Analysts Debunk Idea Of Scientist Shortage, Citing Defects In Current Economic Models

Peter House is chief analyst and director of the Division of Policy Research and Analysis at the National Science Foundation. In one corner of his 12th-floor office stands a bookcase filled with 30 years' work on public policy analysis and technical assessment. So it's somewhat surprising to the 54-year-old House that a single, 34-page unpublished paper written in 1989 and revised several times since then has attracted so much attention and made him the center of a sharp debate over whether the United States faces a shortage of scientists. "We began with the fact that the population of college-age people has been declining since the mid-1980s, and will continue to drop until the middle of this decade," he explains. "Then we added in the fraction of Americans earning bachelor's degrees in the natural sciences and engineering [NS&E], a percentage that has been very stable over the past three decades." From those data, House projected the number of NS&E bachelor's degrees that would be awarded over the next 20 years. Then he drew a horizontal line from the 1984-86 period, when the number of degrees was at its peak. The gap that was formed between that line and his projections, he calculated, added up to 675,000 by the year 2006. House says that he never intended the number to represent a shortage, much less to be used as a rallying cry for the scientific community. "It was an analysis, that's all," House says. "A shortage depends on demand, and we don't know enough to do any forecasting of demand. We said to Erich [Bloch, then NSF director], `You decide if it's significant or not.' If someone thought it was going to be a problem, then I suppose they might use the term `shortage.' If they didn't think it was anything to worry about, then they wouldn't [use the word]." House, who holds a master's degree in economics and a Ph.D. in public administration, worked for two other federal agencies before coming to NSF. He's also been president of his own consulting firm, and for the past eight years he's run the policy research and analysis division at NSF, which operates within the science, technological, and international affairs directorate. His division can be compared to an in-house think tank for the foundation, he says, making use of information from NSF's own science resources studies division as well as from the Bureau of Labor Statistics, the Census Bureau, the Department of Education, and other sources. "At the Energy Department and at EPA [Environmental Protection Agency]," he says, "we used to prepare papers that listed various policy options for the agency. Here, it's more important to do analyses of the huge amounts of data that have already been collected." Either way, he says, the goal is to provide his bosses with useful information that can lead to enlightened public policy. "If you can bring an issue to the attention of policymakers, that's very satisfying," he says. "And I guess we've certainly done that in this case." --J.M. In fact, many researchers avoid the words "shortage" or "surplus" entirely in their analyses of future job markets for scientists. "We hold that forecasting and projecting supply and demand for scientists and engineers is largely a wasted effort," report University of Arizona researchers Larry Leslie and Ronald Oaxaca in a recent study funded by the National Science Foundation. Adds Oaxaca, an economist, "Even if you knew all the factors that affect supply and demand, which we don't, there is a randomness about the data that you can't eliminate. And it gets worse the farther out you go." The problem, says Oaxaca, isn't insufficient computing power, or even insufficient intellect. "It's not the fault of the modelers," he notes. "It's the nature of the beast." Until recently, such talk would have bordered on heresy within a government that prides itself on being able to forecast the nature of labor markets and feed the information to a hungry horde of users. Just last week, for instance, NSF officials began sifting through grant applications for a new program that will fund a handful of studies to improve the accuracy of economic models of science and engineering supply and demand. Although the grants will be awarded by the foundation's division of social and economic sciences, the money comes from NSF's education and human resources directorate as part of its burgeoning effort to increase the flow of talent into the sciences. Talk of impending shortages has been in the wind for several years. In 1989 it crystallized around a paper by NSF policy analyst Peter House, entitled "Future Scarcities of Scientists and Engineers: Problems and Solutions." The paper (see Close-Up on page 4) rests its argument on two demographic factors expected to affect supply. One is the shrinking pool of college-age students; the other factor is a declining percentage of students who pursue degrees in the natural sciences and engineering. Those two trends, House predicts, could result in a cumulative shortfall over the next 20 years of 675,000 bachelor's degrees from current graduation levels. House never actually uses the word "shortage" to describe what lies ahead. And the number he cites refers only to undergraduate degrees, not jobs. Furthermore, he warns that any steps to increase participation at the undergraduate level must be taken quickly--or not at all. "If initiated too late," he writes, "[such efforts] may lead to an unnecessary oversupply of people with scientific training." Even so, the idea of a looming, potentially catastrophic shortage of scientists quickly embedded itself in top-level science policy circles. For example, Richard Atkinson, chancellor of the University of California, San Diego, gave his presidential address at the 1990 meeting of the American Association for the Advancement of Science (AAAS) on the subject of "Supply and Demand for Scientists and Engineers: A National Crisis in the Making." In the same issue of the journal Science that carried his remarks was a news story asking, "Who Will Do Science in the 1990s?" (248:433, 1990). At times, the idea of a shortage takes an even more precise form, that of a projected undersupply of doctoral-level researchers. For instance, a position paper issued last month by the Council on Research and Technology (CORETECH), which lobbies for more government spending on research, calls for 3,000 new federally funded graduate fellowships and traineeships. The program is needed, according to a CORETECH background paper, because "the U.S. faces a potentially serious shortage of Ph.D. scientists and engineers." Many experts in the field don't agree. They tick off several reasons why such a labor crisis is unlikely to occur. Chief among them are the tendency of employers to take whatever steps are needed to find scarce talent, the capacity for universities to modify faculty-student ratios to cope with temporary imbalances, the large pool of trained scientific talent not currently working in their chosen field, and the ability of students to alter their career goals to match changing job markets. "There are at least 20 crucial factors that I would put between a demographic analysis of supply and likely outcomes that would be equally valid," says labor economist Robert Dauffenbach, director of the Center for Economic Management Research at the University of Oklahoma. "A strong case can be made that the demographics of 18-year-olds have little impact on supply and demand in the scientific marketplace." "The notion of large shortages is impractical because of the current large pool of scientific and technical manpower in this country that can be tapped," says Michael Teitelbaum, a demographer with the Alfred P. Sloan Foundation in New York. Referring to House's data on student degree choices, he adds, "I've always been skeptical of analyses based on assumptions of the activities of a fixed percentage of a particular cohort. In addition, labor economists are wary of models in which the supply side is not affected by demand. The real world just doesn't work that way." Dauffenbach, born in 1946, cites the experiences of his own student cohort in suggesting how quickly the system can respond to perceived shortages. "Back in the 1960s universities needed bodies in the classroom," he recalls. "Salaries, tenure, and job opportunities were bright. So they squeezed a lot of Ph.D.'s from a relatively small pool, at a time when birth rates were still low." The problem of predicting shortfalls goes even deeper than looking at student behavior, says Jesse Ausubel of the Carnegie Commission on Science, Technology, and Government in New York. "It's not clear to me how meaningful it is to speak of shortages in general," he says. "It may be true, for example, that at a particular point in time there is a shortage of optical engineers in Chicago. But historically, none of the projected shortages of scientists has ever occurred." A lot of labor statisticians are also troubled by a bit of recent history: the January 1991 report issued by AAAS president-elect Leon Lederman that asked 250 academic researchers to describe their funding opportunities. The situation is so bad, Lederman reported, that some senior scientists are thinking of abandoning science, and discouraging their students from embarking on what they see as an increasingly risky profession. Within the human resources community, those comments were read as a sign of too many scientists, not too few. "The Lederman report didn't do anybody a service," says Betty Vetter, executive director of the Commission on Professionals in Science and Technology in Washington, D.C. "By warning that students might choose to abandon their scientific careers and do something where their financial prospects are better, the report seemed to be arguing that we have a surplus of people." Vetter and others admit they are troubled by the declining interest among college students in science and engineering. And they think the U.S. should spend more money on research, both to train the next generation of scientists and to support those already engaged in that work. But they are equally quick to point out that such an attitude is not the same thing as predicting a shortage. "All Peter [House] is saying is that we have a smaller segment of the student population choosing to enter science, and that we need to do something if we want to change that situation," says Vetter. "That's the fact, but is that bad? I don't know." Fewer American college students are studying the natural sciences and engineering, according to 1991 data collected by the Commission of Professionals in Science and Technology (CPST). For the past five years, the number of undergraduates getting diplomas in mathematics, the physical sciences, computer and life sciences, and engineering has been dropping, although the total of students earning bachelors degrees overall in the U.S. has been rising steadily (from 991,477 in 1985 to 1,015,239 in 1989). For minority students, however, the findings are different. "While the proportion of students getting their science degrees has decreased from 21 percent in 1985 to less than 18 percent in 1989, different races show different patterns of science degree attainment," says Betty Vetter, statistician for the commission. Asian Americans have increased their percentage of science degrees earned during this period. Whites and women, on the other hand, have decreased their representation in sciences, while blacks and Hispanics have maintained their levels, although not at the proportion in which they are represented in the undergraduate population. Asian Americans earned 8 percent of the natural science and engineering (NS&E) degrees in 1989, up from 5 percent in 1985 and above their total undergraduate percentage of 4 percent. Blacks earned 5 percent of the NS&E degrees in 1989, up from 4 percent in 1985, but they make up 9 percent of the undergraduate population. Hispanics are 6 percent of the undergraduate group, but their NS&E degrees have stayed at 2 to 3 percent for the past five years. Whites earned 10 percent fewer NS&E degrees in 1989 than in 1985. Of the 25,123 drop in the number of NS&E degrees awarded during that period, 19,363 fewer whites got their degrees in science. A total of 178,941 undergraduates earned NS&E degrees in 1989, compared with 204,064 in 1985. While women earn more than half of all bachelor's degrees, the percentage of women with NS&E degrees fell to 25 percent in 1989, from 28 percent in 1985. Black women, however, fared better, says Vetter. In 1989, 2.3 percent of all science degrees went to black women undergraduates, up 0.5 percent from 1985. These and other findings are in the 1991 edition of Professional Women and Minorities: A Manpower Data Resource Service, published by CPST, 1500 Massachusetts Ave., Suite 831, Washington, D.C. 20005; (202) 223-6995. Price: $100 for nonmembers; $75 for members. --Robin Eisner

Analysts Debunk Idea Of Scientist Shortage, Citing Defects In Current Economic Models

Interested in reading more?

Become a Member of