Here is a brief summary of the flow of control for choice of invention management agent in Bayh-Dole.
First Choice:
Inventors can choose university
Inventors can choose another qualified agent w/university agreement
Inventors can choose any agent w/university and w/federal agency agreement
Universities w/requirement to assign can choose themselves
Universities w/assignment can choose another qualified agent
Universities w/assignment can choose any agent w/federal agency agreement
If no first choice agent is chosen, then:
The agency can choose itself
The agency can choose public domain [perhaps!–more on this soon]
The agency can choose the inventors w/consultation of university
One can see from this list that there is no requirement in Bayh-Dole or the standard patent rights clause that the university must have an agreement in place requiring assignment. It is not a condition of federal funding. The (f)(2) agreement leaves the disposition of title entirely in the hands of the inventors, who may voluntarily choose a qualified agent, or may give up that right at employment, before the research begins, or at any time thereafter. If the right is given up voluntarily, it can be by donation or under contract for consideration. Old royalty sharing schedules reflected the consideration for the assignment of title, such as to research foundations that had no employment relationship with the inventors and provided, generally, no facilities on which to base a claim of ownership. Thus, the outside agent relationship was clear–royalties were consideration for transfer of title.
In the new administrative regime, employment, use of facilities and resources, and royalties are conflated. Assignment is expressly required as a condition of employment (and payment of salary), which makes it appear that faculty are agreeing that the employer is, in essence, purchasing that portion of their scholarship on which the university thinks it can make money–inventions, software, data, know-how–regardless of the form of intellectual property ownership.
The only problem then left for later is just what scholarship has been so purchased, and what happens to the scholarship that has not been fully fingered to find the money in the lining–or even the hope of money, which appears to be enough to support a finding of “commercial potential” as opposed to a demonstration of “commercial value”. “Potential”, here, is something of a weasel word. But more importantly, it is a word of judgment, and if the judgment is solely with university administrators, then the policy actually reads: “anything we decide we want, since if we want it, we say it has ‘potential'”. In a way, it is an agreement to agree. There is no objective standard. All your scholarship is belong to us. You agree to agree to our future determination of “potential”.
My argument is that such a claim is unenforceable. Unless there is an objective standard (receipt of an arm’s length offer of value greater than $1,000 within 30 days of assignment, say), then it’s just an agreement to acquiesce in whatever the other party says at the time. There is no meeting of the minds on the matter. Any review step in a compulsory policy has this defect, if the determination is one of judgment rather than standard or definition.
In a company, the standard is what the company is in business to do, and the facilities are provided for that purpose. Universities, if they reasoned that way, could claim next to nothing, because their charters do not include making money from licensing, they do not direct faculty in much of what faculty and academic employees do, and the facilities and resources are provided for public, not institutional benefit, and public includes *the personal benefit of others, not the fictional “person” of the institution* and “others” includes faculty and students.
Of course, if a university policy demand for ownership is just playing with words, making up things that are impossible, but posing them as official requirements or as veiled threats to one’s employment or character, then that policy unenforceable and is actually equivalent to a voluntary assignment policy, except for the pretentiousness and ignorance by which it is drafted and promulgated.
As university people draft policy in an attempt to tighten demands for ownership, they try to define the scope to be “inventions”–extracting the ownership claim from “is or may be patentable” and transmogrifying the scope into all “patentable and non-patentable” stuff. An invention that is not patentable, however, is not owned under patent law. It is nothing under patent law once “may be” becomes “not”. What then is the ownership theory, since policy drafters are determined to claim ownership anyway? It would appear to be trade secret–that is, the university asserts the right to suppress scholarship when doing so may allow it to make money.
The University of Washington policy almost comes out and claims that:
Where an employee . . . desires to utilize the expertise and/or technology he or she has developed and to assist a business venture . . . in the commercialization of an idea, the employee and his or her supervisor should first seek the early assistance and counsel of UW TechTransfer to aid the employee in distinguishing those things which may be freely shared (e.g., “general expertise”) from those the sharing of which may be restricted (e.g., “technology” or other “intellectual property”).
This is from Executive Order 57, Section 6.C.4. The title of the section pertains to “deeper involvement than consulting” but the statement place in the subsection has no such limitation. Of course, this statement is also found near Section 6.A., which contains one of the most incomprehensibly bad passages of university policy in recent memory. However, this piece in C.4 looks pretty clear–expertise, technology, and ideas may be restricted by the technology transfer office.
How can the technology transfer office enforce such a thing? By asserting a trade secret? Construing a combination of policy statements to amount to a covenant not to disclose, use, or publish without prior administrative approval? It hardly would seem to hold up. Any of these expansive claims beyond an invention that “is or may be patentable” is simply untenable in university patent policy via ownership.
There are other tricks, however. If the university is public, it could assert the power to suppress disclosure. One way is to claim that an exception to the public disclosure law provides a mandate to keep confidential certain information, rather than merely a means by which a request for that information does not have to be honored through the mechanism of the public disclosure law.
The conflation of invention and anything-we-say-is-an-invention goes further. As we chase it down, we see the remnants of voluntary assignment policy and practice, and how these fragments, such as royalty sharing schedules, have been re-purposed. If an invention is made, and full rights are sought, then an application must be filed before there is public disclosure or most foreign patent rights are lost. Thus, there is a reason not to publish in advance of filing. In a voluntary scheme, the choice not to publish ahead of filing is the inventor’s. The idea is, the inventor, as a scholar, has the choice of venues for publication, and the Patent Office is indeed a publisher of technical literature. A patent is a public disclosure made in exchange for a reservation of rights for a limited time to promote the progress of the useful arts. If the inventor-author desires those rights for herself or on behalf of others, then the author will defer other publication until an application is prepared for the patent office. It is done by choice. It is part of the academic decision-making that scholars do. It is not imposed by the university.
However, if the university demands ownership of inventions, then it would appear that the university also is demanding a trade secret in scholarship the moment it “may be patentable”, and is therefore also demanding the right to suppress publication and to choose the publisher–namely, the Patent Office. It may be some premises for doing so are acceptable–say, to preserve a commitment to make inventions broadly available at no charge against predatory tactics by industry players (or other universities, now, come to think of it!). Other premises, such as the desire to make money by taking a proprietary position against collaborating companies (predatory tactic as virtue and public mission, along the lines of the Stanford double-cross), don’t seem to hold up so well. Another way we don’t expect universities to operate “more like a company” where “company” means “crass player”.
In a voluntary approach, when an invention is made, and the inventor desires a patent, the decision not to disclose or publish is the inventor’s, not the administration’s. If Bayh-Dole had required ownership of inventions with the university administration, it would have also had necessarily to authorize universities to assert trade secret over much of scholarship on the premise that it “may be patentable”–that is, until the administration had handled any particular asset, it could not say that it was “not patentable” and therefore, as far as it is concerned, everything “may be patentable”, if the administration, rather than the inventor, has the sole authority to make the determination.
One can see, then, how a university IP policy drafter, having decided to demand ownership, and following a path from information concerning an invention that “may be” patentable to a demand that this information not be disclosed (on the worry that doing so might then ruin an asset claimed by the university), has to expand the definition of “invention” to include everything that has not been shown conclusively not to be patentable. It is then a simple move, once the scope has been so expanded, to claim expertise, know how, ideas, software, data and anything else that is declared to have attributes like “may be patentable”–that is, not disclosed to the university but which might make money via an ownership claim (even if there is, in essence, no ownership possible, as in the case of “know how” in an open university setting). It is an entirely different use of “may be patentable” than what one finds in the patent law definition of invention.
Once this piece of bad reasoning is in place, then any number of rationalizations can be constructed for showing that it is proper, that it reflects a true commitment to the university (and therefore to the public) and any resistance is a sign of selfishness, greed, defiance of authority, and violation of the public trust. The actual violation of the public trust comes when the conflict of commitment/conflict of interest people fail to fully investigate the claims for ownership before constructing a rationale to underscore the university’s “rights”.
The old royalty sharing schedules were in place so that when an agent such as Research Corporation, or a local affiliated research foundation, received payment for a license and sent a check to the university, everyone was clear on how the money would be divided. This strategy still holds in a voluntary scheme, and provides the standard for what an inventor might receive in exchange for, in essence, granting the invention to the university. It’s of the form of a non-negotiable deal, if you request that the university act as your agent. Of course, in an open market for agents, an inventor could negotiate the deal with the agent–how much does the inventor get, how much the agent. This is true even where the agent also has a deal with the university. It gets the university out of having to cut any checks to the inventors, and still gets a benefit by designating some agents as “preferred”. In the Bayh-Dole scheme of things, this amounts to agents paying a share of income in exchange for being pre-approved by the university for assignment of Bayh-Dole obligations and title to subject inventions. If there is money to be made by the university, this is the approach that minimizes expenses and liability, maximizes net income, and preserves the university’s primary role in dealing with research integrity and competing interests among academics.
However, once invention ownership becomes compulsory, then the royalty schedule looks less like consideration and more like something else. If title passes as a condition of employment, as the University of California makes clear in its Patent Acknowledgment, then the royalty schedule is not consideration for assignment. Employment is. Salary is. Royalty sharing isn’t regular pay. It isn’t reported that way, generally. It goes on a 1099 as if it’s “royalty income” for the inventor, when it’s not royalty income at all. It may be royalty income for the university, but a “royalty” is a payment made in exchange for the grant of rights in a patent. Universities making a demand for ownership exclude such a thing. So what is this payment?
Some administrators aim to construct the payment as administrative largesse–a kind of incentive or reward for being a good, dutiful employee, or to show the university’s positive view of technology transfer, patent licensing, and patent litigation. Others recite Bayh-Dole, standard patent rights clause, section (k)(2), that requires nonprofits to share royalties with inventors. For subject inventions, this might be sufficient–the payment is not part of a deal with the inventor at all, but rather a requirement of federal funding. That does not help, however, with any other inventions. I don’t know of a single industry sponsored research agreement where the sponsor requires the university to share royalties with its inventors. It would be an interesting clause to see. Perhaps the HHMC investigator agreements with universities come the closest to doing this.
The royalty schedule, an artifact of voluntary days, comes to stand for a general obligation to share royalties with inventors, but no longer reflects a choice or negotiation. It is not a condition of assignment, so it is not compensation for assignment. It is not “just compensation” at the time of assignment, because there is no compensation at that point, and in some policies not even any obligation by the university to seek licensees or receive payments for licenses. If a university argues that it “technically” hires all faculty “to invent” and then happily and generously waives that requirement for all stuff in which it does not see profit-making, then it is just playing games with words. The university does not direct the work, the university has no “business direction” that involves what faculty choose to study. It is all make-up administrative property grab. It is what the laws against employer invention vampiring are all about.
If a university claims ownership of inventions as a condition of employment, and expands that definition beyond patentable inventions as determined by inventors, then it is necessarily claiming trade secret control of scholarship. But if it does not in practice–or cannot because of other policies–actual make such a claim, or enforce such a claim–then its claim to ownership fails as well. It cannot play with words. It cannot claim something as owned when its own policies demand conditions that cause the ownership claim to fail, and to be contradictory, and, in a way, offensive to academic freedom and the premise of a university’s open research.
Unlike a company, a university has substantial constraints on how it can claim ownership of assets made by its faculty and students. The claim of employment has been demolished, and certainly it cannot extend beyond inventions that “may be patentable” in the judgment of the inventors, as part of their experience of “conceiving” the invention. If the inventor understands his work to be an “invention” then it is also, at that point, reasonably within the scope of “may be patentable”. If the inventor does not recognize her work to be inventive, then it is “not patentable” even if on analysis later someone could argue that it should have been.
A university has to deal with academic freedom, tenure, self-directed scholarship, extramural research, provision of facilities for public purposes (not merely its own self-interested purposes, invested with publicness), public disclosure law (in the case of state universities), and broad conventions of open access, encouragement to publish, and avoidance of claims that would bring export control and with it discrimination against foreign nationals participating in research activities on campus. A technology transfer office writing policy to claim ownership in all things runs up against a great many factors that prevent it from turning the university into a closed-shop commercial concern. It simply cannot adopt such “corporate methods” to “protect its intellectual property”. First and foremost, much of what it seeks to “protect” is not its intellectual property at all–not by right, not by operation of law, not by equity, not by public policy.
There simply is no rationale in universities for an expansive claim on inventions to include expertise, know how, and the like. At best, it is a matter of incompetence in drafting. At worst, it is a cancer on the health of scholarship in the university, and if not removed, will do its part to damage the university and its effectiveness as an agent of public training and discovery.
The overlay of compulsory ownership claims onto an existing apparatus for voluntary, cooperative management of patentable inventions has created a worst case of conflating old definitions and demands to new, inappropriate purposes. It may not be obvious on the surface, where an employer’s claim for ownership of an employee’s work product seems altogether equitable, and where a public university’s desire to protect its facilities from the intrusion of commercial interests seems high-spirited, and its desire to generate new sources of revenue by acquiring assets from its truth-seeking, company-naive, context-lacking, resource-poor faculty and putting those expertly into commercial use–all these things seem beautifully well reasoned. Yet in practice, underneath the superficial rhetoric, aspirational to all good things that it is, one finds a fundamental corruption of practice. The claims do not hold up. The words cannot mean what is written, and if they are subject to whim interpretation later, then there is nothing enforceable.
Much of the compulsory system we have today is built on the claim that federal law requires it. That claim is now shown conclusively to be untrue. Yet we have this compulsory apparatus. The great temptation is to preserve it, use it. Yet it has no foundation. The policy move that would allow it was that it was required by a research sponsor–the government. But that’s simply not true. Once you fall back on employment or use of facilities, then where is the consideration? How is the policy enforceable relative to the “non-patentable” stuff?
The compulsory system is like a prison complex that has been built on the remains of a free city. There is no reason to keep the prison. It has been discredited as the wrong side of history. Rather than attempting to save it, because they allowed it to be built, universities should be actively moving to rid themselves of the needless burden, expense, liability, and irritation of having allowed it to be constructed. There is an IP archeology, and it may unearth important findings that would benefit us were they to come to light.
In the place of the prison, this discredited compulsory system that fails all the way down and scatters its failure throughout university activity in mostly disagreeable ways, universities should consider the agent system in Bayh-Dole, which is itself an artifact of the voluntary system that made American university participation in technology transfer the envy of the world. That apparatus is still in place. In many places it is the foundation on which the prison was built. Many of the artifacts are still functional for their original purpose. Very little has to be done to clear out compulsory claims and replace them with a mandate that university inventors are required to choose an agent (which may be the public domain, as Stanford still allows), and universities may provide incentives or place limits on who those agents are, but they need to stay out of the ownership game unless they are chosen–by the inventors, by a sponsor, by industry request. All a university has to do is choose to uncover the free city that it used to be. Then useless things do not clutter its work. Then it has the best years of its life.