Since the late 1970s, social science researchers, the media, private foundations, and policymakers have directed considerable attention to the labor market problems of young adults and their families. Most of this attention has focused on high school dropouts, the poor, minorities, and inner-city youth. But an equally troubling-and broader-problem has received comparatively less notice: the steep and sustained decline since 1973 in the real (inflation-adjusted) earnings of young men and women generally. Even adjusting for demographic and socioeconomic characteristics, the labor market problems of young workers are disproportionately severe-they include higher than average unemployment and relatively low earnings when employed. This sustained drop in earnings has been especially dramatic for young adults with no postsecondary school education.
Most proposed remedies have emphasized the quality of the labor supply. But improving education and training, while often worthwhile and necessary, is not by itself sufficient to raise earnings. If this downward trend, which has persisted through recession and recovery alike, is to be reversed, then policymakers and educators must address the demand side as well as the supply side. Raising young adult wages will require not only better academic performance, training, apprenticeships, and school-to-work programs, but also full-employment policies, changes in the configuration of jobs and careers, and larger young adult union membership.
FALLING WAGES
Prior to 1973, the annual and weekly earnings of both young adults and older workers had been improving markedly. Between 1967 (the year the Bureau of Labor Statistics began tracking weekly earnings of wage and salary workers) and 1973, the real median weekly earnings of 16- to 24-year-olds rose by approx i mately 8 percent. Since 1973, however, the earnings of young adults have fallen almost continuously. Between 1973 and 1979, the weekly earnings of young men working full time fell by 7 percent. Young men experienced a 19 percent decline in earnings (a real value of $72 per week) between 1979 and 1989. This decline cannot be attributed solely to business cycle contractions. About half of the 19 percent decline did take place during the recessionary period of 1979-1982. But between 1982 and 1989, a period of strong overall job growth, the weekly earnings of young men fell by another $33, or 9 percent. Earnings declined still more between 1989 and 1994, dropping yet another 9 percent. The result of all this decline? A young man under 25 years of age employed full time in 1994 earned 31 percent less per week than what his same-aged counterpart earned in 1973.
Annual earnings trends display the same pattern. Using findings from the U.S. Census Bureau's annual work experience surveys, we estimate that between 1973 and 1993 the median real annual earnings of young males employed on a full-time basis for at least 27 weeks fell from $18,600 to $13,700, a decline of 26 percent. Young male high school dropouts experienced a 32 percent decline in real annual earnings, while high school graduates with no college education experienced a 29 percent decline. In 1993 a young male high school graduate earned in real terms only what a comparably aged high school dropout was earning in 1973. And a four-year-college graduate in 1993 earned only slightly more than a high school graduate earned 20 years earlier.
Young women also experienced earnings declines over this period, but of a smaller relative magnitude. From 1973 to 1994 the real median weekly earnings of young women fell 14 percent-compared to a decline of 31 percent for young men. Because the earnings declines among young men have been significantly larger than the earnings declines among young women, the median weekly wages of young men and women have substantially converged. Whereas in 1973 the average young woman working full time earned only 75 percent of what her male counterpart earned, a young woman in 1994 typically earns 95 percent of what her male equivalent does. Near-equality of earnings between the sexes, sadly, has been attained only as part of a broader decline in the earnings of all young adults.
Of course, wage declines during this time have clearly not been confined to young adults. The earnings of full-time employed males older than 24 have also declined. The relative size of older men's earnings declines, however, has been much smaller-only a 9 percent decline for men over 25 compared to 31 percent for men aged 16 to 24. As a consequence, the relative weekly earnings position of young men has deteriorated dramatically since 1967. In the late 1960s, a young man employed full time could expect to earn nearly three-fourths of what an older man earned. Until 1973, a majority of young men could earn enough to support a family. By 1994, however, their relative weekly earnings had dropped to only one-half the earnings of older men, leaving them hard pressed to contend with the costs of raising children.
Declining Weekly Earnings
Median weekly earnings (in 1993 dollars) of 16-24 year olds, 1967-1994
Year Men Women 1967 $386 $294 1973 $417 $315 1979 $388 $299 1982 $349 $293 1989 $316 $286 1994 $286 $271
These earnings declines would be less consequential if workers could look forward to higher earnings growth in their later years. Unfortunately, the earnings of most older males have also failed to keep pace with inflation over the past two decades. Using annual earnings data from the Census Bureau's March Current Population Survey, we can project that the expected lifetime earnings of an employed male in 1973 would have been $1.482 million. By 1993 an employed male's expected lifetime earnings had declined to $1.312 million, a drop of 11 percent, despite the fact that the pool of employed males is now better educated. These declines have had the strongest percentage effect on least-educated workers: Employed males with fewer than 12 years of schooling have suffered lifetime earnings losses of an estimated 33 percent. Earnings of high school graduates have declined 26 percent.
Declining Lifetime Earnings
Young adults beginning work now will earn considerably less over their lifetimes than their counterparts 20 years ago.
Years of Education 1973 1993 Percent Change All $1,481,690 $1,312,397 -11% Less than 12 Years $1,119,503 $750,072 -33% 12 Years $1,468,467 $1,093,488 -26% 13-15 Years $1,634,382 $1,273,995 -22% 16 Years $1,976,460 $1,678,596 -15% 17 or More Years $2,161,950 $1,995,593 -8%
ECONOMIC ADOLESCENCE
The steep downward trend in the earnings position of young men has lengthened the period of "economic adolescence," during which young adults are working but not earning enough to be economically self-sufficient or capable of supporting a young family. This development has, in turn, had a number of damaging consequences for young men and for society at large. Among the effects of this protracted adolescence are:
- a sharp increase in the age of first marriages;
- lengthier stays in the homes of parents;
- a rise in young single-parent families;
- reduced economic support of children;
- the increased economic attractiveness of drug sales and other illegal activities;
- the sustained rise in the numbers of young men incarcerated in jail and prison.
While we do not believe that economics is destiny, we do believe that changes in the labor market can in large part account for these wider social phenomena.
Between 1967 and 1973 the real median weekly earnings of young men grew by 1.3 percent per year. If the American economy had continued to generate weekly earnings gains at this modest rate from 1973 until 1994, the median real weekly earnings of young men would have been nearly twice as high in 1994 ($551, in 1993 dollars) as the actual observed earnings were during that year ($286). One does not have to be an economic determinist to believe that doubling real wages might alleviate some of our current social problems.
POSSIBLE CAUSES
Explanations of wage trends typically fall into three general categories: supply factors, de mand factors, and institutional factors.
Supply-side factors. It is commonplace to attribute the declining wages of the young mainly to the quality of the labor pool. Too few of the young, supposedly, are equipped to compete in today's job market. But conventional supply-side explanations do not account very well for declining wages. If anything, supply-side forces ought to have led to increasing wages. Young men are better educated than they used to be, and there are fewer of them relative to the overall population-so their earnings should have gone up.
The number of 18- to 24-year-olds in the civilian noninstitutional population peaked in 1981, at just under 29 million. By 1993 this number had declined to 24.1 million. Young adult workers comprised a smaller share of the workforce in 1993 than at any time since before the baby boomers began entering their young adult years, in the early 1960s. So the quantity of young adult workers was not a factor in these wage declines.
What about the quality of the young adult labor pool? Despite common misconceptions, young workers employed full time for 27 or more weeks in 1993 were considerably better educated than their peers in 1973, with a lower fraction of school dropouts and a considerably higher share with at least one year of postsecondary schooling. Moreover, test score findings by the National Assessment of Educational Progress (an assessment of the reading, math, and writing proficiencies of primary and high school students funded by the U.S. Department of Education) has revealed a moderate upward trend in reading proficiency and no significant change in mean math proficiency in American 17-year-olds between the early 1970s and the late 1980s. Even Charles Murray and the late Richard Herrnstein conceded, in The Bell Curve, that the average verbal and math skills of recent high school students were as good as, if not better than, the skills of equivalent students in the 1950s. Though even higher skills may be required in this information age, young adults' skills have not deteriorated over time.
One supply-side development that may, in part, account for the relative earnings deterioration of some young adults is the increase in immigration of the 1980s and early 1990s. Since many immigrants are good substitutes for younger, less-educated U.S. native-borns, increased immigration may explain part of the decline in the earnings of young dropouts, especially males. Immigrants comprise a particularly high percentage of younger, lower-skilled workers. In March 1994 nearly 12 percent of 16- to 24-year-old males were foreign born, including 29 percent of high school dropouts. Although the evidence from economic studies concerning the impact of immigration on the earnings of native-born citizens is mixed, some academic analysts (such as George Borjas, Richard Freeman, and Lawrence Katz) attribute a substantial portion of the declining earnings of less-skilled Americans to the increased supply of immigrants. If their analysis is correct, this raises questions about the advisability of maintaining current immigration flows.
Demand-side factors. Prior to 1973, many young men worked in manufacturing industries, which have historically paid above-average wages. But because of structural changes in the domestic economy and increasing import competition, young men are now much less likely to be employed in mining and manufacturing, and much more likely to be working in retail trade and private services. The share of young men employed in goods-producing industries fell from 54 percent in 1973 to less than 39 percent by 1993. In contrast, approximately 27 percent of all young men employed full time in 1993 worked in retail trade industries, with another nearly 20 percent working in private and public service industries. More than 70 percent of all net new jobs created between 1992 and 2005 will be in the retail trade and service industries, according to Bureau of Labor Statistics projections. Clearly, the future earnings power of today's young adults will be critically influenced by the salary and wage structures in these industries.
However, it is easy to overstate the earnings impact of the shift in young male employment from manufacturing to trade and services. We estimate that only 13 to 16 percent of the actual decline in the annual earnings of young men can be explained by changes in the industrial distribution of jobs. Even though manufacturing jobs have paid above-average wages in the past, the differences between what young men earned in different industries in 1973 were actually fairly small-so the effect of the shift away from manufacturing should not have had a substantial impact on earnings. Over the past two decades, young male earnings have declined within virtually every major industry, including goods-producing industries, so that even those employed in manufacturing are earning less in real terms than their 1973 counterparts.
It turns out that much more important than interindustry shifts have been the declines in earnings within industries. Only those employed in public administration posted gains in real earnings between 1973 and 1993, and they accounted for only 2.5 percent of employed young men in 1993. Between 1973 and 1993, the mean real annual earnings of employed young men fell by 22 percent. Earnings declines varied considerably by sector, ranging from lows of 2 percent in personal and entertainment services and 6 percent in professional services to highs of 23 percent in manufacturing, wholesale trade, and business and repair services, and 28 percent in retail trade.
Might some portion of these declines be attributable to decreased labor productivity? In part, yes: The Bureau of Labor Statistics has estimated that output per hour in such key retail trade industries as food stores, grocery stores, retail bakeries, variety stores, and eating and drinking establishments actually declined between 1973 and 1992. But while labor productivity was declining in the retail sector, it actually rose strongly in most manufacturing industries. Yet young men's real earnings in manufacturing industries also declined considerably-by 23 percent-over this period. Clearly there is no longer a guarantee that the benefits of increased labor productivity in manufacturing will be passed on to young manufacturing workers, a disconcerting development considering that a strong, direct relationship between rising productivity and rising young workers' earnings prevailed between the 1940s and the early 1970s.
Some economists have attributed these within-industry earnings declines to "skill-biased technological change"-shifts in demand in favor of more highly skilled workers capable of handling the requirements of sophisticated and computer-based technologies. But studies have shown that the widespread introduction of new technologies from the late 1970s onward was not accompanied by any major increased demand for highly skilled workers [see David Howell, "The Skills Myth," TAP, Summer 1994]. More likely, technology affects the young adult labor market by making it easier for large companies to avoid hiring young workers, to globalize their operations, and to contract out to small, low-wage suppliers-in short, by enabling employers to pursue low-wage strategies that depress the wages of young adults.
Institutional factors. At least three institutional forces explain, in varying proportions, the decline of wages among lesser-skilled young adults: the decline of trade unionism, the drop in the real value of the federal minimum wage, and the large-scale "restructuring" of American corporations.
Unions boost wages but today's young workers are seldom represented by labor unions or other employee associations at the workplace. During 1994, fewer than 8 percent of employed men under the age of 25 were members of labor unions, a membership rate only one-third as high as that of their older male counterparts (aged 25-64).
Erosion of the federal minimum wage has contributed to rising earnings inequality among young men, particularly between those with earnings in the top and bottom 10 percent. Even so, the Clinton administration's modest proposed increases in the federal minimum wage would have no measurable effect on the real median weekly earnings of full-time employed young men. A hike in the minimum wage to $5.15 per hour over the next two years, combined with projected annual consumer price index increases in the 3 to 4 percent range, would yield a real weekly wage equal to only 44 percent of the median real weekly earnings that young men obtained in 1973 [see John Schmitt, "Cooked to Order," and Barry Bluestone and Teresa Ghilarducci, "Rewarding Work," TAP, May-June 1996]. The bottom layer of young, full-time earners would benefit from a minimum-wage increase, but the median weekly wage would remain unchanged. The entire wage distribution for young workers must be lifted-not simply the bottom layer-if earnings are to be restored to 1973 levels.
Finally, efforts to boost labor productivity and economic competitiveness have led to the adoption of new technologies, capital investments, labor deployment strategies, corporate restructuring efforts, and compensation policies that have substantially reduced the demand for and pay of semiskilled blue-collar workers and lower-level white-collar workers. Increasingly, younger adults without a college education have been relegated by companies pursuing short-term profits to being part-time, temporary, or just-in-time workers. The labor costs of short-term workers are more subject to management control than permanent workers.
WHAT MIGHT WORK
America's labor market crisis for young adults is not limited to minorities, dropouts, or the products of single-parent families. It is, to use Wallace Peterson's expression, a "silent depression," a set of deep and sustained earnings declines. Moreover, there is no evidence that these youth will "grow out" of their lower-earnings trajectory as they get older. Efforts to improve real earnings prospects for young adults will only become more challenging in the years ahead as demographic changes yield increased supply pressures. What can be done to address this problem?
Full-employment policies. The federal government's efforts to move the nation's labor markets to true full employment (a 4 to 4.5 percent unemployment rate) would benefit young adult workers by expanding overall job opportunities and increasing access to year-round, full-time jobs. Greater competition among employers for younger workers would reduce their ability to offer only part-time and contingent positions and would lift earnings at the low end of the wage distribution. Support by the Federal Reserve Board for higher growth rates would ideally be complemented by coordinated efforts to create both private-sector jobs and subsidized jobs for out-of-school youth in high-poverty neighborhoods in central cities and rural areas. The U.S. Depart ment of Labor is beginning to fund several demonstration programs in this area, but pilot programs are only a modest first step.
Improving education. Academic skill level remains a strong predictor of future earnings, strongly influencing access to both apprenticeship training and formal employer training. Employers want high levels of basic academic competency in their entry-level workers. A high school diploma no longer guarantees employability.
School-to-work and apprenticeship programs. Schools cannot solve the labor market problems of young adults by themselves. If they could, the rising proficiency levels among high school graduates and the reduced numbers of school dropouts would have translated into higher wages. Ultimately, it is the nation's employers who control the critical hiring, training, and compensation policies that determine the labor market fortunes of young adults. High schools, community colleges, and technical training institutes need to work more closely with employers to help students gain access to jobs, work-based learning, and formal training opportunities, including apprenticeship programs. In recent years, fewer than 1 percent of the nation's high school juniors and seniors have been involved in regular or youth apprenticeship training programs, and only 1 to 2 percent have been enrolled in an apprenticeship program three years after high school graduation. To this end, we recommend a number of different programs:
First, states should make sustained efforts to expand the number of work-based learning and youth apprenticeship programs with monies from the recently enacted School-to-Work Opportunities Act. This act, passed in 1994, was designed to encourage an integration of academic and vocational education and to increase opportunities for work-based learning. Expanding in-school employment opportunities, especially for those students not expecting to immediately enroll in four-year colleges, is a desirable strategy. Recent research by Chris Ruhm of the University of North Carolina at Greensboro has revealed that work experience in the senior year favorably influenced the annual earnings of young noncollege-bound adults seven to nine years after graduation.
Second, state and local education and training agencies should leverage funds from state workforce development budgets and unemployment insurance trust funds to pay for programs that train young workers in transferable occupational skills. Paying for this training with state money gets around the problem of companies not wanting to pay for the training of a young adult who might end up employed by a competing firm.
Third, federal Pell grants (currently available to low-income youth to finance postsecondary education and training) can be broadened to purchase slots in occupational skills training programs. Research over the years has consistently shown that the economic impacts of vocational and technical education are critically influenced by the ability of graduates to secure jobs that are closely related to the occupational skills they acquire.
Fourth, greater funding should be provided to federal and state apprenticeship boards. Appren ticeship and employer-provided training, particularly formal training, has been consistently found to significantly raise the real earnings of workers. A summary of the findings of an array of studies on the impacts of training by Lisa Lynch of Tufts University revealed that the average hourly wages of workers receiving training rise by 7 to 13 percent. This training also benefits employers by raising the productivity of such workers within their firms.
What clearly don't work are employment and training programs that involve only limited, nonintensive interventions and that lack close ties to jobs and employers. Studies of such programs reveal that they have no lasting effects on earnings. High school diploma equivalency programs not tied to either postsecondary educational opportunities or intensive job placement efforts also have limited effects on the earnings of former dropouts. There is no cheap human capital solution for the steep earnings declines of young men with no postsecondary schooling.
Finally, economic policymakers, employers, and labor unions need to consider seriously the extent to which a number of the labor market problems currently confronting young adults have been caused by past "solutions" to other problems. Efforts to boost productivity, especially in manufacturing, have had the unintentional side effect of lowering the demand for blue-collar and low-level white-collar workers. New corporate strategies, in pursuit of lower labor costs, have eroded the wages of young adults. The key to reversing young adult wage trends requires not only full-employment policies but also effective education and training programs. Only by providing more opportunities for year-round, full-time work-and by increasing the productivity of this work-can we hope to effect a sustained rise in the market wages of young adults.
