University Patent Policies, Past and Present, Part II

The first part of this essay showed some of the diversity of mid-century university patent policies. There were a number of approaches, from discouraging patenting to embracing it, from university direct control to the use of external agents. Almost all universities used some sort of faculty-led committee to advise on patent decisions.  The general sense was that inventions came to the university rather than the university used its position as employer or trustee for sponsored research fund accounts to also claim inventive work.

Much has changed. Especially since the Bayh-Dole Act was passed in 1980, university patent policies have converged on “require assignment of inventions.” External agents are restricted to affiliated research foundations, and many of the foundations that flourished prior to Bayh-Dole have been replaced by internal administratively run offices on the model of the University of California, Stanford, and MIT. National invention management agents such as Research Corporation have not nearly the position they once had. Their fall appears to have been an intended outcome of some of the advocates for the Bayh-Dole Act and coincides with the idea that a university that hosts inventions, no matter how little it may have contributed to the work leading to an invention, or its subsequent development, should own the invention and seek to gain the lion’s share of any proceeds from patents taken on the invention.

The transformation has taken other forms as well. Where in mid-century practice, inventions were referred to the university and/or invention management agents for review—of various sorts, technical, social, equity, contractual, and future business–in present practice a university asserts ownership of inventions, making a comprehensive claim that includes employment, use of materials, participation in research, or even simply association with the university. The expansion has other properties as well.  While there are instances in mid-century policies of claims that go beyond inventions that are or may be patentable, present practice makes such “beyond reasonable” claims routine, despite the problem presented by an invention that is not patentable. Just what is the ownership theory for an invention that is not patentable? Beyond grasping at trade secret, or copyrights in the expression describing an invention–neither of which is adequate–owning an invention that isn’t patentable is not really owning anything. Rather, it is a form of restriction on the actions of inventors (and others) with regard to teaching the invention, disclosing it, using it, and assisting others in the practice of the invention. Even though a university generally does not direct faculty in what to study or how to teach, apparently when it comes to invention policy, university administrators are now comfortable telling faculty what they cannot do with their teaching and publications. This change is a substantial reversal from the mid-century university policy, which, if it restricted anything, it was the seeking of patents over open publication and access, especially in medicine.

  • Possible university equity in inventions has been replaced by an outright, arbitrary ownership claim. Where before it might be equitable, based on a review of circumstances, for a university to be repaid for support, or enjoy a license, or share in the financial proceeds from management of a patent, now administrators simply assert an entitlement to own and control inventions. No review is wanted.
  • Committee review of inventions has been shifted to committee advice on policies, with emphasis on how to soften the blow of compulsory institutional ownership rather than to question the basis for the claim of compulsory institutional ownership.
  • Review of inventions has been shifted from before assignment to after assignment. And review is now mostly administrative, sometimes with participation by local business executives and candidate investment funds.
  • The scope of claim has also expanded. The MIT-style “inventions and other developments” has come to include what amounts to “inventions whether or not patentable,” which is a much to say “inventions and non-inventions” or “inventions are whatever administrators use the word ‘invention’ to mean.” Thus, universities claim ownership of data, software, know how, information, works of authorship, and research materials.
  • The idea of “scope of employment” has been expanded from “official duties” to “anything that one is permitted to do” or “anything related to the field in which one is hired.”
  • Use of resources has broadened from “resources provided in support beyond those of the normal academic environment” to “anything more than de minimis use.”
  • Participation in research has expanded from those funded under a grant to anyone who might collaborate or volunteer assistance on a project.

Part of the motivation for the expansion comes from research contracting practices. If a sponsor makes ownership claims in a research agreement, one then has a compliance issue that trumps, according to any number of policies, the university’s own patent policy. So the way to get more without changing the patent policy is to demand more of one’s own personnel by means of a sponsored research agreement. The argument goes, the default patent policy is trumped regardless of whether the sponsor requires the ownership of inventions or the university administration does.  In this way, university administrators effectively change the patent policy without touching the text of the patent policy. Once the policy is thus changed in practice, then the text of the policy can be later revised to “reflect the reality of the situation.” This was one of the exploits used to make the Bayh-Dole claim of vesting move through a university policy architecture without requiring changes to the local patent policy. Thus, everything has to be claimed because anything could be implicated as an “invention” under some research agreement.

A second motivation for expansion comes from conflict of interest rules, especially at public universities. This is often where the “de minimis” standard comes in. Conflict of interest rules generally restrict the use of state resources for “official duties.” Often these rules were drafted for state agency leaders, expanded to include all state agency employees, and thus imposed on faculty as if they, too, were mere state agency workers, with no regard for the importance of their independence from state control in teaching and research, and the academic freedom that serves as a basis for that independence. Thus, if one is off inventing using one’s lab equipment, and it’s not part of official duties, then according to the conflict of interest laws, it is a misuse of state resources.  There is a deep disconnect in this line of thinking.  It even extends to students who happen to also be employees. Let’s say a student is a research assistant in project X, which studies, uh, rat brains.  The student is also doing dissertation research in computational imaging, which could apply to rat brains but also to a lot of other things. If the student wasn’t working as a research assistant, then she would have the free run of the lab for her dissertation work. But since she is employed, her official duties are in the rat brain lab, and her work on her dissertation, which is “more than de minimis use” is therefore either a misuse of state resources, or claimable by the state under an expanded description of official duties. In this line of reasoning, the university’s expanded claim of ownership generously shields the student from the charge of illegal behavior. Strange, isn’t it?

The problem of course is with the idea of how resources are provided for use within a university, and more to the point, with a mindset determined to claim that what one does creatively has been purchased by the state, either by payment of salary for limited duties (which must be broadened without any agreement with the individual over doing so) or by granting access to resources (which previously were provided as part of the inducement to take the appointment). The free parking space is now an inappropriate perk that should be paid for, or part of one’s official duties. There is no commons, no space in which private activity may take place without a state claim. The logic, if there is any, is that of demanding control of what another does  creatively. In a company setting, such claims would be restricted to the company’s line of business. Here, the university claims all lines of business–its business is selling whatever gets created by any of its personnel. Further, the university claims that its creative folk, otherwise recruited to be independent and exploratory, cannot work beyond what they are assigned to work on. If they branch out, then this branching must be constructed as a request for the expansion of their official duties, at no more pay, but with a claim, now, on anything they might discover, realize, learn, build, compose, create, develop.

What has been lost is the idea of the university as public commons. The word “campus” has “park” roots–back to the use at proto-*Princeton in the 18th century to describe, apparently, the clearing in which the university buildings were placed, to be distinguished from the school’s yards (areas around buildings) and the school’s fields (ultimate frisbee!).  The “campus” is a commons, like a public park or a public library. Or at least it was. Now the idea is that it is a proprietary space owned by the state-as-company, and that use is conditioned on the assignment of creative work to the state.  If one is into government control of ideas, then this would appear to be a great way to do it. It really does not change things if the premise is to find a way to make industry pay in order to use ideas in any practical way. So much for liberty, for faculty independence, for academic freedom, for innovation other than government-approved innovation.

If one were building a campus patent policy directed at innovation, the starting point would be to restore the commons and reject the idea that the state, via the university, is a corporation with a right to control the creative work of faculty for the purpose of money-seeking. Even if such a right is possible–that is, one has better paid lawyers than the other side has–the strength of a public university is that the state does not exploit that right. Otherwise, we are back to Thrasymachus’s argument in The Republic, that justice is whatever those in power say it is. The “de minimis” argument would suggest that those in power are ones who believe that the state should control creative work for its own profit-seeking adventures. That, for them, is “justice” and they have been able to hammer the point home in policies–not so much the patent policy directly, but research policy, conflict of interest policy, and employment policy. Thus, to rebuild a patent policy now, one has to expand it to include inventions that are non-inventions, research contracts that claim inventions that are non-inventions, and conflict of interest rules that make it appear that it is a misuse of public resources to have a private idea of any value that is not given over to the university for a whack at profiteering from it.

In a university policy directed at innovation, getting things as quickly as possible into the hands of the established order—a company, a wealthy investor–is not one’s number one priority. In a policy directed at innovation, all one needs is a basic framework under which folks acknowledge what is equitable, or fair, under the circumstances, relative to their commitments to the public, to sponsors, and to their faculty colleagues. If we have lost the idea of “fairness” in a commons, then we have lost much.

University claims of ownership in inventions (and non-inventions) have also been expanded in terms of the timing of the claim of ownership. In past practice, the claim of ownership might have been made once a patent application was prepared and claims could be considered. After all, it is the claims that define the patent, and thus whatever contribution that the university has made should be reflected in the claims, not merely the specification. However, in recent years, university administrators have moved first to abandon review–ownership comes first, then review for a patent desired by the university administrators. Second, university administrators have introduced present assignments, so that ownership of any invention within scope is assigned automatically, when it is made–or so the theory goes–without the bother of having to have notice of the invention and obtain a paper assignment specific to that invention. The university claims then to own everything—even stuff that it does not know that it owns. In this work up, the assignment that actually conveys legal title–a written instrument identifying the invention and signed by the inventor–is merely a ceremonial confirmation of what the university has claimed it owns “equitably” simply by having made the assertion by means of a required present assignment.

A common past approach was review for equity interest, often via arrangements with external agents. It was this approach, for the most part, on which the rhetorical quantitative arguments for Bayh-Dole were based, that universities were doing a better job than the government in commercializing inventions made in research. These arguments left out that many government-owned inventions were weapons-related and that the contractors had passed on wanting to commercialize theme as products. Also left out was that the university agents waited for inventors to decide an invention was worth patenting and managing privately, or if the cases where the university demanded ownership, the agents cherry-picked the inventions that had the greatest likelihood of commercial success–that is, the inventions that probably did not need an external agent and especially did not need a university administration deciding what that agent would be. That is, the common past approach was Diverse, largely Voluntary, and highly Selective (DVS). These were the components of its apparent success.

A federal law that aimed to make more federally supported inventions available to this diverse, largely voluntary, highly selective approach might therefore be one that does the least possible to disrupt the approach, other than to expand the pool of inventions that it might draw from to do its thing. There would then be two classes of invention made with federal support: those that were drawn into the DVS approach, and those that were not. The inventions that were not drawn in could then be considered by the government for ownership, not simply for “commercialization” but for all the other things that a patent might provide, such as publication in the literature, recognition of government supported productivity, provision of common access on royalty-free or reasonable terms, development of standards, quality control, mitigation of unfair procurement positions, mitigation of unproductive proprietary positions (as in improvements), and competitive development.

Bayh-Dole, however, added a third option, which was originally part of the first option: that is, if inventors didn’t assign to an agent, a federal agency could waive its own claim and allow the inventors to do what they wanted–which could include working with an agent anyway, or assigning the patent, or developing the invention themselves, as a small company. Universities, too, often have such a third option built into their patent policies. Originally, this policy provision went something like this. If the university agent chooses not to file a patent application, then the inventor does not have to assign to the university, or the agent.  In variations, this became: if the university gets assignment and does not file a patent application within a given period, then it will return the invention to the inventors. Here, for instance, is the University of Florida (1950):



Kentucky (1947) is similar:


It is worth pointing out the various dynamics of these situations. First, the need to place a return requirement in policy follows from a requirement in policy that inventors submit inventions for management to the university. If there is no such requirement–the submission of inventions is voluntary–then the arrangements for return may be negotiated in the agreement on management.  Second, the return requirement follows from two kinds of actions. The first is a selectivity review.  It is based on a decision, typically by a faculty-led committee, regarding the technical merits of the invention, the equity of the university’s contribution to the development of the invention, and the public purpose (including, inevitably over time, the financial prospects) of the university’s management of the invention. The second is a failure to act. If the university fails to file a patent application, or in some policies fails to secure one or more licensees in a reasonable period, then the invention and associated patent positions may be re-assigned to the inventor(s).

When a university expands its scope of claim in inventions, so that it claims outright all inventions, it necessarily puts a great deal of pressure on a return clause. If that clause is based, not on selectivity (which the university abandons by claiming inventions wholesale) but on inaction, then it also creates a nasty situation: if by inaction the university has damaged the value of the invention, then offering to return the invention appears to be more in the way of a settlement or remedy for the harm, rather than simply acknowledging (preferably relatively quickly) that the university is not effectively patenting, or marketing, or licensing, the invention.

Note as well the difference in such a situation if the university uses an invention management agent. If the agent fails to patent, or market, or license, then the university acts to hold the agent accountable for both the university’s interest and the inventor’s interest. If the university is acting as the agent, however, there is no institutional accountability.  The university licensing office will present as “successful” and any lack of activity is the result of an invention being “worthless” from the start. The university has a much more difficult time acting as both arbiter and agent. The consequence of this arrangement is the introduction of appeal processes for university ownership claims, but these appeal pathways are generally administrative: if an inventor does not like the decision of a lessor administrator, the inventor can appeal to the administrator’s boss. That’s not much of an appeal path, if both administrator and boss have a vested interest in maintaining the “success” and “rightness” of their invention management program.

If a university does not advocate the use of independent invention management agents, then it gives up its mediating role. It no longer is an advocate to review an agent’s practices and effectiveness on behalf of the faculty, the public, and its own interests. Review discussions become internal, political, and suppressed.

Another layer of complication arises when a university pushes assignment ahead of review. If review takes place first, then if a university (administrator or committee) does not identify an interest in the invention, then the action that is indicated is the release of a claim. A release is a relatively simple thing. No ownership changes hands. The university announces that it will not act on a claim that it otherwise might act upon. In the case of inventions made outside the scope of the university’s proper claims, the release is nothing more than a confirmation that the university had no claim in the first place.

If, however, the review takes place after assignment, then the university must re-assign the invention to the inventors if it is not going to take a responsive action with regard to patenting, marketing, and licensing. Re-assignment means transfer of title. If the basis for the assignment was voluntary, again, the transaction, while involving a substantive formality, is still direct. However, if the claim has been made in policy that the university was entitled to ownership, or held equitable title in the invention simply on the basis of employment or provision of resources, then the re-assignment takes on another color altogether. The invention in this context comes to appear to be an asset of the university and re-assignment appears as the provision of a potentially valuable asset to private hands.

For state universities, this sort of transaction may even appear to involve favoritism or conflict of interest. How can a state asset be given to any private person? Why should an inventor, who owed the invention to the state the moment it was created, have any special claim on the invention over that of others? This line of reasoning ends up second-guessing the decision to re-assign the invention. The upshot is that a re-assignment is a substantially different situation than that of a release of a potential claim.

Clearly, in either case, if the invention is not patentable, then the discussion of ownership ought to be over, as the whole point of owning an invention is for the right to file a patent application and have a patent issue to the owner of the invention.  As university invention policies expand to take in inventions that are “not patentable” the release of rights grows murky. What is the ownership theory for an invention that cannot be owned by patent? Is it trade secret?  Is it some common law theory of ownership of ideas? Is it a claim to copyrights in the expression of the idea? Even here, if there is no assignment, then the release of a claim is straightforward. But if the university takes assignment of “all rights” in a “non-patentable” invention, and than does not act on the invention (it is, after all, not patentable), then what is this the standard under which the university must return the invention? And what does it mean to return “rights” if there none, other than murky ones posited by university policy?

These are the sorts of things that take a fundamental confustation regarding patentable inventions, creative works generally, when to review, and when to take assignment and turn a relatively simple policy situation into mound upon mound of poorly conceived paragraphs that grapple with an intractable situation, often retreating to the dark closet of demanding an appeal process that then is drafted to force the burden of the argument onto the “problem” inventor who feels abused. The form of the policy, at such a point, is that the university’s policy is that administrators should decide things with impunity, and inventors can reveal their ill-will if they want, but that only prolongs the inevitable and runs up costs that hurt everyone else.  It would be easier for a policy to simply read something like:

The university owns all work produced by its faculty, staff, students, and others using its resources or acting under its auspices without the need for any specific agreement has been entered into providing for such ownership.  If such an agreement is required, then this policy statement is that agreement.

The university may seek to make money from its ownership positions in the works it owns but has no obligation to do so. Producers of work owned by the university shall not seek to influence the disposition of such work, but are required to assist the university as it may require in its management of such work.

If the university makes money from the management of such work, it will share a portion, after expenses, with those that produced the work so exploited.

To be clear, I am not advocating universities adopt such policies. My point is, they already have.  They just cannot bring themselves to draft a policy statement that lays it out.

In the mid-century case, if the university agent chooses not to file, or files and does not license, the inventors are restored to their right to act on their own behalf, less whatever damage the university may have inflicted, plus whatever the university has done to advance patenting, if patenting was at all desired by the inventors. The situation presently is that if a university has taken ownership and incurred expenses, then the inventors should have to reimburse the university if they are to be re-assigned the invention.

In all of this discussion of ownership, it must be emphasized that just because some research finding is “inventive” does not mean that a patent is worth seeking. Very often seeking a patent on such an invention is a bad decision—expensive, complicating, distracting, disempowering, creating vain or unrealistic expectations, introducing grave uncertainties, unsettling relationships among a research team (some are inventors in line for royalties and others are not), demanding payment to recover expenses, and restricting available options to ones that focus on exploitation and valuation of the patent right, not the invention.

In a number of mid-century university patent policies–the ones that were by comparison short and reasonable–first an inventor decides something is worth patenting, and the act of preparing a patent application triggers the question of university equity in the invention, which may include a claim of ownership. If an inventor wanted to remove uncertainties, then the inventor might approach the university early on, to reach an agreement on equity in the invention.

In the post-Bayh-Dole style university patent policies–typically long, rambling, and incoherent–the idea is that the university must own anything that is inventive–even if it is not patentable–as the opening position. There is no support for such a position in Bayh-Dole, but Bayh-Dole is routinely cited, either for the authorization to own with a mandate to “commercialize” or as a “just in case” compliance claim, as if Bayh-Dole made uncertain expansive claims when it does just the opposite (despite the NIH’s apparent effort to confuse the matter). Such a position smacks of administrative folks and mercenary legal advisors who don’t know what they want and fear they will lose some future windfall, so claim everything and leave it to others to sort out later what they have trawled up in their policy. It’s a kind of factory ship that strips the ocean of faculty research of its life, in the hunt for anything saleable. There appears to be not even a Greenpeace zodiac about to challenge the methodology.

It is clear that just as the march-in regulations authorized by Bayh-Dole were drafted to make it nigh unto impossible to be ever used, so also the university patent administrators have set up policies that ensure that the third option, whether under Bayh-Dole or under their own policies, never operates to allow faculty inventors to manage their own work. The reasoning is right out of Monty Python: if an invention is worthless, then no one will ask for rights to it. If someone asks for rights to it, then it is not worthless, and so i) whoever asking should have to pay; and ii) if someone is asking, then there may be others who would pay more, if only to keep it from the first party doing the asking. Thus, university administrators do not routinely hand back inventions their policies have claimed.  The patent policies become arbitrary, non-selective, saddled with dealing with everything.

The policy effort then becomes one of deflecting criticism for failures to act by claiming that inventions that are not licensed have no value.  If there were a properly managed “release” procedure–that is, before assignment, to avoid needless overhead–then there could be a third way, with inventors working with their scholarship as they choose.  They may release to the public, work with an agent, start a company, develop a standard, separate make/use rights from sell rights–any number of options, and generally more options and more readily available options, than a university licensing office on its own can contemplate or be capable to undertake.

Returning rights is a difficult task for a university administration in the context of the policies that have been implemented.  Doing so is not nearly as simple as one might expect. To “clear” an invention for hand-back, one has to clear all outstanding possible encumbrances–including terms in sponsored research agreements (often inserted by the university administration) that the university must own the results of the sponsored work or offer a license to the sponsor.  Then there may be letter agreements or options or slow moving negotiations that might have offered a right to negotiate a license and these have to be closed out.  These could be embedded in material transfer agreements, in grant applications, in affiliate programs, in publications, on web sites, or coming to a theater near you.

Of course, all these things are also in play with *any* license granted, but university administrators often do not worry the context nearly so much for licensing as they do for return of an invention. Even if the claims of others are cleared, there is the issue of whether federal funding might have touched the invention at any point, in which case universities argue that they must offer the rights to the government, rather than back to the inventor.  And then there are the claims of the university administrators themselves. They may want a shop right, whether it would otherwise be equitable, but often a “shop right” is seen as archaic and what they really want is a non-exclusive license for at least “non-commercial or educational” purposes without giving much scope to what that means. What they really, really want is to be reimbursed for their out-of-pocket expenses in filing a patent application, as if in a compulsory ownership scheme the expenditures are on behalf of the inventors, so they are really bills that the university agrees not to send out, in exchange for keeping the patent.

If any agent takes an invention and cannot do anything with it–that is, cannot fulfill the purpose for which it is an agent–then the value of the invention to that agent is exactly $0. Making an inventor pay to get back rights is worse than eminent domain. The logic is: “I get to take your house, and if I ever resell it I will pay you a share of the profits, but I don’t have to sell it ever. And now that I am not selling it, and you want it back, you have to pay me what it cost me to take the house from you in the first place, as if that is an improvement on the house that’s an expense you owe me!” It does not matter what an inventor might do with a patent, or patent application–make money, release it, assign it, waste time. The university should let the invention go, and should set up its operations so that it is easy, not next to impossible, to let it go.

The patent policies of American universities have been modified from their mid-century roots to become a mess of problems.  The expansion of scope, the shift in claim from possible equity to arbitrary ownership, the change in the nature of review, appeal, and return of inventions have introduced intractable problems for current patent policies, which are often incoherent beyond the obvious theme that administrators demand ownership while minimizing accountability.  In these policies, which are shanties built amidst the ruins of the policies they destroyed, there is still the traces of the “third option” present in Bayh-Dole.  If faculty inventors are allowed to retain their ownership, with an obligation to account for the disposition of that ownership, then there is still an opportunity to rebuild the DSV approach–diversified, selective, and voluntary–under which universities and invention management agents might be asked to assist in developing an invention for broader use.

To accomplish such a thing, however, the present policy structure has to go.  Wordsmithing is not enough.  Adding more text or better definitions will not do.  The architecture, the fundamental logic, is broken.  The problem is, for a bureaucracy, the outcome–ownership with impunity–is attractive.  The premise–public benefit from expert management–is attractive.  The successful licenses–a big hit a decade–are attractive.  We have been primed for a model that has failed.  The attraction should be, not to hold onto these past rhetorical pleasantries, but rather to practice some creative destruction and start over with patent policies that encourage development (but not of everything), that offer institutional grade resources (but do not impose them) and that allow for diverse choices (rather than fixating on investment-back commercial product speculation).   So, yes, burn down the barn, and the rats will leave.  Maybe get the cows and hay out first.  Then build again, for a new era, having seen what is possible and what is not desired.

This entry was posted in Bayh-Dole, History, Policy, Present Assignment, Technology Transfer. Bookmark the permalink.

1 Response to University Patent Policies, Past and Present, Part II

  1. Pingback: University Patent Policies, Past and Present, Part I | Research Enterprise

Comments are closed.