Innovation and the State

[Should the state back science or even plan its progress? In this excerpt from a longer and previously unpublished study written in 1959—two years after Russia’s launch of Sputnik and two years before the first manned space flight—Murray Rothbard (1926–95) argues that socially beneficial scientific innovation comes from independent thinkers working within the market economy, while the state distorts science and produces no really important innovations. This paper is not only remarkable for its theoretical rigor but also for its enduring quality: it was written at a time when the idea of state planning for science was at its height. See the full text here.]

Charges abound that scientific research, left to the mercies of the free market, would be insufficient for modern technological needs. The general principles of government policy in this field we have already set forth: (a) leaving the general allocation of resources purely to the free market—the profit and loss incentive and test of the free market being the only efficient way of allocating a country’s resources in the way best calculated to satisfy consumer demand. This principle applies fully as well to scientific research as to any other sphere; and (b) for the military needs of research, acting only as a consumer rather than as a producer using funds to pay for private scientific contractors. In actual practice, the Federal government is already doing a great deal (although, as we shall see below, it can do much more) in this direction, by channeling most of its military research funds into private contractors, whom the military sees to be more efficient than government operation.1

Let us first turn to the problem of general research, however. Is it really true that such research will be deficient on the free market?

We have, first, been hearing a great deal of how much resources the Soviet Union has been putting into scientific research, and how we must redouble our efforts in order to catch up. But the National Science Foundation has estimated that the Soviet Union has been putting a little over 1% of its national product into research and development. The Steelman Report of 1947 called for the United States to place 1% of its national product into research and development, in the years ahead. Yet, we now have 2% of our product going into “R and D,” and out national income is far, far higher than that of the Soviets.2  In 1953–54, private sources contributed $2.6 billion to R and D; this contrast to a total of $530 million of private funds in 1941. In fact, with the exception of pure, or basic, research (which we will study further below) the National Science Foundation’s study conceded the sufficiency of private scientific research in American industry.

The flourishing of private research in our modern age had been eloquently hailed by General David Sarnoff, board chairman of RCA:

Today, science and industry are linked by arteries of progress and their lifeblood is technical research…. The patter of our industrial progress … lies in a partnership between those who create good things and those who produce and distribute and service them. It lies in teamwork between research and industry.3

We have seen that government subsidization or operation of (non-military) research would distort the efficient allocation of resources of the free market economy. It would do more; as Sarnoff pointed out, government aid would inevitably mean “increased government control of the daily lives of all the people.” Secondly, government control would tragically bureaucratize science and cripple that spirit of free inquiry on which all scientific advance must rest: “government control of research would destroy the very qualities that enable researchers to make such an important contribution to society. For government control means that rigid lines would be set for research; and these lines may not meet changing requirements. Certainly industry is best qualified to define its own research needs. And the partnership between research and industry loses its meaning when government can dictate the subject and objective of research in any competitive system of private enterprise.4

The myth has arisen that government research is made necessary by our technological age, because only planned, directed, large-scale “team” research can produce important inventions or develop them properly. The day of the individual or small-scale inventor is supposedly over and done with. And the strong inference is that government, as potentially the “largest-scale” operator, must play a leading role in even non-military scientific research. This common myth has been completely exploded by the researches of John Jewkes, David Sawers, and Richard Stillerman in their highly important recent work.5, 6    Taking sixty-one of the most important inventions of the twentieth century (excluding atomic energy, which we will discuss below), Jewkes et. al. found that more than half of these were the work of individual inventors—with the individuals working at their own directions, and with very limited resources. In this category they place such inventions as: air-conditioning, automatic transmission, bakelite, the ball-point pen, catalytic cracking of petroleum, cellophane, the cotton picker, the cyclotron, gas refrigeration, the electron microscope, the gyro-compass, the helicopter, insulin, the jet engine, kodachrome, magnetic recording, penicillin, the Polaroid camera, radio, the safety razor, titanium, and the zipper. The jet engine was invented and carried through its early development, practically simultaneously, by Britons and Germans who were individual inventors, either completely unconnected with the aircraft industry or not specialists in engines. The gyrocompass was invented by a young German art historian. The bulk of the basic inventions for radio came from individual inventors unconnected with communications firms, some of whom created new small firms of their own to exploit the invention. The cyclotron was invented and partly developed by a university scientist, using simple equipment in the early stages. Penicillin was invented and partly developed in a university laboratory, and insulin was invented by a general practitioner who used a university laboratory.

Of the inventions studied that were achieved in industrial research laboratories, some arose in small firms, others were more or less accidental by-products of other work rather than preplanned and predirected. Terylene, the synthetic fibre, was discovered by a small research group in a firm not directly interested in fibre production. The process of continuous hot strip rolling of steel sheets was thought up by an individual inventor and then perfected in a small steel company. The LP record was invented by an engineer working on it as an individual sideline, and then was developed by another corporation.

In other cases, inventions in the research laboratories of large companies were made by small research teams, often centered around one outstanding man. Such was the case with Nylon, at the DuPont laboratories.7

The twentieth century has produced some great independent inventors, creators of many important new devices. One of them, the Englishman S.G. Brown (components for telegraphy, telephony, radio, and gyro-compass) declared: “if there were any control over me or my work every idea would stop.” Brown never accepted financial aid for experimental work, or for producing a new device. How would such a man fare under the control of a government-directed research team, or one that was government-controlled? P.T. Farnsworth, great television pioneer, has always preferred to do his research on a small scale and with simple equipment. F.W. Lanchester, great British inventor in aerodynamics and engineering once wrote: “the salient feature of my career … (is that) … my work has been almost wholly individual. My scientific and technical work has been almost wholly individual. My scientific and technical work has never been backed by funds from external sources to any material extent.” Lee de Forest, eminent inventor of the radio vacuum tube, always found it difficult to work under any conditions short of complete autonomy. Sir Frank Whittle invented the jet engine with very slim resources.

C.F. Kettering often positively preferred simple equipment. And R.M. Lodge recently warned:

The trend towards more and more complex apparatus should be carefully watched and controlled; otherwise the scientists themselves gradually become specialist machine-minders, and there is a tendency, for example, for analytical problem to be passed from the microanalytical laboratory to the intra-red laboratory and from there to the mass spectrographic laboratory, whereas all the time all that was needed was a microphone and a keen observer.8

The worthy individual inventor is far from helpless in the modern world. He may, in a free enterprise system, become a free-lance consultant to industry, may work on inventions on outside grants, may sell his ideas to corporations, may form or be backed by a research association (both profit and non-profit), or may obtain aid from special private organizations that invest risk capital in small speculative inventions (e.g. the American Research and Development Corporations).

One very important reason for the success of the independent inventor, and his preservation from the dominance of large-scale government-controlled projects, stems from the very nature of invention: “The essential feature of innovation is that the path to it is not known beforehand. The less, therefore, an inventor is pre-committed in his speculation by training or tradition, the better the chance of his escaping from the grooves of accepted thought.9 There are many recorded instances of the inventor winning out despite the scoffing of the recognized experts in the field, perhaps even emboldened because he didn’t know enough to be discouraged. One authority maintains that Farnsworth benefited from his lack of contact with the outside scientific world. Once, a professor gave him four good reasons why his idea—later successful—could not possibility work. Before the discovery of the transistor, many scientists claimed that nothing more could be learned in that field. Eminent mathematicians once claimed to prove logically that short-wave radio was impossible. Government-controlled research would undoubtedly rely on existing authorities, and thus would snuff out the searchings of the truly original minds. Many of the great inventors of recent times could not have gotten a research job in the field for lack of expertise: the inventors of Kodachrome were musicians; Eastman, the great inventor in photography, was a bookkeeper at the time; the inventor of the ball-point pen was an artist and journalists; the automatic dialing system was invented by an undertaker; a veterinarian invented the pneumatic tire. Furthermore, there are many inventors who are part-time, or one-shot, inventors, who are clearly more useful on their own than as part of a research team.

As the eminent British zoologist John Baker points out, the life of an independent researcher involves the willingness to bear great risks: “The life is too strenuous for most people, and the timid scientist hankers after the safety of directed team-work routine. The genuine research worker is [an] altogether different kind of person.”10  Darwin once wrote: “I am like a gambler and love a wild experiment.” The importance of self-directed work to great scientists is stressed by the Nobel prize-winning chemical discoverer of vitamins, Szent-Gyorgyi, who wrote: “The real scientist … is ready to bear privation … rather than let anyone dictate to him which direction his work must take.”11

Not only inventors, but many types of scientists benefit from the work of independent researchers in their fields. Einstein said that: “I am a horse for single harness, not cut out for team-work,” and suggested that refugee scientists take jobs as lighthouse-keepers, so that they could enjoy needed isolation. The fundamental discoveries in valence theory, cytogenetics, embryology, and many other fields of twentieth-century biology, were made by individual scientists.12  Scientific discoveries, furthermore, cannot be planned in advance. They grow out of apparently unrelated efforts of previous scientists, often in diverse fields. The radium and X-ray treatments for cancer owe most, not to planned research, on cancer cures, but to the discoverers of radium and X-rays, who were working for quite different goals. Baker shows that the discovery of a treatment for cancer of the prostate emerged out of centuries of unrelated research on: the prostate, phosphatase, and on hormones, none of which was aimed toward a cancer cure.13

Apart from individual scientists and inventors, there is also great need for the existence of small research laboratories in small firms as well as in large ones. There is inevitably a clash between practical administrators of research and the scientists themselves, and the evils of bureaucratic administration and crippling of scientific endeavor will be infinitely greater if science is under the control or direction of the Ultimate Bureaucracy of government.14

O.E. Buckley, when President of the Bell Telephone Laboratories, stated: “one sure way to defeat the scientific spirit is to attempt to direct enquiry from above. All successful industrial research directors know this and have learnt by experience that one thing a director of research must never do is to direct research.” Similar views have been expressed by C.E.K Mees, of Eastman Kodak, and Sir Alexander Fleming, discoverer of penicillin, who said: “certain industrial places … put up a certain amount of money for research and hire a team. They often direct them on the particular problems they are going to work out. This is a very good way of employing a certain number of people, paying salaries, and not getting very much in return.”15  Jewkes and his colleagues, describing the best ways of crippling a research organization, might have had a typical government operation or control in mind:

The chances of success are further reduced where the research group is organized in hierarchical fashion, with ideas and instructions flowing downwards and not upwards … where the direction to research is … closely defined … where men are asked to report at regular intervals … where achievements are constantly being recorded and assessed; where spurious cooperation is enforced by time-wasting committees and paper work.16

In gauging the effectiveness of large vs. small-scale research, we should remember that whether or not a firm engages in research at all (apart from government contract) depends on the type of industry it is in. The great bulk of manufacturing firms, for example, do not engage in research and development at all. The one-tenth that do, are mostly in technologically advanced and advancing industries, where expanding scientific knowledge is needed, and where many scientists must be hired anyway for test and control work. On the other hand, industries that rely more on empirical rather than scientific knowledge do less research. Some large-scale industries, like chemicals, do a great deal of research; while others, such as iron and steel, do much less. Some small-scale industries do little research, while others, like scientific instrument firms, do a relatively great amount. And while the bulk of industrial research is done by the very large firms, we have seen the vital role of the independent inventor (and later we shall see further the crucial role of the university laboratory in basic research). Furthermore, it has been found that in those firms that do conduct research, the number of research workers per 100 employees is higher for the small, and lowest for the large firms.16

It should be noted that few of the Nobel Prize winners since 1900 came from the large industrial research laboratories. Furthermore, many of the current research labs of the big corporations originated as small firms, which were later bought by the big corporation. This happened with General Motors, and with General Electric. The large corporations also make a great deal of use of outside consultants and independent research organizations (both profit and non-profit making). This certainly must confound the partisan of organized, large-scale government-controlled and directed research: for if organized, large-scale research is invariably more efficient, why do these big corporations bother with small outside firms? Here are some of the reasons given by the big firms themselves:

They may be short of trained people. Or they may be confronted with a task of a non-continuing nature which they prefer to have out to others … or they may be confronted with a type of technical problem new to them which they feel they cannot handle at all. Or, having been continually defeated by some technical problem, they may hand out the task to others who will come to it with fresh minds and no preconceptions.18

Resistance of an organization to new ideas has occurred significantly even in efficient, alert corporations—how much more would it occur in government, where there is neither the incentive nor the possibility of a profit-and-loss check on its efficiency! Thus: the telephone, cable, and electric manufacturing companies were originally apathetic about the possibilities of wireless telegraphy; RCA resisted Armstrong’s FM ideas; the Edison Company, at the turn of the century, scoffed at the idea of a gas motor for transportation, insisting on the future of the electric motor for that purpose; the established aircraft-engine firms scoffed at the jet engine and at the retractable under-carriage; the British and American chemical firms were highly critical of penicillin, and almost refused to take part in its development; The Marconi Company expressed no interest in television when it was brought to their attention in 1925; the manufacturers of navigational equipment took no part in the invention of the gyro-compass. When the Ford Motor Company sought to introduce automation in their factories, they turned to the small specialized firms in the machine-tool industry, “The small uninhibited firms with no preconceived notions.” And even Henry Ford resisted the thermostat, or hydraulic brakes.

Furthermore, in many of our biggest industries, the critical innovations of the twentieth century have come from outside the big firms. Of the three big inventions in the aluminum industry up to 1937, two came from men outside the industry—despite the fact that ALCOA had an aluminum monopoly during those years. The two significant new ideas in steel-making in this century cam from a newcomer and from one of the smaller steel firms (continuous hot strip rolling), and the other from an individual German inventor (continuous casting). The large-scale, progressive automotive industry has benefited a great deal from outside ideas—including automatic transmissions and power steering, and small firms and accessory manufacturers have contributed new systems of suspension. In the progressive, large-scale petroleum industry, which devotes heavy expenditures to research, many leading ideas have come from small firms or outside individuals including catalytic cracking: “Looking back dispassionately we find that (the major oil companies) mainly took up and developed ideas, which were brought to them by men who did not, in the first instance, belong to their own team.”19

Another important point is that most industrial research laboratories, even in the large companies, are themselves small; more than one-half of the laboratories in the U.S. employ less than 15 scientists, and most of these are for routine or development work, rather than research. The average operating cost of a laboratory per research scientists is about $25,000—not a prohibitive sum for an average sized firm. Moreover, 49% of all firms holding patents, in 1953, had fewer than 5000 employees all told.

Many laboratories, while remaining at the same size, have fluctuated greatly in their failure or success over time, depending on the qualities of their personnel and, above all, their leadership. The leading inventors in these laboratories themselves stress the virtues of small groups. Fermi has said: “Efficiency does not increase proportionately with numbers. A large group creates complicated administrative problems, and much effort is spent on organization.” And, in a striking anticipation of Parkinson’s Law of Bureaucracy, S.C. Harland wrote this about the large lab:

You see crowds of people milling around with an air of fictitious activity, behind a façade of massive mediocrity. There is a kind of Malthusianism acting on research institutes. Just as a population will breed up to the available food supply, research institutes will enlarge themselves as long as the money holds out.20

We may proceed now from research proper to the field of development. It has been argued that, while small scale basic research may continue to be important, the cost of developing already-created inventions is growing ever-greater, and is therefore peculiarly susceptible of large-scale organized and directed effort. Most of the technological work in the industrial laboratories, indeed, is the actual development of new methods and products, while university and other educational laboratories have relatively concentrated on pure research.

Development costs have grown more expensive especially in the chemical industries, where a new idea is taken and run through very large-scale empirical experimentation (e.g. the trial-and error searching for a better strain of penicillin among a large number of possible molds). Increased caution in developing products, greater testing for quality and safety, a heavy initial advertising campaign to introduce new products—all these factors have increased the costs of development in modern times (although, with technological advance cheapening everything else, we may expect it to lower costs of development as well).

But a crucial point about development has been often overlooked: how many resources to put into development as against other things, how fast to develop at any given time, is a risky decision on the part of a firm. The decision depends upon the firm’s estimates of future costs, sales, profits, etc. Government, crippling or eliminating the free market signals of prices and costs, would be lost without a guide to efficiency or allocation of resources. Further, the main reason deciding a firm to devote its resources in an attempt at speedy development is the spur of competition. And competition means the free, unhampered market. Even in the case of Nylon, the most cited example on behalf of large-scale monopoly research and development, DuPont had the competitive spur of knowing that German scientists were also working on similar synthetic fibres.

Where the competitive spur is weak, or especially non-existent (as in government), development will be slowed down. Furthermore, the existence of many firms, many centers of development, make it far more likely that new ideas will obtain a hearing and a trial somewhere. General Electric, when dominant in lighting, was sluggish in developing fluorescent lighting, but once other firms entered the field, it sprang to life and regained a dominant position through its newfound efficiency. As Jewkes and his associates sum up:

Against the claim that the prerogative in development should always rest with the biggest and the most securely established industrial organizations, may be set, therefore, the advantages of the attack from many angles. The tasks of development are themselves of such diversity and of so varying a scale that it may be a … dangerous oversimplification to suppose that they can always be best handled by any single type of institution.21

The best condition, they add, is a variety of firms, in size and in outlook—some bold and others cautious, some leading and others following.

Even in the field of development proper, in fact, many important new products have come from small-sized firms, or even individuals. These include: air conditioning, automatic transmissions, bakelite, cellophane tape, magnetic recording, quick freezing, power steering, crease-resistant textiles, and ram-jet aircraft.

Professor Baker has preferred another important refutation of the statist claim that governmental monopoly direction of research would eliminate “wasteful overlapping” of effort. Baker points to the enormous importance for scientists, in having two or more mutually independent scientists or laboratories confirming each other’s conclusions. Only then can the world of science consider the experiment truly confirmed.22

Soviet Science

“Planned” science sounds impressive; actually it means prohibited science, where no scientist can follow the leads of his own creative ideas. We have heard a great deal recently about the alleged glories of Soviet science, and about the necessity of the United States catching up with such wonders as sputniks. What is the real record of Soviet science? Professor Baker, analyzing this record, shows that, at the beginnings of the Soviet Union, the old pre-revolutionary scientists continued to do well, largely because science was not yet under government planning. That came with the Second Five-Year Plan, in 1932. The Plan set forth very broad subjects for investigation, but, by the nature of such a plan, many important areas were excluded from the required agenda. “Take almost any branch of non-revolutionary biological science in which outstanding discoveries were made in the outside world during the years of the plan, and you are likely to find that the whole subject was excluded from study.”23  For example; the study of hormones, and genetics. The Lysenko controversy, the use of the State to eradicate the science of genetics in Soviet Russia, and the compulsory twisting of truth by the Soviet State to fit the ideological myths of its rulers, are well-known, but can hardly be overstressed. It is important to realize that it is not simply because the Soviet or Nazi leaders were particularly perverse men that they reached out to prevent or cripple science’s drive for truth; but because such actions are inherent in the very nature of statism, and central planning. Power, and its promotion, advancement of the ideology of power, become the highest social goal, before which all truth, all integrity must give way.

Government control of science, government planning of science, is bound to result in the politicization of science, the substitution of political goals and political criteria for scientific ones. Even pro-Soviet scientists have admitted that Soviet research is inferior to American, that basic, as contrasted to applied, research, is neglected; that there is too much red tape; that little fundamentally creative work has been done; and that science is unduly governed by political considerations—such as the political views of the scientist propounding any given theory. Scientists are shot for taking the view that happens to be in political disfavor. And, as Baker concludes: “If the selection of scientific personnel is left to the State, the wrong ones are likely to be given important posts, because those who are not themselves scientists will be let astray by … false claims and pretences … (and) scientists may exhibit a servile obedience to their political bosses.”24  No wonder that in a list, drawn up by seven scientists, of the two dozen most important scientific discoveries made between World Wars I and II, not one came from the U.S.S.R.

• 1In 1953–54, the Federal government spent $2.81 billion of its funds on scientific research and development; of this amount, only $970 million was spent on programs within the government itself (and most of this was development rather than research); the remainder was channeled into private hands to pay for privately-conducted research ($1.5 billion in industry, $280 million in colleges).
• 2See Basic Research, A National Resource (Washington, D.C.: National Science Foundation, 1957); and John Steelman, Science and Public Policy (Washington, D.C., 1947).
• 3Brig. Gen. David Sarnoff, Research and Industry: Partners in Progress (Address, Nov. 14, 1951), pp. 6–7.
• 4Sarnoff, op. cit., pp. 12 ff.
• 5John Jewkes, David Sawers, and Richard Stillerman, The Sources of Invention (New York: St. Martin’s Press, 1958).
• 6Typical recent expressions of the myth may be found in John Kenneth Galbraith, American Capitalism; W. Rupert Maclaurin, “The Sequence from Invention to Innovation”, Quarterly Journal of Economics, Feb. 1953; Waldemar B. Kaempffert, Invention and Society; A. Coblenz and H.L. Owens, Transistors: Theory and Application.
• 7For other experts who believe that a highly important role still remains for the individual independent inventor, see Joseph Rossman, The Psychology of the Inventor; the late Charles F. Kettering, New York Times, March 12, 1950; W.J. Kroll (the inventor of ductile titanium), “How Commercial Titanium and Zirconium Were Born.” Journal of the Franklin Institute, Sept. 1955; and H.S. Hatfield, The Inventor and His World.
• 8R.M. Lodge, Economic Factors in Planning of Research, 1954. Quoted in Jewkes, et. al., p. 133. On other cases of great scientists preferring simple equipment, see: John Randal Baker, The Scientific Life, P. Freedman, The Principles of Scientific Research, J.B.S. Haldane, Science Advances.
• 9Jewkes, et. al., p. 116.
• 10John Randall Baker, Science and the Planned State (New York: Macmillan Co., 1945), p. 42.
• 11A. Szent-Gyorgyi, “Science Needs Freedom.” World Digest Vol. 55 (1943), p. 50.
• 12See Baker, op. cit., pp. 49–52. Baker comments on the lack of originality of research teams, who tend to be better at following up the leads of others than at originating ideas themselves.
• 13“Our modern knowledge of how to control cancer of the prostate is due to the researches of these men—of Hunter, Gruber, Griffiths, Steinach, and Kun on the prostate; of Grosses, Rusler, Davis, Baaman, and Riedell on phosphatase; and of Kutcher and Wolbergs on phosphatase in the prostate. Not one of these men was studying cancer, yet without them, the discovery of the new treatment could not have been made . . . what central planner, interested in the cure of cancer, would have supported Griffiths in his studies on the seasonal cycle of the hedgehog, or Grosser and Husler in their biochemical work on the lining membrane of the intestine? How could anyone have connected phosphatase with cancer, when the existence of phosphatase was unknown? And while it was yet unknown, how could the man in charge of the cancer funds know to whom to give the money for research? No planner could make the right guesses.” Baker, op.cit., pp. 59–60.
• 14On the inevitable clash between research administrators and scientists, see: Jewkes, et. al., pp. 132 ff.; K. Ziegler, The Indivisibility of Research, 1955, S.C. Harland, “Recent Progress in the Breeding of Cotton for Quality.” Journal of the Textile Institute (Great Britain), Feb. 1955; R.N. Anthony, Management Controls in Industrial Research Organization.
• 15From L.J. Ludovivi, Fleming, Discoverer of Penicillin, cited in Jewkes, et.al.
• 16Jewkes, et.al., pp. 141–42.
• 16This is borne out in separate studies by the U.S. Department of Labor, Scientific Research and Development in American Industry, Bulletin #1148, Wash. 1953; and the National Association of Manufacturers, Trends in Industrial Research and Patent Practices.
• 18Jewkes, et. al., pp. 188–89.
• 19P.H. Frankel, Essentials of Petroleum, 1946, p. 148. Quoted in Jewkes, et. al.
• 20Harland, loc. cit. Also see Laura Fermi, Atoms in the Family, p. 185. Quoted in Jewkes, et. al., p. 162.
• 21Jewkes, et. al., p. 222.
• 22There is one occurrence . . . which helps the scientist form a valid judgment better than anything else. This is the . . . publication of the same result by two entirely independent workers. Central planners are inclined to consider that one of the two independent workers has been wasting his time. The actual research worker knows that this is not so. It is the very fact that the two workers are independent that inclines others to accept their findings. Scarcely a working scientist will deny that two independent papers containing the same result are very much more convincing than a single paper by two collaborators . . . (also) each paper has a different outlook, and the reading of the two papers is far more stimulating and suggestive.” Baker, op. cit., p. 49.
• 23Baker, op. cit., pp. 66ff.
• 24Baker, op. cit., pp. 75–76.