In the past I’ve tried to outline how Bayh-Dole works. What this means is how the standard patent rights clause in federal funding agreements works. This in turn requires us to get at why the (f)(2) agreement is so essential.
Recall that the (f)(2) agreement is required by 37 CFR 401.14(a)(f)(2). Contractors are required to require it. It makes potential inventors party to the federal funding agreement and obligates them personally to protect the government’s interests in subject inventions, which are inventions “of the contractor” because until they assign rights to someone else, they are also contractors for the purposes of the patent rights clause.
Here I would like to lay out how the (f)(2) agreement sets up agent decisions. The inventors are not “free agents” under federal funding agreements, but they do have choices, until they agree otherwise. In the book of innovation support, choice is good. I rather like it that way.
In essence, the standard patent rights clause, or SPRC, sets up three trees of invention support that an inventor may pursue. I’ll put these on a slide, but my WordPress install is having some problems with its media library, so I’ll do it in text first and work that later. [I have now added citations]
Private Agents Inventor(s) Funding Agency
1st option Initial title If 1st option not taken
35 USC 261
University-managed Agency-managed
SPRC (b) SPRC (d)
University-approved Public domain
SPRC (k)(1) <no obligation to file?>
University-approved and Inventors as own agents
Agency-approved 37 CFR 401.9
SPRC (k)(1)
37 CFR 401.13(b)
Under federal patent law, the inventors own title to their inventions. The (f)(2) agreement makes them party to the SPRC and that agreement supersedes any claims that the university might otherwise attempt to make under policy or private agreements with its employees. That’s because the university cannot require its employees to do two contradictory things within the same subject matter. It cannot require employees to have the right to establish the government’s rights in subject inventions at the same time that it claims to take those rights for itself. Besides, (f)(2) is part of a federal agreement, and that trumps agreements enforced in state court, including the university’s own agreements.
The thing to notice is this. The SPRC presents the inventors with a fundamental choice:
Find an acceptable agent to manage your invention, or the funding agency will get to choose that agent.
That’s the core of it. That’s the genius. That’s where all the flexibility, responsiveness, and opportunity in Bayh-Dole gets baked into the relationship with the universities that finds its way to what the federal agencies really want, and that is the special, distinctive expertise of the university investigators. There is next to no interest in what administrators may say or do. Federally funded research at universities is about the investigators. The university serves as a steward, a big institutional butler, to handle the grant logistics, for which it is decently but not amply paid, and otherwise, it is to pretty much stay out of the way of the work. The (f)(2) agreement is the government’s way of ensuring that the patent rights clause makes its way all the way to the inventors, and is not intercepted by the university and manipulated for its own purposes without the inventors having a say.
How does this all work, then? When there’s a subject invention, the inventor is asked to make a choice of agent. If the inventor chooses the university, then that’s a pre-approved arrangement and the agency has nothing to say about it, provided the university performs its obligations. Bayh-Dole relieves the agency of the responsibility for verifying the condition of the university’s tech transfer office, and whether it will be up to the task.
Further, the inventor could choose any other agent that has invention management as a primary function (per SPRC (k)(1)). But this requires the approval of the university. The inventor could also choose any other agent, one that does not have invention management as a primary function, but this would require the approval of both the university and the federal agency. The approval of the university may take either of two forms. The inventor may assign directly to the agent, with the agent accepting the conditions of the SPRC along with title, or the inventor may assign to the university, which then assigns to the agent.
If the university obtains assignment from the inventor, it may assign to a pre-approved agent without the approval of the inventor, unless of course the inventor and university have an agreement that gives the inventor a say–which might be a very good thing, for the continued assistance of the inventor!
Some folks have argued that the university can only assign to a nonprofit, but this is not in the SPRC. There is no limitation there on whether the agent is nonprofit or for-profit, only that it have a primary function managing inventions. Thus, Intellectual Ventures is as good as Research Corporation or the local research foundation. The important thing is that the assignment carries the SPRC obligations of the *contractor*. It’s not just to “title” but to the “entire right, title, and interest” and that includes the SPRC obligations along with everything else.
If the inventor cannot find an acceptable agent, then things move to the right side of the diagram, to the federal agency. The agency then must decide whether to request title and be the invention agent, or allow the inventors to be their own agent. This choice is set out in 37 CFR 401.9 and follows from 35 USC 202(d). Under 37 CFR 401.9, the agency, if it chooses the inventors as its agent, will treat the inventors as if they were a small business contractor under Bayh-Dole. Essentially, the agency writes a new patent rights clause for the inventors, different from the one university/nonprofit SPRC, eliminating the assignment restriction, the use of royalties, and the requirement to try to find small business licensees.
If the agency doesn’t select the inventors as its agent for developing the invention, then it can choose to develop the invention itself, nominating itself as the agent, or it may allow the invention to enter the public domain, without any patent ownership claims.
By recognizing this agent choice structure in Bayh-Dole, one can understand how the law is at once uniform for agencies and wonderfully flexible for inventors–provided it is allowed to operate. The law, in essence, says: “If an inventor can find an agent to develop a subject invention, then within bounds, let the inventor do this. It’s not your problem, agency, to second-guess the inventor’s choice.” The requirements of Bayh-Dole and their distribution under the patent rights clauses and assignment agreements does everything needed to protect the government’s interests and further the public objectives set out in the law.
It now may be obvious how universities, acting in their own self-interest are interfering with the inventor’s choices as provided by the (f)(2) agreement and Bayh-Dole. The universities failed to implement (f)(2) and swapped in instead assignment obligations to the university. They interpreted “elect to retain title” as “entitled to take title” or “have title outright” and proceeded to change policy and education to faculty to make it appear that university ownership was federal law. Stanford v Roche made it clear that it is not.
The universities told the public that Bayh-Dole was about their ownership, when it was really about allowing inventors to choose the best agent to develop their inventions. They told each other that Bayh-Dole required policies and employment agreements with compulsory assignment of inventions in the case of federal funding, when Bayh-Dole did not require assignment to the universities. And the universities told themselves that (f)(2) was satisfied as long as the government got its license to inventions and all the reporting went on as specified and royalties got shared with inventors and any remaining funds went to scientific research or education–and most of the funds seem to have gone to research, and research only at the universities claiming ownership, and not to anyone else.
How to account for such a systemic, long-lasting misinterpretation of Bayh-Dole? I’m not sure on the origin and history. But it does appear that the prospect of ownership gave universities a huge bolus of organizational conflict of interest. Should they stay on mission and allow Bayh-Dole to operate as designed, leaving decisions among these various options with the inventors? Or should they use the power position of employer and contract host to force ownership to their own technology transfer offices, for the purpose of making money while selling the public on the benefits of doing so? Clearly, all but the strongest universities caved in and chose themselves. Hats off to Wisconsin for keeping at least the minimum choice, between WARF and the funding agency, as one the inventor makes. And hats off to all the universities that left their policy to say they would only take ownership when required by an extramural sponsor. Those policies comply with the full choice spectrum under the SPRC, because the SPRC does not require university ownership.
One might see, then how the move to present assignments is another outgrowth of the deep-seated organizational conflict of interest that has been allowed to operate mostly unchallenged for up to 30 years. However, the present assignment does not improve a university’s position with regard to compliance, but rather deepens the problems. For one thing, a present assignment is about how an assignment is executed, and does not address whether the assignment in fact operates. An open present assignment, such as the ones universities are using, is very strange indeed. It purports to be self-executing for anything the university decides later it applies to, but then it apparently is intended to have already self-executed at the time whatever that thing is came into existence. Given the uncertainty with regard to what possibly could be involved in such an assignment, it is no wonder that the University of California and others in an amicus brief to the CAFC on Stanford v Roche argued that present assignments used this way may not hold up in court. No wonder. On that point, the University of California was very right.
A second problem with using present assignments on faculty, in particular, is that it is they, not the university, that decides what is going to be in scope of a university claim and what isn’t. Long before someone in a tech transfer office gets uppity about things, it has already been largely decided. If a faculty member has an idea and decides to use university resources, they have already walked into the restaurant with the intent to pay for the meal. If they don’t want the university involved, they work the idea off campus, outside of university funding, often through consulting deals, but maybe in a startup, or maybe just in their own space. The present assignment changes how an assignment happens, but it doesn’t change the scope of the claim.
But there’s another problem with present assignments. They do not operate at all relative to (f)(2) agreements until the subject invention has been made, and the inventors have decided that the university is to be the agent. That is, the obligation to assign is distinct from the means by which the assignment is executed. (f)(2) says that the inventors are required to have a written agreement that includes the power to establish the government’s rights in inventions. How can they do that without actually having those powers? The university, if it requires the (f)(2) agreement waives any standing claim to ownership under its own policies in favor of the rights established by the SPRC.
The university institutional conflict of interest has led universities to work to prevent Bayh-Dole from operating. While aspiring to all good things–public benefit, use of inventions in the public interest, gaining resources for the university, showing the benefits of technology transfer–they have worked against the very law that gives them livelihood.
Nothing in Bayh-Dole requires compulsory ownership policies. Nothing requires present assignments. Nothing mandates that universities prevent by contract any number of choices from being considered by inventors. It is, after all, the inventors who have the most to say about who they will work with, under what terms, and for what purposes, and with what registers of engagement. Efforts to prevent them from these choices work against Bayh-Dole, against federal innovation policy, and against the universities’ own best interest.