Bayh-Dole disruptions of research and innovation

Bayh-Dole allows nonprofits and small businesses to retain title to inventions made in federally funded work. That is, the nonprofit or small business must first obtain title on their own.

For businesses, this is straightforward as businesses routinely assign employees to work on research tasks and require patent agreements that promise to assign any inventions to the employer. That’s not the case with universities.

Universities do not generally assign faculty or students to research tasks, do not direct or supervise or approve their research and do not dictate when, where, and what they publish. Bayh-Dole does not vest invention ownership with universities, does not require that they take ownership, does not give them any special advantage or privilege to get ownership. They have a right to keep inventions that are assigned to them. But the basis for that assignment has nothing to do with Bayh-Dole.

What has happened, then, at universities? University patent administrators, led by the Association of University Technology Managers (AUTM), misrepresented Bayh-Dole as an invention vesting law–universities had a right to “elect” title. By pointing at an invention made in federally work and saying “the university chooses to own that one,” so AUTM would have it, Bayh-Dole vested title to that invention with the university. University patent administrators then demanded that universities change their policies “to comply” with Bayh-Dole, which universities dutifully did.

Prior to Bayh-Dole many universities had policies that recognized that faculty inventors owned their inventions. Universities did not demand ownership except in those situations in which the university had expressly commissioned or assigned research work–and for faculty, that usually meant a voluntary arrangement outside normal work, for extra pay or the like. The ownership of inventions made in extramural research, “outside the walls” of normal work, was determined by the contractual requirements of that work, if there were any.

One of the leading university patent management organizations, the Wisconsin Alumni Research Foundation (WARF) needed a way to get access to faculty inventions. The University of Wisconsin not only did not demand ownership of faculty inventions, it did not even have a patent policy on the matter. Thus, WARF had to depend on faculty choosing to assign inventions to WARF for management. But WARF figured out a way around this problem. The director of WARF, Howard Bremer, worked with Norman Latker, the NIH’s patent counsel to create a new “Institutional Patent Agreement” program. In this scheme, the NIH would approve nonprofit organizations (such as universities) for IPA participation and sign a master patent clause, an institutional patent agreement, that required the university to take ownership of any invention made under NIH contract that the university chose to patent. Faculty had no say in whether a university joined the NIH IPA program. But once a university signed an IPA with the NIH, it had in effect changed its own patent policy and now just by looking at an invention made in work funded by the NIH and deciding they wanted to patent it, the university patent administrators could demand assignment. You can see here that idea of “electing” title. With the IPA program, universities did not have change their patent policies from faculty-ownership to university-ownership. They just signed an agreement with the NIH that took precedence (so they claimed) over whatever policy they had.

It was clever. Find someone to contract with to void your own policy, claiming that contractual commitments must take precedence over policy. It’s like a government signing a treaty and claiming the treaty then takes precedence over laws assuring civil liberties.

The NIH IPA program ran from 1968 to 1978. It had expanded to include the NSF and the Department of Commerce. But when Latker tried to make the IPA program government-wide, Congressional review shut the program down as ineffective and contrary to public policy. Latker immediately began working on the bill that would end up being called Bayh-Dole, which was introduced in 1979 and passed under strange circumstances in 1980 after being initially voted down. (There was another, more comprehensive bill in the works, with even broader support, that Bayh-Dole was competing with–a story for another time). Bayh-Dole, according to Latker, was based on the IPA program.

This misrepresentation was widespread but wrong. You will find the misrepresentations everywhere–at university tech transfer web sites, at law firm summaries of Bayh-Dole, even in academic papers that critical of the law. It all came to a head at the Supreme Court in Stanford v Roche, decided in 2011. AUTM and scores of university lawyers all claimed that Bayh-Dole vested title with universities, or gave universities a first right of refusal, or prevented inventors from assigning to anyone other than the university that hosted the federal research. Even the fact they couldn’t get their story straight showed how off they were. Even Senator Bayh got into the act with an amicus brief that argued inventors were the last in line to own their inventions, after universities and federal agencies had got first sniff. The Supreme Court swept away all such arguments, read the law carefully, and ruled that Bayh-Dole did not vest title. It just allowed federal contractors to keep title if they had got it, subject to various requirements.

But the damage caused by the misrepresentation of Bayh-Dole had been done. Universities have changed their policies to require assignment of all inventions. The policies made broad claims–not just patentable inventions made in federally funded work, but rather any inventions, and also unpatentable inventions, software, data, tangible objects, information, know-how, ideas, improvements, anything that could be owned, anything related to the subject area under which someone had been hired regardless of whether made on campus or under a work assignment, anything made during the time someone was employed, anything that could have been made at the university but wasn’t, anything that was in the foreseeable direction of any research activity at the university. Even when it became apparent that Bayh-Dole did not give universities title to inventions made in work with federal support, university administrators doubled down and argued that such vesting was what “was intended” by the law (even if it wasn’t the law), and so “to make the law work as intended,” universities must demand assignment of all “inventions.”

This change in university invention policies represents a substantial change–and has caused a great deal of damage. In effect, university ownership claims abridge academic freedom–the freedoms of research and publication that are fundamental to many university policies on faculty governance. It’s more than that faculty should study what they want, how they want, and publish as they choose. Faculty-led research has been the last great space of unencumbered research. Vannevar Bush, in proposing what led to the National Science Foundation, called it the “free play of free intellects” and argued it was important to any effort to expand the frontiers of science. Anything mission-directed, tied to corporate or government interests, was less likely to discover new science.

Bush had led the development of new technology for use by the military during World War 2, operating a set of projects akin to skunk works. He brought together academics on top of cutting edge science with engineers from industry who knew design and manufacturing methods and “gadgeteers”–what we might now call “makers” or “DIYs”–people who could build things and get them to work. In proposing what would become the NSF, Bush argued that the next area for development should be medicine, and for his scheme to work, the starting point was new science, and from there teams of developers outside the established order (government, companies) developing fundamentally new technology and methods.

You might see how such a proposal threatened universities, federal agencies, and industries that enjoyed their status in the established order. There was a great effort to rein in Bush’s proposal, which is what happened. Instead of a national research foundation, the new National Science Foundation was just one of many new research initiatives. Every federal agency wanted its own control of research. Each federal agency put out its own requirements for inventions. This diversity of things was organized into a coherent policy in the Kennedy administration, resulting in an executive branch patent policy in 1963. Under that policy, company contractors could own inventions made under federal contracts, except in a few areas. Most notably, inventions made under federal contract and directed to public health would be owned by the federal government, and released for all to use and develop, except when a contractor made a case that private patent exploitation would result in a greater public benefit than open access. For nonprofits and contract research organizations that had no production capability, the federal government would take title to inventions for open access unless the contractor could make that case for greater public benefit.

In effect, the federal government operated on a policy of open innovation for matters of public health–unless someone could show that exploiting a patent monopoly would be better for the public than open innovation.

For most everything else, the federal government let contractors with actual products and markets acquire patents and exploit them, so long as they did use the inventions the patents covered. If a contractor didn’t want to patent a given invention made under federal contract, the federal government could do so–but only to promote open innovation, not to turn around and license that patent exclusively to a competitor of the company choosing not to patent.

For contractors without actual products and markets–contract research firms that just did research, and universities, that handled the money for faculty who just did research–the federal government started with open innovation and let these non-producing contractors make a pitch for exclusive patent control–just show how one’s exclusive patent control in the circumstances better serves the public interest.

The NIH and NSF used their IPA programs to circumvent these arrangements. They allowed nonprofit contractors–non-producers–to take ownership of inventions whenever they chose, without having to make a case for how the public would benefit if they did so. The IPA program was ineffective and was shut down. Then came Bayh-Dole, written by the same folks, to do the same thing, but this time as part of federal patent law, out of easy reach of the executive branch.

So to drill down, here’s what Bayh-Dole has done. First, it has dropped the requirement that nonprofits make a case that their exclusive dealings in patents on inventions made in public-interest work is better for the public than open innovation.

Second, Bayh-Dole authorizes federal agencies to grant exclusive licenses. That means that if a company waives its interest in patenting a given invention, the federal agency can take ownership of the invention, patent it, and license it exclusively to a competitor rather than help to make the invention useful as a common tool, platform, or standard. Companies now have to patent inventions made under federal contract to defend themselves from federal agency patent practices.

Third, Bayh-Dole has been used by universities to disrupt freedom of research and publication. By asserting ownership of inventions made in federally funded work, university administrators also take control of publication (patent applications are a form of publication)–they force inventors to publish in the patent literature. That’s a usurpation of academic freedom to publish. It says something about our norms if in generally we don’t care about the freedom to publish.

But the more problematic thing for innovation has to do with what university administrators do with patent rights. Remember–there was virtually no reason for university administrators to take ownership of an invention except for the exploitation of exclusive patent rights. (I write “virtually” because there *were* good reasons to take control of inventions–but those reasons were the same as the federal government’s old reasons, to support open innovation). Once a university asserts a patent monopoly, then even the university’s own inventors have no right to make or use the invention, except at the university. They cannot help others use the invention, and they cannot leave the university and expect to have access to the invention outside the university.

The university, by seeking to exploit the patent monopoly, shuts down its own inventors’ ability to help others. All inventors can do is help an exclusive licensee, if there ever is one–and usually there isn’t. More recently, universities have given up on trying to find a real licensee and instead create paper startups, and license exclusively to these shell companies. Often, the principals at the shell company turn out to be the inventors. The inventors have to license back their inventions from the university, pay the university for the patenting costs involved in taking the inventions from the inventors, and then make a show of running a company to make it look like the university supports “economic development.”

University inventors cannot even move to new universities with their inventions. Yes, often a university won’t sue another university for infringement, but the key thing–technology transfer–is shut down. Whatever the inventors might do with their invention at their new university, they cannot teach others how to use their new thing because they do not have the authority to license the old invention at the same time. They are, as it were, screwed over.

There’s one more thing that happens when a university stakes out an exclusive patent position. Everyone in industry that does not need an exclusive position or does not expect to be the one to get the university’s exclusive license on offer has to avoid the invention, design around it, ignore it, refuse to use it, exclude it from future plans, exclude it from standards. If you want to make sure that early adopters and industry folks will avoid your invention–file a patent on it and insist that you will license it only exclusively to a company that will make a commercial product–thereby cutting out all other uses (such as research use and professional uses that do not involve making product to sell to others) and cutting out all other potential collaborators on further research.

A faculty inventor may be free to choose research, but the moment a university asserts ownership of that inventor’s inventions, the inventor loses the freedom to choose research involving the invention and expect to be able to publish that new research for anyone to take up and use. In this way, practice under Bayh-Dole suppresses not only the diversity of research but also the adoption of new findings.


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 *

This site uses Akismet to reduce spam. Learn how your comment data is processed.