I came across an email from November 2009 that rather lays out the heart of the public debate around free agency. It’s from Joe Allen to Robert Hardy (at COGR) and Howard Bremer (long time at WARF). Robert Hardy has sent Joe an email that’s concerned about a “potentially dangerous” “line of thinking” as set out in a post I made on Techno-L (a discussion service since acquired by AUTM) and posted here in slightly revised form (and now extended). Hardy appears to agree with my assessment that the (f)(2) agreement runs to disclosure not assignment, and asks Allen for legislative support for the idea of vesting.
Allen’s response is about how to make sure that Bayh-Dole gets interpreted to support university claims to ownership of inventions, following Stanford’s and AUTM’s arguments in Stanford v. Roche. The email is from 2009, well before the Supreme Court decision made clear that my “convoluted” “theory” “quoting various statues [sic] and regs” actually is what the Bayh-Dole Act requires.
Allen worries that reading Bayh-Dole and its implementing regulations the way they are written is an “alternative universe” that has to be stamped out:
Since his theory is so convoluted, quoting various statues and regs, it can sound plausible to the uninitiated –or to “evil doers” with their own agendas. This may be another case where the Three Amigo’s (Howard, Norm and me) need to saddle up and ride into Dodge City to run out the bad guys. Luckily, since the drafters of Bayh-Dole are still around, that should have some clout.
At least the agenda is clear–there are good guys and “evil doers”, and the bad guys have “their own agendas”. Well now. The agenda here at Research Enterprise is to get it right with regard to the wide range of ways that university faculty, staff, and students can participate in, catalyze, and succeed in discovery and innovation. If institutional ownership of all research assets funded by the government was clearly the one and only way to accomplish this, then Research Enterprise would aim to show how this was indeed the case. But clearly there is lots of room for alternatives, not the least of which is that prior to Bayh-Dole, most universities did not claim ownership of faculty research, period. And we did all right with innovation.
Allen’s email declares (his emphasis):
What these folks skip over (or probably don’t even know) is that B-D is a statutory form of the IPA program, which arose from the premise that universities with an established tech transfer capability–i.e, professionals in licensing, not individual researchers, could be trusted to own and manage their inventions.
The point about IPAs was of course well known to Research Enterprise. But let’s get at the substance here, and that has to do with the premise for Bayh-Dole. Allen argues that Bayh-Dole gives institutions ownership of inventions because they employ “professionals in licensing” who “could be trusted to own and manage” inventions. This may have indeed been the premise in Allen’s mind, and motivated his efforts to get Bayh-Dole passed. The Three Amigos may have believed this. However, it is not the law.
1) Bayh-Dole does not require a university contractor to have a technology transfer office or licensing professionals or even a patent policy.
2) Nothing in Bayh-Dole passes ownership of inventions made with federal support to the contractor–and it would have made no public policy sense to do so at the time since many universities expected affiliated research foundations, not university administrators, to deal with faculty on inventions.
3) The IPAs depended on an invention management agent obtaining ownership of an invention and then working out with a federal agency the basis on which the invention management agent could “retain” ownership of the invention over agency claims to ownership.
Now, for a plain English expression for how Bayh-Dole operates.
Bayh-Dole makes uniform federal agencies’ policies on claims to inventions made with federal support at universities.
Bayh-Dole authorizes the creation of standard patent rights clauses to be included in funding agreements with universities. It is by means of these patent rights clauses that universities agree to the conditions established by Bayh-Dole. As part of this agreement, universities agree to require their research personnel to make certain written commitments, personally, to the government with regard to inventions.
The standard patent rights clauses pre-approve assignments made by inventors to any qualified invention management organization, provided that organization meets the conditions specified in the patent rights clause of the funding agreement governing the invention.
The university as contractor is pre-approved for such assignments, regardless of its capabilities to manage inventions. Other assignees are pre-approved if they have as a primary function the management of inventions.
If the inventors do not assign their inventions, or wish to assign to anyone that is not pre-approved, then the funding agency must approve. If the agency allows the inventors to retain ownership, then the agency is to treat the inventors as if they were a special class of small business contractor rather than as part of a non-profit university.
Bayh-Dole does not vest title to inventions with the university. It does not require universities (or nonprofits, or small businesses) to have licensing professionals. It does not premise its pre-approval on a requirement that universities have technology transfer practices. It does not even assume such a thing–given that at the time of Bayh-Dole’s passage, most universities did not have a technology transfer office with a licensing function, or even an affiliated research foundation, and used instead services of companies like Research Corporation.
When the University of Washington set up its technology transfer office in response to Bayh-Dole in 1981, the “Office of Technology Transfer” was designed as Research Corporation had long advocated–to gather invention disclosures and pass them to an external invention management agent, which in the case of Washington was the newly formed Washington Research Foundation, an independent, university-affiliated foundation built on the models of Research Corporation and the Wisconsin Alumni Research Foundation). If Bayh-Dole were to vest ownership of federally supported inventions with universities on the premise that universities had offices with licensing professionals in them, then it was a horribly drafted law, since it made no requirements for compliance with such a premise. But Bayh-Dole makes no such requirements. It is a reasonably drafted, if rather obtuse, law.
Perhaps what Joe Allen means is that he and others intended the Bayh-Dole Act to motivate universities to convert their technology transfer offices into licensing operations, to be more like WARF and do less business with Research Corporation. No doubt he and others envisioned the growth of a new licensing industry built around the opportunities created by the law, not the least of which would accrue to organizations like WARF, which was a leading advocate for the law. Having a lot more licensing expertise available to help university inventors across the country is a very good thing. As a 20-year veteran of university technology transfer work, I am a beneficiary of that vision, and while I don’t take all that well to Joe Allen’s treatment of my work, I still appreciate his work and that of others to make Bayh-Dole a reality.
The issue, however, is not personal allegiances or political jockeying, but what Bayh-Dole actually does–legally does–and what policies will benefit creative folks in the here and now of university research and entrepreneurship. If we drill down into Bayh-Dole and its implementing regulations, we necessarily have to cite “statutes and regulations”. It comes with the territory. Anyone who wants to get on with this field has to do it. Everything else is hearsay, folk wisdom, and urban legend. Joe Allen and others are now the proprietors of one such urban legend. It is one they apparently truly believe, and have got others to believe also. Their account claims that their purpose behind Bayh-Dole was to get invention ownership into the hands of universities, and for that it vested that ownership with the universities. They ascribe that purpose to Bayh-Dole itself, though that purpose does not show up in the objectives of the law at 35 USC 200. As Joe Allen so clearly puts it, individual university researchers could not “be trusted to own and manage their inventions.” That is what I have labeled “inventor loathing”. Joe Allen makes distrust of inventors the core motivation for his vision for Bayh-Dole. This inventor loathing has been taken up and repeated by university licensing offices. It has been really very damaging to the public reception to university research.
Nothing in Bayh-Dole, however, expresses such a distrust for inventors. Broadly, Bayh-Dole is of the form of a pre-approval for private management of inventions, with an extensive apparatus providing protections for agencies, industry, small business, and the public in the form of obligations to report, to file patent applications, to promote inventions for public use, to grant the government rights, and permit government intervention when indicated. If anything, Bayh-Dole offers protections for inventors and the public relative to a certain distrust of universities in their management of inventions assigned to them–which if you will forgive me a reference to a regulation, shows up as section (k) of 37 CFR 401.14(a). If one compares the requirements imposed on university ownership in 37 CFR 401.14(a) with the minimal requirements on inventors when they retain ownership in 37 CFR 401.9, it is clear that inventors, when their ownership is approved by an agency, are *more trusted* than universities, and even *more trusted* than small business contractors, if requiring fewer protections is any indication. One might therefore argue that of all the possible owners of inventions under Bayh-Dole, university inventors are the most trusted by the Act.
Bayh-Dole allows individual inventors to assign their rights to a wide range of private invention management organizations. This is the genius of Bayh-Dole. Bayh-Dole is, at its heart, a law of invention free-agency for the creative class. Bayh-Dole made this opportunity available thirty years ago in a context in which most universities did not claim ownership of faculty scholarship, including inventions (which are also often scholarship). Bayh-Dole respected the environment in which faculty worked, with conditions that varied from university to university. Some had no interest in IP. A few required assignment. Many directed assignment to an external agent or encouraged development outside of the university. It makes sense that Bayh-Dole would not dictate as law of the land that the personal property of independent university scholars be handed over to the scholars’ employers as a condition of receiving the benefit of federal support for work the scholars, not the employers, had proposed to the government. To do so would have been to merely recreate the federal system of procurement of inventions and accumulation of patents, but now fragmented yet further and put into the hands of organizations without any requirement for policy, competency, resources, public accountability, or preference for non-exclusive licensing. To do so also would have raised issues of eminent domain (taking of private property for public use), requiring due process and just compensation.
What Joe Allen and others envision is a substitute idea. If universities are to exploit Bayh-Dole and its freedoms to their best advantage, then they should set up licensing offices and demand ownership of all inventions they can. This substitute idea is not in Bayh-Dole. It is a great instance of what is called “institutional capture”: orient the institution to exploit available government resources to its best advantage–even if such exploitation comes at the expense of the stated goals of the government program, and comes in competition with the interests of others, which may be weaker or less well positioned to take advantage of the government program–and therefore more likely in need of the program and better stand to benefit from it.
One might then be able to see the agenda Joe Allen and others have set for the now $40b a year in federal research support to university faculty. Their goal is to find ways to strip faculty of invention ownership as expeditiously as possible, and hand that ownership to university technology licensing offices, which will then exploit these inventions to make money for their institutions (and their operations) through royalty-bearing patent licenses. It is little wonder, then, that AUTM collects and publishes licensing information that champions the importance and growth of university technology transfer offices rather than the outcomes of federally supported research. The point of the “metrics” is to chart the growth of this proxy idea, of how an institutional industry is forming around new legislation.
Bayh-Dole envisions something rather different. It envisions an environment in which faculty (and other) university inventors can make choices about who manages their inventions. Bayh-Dole does not care whether a university takes ownership of an invention, or whether the arrangement by which it takes ownership is by the choice of the inventor or as university requirement. What Bayh-Dole cares about is that there is a private arrangement, and that it meets the conditions for uniform pre-approval by federal agencies, so the agencies do not have to fuss over each and every such transaction and be responsible for each assignment if it fails to result in anything being practically applied.
Bayh-Dole makes uniform government pre-approval of university inventors’ choice of invention management agents when their inventions are federally supported.
That is the core of it, in plain language anyone can understand. The rest of the Act is apparatus to implement this fundamental premise. Bayh-Dole is a free-agent model.
By contrast, the Joe Allen version of Bayh-Dole has become a shadow of the previous federal agency patent accumulation model, emphasizing the same outright institutional claim to invention ownership and the same administrative desire for order and self-interest. This shadow model, the “secret fingers crossed” version of Bayh-Dole, has had a silent, substantial, adverse effect on research asset management, for all the claims made for its success. The stark reality–and anyone involved for more than a few years in university licensing knows–is that most research inventions claimed by universities are moribund. If licensed to companies, they are not developed into products. If licensed to university startup companies, those companies do not obtain sufficient funding. But most are not ever licensed, and in not being licensed the patent rights covering these inventions stand not as unrequited offers but as institutional threats to anyone who would use technology covered by their claims but is not willing to take a license that commits to developing a commercial product and paying royalties for the right to do so.
The number of patents issuing to universities has little to do with the public impact of federally supported research. The number of licenses universities have granted, in aggregate, also has little to do with that impact. Even the licensing revenue from a very few lucrative deals also means little for that overall impact or for the disposition of all the moribund research assets caught up in institutional claims of ownership.
It is this problem that various observers–ones who really do want university licensing offices to succeed–have been trying to point out. It is this problem that Joe Allen and AUTM leaders are unwilling to address and would just as soon throttle, or chase out of Dodge City. Joe Allen’s vision is of an institutional monoculture of invention practice, run by administrators efficiently capturing the research output of anyone involved with university research and making money on some few deals per decade. The bulk of WARF’s licensing income has come from four inventions over the course of 80 years. Stanford has had three really lucrative inventions in about 40 years. The University of California holds some 5,000 patents and 5 of these account for half of its current licensing income, and 25 account for nearly three quarters.
Whether a university obtains ownership of inventions made with federal support is not the big issue. How the university obtains that ownership, however, does matter. A voluntary deal is way better than a compulsory one. But all this aside, the issue that matters most is separating institutional ownership from institutional control. A university may come to own an invention, but that ownership need not dictate the choice of agent to manage the invention. This concept is crucial to intellectual property management, and in particular to working with inventions. Bayh-Dole provides for private choice of agent, so long as the invention rights are assigned by the inventors. It is the act of assignment to an invention management agent or to the contractor that triggers the pre-approval apparatus. Otherwise, everything is back to agencies dealing with inventors, as in the days prior to Bayh-Dole. The contractor, in turn, may not be an invention management agent at all, and may not have a technology transfer office with licensing professionals. Small companies, for instance, are also contractors under Bayh-Dole patent rights clauses, and generally do not have such offices.
For university administrators, this separation of ownership and management agent should be liberating. It provides universities with ways to manage the separation of academic and commercial interests without creating huge institutional conflicts of interest by trying to do all the patent and license work directly. Separation of ownership and control provides universities with relief from having to deal with every invention across all areas of endeavor, whether chasing money or dutifully playing out a policy requirement to treat everyone the same, even if that means foreclosing opportunities, imposing needless costs and delays, and incurring undesirable institutional liabilities. In short, agent choice allows university licensing offices to focus on what they are good at, work with inventors that have chosen them, and devote greater resources to promotion of inventions and less to compliance and processing inventions.
While university inventors may not have all the skills of a “licensing professional”, they may well have–and often do have–the best insights into what sort of invention management agent will work best for their inventions, their research, their collaborators, and their objectives. A university licensing office cannot serve everyone. It just isn’t possible. This is not a criticism of licensing folks. It is a realistic assessment of the situation. University research is far too diverse in its subject areas, its goals, and its opportunities. The very things that make university research a national asset–its independence, its access to top-notch resources, its informality of collaboration, its mandate to publish and provide public access–are things that work against a single licensing office being able to work with everything produced at a given university. Bayh-Dole sets up federal funding agreements to reflect these properties. It does not require a comprehensive claim to everything inventive, and doesn’t even say that doing so would be a good thing for universities to do. Bayh-Dole says, instead, to let inventors choose their invention management agents, so long as those agents agree to some basic rules.
A university licensing office that develops expertise and working relationships with industry in a given area is a national resource. It should expect to receive work from around the world to place new technology in that area of expertise. It should expect to refer work that arises at its host university to other agents when the subject matter is not its immediate expertise, and is for someone else. All this can happen and still advocate for regional economic interests, institutional money interests, and matters having to do with institutional control and liability angst. To make each licensing office serve only institutional self-interest is a tremendous limitation on what is possible, and from a bit of distance seems provincial, if not institutionally selfish, if not self-destructive.
The idea that universities should create policies that eliminate that choice runs against Bayh-Dole. It is perfectly appropriate for a university to create a policy that moves patent issues away from instruction and research, and requires university employees who claim an invention to assign their inventions to an invention management agent or separate from university employment to pursue their commercial interests. It even works for a university policy to require assignment of inventions to the institution, if the inventors still retain the right to choose a qualified agent to work with them on managing rights and can negotiate the terms of that management. That agent does not have to be the university’s own administration. Something similar happens all the time for copyright–faculty create books, software, and data, and regardless of institutional ownership, decide where it is to be published and how and when. Faculty appear to be quite capable of dealing with copyrights, with authorship issues, with multiple collaborators even across organizations. While the details and expense of patenting often require the services of a patent agent or attorney, copyright licensing can be every bit as complicated as patent licensing. It is good to have quality help available.
Free-agency is not, then, a “dangerous line of thinking”. It is a liberating line of thinking. It is the line of thinking that Bayh-Dole acknowledges and supports. It is only dangerous to a vision rooted in using federal law to secure special privileges for patent administrators at universities. Free-agent models are a boon to university-based licensing operations. University licensing on the agent choice model flourished pre-Bayh-Dole, and was in large part the motivating reason for Bayh-Dole. Times have changed, but not so much. Freedom and choice are still a better foundation for innovation policy than is compulsory institutional capture, no matter how sincerely and kindly that capture is administrated. It’s time to re-open university practice around Bayh-Dole. A good start is to restore to university inventors their choice of invention management agent, the native condition established by the Bayh-Dole Act.