Where will all the smart kids work?

Authors:  Levy, Frank
Murname, Richard J.

 

For nearly twenty years, the United States has suffered from slow growth in labor productivity and, as a result, very slow growth in wages. With the publication of Education 2000, President Bush's program for school reform, better education has become one official mechanism to restore economic growth and rising living standards.

A skilled labor force is one element in economic growth, but there are others elements as well. And so we can ask whether it is realistic to use growth and competitiveness to sell school improvement. Simply put, if school reforms prove successful, will all the smart kids get good jobs? Or will we come to resemble the world described by Schumpeter in which an oversupply of highly educated people leads to underemployment and disaffection?( n1)

One can approach this question from many directions. Even within our own discipline, economics, more than one approach is possible. When the question is posed in Economics 1, the answer is that an increased supply of educated people unambiguously leads to economic growth. But in more advanced courses, a new body of theory is leading to a richer, more ambiguous story. This new view--a view which, to our mind, better reflects reality--shows the complexity and subtlety involved in industrial restructuring.

Better Schools, Faster Growth?

Before proceeding to theory, it is useful to provide some context for the school reform issue. Economists date the slowdown in productivity growth to 1973. Between 1946 and 1973, output per hour of work (labor productivity) had grown between 2.5 and 3 percent per year. Between 1973 and 1979, it grew by .9 percent per year. Since 1979, it has revived slightly to about 1.25 percent per year.( n2)

Had this slowdown not occurred, the average forty-year-old man would now be earning (conservatively) about $48,000. In fact, he earns about $34,000, not much more than similar men earned twenty years ago. One need not be an economic determinist to see this stagnation as one force behind the taxpayers' revolt and fears of a vanishing middle class.( n3)

The problem of economic stagnation has only recently entered public debate,( n4) but it has been discussed in professional journals for at least a decade. Attention to this issue has jumped from one to another unicausal explanation including the poor quality of American management and the low U.S. rate of savings and investment (Hayes and Abernathy 1980; Peterson 1987).

The connection between stagnation and education was first made in a dramatic way in Workforce 2000 (Johnston and Packer 1987), a report prepared by the Hudson Institute for the U.S. Department of Labor. In it, the authors argued that the U.S. faced a large and growing mismatch in which the occupational structure would require increasing levels of skill while new workers--particularly the growing proportion of young adults from poverty and minority backgrounds--would enter the labor force with low skill levels.

Other studies were addressing the problem of U.S. educational quality: A Nation at Risk (1983), The Forgotten Half (William T. Grant Foundation 1988), and periodic test reports from the National Assessment of Educational Progress. But these reports often focused on the adverse consequences of poor schooling for today's students. By contrast Workforce 2000 made the broader point that a poor quality labor force would impede general economic growth and thereby affect all of us, not just new graduates. It was a short step to the presidential connection between better schools and faster growth.

Academic Exercise

To proclaim such a connection does not mean it exists. And so it is useful to see what both theory and data tell us. In the world of Economics 1, the theory is straightforward. Profit maximizing managers are constantly searching for new and more profitable ways to produce desired products. The competition and bidding among these managers results in two outcomes: Each resource is put to its most productive use, and each resource is paid for its marginal product in that use--i.e., the value of its contribution to the firm's output.

In this perspective, the wage trends of different groups of workers provide important readings on the economy's evolution. Here, too, the results appear to be straightforward. The most dramatic feature of the 1980s has been the increased earnings premium for attending college. Today, for example, a thirty-year-old man with four years of college earns an average 50 percent more than a thirty-year-old man with a high school diploma, the highest that ratio has been in the post-World War II period.

This rising wage ratio suggests that college graduates, and by implication, well-educated workers of all kinds, are in scarce supply. One can imagine, for example, that managers are foregoing the introduction of advanced technologies in manufacturing and service industries because the skilled labor required to run them is now simply too expensive. With an increased supply of highly educated workers, innovation would proceed apace with a resulting spur to economic growth.

Note that in this story managers are not appropriate subjects of study. Rather, they are passive agents whose behavior reflects efficient responses to market forces. In the process, they are continually forced to do the most with what they have. Do employers provide enough training? In this story, employers provide the amount of training that maximizes their profit--no more and no less. In the same way, a greater supply of skilled graduates will make it profitable for employers to introduce new technologies more rapidly. The result will be more good jobs and faster economic growth.

Economic Inertia

The story of competitive forces is elegant in its simplicity and power. And in fact, recent studies have shown that at a point in time, technologically progressive firms have relatively skilled employees (Barter and Lichtenberg 1991). But these correlations do not speak to causation: Were firms induced to adopt progressive technology because skilled workers were available? At least two kinds of data call this causation into doubt. The first comes from the labor market experience of the 1970s. Throughout the post-World War II years, the nation had urged young people to go to college and young people took the advice. During the 1970s, in particular, the large, well-educated cohorts of the baby boom began their careers. Thus the labor market received a flood of highly educated workers--one version of what current educational reform is attempting to achieve.

The results were not extremely encouraging. The rates of productivity growth and wage growth, which had collapsed in 1973, showed no sign of revival. The previously mentioned college income premium for thirty-year-old men fell from about 35 percent in the mid-1960s to 15 percent by the end of the 1970s.( n5) The situation was best described by labor economist Richard Freeman in his book The Overeducated American (1976) in which he argued that we had achieved an oversupply of college educated workers such that college was no longer a good economic investment. In sum, the 1970s offered little evidence that the increased supply of skilled workers available at relatively low wages was the catalyst the economy needed to attain rapid productivity growth.( n6)

The second piece of data comes from the fast-growing literature on trends in earnings inequality. As is well known, earnings inequality among workers increased substantially in the 1980s. Normally, we think of inequality growing between different groups of workers--between blacks and whites or between high school graduates and college graduates. But in practice, increased inequality between groups has accounted for only about one-half of the growth of inequality. The remaining increase in inequality comes from growing earnings differences among apparently similar workers, for example, among twenty-five- to thirty-four-year-old white male high school graduates. Detailed studies by economists, including Erica Groshen (1991), suggest an important source of this second variation involves wage differences among plants in the same industry that are stable over time.

Interplant wage differences run counter to the Economics I model described above. Recall that model suggests market forces will push all employers to adopt the same best technology. If that were true, we would expect all firms in an industry to pay roughly the same wages.

The existence of large interplant wage differences suggests that employers with quite different technologies (and quite different wage structures) can survive in the same market and that day-to-day market forces may not automatically cause employers to change their ways. This fact, like the historical experience of the 1970s, suggests the economy involves a large dose of inertia, which can limit its ability to translate a surge of educated workers into a surge in economic growth.

The Incremental Block

What is the source of this inertia? Or, put differently, where is the Economics I story wrong? There are, of course, many possible answers but one, now attracting increasing attention, is the ineffectiveness of incremental restructuring within a firm. The point is best made through examples.

Alexandria Life Insurance( n7) is a highly successful national life insurance company, which prides itself on never having laid off an employee in its 120-year history. For much of the post-World War II period, the firm's market was buoyed by the baby boom and the ever larger number of people who turned twenty-one each year and who needed life insurance policies. But by the early 1980s, the number of people turning twenty-one was declining each year and the company recognized it was competing in a saturated market.

After much study, it decided to approach this market through improved customer service. Previously, most Alexandria employees had learned one or another compartmentalized skill--how to change a policy beneficiary or discuss a particular policy's benefits. As a result, employees often had to respond to customer questions by saying that the question was not in their area, that they would have to transfer the caller to another department, or that they would have to get back with the answer in a few days. By the mid-1980s, large numbers of employees were being trained in multiple skills so that most customer questions could be answered on the spot.( n8) Employees were also equipped with a computer tracking system so that they could give inquiring field agents the status of each pending policy: where it was in the approval process, what information was outstanding, and when the policy might be approved and the commission issued.

For purposes of our argument, the point is the way in which the company's job tenure and training reenforced each other. An insurance firm in a changing market could not afford to preserve employee tenure without retraining its employees. A failure to improve customer service would have ultimately lowered the firm's market share and led to layoffs. But neither could the company afford extensive training in the face of high employee turnover. It would have been too expensive to constantly train new people from scratch. If a firm had attempted to move to this structure incrementally--adding tenure without training or adding training without tenure--it would, if anything, have made things worse.

The structure of Philadelphia Motors, a joint U.S.-Japanese automobile plant makes a similar point. The plant uses a number of features from the Japanese production system for car manufacturing--features that differ sharply from traditional U.S. practices. One feature is its just-in-time inventory system in which suppliers make frequent deliveries of small numbers of parts. As a result, each worker only has access to parts sufficient for, say, the next two hours of production. This stands in contrast to the traditional U.S. worker who has one or two weeks worth of parts next to the work station.

With two weeks of inventory, a few bad parts can simply be discarded. With two hours of inventory, a few bad parts can stop production. In this way, just-in-time inventories create constant pressure to improve part quality.

A second difference involves the definition of worker responsibilities. In a traditional U.S. automobile plant, workers occupy narrowly defined jobs and perform routine, repetitive tasks. Major responsibility for overall quality control lies with inspectors at the end of the assembly line, a group who account for, perhaps, one-third of all production line labor. Philadelphia Motors (and many Japanese plants) organize workers in teams where each worker rotates through a series of jobs and bears a heavy responsibility for error detection and correction.

Here, too, these differences (and others not mentioned) reenforce each other, making incremental change difficult. A just-in-time inventory system grafted onto a 1975 Chevrolet plant would have gained little and, again, might have made things worse. A two-week inventory system coupled with modern production teams would have been better, but poor quality parts would have sharply reduced the team's impact on productivity and car quality.

The Geography of Change

The ineffectiveness of incremental change--the fact that incremental change may even make things worse--explains why the Economics I model can break down and firms may be slow to take advantage of more skilled workers. More precisely, the Economics I story relies on a number of strong assumptions, including the following: If two technologies, A and B, are both profitable, a mixed technology involving a weighted average of A and B will also be profitable.( n9) Through this condition, the problem of choosing the best technology looks like climbing a single peaked mountain without a map. People move step-by-step as long as each step moves them a little higher. When they reach the point where no step can take them higher, they know they have reached the top. In Economics 1, it is the price system that guides the employer step by step in search of greater profits until the most efficient (and most profitable) technology is selected.

We have argued that for Alexandria Insurance and Philadelphia Motors, technologies violate this weighted average condition. For each firm, the attempt to combine elements of a low-skill, low-productivity technology with elements of a high-skill, high-productivity technology could easily have lowered profits.

Here, the geographic analogy is a mountain with multiple peaks. One peak represents low-skill, low-productivity technology. A second, higher peak represents high-skill, high-productivity technology. But there is no step-by-step algorithmic path to reach the second peak from the first. To the contrary, going step-by-step requires descending from the smaller mountain before beginning to climb the larger one, a journey that could easily go off course (unless one has a map). Given this situation, managers understandably are tempted to refine the technology at hand rather than to leap into the void. But the motivation for educational reform begins with the idea that the technology at hand is not designed to employ large numbers of skilled workers.

While the economics of "multiple peaks" or multiple equilibria is gaining increasing currency within the economics profession, it is (implicitly) old hat to geographers and planners. It helps to explain, for example, why the financial industry or the steel industry is heavily concentrated in a few locations rather than spread evenly across the country (Krugman 1991). But the theory has a variety of other applications including, as we have seen, how one thinks about industrial restructuring.

Beyond Tradition

If the economy is more complex than its Economics I version, that does not obviate either the originally stated problem or the need for solutions. Historically, steadily rising living standards have been an important part of the glue that holds the country together. For the past eighteen years, wages have, on average, been stagnant and society increasingly feels the strain of that stagnation. We see ourselves as too poor to deal with problems ranging from creating national health insurance to restoring a measure of safety in our big cities.

So restoring economic growth remains at the top of the national agenda. But growth involves more than simply improving school quality and letting the market take its course. The U.S. has had a tradition for many years of turning to schools to solve problems: to integrate large numbers of immigrants into society, to respond to the challenge of Sputnik, to deal with civil rights. Activism toward education is part of the American tradition and so it should not be surprising that we now turn to the schools to deal with the problem of stagnant growth. The evidence, however, as we read it, is that school reform by itself has little power to unravel national productivity problems.

Equally strong is the U.S. tradition that firms and industries should be left alone to make their own decisions. Incremental decisions are by far the easiest to make and the ones most firms do make. But there is increasing evidence that, in many cases, the accumulation of incremental changes will not produce the adoption of new technology needed to substantially boost productivity. A major question concerns what, if any, public policy actions would encourage firms to make the nonincremental changes needed to move to high-skill, high-productivity, high-wage techniques. It is this question that should be at the center of the debate about planning and industrial policy in the years ahead.

NOTES

(n1.) All those [university graduates] who are unemployed or unsatisfactorily employed or unemployable drift into vocations in which standards are least definite or in which aptitudes and requirements of a different order count. They swell the host of intellectuals in the strict sense of the term whose numbers hence increase disproportionately. They enter it in a thoroughly discontented frame of mind. Discontent breeds resentment. And it often rationalizes itself into that social criticism which as we have seen before is in any case the intellectual spectator's typical attitude toward men, classes and institutions especially in a rationalist and utilitarian civilization (Schumpeter 1962, 152-3).

(n2.) Because the slowdown began in 1973, it is tempting to attribute it to the OPEC oil- price shock, which occurred in that year. But most careful studies have failed to prove this connection and the early 1980s collapse of oil prices did little to restore productivity growth.

(n3.) During the 1950s and 1960s, we perceived the middle class to be growing rapidly. But family incomes were not becoming radically more equal. Rather, family incomes were growing rapidly throughout the distribution. The middle class was growing because more families could afford a single family home, a car, and the other elements of a middle-class life-style.

(n4.) Stagnation has entered the current presidential campaign through two questions: Will our children live better than we do? And will the U.S. economy be competitive in future years? In earlier elections, most economic issues involved inflation, unemployment, and other short-run (i.e., business cycle) phenomena.

(n5.) The declining return to a college education in the 1970s helps explain the rapid increase in the return to a college education in the 1980s. During the late 1970s, male high school graduates slowed the rate at which they went on to college, a partial reaction to the low return. As a result, the number of thirty-year-old male high school graduates grew faster than the number of thirty-year-old male college graduates during the 1980s. This helped to put male high school graduates in excess supply.

(n6.) In fairness, one could argue that the strong 1970s inflation undermined most kinds of rational business decisions including the decision to adopt advanced technology.

(n7.) Names have been changed to protect company identities.

(n8.) Employees were not expected to keep everything in their heads, but they were equipped with a large number of manuals and they were expected to know where an answer could be found.

(n9.) This is a rough statement of the mathematical condition that the technologies form a convex set.

PHOTO (BLACK & WHITE): James Michael Newell, Evolution of Western Civilization, 1936, fresco mural panel, Evander Childs High School, Bronx, NY.

REFERENCES

Bartel, Ann P., and Frank R. Lichtenberg. 1991. The Age of Technology and Its Impact on Employee Wages. Economics of Innovation and New Technology. UK: Harwood Academic Publishers, forthcoming.

Freeman, Richard. 1976. The Overeducated American. New York: Academic Press.

Groshen, Erica. 1991. Wage Variation Among Establishments Within Industry. Quarterly Journal of Economics 106, 3 (August): 869-84.

Hayes, Robert H., and William J. Abernathy.1980. Managing Our Way to Economic Decline. Harvard Business Review 58 (July-August): 67-77.

Johnston, William B., and Arnold E. Packer. 1987. Workforce 2000, Work and Workers for the 21st Century. Washington, DC: Hudson Institute.

Krugman, Paul. 1991. Geography and Trade. Cambridge, MA: MIT Press.

National Commission on Excellence in Education.1983. A Nation at Risk: The Imperative for Educational Reform. Washington, DC: U.S. Department of Education.

Peterson, Peter J.1987. The Morning After. The Atlantic (October): 43-69.

Schumpeter, Joseph A.1962. Capitalism, Socialism and Democracy, 3rd ed. New York: Harper Torchbooks.

William T. Grant Foundation Commission on Work, Family and Citizenship. 1988. The Forgotten Half: Non-College Youth in America. Washington, DC: William T. Grant Foundation.

~~~~~~~~

Frank Levy and Richard J. Murnane

Levy and Murnane are economists at, respectively, the University of Maryland's School of Public Affairs and the Harvard Graduate School of Education.

 

This article is from The Journal of the American Planning Association

Summer 1992, Volume 58

 

Views: 25

Reply to This

© 2025   Created by Rob Sullivan.   Powered by

Report an Issue  |  Terms of Service