I’ve put together a graphic that shows a cascade of possible places where a university and faculty might consider the matter of ownership of inventions and works of authorship.
I’ve arranged things into various rows, each with a corresponding letter for reference. The cascade starts with everything faculty might do (a), and divides these activities in ones involving university resources or not (b), with the assumption that consulting and outside work do not use university resources. The next row (c) identifies work that is directed by the university, commissioned by the university in exchange for additional compensation or other benefit, and self-directed work, which includes research, conference talks, and the like, some of which may be with university resources, and some not. Row (d) shows work that is generally considered to be university owned as a matter of employment, law (such as work for hire), and contract (such as sponsored research agreements with a university ownership requirement). The last row (e) calls out work that is commissioned, which involves a voluntary agreement on ownership, and self-directed work that is voluntarily offered to the university, and for which terms might be negotiated.
The blue circle shows where the “scope of employment” is for faculty as employees of administrators, capturing those directed duties that are for the administration, as distinct from duties that are directed to benefit, say, students, such as advising.
The pinkish oval takes in a set of conventional mechanisms by which a university acquires ownership of IP outside the limited range of “scope of employment”. These areas cover express conditions made outside normal employment as well as voluntary assignment, which at one time was the dominant practice. Conditions outside normal employment include commissioned work, in which the university can set out its ownership expectations and faculty may choose to accept those conditions or not, without any consequence to employment status. Also included are departmental research and sponsored research with ownership requirements. Departmental research conditions are established by the university, typically involving access to facilities or resources that carry restrictions on use imposed by donors, sponsors, or laws such as tax laws applicable to tax-free bonds used to support construction.
Sponsored research may also involve restrictions, which may be imposed by sponsors, or as in recent practice, imposed on sponsors by administrators as a way of also imposing those restrictions on faculty investigators. Sponsored research itself, however, is also voluntary for faculty. Faculty members are not usually directed by administrators to participate in any particular extramural project. After a faculty member achieves tenure, sponsored research participation is not even a matter of keeping one’s job. Faculty generally propose extramural research, and the university generally has a process by which faculty are approved to participate–which shifts a portion of their salary from the university to the grant funds.
The above diagram is by no means complete. It is meant to show a general geography in some of its complexity. There are further details not represented. For instance, consulting might use university resources (and with the approval of the university), and might also take advantage of expertise or data gained with university resources, even though the consulting does not use those resources.
The point of the diagram is to try to show the differences between various strategies for administrative ownership of inventions. The conventional architecture was to take voluntary assignment, take assignment when required, and take assignment when there is an express arrangement beyond normal duties. This approach has all but disappeared in most research universities. It is an architecture, however, that is wonderfully suited to a university’s role as logistics support for research without taking any active role in asserting any claims on faculty as a condition of administrative authority.
In such an approach, a policy may still require disclosure of inventions more broadly so the university may deal with possible differences of interpretation of sponsored research awards and may manage potential conflicts of interest that involve the academic community. One may call this an institutional “scope of interest” in inventions.
The reason behind trying to sort out these various aspects is to put some context around the expansion of the scope of interest, claims for scope of employment intruding into self-directed work and work done using university resources or arising from that work, and finally, an equating of an expanded scope of interest with a claim of institutional ownership. IP architectures that claim everything and release the worthless tend toward this expanded scope of interest.
We can briefly work through these expansions of institutional claim. The first is to claim inventions because they involve the use of university resources. This is a tough one, because it is not at all clear that in all cases an invention arises because there is use of resources. For instance, a research effort may be made to understand how a particular type of cell behaves, and in doing so, a researcher could imagine a new kind of sensor. The work in the lab involved cells. The invention involves something that’s not in the lab. While the invention arises in the context of using the lab, it’s not what the lab was used for. It might make more sense to consider whether actual reduction practice takes place with university resources. Even here, there is no particular reason to claim ownership of inventions for use. University copyright policies, for instance, may require faculty to reimburse the university for use of facilities (such as computer labs) in creating personal work (scholarship, artistic works) without also making a claim to own resulting IP.
The second expansion is from use of facilities to inventions arising from that use. Thus, the sensor idea in the previous example could be argued to arise from the work in the lab. The sensor idea could arise, as well, from any number of other activities, such as outside consulting work at a company that makes sensors, but the lab and the inventive idea, even accidentally coincident, could create an institutional claim that the work in the lab led to the invention. Similarly, work of the same nature as the laboratory work–whether cell study or sensor study–conducted outside the lab, such as in consulting, could give rise to an institutional claim. Certainly some broadly written university IP policies make such a gesture.
The third expansion is in the “scope of employment”. For IP “employment” is not necessarily equivalent to what an HR department might say is an employee’s job description. The effort of this expansion is to claim that anything faculty may do as faculty is within the scope of employment, unless the university provides an express release of its interest. Thus, outside consulting in some policies cannot include IP assignment conditions that would take precedence over university claims–which is dramatically beyond a reasonable restriction, say, that faculty involved in sponsored research with obligations to the university (or sponsor) cannot contract with others for those same IP rights.
These three areas of expansion lead to questions of how they hold up in an open laboratory, limited supervision, largely self-directed faculty activity. All depend on some sort of review to establish the circumstances that might control a decision as to ownership. Was an invention made using university resources? Do those resources include sponsor-provided funding? How does sponsor funding become university funding? How does work on behalf of a research sponsor, or for the benefit of others, become work for the university’s administration? One answer is quite simple: these things happen because a policy statement asserts them. Yet the mere assertion in a policy statement does not provide guidance as to actual circumstances.
Similarly with the idea of “scope of employment”. Faculty are appointed. Often they have no job description, as administrative employees might have. They are faculty, after all, and do what faculty do–teach and study, publish and lecture. They choose what to study, and where to study it. They choose their sources of funding, and their collaborators. They choose what to publish, and when, and in what forum. In many ways, faculty decide much of what constitutes the scope of their employment. Aside from expressly directed duties, most everything else comes within the scope of employment when a faculty member decides it does, not when an administrator demands it. It is the scope of employment, not the fact of employment, that shapes IP ownership and rights determinations. This is a tough lesson for administrators who are not grounded in IP. A work for hire, for instance, is made by an employee acting within the scope of his or her employment, with a multi-point test for agency to determine if the employer is the author, or the employee. The “employment” is not merely the paying of salary, but includes whether the employer supervised and directed the work, approved it and required the employer’s resources to be used in preparing it. That sort of thing. That is, for a work to be made for hire, the employer has to author the work. That’s hard to imagine for most faculty works of authorship–not only because administrators do not direct and approve such work, but also because it makes no sense for academic freedom and independent scholarship for administrators to seek to control, let alone own, the substance of what faculty write and publish.
A similar argument, with variations for patent law conventions may be developed for faculty inventions arising in self-directed work. Yes, a demand can be made based on use of resources, if that use can be objectively established, and that demand can be embedded in an employment agreement. Employers make such demands all the time. The question is why a university, and especially public universities, should make this sort of expansive demand, especially given their history of not making such demands.
Resources that once were provided for public use, meaning for the general use of faculty as members of the public, have now become more privileged than privately held resources. The public mission of the university has become the self-interest of a public institution. A university adopting a comprehensive, compulsory claim to ownership makes the claim that faculty are not engaged in public service when doing their work as faculty, but rather are working on behalf of the university and its administration. It’s an odd idea, but not impossible.
Why have these expansions happened? My argument is that the mischaracterization of Bayh-Dole as a vesting statute has played a pivotal role. For thirty years, commentators claimed that Bayh-Dole required universities to take title to federally supported inventions, or some variant of this, such as that title vests outright with the university not the inventors or that inventors are forced to offer title to the university. Such claims were rejected by the Supreme Court in Stanford v Roche. We live in a new world of IP, which is actually an old world that once was well established, but now carries the ruins of policy built on the mischaracterization of Bayh-Dole.
For thirty years, administrators relied on “required by law or contract” to claim the majority of inventions made in self-directed research, and all made with federal funding. At many universities, federal funding is two-thirds or more of all extramural funding. It was this claim that led to the build out of technology transfer offices to handle the volume of disclosures–which would be quite simple if the university merely sent them along to the funding agency and allowed the inventors to decide what agent they might use, or asked the funding agency to allow them to retain ownership of their inventions.
The idea of federal funding providing a means for administrators to claim title was appealing, and it was sold to administrators that Bayh-Dole was a law meant for them. While the implementing regulations for Bayh-Dole anticipate that contractors might hold back some inventions and claim they weren’t federally supported, and provided examples intended to clarify what was within the “planned and committed” work under a grant and what was merely closely related but not subject to the standard patent rights clause, university administrators did just the opposite. They expanded the claims of federal funding–“just in case” they said–because no one wanted to be accused of breaking federal law. Yet, as is now clear, Bayh-Dole made no requirement for university ownership, nor did the standard patent rights clause. But the infrastructure of expansion remains. Because so much research at most universities does involve federal funding, an expansive claim on work that benefits from or arises from such funding means that administrators can claim ownership of just about everything, and put the burden of proof on inventors to show why a claim is not justified, and further put the burden of proof on the inventors to show that they are not working against the best interests of the institution by objecting to institutional claims.
This claim to university ownership of federally supported inventions also led to a substantial restriction on the flexibility of industry supported research and collaborations. To comply with federal law, university administrators argued, universities cannot allow companies to have outright ownership of inventions made with their support, or even a right to negotiate an exclusive license, without a thorough review for possible involvement of federal funding–which of course was everywhere. With their expanded, “just in case” arguments for federal funding, administrators necessarily were forced by their own logic to restrict industry interest in new technology created with industry funds. This self-imposed restriction, blamed on “Bayh-Dole”, has caused no end of misunderstanding, cross-purposes, and bitterness in university-industry relations.
The irony, of course, is that an objective of Bayh-Dole is to encourage collaboration between universities and industry–indeed, this is the only objective in 35 USC 200 that mentions universities. It would appear under Bayh-Dole that the federal government is quite pleased if inventions made with federal support come into the hands of American industry, provided that the inventions are practically applied and the benefits of doing so are made available to the public on “reasonable terms” (see the definition of “practical application” at 35 USC 201(f)). Instead of collaboration that accomplished such transfer of inventions to companies, university administrators insisted that institutional compliance with Bayh-Dole prevented this very thing, and in its place substituted a process of institutional claims to title, patenting, and efforts to license, preferably to a company willing to take a monopoly position to create commercial product.
In this way, Bayh-Dole–through no real fault of its own–has come to stand for the disruption of university-industry collaboration rather than the encourager of such collaboration. The mischaracterization of Bayh-Dole as a vesting statute has humped up administrator claims to invention ownership such that ownership for institutional profit has come to be construed as a public good, because it appeared to be endorsed by federal law and federal research innovation policy. All of this is simply untrue, however real it has become in policy and in the fortunes of inventors and their institutions. Bayh-Dole does not require university ownership, does not introduce ownership barriers to collaboration with industry, and does not penalize universities for allowing inventions to fall into the hands of industry to be–shock of shocks–developed and used.
The upshot, then, is that university administrators are now faced with a choice. They could return to pre-Bayh-Dole practices, and open up their technology transfer offices to a broader set of voluntary collaborations with faculty, with industry, and with invention management agents. Or they could attempt to salvage the existing infrastructure of expansive claims, conveniently defensive positions relative to industry, and policy buildout. The retrenchment movement clearly favors this latter alternative: keep the expansive claims, the stick-it-to-industry mindset, and the faculty “educational materials” on intellectual property just as they are, and find new reasons to prop the current approach up, if Bayh-Dole won’t do.
This then is the second great irony, that in advancing the use of present assignments, as a condition of employment, in an effort to permit the university to attempt to cleverly outwit, if not double-cross, companies that have bargained for invention rights, university administrators cite Stanford v Roche. Even though present assignments don’t work the way administrators claim, and even though law professors and practicing attorneys (generally those representing university administrator interests) don’t care to follow how Bayh-Dole operates, or how faculty are affiliated with universities, it’s clear that if one is going to salvage the present technology licensing infrastructure, one really does have to do something because of Stanford v Roche. That doing appears to be to shift the rationale for tying up as much creative work as possible from Bayh-Dole and compliance with federal law to a simple assertion of ownership as a condition of employment, or use of resources.
There is no thought that the present policy architecture–comprehensive and compulsory and unthinking–might be a poor one for scholarship, instruction, research, collaboration, innovation, or economic impact. There is no thought that past practices, successful by being selective, voluntary, and faculty-initiated, might yet be a guide to present practices. There is no thought to the distinctive position of university faculty as independent actors in a research world dominated by institutional interests. There is not thought that a comprehensive and compulsory IP architecture might frustrate the objectives of Bayh-Dole, and in so doing, cause the federal government, and industry, and the public, to look elsewhere to place research funding, ending the great agenda advocated by Vannevar Bush.
The role of universities as hosts for independent research, collaborations of all sorts, and trading centers for invention management agents, companies, and foundations hangs in the balance. My argument is that Stanford v Roche allows universities to unwind the tight-fisted policies that increasingly stand in the way of technology transfer. If there ever was a time to do it, and not have to apologize for the past thirty years, it’s now. Wait a few more years, add a bit of present assignment language, tie everything down, and the university research system will rupture or stagnate. That’s what this debate is all about. Faculty freedom or institutional self-interest–which will it be?