Here’s a typical slide deck (it opens in PowerPoint–sorry non-‘Softies) [now deleted–but here is a similar slide deck, posted at the University of Tennessee at about the same time, by Lakita Cavin, a staff attorney, and displaying similar problems] talking about Stanford v Roche and offering lessons learned. It looks like it was prepared by the University of Rochester’s Office of General Counsel. It does a nice job on the chronology, ignores Stanford’s policies and practices, and constructs a strange hypothetical that leads to a wonderful set of “lessons” that may be useful to keep in mind but have next to nothing to do with Stanford’s practice. I would go so far as to say it is talks like this that do more to obscure the lessons that should be learned than they help anyone understand practice. And this is a thoroughly typical, serviceable talk.
Here are some things the talk leaves out. Maybe the first lesson is to avoid abstracting a situation to a set of facts that mislead folks to a simplistic lesson about being careful about signing documents when they should be taking a hard look at their attitudes about collaboration, money, and innovation.
1) Stanford’s policy allowed inventors to own “whenever possible.”
The condition of the promise to assign to Stanford was not merely that an employee uses “Stanford resources”. The policy was more liberal than that, and good for Stanford.
2) Stanford personnel approved of the VCA deal–it was reasonable under the circumstances.
I would expect Cetus would never have allowed the post-doc to spend nine months learning PCR and applications at Cetus if he had not signed the agreement that assured Cetus that inventions would be theirs. The post-doc’s supervisor at Stanford was on Cetus’s technical advisory committee. No one needed to be in the dark about the conditions of the collaboration. Stanford at this point essentially waived its promise to assign within the scope of the VCA. If the post-doc had come back to Stanford to visit during his time at Cetus and invented something while messing around in a Stanford lab, and the invention was within scope of the VCA, then it would have been Cetus’s. Stanford had given up its standing to re-assert a promise to assign at that point. It didn’t need to waive anything, anyway, because it has that “whenever possible” condition in its policy. And that “whenever possible” condition was a really smart thing to have there. What it meant was that the issue of compulsory assignment only comes up if the post-doc joins a project that has an IP condition requiring Stanford’s ownership. And what we learn is that Bayh-Dole-based funding agreements are not such agreements.
3) Stanford signed MTAs granting Cetus non-exclusive licenses.
The MTA relationship continued following the post-doc’s return to Stanford. Even though Cetus was not funding the research, Cetus (and later Roche) had the VCA and MTA agreements that it relied upon to protect its interest in the collaboration. The MTA agreement never got much attention in the case, but it was there, waiting its turn. One might recognize that collaborations will create multiple relationships and that it may be easier to manage these relationships if fewer of them are contractual than if all of them are.
4) Stanford failed to screen post-doc on return to an IP controlled project.
Stanford knew of the VCA arrangement and yet accepted funding that was directly in line with its terms, and permitted the post-doc to participate in that research, even though there was a conflict in terms. Why on earth would a university do that? Was it miscommunication between administrative units? Was it indifference? Was it a fundamental failure to understand the scope of Stanford’s promise to assign.
5) Stanford failed to secure an (f)(2) agreement as required by federal funding agreement.
The Standard Patent Rights Clause in the federal funding agreement requires the university to require a written agreement of each employee (except clerical and nontechnical employees) to protect the government’s interests–disclose, sign paperwork to allow patent applications to be filed, and sign paperwork to establish the government’s interest in the invention. That would be a federal agreement. It would force the employee and the university to deal with the conflicting obligations. If the university does not want the prospect of Cetus obtaining title to inventions made within scope of the VCA that may also be within scope of the federal funding agreement, then it must exclude the employee from the research, or renegotiate the deal with Cetus, or reach a business agreement with Cetus, leaving the VCA in place. Stanford does none of these things.
6) Stanford filed patent applications before seeking assignment.
If you want to set up a collision between competing assignments, you need not to know about the prior assignment and record yours in the PTO first. If you know about the prior assignment–as an organization–then you are up against 35 USC 261, since you know. Furthermore, I imagine there are such things such as tortious interference in a business agreement if you try to get an assignment when you know an invention is obligated to another party. In any event, if a court was going to sort things out, Stanford could have helped its cause by getting the assignment first, then doing the application. At least at that point, the assignment would have been made, arguably, before the invention was conceived and reduced to practice, and the courts could have worked through which assignment controlled at the time the invention became and invention to which title attached. But no. When the courts looked at things at the time Stanford filed, there was only one assignment in place, and it was to Cetus via the VCA. The fact that the VCA included present assignment language was the result of following best practice. That the present assignment operated without additional formalities is precisely why one would use it in a situation involving a person whose later actions one does not anticipate being able to control as one would with one’s own employees. Lesson learned. The later assignment attempted by Stanford didn’t operate because there were no rights remaining to assign.
7) Stanford never showed that the invention was made with federal funds.
Roche protested numerous times that Stanford never even produced the federal funding agreement or statement of work under which the invention was claimed to be made. A subject invention arises when it meets the conditions of the Standard Patent Rights Clause, not by assertion by either the inventor or the contractor, and not by agreement between the two. What matters is whether the invention comes within the planned and committed activities of the grant, or distracted or diminished that work.
The slide deck’s “Painful Lesson for Stanford” is shocking in its inability to reason from the appropriate circumstances and practices. Yes, 11 years of litigation–that wasn’t necessary, but Stanford pursued appeals that didn’t help its case–but did, in an odd way, help everyone finally get a grip on Bayh-Dole, even if some tech transfer folks are still in denial.
It is absolutely not the case that Stanford lost millions it should have had. If Stanford had objected to the VCA deal in the first place, how can anyone assume the research leading to the inventions would even have taken place. One can build any scenario one wants, I suppose, but at least make it plausible.
The talk then announces the lessons for the rest of us. “Lesson 1” is that the university has to review and sign private consulting agreements. How is this a good lesson? How is it even possible? Perhaps the lesson is, don’t let employees consult. Force all relationships into a service agreement with the university. The talk fundamentally misunderstands the relationship. It was *outside* the university’s scope of employment. The university has no business reviewing or signing the agreement. The university’s interest starts when the employee comes back and seeks to join a project with IP-controlled terms.
“Lesson 2.” Consulting agreements are often too broad. Perhaps–but how does that arise from this set of circumstances? The agreement set a scope that was specific to the work, and I expect many would say was perfectly reasonable. What folks might question is why Cetus was letting anyone in for that kind of access anyway. Perhaps it has to do with who was on the advisory board, or the fact that Cetus was started by academics who would know how to take precautions.
Clearly, here, the money was not the issue–gaining access to PCR technology and research smarts appears to be the goal. That sounds like pretty decent goals for folks aiming to discover and publish. Is it really about the money? Wouldn’t inventing something that helps save lives be a decent end in itself? Wouldn’t it be even more decent if Stanford folk had got their product on the market first, or ever, rather than burning for a share of someone else’s effort?
It would appear here that the Office of General Counsel has ignored the fact pattern, selected out a few bits, and made a big deal that Stanford “lost millions” and that faculty should think twice about signing anything in case they might lose millions, too. Is it worth learning something cutting edge, if you can’t figure out a way to get a troll position on your collaborator later and leverage out a big wedge of the pie for yourself, or in the public interest, for your university?
No, this is all wrong. It shows just how little most university counsel understand about Bayh-Dole or research or academic freedom.
What then are the lessons that should be learned?
1. Don’t envy success. Celebrate it. The story is a success story except for Stanford’s envy. University and company collaborate. An assignment deal makes it all possible. University gets grants. Company gets product on market. University gets patents to go along with it. People are helped. Company is profitable. There will be a future for this relationship, but for the money claim.
2. Comply with the Standard Patent Rights Clause. Our friend at 37 CFR 401.14(a)(f)(2) [now just 401.14(f)(2)], the agreement required of employees to protect the government’s interest. Don’t try to work around this with policy statements and your own written agreements, to try to protect your institutional interest. Protect the government’s interest, the way you agreed to. It is a simple agreement for each funding agreement, that can be signed off when each person goes through HR to join a federally funded project. No big deal. Do it.
3. Check folks coming back from consulting when the join any IP-controlled project. You don’t need to know what their deals have been, or review them in advance, or load yourself up with expansive claims to scope to create uncertainty in an effort to keep research personnel in open labs from mixing with others. That’s their strength. It is unique in America, to have such independence with institutional resources. If you know about their deals, then you have a 35 USC 261 problem. If you don’t know, then your business is to get to the PTO first. With first to file, this will be the thing that matters. If folks sign a conflicting deal with you, that’s a problem for the other guy–and perhaps for the folks not being up front with you. You will never resolve that sort of uncertainty with paper. It comes with the territory–called humanity.
4. Collaboration with industry can spark great ideas. Get your people out there and mix it up. Sure they will sign rights away, but your office is not going to license most stuff anyway, and they are actually net ahead taking consulting money over what you will offer them. Networked non-market innovation happens daily, and is refreshing and inspiring. To be a little part of something big can be a lot more fulfilling than holding out for the big part that never comes. When a director offers you a walk-on part in a big interesting movie, take it. Talk to your agent when you’ve done that a few times, and see whether you can get a bigger part then next time. But if you never will say yes, you have little chance of starring later.
5. Understand how Bayh-Dole works. It does not vest title, mandate title, or reserve title for universities (or other contractors). Read things. “Elect to retain title” does not mean “elect title.” An invention “of the contractor” is not an invention “by the contractor.” A written agreement required of each employee to protect the government’s interest is not a policy statement or present assignment required of each employee to protect the contractor’s interest.
Bayh-Dole is an amazingly versatile law. It applies to federal agencies. It instructs them how they are to contract for invention rights in federal funding agreements. It establishes standard patent rights clauses to be developed by the Department of Commerce and how they are to be used and how they may be modified. The Standard Patent Rights Clause at 37 CFR 401.14 establishes a relationship between contractor employees and the government, treating them as proto-small business contractors should they invent and retain title to their invention (see 37 CFR 401.9).
As for the disposition of inventions, Bayh-Dole establishes a menu of possible agents. The inventors are free to choose between an agent chosen in conjunction with the government agency (on the one hand) or an agent chosen in conjunction with the university contractor (on the other). If the inventor does not get an agent through the university, then the agent discussion is with the government.
This agent menu is not discussed anywhere university counsel might look for a quick guide to Bayh-Dole. You have it here though. Work through the law. You will see how it works. The Standard Patent Rights Clause says, inventors own their inventions, and they have a choice. They can contract that choice away, or they can hang on to it. It all depends on how their employer takes it all, and on what the federal agency wants to do.
What some universities want to do is prevent an inventor from having a choice. Among major universities, only Wisconsin, please their hearts, still allows this choice between the government and the university (or WARF, actually). Stanford, despite the painful lesson, still allows inventors to dedicate their inventions to the public domain. That’s a choice, too, and it’s honorable that it’s still in their policy. It even works under Bayh-Dole, in an odd way. The universities want to prevent an inventor from offering rights to the government (which could happen under the (f)(2) agreement), and do not want the inventor using any agent except as chosen by the university. For most, that means the university itself. The lesson these universities “learn” from Stanford v Roche is to make that demand to limit choice even greater. Cut off even consulting, if necessary, to make it clear to all that inventors should have no choice, ever, if at all possible. To accomplish such a thing, however, these universities also have to breach the Standard Patent Rights Clause and its (f)(2) agreement requirement, which would drill a hole through all their agreements.
Thus, what we learn about Bayh-Dole from Stanford v Roche is that it provides inventors with a choice of invention agent. If they go to the university, it is whatever agent the university is willing to approve that has as a primary function managing inventions. It could be most anyone. And if the agent doesn’t have as a primary function managing inventions, the agency also has to approve the deal. If the inventor goes to the government, the government can be the agent, or allow the inventors to be their own agent, as if they were a small business contractor. As the agent, the government can either file patents or allow an invention to go to the public domain.
If the inventors choose the university, then it is a matter of working out the arrangements for management, and whether the university is the right organization to do the work. Bayh-Dole doesn’t require assignment to the university or any other private invention management agent. It only says that if a qualified agent obtains title (“of the contractor”) then it can retain that title, provided it behaves. Bayh-Dole also doesn’t prevent contractors from requiring their employees to assign title to inventions. That also works, though it flattens the Act down to its most dull form. Why even have Bayh-Dole with its options if university tech transfer offices are going to devote their whole effort to preventing it from operating? What an awful auto-immune disorder, that the tech transfer cells attack the Bayh-Dole joints and prevent them from moving! And all out of institutional self-interest, rather than public interest, or out of respect for the inventors, or out of deference to the way that Bayh-Dole has anticipated choice of agents. Instead, it is now all about institutional self-interest. In that case, if the universities really do, to the last dean and provost, gut Bayh-Dole, then we ought to get rid of it, and try something new that shakes things up.
The public research enterprise is not about tying everything up for patent licensing, no matter the aspirations and training of those involved. If someone wants to debate the point, come on, let’s have at it. This isn’t about sincere rationalizations for self-interest. The public research enterprise is about finding what we don’t expect, creating what we don’t realize, challenging what we think true, and teaching others what we know, to their advantage. What comes back to us takes many forms, even money, but the most important form is the opportunity to be engaged building new things in the community and confirming the value of those things that have withstood the test of inquiry. Mighty words? Yes, in a way. But it is worth reasoning from them, rather than setting them aside, and by doing so compromising the public trust. That’s the lesson I wish university counsel would learn, and would speak up about it.