There are any number of assertions that the Bayh-Dole Act has been wildly successful. Even many critics of Bayh-Dole start from the position that Bayh-Dole is successful, and then carp about bits of collateral damage and beg for crumbs and mercy. We saw that in the article by Boettiger and Bennett, with all its sad irony–since Bennett at one time was the director of the UC System-wide technology transfer program and did not make policy changes to the licensing program then, when he had the chance.
The problem with these discussions about Bayh-Dole is that the actual information regarding performance of the law is not publicly available, and apparently not even tracked or collected. In 1984, Bayh-Dole was amended to make secret all information regarding use reports, not merely the parts of those reports that were determined by federal agencies to be confidential or privileged.
The fundamental metric set forth by Bayh-Dole is “utilization” of inventions made with federal support–subject inventions, as Bayh-Dole defines them. This goal is announced as the first objective of the law, at 35 USC 200:
to use the patent system to promote the utilization of inventions arising from federally supported research or development.
“Utilization” then becomes the basis for the key definition of “practical application,” at 35 USC 201:
The term “practical application” means to manufacture in the case of a composition or product, to practice in the case of a process or method, or to operate in the case of a machine or system; and, in each case, under such conditions as to establish that the invention is being utilized and that its benefits are to the extent permitted by law or Government regulations available to the public on reasonable terms.
Pinch this down to the essentials:
The term “practical application” means to practice an invention under such conditions as to establish that the invention is being utilized and that its benefits are . . . available to the public on reasonable terms.
When an invention has been practically applied, three things must be established:
- the invention is being used
- the invention’s benefits are available to the public
- the benefits are available on reasonable terms
Bayh-Dole does not argue that one invention in 200 is a good goal for practical application–the stated objective of the law is that the goal is specific for each invention that a contractor acquires. If you own it, then you have an obligation to use the patent system to promote utilization of that invention, to achieve practical application. How many patents you own does not matter. How many licensing deals you have done doesn’t matter. How much money you have made doesn’t matter. How many companies you have started doesn’t matter. Use with public benefits on reasonable terms–that matters.
Now consider the information that is put forward for the success of Bayh-Dole. We have the AUTM licensing survey. But that survey does not distinguish subject inventions from other inventions and does not even distinguish patentable inventions from other “inventions”–anything that generates at least $1,000 a year is considered a “commercial” license. Finally, there’s nothing in the AUTM licensing survey that tracks practical application–use.
A university licensing program could be the best at meeting the objectives of Bayh-Dole–for every invention acquired, use with public benefits on reasonable terms–and have really *low* numbers for inventions acquired, patents obtained, licenses granted, and income received. A university could be selective–obtaining only those inventions that would benefit from patenting, making those inventions available on the condition that those acquiring rights also were diligent in using the invention and in making benefits of that use available to the public on reasonable terms. For that, a university might license royalty-free for use. Imagine that–turning the whole royalty scheme upside down–if you use our inventions and make the benefits of that use available to the public on reasonable terms, then you don’t owe us a cent. But if you don’t use the invention, or you use it without public benefits, or you don’t make those benefits available on reasonable terms, then you owe us a stinking pile of gold. Pay for nonuse. Pay for private use. Pay for unreasonable use. Pay for holding rights just in case. But do what’s indicated in the public interest, and doing that is the consideration for the license.
In such a case, the university’s licensing income would reflect the number of patent licensees that had not achieved practical application. The more money, the less public benefit, the poorer the Bayh-Dole performance.
In the original Bayh-Dole, there were limits placed on exclusive licensing by nonprofit organizations. An exclusive license, other than to a small company, could run for no more than five years from the date of first commercial sale or use or eight years total, whichever came first. The obligation to report on the utilization of inventions, then, in the original version of Bayh-Dole, was linked to exclusive licensing. If a nonprofit organization granted an exclusive license (and not to a small company), then a clock started on the term of that license, and commercial sale or use might shorten the term of the exclusive license. And the overwhelming target of Bayh-Dole was research conducted by nonprofit organizations–universities, their research foundations, and nonprofit contract research organizations. Thus, it was important to have at least these two data points–date of exclusive license with a non-small company, and date of first commercial sale or use. These dates both marked progress toward practical application and set limits on the monopoly rights that a nonprofit could preserve and convey in a subject invention. Beyond these limits lay non-exclusive licensing, and, as Bayh-Dole puts it, “free competition and enterprise.”
The 1984 amendments wipe out the restriction on the term of exclusive licensing, leaving only a protectionist requirement that any owner of a subject invention must, for an exclusive license to use or sell in the U.S., also require product to be “substantially manufactured” in the U.S. But that requirement has its own set of compliance issues and does not depend either on the date of exclusive license or date of first commercial sale or use. Thus, after the 1984 amendments, the reporting obligation in Bayh-Dole carried into the standard patent rights clause is essentially meaningless. The status of development of a subject invention does not matter, because that information is not public; information on licensing and commercial use and sale does not matter, because there are no limitations on the term of exclusive licenses granted by nonprofits to non-small companies. One might predict that to reduce administrative bother, the reporting on status of inventions obligation itself will also eventually be eliminated.
In one sense, then, there are no data on the performance of Bayh-Dole–or what information that exists is a government secret. Academics who use the AUTM survey data as a proxy are digging in what already amounts to a polemic. They cannot separate out subject inventions from other inventions, and most importantly, they cannot find any information on practical application. At best, selected “success” stories, put forward to create an impression that these represent many more such transactions. There’s nothing to show, for a given year’s worth of reported subject inventions, when each of those inventions achieved practical application as defined by Bayh-Dole. Thus, one cannot even compare inventions reported in a year with patents obtained in that year with licenses granted in that year with startups in that year or income in that year. These are almost but not entirely unrelated stacks of information. At best, they report a level of activity–an office handled all those transactions in one year. But increasing activity with poor results means inefficiency, ineffectiveness, and waste. It’s just that universities withhold the results so one cannot see anything other than activity itself. IP bustle, it would appear, is what universities and AUTM put forward as the objective of Bayh-Dole.
It’s just that IP bustle is not in Bayh-Dole. Yes, one might argue that an effect of Bayh-Dole was to induce universities to “invest” more money in “technology transfer.” But that’s not a stated purpose of Bayh-Dole, even if it was given as a motive by patent brokers who participated in pushing the law through. For them, the law was about enhancing the livelihood for patent brokers under the argument that what’s good for patent brokers is good for America. No wonder Howard Bremer was seen as an altruistic leader–he helped to create Bayh-Dole so there would be livelihoods for many patent brokers for years to come. In its way, that’s so heartening.
But we know from the few voluntary disclosures from universities that the overall licensing rates for inventions is rather low. At Stanford, a highly regarded program, only 1 in 5 inventions gets licensed. At University of California, one of the largest programs, only 1 in 200 inventions ever becomes a commercial product. There’s no information on how many of these inventions are subject inventions, but the indications are that very few inventions acquired by universities meet the definition of practical application set out by Bayh-Dole.
We can look at the data that is out there, then, from a different perspective. The number of inventions that are patented is a measure of the invention flow that has been diverted from private hands (the inventors’ personal property) to institutional hands (the universities and their research foundations), and diverted from the public domain (that is, held as proprietary property rather than published, disclosed, and dedicated to science and to the public). Thus, the growth in university patenting, without any other context, is a direct measure of the loss of private control and diminishment of the public domain.
The burden is on the universities to supply additional context to overturn the indication of this metric. The argument universities must make is that whatever loss of private control or diminishment of the public domain there may be is more than made up for by better figures on practical application than are obtained without such institutional control. So far, nothing of the sort has been produced. It’s clear that an increase in patenting marks a decrease in key research findings entering the public domain. It will take significant work to show that an increase in patenting of such findings is a good thing. It will take significant work to show that an increase in university IP bustle is a good thing.
Similarly, the growth in exclusive licensing is a measure of the preservation of monopolies. This may be a difficult concept for you. I will go slow. A patent represents a government grant of a private monopoly to an inventor in exchange for publishing his or her invention. That monopoly has a limited term and limited scope of property rights to exclude others from making, using, or selling the claimed invention. An ordinary inventor of an ordinary invention with an ordinary patent has no obligation to make, or use, or sell the invention, and no obligation to assign, or sell, or license the invention to others. Bayh-Dole shows up as a gloss on this situation. Bayh-Dole requires federal agencies to place (and especially not to place) conditions on what inventors can do with inventions made with federal support–and especially inventors at nonprofit organizations.
If an inventor assigns their patent to their employer–a federal contractor–then the government granted monopoly is preserved in the transfer, along with the Bayh-Dole and standard patent rights clause requirements on that monopoly and the obligations of the patent owner. Under Bayh-Dole, an owner of a patent on a subject invention must use the patent system to promote the use of the invention–the practical application, the use with public benefits on reasonable terms. If reasonable terms meant “any legally permissible terms” there would be no reason to include the limitation in federal law–“reasonable terms” must mean something other than “whatever terms a patent owner agrees to.”
Getting at a standard for reasonable terms is what the current debate on the price of patent drugs made with federal support is all about. Here, we can state at least that price is a component of reasonable terms–“reasonable terms” is not whatever a patent owner agrees to as part of a license transaction–it is a matter of how the benefits of licensed invention use are made available to the public. We can also state at least that this price is not “whatever the market will bear” but something less than that, either because the length of monopoly pricing is limited with regard to the full term of the patent or because the licensee has agreed to limit price or other terms, even if a monopoly would provide an opportunity to raise prices or withhold access.
We might think of price gouging for needed supplies after a natural disaster–a hurricane or snowstorm. We think: there is a public need that now changes the moral character of a “market” and raising prices to take advantage of the desperate or to play only to the wealthy is immoral. Bayh-Dole sets up the proposition that research conducted at nonprofits with federal funding is in this category of products. The natural disaster is the lack of beneficial knowledge, and universities commit to assisting faculty and student investigators discover and establish new, beneficial knowledge. The idea of public benefit on reasonable terms then turns on whether any owner of the monopoly has a right to withhold all use or to use without public benefit or to use with public benefit but on unreasonable terms–that is on terms that otherwise would be legal (we are not dealing with antitrust) but in the circumstances are immoral, contrary to reason, against the purposes of the federal funding, unreasonable.
Now consider what happens when a university, having obtained the monopoly right plus the Bayh-Dole limitations on that monopoly right, conveys that monopoly right to a company. Under Bayh-Dole, the university cannot assign the monopoly right without also assigning the Bayh-Dole requirements under the standard patent rights clause applicable to nonprofits. If a university does assign a patent on a subject invention, it transfers both the monopoly and the requirements–the entire right, title, and interest in the patent includes the Bayh-Dole and standard patent rights clause right and interest in the subject invention.
But if a university grants an exclusive license, then there’s a question whether the exclusive license is in fact an assignment. Some people don’t want to consider this point, and plenty of university lawyers will object out of hand. But then, it is their competence that’s in question, so what would you expect of them? And frankly, a lot of university lawyers have little training or experience in intellectual property law or management. They are good at avoiding disputes and setting up policy so that they win disputes. They are not so good at intellectual property, especially intellectual property to be managed in the public interest rather than so that the institution can win disputes over intellectual property.
So here’s what universities aim to do: they license exclusively, but try to split the obligations between the university and the licensee, so the exclusive licensee does not have the same public obligations that the university has under Bayh-Dole, but the exclusive licensee has substantially all the monopoly rights of the patent–and so can sub-license and can enforce the patent. In essence, the university grants a what amounts to a private patent based on the government patent that the university has acquired. The exclusive license operates as a shadow patent.
For state universities, their licensing programs, where exclusive in this way, are in effect state-funded patent offices, re-issuing patents under an exclusive license, but now granted as it were to selected companies rather than to inventors. The overall effect is to give university “middlemen” the responsibility to decide who should receive U.S. patents. While inventors should receive their patents as a matter of federal law, Bayh-Dole has been exploited to create an institutional patent system dedicated to ensuring that inventors never own their inventions and institutions re-issue patents on their inventions to companies. If it were not for the horrible uncertainty regarding which company will receive the favor of an exclusive license (and often, these exclusive licenses are assignments), it would be simple enough to describe Bayh-Dole as creating hundreds of private patent offices that issue patents on federally supported inventions to the companies of their choice. The government creates the patent rights, and then encourages private patent offices to acquire the patent rights using whatever means necessary, and then cedes control of the patent rights to these private patent offices, which then re-issue the patents as “exclusive licenses” to whomever pleases them, based on, say, ability to pay, or likelihood of additional research funding, or political connections, or happy desperation.
Bayh-Dole works by adding public covenants to ordinary patents. Those public covenants are set forth in 35 USC 200, in the definition of subject invention at 35 USC 201, and in the standard patent rights clause at 37 CFR 401.14(a). These public covenants reach to inventors through the written agreement requirement at 37 CFR 401.14(a)(f)(2). The least restrictive public covenants are directed at inventors. Next come small businesses. Most restricted of all are nonprofits. Nonprofits are in particular restricted from assigning subject inventions (and patents on subject inventions) without also assigning the public covenants in the standard patent rights clause applicable to nonprofits.
The point of all this is that Bayh-Dole insists that the public covenants follow the patent monopoly. Only by breaking up the monopoly does one become free not to pass on the public covenants to others. Thus, in a non-exclusive license, there is no requirement for U.S. manufacture. Nor does the federal government have any march-in rights. There’s no requirement for including a statement of government interest in patent applications (because the application is a matter for the patent owner, not for any licensee). Similarly, there’s no requirement to prefer small businesses–since again that is a requirement only on nonprofit owners of patents on subject inventions. If an exclusive license involves any of these requirements, then it is operating as an assignment, conveying rights and obligations that only an owner (if only a co-owner) of a patent may have.
The only requirements in Bayh-Dole that apply to any licensee are those having to do with reporting on utilization:
The Contractor agrees to submit on request periodic reports no more frequently than annually on the utilization of a subject invention or on efforts at obtaining such utilization that are being made by the contractor or its licensees or assignees.
We might think that the word “exclusive” has been dropped ahead of “licensees”–and perhaps in a future amended Bayh-Dole this will be the case. As it is, there is a significant difference between the reporting that a non-exclusive licensee is obligated to do and the reporting of the contractor, any exclusive licensee, or assignee. If a contractor uses an invention and meets the definition of practical application, then the contractor’s reporting obligation is straightforward–just report the fact. If a contractor assigns (or exclusively licenses, but really assigns) the patent, however, then the contractor must also assign this reporting requirement. Before Bayh-Dole was amended in 1984, this reporting requirement made clear sense–a licensor in privately re-issuing a patent on a subject invention would have had to transfer the obligation to the assignee to report, or in the case of an exclusive license would have had to track the date of the exclusive license and the date of first commercial sale or use, in order to comply with the term restrictions on exclusive licenses.
But if a contractor grants non-exclusive licenses, then the very act of granting the non-exclusive licenses is itself the effort at obtaining the “utilization of a subject invention.” That is, non-exclusive licenses, while not establishing practical application, meet the reporting requirement. There is no need to chase down non-exclusive licensees to determine if they are using the invention with benefits available to the public on reasonable terms–the invention is made available to the public through the non-exclusive licensing program, and the terms of that availability (and any benefit) are the subject of a determination of what’s “reasonable.” There’s all sorts of interesting possibility here, for another time. For instance, a royalty-free non-exclusive license might not, in all situations, be “reasonable.” Funny that.
You might then see that when a university licensor breaks up the monopoly in a subject invention, the university does what Bayh-Dole seeks. But when a university licensor preserves the monopoly in its substantial rights by assigning the monopoly to another (or exclusively licensing it), then it merely profits (if it does) from passing on the monopoly, not from getting to the things that Bayh-Dole seeks–that is, practical application. Re-issuing a government patent as a private patent right can be a decent source of income, but it is not income that arises in recognition of fulfilling the objectives of Bayh-Dole. It is merely playing the factor in a chain of patent hot-potato. Universities working this scheme of monopoly preservation step in to separate patent rights from inventors and get paid to convey these monopoly rights exclusively to a private player, who then takes on the obligations of the public covenant in Bayh-Dole and the standard patent rights clause.
The thing universities deploying exclusive licensing programs then try to get away with is not conveying all the Bayh-Dole obligations in the standard patent rights clause along with an exclusive license or assignment. If an exclusive license is not an assignment–that is, it does not convey the substantial rights of the patent owner (make, use, and sell; enforce the patent and have the exclusive right to sublicense)–then the exclusive licensee has only the obligation to report that all who touch a subject invention must do–report on their use of the invention or efforts to get others to use it. But otherwise, if an exclusive license is really an assignment, then all the university has done, really, is to extract money (or other concessions) for the service of passing along the monopoly to a private entity. That’s not the objective of Bayh-Dole, and if a nonprofit contractor does not pass on the requirements of the standard patent rights clause, then the standard patent rights clause is violated.
It is no mere technicality whether an exclusive license is really an assignment. The difference has everything to do with whether a contractor is meeting the obligations of Bayh-Dole–by splitting up the monopoly patent rights to promote use–or whether the contractor is simply profiting from selling (under whatever cover) a patent right in a subject invention. For small companies, there is no restriction in the standard patent rights clause on selling patents on subject inventions (and that would happen any time a small company got acquired by another company). But small companies (and their assignees) are still obligated by the standard patent rights clause to respond to requests by federal agencies for reports on the utilization of subject inventions or efforts “at obtaining” utilization–either use it or license it for use. The difference is that a small company can satisfy the utilization requirement by selling the patent. The same is true of any inventor of a subject invention who retains ownership–there is no restriction on assignment. Only nonprofits must assign the standard patent rights clause obligations in any assignment of a subject invention.
Thus, if we look at the number of exclusive licenses reported universities in their annual reports and AUTM licensing survey, the total represents the re-issuing of federal patents as private monopolies. If those private monopolies in turn fail to achieve practical application, then the number of exclusive licenses represents an even greater barrier to public benefit than if the university had not licensed at all. If a university has not licensed, then at least at some future time, the university would be free to grant non-exclusive licenses. But if a university has passed on the monopoly in its substantial rights, then the university cannot later grant non-exclusive licenses without terminating the exclusive license, or at least its exclusivity–and that might well require a lengthy fight, even if the university has some sort of termination clause in the agreement directed at nonuse.
There are any number of ways to screw up a termination clause either in the drafting or in practice. A university that, for instance, tolerates a breach for nonuse that would otherwise require termination of the agreement or exclusivity, and does not act on that breach, may find itself facing a laches argument that the licensee has relied on the university’s inaction. Consider the Xtandi exclusive license from UCLA to Medivation. Over 160 compounds are covered by that license, but only one compound resulted in a commercial product. Could UCLA cancel the deal for the other compounds that haven’t been used but now are not available to anyone else? I expect there would be a fight. Nothing is as simple as the text on the page might indicate.
Thus, even the number of exclusive licenses granted by universities, without some additional context, argues for the extent to which subject inventions are not available to the public on reasonable terms. An exclusive license that preserves the monopoly does not comply with Bayh-Dole. Bayh-Dole merely permits it, provided the public covenants continue to apply. Only when a license results in practical application–utilization of the invention–is Bayh-Dole’s primary objective met. The evidence we have, far from showing the success of Bayh-Dole, points to its failure: that many subject inventions claimed by universities are not used with public benefits; that university licensing programs are profitable, largely, from the re-issue of federal patents as private monopolies, and from the once-a-decade or longer “big hit” product, often then sold to the public at monopoly prices, and not on “reasonable” terms.
Back in the days of the Institutional Patent Agreement program, when Congress looked at how universities were doing with subject inventions, they found that despite the Kennedy and Nixon executive patent policies and despite the gestures in the master agreements regarding non-exclusive licenses and the like, the nonprofit patent brokers were granting exclusive licenses and disregarding most everything else, and the patent brokers were determined that exclusive licenses should run for the entire term of the patent, thereby maximizing the possible income to the institution and relieving patent brokers of the obligation to operate a non-exclusive licensing program, especially one offering royalty-free licenses to all. Bayh-Dole is more of the same. The same patent brokers who worked the IPA program to ignore the idea of a public covenant in patents on federally supported inventions have done the same thing with Bayh-Dole. This is their legacy–to create a viable private system of re-issuing federally created monopolies for institutional gain, stripping whenever possible the public covenants that run with the patents through Bayh-Dole and through the standard patent rights clause. Universities do not report their utilization numbers because those numbers stink. Those numbers would show that Bayh-Dole has failed in its essential purpose.
Bayh-Dole has propped up the livelihoods of university administrators, patent brokers, factors, middlemen, selectors of who should benefit from federal research funding. For that, universities receive licensing funds. The rest of Bayh-Dole is inoperative. There’s no compliance (but for reporting and statements in patent applications). The government exercises no oversight. What reporting that gets done is a federal secret. As far as I know, there’s never been a challenge to a university’s small business licensing practices. There’s never been a march-in for nonuse or unreasonable terms. There’s never been a government use of its license to practice and have practiced a subject invention resulting in a meaningful public benefit. The IPA program made a reasonable gesture, to permit others to help the federal government accomplish its public mission. But patent brokers exploited the IPA program, didn’t do any better than the federal government, and betrayed government policy objectives when it came to working with new discoveries in the field of medicinal chemistry. So the IPA program was shut down. And Bayh-Dole, a new rough beast, was born in its place, open to the same exploitation and more.
The metrics we have indicate that Bayh-Dole has failed in its purpose–that of promoting practical application. Instead, we have created a market for re-issuing monopolies, and that market operates by ignoring the substantive requirements of Bayh-Dole. Bayh-Dole requires using the patent system to promote use, not exclude use; and for nonprofits, Bayh-Dole requires assigning the nonprofit obligations when assigning patents on subject inventions, and that means that the assignee must use income just as a nonprofit would. Neither of these requirements is observed or enforced.
Bayh-Dole is used as if patents on subject inventions are ordinary patents, and that institutional IP bustle is the measure of Bayh-Dole’s success. The administrative elements of Bayh-Dole that are left are virtually meaningless in the broader scheme of things–reporting inventions to the federal government, electing to retain title, filing patent applications with statements of government interest, granting the government a license it will never use, fussing over march-in rights the government will never exercise, fussing with requiring U.S. manufacture that the government can simply waive, reporting on use in secret.
Given what practice under Bayh-Dole has come, one might as well repudiate the idea of a public covenant and make Bayh-Dole simply prevent federal agencies from having any claim on subject inventions but for a government license. The government wouldn’t even need a license until it had decided to practice an invention and then it could consider whether there was federal support for the invention. It never uses its licenses anyway. No reporting, no limits, no licenses, no fuss. At least we would save the waste energy of all the useless paperwork.
Current university practice under Bayh-Dole repudiates the idea that federally supported inventions should be managed differently from ordinary inventions. The question for public policy is whether such repudiation is justifiable, desirable, and results in the things that we should want for ourselves, the public, in supporting university research with $40b a year.