The Blame Game

Are public schools holding back the nation's economy? by Richard Rothstein

Despite their disagreements on a wide range of issues during this year's presidential campaign, the two candidates for the nation's highest office seemed to agree on one point: that our increasingly high-tech economy demands radical education reforms.

President George W. Bush stated, "Our new economy requires higher and higher skills." His Democratic opponent, Al Gore, agreed, arguing that we have to reform schools because "60 percent of new jobs will require advanced technological skills."

But despite their consensus on this point, it's not true. Overall, the jobs of the future will require only modest skill growth. In fact, the demand for increased skills is increasing more slowly than in the past.

School improvement may be a good idea, but we need better reasons than a false skills crisis. If we reform schools for the wrong reasons, we will reform them in the wrong ways.

Miscalculated Policy

This misunderstanding about the skills crisis can lead to other more serious policy mistakes. Think back a dozen years ago, when the U.S. economy seemed to be suffering and Americans were enduring widespread unemployment. The United States could not seem to compete with other industrial nations that were exporting higher-quality, less expensive products.

Schools were widely blamed for the country's inability to compete in the global marketplace. Unemployment was high, critics said, because high school graduates lacked the skills of German or Japanese workers.

In 1994, when overall unemployment was more than 6 percent and the unemployment rate for high school dropouts was twice that, Alan Greenspan, chairman of the Federal Reserve Board, insisted that poor schools were the cause. The unemployed, he said, were jobless because they lacked the skills employers needed.

Most economists, Democrats and Republicans alike, agreed, contending that attempts to lower unemployment not only would fail, but would spur inflation.

President Clinton's first appointees to the Federal Reserve Board, Alan Blinder and Janet Yellen, supported Greenspan's tighter money policies aimed at preventing what they believed would be fruitless attempts to reduce unemployment and thus ignite inflation.

Blinder and Yellen reasoned that because all adequately educated workers were employed when joblessness fell to 6 percent, efforts to create new jobs would only cause companies to compete for those already-employed workers by offering them higher wages. Inflation, not more jobs, would follow. The only way unemployment could drop further was if schools improved, producing employable graduates.

A Wrong Assumption

Today, with unemployment hovering around 4 percent, many formerly jobless high school dropouts are working. So we know that it was a huge mistake to have believed that 6 percent—not 4 percent or less—was the target unemployment rate. And we now know that about 2 percent of the workforce, young people we believed were too uneducated to be employed, in fact had skills employers could use.

This 2 percent, some three million Americans, suffered prolonged unemployment because policymakers wrongly assumed they were too poorly schooled to work. Not only did these workers probably have better skills than we assumed, but many jobs require fewer skills than it is fashionable to believe.

Economists attribute today's economic growth with mild inflation not to improved schools, but to factors like greater efficiency, downward price pressure from growing imports, and a willingness to experiment with lower interest rates.

But even as unemployment fell below the level he believed possible without igniting inflation, Greenspan declined to modify his position that schools were responsible for the economy. Before initiating a new round of interest rate increases in 1999, he told the Congressional Joint Economic Committee that "we fall far short" in public education. The nation cannot maintain prosperity, he said, if schools do not improve.

A decade ago, the views expressed by Greenspan, Blinder and Yellen were commonplace. In his best-selling book, Head to Head: The Coming Economic Battle Among Japan, America and Europe, published in 1992, Massachusetts Institute of Technology professor Lester C. Thurow predicted that Europe would "own the 21st century," partly because poor schools make America uncompetitive. "To create the productivity that can justify high wages, American K-12 education will have to improve," Thurow said.

But since then, productivity in the United States has soared, wages have recovered and America's economy has left Japan's and Europe's in the dust. It seems that today's growth in productivity stems from years of accumulated innovation, and that American schools do, after all, produce a work force with adequate skills to use new technologies.

Occupational Outlook

Of course, schools must prepare young people for millions of new jobs. But while the number of technologically sophisticated jobs will grow, the largest portion of employment opportunities will be in services—and not very high-tech services at that.

In the next decade, about five million new jobs will be created for food workers, including kitchen help, waiters and waitresses; four million will be for cashiers and retail salespeople; more than three million will be for clerks; and two million will be for helpers, packagers and laborers. Jobs for truck drivers will abound.

Managerial and professional occupations will also need more workers, but their numbers pale compared with openings requiring less education. Employers will hire more than three times as many cashiers as engineers. They will need more than twice as many food-counter workers, waiters and waitresses than all the systems analysts, computer engineers and database administrators combined.

How did we convince ourselves that schools must prepare an entire generation of young people who know calculus? Partly, because we have confused occupational growth rates with the number of new jobs. Computer engineering and science employment will increase by a whopping 100 percent, while food service will grow by only 11 percent. But computer science is a relatively small field, so a few new positions generate rapid growth rates. There are more waitresses today, so a smaller percentage growth yields more new jobs.

The fact that more employees use computers at work is also misleading. Rather than requiring increased skills, computers often allow reduced skill levels. Many workers may need less education, not more, when jobs are computerized. Consider the supermarket checkout clerk who uses the computer to scan prices and determine change.

Naturally we focus on what is new, overlooking what is unchanged. Amazon.com may market in cyberspace, but it creates more jobs for warehousemen than for Web site designers. It takes no more education to drive a forklift for Amazon than for Kmart.

Many workers, including college graduates, are now overeducated for the tasks they perform. Of course, we do face some shortages in skills. And when we do have shortages, observers are often too quick to conclude that schools are to blame. But labor markets aren't that simple; labor shortages often exist at the same time that workforce skills are adequate.

A Labor Crisis?

Avoiding shortages of workers in specific high-skill occupations requires not only an adequately educated workforce, but also a flow of good information to young people who are choosing careers. This information includes the "signals" given by relative salary levels.

If, for example, large numbers of high school graduates with strong mathematics and science skills, seduced by reports of high salaries for lawyers, choose to enroll in pre-law rather than in engineering, the economy could suffer a shortage of engineers and a surplus of lawyers, even if schools are adequately preparing graduates.

Consider the "labor crisis" alleged by high-tech computer and software producers during the past few years. Manufacturers, claiming there were not enough trained American software engineers, urged the United States to permit the import of more foreign workers. As a result, last year President Clinton and Congress agreed to increase the quota on temporary visas, called H-1Bs, for skilled immigrants, raising the yearly cap for the next three years to 200,000 H-1B visas for foreign workers—up from only 65,000 four years ago.

Lobbying for this change, Microsoft Chairman William H. Gates and Intel Chairman Andrew S. Grove told the congressional Joint Economic Committee that while more H-1B visas were needed, this was only a stopgap until American students' skills improve.

But the labor crisis they alleged was probably a mirage, as are the school failures in mathematics and science that the alleged crisis suggests. High-tech firms portray a shortage of qualified workers, but it is only our memories that are short. Few remember that only a few years ago, we had a glut of science and mathematics graduates in the United States.

As recently as 1992, one in five college graduates held jobs that did not require a college degree. This was no secret: the press was filled with reports of underemployed science and mathematics majors. A 1995 New York Times article, "Supply Exceeds Demand for Ph.D.s in Many Science Fields," cited widespread unemployment of engineers, mathematicians and scientists.

According to Michael S. Teitelbaum, a Sloan Foundation demographer, the United States had "too many engineers [and] too few jobs" for them in 1996. There was, he stated then, "an employer's market," especially in the wake of post-Cold War downsizing in defense and aerospace.

In fields with real labor shortages, wages rise. In fields experiencing a labor surplus, wages fall. Despite accounts of dot-com entrepreneurs becoming millionaires, computer technology wage trends have not confirmed a shortage and don't support the need to import legions of H-1B programmers.

Lawrence Mishel of the Economic Policy Institute calculated that wage offers to new college graduates in computer science averaged $39,000 in 1986, declining to $33,000 in 1994 (in constant dollars). It is only since the mid-1990s that offers to computer science majors have climbed.

Surveys by DataMasters, a computer industry consultant firm, show that the median salary for experienced software engineers on the West Coast was $71,100 in 1999, up only 10 percent (in constant dollars) from 1990. Wage growth of about 1 percent a year suggests no labor shortage.

Norman Matloff, a University of California computer science professor, has claimed that high-tech firms created artificial shortages by refusing to hire experienced unemployed programmers. Fewer than half are still working in the industry at age 50. Luring the rest back requires higher wages.

Industry spokespersons say that older programmers have outdated skills and take too long to retrain in a fast-paced competitive environment. Matloff counters that when they urge more H-1B visas, lobbyists point to vacancies lasting many months as evidence of a shortage. Experienced programmers can be trained in the latest computer language more quickly than visas can be processed for cheaper foreign workers.

Student Choice

Whether or not large numbers of older, experienced software engineers are available to hire, computer firms have not offered consistently higher wages to attract more applicants, young or old. A long-term trend of higher wages is what, in a market economy, we would expect to confirm a shortage of workers with adequate skills for an occupation.

Perhaps a shortage of computer engineers is a more recent phenomenon, not yet reflected in long-term wage trends. But if this is the case, the shortage cannot provide evidence of schools' shortcomings. And importing cheaper foreign labor could be counterproductive.

Admitting more H-1B immigrants while addressing these shortages can have a contrary effect--as pay declines with the entry of immigrant engineers willing to accept lower wages, this lower pay will signal young people to avoid this field in the future. As a result, the shortage will only be prolonged.

Labor markets are much more flexible than schools. If the demand for workers with particular skills suddenly grows, and schools adjust their vocational counseling and course offerings to meet that demand, students qualified to fill the new jobs will not be available until they are able to complete their coursework and graduate several years later. Likewise, if students see a surplus of labor in highly specialized fields, fewer will take the courses to prepare for them.

The labor market for young computer technology workers has tightened recently and wages have started to rise again because earlier, students saw wage declines and stopped choosing math and science as college majors. This created the shortage. In 1985, American colleges awarded 42,000 bachelor's degrees in computer science. By 1996 this fell to 25,000.

The number of computer science degrees did not decrease because high school students' math and science skills declined and they were no longer prepared for such majors. On the contrary, more students are enrolling in high school math and science courses, including advanced placement, and math scores are rising.

Fewer qualified students are choosing to major in computer science for economic, not academic reasons. Last year, 24 percent of the high school seniors who took the SAT scored over 600 in math. However, only 6 percent of all seniors who took the SAT planned to major in computer science in college.

Schools cannot solve short-term supply and demand imbalances for particular professions because of the lag between the time a student chooses a major and the time he or she begins the actual job quest. Even if schools are producing students with adequate skills, specific occupational shortages can be erased only by raising wages to attract those with needed skills who are working in other occupations. For the longer term, increasing wages can prompt counselors to encourage well-prepared math and science students to major in computer science and engineering.

We shouldn't confuse shortages in specific fields like computer science and engineering with an overall lack of academic qualifications. The Labor Department projects an increase of less than 1 percent in the overall share of workers in occupations requiring a college degree from 1996 to 2006.

Unfulfilled Hopes

Our false expectation that all children who get higher math scores will land better jobs encourages us to duck critical decisions about how schools relate to society.

We want schools to equip all students from all ethnic and socioeconomic groups to compete equally for better jobs and to get better jobs. And we expect schools to prepare them.

But this hope can't be fulfilled. We already enroll enough college students to fill foreseeable vacancies in professional fields. Increasing the number of applicants does little to increase the number of vacancies. Increasing the number of students prepared for highly paid technical occupations will, inevitably, cause the pay for those occupations to decline. This is how a market economy should function.

So we face a challenge: If academic achievement improves for all students, what will determine which children win and which lose the race for higher incomes and status in American society?

Imagine that schools further narrow achievement gaps between advantaged and disadvantaged children. If highly paid, highly skilled jobs are limited, and more children from poor and minority families win them, where does that leave children from wealthier families? If more poor children are qualified to become computer scientists, more rich children will have to settle for being waiters.

This seems decent and fair, but we have a lot of introspection, debate and economic reform ahead before all children can have equal opportunity. We can't face these critical issues of 21st century education with minds clouded by a "high-tech jobs of the future" fog.

We seriously mislead our young people if we convince them that if they get a better education, they will get better jobs. It can work for some, but it can't work for all. Education alone cannot solve problems of living standards and income distribution. These are fundamentally economic and political problems and they require economic and political solutions beyond school improvement.

Richard Rothstein is a research associate of the Economic Policy Institute in Washington, D.C., and a weekly columnist on education policy for The New York Times. E-mail: rothstei@oxy.edu. Portions of this article appeared previously in his columns.