There has been a lot of bad advice for universities out there in the wake of Stanford v Roche. It almost appears to be orchestrated talking points on the need for universities to implement present assignments to prevent another outcome like Stanford v Roche. I don’t agree. Implementing present assignments–especially as a condition of employment–makes matters worse, much worse for universities. Not only does it not address Stanford v Roche, but it also creates new liabilities, disrupts existing policy, and worst of all, obscures the real exposures a university has in licensing.
The problem that arose in Stanford v Roche has to do with incoming IP. That is, nothing matters in the case until a person with obligations outside the university with regard to IP joins a university research project, uses university facilities, and benefits from university-administered funds in a sponsored project. These are the things that typically trigger claims to ownership under a university IP policy–employment, use of resources, and extramural support.
It really does not matter whether the individual is an employee or not, or whether the IP is owned by the individual, or is owned by someone else, or whether it has already been created or not. This is a common situation. New faculty coming to a university from another institution, or from a company. Faculty or students bringing in technology that is owned by others for use in research–perhaps a company, or a sponsor, or by their own startup venture. Or investigators developing improvements in the context of background rights held by others. All of these situations result in new inventions having uncertain status relative to university claims and university commitments to research sponsors and to past licensees or even providers of equipment and research materials (who may have received any number of promises–such as to improvements, notice of improvements, or a right to negotiate).
We can distinguish a variety of possible problems:
- inventions are made that are already committed (promised, assigned) to others;
- inventions are made for which essential background rights are not available and thus cannot be practiced, making licensing potentially worthless;
- inventions for which pending applications will count as prior art unless filed as continuations in part with the owner of the pending applications;
- inventions that are made jointly with visitors, informal collaborators–who may be in the same institution–but who have obligations to others, or who may have no obligation to the university or the university’s business partners;
- inventions made by others collaborating with the lab or hearing of laboratory activities–and who may be in the same institution–but are not under the same obligations as those within the sponsored project.
These situations happen easily when one operates with an open laboratory policy. There are arguments that these sorts of situations are a good thing, a source of creative input that is part and parcel of the discovery, invention, and innovation that takes place in universities.
Companies have tools to prevent this kind of thing from happening–both the good and the bad. They lock down their labs, so only authorized people have access. Employees are required to sign non-disclosure agreements, non-compete covenants, and their research findings are subject to company control–on disclosure, work location, publication, sources of funding, collaborators. Incoming work is scrutinized. Notices about new technology developed by others raise potential liability flags–possible treble damages for willful infringement. Company employees generally get this stuff. It’s not all that difficult, really.
When companies implement an assignment policy, they establish a scope of interest, they create an obligation to assign, they often require a present assignment of future inventions that are within that scope of employment and company business, and they may in addition require power of attorney to execute documents (such as assignments) if the employee is unable or unwilling to do so. They combine this assignment policy with company control of research to create an entire package. You can’t just implement a promise to assign and have much chance to get a quality package of rights if you are operating in an open, uncontrolled environment. Adding a present assignment does nothing to strengthen one’s position or reduce title uncertainty. If anything, title uncertainty is even greater because, without a clear scope of claim and control over the circumstances under which research and inventing is done, without a written assignment directed at a well defined invention, one has no idea what is assigned and what isn’t. It’s just claims in the air, without circumstances, without scope, without definition.
Any of these conditions for invention ownership or control outside of institutional claims or institutional obligations to a sponsor or licensee are likely to arise. It comes with the territory. It is an attribute of open labs, academic freedom, training without trade secrets, wild collaboration, personal control over disclosure and publication, receiving visitors, trading information and materials, and getting out a lot to conferences and other institutions. Trying to pack down the assignment element without dealing with scope and control makes no sense. It doesn’t work. But beyond that (as if that weren’t enough), it leads one down the path of trying to close university labs, by limiting visitors, introducing publication reviews, adding non-disclosure obligations, and making it a conflict of interest or commitment (or both) to disclose or publish before allowing the university to review for possible inventions to claim. That in turn means administrative approval for publication of manuscripts, public presentations, showing visitors around the lab, sharing data and tools, submitting research proposals, and even teaching stuff in a lecture. If a university goes that way, it is hardly a university, or if it is, it is one run by administrators, not the faculty. Certainly what has been lost is the freedom and independence by which faculty and students conceive of their work, follow their creativity, and who they draw in as collaborators. The university, which operates as an open innovation environment, fueled by personal choices and commitments, both formal and informal, becomes another closed shop, with all of the challenges that this entails. Pretty good if one knows what one has to build and is off building it. Not at all good if one is on the hunt for new ideas and finds these in engaging conversations, the work of others, and random events that cross one’s path.
For universities, scope of claim is a fundamental issue. Unlike most companies, which conduct research in relatively limited areas, a university may conduct research in a whole host of areas. It is unlike almost any company. It is not even much like a company research center like PARC or IBM Almaden. It is an order of magnitude more all over the place. For a university attempting a comprehensive claim to all inventions, there’s a very real difficulty in differentiating between those made as a consequence of employment and those that are outside employment. This is a first order issue of scope. Employment means much more than receiving a paycheck. What is that paycheck for? Who controls and directs the work? For whom is the work done? What tools and facilities are provided? On what condition? These are matters of agency and the circumstances of the employment relationship. It is all too easy, and too wrong, to say, the university gets everything an employee does because the university is the employer, and employees work for their employer, as if the university has paid for the inventive work along with the teaching, the committee and other administrative work, and coming into the office from time to time to be collegial.
For these reasons, universities early on in this invention business took a narrow view rather than an expansive one. If they were required to own, they would own. If they commissioned work with the intent to own (whether with employees or independent contractors) then they would own. If someone requested that they own, then after consideration, if they decided to own, they would own. There was no general claim to everything. If anything, there might be a prohibition against using university facilities to operate a business or be involved in commercial activities.
(Here, we may add, that a very big problem was introduced when someone decided that using university facilities to invent was equivalent to using those facilities for commercial purposes, unless the university owned the invention. This shift to claim conflict of interest or misuse of university facilities is at the heart of the darkness that has become university compulsory claims to inventions. How this claim arises, and why it is objectionable, has to wait for another time.)
The narrow view of invention was really quite liberating. The university did not have to worry about every employee assigning rights, or trying to review every invention that was made, or contesting each invention that slipped through their fingers or avoided their claims, or trying to do something which each invention whether they have the resources or capabilities or not. The only inventions they get are the ones that they are required to have, not all those that they require themselves to obtain. It is this traditional relief from having to deal with all inventions, all employees, all relationships, all possibilities that university technology transfer folks are now trying to end. The result is heartburn, for which no amount of antacid will avail.
The problem raised by Stanford v Roche for administrators actually has little to do with the circumstances in Stanford v Roche. There, an employee comes back from an extended stay at a company, having assigned inventions that might arise from that stay to the company, and is allowed to participate in research in just this same area as the stay at the company. The obvious management intervention–if one wanted undivided ownership–is to exclude the employee from the research, or negotiate a deal with the company, or require the employee to renegotiate the assignment arrangement. The administrative problem raised by Stanford v Roche has to do with reviewing external commitments when someone returns to do research at the university where there may be conflicting obligations. Just that in the case of Stanford, and federal funding, there are no conflicts with the sponsor–the federal government does not require the university to own inventions, and does not restrict who may collaborate in the research, and does not even demand that all university employees even report their inventions or agree to assign rights to the government if requested.
The flaw, if it is that (it isn’t or, er, wasn’t), in Stanford’s policy was that it assumed that if the government sponsored the research, then Stanford was required to own. The Supreme Court said, no, if Stanford wants to own, it has to assert that claim independently of the government sponsor. Otherwise, Stanford comes to own when someone else–the inventor, say, chooses Stanford. That is, there is nothing in government policy or funding agreements for grants that requires a university to demand to own as a way of getting ownership. University ownership can be chosen by others, negotiated, arranged, worked out so that it meets everyone’s expectations. Again, this is rather liberating, unless one really objects to working things out, likes power with impunity, and doesn’t want to be bothered by the opinions of creatives who happen to come up with things.
If a university wants to screen out conflicting commitments, it should do so at the time of involvement with a research project in which the university has made commitments. Not as a condition of employment or use of facilities or involvement in research generally or sponsored research in particular, but specific to the commitments that have been made, nothing more. If the employee can’t agree because of prior obligations, then that removes the employee from the project until the conflict between those obligations and the university’s commitments are resolved. If the employee agrees to the research commitments and doesn’t tell the university about the prior ones, well there’s 35 USC 261 to help out, so long as the university handles its invention practice properly and gets assignments before filing. You are not responsible for a prior commitment if you don’t know about it. In fact, at the time of joining a project with commitments, where the scope is usually well defined (if folks have done their job with the statement of work), is a great time to introduce a present assignment to the inventions anticipated to be made in the research. When the investigator chooses the university, that’s a good time to secure that choice, if one wants–or has to have–ownership of certain inventions.
University ownership can be chosen! What a concept! It’s nothing new. It is, in fact, quite old. And it is the condition that held almost everywhere before Bayh-Dole. It is one of the cornerstones, even, of Bayh-Dole. Now, it appears some universities are done with liberating feelings created by review, discussion, and choice. And that means there’s a great opportunity–a competitive opportunity, an advantage of the first order–for universities to go the other way, to offer more choice, more resources, better services directed toward collaboration, creativity, initiative, discovery, and innovation. The schools that do this will be the shining lights of economic impact. The schools that don’t will grow dark, dull, methodical in their self-interest, paranoid about the work-arounds and grumpy about the uncooperative faculty and students.
The present assignment effort is a rear-guard action to try to overcome faculty (and staff, and student) freedom to choose their research, their collaborators, their sponsors, their publications, and the disposition of their inventions. Faculty, in particular, are about as far as one can get from “employee” in the sense of “servants of a master” when it comes to their research. Almost certainly they are not employees, unless commissioned, with specific regard to either patents or copyrights. If Stanford had argued that as to research matters, its employees weren’t its agents, weren’t subject to its control, were not responsible to it to report the deals they got into–then perhaps it could also have argued it did not have notice of the employee’s prior obligation to the company, and its own assignment might have prevailed (had Stanford obtained that assignment prior to filing the patent application and not known about the prior deal, or had Stanford even required a promise to assign specific to the research at hand, at the time the employee with the obligation joined the research team–but alas, none of this happened, in part because it appears Stanford treated its promise to assign as a comprehensive claim on everything that might happen at Stanford, even though any number of things–like those listed above–could happen to frustrate such a comprehensive claim when combined with the other elements of Stanford policy.)
All this aside, university patent administrators found something deeply disturbing in the vicinity of Stanford v Roche. It isn’t about the Stanford fact set so much as the following line of thought: “What if,” worry the administrators, “what if the research had been sponsored by a company? And furthermore, what if we had offered to the sponsor, as we nearly always do, a first right to negotiate an exclusive license? Then, we would be toast. Totally hosed. Pwned.” The problem would be that the university would have offered to license property that it did not have a right to license. This, along with double licensing–licensing the same property exclusively to one party and then again to another–is the great concern of all licensing activity. You license what you’ve got, not what you don’t got. And once you’ve licensed exclusively, you don’t license again–not express, not implied, not a non-assert. And if you’ve licensed already non-exclusively, you can’t offer an exclusive license or even the first right to negotiate for an exclusive license. Gotta cover this stuff. It’s not arcane. It’s not complicated or advanced. It’s the fundamental most basic everyday concern of any intellectual property operation. If you don’t get this bit, you simply have no business in the biz.
Somehow, however, the idea that Bayh-Dole required university ownership put a bunch of folks to sleep, as if in a field of poppies. In their poppy dream, federal law requires assignment to the university and voids any other disposition of inventions. This part of the dream is comforting and pleasant. In their poppy dream, however, there are still flashes of nightmare, that they will license, or contractually offer to license, stuff to which they do not have title. Title uncertainty. They wake with a start, sweating, disturbed. No pill for that. The only remedy is to get it right, and for that, one has to be absolutely clear on the conditions that pertain in university research, university invention, and university interest in ownership.
That’s what everything in university patent-based technology transfer is based on. The effort to claim everything to try to resolve this title uncertainty has to fail. It cannot succeed and leave the university a university. The effort to deal with title uncertainty has to go in another direction altogether, one that works, one that makes university administrators happy again, with smiles and fluffiness and public purpose, stardust and unicorns and glitter. One that puts an edge on the game and deals with title uncertainty by only working the IP that one does have title to, not stuff that one hopes to claim in the future.
What happens in this other direction altogether? The university takes a narrow view of claims for ownership, acts as the mediator of interests, not a claimant, and limits its commitments to others–sponsors, vendors, licensees–to those things that it obtains by this narrow view. It is not the tail wagging the dog–it does not make broad commitments to things it doesn’t have and then insist on a broad compulsory claim to assure everyone that it can honor those broad commitments. The dog must wag its tail. That means, happy dog. In this case, happy university. Stardust, glitter, ponies with horns.
What happens then? Universities refrain from claiming ownership based on use of facilities, or based on participation in sponsored research, or based on inserting a claim of ownership in sponsored research agreements. For federally sponsored research, this makes no difference, because ownership in federal funding agreements is established by the standard patent rights clause, which universities do not get to modify, and which university agree to, and if they agree to it, then they can’t secretly on the side unagree to it and assert it is something different to their employees. For the five or ten percent of research agreements with industry sponsors, a university can disclaim any interest in managing inventions except those that are assigned to the university. The sponsor then can make arrangements for inventions directly with the investigators. That is, the sponsor can commission inventions and investigators can agree that those inventions are deliverables. If the arrangements do not meet university requirements for use of facilities and publication, then the university can refuse to host the work, or require that it move to off campus facilities out of reach of students and visitors. This was the model that the University of Wisconsin used, and arguably Wisconsin is one of the top five practitioners in the area of university invention management. Other universities would do well to pay attention, think it through, and follow this model.
This approach would address the problem of intractable negotiations with industry sponsors over prospective invention rights that the university doesn’t have and has set up with academic freedom to make it difficult for the university to get those rights unless the inventors choose to offer them. Instead, investigators can do the work off campus, outside the grant work, or after the grant work, or collaborate with others not constrained by the grant limitations–any of these things would prevent the university from obtaining title to inventions that it might commit itself to get if it were doing the negotiating.
What does a sponsor then do? It does what the federal government does. It requires the equivalent of the (f)(2) agreement. The university requires its investigators to report inventions, assist in patenting, and assign rights to the sponsor if they do not offer rights to the university or another agent that accepts the same conditions as the university. Thus, the sponsor gives the investigators the opportunity to deal direct, or work through the university to represent their interests, or through another qualified agent who will represent their interests and accept the conditions that the university accepted in entering into the research agreement. It’s a kind of special performance, a participation agreement, a personal commitment by those doing the work, who are the best judge of what they have planned to do, and committed to do, and used the funds to do. If they try to cleverly avoid an obligation to the sponsor, it is their reputation and careers in research that are at stake. And it is the university’s business to help them make good decisions, not to try to make those decisions for them.
If a university disengages from pre-commitment of IP in sponsored research–to make its practice consistent with federal funding–then we can focus on the remaining issue: how badly do university administrators want to make money from licensing inventions? If they very badly want to make money from licensing inventions, then they should consider licensing any inventions that they can get that are within their expertise to license, regardless of their origin. If they don’t care all that much, then they should get out of the business and allow others to manage licensing, at least be designating one or more research foundations or agents as possible partners for those with inventions and needing the services of an agent. There is no point, however, for being caught in the middle, wanting to license inventions to make money, but only these inventions of one’s own employees, even if those inventions are all over the place so one can’t possibly have the right resources and expertise at the right time for many of them. No, that’s neither good avarice nor good service. It’s the impossible middle ground, the half-way commitment that is the failed compromise that wastes time, money, goodwill, and opportunity. If one is going to be in the middle ground, it has to be a voluntary, considered middle ground. For that, the policy should be: if you invent, and you want us to consider handling it, disclose it to us, and we will decide together whether it’s something we can take on, and whether we can work together, and what the arrangements ought to be, given the circumstances. If the arrangements are set by a research agreement that you negotiated and we all agreed to, well there it is.
It is this narrow view to compulsory matters–for commissioned work, for work in which investigators have made commitments that include the university–that opens up the administration of inventions where others want the university’s involvement, look forward to it, are willing to accept the uncertainties and turns of fortune that go with efforts to place new technnology, whether openly, or via a startup, or to existing companies, or to form a standard. None of this is easy, unless you are lucky, and then it is lucky, not easy, so it’s still not easy. But it is a lot better when people choose their partners in innovation matters, and that is the fundamental condition for innovation success in my book.
Any university that forgets that for a moment has just made technology transfer a whole lot more difficult, uncertain, and unlucky than it ought to be. There can still be money made, and “success stories”, and claims about impact, but these are coming at the expense of a great many things–freedom, collaboration, personal initiative and responsibility, goodwill, imagination, the first-call list, and public support. It is these great many things that are at the heart of the university research enterprise. It is for these things that the dog wags its tail. It is for these things that the world is a world of scents and bits of food on the floor and old cushions and wet paws, the hole under the fence, slow moving squirrels and hubcaps going around and around. It is not the collar or the chain or lonely nights in the dog house wishing for the pack, or the moon, or freedom. Oh, to be the happy dog again, and build policy to support the things that really matter to university research.