How the US R&D model was weakened

This is the second article in a three-part series. Part 1 is When the US risks being leapfrogged.

An examination of the United States’ historical experience with infrastructure and government spending on research and development sheds light on what went wrong – and what the remedies are. 

Consider the late president Dwight Eisenhower’s “Don’t Sputnik me” debates in 1958.  When asked about overcoming recession then, his answer was: “Let’s not just get a Sputnik attitude about everything.” 

He decided that the $1.7 billion pork-barrel Rivers and Harbors Bill with its 154 projects (when a billion was still a “billion”) was not an anti-recession remedy.

The major expenditure in the proposed budget was $7.2 billion for highways. Eisenhower, a Republican, wanted to enhance highway spending by only $2.2 billion over four years, but the Democrats managed to get $1.8 billion for the 1959 fiscal year alone. 

History is rhyming again, with now hundreds of billions and trillions planned to be spent on a range of what is claimed to be “infrastructure” investment. So far nobody in Washington has come up with a “Don’t Covid me” line, though the debates about the stimulus sound the same as in 1958, when we adjust the numbers for inflation and population.

Part of the motivation to ramp up highway expenditure was the perceived need to allow a quick escape from densely populated cities after the panic over possible Russian nuclear missile or bomber attacks following the launch of Sputnik in 1957. The attacks did not happen, but many city dwellers and manufacturers moved to suburbs. 

The highways thus had the unintended impact of leaving poorer people behind (many of them blacks), many now jobless, in what became the disastrous “inner cities” infested with crime for decades. Such long-term impacts were out of sight and mind during the years of massive spending.   

The 1958 National Defense Education Act was another reaction to Sputnik, leading to massive spending on universities in the hope for advances in rigorous education. That turned out to be a failure too, gradually bringing academia into the credential-creating institutions many have now become, backed by much jargon and few skills – which will be discussed in Part 3 of this series.     

Public debate in the US after 1957 led to major changes not only in domestic matters but major foreign-policy matters too, related to the military and R&D. 

The historian Walter McDougall documented how in his book … The Heavens and the Earth: A Political History of the Space Age (1997), the US having adopted in the decade following Sputnik a Soviet-style “Five Year Plan” to prove to both NATO allies and what was then called the Third World that they should not worry that the US model of society would not prove superior to the Soviets’ and that its security guarantees could be trusted.   

But as McDougall and many others have noted, in spite of a few successes, the resulting large-scale “technocratic” approach was not a success: In McDougall’s words, none of the plans “came to pass. Instead, the dream of limitless progress through government-sponsored R&D began to fade.…

“By the time Apollo was history, NASA’s budget was plummeting, many of its best engineers had left, and the agency had lost the ‘institutional charisma’ of its early years.” Once the “best engineers” left, the outcome was predictable: mediocrity. 

Already in 1991, a 400-page book of detailed case studies published by the Brookings Institute, The Technology Pork Barrel, each chapter devoted to a special R&D project, five scholars from Stanford, Columbia and other universities and institutions summarizing extensive evidence, had made observations similar to McDougall’s years later.  

They concluded (page 365): “On the basis of retrospective benefit-cost analysis, only one program – NASA’s activities in developing communication satellites … can be regarded as worth the effort. But that program was killed.… The photovoltaics program made significant progress, but it was dramatically scaled back for political reasons.… The remaining four programs were almost unqualified failures. 

“The supersonic transport (SST) and Clinch River Breeder Reactor were killed before they produced any benefits, and Clinch River, because of cost overruns, absorbed so much of the R&D budget for nuclear technology that it probably retarded overall technological progress. The space shuttle cost too much and flies too infrequently. 

“The synthetic fuels program produced one promising technology … but billions were spent on another pilot and demonstration facilities that failed.”

In 1997, McDougall summarized the politics behind such failures, noting that the belief was that “politicians could mobilize science and technology for the achievement of technical and even social goals without politics itself skewing the process.” 

Yet the space program, for example, “offended not only the cultural critics of the 1960s, who damned the military-industrial complex, the arms race and the mentality behind the Vietnam War, but also the environmental and feminist movements, which … identified technology with pollution and ‘missile envy.’” 

The combination of talented staff leaving the institutions combined with the above political/cultural compromises brought about the gradual decline deplored now.       

The above experience in the US with R&D and other massive government spending now raises the same questions that McDougall raised in 1997. 

Could massive misspending be remedied “if only the right people were put in charge, the right decisions made, even more money spent, and the experts insulated from the … impulses of the Congress and the public?” 

Or, “Is the lesson that technocracies … are inherently flawed because there are no ‘right people,’ poor decisions will be always made … and because bureaucracies, the interest groups that align with them and the political leaders that yearn to steer them to their own self-interested ends, are every bit as ‘irrational’?”  

A closer look at the evidence, following the reasoning implied by McDougall’s quoted options, suggests that there are ways to restore US prominence in R&D: The guiding principle being to restore institutions and agreements that would lead to cooperation to better match talent and capital accountably, correcting mismatches faster. 

It was the sequence of events that let the institutions and arrangements that matched the best-skilled and money by government, private companies and educational institutions in the US lapse or even disappear that brought about the decline.         

Which matching was weakened and which strengthened?

In a sequence of books and articles, the first published in 2007, Henry Kressel, Norman Winarsky, Thomas V Lento and David P Goldman summarized extensive evidence to suggest that the biggest change explaining the decline in long-term R&D projects is the disappearance of corporate laboratories that were partially, directly and indirectly, funded by the US government. 

A prominent case of “indirect” funding was that of AT&T, which, having monopoly powers, could pay for Bell Labs’ employment of thousands of engineers and scientists.  

True, these labs brought together the best interdisciplinary teams for extended periods. But as the experience of William McGowan, chief executive of MCI (Microwave Communications Inc), showed, this did not necessarily lead to the best way to commercialize innovations: AT&T did everything in its power to stop them. 

MCI’s plans were to build microwave radio-relay stations that truckers could use for two-ways communication on highways and waterways and also to offer voice, computer information, and data communication services for business customers unable to afford AT&T’s TELPAK monopoly-priced services. 

But it took years for MCI to fight AT&T in the courts when the latter blocked the former from its phone network. It was McGowan who brought about the breakup of AT&T, and the phone and communication sector were then transformed, MCI and McCaw Cellular leading the way (both financed by Michael Milken’s “junk bonds”).  

In fact, MCI Mail, launched in 1983, was one of the first e-mail services and MCI.net was an integral part of the Internet backbone. Both breakthroughs came about only after the courts decided that AT&T’s monopoly was de facto, not de jure: For years McGowan jokingly referred to MCI as a “law firm with an antenna” – having to employ so many lawyers, and needing its headquarters to be located in Washington. 

Kressel et al point out correctly, though, that “maintaining a leading-edge economy the size of the United States requires combining the skills of the most talented people with appropriate resources to build market leadership, and corporate labs were critical components of this process.” 

No doubt some on the list, including labs at Westinghouse, Hughes, Xerox and RCA, brought about a range of important innovations. Although Kressel et al do not mention that monopoly and government-linked financing had their downsides too in matters of both R&D and commercialization of inventions, they do conclude that these corporate labs cannot be restored now even if governments increase R&D funding. 

(In fact back in early 1990s, observers were already skeptical that the US R&D laboratories would be able to “influence America’s prowess.”) 

While the big corporate labs were thus a gradually disappearing act, both because of the labs losing top talent and Washington’s politically motivated decisions, as McDougall and others emphasized, another sector was entering the void.

VC, financial sector and military step into the void

Immediately after World War II, entrepreneurs and scientists, noticing the lack of funding for startups, initiated discussions with US government agencies. They persuaded the Securities and Exchange Commission to make precedent-setting decisions to approve funds entering the void under a “public interest” exemption. As a result, small funds were created, though these remained of marginal importance well into the late 1950s.

The big change in the venture capital sector came in 1958 – no surprise – correlated with the Sputnik-induced hysteria. The first VC limited partnership (Draper, Gaither, and Anderson) was formed in 1958, and the US government set up the Small Business Investment Companies, a federally guaranteed risk-capital pool, establishing the conditions for a competitive angel and venture capital sector.

The same year, as noted above, the government passed the National Defense Education Act, allocating a large amount of taxpayer dollars to expanding enrollment in universities’ science and engineering programs – though the execution of the plan had disastrous, compounding impacts over the decades since, as events on campuses now show. 

Focused spending on STEM (science, technology, engineering and mathematics) rather than indiscriminate, massive direct and indirect spending on academia would have been a better policy.  

It was in what became Silicon Valley that the interaction among the expanding universities (where many scientists from the dwindling corporate labs found refuge), venture capital firms, the military, and the tech sector became successful, also attracting qualified and ambitious immigrants from around the world (disproportionately represented in the top echelons of Silicon Valley). 

Although many governments spent massive amounts to replicate Silicon Valley’s success, spending on “centers of excellence” within academia, they did not succeed. Contrasting Canada’s with the Israel’s experience sheds light on the critical role educational institutions (and I include the military among them) and venture capital and the financial sector play in sustaining R&D prowess. 

Solid cooperation leading to better matches between talent and sources of capital brings such prowess about, whereas weakened cooperation does not: As shown before, this is what doomed US national research laboratories with the best quitting and politics then skewing the selection process, as McDougall put it.

Here is Canada’s experience in a nutshell. After its creation in 1997 and until 2004, the Canada Foundation for Innovation (CFI) committed $3.8 billion in support of 5,585 projects at 128 research institutions in 64 municipalities. 

In its 2004 Commercialization Report, the CFI showed that in Canada license income received per $1 million of research expenditure had gone from $10,000 to $20,000 since 1997. In the US during the same time, it went from $30,000 to $60,000.

The performance of startup companies formed per $1 million in research expenditure is reflected by these numbers: The rate of startup formation since 1997 in Canada went down from roughly 0.06 to less than 0.04. In the US the number stayed stable at 0.02. 

The CFI report concludes that both results “can be explained by the lack of private-sector receptor capacity in Canada” (page 8) – referring to lack of cooperation with venture capital to select among the projects.    

Here is one concrete example of what “lack of sector receptor capacity” means. David Huber, an engineer working for General Instruments, invented a crucial part of the technology. For years he tried to promote developing it within his company but without success. 

In the end, it was an “angel” group that gave him a few hundred thousand dollars.  He left his job at General Instruments and established a company called Ciena. When that company made an initial public offering (IPO) in 1997, Huber’s stake alone was valued at $200 million.  

The Israeli experience stands in sharp contrast with the Canadian one and shows where the remedies to bring back both US R&D to its previous glories and some of its manufacturing base (whose loss Kressel et al justly deplore) – the “secret sauce” being bringing back institutions to make better use of its talent pool from an early age, and a better matching process between it and the different sources of capital.

In 1985, a team including a former chief of staff of the Israeli Air Force and people from the US experienced in venture capital created Israel’s first VC firm, Athena Venture Partners. In 1993, Israel offered tax incentives to foreign venture-capital investments by matching any investment with government funds. 

As in the United States after World War II, this policy was initiated as part of Israel’s transition from its centralized model of society (which it had adopted at its founding in 1948, centralization being a necessity as in all war-wrecked countries) and disastrous inflation in the 100% range for few years, reaching 450% in 1984. 

Even more than in the US, immigration played a crucial role. Among the 1 million Russian immigrants in the late 1980s (a 20% increase in Israel’s population at the time), more than 55% had postsecondary education; 15% were engineers and architects; 7% were physicians; and 18% were technicians and other professionals. 

By 1998, Israel had 140 scientists and engineers per 10,000 in its labor force, becoming the world leader in these terms, followed by the United States with 80 and Germany with 55.

Networks formed during compulsory military service helped shape successful business teams, such as a unit within the Intelligence Corps responsible for collecting signal-intelligence decrypting codes. The founders of Nice, Comverse, Stylit and Outbrain all emerged from this unit. 

Jason Gerwitz’ 2016 book Israel’s Edge describes Talpiot, a unit created 30 years ago as a collaboration among the Israeli Air Force, a weapons and technology arm of the Israel Defense Forces, and the Hebrew University, which scrupulously selected young people who had to sign up for a 10-year commitment. 

The unit’s graduates ended up creating companies around the world. The mandatory military experience, followed up by annual service in the reserves, allowed such a “vital few” to surround themselves with a disciplined team and create a relatively hierarchy-free culture that bore similarities to California’s Silicon Valley culture.    

Briefly, as in the US, the association among the military, universities and venture capital, attracting critical masses of top talent, surrounding themselves with disciplined teams, brought about successful innovations – without reliance on large R&D laboratories within large corporations. 

The cooperation among extremely rigorously selected people, properly financed by a range of cooperating institutions – military, universities and private firms – were the keys to success.  

French President Emmanuel Macron alluded to that on June 15 this year when stating the goal of having “10 companies worth €100 billion by 2030” in Europe. He admitted, though, that “entrepreneurs need to push governments to be more efficient” to get there and prevent the US and China from leapfrogging Europe. 

Since Macron said nothing about why entrepreneurs have not been able to do this for decades, perhaps Oscar Wilde’s observation that “big words mean so little” applies.

But drawing on the facts and events presented until now, the remedy for the United States’ technological prowess lies indeed in having entrepreneurs cooperate with governments, the financial sector, universities and other educational institutions – and I would include the discipline-requiring military among them – to bring back the better matches between the United States’ (or for that matter any country’s) pool of talent and its sources of capital. 

Part 3 of this series will focus on evidence that such a matching process has been weakened – and, more important, how it can be remedied.    

This article is the second in a three-part series that draws on Reuven Brenner’s books History – the Human Gamble (University of Chicago, 1983), Labyrinths of Prosperity (University of Michigan Press, 1993) and Force of Finance (Thomson/Texere, 2002), and a series of recent articles in American Affairs and Law and Liberty.