hiring. As one measure of pent-up demand, job opening rates (the availability of unfilled jobs relative to employment) from the Job Openings and Labor Turnover Survey (JOLTS) remained close to new cycle highs in April, especially in professional business services, education and health, manufacturing, and state and local government. Thus, prospective monthly job gains may slow to 100, 000 to 125,000.

Labor demand has lagged cyclical norms in this long expansion, but labor supply has lagged even more. The labor force participation rate has declined by 1.2 percentage points from its peak seven years ago, following a four-decade-long uptrend. Whether this decline is a temporary or a more lasting development is critical to the analysis of labor-market slack. There were certainly cyclical elements in the falling participation rate, which gathered momentum in the 2001 recession. Those are well and truly behind us. In contrast, I’ve long felt that the decline in the participation rate is mainly secular (see “Are Labor Markets Tight?” Global Economic Forum, July 22, 2005).

At work are three significant structural changes in labor markets. The first, the decline in teenage labor-force participation, has actually been underway for nearly 30 years, challenging the notion that the recent decline is a new, cyclical development. Rather than looking at monthly data, labor-market analysts typically view the July participation rate for teens as a benchmark, because participation during the school year is erratic. So measured, and before seasonal adjustment, teen (ages 16-19) participation rates plunged to 53% in 2005 from a peak of 72% in the mid-1970s, and actually rose to 53.5% last year, perhaps as labor markets continued to firm.

The reasons for this secular decline aren’t clear, but a study by the Bureau of Labor Statistics suggests that, at least since 1994, teens are putting more emphasis on school, both in the summer months and during the school year (see “Declining Teen Labor Force Participation,” Bureau of Labor Statistics, Issues in Labor Statistics, September 2002). While those data may exaggerate the focus on scholastic endeavors, summer school teen enrollment rates jumped from 19.5% to 27% between 1994 and 2000 — a period of booming job growth. Work by economists at the Federal Reserve Bank of Chicago confirms this secular explanation (see Daniel Aronson, Kyung-Hong Park and Daniel Sullivan, “Explaining the Decline in Teen Labor Force Participation,” January 2007). Consequently, I don’t think this trend has much to do with jobs being easy or hard to find.

That view isn’t universally shared however; some view the recent plunge in the teen participation rate and the concurrent rise in the teen unemployment rate over the past five months to 15.7% as an ominous sign that teenagers are dropping out because they can’t find work. I’ll concede that some jobless teens might give up looking for work and go back to school, but that’s unlikely in the middle of a school year. And I simply don’t trust the monthly teen unemployment and labor force data enough to base strong conclusions on them. The CPS sample from which the data are derived covers 60,000 households, and the month-to-month variation can be substantial. As a result, we must wait for July 2007 data to see whether the teen participation rate has really plunged and whether the teen jobless rate has risen.

A second secular trend: The aging of the population is boosting the share in the population of groups that historically have had lower participation rates. This shift in composition is depressing aggregate labor force participation. A Federal Reserve study calculated that such shifts added about 0.6 percentage point to the aggregate participation rate between 1980-95, and subtracted about 0.4 percentage point between 1995-2005 (see Stephanie Aaronson, Bruce Fallick, Andrew Figura, Jonathan Pingle, and William Wascher, “The Recent Decline in Labor Force Participation and its Implications for Potential Labor Supply,” Brookings Panel on Economic Activity, March 2006). A recent update strongly hints that those demographic shifts will continue to depress labor force participation, as it is clear that only heroic increases in geezer labor force participation will offset population aging (see Jonathan Pingle, “The Outlook for Labor Supply in the United States,” presented at “Labor Supply in the New Century,” Chatham, Mass, June 18–20, 2007).

Nonetheless, I think the debate about labor force participation for older cohorts in the future is still wide open because several factors may influence change. There are three traditional legs to the retirement saving stool. Working longer, in my view, is the fourth critical leg, and future policy changes in Social Security and other retirement saving incentives matter for labor force participation. And there is a fifth leg: Access to health insurance. Many stay at work to get it, and retire once they are eligible for Medicare. Thus, any changes to health care financing or Medicare have the potential to trigger significant change in labor force behavior (see Alicia Munnell and Steven Sass, “The Labor Supply of Older Americans,” presented at “Labor Supply in the New Century,” Chatham, Mass, June 18–20, 2007).

However, a third and final secular factor reducing the labor force is unlikely to change. The long upswing in female labor-force entry began to abate in the late 1980s and ended in the late 1990s. The participation rate for women 20 years and older rose by more than 20 percentage points in the 40 years ended in 2000 but has since flattened out. The peaking is especially obvious among women aged 35-44: Relative stability between 75-78% in that participation rate followed a rise of more than 30 percentage points through the 1990s. That secular plateau in female participation probably won’t reverse. Until 2004, many women apparently decided to leave the labor force either to start families or to spend more time with their children. More recently, the strength of labor demand has lured some back. Women with their own children under 18 years of age numbered about 36.5 million in 2006, and account for roughly half the female labor force between the ages of 20 and 54. Even with the recent rebound, their participation rate has dropped two percentage points in the past six years, and participation for those with kids under 6 years old has declined by nearly as much. Whether due to poor job prospects, greater affluence, or other reasons, women in greater numbers have chosen families over paying jobs.

In sum, these labor-market trends make this expansion — and those to come in the future — different from those in the past. Teens are less likely to be a ready source of summer labor. An aging population and the end of the long upswing in female participation probably have reduced — and will continue to reduce — growth in the labor force. Thus, the norms for job growth consistent with a stable jobless rate will be lower than in the past.

Cyclical factors also matter, however, and I still think labor force participation will get a cyclical lift. But most participation rates will only continue to rise if labor markets remain tight and pay rises enough to make the payoff from looking for a job worthwhile. An upswing in overall participation actually began in 2005; today the participation rate for adult men is up about half a point from its recent trough. The increase has been concentrated among men 35 years and over. In addition, Fed authors cite evidence that labor force participation may have gotten a temporary cyclical lift in the 1990s from welfare reform, and the end of that boost may have unmasked the underlying trend in participation. In all, while cyclical factors may boost labor supply, labor markets likely will loosen only gradually and wages could accelerate further.

Good inflation news and renewed subprime mortgage woes have helped contain bond yields over the past week, although markets remain volatile. The decline of inflation to low levels is a key reason that the hurdle to tightening monetary policy is high. Nonetheless, Fed officials likely will agree that there is scant evidence of increasing slack in either labor or product markets, and thus there remain upside risks in an uncertain inflation outlook. That means the hurdle to easing monetary policy is also high.

These changing trends add risk to the outlook for slack in labor markets. The unemployment rate still seems likely to rise as job growth slows and labor force participation either holds steady or rises slightly. While job growth could slow more abruptly than we expect, there is also a risk that a slowing labor force would prevent the unemployment rate from rising.