How should a federal government deal with ownership of inventions made at universities with federal support? Consider the situation that existed at the time the Bayh-Dole Act was being implemented. Many universities did not have technology transfer offices of the form we have today. A number used research foundations. Many others had deals with invention management agents such as Research Corporation.Research Corporation, facing logistics challenges in dealing with many universities, advocated creating local technology transfer offices to work with inventors, obtain disclosures, and then send these along to RC for review and possible management.
The technology transfer office was a finder, not a keeper. Ownership and licensing were matters for the invention management company. At the University of Washington, for instance, that’s how the Office of Technology Transfer was originally conceived, with the Washington Research Foundation, modeled in part on Research Corporation, serving as the licensing agent. Very few universities–notably MIT and Stanford–had internal licensing offices.
For those working on implementing the Bayh-Dole Act, the challenge was dealing with this existing diversity of practices. What was argued–persuasively–was that this diversity of practices was doing better at placing inventions than were the federal agencies. The agencies appeared to be limited in their operating models, less motivated to take risks, unwilling to offer exclusive licenses (perhaps for good reason), and burdened with regulatory matters that made it difficult to transact deals. It was even difficult for agencies to set up arrangements to allow university inventors to manage inventions made with federal support. Each case was reviewed on its own. Some agencies set up patent agreements with patent management organizations that worked with inventors. Some of these agreements were with universities, others with foundations affiliated with universities. It made sense to create a coherent approach to agency interests in federally supported inventions made by investigators at universities.
The reasoning for the implementation of Bayh-Dole appears to have gone something like this:
- IF federal funds are to augment the work of independent university investigators who have shown a keen insight for selecting which inventions to develop, who they choose to partner with, and how to manage the timing of all this, to therefore do things ten times better than government agencies do,
- AND if the existing circumstances for doing this include a range of approaches by various universities–some entirely hands off, some with research foundations, some with contracts with Research Corporation, a few with internal technology licensing offices–
- THEN the implementation should make uniform agency requirements while allowing a diversity of approaches to flourish, with the only common elements that the government may obtain a license to practice federally supported inventions, and may step in to file patent applications if the inventors and their agents decline to do so, and may step in to require licenses or take title if the inventors and their agents fail to develop an invention to the point of practical application, and others appear ready to do so.
The last thing the federal government appeared interested in doing was to impose a single model of ownership on universities. That would have destroyed the very selectivity and diversity that appeared to make university inventors so good at picking winners and working hard to see that these inventions got used.
If the government intended to impose a single model, such as demanding institutional ownership, it could have easily required all universities seeking federal funds to establish an invention ownership policy that included a requirement that all employees assign title to inventions to the institution. Or they could have followed a European model and decreed inventions made by employees to be owned by the employers, regardless of the limitations of federal authority to do so, and heedless of the relationship of faculty to their universities–which clearly is not one of mere employment for the benefit of the employer.
Creating a nice, clean, monopoly practice would have been a foolish thing to do, however, as far as creativity and innovation are concerned. Fortunately, the Bayh-Dole Act is shaped entirely differently, to let university inventors’ relationships with their support communities be what they may be. Doing so makes entirely good sense if the purpose of funding university-hosted research is to gain access to independent talent and see what it discovers and produces. Bayh-Dole does not anticipate a system of monoculture technology transfer offices operating within universities, each claiming outright ownership of every possible invention, releasing only what is worthless, unable to work with each other and unwilling to cede much of anything to industry. Bayh-Dole instead anticipates a diversity of private practice with the possibility of further innovation of that practice, freed from government requirements that it be one way or another. This is truly insightful legislation.
The incredible thing then is that autocrats have come to dominate university policy, and have exploited the liberty of Bayh-Dole to transform an institutional support community into a dominating administrative system, one in which faculty are dutiful workers discovering new things for the financial benefit of the university, motivated by the idea that the university will share (“generously”) some bit of that financial windfall with them personally. That condition, for the autocrats, is progress, and that is what the AUTM licensing survey statistics track–the transfer of inventions that would have been deployed previously in any number of ways into ones now “managed” by institutions. AUTM records the “capture” of federally supported inventions by institutions, not productivity of research in improving the quality of life in America.
This position is made forcefully clear by AUTM and other organizations in their amicus briefs to both the CAFC and to the U.S. Supreme Court in Stanford v Roche. They equate the Bayh-Dole Act with the building out of this administrative capture of independent faculty discovery. The brief to the District Court puts it succinctly. The “success” of technology transfer after Bayh-Dole is an administrative success:
The success of the Bayh-Dole Act stems from its carefully crafted scheme: universities and other research institutions receive ownership of federally funded inventions in exchange for a commitment to use their best efforts to commercialize the inventions for the benefit of the public.
The amicus brief is accurate on the point of a “carefully crafted scheme.” But that scheme was not Congress’s scheme: it was a scheme that came afterwards, by misrepresenting the law, its objectives, and the nature of the bargains it proposes. This single statement is wrong on multiple fronts. Universities do not “receive ownership” of federally funded inventions. There is no such “exchange” stated in the law or the standard patent rights clause. There is no commitment to use “best efforts”–there is only “efforts that are reasonable under the circumstances to attract licensees of subject invention that are small business firms…” (37 CFR 401.14(a)(k)(4)). And there is no requirement that the university administrations are the ones to do any of this.
The Bayh-Dole Act normalizes federal agency requirements for procurement of invention rights in research. Bayh-Dole does not normalize how those rights are managed privately. The Act ensures that everyone who might be involved makes a commitment to the government with regard to the government’s rights, and accepts that if anyone involved doesn’t work diligently to promote practical application of an invention, the government can march in to enable such practical application. If there is a bargain, that is it.
The amicus brief then uses its articulation of a “carefully crafted scheme” to argue that if universities do not own outright, then the law itself is at risk:
The fundamental premise of the statute–that clarifying who owns and can license federally funded inventions will encourage their development and exploitation–is now in jeopardy.
This is the “title uncertainty” argument that is being put out now as a call for “present assignments.” It’s the same people, following up on their same plan to capture independent faculty (and student, and staff) work and grind it through their technology transfer offices to make money. They then conflate making money and public benefit: We can make money so long as we ascribe the intention to do so to “public benefit”. Public benefit becomes a motivation to make money, not an outcome of activity. All this is pinned on Bayh-Dole as a roaring success. It’s not Bayh-Dole’s fault, but in the end, it may be Bayh-Dole that pays the price for it.
Lost in all this is that Bayh-Dole has nothing to do with title certainty. The issues that motivated Bayh-Dole are not ones of confusion over who owns title to inventions made with federal funds. Everyone knew that the patent law placed that title with the inventors, and that any transfer of that title to anyone else was a function of written instruments of assignment. That is still true today. If there now is title uncertainty, it has been introduced by the uniformly wretched university policy statements and invention agreements created by university bureaucrats. Bayh-Dole does not require any such policies or agreements, and certainly does not mandate that they be uniformly wretched. It requires only the (f)(2) agreement, with its three simple obligations–report subject inventions, assist with patent applications, establish government rights.
The autocrats made their inroad by claiming that Bayh-Dole required university ownership. This claim went largely unchallenged. The claim starts a progression toward administrative control of all university-hosted inventions, in the name of consistency, orderliness, managing “conflict of interest”, money, risk, superior resources or diligence or expertise. An impressive list, if one believes these are the virtues of research and innovation. But in this list, is there anything that sparks the imagination as supporting cutting edge creativity? Hardly. These are the virtues of bureaucracy. Yet when law professors discuss Stanford v Roche, it seems all they can do is find ways to help the autocrats complete the institutional, bureaucratic pattern of ownership that has been set up. Where are the law professors that value diversity, freedom, limited intervention by government, and a certain degree of self-determination when it comes to research, discovery, and innovation? Ah, yes, I see.
The amicus brief hammers home the idea that Bayh-Dole was about building up an administrative infrastructure in a footnote: “As one study puts it, ‘growth during the 1970s in patenting, licensing, licensing income, or in the establishment of independent technology transfer offices, was dwarfed by the surge in all of these activities after 1981.” The amicus brief does not bother to point out that in the 1970s, the universities were not generally involved, the agents operating on behalf of faculty inventors were highly selective, and the great majority of inventive stuff moved into public use directly (esp if funded by the NSF) or into government agency management (if a mission-directed agency). The studies cited do not show that government funded research was more productive after Bayh-Dole, but merely that universities moved from a supportive role to a dominating one, that they became less selective in what they claimed, and that an American model based largely on broad public access to research discovery was brought, over the course of 25 years, behind a licensing paywall managed by university administrators. The great advantage of university management cited by the proponents of Bayh-Dole, that of something like a 30% licensing rate compared to the federal agencies’ 4%, depended on selectivity, specialized invention agents, and faculty goodwill.
The “carefully crafted scheme” replaced these attributes with comprehensive claims, generalized administrative offices, and an obligation to assign under the threat of job action or criminal proceedings. The “growth” of university “technology transfer” tracks the capture by administration what once was independent of university controls. This outcome is not among the stated objectives of the Bayh-Dole Act, but it clearly is the objective of AUTM and other organizations that front for university administrators.
We can summarize the sequence involved in the administrative sequence of bureaucratic capture of research inventions at universities:
1. University ownership only when required or offered, otherwise use agents
Bayh-Dole Act passed
2. Bayh-Dole requires university ownership (untrue but effective trigger)
*Bayh-Dole requires “commercialization” (also not true)
3. University build out of “technology transfer”
*expenditures to “comply”
*expansion of policy to “make efficient”
4. Conflict of interest transformed
*any personal benefit steals from institutional benefit
*personal benefit without institutional benefit is unfair
Stanford v Roche: Bayh-Dole does not require ownership or commercialization
5. University response: make ownership a condition of employment
*present assignments
*no review for interest
*capture consulting and collaboration
Stanford v Roche therefore represents a double threat to the autocrats. First, it removes the rationale for university ownership and the mechanism for that ownership. Federal policy does not mandate university ownership, and federal law does not vest that ownership or require assignment.
There is a scramble on now to create a new rationale and mechanism. A present assignment as a condition of employment appears to get at both, but it actually fails both regard to employment and in the context of university policies on instruction, publication, academic freedom, and open laboratories. The autocrats are desperate to preserve the administrative tower they have built, even to the point of being willing to undermine the faculty freedom, initiative, and goodwill that demonstrated their capacity for discovery, the importance of their independence (underscored by Vannevar Bush), and the fact that university faculty, acting outside of university administrative demands, created the practice of technology transfer in the first place–the idea of using patents to promote research and industry adoption, the creation of foundations as invention management organizations, the idea of establishing a collaboration with industry over innovation.
The carefully crafted scheme has run its course. Stanford v Roche is its Waterloo (and ironically, it is the University of Waterloo, in Canada, that shows how things can be done without an arbitrary institutional demand for ownership of inventions).
University faculty and administrators alike now have a chance to step out from under its monopoly-like controls. America has the chance to rediscover the diversity of practices that made university research so exciting in the 1960s and 70s, and add new practices. University administrators can step up and be leaders in this transformation. Keep the old process-based technology transfer infrastructure in place, but snip the requirement that anyone use it, so long as university inventors use invention management agents unless sponsors and the university concur otherwise. That is, make the university’s direct involvement in patent management investigator-voluntary. That was how university invention licensing came to be such a respected activity, and how it can rise again to that standing. Bayh-Dole still serves an important role, but it is up to university faculty and administrators to meet their end of the bargain and value the freedom to innovate over institutional self-interest.
[updated 4/5/13 to fix broken link and distinguish briefs to the CAFC and Supreme Courts.]