The Portland International Conferences on Management of Engineering and Technology (PICMET) occur in Oregon in odd-numbered years, and in diverse locales in the even. I write from lovely Phuket island in southern Thailand, as I listen to bird- and cricket song and the crashing surf of the Andaman Sea, and gather my thoughts following the close of PICMET-2010.

Earlier today the thoughts said, "Go to the beach!" So I did. But this evening I’m thinking about government technology policy. The conference shows it is enjoying a resurgence in Asia, following a brief flirtation here with neo-liberal economic policies. I’ll tell you about a potentially important consideration, raised by researchers in my conference sessions, involving technological spillovers. Don’t mind the jargon, I’ll make this palatable and at least semi-interesting, I promise.

One PICMET presenter explained how the Thai government adopted his dissertation research, which applied the Analytic Hierarchy Process  to prioritizing funding for the scientific and technological underpinnings of Thailand’s goal of being a leading foodstuffs-based economy. (There’s gonna be a lot of genetic and biotech stuff going on in Thailand.) His co-author explained how the government of Bhutan implemented a similar approach for electrifying every region in that country. The research I presented, with my co-author Mei-Chih Hu of Taiwan’s National Tsinghua University, had to do with China’s policy-driven initiatives in biofuel development.

In the United States we do not do this kind of thing at the national level. (Local and regional initiatives abound, of course, as various metro areas try to create echoes of Silicon Valley. But that’s a topic for another column.) President Ronald Reagan declared that government was incompetent to pick technological winners, and that technical directions were best chosen by the private sector. Since the Reagan administration, then, government technology policy has been unfashionable in the US, with the notable exception of military technologies.

Any federal agency whose activities smacked of "picking winners" was refocused. The National Institute of Standards and Technology (NIST)  was allowed to continue co-funding high-cost, high-risk engineering initiatives if industry would pony up the balance of the dollars. A 1990s Republican Congress de-funded the US Office of Technology Assessment as part of Newt Gingrich’s Contract On America. (OK, I know he called it Contract With America, I’m just letting my bias show.)

But can the private sector really do it better? There has been, I believe, no rigorous study of this question. The US federal government has difficulty dismantling a bureaucracy, once built. So I would concede, on the opinion level, that government may pick beneficial technological directions less efficiently than the private sector, because when the government is wrong, it’s expensive to recover. But I see no reason to believe corporations can pick winners more consistently. That is, out of ten chances to pick winners, I’d bet governments and corporations would be right about the same number of times.

Even as Reagan dissed government foresight, Japan and the Asian Tigers (South Korea, Singapore, Taiwan, and Hong Kong) were experiencing tremendous economic success and growth under government-driven tech strategies, Japan’s focused on memory chips, automobiles, steel, and so on. Later, in the 1990s, Japan’s ”lost decade“ began, the Thai Baht crashed (due to lending practices similar to those behind the 2008 US crash) and dragged other regional currencies down with it, and the International Monetary Fund (IMF) imposed a severe neo-liberal regime on South Korea in return for loans to keep that nation’s economy afloat.

No doubt that Japan and the Tigers over-spent and over-lent in support of their economic-technological strategies. This does not imply a blanket condemnation of government technology policy ("GTP"), however. In fact, Koreans still believe the IMF did them more harm than good.

That was a bit of digression; what we really want to talk about is Reagan and his Strategic Defense Initiative. Popularly known as "Star Wars" due to its surface implausibility, the SDI was to combine laser, magnetic, electro-kinetic and missile weaponry with computer control, to create an impermeable anti-missile shield over the United States. It was one of those military exceptions to GTP that I mentioned earlier. It didn’t work. It was a massive flop.

No, wait, maybe it’s not fair to say it didn’t "work." Here’s what happened. The global shield was never developed, never even got close. Under President Clinton, SDI was renamed and redirected toward regional and theater-level shields, and I believe some progress has been made thereto. (NATO is now looking at a missile shield for Eastern Europe.) SDI spurred much useful work in computer science - useful in non-defense arenas - and basic research in the sciences underlying SDI is still going on in the US national  laboratories, using funding from SDI’s successor programs. So there has been benefit from SDI.

On a grander scale, it is widely thought that Reagan’s resolve on the SDI front, and the capability (at that time) of the American economy to absorb the huge cost of developing SDI, accelerated the fall of the Soviet Union. The Soviet economy was in a shambles (though the CIA did not fully apprehend that fact), and could not escalate weapon research and production to match SDI, a project which American wealth, resolve, and technology might - just might - pull off. The Cold War balance of terror could not be maintained. The Soviets gave up, the Berlin Wall came down, and that was that, at least in this simplified retelling.

And now, at long last, we come to the point of this column. Defense and non-defense SDI-related scientific advances, and the impact of the SDI project on geopolitics, are called spillovers. Spillovers are knowledge, skills, and behaviors that "leak" from a series of techno-economic transactions. In a classic example, Corporation A produces widgets in City X and has a fixed set of suppliers. Employees of Corporation A and suppliers shoot the breeze with non-employees at bars and churches, talking shop but not divulging trade secrets. This free-floating information suffices to inspire other residents of City X to start related companies and even compete for supplier roles to Corporation A.

Spillovers are externalities, an externality being any effect an economic transaction has on a person who is not a direct party to the transaction. Externalities include air pollution from a factory, as that pollution afflicts non-customers. "Dual-use" of military-developed technology in civilian and commercial applications is an example of a spillover externality.

So SDI, a government-chosen technological direction, was a flop, but the spillovers were substantially beneficial.[1]  Intriguing. Might there be other examples where this is the case?

The French Minitel was a to-the-home teletext service that worked over the phone lines. Backed by the government of France, Minitel had penetrated most French households and businesses at the moment it was obsoleted by the World Wide Web. Minitel itself was a flop, at least in that it had failed to anticipate the Internet. (Though, hey, as we all recall, Microsoft failed to anticipate the Internet.) But Minitel had prepared French businesses for e-commerce, and by virtue of its text interactivity had laid a path for accelerated uptake of Internet technologies on the part of the French public, with all the economic efficiencies that implied. Spillovers ruled.

China’s latest 10-year plan makes biofuel development a national priority - though bio-energy will remain small compared to coal and oil. Dr. Hu’s analysis of Chinese patent cross-citations shows that biofuel advances in China, though originally closely related to advances in foodstuffs and chemicals, has since 2000 most closely benefited and benefited from pharmaceutical advances.

Technically, with the possible exceptions of cellulosic, algal and sugarcane ethanol, biofuels appear to be a dead end. Skeptical re-evaluations of biofuels direct criticism at questions of energy budget, water budget, disruption of the soil nutrient cycle by removal of crop residue for biofuel production, economies of scale, and externalities including indirect emissions, social exploitation, competition with food crops, and the health effects of concentrated mycotoxins in the by-products of biofuel processing. Ausubel^[2] notes,

...in biomass the lack of economies of scale loom large. Because more biomass quickly hits the ceiling of watts per square metre, it can become more extensive but not cheaper. If not false, the idol of biomass is not sustainable on the scale needed and will not contribute to decarbonisation. Biomass may photosynthesise but it is not green.

Again the direct government initiative for biofuel may prove to be a flop (more complexities are discussed in our paper[3]), but the spillovers - faster advances in pharmaceuticals, to name one - may be substantial.

Countries that have been committed to government intervention, and those that have been committed to non-intervention, may be in the process of reaching compromises. This week’s newspapers report a rapprochement between the Tigers and the IMF. In the US there is a groundswell movement to bring back the OTA. The European Union seems committed to "National Innovation Strategies." As it seems nonsensical for governments simply to tell citizens "Innovate!", one would correctly assume that such strategies do have more or less specific technological focuses and are simply avoiding the more inflammatory phrase "technology policy."

As that may be, we have come to a hypothesis: Even if a government-driven technological direction proves to be a wrong one, the spillovers may provide substantial benefits and even pay back the social cost of the central failure. Research, naturally, will bear this out or not as the evidence dictates.

Spillovers are closely related to serendipity and making lemonade from lemons, and these things happen in industry as well as in government. I’ve provided only anecdotal evidence in this column, but I hope to have argued effectively that government technology planning is not always a bad thing.

-------------------------------------------------------------------
[1] Strangely, I had an intuition of this in the early SDI days, and it might nearly have cost me my job. I was working for a billionaire technology entrepreneur who brought up SDI in conversation. I remarked that I was sure it wouldn’t work, but that we’d get some good advances in computer science out of it. The boss stared daggers at me. The explanation for his pique? First, he was the son of Russian refugees, and had personal hopes that SDI would cow the Soviets. And second, no one who hoped to gather in some SDI money (as our institute did) could afford to pooh-pooh the plan. The man was interested in my opinion, but wanted to make sure I wouldn’t repeat it.