In today’s Forbes.com, Sramana Mitra has a column discussing the role of universities and taxpayer funding in innovation that contributes to economic growth. She is a thoughtful and insightful commentator, and I recommend the article, despite having a couple of disagreements with her argument. I do think she overstates the case against innovation arising from market processes in her introduction:
In an ideal world, a capitalist framework would facilitate all the economic progress essential to civilization. Unfortunately, we don’t exist in an ideal world. Innovation, however, is among the areas where capitalism as we know it–private, free market, borderless and without government intervention–fails.
For one thing, “innovation” is a large and amorphous category, and that vagueness makes her general claim less plausible. What exactly is innovation? To condemn the ability of market processes to lead to some notion of the “optimal” amount of innovation requires a more explicit definition and scope. I think a more plausible claim, and one that adds to the persuasiveness of her argument, is that basic science research that has long lead-times and is not directly ex ante connected to some possible commercial product has some public good characteristics, and is therefore more likely to be underprovided via market processes. Note here that I am narrowing the scope of “innovation” to not-obviously-commercializable activities.
Her argument then uses DARPA as an example of the various useful roles that academia, government, and industry can play in generating value-creating innovation, and she makes this statement that is consistent with my framing of the issue above:
But does that mean capitalism is not key to innovation? Absolutely not. Once innovation is ready to be brought to market, it is best for academia and government to wrap it in the right capitalist packaging and hand it over to the market.
It is an intricate dance–this tango between industry and academia–with the government playing DJ in the background. Few have learned to dance it well. MIT, Stanford, Berkeley and Carnegie Mellon belong in an elite list of about a dozen universities that do a truly professional job of consistently bringing university-led innovation to market.
I think this is correct. In fact, I spoke on a panel at Northwestern earlier this week where we tackled precisely these issues. There are several challenges in pulling off the healthy academia-government-industry collaboration that Sramana describes, and I think many of those challenges reflect the difficult political economy that accompanies taxpayer-funded research. One effect of the politicization of such research is the mismatch of the justifiable desire for accountability and low risk with the nature of basic research — much research leads to “non-results”, which from a scientific perspective are still valuable results, but from a political perspective do not necessarily count as meaningful and valuable results. Another effect is the timing mismatch; payback periods on research investment in basic science are typically long, while political cycles are short (2-4-6 years), so the incentive to support research with longer and more unknown timeframes is reduced. Finally, the future-taxpayer debt-funded research funds in the current stimulus package have a short timeframe, 20 months typically, while the funding stream that is most consistent with generating valuable results is longer and slower and less episodic.
Taxpayer-funded research also has crowding out problems, where government funding induces private sources to reduce their funding, and empirically the return on taxpayer investment is lower than the return on private investment in research. In combination with the political economy issues this leads me to prefer private funding on both philosophical and pragmatic grounds, despite the public good characteristics of basic science research (what a surprise!). Not a very popular or plausible stance right now.