We are working through the University of Pittsburgh’s account of the Bayh-Dole Act since AUTM has called it out as worth reading. We need to pause and consider some real history to work out of our imaginations the fake history implied by Pittsburgh’s account.
The government’s position on the use of patents is captured by this Congressional report by the National Patent Commission in 1945:
The Commission recommends that the Government as a general rule continue to pursue the historic policy of not exercising the right to exclude conferred by patents which it owns; of not attempting to exclude its own citizens from engaging in any enterprise; of not seeking to derive revenue from patents, and of not undertaking control by means of patents. Inventions covered by patents owned by the Government should be available for commercial and industrial exploitation by anyone, with, however, the recourse open to the Government to take different action in exceptional cases.
The National Patent Commission then goes on to discuss exceptional cases, such as where general access is not sufficient to attract the investment necessary to move an invention from its initial condition to a form from which the public can benefit. Much of what then becomes a quest for a “uniform” patent policy is one of debating whether the default should be open innovation, with some few things controlled by patent by the government (and its contractors), or whether the default should be monopolies in inventions, with only those things judged absolutely worthless permitted to fall into the public domain. The Kennedy patent policy describes the conditions under which the government should allow contractors to pursue monopoly positions–namely where private risk capital is necessary to bring an invention to the point of practical application, and then for only so long as is reasonable for that contractor to recover that risk capital from the practical application of the invention.
The implied argument around those 28,000 government patents is that the default is wrong, that these inventions are wasting assets because, apparently, speculative investors would have purchased a monopoly interest in these patents (by exclusive license or assignment), allocated risk capital, and created valuable products. Bayh-Dole’s “uniform” policy replaces the default and very uniform government policy that government-funded inventions should be open to U.S. citizens and companies, but for exceptional circumstances. In its place we get a uniform policy of allowing universities to speculate on the future value of monopolies on inventions made with federal support. Unlike inventions held by commercial contractors who already have the “competence” and have “an established non-governmental commercial position” (to use the terminology of the Kennedy patent policy), Bayh-Dole focus on university and non-profits who generally will not have either of these qualities–they will lack both technical competence to produce product and will not have an established commercial position from which to reason about development.
In short, Bayh-Dole creates about the worst possible uniform alternative to the uniform policy established by President Kennedy. The Harbridge House report in 1968 demonstrated how poorly the universities were doing with their patent management compared to other contractors. When contractors had experience and ownership, they used inventions made with federal support about 25% of the time. And a majority of these contractors were using the invention within three years of the date of disclosure–often before any patent ever issued. Here’s the key table from the Harbridge House report:
The worst possible outcomes came from inventions owned by organizations without experience and licensed to other organizations. Here’s the key table from the report.
For universities in particular, Harbridge House’s sample showed no more than 10% were “utilized” (See Figure IV-1). An invention management firm estimated that less than 5% of its inventions were licensed, and only 0.5% of its inventions were commercially profitable. Universities slot in there at less than half the use rate of commercial contractors. Licensing–“contractor has no title”–performs at about half the rate of assignment. Think about that. Universities have adopted a licensing approach rather than an assignment approach, and point to Bayh-Dole as the reason, but the evidence indicates that an assignment approach–place ownership of the patent with a commercial concern that has prior experience–is by far the much better way to go.
Far from persuading Congress on the best way to realize technology commercialization, university patent brokers successfully made Congress think that the best way to do things was the worst way. From the university patent broker point of view, things couldn’t have gotten much better, and for good reason, since they had made it the law that the worst way to do things was now the default way to do things. It must have felt magical to realize that a law had just enfranchised a livelihood–patent brokering university-supplied inventions–that had no compelling justification in the broader scheme of federal goals for national innovation through the results of basic research. Perhaps the only thing that can be said of Bayh-Dole in its favor is that it did not make compulsory the assignment of inventions from inventors to the universities that hosted their work–and of course this one decent thing in the law is just what university patent brokers in a very big heap attempted to eradicate, only to be stopped by the Supreme Court in Stanford v Roche. Not to be deterred, they press on with an assignment clause requirement they hope NIST will adopt, despite the Supreme Court ruling.
The Harbridge House report concluded that a uniform title policy was not indicated: “a balancing of government objectives appears necessary to ensure that the net effect of the patent policy promotes the government’s overall goals.” The only way that Bayh-Dole is successful is if the government’s overall goal has been to make jobs for university patent brokers.
At the time, the government was not willing to grant licenses to the private sector.
This is pure bunk. First, with most federally owned patents, there was no need for licenses–the government refused to exercise its right under patent to exclude its own citizens. National Patent Commission, again:
The general policy of the Government in the past has been not to exclude its own citizens from engaging in any commercial or industrial activities; it has not attempted to exercise the right to exclude conferred by the patents which it owns. As a rule such patents have been open to licensing to anyone who applied, without payment of royalty or other charge and mainly on nominal conditions. Indeed, patents owned by the Government have been open to use by anyone, with or without an explicit license.
The lack of an “explicit” license from the Government had nothing to do with whether an invention was being used. Nor would it have been generally true that where there was no use, the reason was that a monopoly had not been first created, or that this monopoly would have been better managed in the public interest by private speculators rather than a government agency. All that is fool’s talk.
Under the Kennedy patent policy,
Government-owned patents shall be made available and the technological advances covered thereby brought into being in the shortest time possible through dedication or licensing and shall be listed in official government publications or otherwise.
That is–license or don’t license. Do whatever gets stuff “brought into being” or developed “to the point of practical application” as quickly as possible. Not for the most profit. We might remind ourselves of Kennedy’s inaugural challenge: “Ask not what your country can do for you–ask what you can do for your country.” This challenge is embodied in the Kennedy patent policy. If contractors are to take up ownership for a time of inventions made with federal support, for federal purposes, then the question for those contractors is how they contribute to the country’s goals for those inventions. The answer to this challenge was not “to make as much money as possible by excluding all other citizens for the life of the patent.” It was, to develop inventions to the point of practical application and enjoy a limited monopoly in non-governmental markets for three years from the issue date of the patent to recover some or all of the risk capital necessary to develop the invention. After that, we again compete based on quality, price, service, and brand. Bayh-Dole destroyed this ethos, for reasons that the folks at the University of Pittsburgh must somehow find virtuous.
There were two areas that the Harbridge House report found where an agency’s implementation of the Kennedy patent policy had caused problems–medicinal chemistry and some areas of research supported by the Department of the Interior, such as desalinization technology. In medicinal chemistry, HEW moved in on territory that had been staked out by the growing drug industry. HEW insisted on interpreting the Kennedy patent policy to require non-exclusive licensing of inventions in medicinal chemistry made with HEW support. The drug companies organized a boycott. It was not that the government would not offer licenses–it was that the government would not grant monopoly licenses. NIH launched the IPA program to grant monopoly licenses anyway, by having universities do the deals. When the NIH tried to make the IPA program government wide in an effort legitimize the monopoly business it was supporting with the drug companies via the universities, that effort was blocked. The NIH’s second try was Bayh-Dole, put into federal patent law to prevent the executive branch from easily reverting to a more flexible practice.
The only bits of federal research contracting that we might consider for Bayh-Dole treatment, if we follow the Harbridge House report, were those where
- the invention is commercially oriented but requires substantial private development to perfect it,
- applies to a small market, or
- is in a field occupied by patent sensitive firms and its market potential is not alone sufficient to bring about utilization.
Inventions in this category may arise with any agency and may have had only limited government development toward a commercial application.
In other words, what should have been made “uniform” in federal research contracting was a breakout for these inventions. What would be necessary is a rapid determination that an invention is “commercially oriented” but requires “substantial private development” or is in a “small market” that cannot support multiple providers anyway or is in a field of “patent sensitive firms” and lacks sufficient “market potential” to motivate these firms. Those might be difficult determinations to make in a timely fashion. The alternative, however, to treat all inventions as if these conditions are necessarily true by default does a great disservice to the majority of university-hosted inventions made in basic research, which have their first application as research tools or as methods embedded in software or involve compounds, biological materials, and techniques readily adopted with little additional development work.
The problem for HEW’s medicinal chemistry program was that it operated in a field of patent-sensitive firms and failed to provide the follow-on screening of candidate compounds that would have reduced the private development requirements, aggregated additional IP rights into a single platform, which then could have been made available to industry to further test and manufacture in competitive formulations. We might say that Bayh-Dole is a bad response to a poorly conceived program of government-supported research–funding “basic” research but failing to be responsible (or even care much) about the findings. The NIH program in medicinal chemistry stales in comparison to the government work done in support of treatments for malaria and leukemia.
Thus, even in this problem area of medicinal chemistry, the problem was not that the government would not offer licenses. The problem was that the government would not offer monopoly licenses–and the reason that monopoly licenses were indicated was that the government was not willing to subsidize even the initial screenings for activity and efficacy. That this problem could be made into a general case applicable to all federally supported inventions is just poor reasoning and self-interest wrapped up into a political turd and made to smell like a pork roast of technological development to maintain a dominant global position.