CHAPTER 1

BACKGROUND

Issues in the Prediction of Demand and Supply for Doctoral Scientists and Engineers

Interest in predicting demand and supply for doctoral scientists and engineers began in the 1950s. Arrow and Capron (1959) found the academic labor market interesting, in part because it was characterized by lengthy training periods for Ph.D.s. This meant that the market students saw when they began their graduate training was often very different from the market they encountered when they finished. These sorts of lags led to alternating spells of oversupply and undersupply (often called cobweb adjustment) in some fields. Freeman (1971) and Cain et al. (1973) carried out empirical explorations of these models in the 1970s. At the same time, Cartter (1970, 1976) made demographic models of the demand for faculty and forecast an end to the booming market for Ph.D.s in the late 1960s and early 1970s, which, in part, had been driven by the growing demand for faculty to teach the baby boom generation of students.

In addition to the surging demographic demand for doctoral scientists and engineers, the late 1950s through the late 1960s saw a tremendous increase in federal funding for research in the academic, industrial, and government sectors, which added additional fuel to demand and a growing recognition that this sector of the work force is critical in meeting national needs. The same period saw steep increase of graduate fellowships and traineeships supported by NSF and federal agencies. In particular, the National Aeronautics and Space Administration (NASA), the Department of Defense (DOD), and the National Institutes of Health (NIH) funded predoctoral and postdoctoral fellowships and training grants in the biomedical and behavioral sciences. As a result, the number



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 9
FORECASTING DEMAND AND SUPPLY OF DOCTORAL SCIENTISTS AND ENGINEERS: Report of a Workshop on Methodology CHAPTER 1 BACKGROUND Issues in the Prediction of Demand and Supply for Doctoral Scientists and Engineers Interest in predicting demand and supply for doctoral scientists and engineers began in the 1950s. Arrow and Capron (1959) found the academic labor market interesting, in part because it was characterized by lengthy training periods for Ph.D.s. This meant that the market students saw when they began their graduate training was often very different from the market they encountered when they finished. These sorts of lags led to alternating spells of oversupply and undersupply (often called cobweb adjustment) in some fields. Freeman (1971) and Cain et al. (1973) carried out empirical explorations of these models in the 1970s. At the same time, Cartter (1970, 1976) made demographic models of the demand for faculty and forecast an end to the booming market for Ph.D.s in the late 1960s and early 1970s, which, in part, had been driven by the growing demand for faculty to teach the baby boom generation of students. In addition to the surging demographic demand for doctoral scientists and engineers, the late 1950s through the late 1960s saw a tremendous increase in federal funding for research in the academic, industrial, and government sectors, which added additional fuel to demand and a growing recognition that this sector of the work force is critical in meeting national needs. The same period saw steep increase of graduate fellowships and traineeships supported by NSF and federal agencies. In particular, the National Aeronautics and Space Administration (NASA), the Department of Defense (DOD), and the National Institutes of Health (NIH) funded predoctoral and postdoctoral fellowships and training grants in the biomedical and behavioral sciences. As a result, the number

OCR for page 9
FORECASTING DEMAND AND SUPPLY OF DOCTORAL SCIENTISTS AND ENGINEERS: Report of a Workshop on Methodology of science and engineering Ph.D.s awarded annually tripled, from 6,000 in 1960 to more than 19,000 in 1971. Federal programs for doctoral training were closed down or sharply reduced in the early 1970s, when demographic trends, reductions in federal R&D funding, and unemployment among new Ph.D.s indicated an oversupply condition. However, Congress kept the supply of biomedical researchers up by passing a special law, the National Research Scientist Act (NRSA), and overseeing its implementation. Congress believed that NIH research training should be related to market demand and mandated the National Research Council to advise NIH on the nation's need for biomedical and behavioral research personnel for the next 10 years. The National Research Council (NRC) has done so in biennial and quadrennial reports since 1976 (NRC, 1994). Consequently, increases in the annual number of science and engineering Ph.D. degrees ended, and the number declined slightly after 1971 before starting to grow again in 1979. The 1972 total was not surpassed until 1986. Also during the 1980s and into the 1990s, foreign-born graduates with temporary visas accounted for almost all the growth in the annual number of Ph.D.s. Although some investigators expressed skepticism (Ehrenberg, 1991), in the late 1980s the NSF and others estimated that there might not be enough new science and engineering faculty to replace those hired to teach the children of the baby boom (NSF, 1989; Bowen and Sosa, 1989; Atkinson, 1990). To date, quite the opposite has occurred. In response to a sharp buildup in military research and development during the first half of the 1980s and widely publicized projections of a shortage of doctoral scientists and engineers, the number of Ph.D.s granted increased steadily until a year or two ago, due in part to an increase in noncitizen doctorates, as shown in Figure 1-1. By the early 1990s, a growing share of new Ph.D.s was experiencing difficulty in obtaining permanent employment upon graduation. In the biological sciences, this has led to a growing pool of individuals in postdoctoral positions (NAS, 1998). Nonacademic

OCR for page 9
FORECASTING DEMAND AND SUPPLY OF DOCTORAL SCIENTISTS AND ENGINEERS: Report of a Workshop on Methodology FIGURE 1-1 Number of natural science and engineering Ph.D. awards by citizenship and year. employment now accounts for over half of the employment of Ph.D.s in most scientific and engineering fields (NAS, 1993). Government demand for new Ph.D.s has been declining. Academic hiring has remained flat because state contributions to public colleges and universities have declined, and academia, both public and private, has come under increasing pressure to slow the growth of tuition and costs. Industry demand, which is cyclically sensitive, has been growing slightly, but many of the large industrial laboratories have been drastically downsized in the past decade. These forecasts of undersupply that did not materialize have led policymakers for graduate training and research support to be highly skeptical of any forecasts and to worry about the self-interest of the forecasters. Models that have predicted an extreme

OCR for page 9
FORECASTING DEMAND AND SUPPLY OF DOCTORAL SCIENTISTS AND ENGINEERS: Report of a Workshop on Methodology oversupply have received favorable media attention (Massy and Goldman, 1995), and such models have sharpened concern in policy circles that the federal funding of graduate education and research may simply be aggravating the oversupply. Conflicting forecasts have cast further doubt in the public mind on the usefulness of any forecasting at all. Still, the role of the nation's human resources in science and engineering remains critical. Against this background, the NSF and the Sloan Foundation asked the NRC to form a committee of model makers and users to assess the current state of the art of models that are used for forecasts, to suggest directions for improvement, and to assist policymakers in the informed use of forecasting models. Issues for the Committee Forecast error may proceed from many sources. Models may be misspecified with respect to overall structure, included variables, lag structure, and error structure. Data used for estimation may be flawed and aggregated at an inappropriate level. Further, unanticipated outside events may occur that can ruin the accuracy of even the best of forecasts. As Leslie and Oaxaca (1993) have described in their thorough review article, virtually all models of demand and supply have been flawed by at least one (and, in many cases, all) of these problems. The task of the committee was not to find fault with past efforts, but to provide guidance to NSF and to scholars in this area about to how models (and the forecasts derived from them) might be improved and what role NSF should play in their improvement. Another issue for the committee was the responsible reporting of forecasts to policymakers. Virtually no forecast is errorfree, and some uncertainty is always associated with it. Policymakers who use the results of forecasting are not usually technically proficient in the arcana of the forecaster's art. If forecasts are to be used responsibly, policymakers need to be informed

OCR for page 9
FORECASTING DEMAND AND SUPPLY OF DOCTORAL SCIENTISTS AND ENGINEERS: Report of a Workshop on Methodology about the assumptions upon which the forecasts rest, and forecasters need to track the validity of their assumptions and the accuracy of their forecasts over time. For example, if there appears to be a permanent decline in state funding of higher education, forecasters need to recast their models to reflect these changed conditions and must inform those who use their models about the implications of the change. Other audiences to whom this report is directed include graduate students in science and engineering and employers of these students—including industrial and academic employers, both federal and private funding agencies, and the scientific community itself. How best to inform policymakers who would be interested in using forecasts of demand and supply of scientists and engineers is also an issue that needs to be addressed. In order to learn from both forecast makers and forecast users about improvements that can be made in understanding the markets for doctoral scientists and engineers, the committee sponsored a workshop at NAS on March 20 and 21, 1998. Papers were commissioned on: (1) the history and problems with models of demand and supply for scientists and engineers, (2) objectives and approaches to forecasting models, (3) margins of adjustment that have been neglected in models, especially substitution and quality, (4) the presentation of uncertainty, and (5) whether forecasts of supply and demand for scientists and engineers are worthwhile, given all their shortcomings. The workshop discussion is summarized in Chapters 2-5. (The agenda and list of participants for the workshop appears in Appendixes A and B.) Following the workshop, the committee met and agreed to a set of recommendations that are presented in Chapter 6.

OCR for page 9
FORECASTING DEMAND AND SUPPLY OF DOCTORAL SCIENTISTS AND ENGINEERS: Report of a Workshop on Methodology This page in the original is blank.