Science has recently published an analysis of the Stanford v. Roche case. AAAS came in as an amicus on the side of Stanford. The account is useful for what it leaves out and for what it spins. My comments aim to supply some of the left out stuff and counter the spin.
1. Holodniy was “visiting” Cetus for nine months, working with proprietary technology in a company, publishing with its scientists. What did he *think* he was doing, just going for a stroll in a company that didn’t care about its own technology? No one thinks that. Not even “young scientists” (except later, of course, when their share of a quarter of a billion is in play).
2. Merigan, his advisor, was on the technical advisory board of Cetus. He also would know that it was not a fluffy gift from Cetus that Holodniy was allowed to be there learning how to use the PCR technology and was free to walk away and own any darn future thing he wanted relating to his work at Cetus. (And we are told that Stanford also signed an MTA granting royalty-free rights to Cetus in anything that came from the use of the materials–that would be fun to see).
3. Holodniy does his work at Cetus *before* any federal funding. There was no Bayh-Dole involvement when Holodniy signed his present assignment for future inventions arising from his work at Cetus. That is because Bayh-Dole is implemented through federal agreements made with universities, agreement by agreement, patent clause by patent clause. Holodniy was under no obligation to a future patent clause in a federal agreement that Stanford has not accepted. And it is Stanford that undertakes obligations per Bayh-Dole via a federal funding agreement–how those obligations reach to employees is another matter.
4. Stanford approved Holodniy’s consulting, and had a policy at the time that encouraged personal ownership of inventions “whenever possible”. This was just such a time. The issue is not whether Stanford has a claim on Holodniy’s work–it doesn’t–but whether it can find something else to make that claim, like using federal law to force the matter.
5. The research that Merigan’s group then went and did was exactly within the scope of what Holodniy had been doing at Cetus. The court said, there’s just no way it is different or outside the scope. It’s clear the federal funding was brought in to work on what had already been obligated, as far as Holodniy was concerned, to Cetus.
At this point, we take a breath. Stanford when it received federal funding had a choice to make. It could have excluded Holodniy from the research, it could have cut a deal with Cetus, or it could have required Holodniy to clarify or renegotiate his obligations to Cetus with regard to scope of future inventions. It didn’t do any of these things. Now Stanford seeks to make Bayh-Dole void the prior deal Holodniy made, after the fact, by creating the idea that Bayh-Dole vests federally supported inventions outright with the university contractor, against the Constitution, patent law, and practice. Even Stanford went back after the fact for assignments.
6. Holodniy has no obligation under federal law to assign inventions to his employer. He has an obligation (if Stanford properly implemented it) under the standard patent rights clause (f)(2) to establish the government’s rights in the invention. That could mean assigning to the government, granting a non-exclusive license to the government, or assigning to his employer or to any agent that his employer might assign its standard patent rights clause obligations to. Bayh-Dole doesn’t care. If Stanford cannot obtain Holodniy’s assignment (since it permitted that assignment in advance so that Holodniy could spend the time at Cetus), then Stanford cannot elect to retain title in Holodniy’s interest in the invention. Nor can it turn Bayh-Dole around and make the law do what it was supposed to do–that is, void the prior agreement and do what the (f)(2) agreement does, even if the (f)(2) agreement is not in place! Wow.
7. The present assignment vs. promise to assign discussion misses the point. It is not merely a contest between competing claims, or that a present assignment somehow trumps an earlier promise to assign but wouldn’t trump an earlier present assignment. The problem is, Stanford exhausted its promise to assign by permitting the consulting and assignment Holodniy made to Cetus. It would not have mattered if Stanford’s policy statement had been a present assignment. Such a statement would have necessarily had to have been just as conditional as was the promise to assign. It is not the mechanism by which the assignment is made but rather the conditional scope of the claim that matters. The claim would have had to be conditional for federal funding (to comply with the patent rights clauses), and it would have had to be conditional for consulting and other activities that Stanford allowed. It would have been conditional for Holodniy, and would not have operated in competition with whatever Holodniy assigned to Cetus.
It goes further. Even if Cetus had used promise to assign language, its assignment deal would have taken precedence if it had transacted it. Stanford had waived its policy claim to permit the consulting, knew of the obligation to Cetus, and therefore cannot later come back and assert its own policy obligations to claim assignment. If nothing else, it’s a 35 USC 261 matter, as the CAFC decided. Well, er, it can come back later and assert whatever it will, but it’s still a 35 USC 261 matter.
8. So what the universities want is for Bayh-Dole to reverse their own practices, vest inventions directly with employers, and void any deals that the employers have already approved. I can see how an employee-inventor, hoping for a share of royalties from Stanford, would get caught up in the regrets.
9. The “system of commercialization” that the universities are now pitching has next to nothing to do with Bayh-Dole. They are pitching making money from monopoly licensing. Bayh-Dole has little to say about that, and a lot to say about collaboration, practical application, and promoting competition. Bayh-Dole applies to federal agencies. It requires the agencies to use a standard patent rights clause developed by the Department of Commerce. That clause is included in each federal funding agreement with universities. It is by agreeing to this clause that universities accept provisions of the standard patent rights clause. Bayh-Dole is not imposed on them. Bayh-Dole requires nothing of them. The universities choose the conditions of the clause when they accept the money.
The problem is, if the university is choosing to do things under a federal agreement, how comes it that the federal agreement suddenly has the force of law to slice through private contracts made prior to the agreement by individuals who are not party to the agreement? Holodniy agrees to future disposition of patents. Stanford approves. Then Stanford goes off to take federal money. Stanford agrees to comply with the standard patent rights clause. But it cannot. And it knows it, or has reason to know it–its employees know it. Holodniy has obligated his invention rights previously. Taking the money does not void the prior transaction. No. It puts Stanford potentially in breach of its federal funding agreement if Holodniy invents. As it turns out, there are co-inventors that could assign to Stanford and did so, so Stanford could grant the government its license. So there is product on the market helping people, the government has its license, and Stanford has patents it is free to license. The only thing missing is that it can’t license a co-owner, and the only reason it wants to is to get money from that co-owner. It’s about the money. No surprise, there, I guess. If it were about Bayh-Dole, then there would have been an open discussion in the tech transfer community about it, but there wasn’t. It was all politics and behind doors stuff to pack the house with tabors and pitchforks.
10. Nothing becomes uncertain if the CAFC decision is upheld. Inventions are owned by the inventors in the usual way, per the Constitution, and employers and others obtain title in these inventions in the usual way, by a written instrument, and Bayh-Dole operates as it has for 30 years, telling agencies what patent rights clause to use and setting bounds on how agencies can vary from the standard and what they can require by way of reporting, approvals, and assignments from inventors supported by federal funds.
All the big changes come if somehow a federal funding agreement can be used to void prior assignments. Stanford argues that Bayh-Dole is a vesting statute, that contrary to any other classes of inventors, federally supported ones never own their invention rights. Even the Solicitor General did not go so far, arguing instead that the inventors hold title, just cannot do anything with it until Bayh-Dole is satisfied–which might be the case if they have agreed in writing to the (f)(2) agreement, which is a university obligation to require under the standard patent rights clause.
As a vesting statute, Bayh-Dole as it stands would be unconstitutional. It lacks due process and just compensation for taking private property for government purposes, even if that purpose is handing personal patent rights to university administrations for “commercialization”–i.e., like suing companies for royalties. In the case of Holodniy, that private property is his invention rights, which he assigned to Cetus. It was a bowl of lentils, and he may regret it now, but the fact he didn’t take it seriously then isn’t the subject matter for the Supreme Court. The CAFC looked at it and said, it’s Cetus’s title. That’s the only valid assignment at the time of the invention. The universities are arguing that somehow Bayh-Dole listens to regrets, voids the personal assignment, and passes title directly to Stanford, no assignment needed.
What’s the point of the standard patent rights clause? Ah, it’s just for show. What’s the point of the (f)(2) agreement. Needless ceremony. What’s the point of the 35 USC 202 (d) provision allowing agencies to let inventors retain title to inventions? Badly written, obviously, and replicated as 37 CFR 401.9, since the agencies could not allow inventors to retain title to inventions if that title transfers outright to the university. It would be the university having to assign title to the inventors. But no. Who cares? Just make up any old explanation so long as you get the money.
If folks want to wonder about things, here are two.
11. When Holodniy invents (let’s say, it really is a patentable invention), he has a known obligation to Cetus and is working under a federal agreement with a standard patent rights clause. The federal agreement reaches to Stanford. But it is only by means of (f)(2) agreements that Stanford imposes an obligation to assign on Holodniy. If there’s no (f)(2) agreement, because Stanford did not obtain it, then what’s the basis on which Stanford should have any interest in Holodniy’s title?
12. If the whole system of “commercialization” is turned on its head if Holodniy owns his inventions until he assigns them in writing, then what do you make of the limits of the standard patent rights clause, which calls out only (f)(2) agreements with contractor employees, and not then even with clerical and non-technical employees (who also may invent), and also not with independent contractors, volunteers, visiting scholars, industry collaborators? Any of these might co-invent as well, and yet the standard patent rights clause doesn’t care. No big problem. It might dawn on folks that Bayh-Dole is not about helping universities make money from licensing, but rather is about making uniform agency requirements in agreements with universities. One key element of that uniformity is a limitation on what the federal agencies can be interested in by way of invention rights–and that is only those rights that are available from research employees. The clause cannot reach to others, and the law does not reach to others either.
The standard patent rights clause does not say: “if you are an employee of a university, you have no invention rights”. Rather, it says, “a university is permitted to step in and take assignment of patent rights from employees who otherwise have agreed to assign to the federal agency on request, provided it does those things that are conditioned upon such assignment.” That is, the patent rights clause says, you may do this if you can comply. It does not say, when you take title, you may invoke the law to force everyone else in the world to support your compliance, even if you are not able to comply, by voiding their personal interests and private transactions.
At the root of this dispute is a love of money and a story retold so often it appears as truth. The love of money is understandable. The retold story is the deeper problem, for it shows that university administrators do not respect private property rights. Somehow they believe Bayh-Dole gives them the invention rights of others, rather merely the standing to negotiate for those rights, if those rights are available. These non-profit institutions are arguing for the federal government to take private property so they can make money. It’s the taking that gets me.
As Hayek has it, private property is at the heart of a free, democratic society. Love of money is at the heart of something else.