Whether our current, relatively low unemployment rates can be sustained without increasing inflation has been a point of contention among economists, financial market analysts, and policymakers. Arguments over trade-offs between unemployment and inflation, however, too quickly gloss over the causes of our falling unemployment. The cause seems relatively straightforward: a sustained upturn in the business cycle. But this simple explanation overlooks another potent factor: the rising rate of labor force participation.
Compared to previous economic expansions, the economic growth of the last few years has actually been quite weak. During the previous two economic expansions, of 1975-1980 and 1982-1990, real economic growth averaged 4.3 percent and 3.6 percent per year, respectively. Real economic growth since the end of the last recession in early 1991, by contrast, has averaged an anemic 2.7 percent per year, even taking into account the recent economic surge. It seems improbable that such slow economic growth is alone responsible for our current low rate of unemployment. But if the business cycle is not solely responsible, what else is?
One deeper explanation is a quiet but fundamental shift in the workforce over the last few decades, reflecting demographic and cultural changes. These shifts, the focus of this article, deserve far more scrutiny. If they are even partially responsible for our current low unemployment rates, they could easily produce a further drop in unemployment and a substantial increase in wages. But given the Fed's aversion to tight labor markets, that happy result could produce more restrictive monetary policy and even slower growth.
While the unemployment rate has waxed and waned with changes in the business cycle, employment as a percentage of the population has increased quite dramatically. In 1963, the last year of the Kennedy administration, just 35.8 percent of the population worked. Today that number is 47.9 percent. This job growth has paralleled a substantial increase in labor force participation. In 1963 just 38 percent of the U.S. population was in the labor force. Today, thanks to a huge infusion of baby boomers and women, more than 50 percent of the population is in the labor force—working or seeking work.*
Given this sizable influx of job seekers, it is not surprising that the U.S. economy has experienced periodic high unemployment. On the contrary, what is surprising is that unemployment rates have been as low as they have been. The American economy has shown a tremendous capacity to put job seekers to work.
But even despite a small jump in recent months, this steady increase in people seeking employment seems to be reaching its natural limits. Growth in labor force participation rates, which began in the early 1960s, peaked in the 1970s at about two-thirds of a percentage point per year. The rate of growth began to slow in the 1980s, dropping roughly in half, and finally slowed to a crawl from about 1989 onward, when increases averaged about 0.07 percent per year, or roughly one-tenth the rate of increase during the 1970s. The influx of baby boomers into the job market, which crested in the late 1970s, has played out, and women's participation rates now nearly equal men's.
Flattening labor force participation rates spell good news for unemployment, as fewer applicants chase the existing supply of jobs. But they spell bad news for economic growth, which has been boosted substantially by the increased output of the new labor force participants. This combination provides a compelling explanation for our otherwise confusing low-growth, low-unemployment economy. Slower workforce growth may also help explain why unemployment is just now returning to levels rarely seen since the early 1970s, when the most dramatic increases in labor participation were just getting started. Employment, which has at times over the last three decades lagged well behind accelerating labor participation, may finally be catching up.
To fully understand its impact, imagine that the labor force growth rate in 1996 was what it averaged during the 1975-1980 expansion. The workforce would have grown by 3.5 million instead of 1.6 million, adding an additional 1.9 million job seekers to the 7.2 million already unemployed. Absent any additional job growth, the unemployment rate would have risen to 6.7 percent in 1996, up from 5.6 percent in 1995, instead of dropping to 5.4 percent as it did. In fact, had Bill Clinton faced the same labor force trends that confronted Jimmy Carter, last year's elections might have turned out quite differently.
Moreover, the aging of the labor force may be just as important as changes in its size. Younger workers possess fewer skills and encounter greater difficulty in both obtaining and retaining jobs. The baby boomers exacerbated unemployment not just because of their sheer numbers, but because of their youth and inexperience. The baby bust could have the reverse effect. As young workers become a smaller fraction of a labor force that is itself already shrinking, unemployment may fall even more.
In a sense, the effect of the baby boom on labor force participation has camouflaged a longer-term trend. Since as far back as World War II, male labor force participation rates have actually been dropping, mostly due to the increasing custom of retirement and the lengthening life spans of retirees. At some point this drop will occur among female workers as well. When that happens, overall labor force participation rates will contract, probably around the time baby boomers begin retiring 15 years from now. So our current labor surpluses could easily become labor shortages. Generalized unemployment might virtually disappear.
WAGE INFLATION AND THE FEDERAL RESERVE
Left unimpeded, unemployment, already hovering close to 5 percent, could easily dip toward 4 percent in the next few years. Four percent unemployment is not unheard of in this country. It once was the level economists considered full employment.
But Alan Greenspan and his compatriots at the Federal Reserve are not likely to welcome that prospect. Fearing inflation, they would feel enormous pressure from the financial community to raise interest rates to halt the downward trend. Of course unemployment has dropped to its current levels with little or no impact on inflation (perhaps due, at least partially, to the restraining impact of international trade on prices and wages). But to expect wage inflation to remain low indefinitely as unemployment continues to drop is overly optimistic. At some point the Fed would feel compelled to act.
Raising interest rates to forestall wage inflation would jeopardize the ongoing health of the entire economy, and for very dubious reasons. In the current economic climate, wage increases are something to embrace, not something to fear. Real wages have been allowed to stagnate for far too long, notwithstanding recent debates by the Boskin Commission and others about the real rate of inflation and its impact on the value of those wages. Even accepting the Boskin adjustments, which remain controversial, real growth has slowed since the postwar boom and income inequality has widened. According to the Economic Policy Institute, from 1973 to 1995 real median wages dropped from $11.02 to $10.13 per hour (1995 dollars), a decline not significantly offset by a slight increase in fringe benefits. Over that same period the percentage of workers earning poverty-level wages—$7.28 or less in 1995 (1994 dollars)—increased from 23.5 to 29.7 percent.
Falling wages are often blamed on a host of factors, including slower productivity gains since 1973, dropping rates of unionization, increased trade with low-wage countries, the loss of relatively high-paying manufacturing jobs and their replacement with low-paying service-sector jobs, and a failure to raise the minimum wage enough to keep pace with inflation. But basic supply and demand at a grander scale has probably also played a role. Over the last several decades, with the labor pool expanding so quickly, the oversupply of potential workers has, no doubt, put significant downward pressure on wages. If a tightening of the labor supply is at long last turning the tables on employers, forcing them to pay more, so much the better. After years of profiting from a surplus of labor, it seems only fair that employers pay more once the surplus has dried up.
True, some of those wage increases will be passed on to consumers in the form of higher prices, and since workers are also consumers the real value of their increased wages will be somewhat diminished. But some of the cost also will be passed on to business owners and stockholders in the form of reduced profit margins. For stockholders in particular, given the rising stock prices of recent years, this seems a small price to pay. Even if real median wages are increased by just 1 percentage point a year, keeping pace with rising productivity, it would be a significant improvement over the record of the past few decades—and an improvement with practically no inflationary implications for the economy as a whole.
Putting aside the possibility of Fed-induced stagnation, an economic future featuring lower unemployment and rising wages sounds rosy enough. But it would be a mistake to assume that demographics alone will solve these problems entirely, even if they do have a positive influence. Even if unemployment drops, the depth of our unemployment problem is much deeper than official figures indicate. In 1996, a year when the official unemployment rate was just 5.4 percent, the total number of unemployed and underemployed was actually 9.7 percent. On top of the 7.2 million Americans who were unemployed, an additional 1.6 million were discouraged or otherwise marginally attached—workers who were neither working nor looking for work, but would have taken a job if one were offered. Another 4.1 million Americans, desiring full-time jobs, were involuntarily working part-time.
Much of this unemployment and underemployment is highly concentrated in inner-city and rural communities that are too often left behind during periods of economic prosperity. Even during times of economic growth, the unemployment rate among 16- to 19-year-old black males, for example, ranges from 30 to 50 percent. If future job growth continues to concentrate in the suburbs, many of our hardest-hit communities may continue to miss out on the benefits of an improved economy.
If the problem were just physical location, empowerment zones and transportation subsidies might be the answer, the first taking jobs to people, the second taking people to jobs. Unfortunately, the situation is more complicated than that. Job openings typically occur all along the skills spectrum, while a disproportionate number of the unemployed are unskilled. Unless enough low-skill jobs or training slots are created, a general labor shortage caused by changing demographics is no guarantee that unemployment will be eliminated.
Another potential problem has more to do with the theory itself. Neoclassical economists would argue that changing labor force participation rates have little effect on unemployment, because prices (in this case, wages) adjust to reflect changing quantities. If the labor supply expands, wages will drop until the excess labor is fully utilized. As labor force participation rises, employment will too—as has happened. If a subsequent contraction of the labor force takes place, as may occur in the next 15 years, it will simply have the reverse effect. A tighter labor supply will push up wages, forcing many employers out of the labor market, until a new equilibrium is reached at a lower level of employment. As employment paralleled the expanding labor force on the way up, it will parallel it again on the way down, leaving unemployment unchanged.
Keynesians would dispute this simple view of labor markets and unemployment, but the potential impact of contracting labor force participation rates on unemployment should be interpreted cautiously. Interestingly, since the neoclassical view predicts rising future wages, it still leaves open the possibility of a Fed-induced recession. Knowing which view is correct will require waiting a few years. But since neither view predicts the elimination of unemployment entirely, government activism on the issue will still be necessary.
What form should activism take? The Clinton administration's primary approach to joblessness is a combination of job training and job subsidies. Unfortunately, these initiatives are probably too limited to have much impact, and may be the wrong policy solutions even if pursued more aggressively. While job training, if performed effectively, could help some of these workers, for many more the real educational challenge is basic literacy, not a lack of higher-order skills needed in our increasingly high-tech economy. Nor will training programs obviate the need to create more entry-level jobs.
Job subsidies, an increasingly prominent fixture in the Clinton administration's arsenal of policy prescriptions since the enactment of last year's welfare reform law, are also a less than satisfactory answer. Studies by the New York-based Manpower Demonstration Research Corporation indicate that job subsidies have had, at best, a limited impact on unemployment. Job applicants who should benefit from such subsidies are too often stigmatized by them instead, and this hampers their job search efforts. When they do find jobs, the subsidy typically rewards employers for hiring someone they would have hired anyway, or subsidizes the hiring of one worker at the expense of another—with little or no net impact on unemployment.
Ultimately, eliminating the last vestiges of unemployment may require both a friendlier Fed and a targeted public jobs initiative modeled after the Works Progress Administration (WPA), a program that em ployed millions of Americans during the height of the Great Depression. The nation's experience in the 1970s with the poorly run Comprehensive Employment and Training Act (CETA) program, our most recent experiment in public service employment, may make the creation of such an initiative more difficult politically. But workfare programs popping up all over the country in response to last year's welfare reform law may give the concept another chance. Unfortunately, workfare poses its own problems, including the potential for subminimum wages, inadequate child care, nonuniversal eligibility, and possible displacement of government workers, to name only a few. But if workfare can be transformed into a true public service employment program with jobs paying a livable wage, it could have an enormously positive impact. Certainly the Clinton administration's Nation al Service program has shown that community service is an idea not completely without merit. And if national unemployment levels continue to drop, the cost of eliminating the last trace of unemployment through such programs will become substantially more affordable.
True full employment is a worthy goal. Too often policymakers are lulled into believing that because an official 5 percent unemployment rate is low by both international and recent historical standards, there is nothing left to do. But "low" unemployment, by these measures, is not low enough. It still leaves millions of Americans desperate to earn a living for themselves and their families. Calling our current levels of unemployment "low" really says more about our standards than it does about the economy.
*Employment and labor force participation rates are both expressed as a fraction of the entire population, not the noninstitutional civilians aged 16 and over, a more traditional measurement that undervalues the impact of baby boomers entering the workforce.
Back to text