The (f)(2) provision in the standard patent rights clause authorized by Bayh-Dole makes everything happen. It is the key to understanding how agency procurement of patent rights works. It gets at how a federal agency, through a funding agreement, by means of a patent rights clause, can reach to the individual inventors within a university or other organization, so that if necessary they can assign invention rights directly to the government, if they haven’t found an approved invention management agent to work with, or that agent screws up. The (f)(2) provision does this by requiring universities, under the funding agreement, as a condition of accepting the money, to include their employees as parties to the funding agreement for the narrow purpose of reporting inventions, assisting in the filing of patent applications, and establishing the government’s rights by assignment or license, if it comes to that.
Since inventions are owned by inventors, it makes a lot of sense for the government, or any sponsor of university research, for that matter, to have a direct relationship with each possible inventor on a project. The federal government has the added advantage that when university investigators invent and become for that little bit of invention management parties to the funding agreement, and therefore contractors for that little bit of management, they do so through the agency of and delegations from their employer, the university, so that there can be no possible conflict between the employer and the government as to the employee’s obligations. (f)(2) creates the patent policy, the tech transfer office, the invention agreement, the exclusion of conflicting employer requirements, and the special performance by inventors all in one provision of the standard patent rights clause. Everything that a university needs to comply with the standard patent rights clause, up until electing to retain title, is built right into the federal funding agreement via (f)(2).
One might ask, fine, brilliant even, but why not just decree that universities own faculty inventions, or require universities as a condition of funding to require assignment of all inventions? It’s not there. At best, it’s a “presumption” that universities will do this. It seems to be a heck of a presumption if the federal policy really had been to demand that universities own all inventions made with federal funds. It can’t be that the idea wasn’t considered. It’s just so obvious. Bayh-Dole would not have to be a vesting statute at all. Inventors would still own their inventions, initially. The standard patent rights clause would just require universities to take that ownership away from them, as a condition of the funding. But something kept Congress from going this direction with the law, and the Department of Commerce with the standard patent rights clause. The law doesn’t require universities even to have a tech transfer office. All that is required is that they have personnel designated as responsible for patent matters. Those folks don’t even have to be employees. What kept the law and patent rights clause from just doing what university administrators are now overwhelmingly, at least for a time, so keen to have–outright title to everything their faculty create, but for those worthless traditional academic works, which are undermined anyway if all people can do is read them and avoid practicing anything they teach, because practicing without a license, and without paying, would be infringement?
Part of the reason must have been the situation with invention management as it existed in the late 1970s at universities. A number of universities had technology transfer offices. These were not licensing offices, but front offices to support the movement of inventions from the faculty to licensing agents such as Research Corporation. Others had affiliated research foundations on the WARF model, including Purdue, Kansas State, and Washington State. Some very few–California, Stanford, MIT, especially–had an internal licensing operation, run by the institution. Some universities, even big research universities, had nothing. For instance, the University of Washington, one of the largest recipients of federal research funding at the time, had no patent policy to speak of, no technology transfer office or licensing function. Washington had an official designated part-time, along with other duties, to deal with patent issues if they came up.
Given Bayh-Dole, what sort of standard patent rights clause will work for all these various ways that universities were dealing with inventions made by their faculty (and others)? It’s one thing to limit how the federal agencies contract to take ownership of these inventions as contract deliverables. It’s another to figure out how to allow all sorts of arrangements at the university level while maintaining a uniform patent rights clause. There is no question that some advocates for Bayh-Dole intended for the Act to hand ownership of inventions, somehow, to patent management organizations affiliated with universities. Would that be expedient, like vesting? or overt, like requiring it in funding agreements? or roundabout, by backing off the federal agencies and letting whomever had the urge to swoop in when there were no remaining protections for faculty inventors? To get Bayh-Dole passed, it would appear that it was politic not to decide, and leave that for the drafting of the standard patent rights clause. That all but ruled out vesting, since that would require a number of changes to patent law, and also to the nature of how faculty work for universities.
If we look at the situation from the point of view of someone charged with drafting a patent rights clause to be used by all agencies, and to be acceptable across a diverse range of university practices, where universities do not direct and control the work of their faculty, it becomes clear that we will have to cover a number of contingencies. There are two phases to the work. First, the standard patent rights clause has to distribute obligations as work, money, and rights are distributed by the university contractor. Then, the government has to obtain the rights established under these obligations, tracking each as the conditions occur that enable the rights.
Here is a diagram of how it all looks from the perspective of the federal funding agency. It is required to use the standard patent rights clause (SPRC, 37 CFR 401.14(a)) in its funding agreements with universities, so the first point of contact in distributing obligations to protect the government’s interest is the university. When the university accepts the funding agreement, then it becomes a contractor. The university has one obligation under the funding agreement, and that is to implement the (f)(2) agreement with its employees. Not all employees, just those other than clerical and non-technical ones. Note, this is much narrower than the typical all-encompassing version that university administrators try to use these days. The government does not care about non-employees, independent contractors, volunteers, visitors, or even employees who could not reasonably be considered as potential inventors–even if these folks indeed could invent. The government is not bargaining for rights to their work. Whatever the demands that drive administrators, compliance with federal grant conditions is not one of them. This is doubly true, because administrators don’t implement the (f)(2) agreement, either, so clearly they just don’t care.
Back to the diagram. When a university implements the (f)(2) agreement with its employees, it sets up a conditional delegation of certain requirements of the standard patent rights clause to those employees. When any of those employees invent, and the invention is within the planned and committed activities of the federal award to which the SPRC pertains, then the employee, now an inventor, gains new standing: the inventor is a party to the funding agreement, has obligations to the government independent of the university (such as reporting subject inventions to the university and establishing the government’s rights in the invention), and is, therefore, a contractor under the definitions of funding agreement and contractor at 37 CFR 401.2. The critical difference is that the inventor, unlike other contractors, does not have the pre-approved right under the SPRC to retain title. This is the cornerstone of the whole operation.
The (f) provision in general implements, as a condition of federal funding, the fundamental minimum requirements that the university must meet to manage inventions. Beyond this primary, absolute requirement regarding making employees have standing to become parties to the funding agreement, for the purpose of performance of certain duties under SPRC, the university may also transfer SPRC obligations in other ways. Under SPRC (g), the university, if it subcontracts the research, must flow down the SPRC to the subcontractor, and if the subcontractor is not a small business or nonprofit, then the SPRC must be modified for the appropriate clause required by the funding agency. In any event, the subcontractor then becomes a contractor, as well, and has independent standing with respect to the government, in the context of the federal funding agreement, to retain rights to subject inventions, providing the subcontractor also implements the (f)(2) agreement and related (f) provisions:
The subcontractor will retain all rights provided for the contractor in this clause, and the contractor will not, as part of the consideration for awarding the subcontract, obtain rights in the subcontractor’s subject inventions.
It is clear that the consideration in a subcontract is entirely covered in the services rendered *to the government*. That is, the money of the subcontract cannot both cover the services and any resulting rights in inventions: not title, not a license, nada. This principle also should guide reasoning with regard to how the (f)(2) agreement operates. When the university requires this agreement, it amounts, in effect, to a subcontract with a statement of work restricted to key aspects of what only an inventor could be expected to do: recognize an invention and report it, sign paperwork required of inventors to file patent applications, and establish the government’s rights in the invention. If the university cannot use access to government funds to claim invention rights from subcontractors, then it stands to reason that they also cannot do so with regard to their own employees who invent. It is not that the university cannot work agreements with employees to obtain invention rights; rather, it is that the university cannot use access to the federal funding as the consideration for that agreement.
Consider it from another angle. A faculty member writes a grant proposal, offering to do research in a given area. This goes to the university to be submitted to a funding agency for review. The university, in this transaction with the faculty member, approves the release of the faculty member’s time commitments to the university to shift a portion of his or her salary to the grant, should it be awarded. At the same time, the university commits to the government that the university has and will provide the resources necessary to support the proposed work. The government, in turn, will pay the full direct and indirect costs involved in the grant. Thus, the university cannot then double back and tell the faculty member that a condition of having access to the federal funding is that the university owns all inventions that might be made by the faculty member or anyone working on the grant. There is no consideration in that deal: the federal money pays for the work, the (f)(2) agreement obligates the government’s rights, and the university’s own rights in inventions arises only *after* it has obtained those rights. It cannot be that the consideration for obtaining those rights is the approval to participate in the funded project, using the government’s resources.
The point of this is that the university does not have standing, either with regard to subcontractors or its own employees, in demanding rights in inventions as a condition of delegating work. For subcontractors, delegating work is optional. For employees, it is not, because the university has already approved that work, and in any event, the (f)(2) agreement is mandatory. To come to have rights in inventions, the university has to use some other method. That method does not have to be in the form of a demand or compulsory requirement, nor does it have to be negotiated up front. An inventor or subcontractor could voluntarily assign or license invention rights, in return for negotiated consideration. There are plenty of ways to get there, and sometimes the most administratively enticing ways are the least effective or beneficial.
Returning again to the diagram, there are two more pathways to explore. In one, the university may also distribute SPRC obligations to an invention management agent. That is, it may assign the SPRC to, say, an affiliated research foundation, so that for the purposes of the federal funding agreement, the contractor identified in the SPRC now is the research foundation. This could be considered a subcontract, but only of the SPRC work, but it also works under SPRC (k)(1), which allows a contractor to assign an invention to an organization that has a primary function in managing inventions, so long as the SPRC follows that assignment. Thus, a university can delegate its management of the SPRC to another agent, which then serves as the “contractor personnel responsible for patent matters” among other things. Depending on the scope of the delegation, the agent also may be required to provide education to university employees with regard to the importance of timely reporting of inventions and the like.
We come, then, to the last pathway remaining. Inventors, too, may assign their rights in inventions. Under (f)(2), they may assign those rights to the government. If left unmolested by administrative outreach, they have no other obligations with regard to assignment, just as small business contractors are not subject to the university limitation of SPRC (k)(1). Thus, inventors, too, can assign their SPRC interest to agents, that is, their (f)(2) arrangements. For the government, so long as it gets the rights in inventions that it is allowed to negotiate for under the SPRC, this is all fine. That’s the whole point of Bayh-Dole: the agency is not supposed to care, contractually, at least, with regard to what the contractor(s) do with the obligations, so long as the transactions involved meet what is required by the SPRC. We end up, then, with four possible classes of contractor:
Thus, it is not appropriate to read “contractor” in 37 CFR Part 400 to mean exclusively “university” or “primary recipient of the federal funding award.” Any number of contractors can be involved, each with a share of the obligations for the proposed work, and for the management of inventions that might be made in the project.
Now we can look at what happens if an invention is made. How does the government get the rights it has bargained for from any of these contractors? Let’s take a simple case first:
When an invention is made in the funding agreement, the inventor(s) own the invention, and have an obligation under (f)(2) to establish the government’s rights. Thus, if the inventors do not assign title to an agent (including the university), then the government deals with them under their (f)(2) commitment and 37 CFR 401.9 (as well as the consultation requirement of 35 USC 202(d)). The university’s role in all this is to report inventions to the government, and notify the government that the university is not electing to retain title, since, of course, at this point, it has no title to retain. This is the simplest, most fundamental form of Bayh-Dole. It is rather attractive. The university is entirely out of the invention management business, and the inventors have the opportunity to deal directly with the government. If the agency wants rights, it gets them. If it doesn’t, then it can treat the inventors as small business contractors, but with fewer restrictions. The university does not exist in all this. It is a bug in the grass.
Now the advocates of Bayh-Dole, at least some of them, no doubt will argue that this wasn’t their intent at all. No matter. The law is not about the intent of some pressure group, even those that got their hands into the drafting–it is about what the law as passed says, and how it is implemented. The base form of the SPRC, implemented via the (f)(2) agreement, creates a relationship between the government and the inventors with regard to invention rights. If the inventors want the government to hold title, and the government requests title, then all is at peace. Same for if the inventors want to retain title, and the government agrees. Any disagreement by the university about all this amounts to interference, meddling, and busybodiness. There is no presumption of university ownership in the law. If anything, it is a presumption of contractor ownership, and since, at invention, contractor does not mean “university” but rather “inventor” for the purposes of the (f)(2) part of the SPRC, the deal is between the inventors and the government.
In many ways, focusing the arrangement between the inventors and the government with no bureaucratic thumbs outreaching might be the best of all possible arrangements. Low overhead, up or down choice. “Do we want this invention?” ask the inventors. “Should the inventors have it?” asks the government. Fine discussion. Yet it is just this discussion that university administrators appear dedicated to never allowing to take place. University invention policies appear determined to prevent the operation of 37 CFR 401.9, as if it is an affront rather than the default, the simplest approach, the one that might operate 99% of the time, with the government taking rights some of the time, the inventors some of the time, and the public domain getting stuff that neither sees the point in wasting time on. In a freedom to innovate environment, in which a university aims to promote innovation by placing thoughtful restrictions on its own institutional self-interest and thus empower creative folks to make their own way, asking for help when they need it, Bayh-Dole would be the great enabler, and 37 CFR 401.9 would be the first thing that any university policy guidance document would champion. (“When the federal government supports your research, you have the opportunity to retain title to your inventions, provided you persuade the funding agent that this is the best thing to do. The university has resources that can help you with this discussion. Call 555-LOVE” etc.)
Now consider the situation in which the government wants title but the inventors feel fond of their invention and don’t think government ownership is the thing. The SPRC says there is a way to pre-empt the government’s request for title under (f)(2); namely, the inventors assign title to an organization that is pre-approved to retain title if it gets title. That is, to a contractor that has the right to retain title. Back to the diagram of contractors. One contractor for sure is the university, since it has to be a contractor to get the whole SPRC thing going. But other contractors are also possible, through assignment of SPRC obligations, or by assigning title to the invention, and with that title, the SPRC obligations that go along with title. That is, the SPRC rights can be distributed before there is any invention, or after. Bayh-Dole handles both cases. In practice this allows a university to stay entirely out of the patent administration business by delegating all that to one or more agents. How the delegation may happen is also a matter of some interest.
In the case that many folks are fixated on, the inventors may assign to the university, as it is necessarily a contractor, and it has the right to retain title even if it has no capability whatsoever to manage inventions. There are gaps in Bayh-Dole. This is one of them. Why should universities be permitted to retain title if they have no invention management capability? Why should the SPRC permit a university from demanding title if it has no invention management capability, or if in the opinion of the inventors, that capability is not adequate to manage the invention that they have made–whether through lack of expertise or resources, or because the institution is compromised by conflicts of interest? Whatever the case, if the inventors assign to the university, then the university can elect to retain title, the government agency gets its royalty-free non-exclusive government license, and the other provisions of the SPRC trigger, requiring filing patent applications, preferences for small business, reporting on utilization, sharing royalties, and the like.
There is, as we have developed the situation, no requirement that the university demand ownership of inventions. It certainly isn’t in Bayh-Dole or the SPRC. The demand bit exists only as a political rhetoric by pressure groups that have been determined for thirty years to find a way to take faculty inventions for speculative management, thereby creating a new industry of administrative bloat, fueled by the prospect of a “big hit” deal which is about as likely as a winning Lotto ticket. The demand-title claim aims to make academic freedom and independent scholarship subordinate to institutional money needs, citing “public benefit.” The defects in such an argument are many. An obvious one has to do with the implicit claim that faculty do not serve the public, only administrators do. A less obvious one has to do with the question of whether trying to make money is the same thing as making it, and when a lot of money is made, does it actually go to benefit the public in a meaningful way, or does it become another form of slush fund to feed administrative bloat.
The point is: the source of the demand for university ownership does not come from federal policy, and relies in large part on the repudiated idea that federal law vests, or requires, or mandates, or endorses, or encourages, or presumes university ownership. The fact that the pressure groups cannot even get straight what verb to use shows how empty their argument has been. It took the Supreme Court in Stanford v Roche to put a stop to it. But the echo of the argument lives on in policy statements, guidance for faculty, handbooks of technology transfer practice put out by front groups for the demand-advocates, and academic articles on Bayh-Dole that don’t show much evidence of having an acquaintance with Bayh-Dole or the SPRC. Rather than give up their position, as this is as much about appearing to be right as it is about preserving power however ill-gotten, the advocates now aim to backfill the demand argument with other reasons. Time will tell whether the faculty, or the public, or legislators, will tolerate the demand-advocates much longer. Hopefully not.
There is a broader, more wholesome point to all this, however, and that is that universities really don’t have to demand much of anything. It is apparent that if faculty inventors wish to resist a government request for title, they very likely, and with happy hearts, will turn first to the university, the host and supporter of their work. Why demand what one will likely offer anyway? The stronger business position, and social position, is to prepare the resources and offer assistance, rather than attempting to dictate ownership and making up dark claims to justify it (like, “employment” or “use of resources” or “right to do research” or “business of the university”). Another good reason for not demanding assignment by policy, for public universities at least, is that a demand for assignment amounts to a taking of private property for a public purpose. That’s eminent domain. That requires due process and just compensation. “Just compensation” is way different from stipulating one has to pay only if one makes more money than expenses–no, just compensation means paying up the value of the asset at the time it’s taken. Royalty sharing schedules created for voluntary assignments don’t work the same way in a demand environment.
Thus, at least if the dark claims are represented as “consideration” for the assignment, one at least has the appearance of a negotiated agreement, so then the institution can, if it does some clever drafting, get away with the idea that it has already paid for the invention (via employment and the like) and therefore any royalty sharing is just a perk, not part of the deal by which the university obtained title. Shifting the theory on which royalties are paid is a big thing, because making royalties a matter of perk rather than privity means that the inventors are cut out of the arrangements. In a corporate world, fine. Do it that way. But when the inventors are faculty, and their scholarship, their careers, their independence is also on the line, then it’s a cowardly, low brow trick to cut them out of the decision making with regard to licensing strategies and execution. The better approach is to offer consideration in a negotiated arrangement to assign. That offer can be a standard deal on royalty sharing. A university might adopt a standard deal so that (a) there is not a lot of time spent haggling; (b) folks don’t get special deals for being special people; (c) everyone, including the public, knows what the deal is for any invention under management. If the inventors don’t like that, they can find another agent, or go back to deal with the government. And, of course, a university could choose to be flexible, and accommodate special circumstances, perhaps with the stipulation that any such arrangements have to be made public so everyone sees what is possible and why.
To add one more point to this whole royalty sharing thing, university administrators on the demand-side of things see university ownership as a right, and the royalty sharing schedule as a form of institutional generosity. This is low brow, but understandable. It assumes that the university is the owner of inventions, and that faculty should be happy for any recognition at all. In a voluntary assignment approach, however, it is the inventors that decide on their own generosity. A university could–though it would take administrators with courage and character and insight–ask only for recovery of its expenses, and allow the inventors to keep the remainder, or offer some or all of that remainder to the university, or to a unit of the university, or to another cause or organization. (This works, as well, under Bayh-Dole and the SPRC, since the use of “remaining funds” is what is constrained to “scientific research or education” and even there, the stipulation is not that the contractor must be the one to expend the those remaining funds *on itself*.)
Now we consider what happens if the inventors don’t want to assign to the university, or the university does not want assignment of title. There other options. The university may assign its SPRC obligations to an invention management agent. The university may even allow the inventors to designate which agent this will be. There need be no monopoly for any one agent. The assignment may be made directly to this agent, or may go first to the university and then to the agent. It all depends on how folks want to set up the arrangements. From the government’s perspective it doesn’t matter, since the SPRC is about how the government gets its rights, not on how universities profit, or how commercialization happens, or how tech transfer offices can make themselves look good and so capture more money to spend on themselves.
It is also apparent that inventors can also assign their inventions without the cooperation of the university-as-contractor. Small business contractors do not have any limitations on assignment of title under the SPRC, unlike universities, which are constrained by (k)(1). Small businesses, if they obtain title, and elect to retain that title, can do with it what they want–assign it, sell it (they would have to, especially, if they were acquired), license it. Inventors, if they hold title, are in the same condition as small business contractors, per 37 CFR 401.9. If they assign title before they are granted the right to retain title, however, there’s a problem, because such an assignment carries with it the (f)(2) agreement and until that’s resolved, it would appear that the government still has a right of approval for such assignments, or one might say, a right to disapprove and demand assignment to the government. This is one of those things that makes one wonder why the message going out about Stanford v Roche is the need for present assignments. A present assignment would have done nothing for Stanford, given the facts of the case. An (f)(2) agreement, however, signed by the post doc upon starting work with the federal funding, would have attached to the present assignment to the company for any subject invention, and while that would not have given Stanford ownership and therefore not helped it regret its policy and business decisions, it would have placed SPRC conditions on the company, which may have involved agency approval of the assignment or a finding that the (k) conditions of Stanford’s SPRC apply, meaning that the company would have an obligation to share royalties with the inventors and use the remainder after costs for scientific research or education. Such an outcome might not make some Stanford administrators happy, since they are still out their quarter billion (which they could have gotten by licensing non-exclusively to others, perhaps), but from the inventor’s point of view, it would all be good. Same for the public, with regard to use of taxpayer supported assets.
The message from Stanford v Roche, therefore, is: get the (f)(2) arrangement in place. That’s what protects university inventors with regard to royalty sharing. That’s what protects the public investment in research when resulting inventions are assigned to for-profit concerns that do good things and show a profit for their management of the invention, and that’s what protects the government’s interest in rights to the invention. Saying that the problem fix is present assignments is gross denial of the situation.
To complete the picture, here is a schema showing a range of ways that the government can expect to obtain the rights in inventions it has bargained for. It can get those rights, initially, from the inventors. If the inventors assign to the university, from the university. If the university engages an agent, from the agent, and if the university or agent licenses exclusively, from the exclusive licensee. If the inventors do the assigning, then if it is to an agent that has pre-approval, then the government gets rights from that agent. If the agent doesn’t have pre-approval, then it has an unresolved (f)(2) obligation from the inventors (“entire right title and interest”) to deal with. That’s the best I can make of it.
Sadly, universities have refused to implement the (f)(2) agreement, and have sought to conflate a private agreement with the (f)(2) agreement, as if it was federal policy to require universities to get ownership, when clearly that is not the case, and the universities furthermore have sought to pre-empt the need for (f)(2) by trying to obtain all inventions with a perfunctory, open-ended, hopelessly ambiguous imposition of a present assignment. All of which is needlessly convoluted, complicated, fussy, iron-fisted, petty, and inappropriate to university affairs. And all that fuss cannot possibly serve innovation.
If all this federally funded research really is about practical application and public benefit, then the university administrators fixated on money and assertion of authority need to stand down, take a breath, and try something refreshing for a change, like complying with the SPRC and giving faculty inventors the opportunity to choose how their scholarship will be managed. That choice fueled decades of productive faculty support for innovation, and that choice laid the foundation for the Bayh-Dole Act itself. It is reasonable to expect that faculty choice, if it is once again allowed, will contribute to research innovation, across a broad spectrum of activities, from public domain to venture backed startups, using conventional and innovative approaches, as they have since at least Frederick Cottrell’s creation of Research Corporation in 1912.