Yesterday I read through the amicus curae brief filed by WARF, AUTM, UC, and others. I thought about it on the commute home, and read it over some more last night and early this morning. I’m bothered, and not just that it asserts title vests with universities and not with inventors. What bothers me is how poorly it lays out this argument. But more so, how poorly it handles the significance of the argument to university innovation practice. That’s what I’ll deal with here. This is longer than I’d like.
The short form is: academic freedom would argue for more choice rather than less to be distributed to the inventors; arguing from extreme cases does not make a compelling case for change, especially when even more proximate and extreme cases can be cited for university-based ownership; Bayh-Dole permits a variety of local choices regarding the disposition of inventions, and the brief appears to disregard these with a melodramatic argument for convenience in which one model–that of “commercialization” of inventions to generate royalties–dominates all other possible uses of the patent system to meet the objectives stated by the law. Without a discussion of these other uses, and the role of principal investigators and faculty inventors in the life of their inventive work, I don’t see how Bayh-Dole can live up to its potential.
Having gotten things taken care of in summary form, if you want to continue on, at least you were warned.
Let’s look at the closing argument of the brief first. “Finally, apart from its economic consequences, the panel opinion threatens academic freedom. As the law stands, if a company entices a faculty member to sign an agreement assigning away ownership of inventions, it may obtain a patent and dictate whether he or she–or anyone else at that or any other university–may continue to conduct research in the field.”
Let’s put aside “entice” as unnecessary. The company could be recruited by the faculty inventor, or they could meet randomly at a mixer put on by the university’s “connect” program or business plan competition. And let’s not put much pressure on the difference between “any research in the field” and “research within the scope of the claimed invention once the patent issues some years thereafter.” That’s just loose language for dramatic effect, let’s say, not part of the actual argument.
How can faculty ownership of their inventions threaten academic freedom? This one has me stumped. Look at university management of copyrights. There, most universities go out of their way to exempt scholarly works from institutional ownership. The argument is, academic freedom is preserved when the scholars decide the content, timing, forum, and consideration for publishing–that is, they enjoy the benefits of ownership in their work. The institution claims ownership only where the work is commissioned, makes use of institutional resources in some abnormal way, or is required by contract.
Perhaps we are on to a strange new version of academic freedom: “Faculty are free when they don’t have to think or do things for themselves.” We can start broadly and look at academic freedom as a freedom from interference by government, industry, or others in the conduct of university research and instruction. This is sort of the worry in the brief. That companies might take candy from babies and this would interfere with the parents’ own interest in the same candy.
We can also consider the academic freedom relative to outside forces that individual faculty and other scholars expect with regard to their teaching, research, and publication. This freedom also might be affected by government (security) and industry (trade secrets) claims. And we might think of the interaction between academics and their own institutional governance.
For what it’s worth, the AAUP focuses its discussion of academic freedom on freedom in research and publication, freedom in the classroom in presenting subjects but not in introducing “controversial matter” that is not relevant to the subject, and free from institutional censorship or discipline for expression made as a private citizen (though potentially held to account for being inaccurate, unrestrained, or disrespectful of others opinions when they are acting in their capacity as institutional officers or members of a “learned profession”. In all of this development, one might think academic freedom arguments would run the other way, that ownership vesting in faculty inventors would preserve academic freedom, and ownership taken away from those producing the invention would appear to be a collusion by government and university administration to interfere in the conduct of faculty research and publication.
So it’s an odd thing, to say that academic freedom is on the line. Look further at this. Outside of Bayh-Dole, faculty generally have broad discretion in what they choose to study, and where and how they do their work. They may consult, within certain constraints, as they please with companies, foundations, or governments. They can sign almost whatever agreements they please, so long as they do not obligate assets that are otherwise obligated to the university or one of its research sponsors. Moving potentially inventive or valuable ideas outside of the university, where subsequent inventions may be owned by a private party, also may jeopardize further research in universities (if one buys the concern set out by the brief), yet universities make no general case to prevent this. They may even offer statements encouraging consulting. In many cases, universities refuse to read faculty consulting contracts or advise faculty on acceptable or prudent terms. This leaves the amicus brief arguing that universities don’t care about the consequences of losing invention assets that universities could most directly claim as work carried out within the scope of academic duties, but if the work is federally sponsored, somehow there is a great threat if rights are held privately rather than vesting outright with the university. The contradiction between practice and the claimed argument about academic freedom is startling in its audacity: universities don’t care about non-Bayh-Dole inventions interfering with research, but they do with federally funded ones. Doesn’t add up, does it?
We might add, Bayh-Dole gives as an objective the building of university-industry collaborations. One might think, faculty consulting with industry to exchange tacit knowledge and move inventive ideas into the hands of people capable of building out useful things would be one way to accomplish this. There are ways under Bayh-Dole for this to happen, in fact. But the amicus brief makes industry holding of patents to be nigh unto evil because it threatens to shut down all of academic research. That this actually has not happened is I guess just a statistical fluke. One might otherwise think of the millions of patents held by industry that presently do not block university research. What might they otherwise block? Perhaps, really, it’s the ability to push out a new inventive technology and have no constraints on its practice. Well, that’s a dream that happens now and then, but not routinely. For every invention that might pass from a faculty inventor to a collaborating company under a consulting agreement, there must be a thousand already there with the same potential for disrupting anyone’s licensing efforts, if those efforts are focused on royalty-bearing licenses covering sales of product. It comes with the territory. An argument from academic freedom would say: more ownership interest to faculty–and perhaps not just to the inventors, but to the principal investigators who are responsible for the choice of funded project, selection of sponsor, expenditure of funds, and reporting on work done. Why not have principal investigators also responsible for inventions made on their watch?
We may observe, also, that some recent studies (and practice experience) indicate that despite the hand-wringing, patent issues have little effect on the conduct of university research, and in any case much less effect than is caused by difficulties in obtaining access to materials and data (which often do not have ownership claims through patent and copyright yet may be managed by universities). That is, if one looks at practice, the disposition of patent rights between universities and industry is rather inconsequential to academic research.
If one wanted to build a case around academic freedom, then one way to do it would be to place invention rights with the principal investigator of the research in which the invention was made. If the university were to have an ownership position, it would be to absorb the cost of filing and take on certain risk in the public interest, but allow the principal investigator and/or the faculty inventors to decide how the patent right would be deployed, relative to the objectives of the grant and their academic sense of the right thing to do. In other words, an argument for academic freedom would involve some arrangement between the university and the research team that gave the team responsibility for the management of the invention and patent rights. That’s pretty much the opposite of the brief’s argument, which is that faculty don’t know what they are signing, so whatever it is has to be preempted by institutional ownership backed by federal law. Think about the disenfranchised principal investigators and faculty inventors in this argument. Academic freedom? Bah.
What the amicus brief might usefully have taken up is whether faculty should have economic freedom with regard to their research inventions. That is, is the integrity of university research (and its claim to a special societal position as independent and focused on accuracy in the public interest), threatened if faculty personally hold patent rights potentially worth millions of dollars and continue to propose new work to the government through their universities? Is it a problem merely that some of them believe their patent rights may be worth millions, even when that is, most of the time, simply not going to happen? That might make for a good public debate on various pressures for financial gain presented to university investigators through consulting and licensing. Donations and research grants, also, can result in personal financial benefits, paying for summer salary, for instance, or additional travel that augments professional standing and development, but nothing like a windfall patent license hitting it big in the product or equity space. In all, an argument based in the integrity of research under the pull of relatively large amounts of money might make a better foundation for protecting the government’s interest in research, and in Bayh-Dole as a means of encouraging universities to engage industry. If the government is going to fund basic research in areas that are deemed to be potentially lucrative, then it makes sense to consider controls that keep the research on task and exclude much of the potential for unfounded claims and speculative dealing in invention rights that preempt broad industry investment to complement government research initiatives.
Let’s turn to a second form of argument used by the amicus brief. The brief develops a scenario of what could happen if four university researchers each enter into personal agreements that created conflicts in claims on intellectual property arising in joint work. “One visits the laboratory of a private company and signs an access/confidentiality agreement with small print addressing invention rights. The second acquires research equipment from another company, unthinkingly signing a form purporting to assign ownership of any inventions using the equipment to that company. The third obtains supplemental research funding under a contract with a third company that also address intellectual property rights. The fourth assigns her rights to a start-up company in which she holds equity. Under the panel’s opinion, it will be virtually impossible to know who has what rights to exploit the invention because any of the inventors’ agreements could override the Bayh-Dole Act’s assignment of ownership to the university.” Here we have the construction of a scenario in which the university fails in many of its functions, but these are ascribed to personal actions by investigators. A university has a responsibility for acquiring equipment used in its facilities, as well as in determining the conditions under which sponsored research is conducted. The rights assigned–even if they are future rights–would only be those of the individual, and not of all those involved, so rather than a fuss over ownership, each organization involved would have a claim on the title held by the individual doing the assignment.
The panel’s decision would create none of the “battle of forms” among the rights holders by assignment. There may be a question about compliance with Bayh-Dole, but that really should start with the university lapses necessarily present in the hypothetical. Why did the university accept government money with a scope of work permitting use of equipment with rights obligated to a company? Why did the university not have a system of reporting personal consulting work that might represent a conflict of interest? Where are the written agreements with investigators required by Bayh-Dole that would, at the time of the federal award, if not prior to any of the private arrangements being contemplated, put investigators on notice of their obligations to protect the government’s interest should they obtain federal funding? However, if the government’s interest is in the use of the patent system to promote the use of federally supported inventions, and in the hypothetical we end up with multiple companies having an interest in such inventions, it’s a bit murky to see how this is adverse to overall government objectives under Bayh-Dole. Companies have access to new inventions! Success!
The only folks on the spot are the university administrators, who clearly screwed up some contracting, created an ineffectual for-show-only conflict of interest program, and failed to conduct an effective program of education and written agreements to protect the government’s interest. The government could look at the situation and void title where the outcomes fail the government’s purpose. The only damage that’s done is to a university expectation that they might make money by throwing their patent ownership weight around. Since universities making money from licensing is a consequence of some forms of licensing, it’s just a matter of losing out on an opportunity by screwing up conditions. It is not a problem of title vesting initially with inventors.
More broadly, the brief does not present any evidence that such conflicting situations indeed occur (though there must be the odd instance), or more importantly, that when such do occur they are not usually resolved by parties recognizing that the various claimed interests are unworkable and that without a vastly superior position to all other claimants, the best way through is to work something out. That is, there are not a lot of cases being litigated with the worst case scenario offered by the brief. The thrust of the argument by hypothetical is: imagine a really complicated situation that rarely if ever happens and only then comes about because a university has screwed up royal and imagine that this situation escalates to a dispute rather than resolves in any number of acceptable ways, and now use that high drama as justification to huddle up and revisit whether Bayh-Dole has anything to say about the case on appeal. Wow. Imagine that one driver of a car had a pregnant cat giving birth, and another was carrying a load of nitroglycerin but had poor brakes, and a third was color blind and talking on a cell phone, and the fourth was… well, you get it. We’d never drive. Someone would have to drive for all of us because it is possible that these four drivers could totally screw up.
But that’s not the end of it. The brief does not bother to point out that just this sort of complex commitment of rights may happen with institutional ownership, and perhaps with some frequency the institutional claim to own makes the potential for problems vastly greater than would personal ownership of inventions. Consider, for instance, obligations made in sponsored research agreements such as first right to negotiate a license, or a promise of a future non-exclusive license, or an obligation to present all findings to the sponsor first for (time-limited) review (for patentable subject matter, yes, yes) prior to publication. These kinds of clauses get accepted all the time by universities. Since university labs may freely collaborate, and do, even across institutional boundaries, it is all too easy for sponsor commitments of the above sort to be become cases of double (or multiple) licensing and be breached. Where an individual holds rights rather than the university, the scope of the obligations is personal, not institution-wide, and the risk of such problems is reduced, as the university can be expected to obligate inventions to which it has a controlling interest.
We can take it further. Consider Apache-style software licensing. In an Apache license, a software developer commits code, copyrights, and associated patents covering the operation of the code to the community of developers. If the developer is an individual, then the rights committed are those that the individual controls. If the developer, however, is a university, then the entire patent portfolio of the university is implicated. One can readily see how an outright claim by the university of ownership of faculty patent rights could make its exposure to downstream problems much greater than if the university allowed, at least in some cases, for personal ownership. These thought-instances don’t prove anything about Bayh-Dole and personal ownership. They do point out how the amicus brief paints one picture and ignores others that might argue that in trying to avoid personal screw ups the institution is any better off.
The aim here is to draw closer to the question of whether the university as a corporate owner of invention rights is, generally, a good thing for the movement of research inventions. In some cases, no doubt it is, and in other cases, it unquestionably is not. The point about Bayh-Dole is not that university choice is decreased–that title vests outright in the university–as the brief would have it. It’s just the opposite. Bayh-Dole provides a common apparatus across funding agencies so that universities can exercise choice in the recognition and management of inventions made in their research environments, not be constrained by incompatible and potentially unresponsive agency requirements. Bayh-Dole provides a university with the choice to manage or not, the choice to assign to another organization or not, and the choice to advocate for inventor management of inventions. The principal investigator also has some choices under Bayh-Dole. The PI can choose to subcontract research to another organization, and can choose what part of the research to place there. Universities almost never resist such subcontracts based on the choice of work. The upshot: the PI can all but decide where inventive work in the research proposal will show up. If not at his/her own university, then at another university, or a private research foundation, or a small company. Under Bayh- Dole (37 CFR 401.14(a)(g)), the host university is out of the picture on inventions made in the subcontract. The deal is between the subcontractor and the funding agency directly. It is these sorts of choices that shape Bayh-Dole–if anything the effect of harmonizing agency requirements is to make the choice of first action not governmental. And for that, a university context might be one thought to represent a good mix of incentives and consideration for the public good.
One point further. Under Bayh-Dole, the university may designate personnel to receive invention disclosures. In 35 USC 202(c)(1) these are “contractor personnel responsible for the administration of patent matters”. These personnel could be PIs. They could be patent attorneys. They could be invention officers at the technology licensing officer or at an affiliated research foundation. The university has a choice under Bayh-Dole regarding who will be responsible. There are other choices: who decides on whether or not to “elect to retain title”? Who decides whether to hold a patent right or assign it to another organization with “one of its primary functions the management of inventions” (37 CFR 401.14(a)(k)(1))? The university could designate these decision points as well in various ways. Again, for instance, designating principal investigators to have this responsibility unless they asked to be relieved of it. Yet we find in the amicus brief apparently arguing against this sort of choice, away from their faculty inventors, based on a constructed fear of individual incapacity to comprehend the import of agreements, invention rights, and the like. Incredible, really, because it appears to be in the spirit and intent of the brief, not an accident in a fumbled argument. I believe, at least, that the drafters of the brief really do want the outcome of TVinC, not something that makes room for additional choices that these folks do not want to allow in their own practice. In a sense, the amicus brief is a melodramatic argument for convenience. The university should be able to claim rights outright and dispense with dealing with its inventors as investigators or faculty and treat them as less than common employees, more like hens in the abstract, laying invention eggs in which they have no further interest. If it does just come down to university convenience, I’m pressed to see why anyone in the legal world would care.
If there’s a point at which Bayh-Dole interpretations matter, it’s not in the little bit of difference between owning outright and obtaining title obligated by written agreement. It’s not about battle of forms among promises to assign and present assignment of future inventions and requirements of federal funding agreement. It’s about the deployment of a selection apparatus that it is hoped will operate to make federally supported inventions more readily available to US industry and build better university-industry relationships. Instead, universities have fixated on a small bit of the many uses of the patent system to focus on product development and company formation, have construed practical application to mean “commercialization generating royalties on sales”, and have developed a robust mythology of excuses for why this doesn’t happen as often as might be expected–lack of “training” of research personnel, a “valley of death” created because not enough stupid, wealthy investors are willing to throw their money away, a diminished “innovation capacity” in industry to appreciate new things and act on them, not enough money for technology transfer to have the staffing necessary to pitch every invention as it deserves. All of these have enough instances in them to be real, but not enough force behind them to be the root causes of the problems in moving research to practice.
Until there’s a decent discussion on how research engages community, patents or not, I don’t see much hope for Bayh-Dole to live up to its potential. There are substantive issues in Stanford v. Roche having to do with the operation of Bayh-Dole. I don’t think, however, the amicus brief has it right on the operation of the vesting of title, and certainly is a long way from a substantive argument about how title other than with the university harms the government’s objectives under Bayh-Dole.