University Research, Industrial Innovation, and the Pentagon
By Lloyd J. Dumas
"The Militarization of High Technology"
Edited by John Tirman. Ballinger Publishing Company: Cambridge, MA. 1984.
It is hardly necessary to argue that the progress of technology has enormous impact on the character of human society for we see the effects of that progress all around us. It has shaped the context of our lives. The physical and sociological environment in which we live and work has been profoundly influenced by the technologies we employ. This influence extends to our long-term prospects for development- even for our survival as a species.
It is important to remember, however, that the particular technological pathways we follow, both in research and in its application, are a matter of social choice and not one of scientific necessity. Technology is not a single-lane road that we must follow; rather, it is a complex, interconnected network of roads offering a wide variety of possible paths that lead into the unknown. It is not the imperative of science, but society's choice that determines the roads traversed and those left unexplored. Given the huge effects of technology on society, this choice is an extremely powerful one.
In modern, market-oriented economies, society directs the progress of technology in two main ways. Private research by industries servicing ordinary consumer and producer markets responds to the incentives provided by the marketplace. New products or the betterment of existing products that "sell" will be favored, as will innovations that lower the cost of production. On the other hand, governmental intervention - through regulation, grants and subsidies, and its own purchases of goods and services - also bends and shapes technological development. Governmental intervention as a social force on the progress of technology will form the focus of this chapter.
The Economic and Social Role of Technology
Before World War I , the largest government spenders in the United States were at .the state and local levels. Government, in general, was comparatively small. Since World War II, however, the federal government has become by far the biggest spender, and its size has grown dramatically. Furthermore, for the first time in U.S. history, the military was not essentially disbanded following a war. Though the levels of military spending have varied considerably, they have remained high, year after year, for more than three and a half decades. Roughly one-half of federal "discretionary" expenditures (those not provided for by special trust funds) is today spent on the military.
According to the National Science Foundation, in the early 1980s, the United States devoted some 30 percent of its national research and development (R&D) expenditures to military and space programs. This figure was two to three times the average percentage of national R&D devoted to these noncivilian activities by other members of the Organization for Economic Development and Cooperation (basically, Western Europe, Canada, Japan, Australia, and New Zealand). In the United States, over the entire decade of the 1970s, the fraction of yearly federal budget obligations for R&D going to the military and space programs averaged more than two-thirds of the total. Fifty percent of the total federal R&D obligations went to the military alone.
By the mid-1970s (the most recent data accessible at this writing), 77 percent of the National Sample of R&D engineers and scientists (excluding social scientists) who received federal support for their work received it from the Department of Defense (DoD), NASA, and the Atomic Energy Commission (AEC). Again, more than 50 percent of this support issued from the DoD alone. Nearly three-quarters of the engineers and scientists employed by business and industry receiving federal support at that time received it from the same three agencies.
When interpreting these and other data on military-related R&D, it is important to look beyond the Department of Defense. A significant fraction of the nation's military research and development spending is not included in the DoD budget. Beyond the obvious fact that much of the space program has been concerned with military activities (witness the fraction of the space shuttle's missions devoted to these purposes), the entire funding for the nuclear weapons program, for example, is located outside the Defense Department's budget. Initially located in the Atomic Energy Commission's budget, it was subsequently shifted to the budget of the Department of Energy.
Clearly, military-oriented technology has been a major preoccupation of much of the nation's research and development efforts for more than four decades. And it continues to be so. It is, therefore, appropriate to question how this particular governmentally driven social force has affected the nature of technological development at two of society's main centers of R&D activities - universities and industry. The impact the resulting choice of technological pathways has had on society will be explored as well.
There has always been some disagreement as to the role of the university in society, and hence to the appropriate nature of university-based research. One view, for example, holds that a university is primarily a teaching institution, not a research center. Accordingly, research activity should be restricted - in nature and in quantity - to that which services the university's teaching function. Another view holds that the university is both a center of higher learning and a research institution on a roughly equal basis. In this view, research should not be strongly antagonistic to the teaching function, but it need not have any direct or immediate connection to teaching. Yet a third opinion holds that, more than anything else, the proper role of the university in a free society is to act as a center for open discussion and inquiry, a kind of institution in which the knowledge developed and transmitted is not subject to the support of any particular vested interest, political, economic, or otherwise. The balance or connection between teaching and research activities is not as crucial as maintaining the neutrality and independence of the university as an institution. Thus, research that compromises this independence or systematically biases the product of either the university's teaching or research functions is to be strongly avoided.
During the latter part of the 1960s, debate as to the appropriateness of performing military-oriented research and development at universities raged on campuses across the country. Primarily as a result of growing opposition to U.S. involvement in Vietnam, pressure grew for universities to divorce themselves from R&D that was supportive of that war in particular, and of war-making capability in general. It was argued that the university, as an institution, had a moral responsibility to house only activities, be they teaching or research, that fostered the improvement and enhancement of life and human welfare, not its destruction. Military-oriented R&D thus had no place at a university.
Opponents of this view held that the university as an institution had no right to exclude any teaching or research activities by imposing the values of any one group. The appropriateness of a particular piece of research was a matter for the individual faculty member or student to decide. And, in any case, the university, as part of society, had an obligation to respond to the needs of that society as reflected in the priorities established by its freely elected representatives. The exclusion of military-oriented research and development fro university campuses was thus not merely unpatriotic but antidemocratic.
It seems clear that no institution and no individual in a society can or should be exempted from moral responsibility for the activities they encourage or directly carry out. This is as true for nonprofit organizations, government agencies, and their employees as it is for universities and the individuals who comprise them. In addition, it is the very essence of a free society to allow for a wide variety of concepts of morality. Nevertheless, there is one aspect military-oriented R&D that is obviously inconsistent with the university's special role in society. That aspect is secrecy.
If the university's contribution to a free society is to serve as a center for open discussion and inquiry, it is quite clear that secret research of any kind - whether military-oriented or for that matter business-oriented - has no place in that institution. One may choose not to discuss openly all aspects of one's research as it is proceeding in order to avoid prematurity, misinterpretation of the implication of work still in its preliminary stages, or even plagiarism or prepublication by other researchers. Such restraints are, however, very temporary and a matter of individual judgment. Work that cannot be openly discussed because of its very nature and purpose, or because of constraints external to the university (let alone the researcher) is an entirely different matter.
If the central role of the university is to teach, covert R&D activity may properly be excluded. In such a concept, research is viewed as appropriate only insofar as it enhances the basic function of teaching. Research that cannot be openly discussed can clearly not be openly taught. But secret research may be considered inappropriate on these grounds even if teaching and research are considered of equal importance. Secret research is not merely unsupportive of the teaching function, it is directly antagonistic to it.
It is an ideal of our society that no one who is academically qualified and economically capable of attending any university should be excluded from doing so. Within the university, any course offered is to be open to anyone possessing the proper prerequisite coursework. As a result, information related to secret research cannot be taught in the classroom. To the extent that a faculty member engages in research of this kind, the instructor must not only avoid speaking directly about the research itself while teaching but also must avoid saying anything that might be interpreted as closely related to that work. Otherwise, the faculty member might well run afoul of security regulations and therefore be exposed to severe penalties. Even having to worry about such a possibility cannot help but hamper teaching effectiveness, especially in more advanced courses that are likely to draw on a faculty member's special expertise.
There are those who have argued that secrecy even subverts matters pertaining to military R&D, that it may be a hindrance rather than a help to national security. Among them, interestingly enough, is Edward Teller, hardly a critic of the military or a continued arms race. Teller, for example, has written:
Science thrives on openness - researchers should, and often must, share their findings. . . . Rapid progress cannot be reconciled with central control and secrecy. The limitations we impose on ourselves by restricting information are far greater than any advantage others could gain by copying our ideas. . . . Adopting a policy of openness . . . would strengthen our relationships with our allies as well as illustrate the advantages of freedom to our Soviet colleagues.
Under present circumstances it is quite clear that the world of military R&D is very much a world of secrecy. On this ground alone, its presence on university campuses seems broadly inconsistent with the role of universities in society.
Nevertheless, there is a great deal of military R&D performed at or administered by universities across the United States. Although direct data on the extent of this type of university R&D activity are not readily available, it is not difficult to produce a rough estimate. As of fiscal year 1980, total federal obligations for research and development were an estimated $31.9 billion, some $4.2 billion of which was allocated to universities and colleges directly. An additional $2.2 billion went to federally funded research and development centers (FFRDCs) administered by universities and colleges. Roughly 24 percent of federal R&D obligations to universities and colleges came from the Department of Defense, the Department of : Energy, and NASA. For university-administered FFRDCs, the fraction of federal obligations channeled through these three agencies was nearly 96 percent. Universities and colleges thus received, directly or indirectly, roughly $3.1 billion in federal R&D obligations in FY 1980 from DoD, DOE, and NASA.
An estimated $8.2 billion of R&D (funded from all sources) was performed at universities and colleges and related FFRDCs in 1980. If it is assumed that the amount of federal R&D obligations is a rough estimate of federally funded research actually performed in that same year, then nearly 40 percent of R&D performed directly or indirectly under the auspices of universities and colleges was funded by DoD, DOE, and NASA. If it is arbitrarily assumed that only three-quarters of the funding from the Department of Energy and NASA could be classified as predominantly military in nature, then the military's fraction of total R&D performed at universities and colleges and the FFRDCs they administer would be slightly over 30 percent ($2.5 billion out of a total $8.2 billion).
Another approach to the same estimate is to calculate the fraction of R&D expenditures at universities and colleges funded by the federal government, and then multiply the result by the fraction of federal R&D outlays in that same year categorized as "defense-related" or "space-related.'' The data and results of these calculations for the years 1968, 1970, and 1972-1979 are presented in Table 7-1. These estimates of the fraction of military-related R&D performed at universities and colleges for these years are considerably higher. There are a number of possible reasons for this, including greater emphasis on basic research at universities and lesser emphasis on funding this type of research by the military-related federal agencies.
Since a more comprehensive analysis is beyond the scope of this chapter, and in order to err on the side of underestimating the influence of military R&D at universities, the lower estimate derived from the first approach (30 percent) will be used. Even so, it is clear that military-related work is a major component of university research and development activity.
University faculty often find themselves under pressure to obtain external research funding, and this is particularly true of faculty in the sciences and engineering. Expensive facilities and equipment and considerable technical assistance have become increasingly prerequisite if research is to remain at the cutting edge of in these fields. Then, too, university administrators, aware that a healthy slice of outside funding can be appropriated for general "overhead" purposes, exert another kind of pressure - pressure in the form of criteria used to evaluate faculty salary increases, promotion. and tenure. These considerations can be extremely powerful for junior faculty who base an academic career on the achievement of tenure. Even tenured faculty, however, are not immune to these pressures, nor to the desire for status and recognition associated with continued funding of their own R&D projects. For these reasons and others, that multi-billion dollar pot of military R&D funding constitutes a powerful magnet for faculty in pursuit of research money.
To the extent that faculty are attracted to military R&D dollars, their creative thinking may be directed along lines different from those that might otherwise be followed. Research opportunities presented by existing projects, as well as those represented by the possibility of achieving new funding, tend to be skewed to follow technological pathways of greatest interest and relevance to the military. Even researchers currently working on projects not oriented toward the military have an incentive to keep an eye open for results that might generate spin-off projects of military interest.
The big pot of military R&D dollars does more than distort faculty research. In order to be properly trained, graduate students must have research projects of their own, and they must have faculty to supervise their research as well. If the ongoing research projects at the school they are attending are heavily oriented to military purposes, and if the faculty's attention and expertise are sharply focused in these same directions, the graduate students will also be drawn into this work. And since the nature of work performed at this formative stage of their careers will tend to define the marketability of their skills, as well as shaping their thinking, early exposure to military-oriented R&D may have a very long-term effect. This is accentuated by the fact that much of the financial aid available to graduate students in science and engineering comes in the form of research assistantships, the bulk of which (over 60 percent in the latter 1970s) are federally funded.
Furthermore, since universities ordinarily consider themselves obligated to prepare their graduates to function effectively in the aspects of their chosen fields currently emphasized by society, they will tend to alter their curricula accordingly. If the highest paying, most prestigious jobs lie in military-oriented R&D, schools of engineering and science will tend to modify their course offerings and requirements to reflect this emphasis. For example, since military R&D tends to be far less cost-sensitive than civilian market-oriented R&D, courses on the cost implications of design and on the economic evaluation of technological projects will be de-emphasized. As an illustration, at the School of Engineering and Applied Science of Columbia University, there is such a course entitled "Engineering Economy." In the 1940s, it was a required course for all engineering undergraduates. By the 1970s, not only had it been removed from the list of requirements, but students were often discouraged from taking it by many faculty who thought it "unserious." Enrollment during the 1970s was typically on the order of 15-20 students per semester. Since military R&D requires a far greater level of specialization than is healthy for market-oriented R&D in the civilian sector, fields and subfields of study will also be more narrowly defined. Thus, even the training of students who do not enter military-oriented areas of R&D is seriously affected by the emphasis on military R&D in the society at large.
All of these problems are far more serious at the universities and colleges whose engineering and science programs are of highest repute. Military-oriented R&D funds are not evenly spread but are instead highly concentrated at a relatively small number of institutions. By way of illustration, in fiscal year 1980 the ten universities receiving the most R&D funds from the Department of Defense accounted for 56 percent of total Defense Department R&D funding td all universities and colleges. For the Department of Energy, the comparable figure was 39 percent.
Because the leading institutions consider their graduates to be those who will most strongly advance the frontiers of knowledge, they are that much more sensitive to where the highest paying, most prestigious jobs are to be found, and to where society's priorities have indicated the frontiers of knowledge to be. It is at the M.I.T.'s and Stanfords of the nation where these problems tend to be most clearly seen.
Thus, the estimate that 30 percent of university-based R&D is military-oriented may be seriously understated as an indicator of the extent to which the nation's university R&D capacity is directed toward military ends. For if this 30 percent funding is claiming the talents of a much larger share of our "best and brightest" university-based scientists and engineers, the real impact on the direction of % technological progress in the United States will be far greater.
Industry accounts for the largest share of the nation's research and development activity. In 1981, an estimated national total of $69.1 billion was spent on research and development, more than 70 percent of it ($49.2 billion) by industry. This effort dwarfed that of universities and their associated FFRDCs that spent an estimated $8.6 billion in that year, slightly more than 12 percent of the nation's total. Although industry provided the bulk of its own funding, some $15.8 billion, or nearly one-third of industry's own spending on R&D, was directly financed by the federal government.
Data on funds provided to industry by the federal government specifically for military-related R&D are not readily available. There are data, however, on the distribution of undifferentiated federal R&D among major categories of industry for the years 1968-79, as presented in Table 7-2. It is interesting to note how large one industry looms in these figures. "Aircraft and Missiles" received an average of more than 50 percent of the total federal R&D funds provided to industry over the twelve years covered. According to the National Science Foundation, "This industry also has the largest share of all its R&D supplied by the Government mainly by the National Aeronautics & Space Administration (NASA) and the Department of Defense (DoD)." It is quite obvious that the bulk of federal R&D funds provided to this industry were for military-related projects.
Similarly, the combined "Electronic Components" and "Communications Equipment" industries received an average of more than 18 percent of total federal R&D funds to industry over the years for which data are given for both (1968-74). Even after averaging in the five years (1975-79) for which data for "Electronic Components" are not provided (i.e., assuming it received zero funding), results in a twelve-year average of nearly 16 percent of all federal funding for industrial R&D. Considering the military emphasis on the high technology components of these industries in the United States - particularly electronics - it is highly likely that much of the federal money provided to high-tech industries for R&D was in support of military-oriented research. In fact, according to the National Science Foundation, "The second ranking industry, both in Federal R&D dollars and in Federal share of all R&D dollars is electrical equipment and communication which also is largely funded by DoD."
Thus, three industries highly important to the Pentagon - "Aircraft and Missiles," "Electronic Components," and "Communications Equipment" - alone received a (conservatively) estimated average of approximately two-thirds of total funds provided by the federal government to industry for research and development purposes over the years 1968-79.
Rough estimates of the overall fraction of industrial R&D expenditures that were oriented to military- and space-related purposes through direct federal funding for the period between 1968 and 1981 are given in Table 7-3. The approach is the same as that used in Table 7-1. For the entire fourteen-year period covered, these estimates average more than 25 percent.
As discussed earlier, channeling technological efforts along particular development pathways is not merely the result of direct funding. It is also, in a market-oriented economy, the result of what "sells." Here the federal government has again stepped in to divert even more of the nation's research and development capacity along military-oriented lines. The government has become a major force in the market through its continuing purchase of large amounts of increasingly sophisticated weapons and related systems. Direct, federal, military-oriented R&D funding to industry thus does not encompass the full influence of military demand on the use of the nation's technology-developing capability. As the National Science Hoard points out: "Complete data and information are not available regarding the objectives to which total R&D resources are directed; only in the case of Federal obligations are R&D resource data reported according to specific areas of national concern such as health, energy, and national defense."
This is, of course, as true for universities as for industry. However, while it is reasonable to assume that universities and colleges would not use their own R&D funds (including state and local government support) to fund military-oriented research, it is not reasonable to assume that this is true for industry. Precisely because some segments of industry are focused heavily on servicing the government demand for technology-intensive military goods and services, it is a virtual certainty that a significant amount of their internally funded R&D is a directed toward the development of military-oriented technology. Thus, though our 30 percent estimate of the share of military R&D in total university-performed R&D must still be regarded as conservative, it is likely to be a more accurate reflection of the importance of military R&D to university R&D effort than our comparable 25 percent estimate is for industry. The true share of military-related u K&D in industry is difficult to estimate given the unavailability of relevant data, but it is surely considerably larger than 25 percent.
The Technological "Brain Drain"
The foregoing analysis of the extent to which the military-related technological effort has exerted a claim on the nation's R&D capacity has been cast in terms of R&D expenditures. The overall National Science Foundation estimate presented earlier was that some 30 percent of research and development spending in the United States in the early 1980s has been devoted to the military and space programs. This is essentially consistent with the conservative estimates derived separately for both universities and for industry.
One might argue, however, it is not so much the money spent on research and development for military purposes as it is the claim of military R&D on the nation's most crucial technology-developing resource - its engineers and scientists - that is critical. It is not unreasonable to expect that there would be some sort of rough correspondence between the share of national R&D expenditures and the share of the nation's pool of engineers and scientists directed toward the development of military-oriented technology. Nevertheless, it is worth looking at the question of "brain drain" directly.
According to data for 1982 provided in the Defense Economic Impact Modeling System (DEIMS) of the Department of Defense, only about 14 percent of the nation's engineers and scientists working in industry are included in so-called "defense-induced" employment. However, the methodology used in DEIMS excludes from the "defense-induced" category all employment related to arms exports: nuclear weapons research, design, testing, production programs (all of which are located outside the Department of Defense's budget), and the military-oriented part of the space program. Perhaps even more significantly, the methodology assumes that the percentage of the workforce in any given industry made up of engineers and scientists is the same in the military-sewing part of the industry as in the civilian-oriented part. Yet it is clear that the technological intensity of the labor force in military-oriented segments of industry is far greater. It is not unknown for military-industrial operations to employ one engineer or scientist for every production worker, as, for example, in the Rockwell International's B-1 bomber plant in the late 1970s. Such ratios of technologists to production workers are unsupportable in civilian-oriented industry. Conservatively assuming a 50 percent greater intensity of technologists in the workforce of military-oriented segments of industry, and making a rough correction for all of the military-oriented activity completely excluded by the DEIMS methodology, it can be safely estimated that at least 30 percent of the nation's engineers and scientists are engaged in this form of activity.
The estimate can be approached from another angle by considering the data for full-time, equivalent R&D engineers and scientists provided by the National Science Foundation. These data are presented in Table 7-4 for the years 1970-80. If we, first, extract the data for three major military-oriented industry categories - "Aircraft, and Missiles" and "Electrical Equipment," essentially as before, plus "Machinery" - ignoring all others (industries ignored would include tank production, ordinance, nuclear submarines, etc.); then, make the arbitrary bur not unreasonable assumption that approximately two-thirds of the R&D scientists and engineers in the first two industries and only one-quarter of those in the third industry are engaged in military-related work; and then finally, divide the military-related total for those three industries by the January totals for employment of R&D engineers and scientists in industry as a whole, we would find that, for the period 1970-80, the fraction of the nation's R&D engineers and scientists employed by industry engaged in military-oriented work averages to just under one-third (Table 7-5).
These data in terms of engineering and scientific personnel are for industry only. The attempt to estimate even roughly the fraction of engineers and scientists in universities and nonprofit institutions so engaged is beyond the scope of this analysis. We shall simply assume that the estimate for industry is roughly representative of the use of engineering and scientific personnel as a whole. This seems at least plausible, given the dominance of industry as a performer of research and development in the United States, as well as the relatively close correspondence among all of the admittedly rough estimates of the military claim on the nation's R&D capacity we have derived.
Though it is inappropriate to rely too heavily on the accuracy of estimates so crudely developed, it would appear likely that a great many of the engineering and scientific personnel in the United States have been devoting their talents to the development of military-oriented technology. It is unlikely that the fraction would be substantially less than 30 percent. In all probability, it is far higher, since the calculations referred to above are conservative estimates it is important to remember that the preemption of technological resources has been maintained at this magnitude for two or three decades of more.
Now, the kind of new technological knowledge that will ultimately emerge from any given research or development project is not wholly predictable in advance. By definition, the researchers are engaged in a quest for new knowledge, and such exploration of the unknown and untried must always involve uncertainty. However, while somewhat undeterminable, the kind of new technical knowledge developed is strongly conditioned by the nature of the problems being studied and the type of solutions being sought. Since 38 percent or more of America's engineers and scientists have been seeking military-oriented solutions to military-oriented problems for the past several decades, it should be no surprise that the development of military technology has proceeded at a rapid pace in the United States. Nor should it be surprising that the growth of civilian-oriented technology has been seriously retarded.
Of course, the "spinoff" or "spillover" argument is often invoked. Claiming that military-oriented technological development produces massive improvements in areas of civilian application and so contributes to civilian technological progress, this argument makes very little sense conceptually, and more to the point, is contradicted by straightforward empirical observation While some transferability of technical knowledge between military and civilian applications would be expected (in both directions), conceptually it is difficult to see how directing attention to one area of technical research would routinely produce an efficient generation of knowledge pertaining to a completely different area.
On the empirical side, a 1974 report of a committee of the National Academy of Engineering stated:
With a few exceptions the vast technology developed by Federally funded programs since World War II has not resulted in widespread "spinoffs" of secondary or additional applications of practical products, processes and services that have made an impact on the nation's economic growth, industrial productivity, employment gains and foreign trade.
The seventh annual report of the National Science Board, governing body of the National Science Foundation, expressed concern over the serious erosion of U.S. predominance in science and technology. In several international comparisons, the empirical indicators behind this concern were detailed:
The "patent balance" of the United States fell about 30% between 1966 and 1973. . . .The decline was due both to an increasing number of U.S. patents awarded to foreign countries and a decline (in 1973) in the number of foreign patents awarded to U.S. citizens. Overall, foreign patenting increased in the United States during the period by over 65%, and by 1973 represented more than 30% of all U.S. patents granted. This suggests that the number of patentable ideas of international merit has been growing at a greater rate in other countries than in the United States.
Further, the report describes the relative production of a total of 492 major innovations by the United States, the United Kingdom, Japan, West Germany, and France over the twenty year period from 1953 to 1973:
The U.S. lead . . . declined steadily from the late 1950s to the mid-1960s, falling from 82 to 55% of the innovations. The slight upturn in later years represents a relative rather than an absolute gain, and results primarily from a decline in the proportion of innovations produced in the United Kingdom, rather than an increase in the number of U.S. innovations.
More recently, the National Science Foundation has pointed to a continuation of these downtrends:
U S . patenting has decreased abroad as well as at home.. . . From 1966 to 1976, U.S. patenting activity abroad declined almost 30 percent in ten industrialized countries. . . The decline in U.S. patenting abroad could be attributable to a number of factors, including . . . a relative decline in the U.S. inventive activity.
The relatively poor showing of the United States is even more remarkable considering that these data do nor specifically exclude military-related technology and hence are biased in favor of the United States.
It is worth noting that Japan and West Germany did quite well in these comparisons.
Since 1963, inventors from West Germany have received the largest number of foreign-origin U.S. patents (83,220). In fact. among U.S. foreign-origin patents, West Germany was first in 11 of the 15 major product fields and second in the remaining four. . . . Japan ranks second in the total number of U.S. patents granted to foreign investors between 1963 and 1977 (61.510). Japan has the largest number of foreign parents in three product groups . . . and is second in an additional five categories. . . . Since 1970, Japan has dramatically increased its patent activity by over 100 percent in every product field except the two areas in which it already had a large concentration of patents.
Not so coincidentally, these two countries spend on defense and space only about 4 percent (Japan, 1961-75) and 20 percent (West; Germany, 1961-76) of overall government R&D expenditures, as, opposed to a U.S. average of about 70 percent over the comparable period.
Recognition of the serious retardation of civilian technologic progress is also widespread in the nation's business community. February 1976, a Business Week article, "The Breakdown of U.S. Innovation," opened on an ominous note: "From boardroom to search lab there is a growing sense that something has happened to U.S. innovation. . ." Apparently that "sense" continued to grow because by July 3, 1978, a similar story had made the cover of that journal. The article, entitled "Vanishing innovation'' began: "A grim mood prevails today among industrial research managers. America's vaunted technological superiority of the 1950s and 1960s is vanishing. . . ." The government also clearly recognized that a severe problem existed, as the Carter administration ordered a massive, eighteen-month long, domestic policy review of governmental influence on industrial innovation that involved twenty-eight agencies.
Given the huge amounts of money and technical personnel that have been poured into military-related research over the past several decades in the United States, the severity of the slowdown in civilian technological progress would not have occurred if the "spinoff" or "spillover" effects had been anything more than marginal.
But if the transferability of invention and innovation between the military and civilian worlds was and is actually low, then the decades-long diversion of at least 30 percent of the R&D effort in the United States to military-related work would predictably have produced precisely the sort of civilian technological deterioration that has in fact been experienced. Furthermore, what spinoff there is may sometimes carry with it other significant societal effects - however unintentional - traceable to its military origins.
If, as earlier argued, the nature of the technological pathways explored is a matter of social choice, not scientific necessity, then the social context in which technological development proceeds must shape that choice. The military looked upon as a society within a society, emphasizes uniformity, obedience, and hierarchy. The proper functioning of the unit supersedes the needs of the individuals who comprise it. It is no surprise, therefore, that military-oriented projects were the historical sources of such technological innovations as interchangeable parts in the nineteenth century and automated metal-working machinery in the twentieth century.
While in no way denigrating the positive effects these occasional spinoffs may have on the economy, it is important to be aware of the values imbedded in them. The military is, of necessity, a highly authoritarian system. As such. it is at odds with the principles of personal freedom, individuality, and pursuit of enlightened self-interest - ideals endorsed by the wider body of society in the United States. To the extent that technologies that originated in service of the military bear the inherent values of that system, care must be taken when applying such technologies in the civilian sphere to avoid the subtle corruption of those very ideals central to democratic society as a whole.
Military Spending and Economic Decay
It is widely recognized that civilian technological progress forms the keystone of improved and economic growth. As the National Science Board puts it ". . . the contribution of R&D to economic growth and productivity is positive, significant and high, such innovation is an important factor-perhaps the most important factor - in the economic growth of the United States in this century."
Civilian technological progress is that which leads to improved consumer and producer products and to more efficient methods of production. Such progress contributes to greater labor and capital productivity through the development of new machinery proved production techniques, and consequently the more efficient use of productive resources in general. Accordingly, as the technological brain drain generated by the military sector led to a deterioration in the rate of civilian technological development, productivity rates began to collapse.
From 1947 to 1967, output per hour grew at an average annual rate of 3.4 percent in the nonagricultural business of the United States, according to the Council of Economic Advisors. From 1967 to 1977, that average rate of growth declined sharply to 119 percent per year. From 1977 to 1982, productivity growth entirely disappeared, the index of output per labor hour bring the same in 1982 as it was in 1977. The deterioration of productivity growth has thus been accelerating.
The improvement of productivity plays a crucial role in countering inflationary pressures, for it is sustained growth in offsets the effects of rising input costs. It is not the separate cost of labor, fuels, materials, and capital that is relevant to the determination of product price, but rather the combined cost of these productive resources per unit of product. Thus, the rise in labor costs, for example, might be at least partially offset by substituting cheaper capital for increasingly expensive labor or by organizing production to use labor more efficiently, or both. As long as the net result is the production of more output per unit of input, rise in input costs need not be fully translated into rises in the cost per unit of product. Correspondingly, the upward "cost-push" pressures on price will be mitigated. But productivity is nothing more than a measure of output per unit of input. Hence, rising productivity permits the absorption of rising labor or fuel prices, without full reflection of these resource-cost increases in unit cost and thus in price.
The deterioration of productivity growth substantially compromises this cost-offsetting capability. In the absence of strong productivity improvement, rising costs of labor or fuels, for example, will be translated into rising product prices. As this occurs over a whole series of industries, a self-reinforcing rise in the general level of prices or "inflation" is generated.
As the price of goods produced in America rose higher and higher (and quality too often failed to meet world standards), the nation's industries became less and less competitive vis-à-vis foreign production. Overseas markets were lost and the U.S. export position weakened. Domestic markets were lost to foreign production and the U.S. import position worsened. The progressive loss of markets induced cutbacks in U.S.-based production, resulting in high unemployment rates. This problem was exacerbated by the flight of U.S.-owned production facilities to cheap labor havens abroad, as one logical response to the inability to offset higher costs in the United States. The declining competitiveness of U.S. industry, the result of decreasing productivity growth, has generated unemployment even in the face of high product demand.
Inflation at historically high levels continued to plague the American economy until the depth of the 1980-82 recession drove the country to unemployment rates averaging nearly half of those of the Great Depression. And as the subsequent recovery has proceeded, inflation has once more become a threat, despite the persistence of high levels of unemployment.
Productivity growth continues to be "the economic linchpin of the 1980s," according to the Joint Economic Committee of the Congress in its mid-1979 analysis of prospects for the economy. Its warning that, as The New York Times put it, "The average American is likely to see his standard of living drastically reduced in the 1980s unless productivity growth is accelerated" is precisely correct.
At the end of 1979, I wrote:
During the decade of the 1970s the dynamic process of deterioration which has been described here has produced unprecedented simultaneous high inflation/high unemployment. . . .For an entire decade, the inflation rate has averaged near 7% at the same time the unemployment rate averaged more than 6%. The economic prognosis for the coming decade is not good. If the arms race continues unabated, and we somehow manage to survive these rates of inflation and unemployment rates that were viewed as horrific at the beginning of the 1970s - will look like economic good times compared to what will be commonplace by the end of the 1980s.
It now appears that this wild prognostication may turn out to have been a conservative estimate.
Summary and Conclusions
Beyond its year-by-year effects, the military budget has had an enormously negative long-term impact on the functioning of the U.S. economy. It has preempted at least 30 percent of the R&D effort administered by universities and colleges (as well as a similar share of industrial R&D activity and has laid claim to the talents of a third or more of the nation's pool of scientists and engineers. Through this "brain drain" effect, the persistence of high levels of military-related technological activity has produced a severe retardation of the growth rate of civilian-related technology. The retardation, in turn, has played a key role in producing a serious slowing of the nation's productivity growth, even to the vanishing point. The failure of productivity pushed industry increasingly away from traditional cost-offsetting behavior and into a cost "pass-along" mode. Input-cost increases were routinely translated into output-price increases. And as the prices of U.S.-produced goods and services rose, domestic production priced itself more and more out of foreign and domestic markets, leading to layoffs and growing unemployment in the United States.
The effects of this damage to the competitiveness of U.S. industry, wrought by more than three decades of persistently high military spending, surfaced with a vengeance in the 1970s. Collapsing productivity, high inflation, and high unemployment have been the sad legacy of our participation in the ongoing international arms race.
It is ironic that, in our blind quest for national security through the expansion of military capabilities, we have undermined the very source of our rise to international prominence and influence - the power of industry fuelled by the spectacular efficiency of technological capability. We have not increased our power by continued expansion of the destructive capability of our armed forces. We have, in fact, diminished it. We should have taken more seriously the warning that General Dwight Eisenhower gave us when he left the presidency in 1961: "The Military Establishment not productive of itself, necessarily must feed on the energy, productivity and brainpower of the country, and if it takes too much, our total strength declines." He could not have been more correct.
The policy implications of the analysis presented here are straightforward. Tinkering with the money supply, interest rates, tax policy, and the like, may be comfortingly familiar, but it will at best produce only temporary and cosmetic improvements in the economic situation. The fundamental deterioration of the economy cannot be undone without a revitalization of productivity growth, and that will not be achieved unless and until we have moved a large fraction of the engineers and scientists now performing military-related work into productive civilian activity. The question is not so much one of economic policy as of political will.
Notes to Chapter 7
 National Science Foundation. National Patterns of Science and Technology Resources, 1981. (Washington, D.C.: Government Printing Office, 1981), p. 14.
 National Science Foundation, Science Indicators: 1978 (Washington, D.C.: Government Printing Office, 1979), p. 182.
 National Science Foundation, Surveys of Science Resources Series, Characteristics of the National Sample of Scientists and Engineers: 1979, Part 2, Employment (Washington, D.C.: Government Printing Office 1975), Table B-16, pp. 128-142.
 Ibid., Table B-15, pp. 113-127.
 Stephen H. Unger, "National Security and the Free Flow of Technical Information," Committee on Scientific Freedom and Responsibility, American Association for the Advancement of Science (September 1981), pp. 12,14.
 Edward Teller, "Secrecy: The Road to Nowhere," Technology Review (October 1981).
 National Science Foundation, National Patterns, Table 24, p. 35.
 National Science Foundation, Survey of Science Resource Series, Federal Support to Universities, Colleges and Selected Nonprofit Institutions Fiscal Year 1980 (Washington, D.C.: Government Printing Office, 1982), Table B-41, p. 148.
 Ibid., Table p. 29.
 Ibid., Table B-41, p. 148.
 Ibid., Table B-9, p. 29 and Table B-41, p. 148.
 National Science Foundation, National Patterns p. 21.
 Assuming the overall average fraction of federally funded R&D that was military-related applied to federally funded R&D at universities would lead to an overestimate of university military research under these conditions.
 National Science Foundation, National Patterns, Tables 63 and 64, pp. 65 and 66.
 Calculated from data in National Science Foundation, Federal Support to Universities, Table B-16, pp. 41-42,
 National Science Foundation, National Patterns, Table 1, p. 21.
 National Science Foundation, Science Indicators, pp. 84-85. This was specifically in the context of a discussion referring to the years 1967 and 1977.
 Ibid., p. 85.
 Ibid., p. 43.
 It is also worth noting that for the same reasons the estimation approach used in Table 7-1 may have yielded higher estimates than appropriate for universities, the estimates in the final column of Table 7-3 may be biased downward. To the extent that federal military R&D funding is oriented mainly to applied research and development (and not to basic research), the emphasis on this type of R&D in industry may imply that a higher than average share of overall federal R&D is "defense-related" and "space-related" there. Thus, the share of federal R&D funds provided to industry for military purposes may be underestimated in Table 7-3.
 U.S. Department of Defense, "Estimates ,of Industrial Employment by Occupation (Engineers & Scientists)," Defense Economic Impact Modeling System (Occupation by Industry Model) (Washington, D.C.: Government Printing Office, 1983).
 National Academy of Engineering Committee on Technology Transfer and Utilization, "Technology Transfer and Utilization, Recommendations for Reducing the Emphasis and Correcting the Inbalance" (Washington, D.C.: National Academy of Engineering, 1974), p. i.
 National Science Foundation, Report of the National Science Board, 1975, Science Indicators: 1974 (Washington, D.C.: Government Printing Office, 1976), p. 17.
 Ibid., p. 19.
 National Science Foundation, Report of the National Science Board: 1979, Science Indicators, 1978 (Washington, D.C.: Government Printing Office, 1979), pp. 20 and 21.
 Ibid., pp. 19 and 20.
 Ibid., pp. 146 and 147.
 "The Breakdown of U.S. Innovation," Business Week (February 26, 1976).
 "Vanishing Innovation," Business Week (July 3, 1978), p. 46.
 An argument along these lines is developed more fully in David F. Noble, "The Social and Economic Consequences of the Military Influence on the Development of Industrial Technologies,- in L. J. Dumas, ed., The Political Economy of Arms Reduction: Reversing Economic Decay (Boulder, Colorado: Westview Press, 1982).
 National Science Foundation, Science Indicators: 1976 (Washington, D.C.: Government Printing Office, 1976).
 Council of Economic Advisors, "Annual Report" in Economic Report of the President (Washington, D.C.: Government Printing Office, 1983), Table B-40, p. 208.
 Clyde H. Farnsworth, "Lag in Productivity Called Major Peril to Living Standard," The New York Times (August 13,1979).
 L. J. Dumas, "The Impact of the Military Budget on the Domestic Economy," Current Research on Peace and Violence, no, 2 (1980), p. 81.