The mistaken assumptions of Bayh-Dole, 1

I know this article by Sean O’Connor on the mistaken assumption in Bayh-Dole is six years old and I have discussed this issue previously, but since it is out there on the web, and as far as I know it hasn’t been retracted, I write another counterpoint to it. Perhaps the purpose served is to create a discussion, at least, with regard to the defects in Bayh-Dole.

The context of O’Connor’s analysis appears to be that university ownership of faculty inventions is a good thing, that Bayh-Dole set out to accomplish that, but screwed up with a mistaken assumption regarding an utterly basic requirement. Bayh-Dole could be fixed and so ensure that institutions rather than inventors control inventive work supported by public money.

Simply: folks drafting Bayh-Dole were mind-numbingly ignorant. Inventors are stupider. University ownership of inventions is virtuous. Bayh-Dole should serve virtue, not stupidity. And it can, but to make the law work, given its origin in ignorance, universities must repudiate academic freedom, tenure, and the freedom of research and publication. Then all will be well.

Sean O’Connor starts “Mistaken Assumptions: The Roots of Stanford v. Roche in Post-War Government Patent Policy” with this claim:

The Bayh-Dole Act of 1980 was built on a mistaken assumption that “contractors”—recipients of federal funding—were securing assignments of inventions from their employees.

O’Connor never identifies anyone involved in drafting Bayh-Dole who evidenced this mistaken assumption. Not Latker, not Bremer. The Bayh-Dole Act was drafted by Norman Latker at the NIH. I’ve talked with the person who typed the draft for Latker so it would not be traced to his government typewriter. Latker is recognized by “Bayh-Dole Central” as one of the “founding fathers” of Bayh-Dole. Another “founding father” is Howard Bremer, then patent counsel for the Wisconsin Alumni Research Foundation and head of the Society for University Patent Administrators (now called AUTM). Latker and Bremer previously had revived the NIH’s Institutional Patent Agreement program, which created master agreements between the NIH and universities concerning research inventions–but which excluded faculty investigators from being parties to that agreement.

The IPA program was used to circumvent both the Kennedy executive branch patent policy and the Public Health Service’s implementation of that policy. Under the IPA program,  NIH-funded inventions could make their way as patent monopolies to private corporations with the assistance of university-affiliated patent brokers, who provided served as “middlemen” to cover the trail from public money to private monopoly. The private monopoly, it was argued, was necessary for there ever to be public benefit from such federally funded research. If university faculty discovered a cure for cancer–so the argument went–no company would bother to create a product based on that discovery unless it was first assured of a patent monopoly.

Latker worked closely with Howard Bremer of the Wisconsin Alumni Research Foundation on the IPA program and then again on Bayh-Dole. In the IPA program, the IPA master agreement required the university to have a patent agreement with each employee under which the employee promised to assign inventions to the university. The sequence was: university received report of invention; university decided to file a patent application on the invention; university then was obligated under the IPA to obtain assignment of the invention from the inventor; university then had NIH’s assurance that the university could exploit the patent right, subject to the conditions of the IPA.

Latker absolutely knew that universities did not ordinarily demand ownership of all inventions made by their personnel, but did routinely have policies that asserted that any contract entered into by the university took precedence over the patent policy. Thus, the IPA at Wisconsin is with the university, not with WARF, and by this means end-ran the university’s own non-policy of allowing inventors to own their inventions (that is–the university had no policy that took ownership of inventions and Bremer reported that the “informal” policy was that the university did not claim ownership).

Latker later asserted that Bayh-Dole could be considered a codification of the IPA program. Despite Latker’s assertion, Bayh-Dole should not be so considered, because it varies from the IPA program in significant ways. But it is true that Bayh-Dole does preserve the IPA’s pipeline of patent monopolies flowing to the pharmaceutical companies via university patent brokers, and in this regard Bayh-Dole does the job of the IPA program. Bayh-Dole goes further, however, and extends the program to all federal funding, not just NIH funding, and thus the pharmaceutical industry can also capture biomedical work undertaken with funding from any federal agency.

Thus, Sean O’Connor’s assertion that Latker assumed universities routinely required assignment of inventions is untenable. Bayh-Dole is not built on any mistaken assumption of this sort. That’s not to say Bayh-Dole isn’t built on all sorts of mistakes and bad drafting.

Here is what happened with Bayh-Dole. First, it couldn’t get passed on its own. It took a last-minute scheme (by a Wisconsin representative–what was WARF’s role?) to attach Bayh-Dole to a bill with amendments to patent law. Thus, Bayh-Dole becomes a part of patent law, not federal procurement law. As patent law, there are problems with the idea that inventions vest with organizations that host federal research rather than with the inventors, as patent law has long provided. The Supreme Court in Stanford v Roche called this point out rather clearly. If Congress had wanted to suspend this fundamental of patent law, it would have done so openly and not relied on idiosyncratic definitions and convoluted interpretations.

The Constitution grants the federal government the authority to grant inventors exclusive rights in their inventions for a limited time, in exchange for the inventors publishing how to practice their inventions. The instances in which the government has by statute refused to allow inventions to vest with their inventors are those in which the government asserted title to the inventions–that is, the government refused to act on its authority to grant a right to inventors. No federal statute has provided that an invention must vest in someone other than the inventor–and not the federal government.

For subvention funding, it is even worse. A university subvention is a grant-in-aid. Faculty propose research projects. Their university excuses them from university duties to pursue the projects. The federal government provides support, and pays the faculty salaries, expenses on the projects, and even reimburses the university administration for its involvement and the use of university facilities. Even though there are situations in which an employer might task an employee to invent and thus have an equitable claim to ownership of a resulting invention, that cannot possibly be the situation in faculty-led extramural subvention research supported by the federal government. For those situations, the university is not an employer, does not direct the research, does not approve it, is not the intended beneficiary of the work, and has no say in the disclosure or dissemination of the results. There simply is no argument that a university has an equitable claim on inventions made in such research.

Thus, for federal subvention research hosted by universities, there never was any presumption that the university would have in place patent agreements that demanded that faculty inventors assign their inventions to the university. There wasn’t even a “presumption of ownership” by federal agencies. They operated–from 1963 on–under presidential executive orders that authorized them to obtain ownership of inventions in research funding contracts unless specific conditions applied–the contractor had a commercial position and capability and the work took the form of procurement, or an agency had reviewed the invention and determined that the public interest would be best served by permitting the contractor asking for a determination of rights to retain ownership of an invention having obtained that ownership by the inventor’s assignment. A federal agency might then “presume” to be entitled to receive assignment of inventions because federal regulations forming the basis for federal research contracts provided for that outcome. Inventors were made parties to the federal funding agreements and therefore had an obligation to assign inventions to the federal government upon request–even if they had no obligation to assign inventions to the university that hosted their work.

For federally funded subvention research hosted by universities, there never was a federal expectation that invention ownership would have to pass from inventors to universities in order for the federal government to acquire its rights. Rather, inventions moved directly from inventors to the federal government, which then made the inventions available to all in the “domestic” market, either by dedication (i.e., not filing a patent application for its own benefit) or by non-exclusive, generally royalty-free licensing. Thus, university inventors inventing with federal support were assured that they, too, would have unconstrained access to their inventive work. The federal government was so open on this matter that it did not bother to keep records of citizen access to inventions that the federal government acquired from its support of university-hosted subvention research.

The arguments behind dedication or non-exclusive licensing were that the federal government should not use the patent system to seek to profit from patent monopolies (through licensing or what’s now called patent trolling); should not put itself in a position to sue its citizens for infringement with the idea of suppressing citizens’ use of publicly funded inventions (other than for national security or public safety); and should not play favorites in dealing in patent monopolies that it acquired. The Biddle report works through these arguments and ends up affirming them as general practice while making the case that there might be times in which a private patent monopoly might be of public value. The Kennedy patent policy takes up this observation in the form of three years of exclusive right from the date of patent issuance for contractors allowed to retain federally supported inventions that they had acquired.

Again, even for the for-profit contractors, there was no assumption that the contractor must require inventors to assign to the contractor to comply with federal regulations. The contractors employees who might invent were subject to the same contract requirement to assign to the federal government unless the federal government agreed otherwise. There was no need to make an assumption about contractor patent behaviors, any more than there was a need to make the assumption that every federal contract and every invention made in every federal contract was within the scope of the contractor’s business or its foreseeable future business. Just wasn’t.

But there is a deeper problem with Bayh-Dole. Actually, it is right there on the surface. Bayh-Dole asserts (35 USC 210) that it preempts all other “Acts” with regard to the disposition of invention rights in the priority chain between a contractor (having obtained ownership) and the federal government (represented by an agency providing funding). Thus, Bayh-Dole cuts off the authority for federal agencies to claim ownership of inventions made in such research as a matter of regulatory authority–but only when a contractor has somehow obtained ownership of such an invention. The only way a federal agency can restore the authority to deal with ownership of subject inventions any other way, once Bayh-Dole is in place, is through Bayh-Dole.

Following the requirements of Bayh-Dole, a federal agency may declare exceptional circumstances before an award is made, or use the DOE portion of the standard patent rights clause for naval propulsion and weapons systems, or rely on the conditions within the standard patent rights clause when a contractor fails to disclose, elect to retain title, file patent applications, or maintain or defend a patent. That’s it. After Bayh-Dole, there is no possibility of a second, secret federal agency authority to acquire outright inventions made in subvention research once a contractor has obtained title. Whatever a federal agency wants by way of inventions from university research it must get through Bayh-Dole or through new legislation that cites Bayh-Dole.

Bayh-Dole, then, did not “reverse” any “presumption of ownership” in inventions made in federally supported university-hosted research. Perhaps that would have been the effect had Bayh-Dole been made part of federal procurement law rather than federal patent law. As part of patent law, Bayh-Dole could not “reverse” a presumption that was never there. Federal patent law “presumes” that inventors own their inventions. Bayh-Dole made no change in that presumption–and the Supreme Court was emphatic on the point. Bayh-Dole did not vest ownership of inventions with universities that hosted federally supported research, did not mandate university ownership, did not give any special right for that ownership, did not even encourage that ownership. Nothing. Bayh-Dole concerns only how a federal agency might obtain ownership or other interest in such inventions if a contractor has first acquired ownership. There is then absolutely no reason why Bayh-Dole should bother with how a contractor first acquires ownership–that has nothing to do with Bayh-Dole, even though the IPA did address that issue.

There is absolutely no need for anyone to assume–even mistakenly–that universities were routinely obtaining assignment of inventions under private patent agreements with their faculty and other research personnel working on federally supported projects. It is fantasy to assert such a thing is the problem with Bayh-Dole. Sean O’Connor co-drafted the AIPLA amicus brief in Stanford v Roche, and that brief played an important role in guiding the Court in its decision–“of the contractor” in the definition of “subject invention” means “owned by the contractor” not merely “made in connection with the contractor, or made in research administrated on behalf of faculty.” But here, when O’Connor attempts his hand at history rather than law, he has it very wrong. It is a bummer to be wrong. It may well be that there are some law professors who have assumed that universities have patent agreements that require assignment–and even some who have worked to make sure that result comes about–but that a coterie of law professors and legal pundits have made mistaken assumptions about the law is not a particularly compelling basis to attempt to project that mistake onto the history of Bayh-Dole and so come to assert that their mistake miraculously also forms the foundation for the law or for Stanford v Roche.

Even in the context of Stanford v. Roche, the claim makes no sense and runs against the facts. Stanford did have a patent agreement with its postdoc. It’s just that the patent agreement stipulated that the postdoc promised to assign those inventions that Stanford’s patent policy authorized the university to have an ownership interest in. And Stanford’s patent policy–since changed to the usual boring crap–at the time insisted that inventors owned their inventions unless Stanford had a legal obligation to acquire ownership. The assertion made by Stanford in taking the case to the Supreme Court was that Bayh-Dole created that legal obligation to acquire. That is, the university claim in Stanford v Roche was not that anyone mistakenly assumed that universities had patent agreements with their inventors so that Bayh-Dole would operate–but rather that by asserting for the law a power that it never had, somehow Stanford could make the postdoc’s obligation to assign operate, or by-pass that obligation altogether and somehow claim to have obtained ownership of the inventions in dispute ahead of Roche, and therefore have standing to sue the company for infringement.

The deeper irony of course is that Roche did not need a patent monopoly on the Stanford-based invention in order to develop commercial product. And while Stanford sat on its licensing hands, Roche was off developing and then selling product, all without any need for a patent monopoly (or exclusive license) from Stanford. Thus, Roche defies at least one other assertion made by Bayh-Dole advocates, that companies will not invest in federally supported inventions unless they get patent monopolies secured and licensed by university patent administrators. Roche is not the exception that proves the rule, but the case that shows the “rule” is fundamentally wrong. Clearly there are speculators and companies that do desire patent monopolies. But their desire does not create a general rule–only political talking points repeated until they are made to seem to be true.

The “failure” in Bayh-Dole is actually not a failure at all. Bayh-Dole does not require inventors to assign to their employers, nor faculty inventors to assign to their university non-employers. While a small company may have a patent agreement that covers an employee’s work on a federally funded project, the patent rights clause precludes that same outcome for university administrators in their dealings with faculty inventors. That’s the effect of the (f)(2) written agreement requirement, though of course university administrators don’t comply with that requirement and perhaps assume–mistakenly–that their gawd awful patent policies–full of contradictions, screwball definitions, assertions, and things that can never be–somehow give them the right to claim ownership of whatever it is they want, and that right means that they don’t have to bother with (f)(2).

If Norman Latker wanted to deal with inventors assigning to the university that hosted their research, section (f) was the place to do it–“Contractor Action to Protect the Government’s Interest.” There’s the (f)(2) written agreement requiring contractors to require their inventors-to-be to make an agreement involving inventions. Instead of requiring inventors to assign inventions to the contractor that hosted the federally supported project, (f)(2) requires contractors to require those inventors-to-be to become parties to the funding agreement–re-implementing the Public Health Service regulation that inventors were subject to the same regulation on inventions as the nonprofits were.

But Sean O’Connor is stuck with a fixed idea–that somehow the goal of Bayh-Dole, and of public policy, should be to find a way to force inventors to give up their inventions to university administrators. Discussing the (f)(2) agreement, O’Connor argues that

Both prongs of this contractual requirement would be unnecessary if Bayh-Dole automatically vested title to the invention in the contractor by operation of law.

That is, if Bayh-Dole did something that patent law had no constitutional authority to do, then things would be just grand for institutional ownership of inventions made by faculty inventors working “outside the walls” of their employed duties to the university. If only the federal government were to exceed its authority, university patent brokers could get everything they wanted without the bother of having to deal with inventors. Why would O’Connor be so flippant about this issue? Or is it a mistaken assumption that patent law can provide for institutional, not inventor, ownership of inventions. Hell, the institutions involved don’t even employ the faculty for the purposes of extramurally funded research. They have no equitable interest in inventions made by faculty. The federal government pays their salaries, pays their expenses, and reimburses the university for the use of its facilities and administrative services. The federal government might have a claim to an equitable interest in inventions–at least a non-exclusive license–but there’s no way in hell that a university that hosts the research does. Sure, a university can bully and blather its way to a claim of ownership–and folks like O’Connor are there to help, it would appear–but a bully-and-blather demand is not an equitable interest.

Nothing in Bayh-Dole requires the (f)(2) written agreement requirement. The public policy logic for having it runs entirely the opposite way from O’Connor’s direction. Both prongs of this contractual requirement are necessary because Bayh-Dole has nothing to do with giving contractors any special interest in the inventions of their employees–especially faculty investigators at universities. Bayh-Dole has been misrepresented as intending to give contractors a special interest, and O’Connor apparently without qualms works toward turning Bayh-Dole into just such a law.

Latker drafted the IPA template with an assignment requirement. Latker also drafted Bayh-Dole. Latker then worked on the implementing regulations. Latker knew exactly how universities such as Wisconsin dealt with faculty inventions. And yet the (f)(2) agreement requirement is silent on assignment of inventions to the university. That’s not the sign of a mistaken assumption but rather of the public purpose of the law (and its limits, given its placement in patent law and its preemption of other law (but for Stevenson-Wydler).

There is no mistaken assumption behind Bayh-Dole. There is a deliberate misrepresentation of Bayh-Dole aimed at making the law appear to provide to universities what it does not provide and can’t provide. The IPA program worked around executive branch policy to contract with universities to violate that policy–and to contract in such a way that faculty investigators had no say in the deal made between university administrators and the NIH, since the IPA deal was not a contract involving research. It was a side deal that the NIH didn’t have the authority to make, and at universities such as Wisconsin, it appears that administrators violated their own policies (or, as they might put it, exempted themselves) to sign up for the IPA program.

Bayh-Dole, by contrast, builds its invention administration into each funding agreement in the form of a default patent rights clause or custom patent rights clause, if the federal agency can justify anything other than the default. Bayh-Dole’s default patent rights clause requires the institutional contractor to require its technical employees to become parties to the funding agreement–and thus to become contractors, to be treated by the federal agency as small business contractors with regard to their inventions–under the inventor patent rights clause (37 CFR 401.9), a reduced version of the small business patent rights clause. Under Bayh-Dole, inventors own their inventions, and have the most liberal requirements on what they can do with their inventions. O’Connor, for some reason, appears to find that outcome objectionable as a matter of public policy.

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

Leave a Reply

Your email address will not be published. Required fields are marked *