Chapter 3: Science & Engineering Indicators 93

Forecasting the S&E Job Market

Forecasting supply and demand for scientists and engineers is an extremely difficult (see Vetter 1992a and 1993 and Fechter 1990), and rarely accurate (see Leslie and Oaxaca 1990), undertaking. For example, how could anyone have predicted the end of the Cold War and its aftermath? Although the end of the Cold War has not caused a major disruption in U.S. labor markets for scientists and engineers, some turmoil is being generated by newly jobless engineers (many of whom have spent their entire careers in the U.S. defense industry) and by scientists exiting the former Soviet Union.

BLS analysts have conducted several studies of the future job market for scientists, engineers, and technicians. Findings from these studies yielded the following conclusions:

Employment in technical occupations will grow at a faster pace than overall employment.
Employment in technology-intensive industries will grow at about the same rate as employment in general.
Surpluses are more likely to be observed in the S& E job market than shortages, but the latter (especially in specific fields) cannot be ruled out.

Every 2 years, BLS analysts prepare employment projections by occupation and by industry for the entire economy. The most recent forecast was prepared in 1991 and covered the period 1990-2005. Data derived for technical occupations are presented in text table 3-2. They show wide variations in employment growth under the three alternative scenarios--which prescribe high, moderate, or low growth for the economy--BLS uses for projecting future employment. (Assumptions concerning the impact of defense downsizing on employment were included in these scenarios.) For all technical occupations, growth over the 1990-2005 period is projected to range from 9 percent (using the low-growth scenario) to 59 percent (in a high-growth economy). The moderate-growth alternative yields a 35-percent increase, a much higher gain than the 20-percent increase in employment projected for the economy as a whole (Silvestri and Lukasiewicz 1991).

Among individual scientific and technical occupations, projections for engineering employment show the widest variation: from a 2-percent decline (using low-growth assumptions) to a 54-percent increase (under the high-growth scenario). Engineering employment is more sensitive to changes in the economy and the defense budget than employment in the other technical occupations. Under each of the three alternatives, computer, mathematical, and operations research analysts are expected to have the highest growth rates, ranging from a 46- to a 97-percent increase. Employment of social scientists shows the least variation in growth--up or down 7 percent--depending on the state of the economy.

As part of this ongoing effort, the National Science Foundation (NSF) sponsored a special BLS study of employment growth in approximately 50 industries that employ the highest concentrations of technical personnel and all levels of government (Braddock 1992). (Click here for footnote 14.)Once again, projections were based on three alternative scenarios. In addition to the economic and other assumptions used in the original BLS model, additional variables that affect employment in high-tech industries--e.g., the propensity for the Nation to spend money on R& D and the export of products with a high technology content--were incorporated in this model.

The results of the analysis show employment in technology-intensive industries increasing about 20 percent (in the mid-range scenario) between 1990 and 2005: This is about the same as the employment growth rate forecast for the economy as a whole. This finding is counter to projections made a few years earlier (BLS 1990) that showed employment in high-tech industries increasing at a faster pace than overall U.S. employment. The change is largely attributable to the turnaround in defense spending that occurred in the late 1980s. The curtailment of military-related expenditures is lowering the rate of employment growth in technology-intensive industries, bringing the rate of increase down to the level expected for all U.S. employment.

These two BLS studies present an interesting anomaly: Although employment in technical occupations is expected to increase faster than overall employment, employment in technology-intensive industries is not expected to increase any faster than employment in nontechnology-intensive industries. One explanation is that in the less technology-intensive industries--e.g., those in the service sector--the proportion of the workforce comprised of scientists and engineers is increasing faster than employment in general.

The BLS analysis of the job market for technical workers was carried another step farther in an attempt to determine whether the supply of new S& E graduates would be sufficient to fill the new jobs created in the near future (1990-2005) and to replace workers who retire or leave S& E jobs. Once again, three estimates of the supply of new S& E graduates were prepared; they were calculated using (high, moderate, and low) percentages of the college-age population expected to earn bachelors degrees in science and engineering.(Click here for footnote 15.)

Next, the number of new S& E graduates derived under each of the three supply scenarios was compared with the number of job openings derived from each of the three employment growth scenarios for all technical occupations. Matching the three supply with the three demand estimates yielded nine possible depictions of the future job market for technical workers. Each of these nine alternatives was then compared with a benchmark determined by BLS staff to be the ratio of technical degrees awarded annually to the number of technical job openings during a time (1984-90) when the supply of new S& E graduates was thought to be equal to the demand for them. In the late 1980s, the ratio of technical degrees awarded annually to the number of technical job openings was about 1.6. That is, of every 16 S& E graduates, 10 took S& E jobs; the other 6 either went into non-S& E occupations or left the country.

Using this 1.6 ratio as the benchmark, it was determined that most of the nine supply-demand possibilities yielded ratios equal to or higher than 1.6--that is, the supply of S& E graduates was greater than the demand. Only in three of the alternatives--those in which high-growth estimates were coupled with low-growth estimates of technical degree production--would there be situations in which shortages might exist. The results of this modeling exercise indicate that although there may be future shortages of technical workers in some fields, overall, there are more likely to be surpluses in the coming decade and beyond.(Click here for footnote 16.)

Footnote 14:

Braddock's definition of high tech industries differs from the Organisation for Economic Co-operation and Development definition used in chapter 6.

Footnote 15:

This method of estimating the supply of new scientists and engineers has several deficiencies cited by Braddock (1992), the most important of which is the omission of other sources of supply, i.e., (1) individuals switching to S& E jobs from other occupations and (2) immigrants.

Footnote 16:

In a response to the BLS findings, Finn and Baker (1993) show that there are likely to be more shortage situations than predicted using BLS' model. They reach this conclusion by showing that the BLS estimates of degree production are overly optimistic.

ContentsSearchContinue