This estimate of the number of companies engaged in research and development is far above the top figure suggested by the most comprehensive previous dataprimarily because this survey included a sample of small firms which were inadequately represented in earlier statistics of industrial research activities. Chart 4 PROPORTION OF SMALL COMPANIES WITH RESEARCH AND DEVELOPMENT PROGRAMS VARIES AMONG INDUSTRIES-NEARLY ALL LARGE COMPANIES HAVE SUCH PROGRAMS As shown in chart 4, the great majority of companies participating in research and development are small. Of the nearly 14,000 manufacturers with research and development programs in 1953, about 86 percent had fewer than 500 employees. However, as the chart also shows, the proportion of companies conducting research and development rises as sharply from one sized group to the next as the absolute number declines. Whereas only 8 percent of the manufacturers with 8-499 employees did research and development work during 1953, nearly 95 percent of those with 5,000 or more employees had research and development programs. As chart 4 also indicates, the proportion of small com panies (8-499 employees) with research and development programs varies greatly among industries. When actual expenditures for research and development are analyzed, however, the concentration of research and development effort in the large companies becomes much more apparent. As chart 5 indicates, large companies bear a greater share of research and development costs than of employment. B. Manpower How large a share do small companies have in scientific and engineering employment? Are scientists and engineers concentrated in large corporations to a greater or less extent than workers in general? Than research and development activities? As chart 5 shows, companies with 5,000 or more workers employed relatively more of the scientists and engineers in manufacturing industries (about three-fifths in January 1954) than of all workers in manufacturing (about two-fifths). The proportion of the estimated total R. and D. cost for manufacturing industries accounted for by these large companies was still greater (over two-thirds). In contrast, firms with 8-499 workers employed only one-fifth of the scientists and engineers and accounted for one-tenth of the R. and D. cost, although they employed over one-third of all workers in manufacturing. Since this is the first time comprehensive estimates have been made of the distribution of scientific and engineering employment by size of company, there is no evidence as to whether or not the moderate numbers of scientists and engineers employed in small companies reflect an upward or a downward trend in utilization of technical personnel by small business. The number of small companies employing scientists and engineers was substantial, however-much larger than the number engaged in research, though quantitative estimates are not available on this point. In general, small companies utilize relatively more of their scientists and engineers in nonresearch activities than large corporations. Thus, the proportion of scientists and engineers engaged in research was approximately 31 percent in companies with 5,000 or more employees, compared with 28 percent in those with 1,000-4,999 employees and slightly under 25 percent in those with less than 1,000 employees. This differential is one of the main reasons for the greater concentration of R. and D. cost than of scientific and engineering employment in large companies, but it is not the only causative factor. Chart 6 indicates the distribution of scientists in various fields of science among companies of varying sizes. It will be noted that physicists are more highly concentrated in large companies than any other professional group for which separate figures were obtained. Four out of every five physicists in industrial employment were working for companies with 5,000 or more employees in January 1954, compared with two-thirds of the mathematics and metallurgists, roughly, three-fifths of the earth scientists, engineers, and chemists, and not much more than two-fifths of the life scientists. Chart 6 LARGE COMPANIES EMPLOY FOUR OUT OF EVERY FIVE PHYSICISTS CHAPTER III. RESEARCH AND DEVELOPMENT FINANCED BY THE FEDERAL GOVERNMENT As indicated earlier in this report, the Federal Government financed over half of the research and development conducted in the United States in 1953 and nearly one-half of R. and D. performed by industry-oriented organizations. It is well to examine in some detail the Federal research and development budget in terms of the kinds of research performed within and outside the Government with Government financing and the distribution of the latter category as between profit and nonprofit organizations. Table III sets forth obligations of the Federal Government for the conduct of research and development in fiscal year 1957, distributed by performance component and agency. With respect to Federal payments to educational institutions for the conduct of research and devolpment, no current data are available as to the distribution between basic research on the one hand and applied research and development on the other. However, estimates obtained from Federal agencies for 1953-54 indicated the following distribution of obligations for research and development at universities, including agricultural experiment stations but excluding research centers managed by universities under contract with the Government. For the major financer of Federal research and development, the Department of Defense, the distribution was: Defense Dept. R&D Basic Dev. 36% Applied Res. For the Atomic Energy Commission, the distribution was: Dev. AEC R&D Obligations at Universities Applied Res. Basic |