Ruttan (2006) argues that large-scale and long-term government investment has been the engine behind almost every GPT [General Purpose Technologies] in the last century. He analysed the development of six different technology complexes (the US ‘mass production’ system, aviation technologies, space technologies, information technology, Internet technologies and nuclear power) and concluded that government investments have been important in bringing these new technologies into being.

Thus argued Mazzucato in her book. So I thought it would be interesting to read this book titled "Is war necessary for economic growth?", and so I read some of it. So far I've read the introduction, the case of the aircraft industry and the conclusion (ch.1,3 and 8). And yay! This book is like two orders of magnitude better than Mazzucato's. You have wide surveys of historical facts, references to studies on the impact of military R&D on productivity and a discussion of the evolution of military research policy.

I will comment here some things from the book's overall arguments, and from the chapter on the aircraft industry in particular.


So first, the stated purpose of the book is

to demonstrate that military and defense-related procurement has been a major source of technology development across a broad spectrum of industries that account for an important share of U.S. industrial production.

He begins his introduction by noting the important role  that military procurement has played in technological development, including the Industrial Revolution (John Wilkinson's cannon-borer applied to condenser cylinders). He then presents the opposite view that arose during the Cold War, that defense related R&D detracted resources from commercial application, and that slowed productivity and technical change. He also points out that there was a historical tendency for defense contractor to insulate their military research from the commercial one, and so one didn't just help the other.

Ruttan says that he is interested in all of this

to the extent that it contributes to commercial technology development. And I am concerned with the impact of military and defense-related technology development and procurement on commercial technology development primarily during the early stages in which commercial technology has most typically drawn on military and defense related research, technology development, and procurement

We then have a discussion of internal and external interpretations of advances in science and technology. Internal explanations would talk about the motives of individual scientist and engineers, culture of scientific and engineering societies, instead that of changes in their environments ranging from natural resources availability to relative factor prices, which would be external explanations. But he points out that this explanations became less compelling as a sizable fraction of research began to be carried out in governmental and industrial laboratories. Nonetheless, this explanations still have strong backers in military history. My own view is close to the internal interpretation, allowing for external factors that modulate the problems and challenges innovators think is worthy to try to think about and solve.

The History of Flight: Military and Commercial Aircraft

The chapter begins from the first successful sustained flight (12 seconds of flight) of a heavier than air self powered flying machine, by the Wright Brothers, in 1903, and their later developments in 1905 (a longer than half an hour flight), 1908 (more than two hours of flight), and the US Army Signal Corps acquisition of its first airplane from them that same year.

Now let's pause here from a second and bring in John D. Anderson's Introduction to Flight, an aerospace engineering textbook in which we can find another account of the early history of aviation that predates the Wrights. Ruttan probably knows this because of the long sausage of adjectives that qualify "flying machine".

As we can read in Anderson

Contrary to some popular belief, the Wright brothers did not truly invent the airplane; rather, they represent the fruition of a century's worth of prior aeronautical research and development. The time was ripe for the attainment of powered flight at the beginning of the 20th century. The Wright brothers' ingenuity, dedication, and persistence earned them the distinction of being first.

Just for a quick overview:

  • Early interest in flight, as seen in the Greek myth of Daedalus and Icarus. All early thinking dominated by the most reasonable sounding way of achieving flight: doing what the fliers do, meaning copying birds.
  • So we get ornithopters, the most famous of which probably being Da Vinci's machines at the end of the 15th century. But all this birdy efforts failed George_Cayley Figure 1: George Cayley, aeronautic badass
  • Then we have mankind's first men who lifted off the ground: Pilatre de Rozier and the Marquis d'Arlandes, in 1783, in a hot air balloon built by the Montgolfier brothers. The idea was conceived the previous year when Joseph Montgolfier was staring into his fireplace, and first did some balloon experiments with animals. This development was followed by a hydrogen filled balloon by J.A.C. Charles, a physicist. Balloons were the only means of human flight for almost 100 years!
  • We get to Sir George Cayley, true inventor of the airplane according to Anderson. He conceived his design in 1799, and included a fixed wing for generating lift, a propulsion mechanism (paddles) and a cruciform tail for stability. This was inscribed in a silver disc that also contained a lift and drag forces diagram, showing that he understood the forces acting on his wing. The breakthrough of Cayley was to abandon the bird-like paradigm, separating propulsion and lift. He didn't stop there. He also built a whirling-arm device to test his airfoils. Basically a precursor to modern wind tunnels: the idea is to have an airfoil on one side of the arm, a counterweight at the other and a point of support in the middle, allowing the contraption to rotate. Rotation would induce relative motion between the airfoil and the air, thus simulating the behaviour of the airfoil when flying. He also built and flew a glider with wings and a tail and wrote a paper in 1810 titled On Aerial Navigation, published in the Journal of Natural Philosophy, where he stated, among other things, the principle that makes wings work: an induced region of lower pressure at the top of the wing causing an upward force. He was called the father of aviaton by W.S. Henson in 1846, but then his name disappeared and his work became obscure to basically all future aviation inventors. He came back from the oblivion grave via C.H. Gibbs-Smith work, who wrote Sir George Cayley's Aeronautics in 1962. He says that if Cayley's breakthroughs had been continued, powered flight would have occurred in the 1890s instead. Twenty years earlier than the Wrights'!
  • After Cayley, 50 years of no major advances, but there were still people trying to conquer the air. Some names in this period are W.S. Henson, John Stringfellow, Felix Du Temple, and Alexander F. Mozhasiski. Mozhasiski built a steam powered monoplane and achieved some seconds of flights, but having made use of an assisted takeoff.
  • Also, in 1866, the Aeronautical Society of Great Britain is constituted to encourage the advancement of the field of aeronautics. This inspired the American Rocket Society, and the Institute of Aeronautical Sciences, which merged in 1964 to form the AIAA, a channel for sectoral information exchange.
  • Jump to 1891 and we have Otto Lilienthal glider: the first controlled glider in history. He made interesting advances, but suddenly died when flying one of his gliders.
  • Then Percy Pilcher, a student of Lilienthal who also went into glider construction, and calculated what engine wold be required to enable powered flight, but such engine was not available, so he decided to build one. But he died in the same way as his teacher, and he wasn't able to use the engine. According to Anderson, had Lilenthal or Pilcher lived for longer, they may have beaten the Wrights to the first powered flight.
  • The action now moves to the US, where French-born Octave Chanute wrote a book compiling everything that was known so far about aeronautics: Progress in Flying Machines (1894). The Wright brothers read this book, and established a relationship with Chanute until his death in 1910.
  • We then have Samuel Pierpont Langley, secretary of the Smithsonian, who attempted to do what the Wrights did, and achieving it just weeks after the Wrights. Like other names mentioned in this list, he had no formal education. In 1898, with the coming of the Spanish-American War, the War Department invited Langley to build a machine for passengers (they probably had other intentions in mind too). They granted Langley $50,000 so that he could continue his research. He designed an aircraft not unlike the one of the Wrights', but he was unsuccessful (That is, it crashed) and the War Department gave up, stating "were are still far from the ultimate goal of human flight". This kinds of remarks are typical of scientists, and it would be interesting to make a compilation of them, I say. Quite a lot of times a top scientist has declared something impossible or extremely hard, and time has proven him wrong.  Unfortunately for Langley, he died unjustly ridiculed for his failures in 1906, but his achievements provided encouragement to others.
  • And then we have the Wrights, in 1903.
  • A final remark of Anderson's is that there was a gap between the state of aircraft manufacturing and engine manufacturing. The early designers only had steam engines at their disposal, and combustion engines wouldn't appear until Lenoir's 1860 engine. Then we have Otto and Langen, and developments in the automobile industry that finally rendered an engine with a sufficient weight to thrust ratio which was appropriate for flight.

Two comments can be made here. One, that Cayley serves as a sort of test of the Great Man Theory of Innovation. He was that Great Man of his time, and yet innovation continued. No "Without Cayley there wouldn't have been aircraft" nonsense. On the other side, his death could have meant a twenty years delay on the industry, so this is some evidence that individuals do matter for when an invention is developed, or instead, that diffusion of information matters. Had others known about Cayley, maybe advances could have been made faster. Other event is the Langley-Wright race for the first flight, and how it would be an easy move to claim that Langley was crucial, or that the government invented the airplane had Langley succeeded, which would have forced me today to talk about a counterfactual where the Wrights succeeded instead of Langley. But by now we know how that trick works.

Back to Ruttan.

He mentions advances done in fluid dynamics and aerodynamic theory in the XIX century, mentioning people such as Wilhelm Kutta, Nikolai Joukowski, Fredrick Lancaster and Ludwig Prandtl of Prandtl's Number fame. Prandtl was a professor at Gottingen University, where his team had a wind tunnel that enabled the university to become the top aerodynamic research site.

Figure 2: Ludwig Prandtl, Boss of Bosses

Also, craftsmen and engineers were also working to understand the laws of flight to build successful flying machines, but, notes Ruttan, there was little contact between scientific knowledge and engineering and mechanical practise. The Wrights drew mostly on craft and engineering knowledge and practise. Early flight owed nothing to fluid dynamics, that is. Most of advances in aircraft design after the Wrights were made in Europe. Why? According to Ruttan, because the growth of the industry was tied to military procurement during the war and in anticipation to it.

World War I

The aircraft of WWI, however, were just evolutions of the original Wright Flyer, and the only revolutionary design was the Junkers J-1, entirely built with electrical steel (the one used in transformers. Duraluminum, a better material, was too recent and still had problems).

World War I saw a surge in military aircraft demand, and when the war ended, a subsequent drop happened. The US industry went from 14020 (13991 military) aircraft in 1918 to 328 (256 military) in 1920. According to Ruttan, it was the Army and Post Office demand for aircraft that prevented complete industrial collapse in the sector. Here are some nice looking charts I made from Ruttan's source, Historical Statistics of the US. Colonial Times to 1970, Part 2 . Things to note in the chart: the surge of manufacturing in WWI, that didn't affect the level of production of civil aircraft, the productive boom in the late twenties and the 1929 crash, followed by a timid recovery, and an implosion in WWII, along with a surge of military aircraft. After WWII, a boom in civil aircraft production (Because those factories had to so something), and then a decline and slow growth. This corroborates Ruttan's thesis that the Post Office's Airmail did reinvigorate the industry, as passenger airlines were still not profitable. The situation in Europe was similar, so it is not implausible to infer that, absent subsidies, commercial aviation would have taken off  much later. However, a case for the contrary is laid out in a later section of this article.

Figure 3: Number of aircraft manufactured (1913-1936)

Figure 4: Number of aircraft manufactured (1937-1961)


We now enter a section titled the NACA era about, well, the NACA, the National Advisory Committee on Aeronautics. Says the book that in 1911, the Aeronautical Society started efforts to obtain political support to establish a national aeronautical research lab, but it didn't happen until World War I, when political, military, commercial and scientific interests united and a bill was passed to establish NACA, initially serving as a consultive board. Its first task was to survey the state of research in American universities and industry.

From the end of WWI, NACA dedicated its efforts to research, and produced interesting results: analyses of propeller designs, and the construction of advanced wind tunnels. Their chief scientist, Max Munk -an student of Prandtl- also worked on airfoil design, producing a series of airfoils.

In 1927, with the appointment of Joseph Ames (From John Hopkins University) as NACA chairman, the agency expanded in budget and facilities, and tried to be more responsive to the needs of the civil sector.

One of the advances of this period is the NACA Cowling, a streamlined cover for aircraft engines to reduce their aerodynamic drag (60% less). The project began in mid-1920s with a meeting between industry representatives, the military, and NACA. They didn't know how the Cowling worked yet, as the work was purely empirical, and a theory had to wait until Theodore Theodorson, NACA head of their Langley facility put one forward. Other NACA facilities were built as time passed: in Moffett Field, another near Cleveland, and another in Edwards (California).

Most of NACA research before WWII was dual use technology: with both military and civil applications. With NACA, the US had went from having an industry in a sorry state to having the world's best airliners, such as the Douglas DC-3. However, says Ruttan,

the aircraft and airline industries weren't much able to achieve a stable economic structure, because many of the firms were established or financed by wealthy hobbyists, designers, and entrepreneurs "whose enthusiasm for aircraft defied rational business calculation" (Vander Meulen 1992, p.57)

So they weren't big enough to conduct independent research according to him.

To finish the section, he stresses that military considerations were a primary motivation for the establishment of NACA and support of its R&D. Demand for military aircraft supposed an important inducement for advances in airplane design.

Before moving on to jet engines, let's fill in some details from another source, Engineer in Charge: A History of the Langley Aeronautical Laboratory, 1917-1958 by James R. Hansen (1986).

Birth of NACA

Initially, government officials opposed the NACA. David W. Taylor from the Navy said that his own facilities were already doing aeronautical research, and opening a civilian laboratory would only duplicate efforts. On the other hand, Richard C. Maclaurin, president of MIT, argued that the lab should be located at or near a university like in Europe. Meaning the MIT. Samuel Stratton, director of the National Bureau of Standards also dismissed the idea, saying that his Bureau could manage to do the research. There was a a feeling that although there was some progress in the field of aeronautics, there wasn't enough. Langley Pierpont's lab was closed after his disastrous attempts to fly. Albert F. Zahm's wind tunnel, in Catholic University in Washington, D.C. was also closed for lack of money. All of this while Europe was heavily investing in aeronautical research.

It is finally in 1915, after

active politicking by dedicated, well-connected scholars and government officers to grease the bureaucratic machinery for the creation of a new federal agency devoted to advancing the state of the art in aircraft design and operation.

that the NACA is born as a committee with members from the War Department, the Navy, the Smithsonian, the Bureau of Standards, and people from industry. The initial idea was

A Main Committee, composed of seven government and five private members, would meet in Washington, D.C., semiannually-and occasionally more often-to identify key research problems to be tackled by the agency and to facilitate the exchange of information within the American aeronautical community. This body would be independent, not under any department, but reporting directly to the President, who appointed its members. These members would receive no salaries.

[...] Military officers would arrive with ideas provoked by recent intelligence reports, university professors with word of yet-to-be-published papers announcing new theories or experimental findings. By pooling their information, which was already high-level, and as a group assessing its practical effects for American aeronautics, NACA members would produce a new body of knowledge even better than the sum of its parts. They could then translate this knowledge into effective technical advice and wise government research policy. In this synthesis, many would say, lay the genius of the committee system.

Jet Engines

Now let's continue with Ruttan's account of Jet Engines

Aerodynamic theory advanced, but it wasn't being applied to aircraft. Even then, aircraft were flying faster and higher. But by mid 1930, the NACA conferences that gathered industrial representatives showed a shrinking gap between theory and practise. These advances, however, were making clear that piston-propeller propulsion was a dead end, and that it was very difficult to keep increasing performance within this paradigm. The way out were jet engine, but NACA had assessed them in 1923, and wrote a negative report, saying that fuel consumption and maintenance costs were too high.

At some point, Frank Whittle in the UK and Hans von Ohain in Germany initiate jet engine development with their own resources.

[28-5.jpg?w=300) Figure 5. Whittle's jet engine. Working model in Cranfield University

Whittle received British Air Ministry support, after they had judged his proposal as too radical. After he had a prototype developed with his and some ex-RAF officers' funds, and having formed a company by the name of Power Jets with the help from what we could call early venture capitalists (O.T. Falk & Partners), the Air Ministry granted him some money to advance his research, and then received a contract to build an engine. According to Ruttan, the Air Ministry was reluctant to give him support because that would detract funds from piston-propeller engine research. He also didn't manage to get support from established companies, as British-Thomson-Houston didn't grant him the 60,000 pounds he requested to develop his ideas.

von Ohain, on the other hand, got support from manufacturer Ernst Heinkel. He was a student from Göttingen University, and had received funds from Siemens after he sold a patent to Siemens. He then built a prototype engine in his garage, that tended to burn out due to using low quality metals. But this impressed his professor, Robert Pohl, who wrote to Ernst Heinkel, telling him about the engine. He managed to convince him, and as an end result, the Heinkel HeS 1 engine was born. Then, the Reich Air Ministry took an interest in the engine, and contracted Heinkel to build engines for military aircraft.

NACA wasn't interested because of war-induced conservatism. This was also the reason almost all aircraft manufactured in the US during WWII were based on pre-war designs. In the end, they changed their minds after General Hap Arnold visited the UK, saw a Whittle engine, and wrote to Vannevar Bush, director of NACA, asking for funds to do jet engine research. In the end, they had to bring Whittle himself to the US and other British engineers to help build an engine in the US.

Further interest into jet propulsion came when, during the Korean War, American bombers were easy targets for Russian jet fighters. The company today we know as Boeing built under contract for the US Airforce an experimental jet bomber, the XB47. This design was the basis for the future 707 airliner.

Commercial jet aviation progressed in Europe and thus we see the Vickers Viscount and de Havilland Comet aircraft being relied on for international flight.

To finish this section, a summary of other NACA achievements: thermal deicing using exhaust gases, refinement of prototypes, and consulting work for aircraft manufacturers.

The GALCIT-NACA rivalry

In the next page we find an explanation of the changing role of NACA through the years, and there we also find the existence of the Daniel Guggenheim Fund for the Promotion of Aeronautics, who made endowments to Caltech (GALCIT), the Georgia School of Technology, the MIT, NYU, Stanford, Michigan University, and the University of Washington.

One of this organisations, GALCIT (Guggenheim Aeronautical Laboratory of the California Institute of Technology, that later became the Jet Propulsion Laboratory), was of particular interest, because according to Model Research: The NACA 1915-1958 (ch. 7) aircraft manufacturers located in California preferred to rely on GALCIT for their research needs instead of NACA, located in the east coast. GALCIT was so successful that they couldn't keep up with manufacturers requests, so they asked the government for extra funds, backed by the manufacturers. NACA wasn't amused with GALCIT's influence

The real threat from Caltech, however, was territorial and functional, and it ran to the very heart of the NACA's justification for continued existence. Understanding it requires a perspective not only on the specific issue of developmental wind tunnels for Caltech but also on the division of research roles in the United States, not just in aeronautics, but in all branches of science and technology. A. Hunter  Dupree has described the situation in the 20th century in the same terms used by Millikan (GALCIT's chief):

In 1900 the universities, grown in one generation from colleges with narrow courses of studies, seemed to have become the natural homes of disinterested, pure science. The broadening of the curriculum, the introduction of' the German seminar and its ideal of research, the creation of graduate schools, and the rapid accumulation of endowment either created new centers of learning or remade old ones. With Johns Hopkins setting the pace, such universities as Harvard, Cornell, Chicago, Columbia, and Michigan became the headquarters of fundamental research in the country.

The result was a division of labor which gave rise to the assumption that basic research belonged to the universities, leaving only applied research to the government. The difference heightened between the disinterested, cloistered seeker for pure knowledge and the grubby civil servant chained to the mundane, grinding routine investigation. Although the split between basic research and the common concerns of society was noticeable fairly earl), in the nineteenth century, after 1900 it became institutionalized in the division of functions between government and the universities.

You can picture this rivalry as one between two students of Prandtl (NACA's Munk and GALCIT's von Karman) and the two organisations themselves. NACA didn't want to disappear:

Now Millikan, by raising the dichotomy again, was endangering the NACA position in two ways. First, by ascribing basic research to the NACA, he was associating the Committee with the universities at one end of the research spectrum, separating the NACA more clearly than it wanted from the applied research that he left to the services and industry. Second, Millikan was proposing that the government help Caltech move into applied research to assist the west coast manufacturers. If that were done, what part of the research spectrum could the NACA call its own? It had always been willing, at least after Munk's departure, to concede to the universities an edge in theoretical work, retaining for itself the incomparable wind tunnels needed to convert that theory into fundamental data useful to the military and industry. If the universities started building similar tunnels with government funds, Congress would soon cry duplication.

To make matters worse, GALCIT was then run by Theodore von Karman, who revealed another chink in the NACA armor. Von Karman was a brilliant aerodynamicist whose career bore striking similarities to and sad contrasts with that of Max Munk. Both had been students and protégés of Ludwig Prandtl, and both possessed the rare ability to comprehend aerodynamics in the abstract and to apply that insight in fruitful experiments and techniques. Though both had been trained in engineering, their real strength lay in theoretical insight that informed and directed their research. It was for these research gifts that both were brought to the United States, Munk to the NACA in 1921 and von Karman to GALCIT in 1930. While Munk's prestige deteriorated after he left the NACA, von Karman's grew through years of productive teaching and research at GALCIT, culminating in election to the National Academy of Sciences and undisputed recognition as dean of American aerodynamics. Of course, all good aeronautical research - whether done in the laboratories of GALCIT, the NACA, the military services, or the industry - required ad hoc mixing of theory, experimentation, testing, and ingenuity, and no institution had a monopoly on any of these ingredients. Still, for the NACA to agree to place government funded research tools in von Karman's hands was to arm a rival and loose him in a field the NACA meant to command.

Going back to GALCIT's proposal to the government of a grant to build wind tunnels (which was backed by the manufacturers), we find out this seemed like an attempt by them to just obtain some benefits from government, as they in the end were perfectly capable of building wind tunnels:

here were signs, however, of changes to come. When the Millikan proposal failed to win army support, Congressman Carl Hinshaw (whose district included Caltech) introduced a bill to fund a Caltech wind tunnel. Commenting on this proposal, Jerome Hunsaker reported that Caltech was appealing to the government only because the manufacturers in southern California were unwilling to support the tunnel themselves, even though they were to be the main beneficiaries. They were happy to endorse proposals to build tunnels at government expense but - unlike manufacturers in other parts of the country - less willing to use their own funds. When forced to it, however, they later began to build their own wind tunnels rather than share university facilities and staffs with competitors. In time, both government and industry would contribute to university wind tunnels.

Also note that, if you read the complete story, rivalry is not a completely accurate word. Guggenheim, founder of GALCIT, was one of the early pushers of NACA, and Millikan, head of GALCIT tried to avoid giving the government the impression that NACA was unnecessary.

There's then a discussion of Boeing and its relation to the military, and how later on NACA and other centres joined to form NASA.

Chapter Conclusion

As a conclusion of the chapter, we have a commencing paragraph that would seem out of place in Mazzucato's book:

The private sector played a primary role in the initial development of aircraft. The Wright brothers self-financed the invention and initial development of the first successful propeller-driven aircraft. During the next decade design and technical improvements were made primarily by flight enthusiasts, craftsmen, and engineers. World War I played a major role in inducing the development of an aircraft industry in both Europe and the United States, but it was not, with the exception of the all-metal airplane, productive of revolutionary developments in aircraft technology and design. The airplanes that flew during World War I were evolutionary from the Wright Flyer.

I would dispute that WWI itself developed the industry itself, because as we have seen, it didn't have an impact on aircraft manufacturing. It did impact the long run industry, though, via forcing the creation of NACA.

He concedes that it is an open question whether without NACA, the industry would have developed as fast, but he favours the interpretation that it would have slowed down the pace of advance:

It is hard to avoid the conclusion that, in the interwar period, commercial aircraft would have been developed and introduced more slowly in the absence of defense-related technology development and military procurement. In the United States, NACA played a particularly important role in the development of dual-use technology applicable to both military and commercial aircraft. It is an open question whether, in the absence of NACA research and military procurement, propeller-driven commercial aircraft would have achieved the level of technological maturity represented by the DC-3 by the mid-1930s (Mowery 2004).

However, he then cites a 1981 book that comes to the conclusion that

Without federal government [support] there would simply be no aircraft industry despite the fact that the commercial market is playing a much larger role than it has in the past. No aspect of the industry, including the commercial sector, could exist without the R&D funds provided by the state, or the state’s purchase of military equipment. (Bluestone, Jordan and Sullivan 1981)

This refers not to the historical industry, but to the contemporary one. That is, this section seems to be implying that if state involvement in the aeronautical sector dropped to zero, the sector itself would disappear. This doesn't seem plausible to me, but it can be discussed.

Finally, the last paragraph remarks that without military intervention, the industry wouldn't have progressed as fast as it did. Recall, Mazzucato speaks in term of no progress, while Ruttan speaks in terms of slower progress.

What would a counterfactual look like? Imagine no World Wars, no Air Mail subsidies, and no NACA. Scientists working for NACA would be elsewhere. Without military pressure for NACA, GALCIT or other Guggenheim funded colleges would have taken the throne of top aircraft research centre. Aircraft manufacturing would have remained low, as research advanced. Eventually, advances in GALCIT, perhaps even funded by manufacturers, would have enabled the construction of profitable aircraft. Following the same route, jet engines would have been integrated at some point. So I am willing to grant that NACA support was very important, and that it could have accelerated the pace of development, but that mass manufacturing of aircraft for the war did not. I thus conclude here that war was not necessary for the appearance and growth of the aircraft industry, but that NACA could have accelerated it. NACA isn't an example of state entrepreneurship, as apparently the ones pushing the case for NACA forward were the Aeronautical Society, who just wanted to see more progress in things they liked, and the initial reaction of government officials wasn't much appreciative.

However, NACA looks like a prototype for what successful government intervention could look like: A research laboratory in touch with industry, doing research as requested by them. This model is similar to Germany's Fraunhofer institution.

Book conclusion: Is war necessary?

The question Ruttan set out to answer is this. Not whether the government had played an important role in technological development, as that was an uncontroversial conclusion in an earlier book of his. He notes that for the cases reviewed (I have only looked into the case of aviation), absence of government intervention would have delayed their commercial development, and in one case (nuclear power), it would have made it impossible.

There are two charts with the evolution of total R&D spending in the US, finishing in 2001. I provide here an updated version, sourcing the data from the NSF.

t2p1 Figure 6: US R&D expenditures

Figure 7: Funding sources for US basic research

Ruttan then makes a concession: massive military R&D programs are not to be evaluated in terms of their impact on commercial technology. They are to be evaluated on their impact on national security. He also says that he's not defending the case that military research spinoffs are an efficient way to advance commercial technological development, granting that much of military spending is absurdly expensive, without taking into account cost-benefit analysis in most cases. His thesis, instead, is that the world would look differently absent all this spending.

He now reviews a series of military innovation policies:


During the first two post–World War II decades, the spin-off issue attracted relatively little attention. It was generally taken as self-evident that substantial spinoffs of commercial technology could be expected from military procurement and defense-related R&D. It was also generally assumed that there was no need for policies to enhance the development of dual-use technology [...] The slowdown in the rate of economic growth in the United States after the early 1970s generated considerable controversy about the role of the military in technology development.  Some critics had argued, even in the 1960s, that defense related R&D was becoming a burden on economic growth Military and space research was viewed as drawing scientific and technological capacity away from civilian applications. It was argued that the effect was to slow the advance of industrial technologies and reduce the rate of economic growth (chapter 1; Solo 1962, pp. 49–60; Kaldor 1981; Lichtenberg 1984; Dumas 1986; Lichtenberg 1989). In addition, defense industry firms, even those with substantial commercial activity, often failed to take advantage of technology transfer opportunities from their military R&D. Murray Weidenbaum (1992) has observed that “those defense firms that do operate in civilian markets often tend to maintain operationally separated insulated divisions that have little contact with each other, merely reporting to the same top manager” (p. 51).


The linear model was deeply embedded in military procurement practice. Procurement followed a “pipeline” progression.


By the mid-1980s this process was beginning to appear increasingly incongruent with R&D practice in the most advanced sectors of the commercial economy. The postwar U.S. economy had witnessed an accelerating transformation of the relationships among science, technology, and production. It became widely recognized that commercial production processes pressed more immediately, and sometimes beyond, advances in scientific knowledge (Alic et al. 1992, pp. 11–22).

A literature review

We are then presented, in page 182 with a box titled Military R&D: The Productivity Puzzle, in which he reviews evidence about that, opening with a very direct question: Is publicly funded military research and development a source of technological development and productivity growth? One could be inclined to say so, given what he has argued for, but

The results seem to be in direct contradiction to a number of very careful econometric studies that show that measured private and social rates of return to military R&D have been very low and have had no discernible effect on industrial productivity growth in the United States (Lichtenberg 1984; Lichtenberg 1988; Lichtenberg 1989, p. 275).

He then reviews the classic consensus that the social rate of return to R&D in general is higher than the private one, thus leading to the conclusion that there was underinvestment in R&D

The results of the large body of firm-level, sector level, and economy-wide studies, combined with studies of the sources of productivity growth, have supported a view that the social rates of return to R&D have generally exceeded the rates of return on almost any other form of investment available to the U.S. economy. These high social rates of return contributed to a consensus that the United States was substantially underinvesting in R&D—and that this underinvestment was a substantial constraint on economic growth. Because of the spillover of R&D benefits in the form of consumers and producers surplus, even privately funded R&D shared the characteristics of public goods—the economic unit that generates the new technology can capture only a portion of the social benefits deriving from the research. The policy implication that has generally been drawn is that the United States should expand public-sector support for R&D to correct private-sector underinvestment.


The generality of this conclusion has been challenged, however, by studies by Lichtenberg (1984; 1988; 1989) and others that have attempted to measure the private rates of return to firms that conduct publicly funded research and technology development and of the firm and economy-wide spillover effects of such R&D. A large number of studies have failed to find significant private or social rates of return from publicly funded research conducted by private firms. However, “privately funded R&D in manufacturing industries is found to yield a substantial premium over the rate of return from ‘own productivity improvements’ derived from R&D performed with government funding” (David, Hall, and Toole 2000, p. 498). It has been suggested that one explanation for these results may be that a high percentage of firm-level federally funded industrial research has been conducted by defense or defense-related firms. Neither the R&D nor the products resulting from such R&D are subject to a market test. The design of technology, firm-level costs, and returns are heavily influenced by bilateral bargaining. Research results and technology development information are usually classified. A substantial share of the products derived from the federally funded R&D are often sold back to the government. Under these institutional arrangements conventional measures of profitability and productivity may not be appropriate (Griliches 1995, p. 82).

He then says that profitability analyses are not fully adequate, because public R&D can act as complementary to private R&D, but there is also evidence for public R&D could also crowd-out private R&D:

A number of early studies contributed to a presumption that much of the R&D conducted or funded by the public sector simply substituted for private sector R&D—that it crowded out private-sector R&D. In an attempt to test the substitution hypothesis, David, Hall, and Toole (2000) conducted a critical review of the large body of econometric research studies that attempted to shed some light on the issue of whether public-sector R&D has been a substitute for, or a complement to, private-sector R&D. After sorting out the subset of studies that were adequately designed to test the substitution hypothesis, they found that the results from about one third were consistent with the substitution hypothesis, while two thirds were consistent with the complementarity hypothesis.

He concludes by saying evidence is uncertain as of the writing of the book, so this is why, at the beginning of the book, he says that he preferred a narrative approach to an econometric one: because he doubts that it even can be done:

My own view is that we do not yet have, and perhaps cannot have, a body of rigorous econometric evidence against which to evaluate the economic impact of defense and defense-related R&D and procurement. David, Hall, and Toole explicitly eschewed any effort to assess the magnitude of the economic effect of complementarity. What are the implications for the attempt that I have made in this book to assess the significance of military procurement on the development of commercial technology? My answer is that careful narrative analysis of individual cases is at present a more effective method of capturing the effects of complementarity than econometric analysis. Paul A. David has also pointed out to me that narrative analysis may be better able to capture the long-term or lagged effects of public R&D investments (David 2004).

I will survey this literature in the future.

Dual use and consolidation

Dual use technology is technology developed with both military and civil purposes in mind. It became the conventional solution for military procurement in the eighties. Ruttan then says that Dual use as a major initiative came to an end in 1993 when Deputy Secretary of Defense W. Perry announced an end to the DOD efforts to maintain rivalry among defense contractors by opposing mergers. Policy changes induced massive sectoral consolidation. In the long run, and illustrated in the book earlier, military procurement and R&D have represented a shrinking share of GDP, although the spending itself has remained relatively constant, due to GDP growth.

Answering the question

The first question is, can the private sector be relied on as a source of major new general-purpose technologies? The quick response is that it cannot! When new technologies are radically different from existing technologies and the gains from  advances in technology are so diffuse that they are difficult to capture by the firm conducting the research, private firms have only weak incentives to invest in scientific research or technology development (Nelson 1959; Markiewicz and Mowery 2003). Each of the general-purpose technologies that I have reviewed have required several decades of public support, primarily in the form of military R&D and defense or defense-related procurement, to reach the threshold of commercial viability.

Ruttan here seems to ignore a crucial fact that he is surely aware of, as he has talked about it in the book. This revolutionary revelation is the mundane fact that the private sector is not composed just of firms. You also have inventors, startups, venture capitalists, philanthropists, universities and research institutes, and they have different roles. Sure, firms will rarely invest in really long term and uncertain research, but that's not their role!

He does say that it is not totally impossible, because things like Bell Labs existed, but that was fruit of a past of monopolistic power, and that the large research corporations of the past are long gone. This is in line with recent research.

Ruttan knows there is a way out of the problem:

As support for basic and even early-stage technology development at large corporate laboratories has atrophied, new institutional arrangements for advancing scientific knowledge and technology development have emerged. These often involve complex relationships among large corporate research laboratories, specialized independent research laboratories, and venture capital “angels” or firms. These organizations have increasingly established close links with university or government laboratories that are involved in the more basic or conceptual investigations associated with technology development (Branscomb and Auerswald 2002, pp. 41–57).

But is not optimistic, nonetheless:

find it difficult to anticipate that the private sector, without substantial public support for research and technology development, will become an important source of new general purpose technologies over the next several decades.

He now considers the role of the public sector, discussing initiatives such as CRADAs, NIST, and SBIR, but concludes in a similar bleak way

In spite of a number of promising initiatives, I remain skeptical that public support for nonmilitary or defense-related technology development can be depended on to become the source of major new general-purpose technologies in the foreseeable future. These programs have generated substantial economic benefits, but even the most successful programs must be evaluated in terms of their contributions to evolutionary rather than revolutionary changes in technology

That is, nonmilitary spending may lead to evolutionary changes, but not revolutionary ones. What about military spending? That would serve, says Ruttan, but it won't happen

It was access to large and flexible resources that enabled powerful bureaucratic entrepreneurs such as Leslie Groves and Hyman Rickover (chapter 4), Joseph Licklider (chapter 6), and Del Webb (chapters 3 and 7) to mobilize the resources necessary to move the general-purpose technologies from initial innovation toward military and commercial viability (Doig and Hargrove 1987). They flourished in a political and administrative environment that accommodated their entrepreneurial energies—an environment that no longer exists for military and defense-related agencies and firms. The rationalization of the processes involved in the allocation of resources to R&D in defense and defense-related procurement, combined with changes in the structure of the defense-related industrial base, has placed serious constraints on the ability of military R&D and defense-related procurement to continue to play a dynamic role as a source of new general-purpose commercial technologies.

Thus he concludes

I have argued in this chapter that the U.S. private civil economy is unlikely to generate the major new general-purpose technologies necessary to sustain rates of productivity and economic growth comparable to the rates achieved during the early post–World War II decades and again during the information technology bubble that began in the early 1990s. I have also argued that in the absence of a major war, or threat of a major war, new general-purpose technologies are unlikely to emerge from military R&D and procurement.

I think I have made the case for optimism at some points during the article. Ruttan's argument is that without a war, there is no incentive for military spending, and that GPTs result from there, as revolutionary changes are required, and the military disregard for cost-benefit analysis could help that. My argument is that there isn't really such thing as revolutionary change, and History, and Ruttan's book proves it. Scientia non facit saltus.

To finish, I want to bring your attention to a 2010 Mowery paper titled Military R&D and Innovation. This paper is intended to be an analysis of the field, going beyond Ruttan. This is not about the effect of war on R&D, but about the effects of military spending itself. The conclusions could be summarised as:

  1. Most economic historians (including, for example, Joel Mokyr) assess the effects of war on technological innovation as largely negative, with a few exceptions (Ruttan being one)
  2. The reason of the above is that war encourages a conservative approach to technology, inducing demand for existing technology. Thus most war technology was developed prior to the war.
  3. The more developed a technology is, the harder for spinoffs of military technology to be successful
  4. More research is required, and understanding the effects of military R&D is plagued with difficulties

As for the effect of war itself on R&D, there is an article analysing the poster child case of war induced increased innovation and post-war prosperity: World War II. In The impact of the Second World War on US productivity growth (2008), A.J. Field argues that evidence is not consistent with this view. Increasing demand for manufacturers or spillovers from military technology don't seem to have been helpful to post-war productivity. He also makes a case that so far I've ignored, and is that war can modify the path of technological advancement, inducing some technologies to develop faster than others, decoupling this path from consumer demand.

The scientific and engineering community, in cooperation with government officials, managers, and workers had, by all accounts, and based on our experience with war mobilization, done a superb job in helping to expand the potential output of the economy between 1929 and 1941. This community was then asked to drop much of what it was doing and focus on challenges central to the war effort. In the process, some discoveries and learning useful for civilian production took place. But these were incidental to the war effort, and entailed opportunity costs in the form of disruptions to the trajectory of technical advance in the civilian economy. On balance, it is unlikely that the stock of economically relevant knowledge (both technical knowledge and production knowledge) was higher compared to what it would have been in the absence of war.

So so far in my reading, having reviewed Ruttan's approach and one of his examples, I conclude that no, neither war or military spending are necessary or even beneficial for innovation or economic growth.

Appendix: General Purpose Technologies

A list of them,from Wikipedia, original source is Lipsey and Carlaw's E_conomic Transformations: General Purpose Technologies and Long Term Economic Growth_.

| GPT | Spillover Effects | Date | Classification | | Domestication of plants | Neolithic Agricultural Revolution | 9000-8000 BC | Process | | Domestication of animals | Neolithic Agricultural Revolution, Working animals | 8500-7500 BC | Process | | Smelting of ore | Early metal tools | 8000-7000 BC | Process | | Wheel | Mechanization, Potter's wheel | 4000–3000 BC | Product | | Writing | Trade, Record keeping | 3400-3200 BC | Process | | Bronze | Tools & Weapons | 2800 BC | Product | | Iron | Tools & Weapons | 1200 BC | Product | | Water wheel | Inanimate power, Mechanical systems | Early Middle Ages | Product | | Three-Masted Sailing Ship | Discovery of the New World, Maritime trade, Colonialism | 15th Century | Product | | Printing | Knowledge economy, Science education, Financial credit | 16th Century | Process | | Factory system | Industrial Revolution, Interchangeable parts | Late 18th Century | Organisation | | Steam Engine | Industrial Revolution, Machine tools | Late 18th Century | Product | | Railways | Suburbs, Commuting, Flexible location of factories | Mid 19th Century | Product | | Iron Steamship | Global agricultural trade, International tourism,Dreadnought Battleship | Mid 19th Century | Product | | Internal Combustion Engine | Automobile, Airplane, Oil industry, Mobile warfare | Late 19th Century | Product | | Electricity | Centralized power generation, Factory electrification,Telegraphic communication | Late 19th Century | Product | | Automobile | Suburbs, Commuting, Shopping centres, Long-distance domestic tourism | 20th Century | Product | | Airplane | International tourism, International sports leagues, Mobile warfare | 20th Century | Product | | Mass Production | Consumerism, Growth of US economy | 20th Century | Organisation | | Computer | Digital Revolution | 20th Century | Product | | Lean Production | Growth of Japanese economy | 20th Century | Organisation | | Internet | Electronic business, Crowdsourcing, Social networking,Information warfare | 20th Century | Product | | Biotechnology | Genetically modified food, Bioengineering, Gene therapy | 20th Century | Process | | Nanotechnology | Nanomaterials, Nanomedicine | 21st Century | Process |

We could now play the game and ask how many of these came from military spending. I would argue Ruttan's story only applied to inventions more recent than the Airplane (and there we could then dispute the account or grant it). But then, you can see there was plenty of GPT discovery going on that had nothing or almost nothing to do with governments and the military.

EDIT: I found Forty Years of Aeronautical Research (1955), by J.C. Hunsaker, Chairman of NACA. There it says that before NACA came to be, Alexander Graham Bell and Charles Doolittle Walcott (Secretary of the Smithsonian) founded the Aerial Experiment Association (in 1907), that helped demonstrate the usefulness of ailerons, and originated the Curtiss biplanes. Bell and Walcott, members of the National Academy of Sciences, also participated in the lobby effort that culminated inthe creation of NACA.


Anderson, J. D. (2005). Introduction to flight (Vol. 199). Boston: McGraw-Hill.

Arora, A., Belenzon, S., & Patacconi, A. (2015).Killing the Golden Goose? The Decline of Science in Corporate R&D (No. w20902). National Bureau of Economic Research.

Eberhardt, S., & Komerath, N. (2009). The Guggenheim Schools of Aeronautics: Where are they today?. In American Society for Engineering Education. American Society for Engineering Education.

Field, A. J. (2008). The impact of the Second World War on US productivity growth. The Economic History Review, 61(3), 672-694.

Glines, C.V. (1996)_The Guggenheims, Aviation Visionarie_s. Aviation History.

Hansen, J. (1986). Engineer in Charge: A History of the Langley Aeronautical Laboratory, 1917-1958.

Mazzucato, M. (2013). _The entrepreneurial state: Debunking public vs. private sector myths _Anthem Press.

Mowery, D. C. (2010). Military R&D and innovation. Handbook of the Economics of Innovation, 2, 1219-1256.

Roland, A. (1984). Model research: The National Advisory Committee for Aeronautics, 1915-1958, volume 1.

Ruttan, V. W. (2006). Is war necessary for economic growth?. Historically speaking, 7(6), 17-19.

United States. Bureau of the Census. (1975).Historical statistics of the United States, colonial times to 1970 (No. 93). US Department of Commerce, Bureau of the Census.

Wikipedia entries: von Ohain, Heinkel HeS 1, General Purpose Technology

Comments from WordPress