Dubilier, university IP policy and, er, inner life

Dubilier set in motion a cascade of things that leads us to, well, to where we are. In Dubilier, the Supreme Court established that inventors own their inventions unless they agree otherwise, even if they are employees, and even if they use their employer’s resources, and even if they work on their invention when they are supposed to be working for their employer. The employer’s compensation is the shop right to practice the inventions made by employees, if the invention is at all useful to the employer. If an employee is hired to invent, or to conduct experiments that might lead to an invention, then a promise to assign title to any inventions is implied in the employment relationship, called “equitable title” by the courts. Thus, the defaults confirmed by Dubilier, and the two concepts–shop right and equitable title–that have been developed to deal with Dubilier.

In the absence of an inventor agreeing otherwise, then, inventors own their inventions and the shop right means inventor employees leave their employers alone when it comes to patents.

After Dubilier, Archie Palmer worked for thirty years to exhort universities to adopt formal patent policies. Palmer published three compendia of university patent policies, providing commentary and context for these policies, summarizing arguments for and against university ownership of patents and involvement in patent licensing and litigation. But what role does a university patent policy play, given the defaults established by patent law and the courts? For the most part, early university patent policies tracked shop right and equitable title. If an employee was hired to invent or to conduct experiments–some universities operated practical laboratories focused on practical problems (not so much these days)–then the university claimed ownership of resulting inventions. In such cases, the policy repeats the equitable title default, but with more words and often introducing ambiguity in the name of greater clarity (just like administrators to think to do that).

In the case of faculty, universities also had to deal with academic freedom–including freedom of research and publication. For faculty, the idea of “employment” was not that of a master-servant relationship. Faculty were appointed. They were paid a stipend. Their duties of teaching and research (if they wanted) and professional development were theirs to undertake and manage. While faculty were expected to teach and study, they did not do so for the benefit of the university. They were not servants of the master university. The university did not review and approve their work for its own benefit, but only to ascertain merit for tenure and advancement. Thus, the Dubilier defaults for faculty ended up limited to the very special cases in which faculty agreed to give up their academic freedom/research freedom/freedom to publish in order to accept special commissionings or tasks proposed by the university and carrying their own compensation. Thus, a faculty member might agree to work over the summer (when not paid a stipend) on a project initiated by and paid for by the university. Then the faculty member would be an employee of sorts (more like a contractor within an appointment), and hired to invent, and it would be equitable that title to any inventions be assigned to the university.

We can identify a number of generally different contexts in which faculty might invent.

(1) There’s the invention within the university–working in what was called “departmental” research, using university labs and equipment. For this work, faculty are not employees. The university does not commission the work or expect the work to be useful for the benefit of the university. Rather, the university openly provides resources for the use of faculty. The idea is to support faculty work as a patron, if you wish, not as an employer expecting some deliverables in exchange for its outlays of funds and materials. In the context of departmental research, then, a university patent policy might step in and establish defaults where the inventor is not really an employee at all. Just futzing around with the toys made available by the university in its thoughtfulness without conditions. The default would be, inventors own their inventions and the university doesn’t have a shop right, doesn’t have equitable title. A new patent policy might then introduce the idea of compensation for the use of university resources if the faculty inventor ever received income from the exploitation of a patent. Or, later–might introduce the idea of a claim of ownership as compensation for the inventor’s use of the university-provided resources.

In terms of a university-faculty relationship, a university policy claim based on use of resources other than when specially negotiated (and should there even be such negotiations?) marks a dramatic change. The ownership claim starts to look like a penalty for using university resources, creating an implied condition that if the faculty member uses any such resource, the consequence is that the university claims the inventions. Early policies set up committees to review inventions made in these circumstances, and determine what would be equitable. When the committee work got to be a bother, administrators had the bright idea to replace a determination of the equities on a case-by-case basis with a general statement of claims, with the opportunity for appeal or waiver of the university claim. This bright idea led to the further bright idea to have the university claim everything by defining an “invention” as pretty much everything (so it becomes an invention because policy calls it that, rather than focusing on inventions that are patented–even slipping from patented to “may be patentable” is a huge slip in what gets claimed–stuff that isn’t patented has to be claimed just in case it could be patented, and stuff that could be patented has to be claimed because the university would then claim the invention even if it then did not go on and apply for a patent or eventually receive a patent, then the university must claim everything until it knows for sure that something will receive a patent.

By the time a university has claimed everything, and sorted out what is patentable, and what gets patented, it is such a bother to deal with releasing what isn’t patented that universities just don’t generally do it. University patent policies that claim everything often add to that everything anything with “commercial” value, whether patented or not. Thus, if an inventor (the “Creator” as one tone-deaf university policy has it–could be the maker of anything, and well, isn’t the Creator the maker of everything?) requests that the university release its ownership claim on some bit of newness, the obvious bureaucratic calculus runs–if the inventor wants this back, it must have value. If it has value, then the university claims the right to exploit that value, and will share some portion of that value with the Creator. Therefore, it is not appropriate (or acceptable) to release the university’s claim. If the invention has no value, then it is a waste of resources to bother to release the invention. If the invention has value, then the Creator must disclose that value so that the university can exploit it. If the Creator refuses to disclose that value, then that’s an ethics strike against the inventor, who will face disciplinary action for not assisting the university to realize the value of invention (for the benefit of the public, ahem, of course).

Yes, very much a Catch-22. You can’t have it back if it’s worth anything. And you can’t have it back if it isn’t.

This is a cascade of policy that moves from restating the already established defaults to creating committees to decide equities to replacing committees with university-favorable (generally) defaults regardless of equities to expanding university claims to cover all potential cases (even if the claims are grossly overreaching), making it impossible for inventors to get anything back (because if they try that is evidence that there’s value, and if they don’t try, that is evidence there is no value). In this way Moloch eats his children and destroys opportunity. But it looks like a great idea to institutional bureaucrats. Even progress.

(2) There’s invention within extramural (“outside the walls”) research–sponsored by a company, by “industry” via an association or consortium, by charitable foundations, or eventually also the federal government. Early on, faculty negotiated the IP rights in these sponsored research agreements, but as the federal government got involved in funding university work, universities created new divisions for research administration, along with policies required by the federal government for how to submit grant proposals, how to manage them, how to report the work, and how to close them out. Eventually (Wisconsin was the last big holdout) universities took over the negotiation of sponsored research contracts and especially the IP clauses.

A general early policy position was that the university would own inventions made in sponsored research unless the research contract specified otherwise (such as, the sponsor would have a right to own–as was eventually the case with federal contracts applied to nonprofits, at least since the PHS policy in the 1950s and generally after the 1962 Kennedy patent policy). The Kennedy policy created the possibility that a federal agency could respond to a petition by a nonprofit to retain “greater” rights than an assurance of a royalty-free license to practice an invention, and this “flexibility” was turned on its head later and depicted as “uncertainty of title”–meaning that a university seeking to own a patentable invention made in federally sponsored work would not know how a federal agency might respond to its request, and this uncertainty along with any delay in the agency’s decision might create difficulties for the university’s efforts to grant a license.

Indeed, just such a delay on a request by the Purdue Research Foundation to the Department of Energy served as the premise to get Senator Bayh involved in fixing the “uncertainty of title” problem. But really what got “fixed” was federal agencies losing control of patent rights developed in multi-contracted projects, with a number of universities and nonprofits all working on bits of a greater work. There’s a provision in Bayh-Dole for restoring the original defaults (“exceptional circumstances”–where “exceptional” means what it says, an exception to the new default established by Bayh-Dole–see 35 USC 202(a)) but it is rarely used, difficult to use, and usually results in a screaming fit by universities and their front organizations.

Thus, universities (and their research foundations, led by Howard Bremer at the Wisconsin Alumni Research Foundation) turned flexibility into liability and used the altered rhetoric to impose even greater claims to ownership of inventions and anything that could be called an invention via their policies. It was a simple but clever loop. Make the federal government require whatever you wanted in federal policy. Then tell the university it has to change policy to comply with the federal requirements. Once the university has changed policy, then it doesn’t matter if you misstated the federal requirements–the university policy has changed and is what it is. The rationale for the change simply falls away. One gets policy that was never debated because the changes were represented as necessary to comply, not because they made sense or were equitable or might lead to greater effectiveness in serving the public through the development of research findings.

The university claim to ownership of inventions then was backed up by Bayh-Dole. It didn’t matter that Howard Bremer at WARF and Norman Latker at the NIH drafted the law and regulations to rig the system to solve Bremer’s problem that Wisconsin lacked a patent policy at all, and then certainly not a policy that claimed inventions so that they would be passed by administrative process rather than voluntarily to WARF for management. So federal patent policy came to displace university patent (and non-patent) policy, to make university patent policy that much more grasping. Very clever in its way, but utterly failed as effective practice.

As a consequence of this change federal policy to force change in university policy to efface the defaults established by Dubilier, university inventors are forced to use the patent system. Nothing in federal patent law forces inventors to use the patent system. Even Bayh-Dole, made a part of federal patent law for some mysterious (and, well, stupid thinking its clever) reason, does. not require inventors to use the patent system (though it is misrepresented to claim that it does). Somewhere, then, university patent policy must force inventors to agree to be forced to use the patent system. There’s another role for university patent policy.

Why would a university want to force inventors to use the patent system? The decision to do so repudiates university policy on freedom to publish and freedom of research. A patent is, at the heart of its Constitutional purpose, a publication. A university policy that forces inventors to assign ownership to the university becomes a policy that forces inventors to publish in a forum and manner of the university’s choosing–in the patent system. Or, we might say the decision to force publication if a new thing comes within the broad definition of invention as “everything” is not authorized by university policy, which typically makes freedom to publish and freedom of research as fundamental to the university itself. University IP policy is not fundamental in this way, but rather something more like a potentially damaging parasite, like a tick or ringworm. But the dag can wag the dog.

(3) There’s the invention made “outside” the university but not “extramurally.” Such inventions, as depicted by university policies, arise with faculty using their own time (as if they obligate their time in some way otherwise except by special agreement–they certainly aren’t employees for their academic work). For equitable title situations, use of “own time” and “own resources” is not material. If someone is hired to invent, then where and how they do it doesn’t much matter. And if they don’t use an employer’s resources or time, then the employer doesn’t even have a good claim to a shop right. In the case of the university, “own time” can also mean consulting or “outside work.” It gets odd. University policies on faculty consulting establish how much time can be spent in outside work and with what levels of reporting and approvals necessary. But that’s only if the faculty member has some agreement with some other organization, typically a company.

There’s nothing, really, in university policy that demands that a faculty member seek approval before thinking about a problem at home, or working on an idea in the garage or coffee shop or wherever. And unlike building a prototype or testing it or adapting it and the like, the actual bit of inventing doesn’t require resources, doesn’t require significant time. In a way, it is another absurdity of university patent policies to worry about inventions “made using university resources.” Inventions are made by conceptions of the mind that an inventor recognizes as entirely new to the world, and yet useful and not obvious. Inventions are never made by university resources. An invention might be “actually reduced to practice” using university resources–shown to operate as conceived. But then a university claim to ownership based on use of resources serves as a disincentive for university inventors to bring their inventive ideas into the university. Only the desperate or unaware would do that. Policy that smells like Moloch spirit.

University officials, recognizing that there may be weaknesses in their Death Star for destroying opportunities on Planet Innovation, came up with a new policy scheme to claim any invention (broadly defined to include non-inventions) made “under the auspices” of the university. Now “auspices” is almost like saying “anything that an administrator thinks ought to be connected to the university.” And officials added a claim to anything that could be considered within the area of expertise for which the faculty member has been hired (ignoring that faculty members are appointed, not hired, except when they expressly agree to be hired). This provision in policy amounts to a non-compete clause with an ownership provision–if you compete with the university’s research programs and invent (make anything that might have value, implied whenever you resist assigning it to the university), then the university claims that invention. Do not resist.

Thus, if a faculty member gets into an outside relationship, university policy claims it has priority over any contract that would place ownership with the outside entity. Again, non-compete, plus overreach, plus squash opportunity and relationships, all in the name of compliance, creating an unimaginably huge number of claims on faculty work (and staff work, and student work, and volunteer work, and informal collaborator work), swamping the IP office with work that’s not just untoward, and useless, and self-damaging but also impossible to manage. Inventors that demand that the IP office manage what tbeir policy says they manage are dubbed “problem inventors.” But the problem is the Molochian policies that they uncritically serve.  Give low status people inordinate power within an institution and they are likely to turn cruel before they turn gracious. There’s another aspect to substituting university patent policy for the Dubilier defaults–faculty inventors end up dealing with low status tech transfer folk who have the power to ruin opportunities. (Hey, I was one for a long time, you have to commit every day to making things work for those that ask for help not to make things work according to some policy demand (which is only a default–that is, the simple way but often the ineffective or cruel way).

Thus, Yale could think to sue John Fenn for failure to follow their interpretation of policy while he was in Sweden receiving a Nobel Prize. Can you think of anything so cruel. Or the University of Missouri, faced with a compelling claim of incompetence in its technology transfer program from a faculty member that sought administrative review, could turn around and sue the faculty member and then claim that no administrative review could be done while there was “pending litigation”–construing a policy statement that was clearly directed at folks complaining to stay within the administrative system if they wanted administrative review as if it was an administrative tool by which low status folks in administration could sue people in retaliation for complaints and so prevent any university review of their own practices.

Dubilier, then, motivated Archie Palmer to get universities to adopt patent policies. Once universities had patent policies, these went beyond the fairness established by equitable title and shop rights. Committees had to decide what was equitable. And when that work got too time-consuming, policy claiming inventions replaced committees. And once compliance replaced equity then policy makers were led to claim more and more, starting with use of university resources, and not bothering to release claims. Once the policy claims expanded from federal compliance to everything should be treated consistently (as if it were in federally funded work, even if it wasn’t), we got to everything even if not using university resources if it was “under the auspices” or used expertise that one was otherwise appointed for or had “commercial” value even if there was no ownership theory to be had. University policies claim ownership of invention and non-invention, patentable and non-patentable, ownable and not ownable.

A better way would be to have no patent policy and rely on the Dubilier defaults, using the highest status folk in the university to decide any close cases. In general, in my experience, people in the university who do new things and feel well treated are generous and share their benefit with the university and others–and the less demands placed on them (with the implication that they won’t share, ever, otherwise) the greater their generosity. I know, generosity is not something that shows up in university policy. That in itself is telling.

Perhaps generosity is an essential element of innovation. Perhaps generosity–including looking out for others, creating opportunity for others, expecting freedom rather than service–is important in innovation, in technology transfer. Where are the workshops on being generous? Oh, here’s one.

This entry was posted in Freedom, History, Innovation, Policy and tagged , , . Bookmark the permalink.