Now the Nolo page turns to Stanford v Roche. Given how Nolo can’t seem to get much at all right about Bayh-Dole, what do you think the odds are with Stanford v Roche?
Stanford v. Roche (2011): The Supreme Court Weighs in on Bayh-Dole
A 2011 Supreme Court case illuminated the power of the Bayh-Dole Act and cleared up some issues regarding assignments: specifically, the difference between a promise to assign a patent, and an actual assignment. (Board of Trustees of the Leland Stanford University v. Roche Molecular).
Totally inaccurate, other than the link to the case.
The Supreme Court did not discuss promise to assign vs present assignment. That was decided by the CAFC. The Supreme Court held that Bayh-Dole does not automatically vest title to federally funded inventions in federal contractors or authorizes contractors to unilaterally take title to such inventions. Justice Breyer’s dissent complained that the Court did not address the CAFC’s distinction between a promise to assign and a present assignment of a future invention. Breyer argued there should not be so great a distinction. Breyer’s point is a good one, but it runs to just the opposite of what is put here. If one follows Breyer’s opinion, it does not matter whether one promises to assign or makes a present assignment of an invention that hasn’t been made yet. Breyer’s argument is that the wording should not make a difference. It is the CAFC’s distinction that causes people to change the wording of their patent agreements.
The case did not “illuminate the power” of Bayh-Dole. Instead, it made clear that Bayh-Dole concerns only the relationship between a federal contractor and a federal agency after the contractor has acquired ownership of an invention made with federal support. Bayh-Dole does not vest ownership of such inventions with the contractor nor authorize the contractor to unilaterally take ownership. If anything, the decision reduced the powers claimed for Bayh-Dole.
The Supreme Court reviewed two agreements signed by a university scientist: one in which he promised to assign rights to the university; and another in which he actually assigned rights to a company that was later purchased by Roche.
No. The Supreme Court did not review these agreements. The majority opinion does not mention present assignments. Breyer’s dissent does, because he wanted the Court to review the agreements and they chose not to. The CAFC reviewed the agreements. The promise to assign agreement concerned only inventions that Stanford was legally required to own. Further, Stanford approved of the later agreement containing a present assignment of a future invention, thus Stanford effectively waived any obligation that might have been created by the promise to assign agreement. And for all that, Stanford failed to comply with Bayh-Dole reporting requirements, amd filed a patent application before obtaining assignments.
The CAFC did not consider all of these elements, but rather ruled that the present assignment took precedence over a later assignment made by the inventor to Stanford. The earlier promise to assign did not operate. People assume that it did not operate because it lacked the magical language of a present assignment, but that assumption is wrong. The CAFC needed only to find that the promise to assign did not operate. The promise to operate would have operated if Stanford had been legally obligated to obtain ownership. Stanford’s argument was that Bayh-Dole created that legal obligation (in some way, perhaps with the use of magic). But even if Bayh-Dole did that, the CAFC ruled, it could not reach back and revive a promise to assign that had long since been superseded by an assignment made to the company (which happened to be in the form of a present assignment). Thus, the CAFC ruled that Bayh-Dole did not vest ownership outright in Stanford or create an obligation in Stanford to obtain title. That’s what Stanford appealed.
If Stanford believed that the promise to assign operated to assign the invention, then it did not need the later assignment. So why did it bother later to obtain such an assignment? The CAFC reasoned that the promise to assign—even if it applied (which it didn’t)—did not preclude the later present assignment to the company. The CAFC found that when the inventor attempted to assign to Stanford, he had no rights remaining to assign, having already assigned them to the company, with Stanford’s approval.
Stanford argued that the promise to assign was in fact an assignment. But promise to assign was conditional and only became operative if Bayh-Dole somehow required Stanford to own the invention. Stanford appealed to the Supreme Court that Bayh-Dole vested ownership of the invention with Stanford, voiding any prior assignment. The Supreme Court ruled against Stanford and in doing so let stand the CAFC analysis of assignments, chiding Stanford that if it wanted to acquire ownership, it might consider using a conventional assignment and not rely on a claim based on a postulated subtle change in federal law.
The Supreme Court ruled that the actual assignment agreement took precedence over the promise-to-assign agreement.
No, the Supreme Court just did not reverse the CAFC’s reasoning about precedence. If Bayh-Dole did not require Stanford to take ownership, then the promise to assign agreement did not operate, and the CAFC’s ruling that the prior present assignment took precedence over the later formal assignment to Stanford was correct. Technically, what the Court found was that the present assignment prevented Stanford from having standing to sue Roche for infringement, not that the present assignment indeed vested the inventor’s ownership of the invention in Roche. It is possible that the Supreme Court would have found the same outcome had the present assignment been worded as a promise to assign. That promise would have had the same effect—consistent with Breyer’s point in dissent.
The ruling won’t invalidate past assignments. But problems may occur in cases like this one, where two assignments appear to conflict. In addition, universities (and employers) should stop using language in which inventors promise to assign and instead use language that automatically assigns inventions.
This is just bad advice. The matter of conflicting assignments is addressed by federal patent law at 35 USC 261. Changes in magical wording should have little to do with it, as Breyer’s dissent makes clear.
In addition, universities (and employers) should stop using language in which inventors promise to assign and instead use language that automatically assigns inventions.
More bad advice. The practice issue is not the form of agreement that gets used, but the management of the scope of claims in inventions and the possibilities for conflicting assignment obligations. If a university extends unreasonably its scope of claim, it will create an environment in which conflicting claims are likely to arise. Universities in this regard are Pigpens, bringing with them their own cloud of administrative problems, which they confuse for complexity inherent in the activity. Patent management is complex, but not as complex as university mismanagement and rotten policy make it.
The practice warning from the Supreme Court is that universities should obtain ownership of inventions by means of negotiated assignments rather than claiming that federal law gives them title or the right to take title. I get why university attorneys don’t understand this–they don’t want to. They were all in on the Stanford side of the case, many joining briefs supporting Stanford as friends of the court. But Nolo has made its market by writing books for regular people–you know, like Patent It Yourself–not Patent It By Surrendering It to University Lawyers. Why Nolo doesn’t bother to get it right is just plain sad.
There is a bigger problem with this whole section. The first is that the point of view shifts from university requirements to inventor issues. If the focus is on what inventors need to know, then what’s listed is pretty much irrelevant.
Here’s the deal from the point of view of anyone involved in a federally funded project at a university, using the Nolo outline as a guide. There are other things to say, if we started differently, but this is a good start.
When a university receives funding under a federal funding agreement, a law called the Bayh-Dole Act (35 USC 200-212) requires the federal agency to use a default patent rights clause in the funding agreement unless the agency can justify an exception.
When a university accepts research funding from the federal government on behalf of a faculty investigator, it agrees to comply with the patent rights clause in the funding agreement.
The standard patent rights clause in these funding agreements requires the university to require potential inventors in a federally supported project to make a written agreement to protect the government’s interest. Under that agreement, if you invent, you promise to
- promptly disclose the invention to the university
- sign papers to allow patent applications to be filed
- sign papers to establish the government’s rights.
By making the written agreement you become a party to the funding agreement and, for the limited actions listed above, a contractor, and your invention becomes a “subject invention.” Unfortunately, most universities refuse to comply with the written agreement requirement and instead insist that you agree to assign any inventions to the university.
Depending on university policy or your agreement with the university (if there is one), the university may require you to assign the invention to the university or to an invention management agent. Any such requirement, however, has nothing to do with the Bayh-Dole Act or the patent rights clause in the federal funding agreement. It is strictly a matter of your university’s policies and your agreements with your university. Of course, a university may simply request that you assign, and you may voluntarily do so, without any compulsion brought by assertion of policy or contract.
The university must report to the funding agency inventions made with federal support within two months of receiving your invention disclosure.
If you assign your invention to the university, then the university must notify the federal government within two years of its reporting the invention whether it will retain title to the invention.
If the university retains title, then it must grant to the federal government a nonexclusive license and file a patent application within one year of notifying the government of its decision to retain title and
- allow the government to own rights in any country in which it does not seek patents;
- include a statement of government rights in each patent application;
- make an effort to attract small business licensees;
- report on utilization of the invention;
- require U.S. manufacture for an exclusive license to use or sell the invention in the U.S.
- allow the government to require it to grant licenses if it (or its exclusive licensee or assignee) fails to require U.S. manufacturing, fails to use the invention, or if its use fails to meet government requirements for health, safety, or public use
- share royalties it receives from licensing with inventors
If you don’t assign your invention to the university, then you may negotiate with the federal agency, which may allow you to keep your invention, if you have disclosed the invention to the government and have elected to keep title, subject to the minimal requirements that you:
- allow the government to own rights in any foreign countries in which you do not seek patents;
- include a statement of government rights in each patent application;
- report on utilization of the invention;
- require U.S. manufacture for any exclusive license to use or sell the invention in the U.S.
- allow the government to require you to grant licenses if you fail to use the invention, fail to require U.S. manufacturing, or if your use fails to meet government requirements for health, safety, or public use.
Unfortunately, most universities do everything in their power to prevent you from owning your inventions made with federal support. Further, most universities don’t license most of their inventions. And even those that they license often don’t make much more than the cost to obtain patents. Your chance of getting any royalties at all on any given invention is about 1 in 200. Your odds of getting rich from royalties is less than 1 in 1,000 at the top research universities.