the primary investing in a given product line. In this respect the holding of patents can be an instrumental device in the competitive struggle.

The managements interviewed in the course of this investigation reported uniformly that the major factor in the operation of their research laboratories and in research decisions has been the extent to which new knowledge gave advantages to their own firm, not the extent to which it resulted in income from patent licenses. Indeed, fees from the latter accounted for not more than 10 percent of annual outlays for industrial research.36

While the opportunity to obtain patents does not appear to have a controlling effect upon the intensity of industrial research activity, it can significantly affect its timing and location. Research on industrial machinery, for example, is typically carried out by the users or by the conventional vendors of such equipment.

Finally, patents are widely used as instrumental devices in competition among firms. Aspects of this use of patents have been recorded in court records, in the hearings and reports of various governmental bodies, and in other studies of the patent system.37

It should be emphasized, however, that the patent is only one of many devices and techniques that have been used in interfirm competition.3

38

5. Industrial research and the promotion of science and technology

There can be no doubt that many of the results emerging from industrial research laboratories are relevant to the promotion of science and the useful arts. It is nevertheless the case that the advancement of knowledge as an end in itself is not an objective of industrial research activity. Possibilities for industrial firms to facilitate the promotion of science and technology by the expenditure of large resources have existed for a long time. The scale of present industrial research outlays and their growth is traceable not to the search for knowledge in itself but to the fact that the production of new technical knowledge has become an increasingly important competitive weapon in the struggle for position in industry.

6. Research facilities as assets

Owing to the critical role played by industrial research organizations in interfirm competition, the very existence of such research organizations becomes an important factor in determining the relative standing of firms. Industrial firms now need engineers and scientists to serve as a reservoir of knowledge for directing the development of new products and new production methods. As a result, the very possession of the industrial research laboratory means possession of the means for producing new knowledge that feeds into product and process development technique. It should be noted that these values in an industrial research facility exist quite independently of whether

It is true that some firms make a specialty of developing patent rights and licensing them at fees which, in the aggregate, are a substantial portion of the firm's total income. This practice, however, was not found among the large industrial research establishments examined during this inquiry, and does not appear to be prevalent.

See, e. g., Vaughan, The United States Patent System (Norman: University of Oklahoma Press, 1956), which contains an excellent bibliography.

"Burns, The Decline of Competition (New York: McGraw-Hill Book Co., 1936).

20609-58

patents are taken out or not. The operation of industrial research facilities thus becomes, in itself, a form of insurance for protecting or advancing the relative position of the firm in interfirm competition.39

Summary

In summary: The discussion in the preceding chapters has indicated that the leading characteristic of the production of technical knowledge under modern conditions is the division of labor, and the necessary integration of work that must accompany it. These modern characteristics of the production of technical knowledge are reflected in the manner of payment of engineers and scientists. This is primarily on a salary basis, that does not vary with output per unit of time.

These features of research activity are also reflected in the problems of determining the cost of performing particular research activity. Such costs are usually overhead in character, and therefore are not necessarily traceable to particular units of research output. This overhead character of the industrial research cost structure stems from the integrated character of much of the creative activity that is involved.

Finally, it is noted that the factors influencing industrial research activity are the changing circumstances of interfirm competition, much more than the availability of patents. Likewise, the conduct of research in universities and other nonbusiness groups is largely independent of patent considerations, since in the nonbusiness laboratories knowledge is produced for its own sake and the search for knowledge irrespective of its commercial and industrial value, is a primary criterion in selecting research projects.

Against this background of conditions surrounding modern technical research, this inquiry now turns to an evaluation of such research and technical knowledge in terms of its use as property. This is the major theme of part II.

On this matter it is instructive to review the testimony given in 1940 by Charles F. Kettering, vice president in charge of research at General Motors Corp. TNEC hearings, supra, note 14, at 16292-16317.

PART II. THE USE OF TECHNICAL KNOWLEDGE AS

PROPERTY

CHAPTER VII. IMPLICATIONS FOR BUSINESS

What are the implications from the business and competitive standpoint of a firm taking out patents? To answer this, one must first examine the extent of patenting activity and ascertain whether it has kept pace with the expansion of business-sponsored research and development. Secondly, the use of patents as a weapon in interfirm competition will be reviewed. Finally to be examined is the question whether extensive patent holdings are a necessary condition for successful competition among firms.

A. PATENT ACTIVITY AND BUSINESS-SPONSORED RESEARCH

Among persons who are active in the patent field, there is substantial agreement that the opportunity to obtain patents has been a potent incentive for research outlays by industrial firms. If this assumption is valid, it supports the contention that patenting is an important factor in industrial research. But do the facts support the assertion? One way to find out is to examine the relation between patenting and research activities. The relation between research and patent activity can be approached in two ways, to wit, in aggregate terms and through the detailed records of particular firms.

For the period 1941-54 there are reliable estimates of the total number of research scientists and engineers in the United States. These data can be compared with the number of patents granted on inventions by the Patent Office. The data of table 1 show a dramatic growth in the number of scientists and engineers from 87,000 in 1941 to 194,000 in 1954, an increase of 120 percent. During the same period the number of patents issued shows a marked fall. Even if allowance is made for administrative problems of the Patent Office and other factors, the larger picture that emerges is plain enough. There has been no growth in the number of patents taken on inventions that matches the increased number of scientists and engineers. in the industrial and other research laboratories in the country.40

#One may properly ask whether it takes more manpower and costs more to make a given invention today than it used to? The data available here do not give a direct answer to this question. The several tables shown in this chapter do show, however, that the differences in rates of development between technical employment (and budgets) and patenting are large, both in the long run (1900-1954) and during shorter periods (1941-54, 1950-55, 1940-55, and 1942-54). It seems unlikely that changes in the state and circumstances of science and production technology during these shorter periods have been sufficiently great to explain the lag in patenting activity shown by these statistics. The writer is advised that there is no indi cation of any steady upgrading, over the decades, in the Patent Office standards of invention which would account for the reduction in patenting-at least to an extent beyond that which would balance the increased competency of technicians. Whitmore, What's Got Into the Office Lately 29 JP.O.S. 869 (1947); cf. Hearings before the Senate Subcommittee on Patents, Trademarks, and Copyrights, 84th Cong., 1st sess., at 72-95 (October 10-12, 1955).

27

TABLE I.-The number of patents granted for inventions in relation to the growth of research scientists and engineers

1 Department of Defense, Office of the Secretary of Defense (Research and Development), The Growth of Scientific Research and Development, at 12 (1953).

Department of Commerce, Historical Statistics of the United States, 1789-1945, at 312 (1949); Statistical Abstract of the United States, 1954, at 520 (1954); id., 1955, at 505 (1955).

Even more striking is the lag that appears when one examines technological effort generally. Activity in patenting as compared with change in the total number of scientists and engineers, not just those engaged in research and development work, is shown by table 2. As this table shows, from 1900 to 1954 there occurred a massive expansion in the number of scientists and engineers from 42,000 to 691,000, an increase of 1,600 percent. In contrast, over this same period the largest increase shown in any given year in the number of patents granted was only 83 percent over the number granted in 1900 and in most years the increase was even smaller than this. The relationship between the growth of the number of scientists and engineers and the number of patents granted is reflected in the fifth column of table 2, by dividing the index of patent growth by the index showing the growth in scientists and engineers. The data show a steady and persistent decline from 1900 to date in patenting as compared to overall scientific and engineering employment. In other words, activity in taking out patents has lagged increasingly behind the total activity in science and engineering, as indicated in the growth of the total population of scientists and engineers, until today it is less than onetenth of what it would be had it kept pace."1

4 It may be suggested that the growth in Government-contracted research work during recent years may involve the employment of technical men on work that is not con lucive to patenting by the contractor firm, since the patents would be Government-owned. "his would not, however, explain the decline shown in table 2 which shows relative patenting activity since 1900. Nor does it explain the situation among the individual firms, as shown in tables 3 to 6 and in figure 1. Of the firms, only one is very active in Government-contracted research and development and the data for this firm have been adjusted to exclude the Government-contract research staff.

TABLE 2.—The number of patents granted for inventions in relation to the growth of engineers and scientists

National Science Foundation, Scientific Personnel Resources, at 9 (1955). Department of Commerce, Historical Statistics of the United States, 1789-1945, at 312 (1949); Statistical Abstract of the United States, 1955, at 505 (1955).

When one turns from the general to the specific and examines the detailed data of the firms that were sampled for this investigation, the same picture is presented, differing only in degree. Detailed statistics covering major firms in the electronics, chemicals, transportation, and petroleum industries are presented in tables 3 to 6. Table 3 shows the data for firm A during the period 1950-55. There was a threefold increase in the number of scientists and engineers engaged in research and development, and the number of engineers engaged in manufacturing (i. e., concerned with the design of production facilities and related activity) increased in about the same degree. During this period there was also an increase in the number of patent applications, but in contrast to the almost fourfold increase in the number of engineers and scientists employed in research, these increased by only two-thirds; in other words, at less than half the rate that research activity increased.

In firm B (table 4) the available data show the development from 1940 to 1955. Although detailed and exact statistics were not available, it appeared that the number of patent applications on inventions made during this period fluctuated between 300 and 350 each year. During this same period, however, the number of scientists and engineers in the laboratories increased by about 40

percent.

The data for firm C are more complete, insofar as they show both the patents issued for selected years and patent applications filed. By both criteria, patenting activity showed a manifest decline over the period 1940-55, while the number of research scientists and engineers in the firm increased about 50 percent; again, showing patent activity increasing at less than half the rate that research activity increased.

In the case of firm D, the picture is different. Here, there was a doubling in the total research and development staff from 1940 to 1955 which was accompanied by a proportionate, and in some years even greater, increase in patent activity. The management of this firm includes a large patent staff, with an elaborate file of patents, both domestic and foreign, that relate to its sphere of activities. This