Tuesday, March 11, 2014

Local Commercialization Incentives

Guest Blogger

Camilla A. Hrdy

For the conference on Innovation Law Beyond IP at Yale Law School

The Innovation Law Beyond IP Conference will explore how forms of law and governance beyond intellectual property rights (IPR) promote innovation. In my new paper, Local Commercialization Incentives,[1] I make two contributions to this discussion. First, I question the wisdom of proposals to introduce new forms of patents to facilitate commercialization of un-worked inventions based on the observation that the U.S. government already provides financing for commercialization at both the federal and local levels. These incentives, sometimes called commercialization awards, may be as efficient or more efficient than so-called commercialization patents. Second, I question the common presumption that national incentives are preferable to state and local incentives for promoting innovation due to the existence of inter-jurisdictional spillovers. As with the argument for IPR maximization, this argument is based on the questionable premise that containment of spillovers is always socially beneficial.[2] 

In this post I’ll briefly summarize the main points I make regarding each of these two contributions.

[1] Rethinking the Case for Commercialization Patents As we know, the patent system protects inventors' exclusive rights in their "novel" and "nonobvious" inventions. But some scholars have argued that we also need new forms of IPR to promote commercialization of those inventions. Just as technical information is subject to free-riding, Michael Abramowicz and John Duffy write, so is information generated in the process of commercializing known but “economically nonobvious” inventions; and just as patents encourage “risky but ultimately beneficial technological experimentation,” some form of IPR for commercialization “could result in a socially beneficial increase in market experimentation and entrepreneurial activity."[3] Controversial proposals of this ilk include Ted Sichelman’s recent suggestion that Congress start offering short-term “commercialization patents” for inventions that go uncommercialized for three years after patenting,[4] or a similar form of IPR called “innovation warrants,” which was suggested by William Kingston a while back but never adopted.[5]

At first glance, the basic reasoning behind IPR for commercialization makes sense. Innovation is a long process and "can be seen broadly as encompassing all forms of risk-taking and entrepreneurship, from new technology to new business organizations to minutiae-laden everyday business decisions."[6] Economists, business management specialists, and investors have long recognized the parallels between spillovers generated from technological experimentation and market experimentation, both of which may diminish or eliminate incentives to invest. And, as scholars like Edmund Kitch observe, patents already arguably do more than simply “reward” technological experimentation.[7]  So given these similarities, wouldn’t it “highly surprising if the optimal policy choice for encouraging market experimentation were always to rely upon whatever natural first-mover advantages exist in a particular market …”?[8]

I agree with Abramowicz and Duffy that the answer to this question is “yes, it would be surprising.” However, I disagree with their subsequent conclusion that the solution to this problem is to “deploy some form of exclusive rights, since society generally relies on exclusive rights to encourage technological experimentation."

I have two main reasons for this disagreement and explain each below.
1. Commercialization Patents Lack Historical Basis in Modern U.S. Patent Law. First, it is not clear that IPR to promote commercial experimentation has a strong historic basis in patent and intellectual property law at all – at least not as it developed in the United States circa 1787. Based on my work on historic state and colonial patents, I agree that governments once commonly employed “exclusive privileges” as a cheap and efficient way to promote both new invention and commercial risk-taking.[9] But as Mario Biagioli has observed, early "privileges” in England and the Venetian Republic were quite different from the patent system that developed under the U.S. Constitution.[10] In the first Patent Act of 1790, Congress expressly limited patents to universally novel inventions that fell within a specific category of subject matter (specifically, “any useful art, manufacture, engine, machine, or device, or any improvement therein not before known or used"), and instituted no requirement that inventions be practiced.[11] Over the course of the next century, U.S. patent law would come to value generation and disclosure of new technical know-how far more than the actual working – let alone the commercialization – of inventions.[12]

This is not to say the idea of patents to promote commercialization was never considered.  Alexander Hamilton, in particular, wrote in his Report on Manufacturers (1791)[13] about the need for more commercial risk-taking and entrepreneurship in the infant nation, and recommended a variety of incentives, including “for a time, exclusive privileges,” in order to encourage “new intentions and discoveries at home, and the introduction into the United States of such as may have been made in other countries." Yet Hamilton backed off on his proposal for patents for “introducers” for two reasons. First, he apparently thought Congress might lack power under the IP Clause to grant patents for inventions that were not new. Second, he recognized that patents were not the only way to encourage “new intentions and discoveries.” In fact, Hamilton wrote that, along with “exclusive privileges,” “[t]he usual means of … encouragement are pecuniary rewards.” 

2. We Probably Don’t Need Commercialization Patents. This leads nicely to my second basis for objection: we probably don't need more IPR to promote commercialization because non-patent incentives already perform a very similar function. Obviously, the private sector supplies venture capital for a select group of lucky entrepreneurs who are developing risky but potentially profitable new technology.[14] Although I disagree with Mark Lemley that market incentives are enough,[15] in the paper I demonstrate that the government already provides the modern-day equivalent of Hamilton’s “pecuniary awards”: risk capital for businesses engaged in commercialization or early-stage technology development that cannot raise sufficient funding to survive into profitability due to the high cost and risk involved in developing and marketing unproven inventions and the chance that competitors will free-ride off first movers’ efforts.[16]

As I discuss at length in the paper, an array of federal research agencies, such as the Department of Defense and the Department of Health and Human Services, provides around $2 billion a year for small businesses that are developing innovations with commercial potential through the Small Business Innovation Research Program (SBIR), with awards ranging from around $150,000 to $1 million in size.[17] Perhaps more surprisingly, many U.S. states – including Connecticut, Illinois, Iowa, Indiana, Kansas, Kentucky, Maryland, Michigan, Minnesota, New York, Oklahoma, Texas, and Wisconsin – offer cash, loans, and/or equity awards ranging anywhere from $10,000 to $2 million for high technology companies that are in or near the commercialization stage. Although state awards – which are called things like “commercialization awards”[18] or “innovation acceleration” awards[19] – may seem like chump change compared to the Sequoia Capitals and Googles of the world, research suggests that start-ups and entrepreneurs pursue them,[20] and that they can significantly enhance award winners’ chances of securing financing from private investors with deeper pockets.[21]

My skepticism of proposals for IPR for commercialization grows deeper still in light of the fact that an initial comparative analysis of commercialization patents and commercialization awards suggests the latter may be as efficient or more efficient than the former. Critics of rewards for inventions such as Scott Kief[22] stress the administrative difficulty involved in efficiently administering such systems. But drawing on prior scholarship on prizes[23] and venture capital strategies[24], I suggest that, when commercialization awards are made contingent on private sector matching or other certifications of technical and commercial viability – which most state awards are – they can reduce the risk of government wasting public money on bad investments and avoid some of the concerns about politico-economic goals like instant but unsustainable job creation disrupting awards’ utility at promoting true innovation.

That said, I am not sure it is even necessary to go this far. The upshot of Abramowicz’ and Kieff’s work in this area is that granting prizes for inventions is a feat that should not be undertaken lightly since we already have a patent system in place. But when it comes to government awards for commercialization, the opposite is true: no government in this country offers commercialization patents, but many U.S. governments offer commercialization awards. In light of scholarly debates over the patent system’s efficacy and the relatively anti-patent political climate, it seems wiser to try to improve existing non-patent incentives rather than trying to design and implement entirely new forms of IPR for commercialization.

[2] Do We Have A Bifurcated Patent System? The second main insight I make in the paper is that, once the field of innovation incentives is expanded to include non-patent options like commercialization awards, we can and should begin to think about patent and innovation policy from the perspective of federalism and decentralized governance. As I’ve discussed elsewhere, the reason the national government is generally held responsible for sponsoring generation and disclosure of inventions is that smaller jurisdictions like states and cities cannot easily internalize the benefits of new information produced as a result of inventions made on their dollar.[25] In contrast, I argue, state and local governments can reliably internalize many of the benefits of commercialization. These benefits potentially include short-term politico-economic wins, such as company location, in-state property purchases, increased tax revenues, and hiring of local workers. But they can also include long-term productivity gains that result from information generated during the actual practice, production, and marketing of inventions.[26]  Although this information may “spill over” to other jurisdictions, a line of research by Eric Von Hippel,[27] Gary Pisano,[28] and Suzanne Berger[29] suggests that information related to the actual practice and marketing of inventions – in particular, to manufacturing – is likely to remain geographically “stuck” to the location in which it originated and cannot easily be transmitted to other locales.[30] 

Therefore, drawing on Robert Cooter and Neil Siegel’s “collective action federalism,”[31] I suggest that the patent system may be to some extent bifurcated, with the national government solving collective action problems through patents and grants for basic research, and state and local governments supplying non-patent financial incentives for applied research that is likely to be commercialized in their jurisdictions in the near future.[32]  I argue that this bifurcation is not only accurate at a descriptive level, but that it may also be efficient. According to the theory, when it comes to supplying “local” goods whose effects are “internal” to a state, local actors possess superior incentives to supply those goods and superior information to do so effectively. Moreover, when states experiment in designing commercialization strategies and copy one another’s successes, this can lead to innovations in law and policy that would not happen with full nationalization.[33] Thus, local commercialization incentives could lead to more commercialization of inventions and greater overall productivity than would occur if innovation policy were left exclusively to federal actors and institutions.[34]

Camilla A. Hrdy is a resident fellow at the Information Society Project. She can be reached at cahrdy at

[2] The notion that spillovers should be contained is presently being challenged both by scholars of intellectual property law like Brett Frischmann and Mark Lemley, and by scholars of federalism like Heather Gerken.
[12] This generalization is not without exceptions.  For instance, as Duffy has shown, early U.S. courts distinguished between patents that were actually worked and those that remained "mere pieces of paper." Today U.S. patents are generally available for business methods and for “nonobvious” variations of known inventions.
[23] As Duffy has shown, early American courts also used the “paper patent doctrine” to reward commercial risk worked patents more favorably than those that remained “mere pieces of paper.”
[26] John Duffy emphasizes the importance of new information generated through actual practice of inventions in his new paper on the “paper patent doctrine.”
[32] By “basic” versus “applied” research I loosely mean “research with no commercial value which lays a foundation for commercial products.” But I recognize that this distinction is “shaky,” especially when so-called basic research is patentable. See Suzanne Scotchmer, Ideas and Innovations: Which Should be Subsidized 18 (NBER working paper, January 11, 2011),
[33] Lisa Larrimore Ouellette makes a similar argument about the benefits of local knowledge and policy experimentation in patent law.
[34] Clayton Gillette has made a similar argument in favor of state business incentives, generally,

Older Posts
Newer Posts