The Bayh-Dole Act has gone from being a golden goose to a weapon of mass destruction. The basic idea, to make more federally supported inventions available for management through university-affiliated agents, is fine. More to the point: as the federal government has expanded its funding of university-hosted research, the federal government has come to dominate the topics and directions that research takes. In essence, the federal government has bought out many of the best and brightest in university research. Some commentators, such as Michael Crow, president of Arizona State University, are concerned that federal funding has diverted university faculty from their independent pursuit of inquiry (really, watch the video–it’s under 4 minutes). President Eisenhower worried just this sort of thing fifty years ago.
Thus, something like Bayh-Dole is essential to make it clear that just because the government provides support for research proposed by faculty, the government is not as well buying out and controlling the inventions and discoveries made in that research. The idea is to get the discovery work done, not to procure for only the cost of supporting the research, everything that faculty (and others) might discover or invent. In the same manner, Bayh-Dole is also not about putting the state government (in the case of public universities) or private universities in charge of this same buy-out of faculty inventiveness. Something like Bayh-Dole is essential for this point as well. Stanford v Roche is a great start. But we need a something like a bill of rights that restricts the power of government (whether state or federal) with regard to research inventions.
However, Bayh-Dole goes too far. It disrupts a perfectly effective federal policy of using flexibility in the management of funding agreements in favor of an arbitrary rigidity. There are times when it may well be advantageous for the government to own patent rights in inventions made with its support, where doing so allows standards to be developed, provides for interoperability in systems, levels the playing field for competitive bids, and provides for broad access and competitive application. We might then make a distinction between government grants to universities and non-profits (and perhaps even small businesses in some situations) and contracts under which the government stipulates the statement of work. Or, another way, distinguish Circular A-110/2 CFR 215 from the FARs and DFARs. If Bayh-Dole is restricted to grants and collaborative agreements, then federal agencies could recover greater flexibility in their management of inventions in contract research. To my mind, that would be an advance.
Further, Bayh-Dole destroys the apparatus for public accountability that used to exist in federal contracting with universities on behalf of faculty projects. Bayh-Dole guts reporting, limitations on exclusivity, agency discretion to intervene, and proposes a march-in procedure so unattractive and unworkable that it has never been successfully used. It is my impression that the failure of march-in is, in fact, one of the unheralded “successes” intended by those controlling the drafting of Bayh-Dole. They didn’t get all that they wanted, but they sure did manage to destroy public accountability for the non-federal management of subject inventions. This accountability, far from introducing more overhead and uncertainty, is essential to the overall management of subject inventions by private entities.
The one place the accountability still exists is in the grant regulations governing funding agreements with universities in the form of grants. There, at 2 CFR 215.37, one encounters this statement:
§ 215.37 Property trust relationship.
Real property, equipment, intangible property and debt instruments that are acquired or improved with Federal funds shall be held in trust by the recipient as trustee for the beneficiaries of the project or program under which the property was acquired or improved. . . .
Intangible property is given a broad definition:
(s) Intangible property and debt instruments means, but is not limited to, trademarks, copyrights, patents and patent applications and such property as loans, notes and other debt instruments, lease agreements, stock and other instruments of property ownership, whether considered tangible or intangible.
When a university claims ownership of inventions made with federal support as a condition of employment, and the employment involved is the work of faculty (and others) paid for with federal funds, then the university is stipulating that it (“the recipient” in 2 CFR 215 terminology) is acquiring the invention (and therefore rights to patent) with federal funds. When a university takes ownership in this way, according to the regulation, the university must act as a trustee of the patent on behalf of others. The university does not have the right to act on its own interest–or, another way, the university has agreed not to act on its own interest, but rather manage what it has acquired and improved for the benefit of others. Where a university has a policy that says it will manage inventions for the public benefit and its own benefit, the latter purpose cannot operate. Despite the presence of this requirement in federal grants, I do not know of a single university that acknowledges this provision in its management of subject inventions. It certainly does not arise in policy statements or in guidance to faculty.
I would expect that university licensing officials would in general disclaim it, arguing that it doesn’t apply, that it is a mistake, or that Bayh-Dole supersedes it. In doing so, of course, they also would be arguing that universities do not have a role as trustee of inventions they obtain, but may behave “like a company” and use patents to attack anyone who has money and ought to pay, or threatens a stream of money that university administrators believe they should get. The reality is, the folks that claim possession of the idea behind Bayh-Dole appear to be quite happy suppressing the idea that universities have special obligations to the public when they obtain ownership of subject inventions by requirement.
This is an important point. 2 CFR 215.37 does not place restrictions on inventions that are assigned to a “recipient” voluntarily, outside the context of the use of federal funds to acquire or improve them. Thus, when a university imposes a compulsory invention assignment policy that extends to inventions made with federal support, it chooses to come within 2 CFR 215.37. If the university does not require assignment, then if it obtains assignment because an inventor chooses it to be an agent, then the university has not used federal funds to acquire the invention, and thus is not subject to the requirement that it act as a trustee–though perhaps that role would still be a valuable one not to give up readily.
Bayh-Dole did accomplish one important thing, not typically discussed by the self-proclaimed defenders of Bayh-Dole, because it represents their greatest failure. In the institutional patent agreements that Bayh-Dole was to replace, the government stipulated that the university must take ownership of inventions made with federal support. Bayh-Dole, however, does not stipulate ownership. The law applies to federal agencies, not to those who the agencies might contract with. The law shapes how federal agencies construct and manage patent rights clauses in funding agreements.
Thus, Bayh-Dole and the IPAs are very different in character and reach. It is one thing to contract for procurement of inventions; it is quite another to pass a federal law that strips inventors of ownership and passes it by statute to another owner. This was what Stanford v Roche at the Supreme Court level was all about, not that ghastly canard about present assignment. There was no way to get Bayh-Dole passed in a lame duck Congress if it represented a wholesale change in federal patent law regarding ownership, which in turn would have raised a substantive constitutional issue with regard to whether Congress had the power to vest ownership of inventions in anyone but inventors.
But there was still a way to put the ownership requirement into the standard patent rights clause. This clause, at 37 CFR 401.14(a), could have replicated the requirements of the IPAs. But the ownership provision is not there. Why did Norman Latker not do this? Instead, there’s the (f)(2) provision requiring a written agreement, but that written agreement is not stipulated to be between the university and its research personnel. It is an agreement that delegates specific performance responsibilities to those research personnel, regardless of what sort of patent policy (if any) or patent agreement (if any) a university might otherwise have. Those responsibilities are to report inventions, sign papers that allow patent applications to be filed, and to establish the government’s rights in inventions. For the latter, an inventor cannot do this if the inventor has assigned ownership of the invention to the university. Clearly, if the Bayh-Dole Act secretly and mysteriously vested title in the university host automatically, then there would be absolutely no need for (f)(2) to require an agreement under which research personnel agree to establish the government’s rights. It’s clear that whoever drafted (f)(2) knew that Bayh-Dole did not vest title of subject inventions with universities.
Further, it’s clear that whoever drafted (f)(2) also understood that the standard patent rights clause was not going to require universities to take ownership of inventions. I suggest that one reason why the standard patent rights clause was not going to do this was because many universities had policies under which they disclaimed an ownership interest in inventions or had no policy at all on the matter, and were not prepared to create an internal invention management operation simply to be eligible to receive federal funding for research. Given that many universities designated an affiliated foundation, Research Corporation, or another invention management agent to handle inventions, it would be a stretch for a federal law or a patent rights clause intended to make more federally supported inventions available to such an approach to instead destroy this approach by requiring all subject inventions pass through the ownership by the universities that happened to host the work leading to the invention.
Bayh-Dole is silent on ownership. The standard patent rights clause authorized by Bayh-Dole does not stipulate ownership and instead requires that universities delegate duties to research personnel. It is left to the universities–faculty and administrators–and to principal investigators and inventors–to sort out what agent might be used, if any, to undertake patenting of any given invention. The essence of Bayh-Dole is, if no invention management agent is chosen, then the federal government decides whether to take ownership (relying on the (f)(2) commitment by inventors to establish the government’s rights) or allow the inventors to retain the ownership that federal patent law grants them. If the latter, then the government relies on a separate standard patent rights clause, at 37 CFR 401.9, that applies only to inventors when allowed to retain title.
Bayh-Dole’s silence on ownership has led some to believe that this is a defect in the law, or in the implementing regulations. I don’t believe that for a moment. The silence is an artifact, not a defect. It is the result of a recognition, at the level of the law, that the law cannot dictate an ownership position other than that provided by the constitution. At the level of the standard patent rights clause, it is the recognition that it cannot impose across all colleges, universities, institutes, and other nonprofits a demand that they own and manage inventions for patenting, regardless of their institutional policies and capabilities on the matter. Despite the possible desire of some university administrators and research foundation folks to impose such an ownership requirement, it did not and could not happen. They got as close as they could, but they did not seal the deal. Thus, Bayh-Dole, and the standard patent rights clause, do not stipulate ownership, and there is no need to stipulate ownership, provided that (f)(2) agreements are in place.
The on-going effort, then, by university administrators to “fix” Bayh-Dole by imposing a compulsory assignment obligation on faculty (and others) is both strange and horrifying. Administrators first claimed that Bayh-Dole required assignment, playing on a deliberate confusion over the meaning of “retain” as in “retain title.” The standard patent rights clause is clear on the usage, but administrators took the “retain” out of context and made it appear to mean “take” as in “take title.” Some policy statements flat out state that Bayh-Dole, or federal regulations, require the university to take title, or give the university first right of refusal, and the like. All this was tossed by the Supreme Court, but university administrators have generally not changed their policy statements. Instead, they just moved to change the rationale on which they imposed the ownership claim to inventions–if not Bayh-Dole or the standard patent rights clause, then employment or use of resources or participating in extramural research or association with the university–anything that sounded plausible and lawyers were willing to be paid to defend.
Thus, as an instrument in the hands of certain university administrators and their patent profit-seeking allies, the Bayh-Dole Act has been used to destroy the diverse approach to invention management that existed prior to Bayh-Dole; to replace largely voluntary assignment with compulsory assignment; to force out the invention management agent approach and turn universities into conflicted profit-seeking operations, giving up their role as advocates for public use and proper handling of inventions; and to attack the relationship between faculty and administrators, introducing first in the guise of a legal or contractual requirement, and then when that failed as just a pretty keen idea, a demand for ownership of faculty scholarship when it was at all “inventive”–even when it wasn’t “patentable.” The result has been the destruction of an approach that was “working”–the success of university-affiliated patent licensing prior to Bayh-Dole was the fundamental basis for the passage of Bayh-Dole–destruction of a competitive infrastructure for invention management available to faculty inventors; and an attack on the academic freedom of faculty–that they cannot teach, or use, or disclose their inventive work or use the inventions published by other faculty, without the permission of a university bureaucrat.
That’s some national innovation system, don’t you think? No innovation unless a bureaucrat has control and gets paid first. It’s great if you think the world was made for bureaucrats and that all the federal funding to universities is so there will be a steady stream of opportunities for bureaucrats to seek profits from patent positions. It’s great if you believe that when bureaucrats make money, the public interest is served. Otherwise, it’s horrifying. The stuff that gets out of the present system (and called a “success”) is doing so largely despite the bureaucrats. They are not a catalyst so much as a flow restricter, a filter, a glob of fat in the pipe reducing the overall flow. Yes, it is remarkable and a cause for joy when something does get through. If you understand the plumbing, however, you know what is possible, what is happening, and that something ought to be done about it.
Not all is lost. The Bayh-Dole Act has some good features–especially that it does not dictate ownership of subject inventions. It can be saved. But there is work to be done. Here are five steps to get things going in the right direction: to stimulate innovation, to restore academic freedom, to foster diversity and competition, to rebuild the capacity to manage new stuff. Here they are:
1. Implement freedom to innovate. Freedom to innovate means that universities do not require the assignment of inventions. Not as a condition of employment. Not for using resources. Not for being involved in sponsored research. Not by association. Freedom to innovate means that faculty are free to disclose, teach, and use their discoveries and inventions without having to get the permission from a bureaucrat. For federally supported inventions, the university has no equity interest: the government pays for the research, and for the indirect costs. The university therefore should have no need of a financial interest, a shop right, a license, nothing. It may well be that principal investigators or inventors want the university to have one or more of these things–and a well run university administration is likely the greatest way to create the goodwill that makes such desires a reality. Indeed, a good metric of a well run university administration is the extent to which faculty inventors dedicate a portion of their income, or grant licenses to, universities in recognition of their support and common purpose.
At the state level, legislatures can prohibit public universities from compelling assignment of inventions made with federal support. This would be a tremendous step forward for research invention management. Bayh-Dole need not change at all. The states that do so will be back on the cutting edge, will reduce the costs spent on muddled take-everything trawling for inventions operations, and will stimulate competition to develop strong, responsive technology transfer programs. The states that implement freedom to innovate will have a competitive advantage over other states in the recruitment of faculty, development of collaborative industry arrangements, and public support for instruction as well as research.
Alternatively, the change can be made at the federal level, with a simple modification of the standard patent rights clause at 37 CFR 401.14(a). At section (k), make it clear that nonprofits cannot require assignment of subject inventions as a condition of federal funding, in parallel with the prohibition regarding doing so in the subcontracting provision at section (g)(1): the choice is with the inventors whether to assign to the university, to another qualified agent, to the government, or request that the government allow the inventors to retain title. Innovation opportunities depend especially on such local, not bureaucratic or arbitrary, choices.
2. Eliminate the FOIA exemption in Bayh-Dole. Require public accountability for all subject inventions for which patents are filed by nonprofits. Again, put the change in section (k) of the standard patent rights clause, but here one also has to amend Bayh-Dole proper, where the FOIA exemption is granted. Require annual public reporting. Such reporting does not have to reveal actual sales figures or non-public development information, but it does have to show what has been licensed, scope of that license, date of first commercial sale or use, and within a range, the money received. Public accountability will not cast a pall over private efforts to develop publicly supported inventions. It no doubt will give a good whack to the hornet’s nest of bureaucracy. We deserve to know what is going on, who has got monopoly rights, and what they are doing with them.
In any event, we need audits to determine what the actual situation is after 30 years of university accumulation of patents without the requirement to account publicly for these holdings. For Bayh-Dole, audits should consider only in subject inventions. For each, the audit should determine if the invention really is a subject invention (or whether the university simply asserted that it was to better position itself for making an ownership claim), whether the invention’s assignment was based on deceptive or wrong premises (such as that federal law requires assignment), and for each subject invention show its patent status, licensing status, date of first commercial sale or use, if any. That would provide a good basis for a discussion of how management practices might be improved.
3. Add public domain to Bayh-Dole. Allow faculty inventors under the standard patent rights clause authorized by Bayh-Dole to dedicate federally supported work under grants to the public domain or to a FRAND standard. Keeping an invention free of patents, or using patents to contribute to a standard is an important way of promoting use, and laying the groundwork for commercial development. Using the patent system should include publishing without using the patent system. This is especially true of methods generally and work of any sort that ought to become part of a standard. Again, a simple modification of the standard patent rights clause is all that’s needed.
4. Implement the (f)(2) provision. Obtain the required written agreement in the form specified by the standard patent rights clause. That form delegates assignment rights to inventors of subject inventions, regardless of any university policy or patent agreement to the contrary. Federal contract requirements take precedence. If universities refuse to do this, then, again, state legislatures wanting to open up innovation activity around research can require public universities, at least, to comply with the requirements. Federal agencies may require proper (f)(2) agreements as a condition for releasing funds to a university on a given grant. That does not require any legislation at all–just a commitment to create the relationship between research personnel and the funding agency that the standard patent rights clause stipulates, and relies upon.
5. Reaffirm the university “trustee” role for inventions made in federal grants. The university’s primary role in the management of inventions should be that of trustee. If a university requires ownership, the trustee standing is even more important. If the ownership is negotiated, then the university’s role is also conditioned on the nature of the agreement by which title is assigned to the university. As a trustee, a university has to drop the idea that its first, and perhaps only goal, is to make money from patent licensing, litigation, and speculation. The emphasis must be on promoting the use of inventions, using the patent system as necessary for the purpose. The emphasis is not on promoting the use of the patent system, actual use of the underlying inventions be damned (or dammed, as the case may be).
No litigation for infringement unless the infringement is causing harm to the public, and especially to the beneficiaries intended by university research projects. Instead, manage inventions within the patent system in a manner distinct from that of the patent troll and the speculative investor always looking for a monopoly position from which to make money. A great success in the use of an invention might not involve a university, or a faculty inventor, receiving a dime licensing. Free and open source software has used this approach for almost as long as Bayh-Dole has been around. Funny thing, it seems to be a useful, well established alternative.