One can read Bayh-Dole to be a means of stripping university research personnel of their invention rights. This is the tornado view of Bayh-Dole. Wherever there is federal funding to universities, because Bayh-Dole says the university may elect to retain title title, this means that when the university elects to do this, it actually gets title, sucked out of whatever other commitments anyone in the little towns below might otherwise have. They should know this, and not make commitments that they would regret later–perhaps years later.
If one is intent on getting title and one routinely gets federal funding, then the tornado view makes a lot of sense, and one reads the law looking for those bits that wind up the tornado. My observation is, if one thinks it is possible to invoke Bayh-Dole to call down a federal invention rights-sucking tornado, for it to be Constitutional, one has to be ready to pay for the damage this causes as well–regardless of the aspiration to public benefit.
However, one can also read Bayh-Dole as providing a means for universities to gain access to patent rights that are available to be had, ahead of an agency’s interest in those same available rights. This is where the CAFC is, and what the AUTM hornet’s nest is all so upset about. We can call this the town view. Imagine there’s a town of invention commitments. They come with the territory. Employers ask for inventions from employees, or don’t. Same for universities. And there are independent inventors making deals with companies or with investors. Lots of places for future invention rights to be secured with contracts. All normal and healthy and expectable in an innovation society.
In the part of town called university research, each university has its own responses to these commitments. Some allow their employees to hold title and others do not. Some allow research sponsors to own inventions and others do not. In the town view, Bayh-Dole doesn’t say, cancel all those things. It says, if you want title the way federal agencies want title, then you have to do these things. And what is it that federal agencies want out of title? Use leading to public benefit. If that happens, then it is all good. It doesn’t much matter who does it. In the town view, Bayh-Dole shows up as another set of commitments, not to displace prior commitments but to work with them. No tornado. But federal deal-making agreements.
We can ask, what does the government want out of university research? From the tornado view, it wants title to inventions, lots of title, entire title, all the time, everywhere. It’s the university’s job to stop this voracious appetite for title and take title instead. But in the town view, we might notice that there is nothing in Bayh-Dole that indicates that agencies must have title. There, it is a choice agencies make.
If an agency wants title, it has options. First, it can contract with a large company, and kick in a FAR 52.227-13 clause. Combining this with a Rights in Data — General clause with an extension to require no disclosure beyond contractor’s employees without government approval gets a lot of the way there. Second, it can declare exceptional circumstances and throw in the requirements it needs to get what it wants. Universities can have trouble with this because it will look a lot like classified research and require control of access to labs. When they accept this work, universities may place it off campus, with a subcontractor, or in a lab that already has a security arrangement. Finally, an agency can wait for the university to waive and get title rather passively. Perhaps one contracts with a smaller university, or one requests that the university simply not take title up front. Doesn’t hurt to ask. I’ve seen this done.
The point is: in most university grants, the federal agencies don’t particularly want title. What they want is for the research they support to be productive. What they want is for rights not to mess up other initiatives. What they want is for folks to get on with it and quit screwing around. That is, if the program officers want anything like that at all. If they want rights, they want those government purpose rights, not title. They don’t want anyone suing the government for infringement, especially if the research is in a strategic line of work with a lot of government projects going on.
In the town view of Bayh-Dole, agencies want a non-exclusive right to what has not been developed entirely at private expense. But they rely on what is already going on in the town to get at it, not a tornado to cancel all the deals in town to take privately committed rights.
How does this recognition of tornado and town views of title to inventions affect our development of Bayh-Dole? First, it reminds us that if we have a desired outcome in mind–title to universities, say–then we have to be careful with our reading. Text can reward our desires, but it may not be all that good for the text as a social asset. That is, we get our way, but at the expense of the text that gave us the leverage. In the end, we end up with our own assertions, our power over things, rather than a shared environment in which power is distributed, aided by text, throughout the community. This latter situation, I argue, is what enables markets, where there can be barter among folks with similar positions of power in the form of buyers and sellers.
Extended to technology, we might posit that a vital role for universities is to develop the assets that distribute power broadly in the economy, so that markets can form. This, rather than aiming always with new things to be *in a market, with power*; that is, holding patent rights to be used as a monopoly to attract investment that aims to gain power in a market that allows a technology to be held in favor of a product sold at margin without competition, rather than to use the patent rights as a means of distributing power so that a market might form around the technology, resulting in time in products.
In this, one might begin to see, if it is not totally obvious, why referring each reported research invention to a technology transfer office committed to “commercialization” is at best only half the effort, and for universities, is more problematic than that because society does not need universities to be the source of products or to take sides in a market, but rather to be the source of revitalizing markets and creating new ones by distributing new technology throughout. This is the lesson of the internet, of agriculture extension, of open source software.
Back to Bayh-Dole then. Bayh-Dole focuses on practical application. That’s what gets a definition. Commercialization is mentioned. Yes, there are pathways for it. But the decision to focus on commercialization, to require the entire title, to license to monopolists, to make money–these are choices of university administrators, not a requirement of the law. The universities ought to come clean on this and say, “we have equated monopolists with public benefit because we think we can get more money from this than anything else we can do with patents.” At least stand tall about it, if one is going that way.
Again, it is a university choice, not a requirement of Bayh-Dole. Bayh-Dole does not care if government agencies get the entire, undivided title to research inventions. If agencies want this, then they can contract for it and protect their interests. If they want title, then they shouldn’t use a grant mechanism with universities that requires the standard patent rights clause.
Universities, if they want title, get the benefit of the standard patent rights clause to step in for the agencies, but the standard patent rights clause has only as much urgency about title as agencies in general want, which can’t be much by the time an agency has chosen to use a funding instrument that involves universities and 401.14(a). If universities want the entire right, title, and interest, then *they* have to set up their arrangements so that when (f)(2) kicks in, *they* do get everything. Bayh-Dole doesn’t do this for them, in the town view.
In the tornado view, of course, Bayh-Dole gets summoned down to destroy private arrangements at will. The university proposes work in the pathway of the tornado, gets the grant, pulls in the folks with private commitments, and destroys the interests of those on the other side of those commitments. It is heady stuff to folks into absolute power destroying absolutely. One might feel some caution, however, in desiring government rights to have such power, and for that power to be used *automatically* by private (and state) entities with such disruptive force, and with a premise of *winning* in markets with monopoly powers rather than *creating new markets* with them. Just takes my breath away, what AUTM and its 40+ universities have proposed for themselves, taking the tornado view.
In the town view of Bayh-Dole, inventors are connected to agency interests via (f)(2) written agreements. Universities are required under the standard patent rights clause to get the (f)(2) agreement. If the universities don’t do this, then the universities don’t have a mechanism to retain title having elected to do so. That is, in the town view, electing to retain title is not the same thing as electing title. In the tornado view, when one gives notice of electing to retain title, one has title. One has called down the tornado and gotten title.
In the town view, all one gets in electing to retain title is the privilege under the standard patent rights clause of directing one’s own research employees as to how to meet their commitments under (f)(2). In essence, in the town view, a university under the standard patent rights clause agrees *not to invoke its own IP policy* until it has elected to retain title, and only *then* can it use its IP policy to provide the written agreement required in (f)(2) to meet its own, rather than agency, requests for title. If its research employees cannot comply with (f)(2), that is not time for a tornado, but rather an expectable consequence of living in an innovation town where rights haven’t been handled the way one wants. The agency doesn’t–or at least shouldn’t–care, provided the government doesn’t get sued and the patent rights are not used to prevent the use of the invention, disrupt further research, or to encourage an unreasonable use, and the like.
In the town view, the government bargains for what it can get in dealing with universities. The deal at (f)(2) is pivotal. The university doesn’t get anything more than the university is able to get for the government. The university is not required to exclude others from the research. There can be volunteers, company employees, visiting scholars, students. Any of these are not required to make an (f)(2) commitment. So for any of these, the university does not have their invention rights by electing to retain title. If the university does not deal with both the (f)(2) commitment and the commitments of others, the agency cannot care. It has no mechanism in a standard patent rights clause based grant to request title. So neither does the university under anything in 401.14(a). It will have to be something else the university has, outside the funding agreement, outside a uniform deliverable to the government.
Now we can ask, if Bayh-Dole lets a lot of title claimants slip through–at least in the town view, the non-destructive view–then is a research employee who obligates rights sometime before joining a federally funded project more like a volunteer or company collaborator, outside the (f)(2) reach, or more like a non-compliant company employee bringing the university into breach of its obligations under the standard patent rights clause? In the town view, this is one’s options.
In the tornado view, (f)(2) is a cleaver that says whatever your other commitments, even if we authorized them, even if they are in good standing, even if you were compensated for them, the tornado takes that all away and you should consider that next time you deal with any of our employees, because they–or we–can break any such commitments to anyone simply by inventing or actually reducing to practice in a federally funded grant. Federal grant = 401.14(a) = (b) = (c)(2) = (f)(2) = IP policy = is imposed now with the force of federal law = all your inventions are belong to us, and we don’t have to do anything more than elect title under (c)(2) and you can’t do anything about it. Most excellent tornado indeed.
What does one do with situations such as: one’s research employees make consulting commitments and then participate in federally supported research. One’s research employees participate in federally supported research *and* during that time make commitments to others. One’s research employees participate in federally supported research and collaborate informally with others. In the town view, this is all good, just a matter for management for what one cares about. In the tornado view, this is sin to be wiped out and put right when federal power is called down by electors.
The Supreme Court will decide between the tornado view and the town view. The CAFC says, town view. You get the rights available, and it’s your own darn fault if these aren’t the rights you discover you want. There’s no mandate to by the agency to get all those rights, and the agency remedy for not getting those rights is that you don’t have them either. AUTM says, tornado view, please. We would rather our inventors were imprisoned by invoking federal power than that we had to work with them ourselves and risk not getting the rights we desire.
What are our options, then, with regard to a research employee with prior invention rights commitments (like a present assignment of future inventions) who then invents in a federally supported grant? In Stanford v. Roche, the CAFC threw out arguments such as laches and non-compete and bona fide purchaser. The CAFC found that the inventions clearly were in scope of the present assignment.
One outcome: just like the CAFC says–that present assignment operated, there was no other assignment to compete with it, the university’s own IP policy failed because its re-imposition at the time of the federal funding lacked precedence over the present assignment, and whatever the problems folks may have with their Bayh-Dole compliance, Bayh-Dole does not save them from these problems by making it impossible for inventors to do things that universities later regret. That’s the life is hard outcome.
A second outcome. The university research employee, having present assignment to future inventions, acts as if an employee of the company rather than the university for the purposes of the federal award. (f)(2) doesn’t apply unless the company gives the employee a release, and the university cannot get at title via an election of title or a later assignment because it lacks standing under (f)(2).
But this does not change the fact that the invention is a subject invention for everyone involved. It’s just that the university cannot take title to everything. It is as if the university also *subcontracted* work to the company via its shared representative, the employee with a present assignment. Under the standard patent rights clause at (g), this means that the company if it takes title also accepts Bayh-Dole obligations. One might argue that while the university doesn’t get the entire title and therefore lacks standing to sue for infringement, the company ends up with Bayh-Dole requirements, which it may not like–such as how royalties after invention administration expenses are to be used.
In this, Bayh-Dole functions rather like work made for hire in federal copyright law. The definition pre-empts anyone else’s formulation. It’s a work for hire when it meets the federal definition. One can move ownership around however one wants, from the defaults, but one is stuck with the federally assigned attributes, such as the term of copyright and whether there’s a termination right. Those persist regardless of whether folks move title around. Furthermore, one cannot say something is not a work made for hire when it is, or say it is a work made for hire when it is not. One has to follow the federal rules. Thus, if it’s a subject invention under Bayh-Dole, then that is what it is, because it meets the definition. It is that way for the company, too. The company has to deal with it as a subject invention, no matter how they get rights in it–by assignment of title, by assignment of obligations via a subcontract, or by license (which actually carries the least obligation when it is non-exclusive). The issue then is, what are the company’s obligations for holding rights in a subject invention?
A third option: the present assignment represents not only a commitment of the research employee but also has been approved by the university’s policy, so it is also a commitment of the university. Thus, when the federal research agreement comes into play, the operative clause is (e)(1), which provides that the university can grant sublicenses to subject inventions provided these agreements are in place before the federal funding and provided that the university has reported the invention. These are both true, so the guidance of the law is that the university can honor the transfer of title under the present assignment made by its employee, even though that employee may regret the deal now. That is, there is no obligation under Bayh-Dole to break prior private deals, even when the government takes title. This might be enough on its own to dismantle the tornado view. The decision to break or honor a deal is entirely within the town, made in the university, not forced on it or made for it by a federal contracting clause.
If I were dealing with things at this point via the town view, and I wanted to stay at the idea that the company has done something wrong relative to Bayh-Dole, I would consider pushing for the company to end up with Bayh-Dole obligations if it has the benefit of the present assignment, with the expectation that the company would rather take a license than have title, and be willing to negotiate a financial settlement to get there. I would argue that the invention is a subject invention regardless of title, that the present assignment acts as a subcontract forcing Bayh-Dole obligations on the company, and that under those obligations, the company has to account for its earnings, including, say, sharing with the inventors. That might not get the university money, but it might make the inventors happier with the deal.
In the town view, I don’t see other options. There must be others. Inventors have personal rights in the town view. Universities can require obligation of those rights into the future, and can release employees from those rights at will. Any tools available in the conventional treatment of inventions are available, including commitments individuals ((f)(2)) and organizations (4014.14(a)) make, state laws, tort claims, threats and promises.
Most university IP policies do this, not as a weakness but as a strength. It is in the nature of universities for the expertise to be at the edges. Universities are not command and control organizations. They are so long-lived because they are adaptable, because their faculty at the edges respond and lay out new directions, not because central administrators have superior vision and authority and assign people their roles. The IP policies say, the university only claims rights that it has to under extramural contracts, because its resources are being used, to protect against conflicts of interest and especially integrity in dealing with patients and others in research environments, and when the university specially contracts for services in exchange for benefit other than normal pay and duties.
With these sorts of conditionals, aimed at recognizing the liberties desired of faculty, it wouldn’t matter if the baseline form of the deal on inventions is a promise to assign or a present assignment of future inventions–because the scope of that commitment would be subject to change largely at the discretion of the individuals involved. One might also see that even with a present assignment to future inventions, in a federally funded agreement under the standard patent rights clause, the university is not on the other end of that present assignment until it obtains standing by electing to retain title and then invoking (f)(2) to trigger that present assignment, on the assumption the university will take title rather than direct that title elsewhere. And if the university does not elect to retain title, then it has no standing under (f)(2) except to see that title is conveyed at agency request, and the present assignment is totally out of scope and therefore does not operate to the benefit of the university–and something else had better be in place to ensure (f)(2) compliance for research employees to establish the government’s interest in their inventions. By the time one has a present assignment to future inventions that is merely conditional as to assignee–I hereby assign to anyone who has standing to be on the other end of this assignment, which may be nobody or may be myself–then one wonders if this kind of legal gesture has eaten its tail and does nothing really at all, for all its effort to transfer an equitable interest in future inventions.
Thus, in dealing with the issues of Bayh-Dole, it does not appear to matter, really, whether universities use a promise to assign or a present assignment. The scope matters more than the form. And the flexibility of a promise may have advantages over always being the named assignee (one thinks, organizational conflict of interest, meeting the obligations of other sponsors and licensees). In a tornado view, the form doesn’t matter because a federal twister forces compliance by elevating any (f)(2) compliant written agreement to take precedence over any other agreement. So, again, even in the cruel world desired by some university patent administrators, changing to present assignments doesn’t do much.
What we ought to get out of Stanford v. Roche is a clear sense of whether Bayh-Dole represents a town view or a tornado view, whether federal research policy is to trump all private arrangements for invention or to work within and among those arrangements, following conventions already there. We ought to find out if the spirit of Bayh-Dole is that all research innovation should run through university offices, or only that research innovation that university offices have sufficient interest in to manage to their advantage. We ought to find out whether a quarter billion dollars of value rides on a technical detail in the law, like the meaning of the word “retain”, which then triggers a tornado to suck up rights, or whether Bayh-Dole has a generally clear effect of accepting and respecting private agreements, and providing universities with a role only when that role is available, whatever role a given university may choose to have, whether licensing for monopoly advantage or stimulating the formation of new markets with distributed power.