Hah–thought I was out of Misery articles? No, not at all. We have a few more!
The distinction between conception and reduction to practice becomes important in university patent policies such as the University of Missouri’s because university administrators attempt (badly, or self-servingly, or both) to accommodate the Bayh-Dole Act as a matter of policy. Nothing in Bayh-Dole requires a change in any university patent policy. The actual requirements for compliance are to be found in the patent rights clauses placed by federal agencies in federal funding agreement. Those requirements amount to educating personnel on the importance of prompt reporting of inventions, designating personnel for patent matters, reporting inventions to the government that get reported to the university, and requiring research personnel to make a written agreement to protect the government’s interest in inventions made under the funding agreement.
Given that these requirements are a matter of federal contract, there’s not much reason to make them also requirements of university patent policy. The university, by accepting the federal funding agreement, has agreed to the requirements. The policy that’s required is one that states “the university will comply with all requirements in contracts to which it is a party.” The flow down of this requirement to individuals takes the form: “individuals will comply personally with any provisions involving inventions made in the performance of such contracts. But that’s what the written agreement requirement does in the standard patent rights clause authorized by Bayh-Dole. So, again, the university does not need new patent policy language–it just needs to comply with the federal contract that it has agreed to. All this simplicity has overwhelmed university patent administrators, however, and they have rushed to deal with the complexity in their minds by introducing a matching complexity into their patent policies, ruining the policies but satisfying administrative emotional needs for complexity beyond reckoning.
Bayh-Dole introduces a new category of patentable invention into federal patent law, the “subject” invention. Previously, “subject invention” had been a creature confined to federal research contracting, and then to executive branch patent policy under presidents Kennedy and Nixon. But Bayh-Dole is part of patent law, and thus so now is subject invention. It’s a significant change, really. A subject invention is no longer a definition that merely scopes a government license or even an invention deliverable in a federal contract–it’s a new category of patentable invention in patent law. That new category carries the new policy requirements set forth for it in patent law at 35 USC 200. These policy requirements apply to any patent that issues on a subject invention. All of this apparatus is very, very different from that of typical federal contracting with a statement of scope for deliverables–whether inventions or licenses to inventions.
Here’s the definition of subject invention in federal patent law (35 USC 201(e)), with the “conceived or first actually reduced to practice” wording highlighted:
The term “subject invention” means any invention of the contractor conceived or first actually reduced to practice in the performance of work under a funding agreement: Provided, That in the case of a variety of plant, the date of determination (as defined in section 41(d)  of the Plant Variety Protection Act (7 U.S.C. 2401(d))) must also occur during the period of contract performance.
This definition is similar to the definitions that had been used in the past–first in military contracts, and then in the Kennedy patent policy. But in these past uses, the “conceived or first actually reduced to practice” concerned the scope of invention deliverables to the federal government, regardless of who owned rights to the invention–inventor or host university or assignee from inventor or host university. The definition in Bayh-Dole differs from the Kennedy definition of “made” (and the IPA program definition of “subject invention” by adding the restriction “of the contractor.” That means that the elements of the definition are not met unless the invention is owned by a contractor. Nothing happens in the law or its implementing regulations or the standard patent rights clauses it authorizes until an invention made with federal support is owned by a contractor. This is a huge change in definitional practice. While the Supreme Court was spot on in finding that Congress could not have intended a sea-change in patent law by hiding a vesting requirement in ambiguous terms with an interpretation that ran against long-established patent law, there is no reason whatsoever to require Congress to worry the nuances of past executive branch patent practice in establishing a law to change that practice.
If Bayh-Dole were left alone, just as it is without the added complications of requirements put into the implementing regulations and standard patent rights clauses that were not authorized by Bayh-Dole, then Bayh-Dole would clearly set out a voluntary invention management scheme. If inventors wanted some benefit they perceived arising under Bayh-Dole, they would assign their federally supported inventions to the organization that hosted their research or to an invention management firm, or with federal agency approval to most anyone. If they did not assign an invention in this way, then Bayh-Dole simply doesn’t apply. Given Bayh-Dole’s preemption clause, it’s not at all clear whether federal agencies have any right to claim ownership or even a license in inventions made with federal support that do not become subject inventions. Bayh-Dole stipulates what must be in federal funding agreements by way of patent rights clauses. Bayh-Dole does not stipulate a federal right to own or receive a license clause for inventions that are not subject inventions. Bayh-Dole excludes all other statutory authority to dictate invention terms in federal contracts. There’s nothing in federal contracting, then, post-Bayh-Dole that gives a federal agency a right to demand ownership of any invention that’s not a subject invention.
The standard patent rights clause authorized by Bayh-Dole attempts to deal with this problem by requiring a clause that’s not authorized by Bayh-Dole. This is the (f)(2) written agreement requirement. Were the (f)(2) requirement complied with, then every potential inventor for which the federal government might make a claim would be a party to the funding agreement for the limited purpose of reporting inventions, signing papers to permit patent applications to be filed, and signing papers to establish the federal government’s rights in subject inventions. If these potential inventors are parties to the funding agreement, then they are, within the definitions of Bayh-Dole, also contractors, and since under federal patent law inventors own their inventions, then all their inventions made with federal support become subject inventions, and Bayh-Dole then operates on these inventions, with the caveat that the “contractor” in the standard patent rights clause refers to the inventor, not to the organization hosting the research (which is of course also a contractor), unless and until the inventor assigns rights to the host organization.
It’s a clever scheme. Too bad it doesn’t operate because university administrators refuse to implement the (f)(2) requirement as written. Instead they substitute a patent agreement that runs outside of the federal funding agreement, often as a condition of employment rather than as a delegated condition of a specific federal agreement, and thus they do not make potential inventors into limited contractors and their inventions are not subject inventions until the university establishes ownership of them. Thus, the concerted (and ugly) effort by administrators to find some way to make all inventor inventions vest with the university, without regard for scope of employment, assurances of freedom of research and publication, and the conditions of federal funding which provides compensation to universities for the release of their research personnel from regular duties to work “outside the walls” on extramural research contracts with the federal government.
When placed in patent law, however, the definition of subject invention takes on two new roles it didn’t have as a creature of executive branch contracting. First, it separates from ordinary patentable inventions a new class of inventions that meet the now statutory definition, and for this new class of inventions patent law now applies a new statement of policy regarding patent property rights. Second, it makes inventions in this new class appear to be deliverables to the organization that hosts the federally supported research. The idea that Bayh-Dole somehow vested or entitled or gave a right or gave a mandate to universities to acquire title to subject inventions is at the heart of the Stanford v Roche case, where the Supreme Court ruled that this idea was absolutely without foundation. The Supreme Court noted the ambiguity of the wording–not only the scope of “conceived or first actually reduced to practice” in this new context but also “retained” and “of the contractor.”
What the Supreme Court did not bother to point out was that this (purposefully?) ambiguous drafting created a situation in which thousands of university inventors were denied their federally assured right of invention ownership by university administrators and patent attorneys all hot and lustful to turn Bayh-Dole into a vesting statute. And the Supreme Court had no need to go still further and point out how truly awful such a comprehensive vesting statute would be for the public purpose of supporting projects committed to making new discoveries and inventions broadly available for use–not merely as monopolies for exploitation and speculation by the moneyed favorites of patent brokers.
We are dealing with sleight of hand in Bayh-Dole. If you want to follow it, you will have to look carefully because by its nature it is designed to fool the unobservant and also those that enjoy being fooled. Let me try it with scribble diagrams.
In the pre-Kennedy policy regime, the government contracted for invention deliverables. It defined its interest in inventions broadly, using the conceived or first actually reduced to practice language–and with the expansion from “in the course of or under” the contract to include work done in anticipation of the contract. If the government claimed such invention rights by contract, then there was no point in a university demanding assignment of those rights from its inventors only to assign those rights over to the government. The flow of control went directly from inventors to the government. If the inventors did happen to assign their inventions to the host university (rarely) or to an invention management agent (more frequently) or to their own startup (more frequent that you might think), then the government claim followed that assignment.
R -> G
The government G provides money to university U to support a project proposed by research R. If R invents, then R assigns the invention to G, not to U. If G assigned to U, then U in turn would have to assign to G. Or, alternatively, R has an obligation to grant G a license to any invention R makes. If R assigns to U, or to an invention management agent, or to R’s own startup company, or to any other company, R still have an obligation to grant G a license, and that obligation goes with any assignment of the entire right, title, and interest to the invention. If R does not assign the entire right, title, and interest, of course, then R still has the right (and obligation) to grant to G its license.
Where the federal claim was to a government license to inventions, the claim’s purpose was to defeat any claim for “reasonable and entire” compensation under 28 USC 1498 for any government practice of the invention.
Now consider the IPA program variation.
U A -> G
R -> G
In the IPA scheme, the federal government provides funds to university U to support researcher R. U is required to make R agree to assign to U any invention R makes that U wants to manage. U then may require assignment of an invention, or require R to assign the invention to a patent management agent A. If U requires assignment, then U or A grants the government G a license. If U decides not to manage a given invention, the U does not require assignment and R assigns to G or licenses to G just as R would have without the IPA.
If we wanted to put this all in simple words, it would go like this. The federal government supports the work of researchers who propose projects that will expand scientific knowledge and address public health and safety issues. The government expects that for this purpose the patent system generally gets in the way. Generally, inventions like everything else should be made broadly available and not held back by a monopoly right. After all, these projects are supported because they are proposed in the public interest. If the researchers had wanted to pursue their own self-interest or that of investors, then they should have sought funding from those sources rather than from the federal government.
But, the university patent administrators argued two points. First, for making inventions available at no charge, universities are better at it than is the federal government because their patent personnel are closer to the inventors and have greater incentives to make an impact. So the federal government should let universities do what the federal agencies would do anyway–use the patent system to grant non-exclusive licenses, but make sure that any licensees are qualified and produce products of good quality and don’t misrepresent things to customers.
Second, there are some situations in which a patent monopoly is necessary to attract private investment to create from an invention a product that the federal government won’t underwrite and won’t come into existence otherwise and the invention can’t be used except as a commercial product. Well, that’s a special case. The university administrators argue that they can do a better job than the federal government here, too, by being closer to the inventors and having greater incentive to find an appropriate licensee. Furthermore, universities can do exclusive licenses with less overhead than can federal agencies, and being motivated by money, they will have greater interest in licenses resulting in commercial products. Oh, and the money that they do make will get reinvested in the licensing program to make it even better than just better than the federal programs, and any money beyond that will go to extend research programs, adding more money to university research than Congress allocates to the funding agencies for this purpose.
With all these arguments, how could anyone object? Except that it all falls apart in practice. Universities do a worse job now with invention licensing than they did before Bayh-Dole, and even then they were doing worse on biopharma licensing than was the federal government. Universities for the most part don’t make much money on their licensing programs, and so their expansion of federal research efforts is inconsequential. They operate on a portfolio model rather than an agent model and thus don’t make available nearly all of what they claim for management, locking everything up behind a patent paywall. And they almost never do non-exclusive licensing. They would rather sit on a pile of unlicensed inventions behind a paywall that might become a troll rock than to release inventions for general access and use. So it’s a pile of crap, but to a university administrator who can’t give it up, it sure is a sweet-smelling pile of crap.