Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
ANTICIPATING THE FUTURE The findings from earlier chapters of this report indicate that the resiliency of the engineering labor market has enabled it to adjust to the post-WorId War IT shocks of external forces with a minimum of disruption. These shocks include wide fluctuations in government programs, in particular the increase in defense expenditures that occurred in the 198Os. Although the market survived these fluctuations, the question remains as to whether it could accommodate further growth, regardless of its origin in the defense or the nondefense sectors of our society. This suggests the larger, and perhaps a more basic, question: Can we determine the point beyond which further growth strains the system? Our ability to address this issue depends in part on the adequacy of existing methods for predicting the Impact on supply and demand of fluctuations in government programs. This section of the report briefly renews the methods by which demand and supply are anticipated and evaluates their reliability. Findings with respect to these methods draw heavily from the papers commissioned for this study. Although demand and supply interact arid influence each other, they are Heated separately for convenience of discussion. Anticipating Demand In a paper entitled "What Can Demand and Manpower Requirements Tell Us About the Impact of Defense Spending on the Labor Market for Scientists and Engineers?" (see Appendix A) Lee Hansen identified a number of survey and analytic approaches for estimating the future need for scientific and technical personnel. He also identified an ad hoc method for est~madng demand associated with a specific government program. Surveys Questioning industrial officials about the changes they expect in labor demand is a technique long used by personnel departments. For example, state employment services used a survey to try to anticipate future employment changes in their areas. The method was abandoned more than 10 years ago as too inaccurate and too costly. A more sophisticated version has been developed more recently by the American Electronics Association and used for a 5-year forecast (National Research Council, Office of Scientific and Engineenng Personnel, 1984, pages ~ I-251. If it were possible to tap the information and judgment of knowledgeable industrial leaders, we might improve estimates of future changes. It is not as clear, however, that the summation of opinions obtained In a survey gives accurate results. This accuracy could be evaluated retrospectively by comparing forecasts with actual events. The evidence with respect to such evaluations is extremely Aide. The Pane! is not able, therefore, to judge the adequacy of this method of anticipating demand and is skeptical of the value of such surveys In anticipating demand more Han one or two years into the future. 13
analyac Methods Other, more analytic, methods estimate demand for workers in any sector of the economy by linking them to the rate at which the goods or services produced in that sector will be expanding--not necessarily a proportionate relationship. Some analytic methods are "m~croecono~ruc," focusing on linkages to a single sector or a few sectors of the economy. Others are "macroeconomic," seeking to anticipate how requirements for each of many different occupations in many different parts of the economy rely on estimates of the growth of the entire economy. These models are limited in their ability to anticipate changes in engineering requirements by (~) the difficulties involved in accurately forecasting future changes In the levels and composition of the nation's output of goods =d services and (2) methodological problems inherent in the estimation of the parameters used by these models to link requirements to output. , , Ad hoc Methods Hansen also reviewed an ad hoc approach developed recently to predict demand for scientific and technical personnel in a component of the defense program--the Innovative Science and Technology Office, a small part of the Strategic Defense Initiative Organization (Sterling Hobe Co~porabon, 19861. To estimate personnel requirements for this pioneering R&D program, the analysts calculated the ratios of personnel to dollars spent by organizations engaged in each of the types of research involved in the program. These ratios were used to estimate tote] personnel requirements arising from budgeted growth; ratios of professional to total personnel were then used to estimate professional personnel requirements. This ad hoc approach was well adapted to estimating broad manpower needs for a highly specific R&D program. This approach, however, does not give the entire picture. It neither evaluates the effect of the program on supply of scientific and technical personnel throughout the economy nor determines its feasibility In terms of manpower. Anticipating Supply The supply side of the labor market for an occupation like engineering may be characterized by flows of individuals who enter or leave through various channels. Included in this characterization are students graduating from engineering schools, expenenced workers moving between engineering and nonengineering occupations in the United States, and international flows of workers migrating between foreign countries and the United States. As individuals enter or leave the engineering labor market through venous channels, their responses to the market are affected by a number of social and economic variables. Analysts of the engineering labor market try to take these flows into account in predicting future labor supply. In an attempt to develop a method that systematically relates these flows to labor market conditions, Robert DauffenBach and Jack Fiorito developed a labor supply mode! that starts with the stock of science and engineering personnel in one year and projects the stock for the following year, using past data on the relationship of each flow (student choices of careers, immigration, occupation mobility) to various indicators of the labor market situation (DauffenBach and Fionto, 19831. The indicators used include (1) the level of employment in the occupation and the projected rate of employment change as variables affecting student course choices and (2) the estimated gap between projected demand and projected supply of graduates as a variable affecting the inflow from over occupations. 14
This mode! was evaluated by Michael McPherson, who calls it "the most comprehensive and analytically challenging among recent projection models of the supply side of technical labor markets" (see Appendix A). He notes that, like most other work in the field of labor supply, it is quantitative only, dealing with numbers of people but not with their quality or productivity--a subject difficult to evaluate or measure. He also notes that while supply is made responsive to demand, in order to make the mode} manageable, the reverse feedback--making demand responsive to supply--is not dealt with. McPherson credits the DauffenBach-Fionto mode! with comprehensively evaluating the supply system for scientific and technical personnel as a whole. He notes, however, that its very comnrehenc.ivenecc. limits the m~ulP.1 to H~tn that Ron he ~cc~n~hl-~1 wraith -A ~ rat ~- ~^ of_ ~~_---v-~ ·, ~~1 · ~ ^. , . . ., ~ ~ . . . . ~ . . . _ consistency across many rlelos and thus excludes the institutional peculiarities of indi- vidual fields--for example, the availability of NTH traineeships and postdoctoral fellowships as a variable affecting the supply in the life sciences. McPherson also points out the desirability of incorporating more specific adjustment processes in the model--for example, accounting for the various channels through which information on labor market conditions reaches the people who are presumably affected. The focus of research should not be whether supply and demand will be brought into balance but instead how and with what costs. The DauffenBach-Fionto effort points to valuable directions for future inquiry to get better information on the dynamics of interfield movements and on the decision processes of colleges and universities affecting the production of graduates. A way of examining qualitative dimensions of this production may be to conduct follow-up surveys of the work experience of graduates, analyzing the results according to school records of grades or honors. An Evaluation - Because of the time and resource constraints imposed on this study, the Pane! did not attempt to experiment with any of these methods of anticipating supply and demand. A study undertaken for the National Science Foundation in the early 198Os, however, found that the market would accommodate a real increase in defense expenditures from 1982 to 1987 Hat ranged from 3 to 8 percent per year, but that this could result in serious stress in the markets for computer specialists and aeronautical and electrical engineers.4 However, the Pane! could not forecast with a tolerable degree of certainty the labor market implications of any further growth in defense expenditures. This conclusion was based on (~) the limitations in the Panel's ability to anticipate accurately future changes in the level . . . ~ . , ~ ~ . . . . . . . and composition or outputs, tz' the snortcommgs in estimates or the linkages between eng~neenng demand and supply, and (3) the current absence of a suitable data base against which estimates of anticipated engineering supply can be evaluated. The capability to illuminate this type of question needs to be developed. Further Issues Although the Panel's general conclusion is that the market is adaptable, it is not perfectly so. Engineering labor markets are characterized by their cyclical nature, 4The National Science Foundation (1984a) found that indicators of labor market conditions (i.e., anticipated differences between supply and requirements) suggesting possible shortages for computer specialists and aeronautical engineers were relatively insensitive to variations in the rate of change in defense expenditures or Gross National Product. Comparable indicators for electrical engineers were sensitive to changes in these variables. 1 5
suggesting the existence of overreactions to any given perturbation. Moreover, some changes can be so massive or sudden that either the entire engineering labor market or the markets for some engineering subspecialties will adapt more slowly or imperfectly, sometimes taking several years to regain a sensible steady state. The results of such imperfections can produce significant costs. Sudden and dramatic increases can result in upward pressure on engineering wages or significant deterioration in engineering productivity. Either of these can inflate cost. Moreover, given the lags that exist in the system, some adjustments may occur at the wrong phase of the cycle. Large unexpected decreases can have significant adverse effects upon individuals caught up in these changes. Occupational mobility and substitution between engineers and noneng~neers reflect the willingness of both employers and employees to modify their standards in the face of changes in labor market conditions. The evidence indicates that such movement and substitution is substantial (see Appendix Tables 10 and ~ I, pages 77-781. Roughly one- s~xth of the 1984 engineering work force had degrees in fields other than engineering, suggesting a nontrivial amount of inflow. The proportion of computer specialists with degrees in other fields (roughly four-fifths) indicates substantial amounts of inflow. Similarly, more than one-th~rd of those with engineering degrees reported that they were employed in nonengineenng occupations, suggesting a substantial amount of outflow. The efficiency with which this mechanism operates, however, can be further influenced through retraining or other means of increasing the fungibility of the existing engineering work force. New technologies and changes in institutional structures affecting engineering utilization are stimulating the need for such adaptability of the work force. Such adaptation win facilitate exploitation of these technological and institutional changes. Engineenng degree production responded quite well to past cyclical swings in engineering demand. These responses, however, occur with a considerable time lag, creating the potential for situations in which policies influencing this adjustment mechanism could contribute to further instability, rather than acting as a stabilizing force. The efficiency with which this mechanism operates can be influenced through financial aid mechanisms (in the form of fellowships, scholarships, research or teaching assistantships, and postdoctoral appointments) that affect the extent to which students are encouraged or discouraged from pursuing careers in engineering. Furthermore, career choices of high school students are probably heavily influenced by the media. Those choices made on the basis of media presentations of the market for engineers at one point in time may be inappropriate for the market at the time Hey complete their education. Varying the extent to which these mechanisms are used can help damp the swings in production of undergraduate engineering degrees and influence the decision of degree recipients to continue their engineering education at the graduate level. In addition to mechanisms designed to influence degree production, measures aimed at more effective use of scarce engineering resources can also be considered. This report notes Me increasing tendency on Be part of employers to employ computer-onented technologies such as CAD/CAM and "expert systems" to reduce the size and increase the efficiency of their technical work forces. The dominant motive for these efforts appears to stem from the desire of Cons to improve them competitive positions In world markets. If so, these efforts can result in a future slowdown in the growth of industrial employment of engineers. Some of the adjustments required to accommodate changes in supply arid demand arise from the nature of preemployment engineering education. One way of ensuring a highly fungible engineering work force capable of responding to fluctuations in supply and demand and to rapid technological change is to offer a broad engineering curriculum with many core engineering courses shared by students in all disciplines. The Committee on Me _ _ _ · 16
Education and Utilization of He Engineer recommended such an approach to undergraduate engineering training. It noted that ". . . increased specialization of eneineenne curricula. rn,~nl-A with H-rr~-A interact _~ ~~ as. __._ ^^w^_v.. . . in degrees in basic sciences and mathematics, win lead to future difficulties in our ability to respond quickly to new technological challenges" (National Research Council, CEUE, 1985a, page 681. It further observed, "Extensive, in- depth disciplinary specialization does not belong in the undergraduate curriculum and should be postponed to the graduate level" (CEUE, 19SSa, pages 68-691. Such a shift in the structure of engineering education not only would introduce the possibility of more fungibility among sub-specialties inside and outside of engineering, but also would provide new possibilities of policy intervention to alleviate temporary conditions of over- or undersupply. It is probable that the concentration of a significant fraction of the professional education at the graduate level for physicists, chemists, biologists, etc., makes these occupations less susceptible to intense swings in supply/demand relationships characteristic of engineering today. We are concerned about the state of our understanding of quality factors in engineering supply and demand in both military and commercial markets. The study was unable to find adequate information to enlighten us on the quality dimensions of engineering labor markets--i.e., dimensions describing (a) the productivity and performance of engineers and (b) the characteristics of the engineer and the work en- vironment that contribute to high productivity and excellent performance. Nevertheless, these quality aspects of engineering labor markets are particularly important because many of the adjustment mechanisms discussed earlier in this chapter can dramatically alter these properties. In addition, although there appears to be little cause for concern about engineering bottlenecks arising from Me current buildup, it would have been difficult, given the state of the art in modeling engineering labor markets, to assess the labor market implications of further dramatic increases in real military expenditures should they continue. Past buildups were accompanied by programs designed to increase the supply of engineers. These programs are much smaller now and, with undergraduate engineering enrollments beginning to decline and the shortage of engineering faculty continuing, expansion of these programs to accommodate further growth in defense or nondefense activity may -not be as feasible as it once was. The market has been able to accommodate past changes in engineering demand and supply through the wide range of adjustment mechanisms that exist within its institutional structure. Our ability to assess potential future problems will depend strongly on the development of a suitable knowledge base about these mechanisms. 17
