occupations,3 and STEM technicians are included in one or more of the estimates. The largest estimate of approximately 8 million is found in the Bureau of Labor Statistics study, which includes the broadest set of occupations defined as STEM and imposes no education requirement.
Observation 3-1. Estimates of STEM employment in the United States vary across studies due to differences in the definitions, assumptions, and data sources utilized.
This discussion of historical trends in the overall STEM workforce relies on information available in a 2006 study released by the Commission on Professionals in Science and Technology (CPST) that examined the U.S. STEM workforce from 1950 to 2000 (CPST, 2006a). The benefit of using a single study is that it applies the same methodology over time. However, since the estimated size of the STEM workforce varies substantially across studies, the focus here is on identifying general trends in the historical STEM workforce.
The CPST study takes a broad view of the STEM workforce and includes the following occupational groups in its examination: life sciences; physical sciences; engineering; mathematics and information technology; social sciences; and science and engineering technicians. The study utilizes the decennial U.S. census to estimate the size of the STEM workforce in each decade from 1950 to 2000. These estimates are provided in Table 3-2 in the annex and illustrated in Figure 3-1. For comparison purposes, the first row of Table 3-2 provides information on the size of the overall U.S. workforce in each decade. Over the period 1950-2000, the U.S. workforce grew at an annual rate of 1.7 percent. In comparison, the overall STEM workforce grew at a considerably larger annual rate of 4.2 percent. By 2000, the overall STEM workforce was 7.7 times larger than it was in 1950 (the comparable figure for the overall U.S. workforce was 2.3). Moreover, in 1950 the STEM workforce accounted for 1.5 percent of the U.S. workforce; by 2000 this figure had increased to 5 percent.
The STEM occupational group with the largest growth during this 50 year period was mathematics and information technology, which grew at an annual rate of 10.1 percent. In 1950, mathematics and information technology constituted only a small percentage of the overall STEM workforce (2.9 percent); this figure increased to 47.5 percent by 2000. Another notable change is the steady decline in the share of the STEM workforce in engineering occupations (from close to 63 percent in 1950 to less than 27 percent in 2000). The dramatic changes in the mathematics and information technology workforce and the engineering workforce are not unrelated. These changes may be due in part to an increase in the demand for software developers relative to hardware engineers, most notably in the 1980s. Albert Endres, in an essay on the history of software engineering, notes that with the arrival of the personal computer the “traditional dominance of hardware over software ended” (Endres, 1996).
Moreover, CPST (2006a) suggests that the observed changes in STEM occupational employment over the period 1950-2000 may be due in part to changes in the way the census defined occupations. For example, according to another CPST study, computer-related occupations were added to the census in 1970 and were expanded in the 2000 Census by reclassifying a large number of electrical and electronics engineers as computer scientists (CPST, 2006b). The CPST (2006a) also notes that changes in the representation of technicians may be due to changes over time in the way these occupations are defined in the census. These occupational changes are illustrated in Figure 3-2, which shows the distribution of the STEM workforce by occupational group from 1950 to 2000.
Observation 3-2. STEM employment in the United States over the period 1950-2000 saw dramatic shifts in the distribution of the workforce across occupational groups, most notably the shift away from engineering occupations and into mathematics and information technology occupations; these changes were likely due to a combination of changes in the demand for software developers relative to hardware engineers and changes over time in the way occupations in the census were defined.
3 Includes sales engineers and sales representatives for technical and scientific products.