The IPA and Bayh-Dole on nonprofit assignment of subject inventions, 1

Norman Latker, patent counsel at the NIH, drafted Bayh-Dole on the sly, working against HEW policy on inventions to create an easier pathway by which nonprofits could pass exclusive control of inventions made in work receiving NIH funding to the pharma companies of their choice. Latker based Bayh-Dole on the Thornton bill (according to Betsy Ancker-Johnson) and on the Institutional Patent Agreement master (according to Latker himself). Really, Bayh-Dole is a Frankenstein’s monster of body parts snatched from the IPA, Thornton, and the Federal Procurement Regulation.

Let’s look at how the IPA and Bayh-Dole handle one provision–the assignment of subject inventions. We will go a long way around to provide context about how both the IPA and Bayh-Dole operate. Short form–both the IPA and Bayh-Dole require assignees of nonprofit-acquired subject inventions to follow the nonprofit requirements of the funding agreement.

Here’s the IPA:

VII. Patent Management Organizations

The Grantee shall not assign any subject invention to parties other than the Grantor in circumstances as set forth in this Agreement except it may assign rights in the invention to a nonprofit patent management organization, provided that the patent administration agreement between such organization and Grantee is approved by the Grantor. Any reference to a Grantee in this Agreement shall also include a patent management organization when applicable and an assignment to such an organization shall be subject to all the terms and conditions of this Agreement.

Grantor is the United States of America, represented by the Secretary of the HEW. Grantee is the nonprofit. Let’s work through this provision.

Patent Management Organizations

The heading indicates the primary concern of the provision is assignment of inventions to patent management organizations. This provision is not some off-hand, out of the way, maybe it might happen so it ought to be covered thing. Almost all universities worked through patent management organizations. Only a few universities–notably University of California and MIT–ran their own patent licensing operations in 1968 when Latker relaunched the IPA program. Mostly, universities used Research Corporation or a campus-affiliated “research foundation” modeled (somewhat) on the Wisconsin Alumni Research Foundation. Dealing with patent management organizations, then, is for the IPA a central concern.

The Grantee shall not assign any subject invention to parties other than the Grantor in circumstances as set forth in this Agreement

First, a general prohibition regarding assignment of subject inventions, to be followed by the exception. A Grantee may assign only to the Government and only then when the IPA conditions pertain–that is, if the Grantee does not elect to file a patent application and administer patent rights, then the Government has the right to dispose of the invention according to its regulations, which track the Kennedy patent policy (and later the Nixon modifications). For inventions made in federally supported research in public health, the Kennedy policy specified federal ownership of inventions unless a contractor made a persuasive case that by retaining exclusive control over an invention (but for the Government), it would better serve the public.

Note that the standard in the IPA for retaining rights to a subject invention is that the Grantee elects to file a patent application, not, as in Bayh-Dole, to retain title. In the IPA master, the crucial decision by the Grantee is to file a patent application. Under the IPA, when a nonprofit reports a subject invention to the NIH (the key part of HEW for our purposes), it must at the same time provide notice whether it will file a patent application. If so, then the IPA requires the nonprofit to have a patent agreement in place by which the nonprofit receives assignment of the invention from the inventor. Here’s the IPA sequence, then:

1) Nonprofit agrees to the IPA master agreement.

2) Nonprofit requires for all HEW research funding potential inventors to agree to assign each subject invention that the nonprofit chooses to file a patent application on the subject invention.

3) Nonprofit receives an HEW research award (no mention of development in the IPA scope).

4) Inventor makes subject invention.

5) Inventor reports subject invention to nonprofit.

6) Nonprofit reports subject invention to NIH and informs NIH whether the nonprofit has decided to file a patent application.

(a) If not, then NIH regulations apply. There’s no authority in the IPA program or NIH regulations for the nonprofit to require the inventor to assign to the nonprofit the subject invention. The NIH takes ownership of the subject invention unless the inventor requests a determination of greater rights and the NIH concurs. And the NIH then can choose to patent or not–and if the NIH does not choose to patent, in effect it releases the invention back to the inventor and the nonprofit (because who else knows about it?), but now shorn of any patent rights.

(b) If so, then the IPA applies and the nonprofit is required to take ownership of the subject invention under the required patent agreement between the nonprofit and inventor and file the patent application it has asserted that it will file.

The sequencing is important. Under the IPA, a participating nonprofit agrees to a general provision covering all HEW funding. The nonprofit agrees to have an agreement with employee inventors under which they must assign subject inventions that the nonprofit chooses to patent. The assignment requirement is general–applicable to all HEW funding agreements–but also conditional–it applies only when the nonprofit chooses to file a patent application. Otherwise, control of the invention passes back to HEW standard policy. There’s no general requirement in the IPA that inventors must assign outright all subject inventions to the nonprofit.

In effect, however, the IPA operates as shadow HEW regulatory policy without going through a formal process to change the regulations. Indeed, when Latker attempted to get the regulations changed, he was blocked by Congressional oversight and HEW shut down his rogue IPA program (the NSF and Department of Commerce had also created IPA programs on the NIH model, and these were shut down as well). Not only were these programs rogue (not established through federal agency rule-making procedures) but also they were counter to executive branch policy and–worse of all, perhaps–ineffective.

Now we get to the patent management organizations. If a university was going to file a patent application, generally it would ask a patent management organization to do the filing (and pay for the filing). Only a handful of universities–most notably the University of California and MIT–did their own patent applications. The patent management organization then has to review the offered invention and decide whether it will take management (and, usually, ownership). Research Corporation’s acceptance rate was 10% to 15% of submitted inventions. Generally if Research Corporation declined to manage, the university would not file (and pay for) a patent application on its own–the invention rights would revert to HEW policy, which followed as it must the Kennedy patent policy, meaning that the invention would be “dedicated” to the public or if the NIH obtained a patent, the invention would be licensed non-exclusively.

From the get-go, then, there were two points of selectivity for subject inventions. If a university did not think that an invention was worth patenting, or submitting to Research Corporation (or another research foundation that managed inventions), then that was that–no notice under the IPA, no reason for the university to bother obtaining title to the invention merely to assign it over to the federal government. The inventors could do that themselves–more direct and a matter between the agency and the inventors.

The second point of selectivity was with Research Corporation. It declined 85% to 90% of the inventions it received. In the IPA program, then, no more than 1 in 10 subject inventions would end up with the university requesting assignment from the inventor, or assigning the university’s side of the IPA-required patent agreement with the inventor (substitution of parties) to Research Corporation (or another research foundation) so that the patent management organization could require the inventor’s assignment and so comply with the IPA. Most inventions, then, went to the public domain or were available non-exclusively (and royalty-free).

Once past the prohibition on assignment, then, we get the exception for patent management organizations:

except it may assign rights in the invention to a nonprofit patent management organization

Not just any patent management organization qualified–it had to be a nonprofit. The effect is that a nonprofit that has been accepted into the IPA program can add additional nonprofit participants to do the patent managing. But there are conditions, just as there are to get into the IPA program:

provided that the patent administration agreement between such organization and Grantee is approved by the Grantor.

The deal under which the university assigns a subject invention to a patent management organization has to be NIH approved. There’s no guidance in the IPA with regard to what constitutes a patent management organization. It’s a nonprofit. It files patent applications instead of the university. Could be most anyone. But mostly it was Research Corporation or, for University of Wisconsin, WARF (which did the first IPA agreement with Latker). Again, it is the patent administration agreement–not just the assignment transaction–that is subject to agency approval. The patent administration agreement may cover all sorts of things that may not be in an assignment transaction–patent and licensing diligence, royalty-sharing, administrative costs, use of income, remedies and termination for inaction, liability, dispute resolution, and the like. Once you slip in a puddle of bureaucracy, you can reasonably expect to slide for a long, long way.

Any reference to a Grantee in this Agreement shall also include a patent management organization when applicable

The definition of Grantor then is expanded here to include any patent management organization once (i) the patent administration agreement has been agency approved and (ii) the university (or inventor) has assigned the subject invention to the patent management organization. Thus, “Grantee” doesn’t mean, generally, grantee–it means the organization that hosts HEW research and anyone that gains ownership of a subject invention under an HEW approved patent administration agreement. “Grantee” is just cognitive dissonance. Really, it’s “conditional nonprofit institutional owner of a subject invention.” “Grantee” is shorter, but wrong. CNIOSI is about as short, but ugly. This is how clever works. Pick something known to call what you are cleverly doing, instead of calling attention to your new thing with a name that calls it out.

This entry was posted in Bayh-Dole, History and tagged , , , , , . Bookmark the permalink.