Patents the government issues to itself
The attributes of ordinary patents make little sense in the context of the federal government issuing to itself a patent. The government has no profit motive from the patent system. The U.S. patent system was set up to preclude the government from taxing technological inventions as a source of revenue (maybe the government should revisit that–the government makes millions from patenting fees alone–imagine what the government’s income would be with a 5% tax on the profits from patented products–for licensed patents, the government could double dip and tax the seller first and then the licensor as well. Given the patent marking requirement, it’s a wonder this hasn’t been seriously considered). If the patent system was set up so that the government did not have a financial interest in the use of any given patent, then how could it be that when the government issues a patent to itself, this principle should change? (It has changed, of course, but that change came with Bayh-Dole, thirty-five years later.)
Patents issued by the government to itself should be used to promote use rather than prevent use, should be used without an immediate profit motive by the government, and in terms of public purpose should give access to all and not play favorites, should break up monopolies where those monopolies held onto obsolete or overly expensive products, and should selectively invite private initiatives to participate in the development of new things, but with only such exclusive rights as needed to justify the private investment–meaning, the prospect of a reasonable return that justifies the commitment of the money and recognizes the risks undertaken and the other opportunities set aside.
If the government issues a patent to itself and the licenses that patent exclusively (or even assigns that patent) it is in essence re-issuing the patent, but now not to the inventor (or even the inventor’s employer, if that would matter), but to a favorite, to someone chosen by the government to undertake whatever dealings the government believes to be in the public interest–use the invention, develop the invention for public benefit, break up the patent monopoly faster and better than the government could, and once the work is done, release the exclusive position, the patent property monopoly, and get back to work in a better, richer public commons.
If you want trails through a trackless land, and there is a pressing need for trails and not enough time and money to do things yourself, then you might allow others to build some of those trails, and charge tolls for a time to recover their costs, since having trails is better than not having trails–so long as once the private expense of building the trails has been recovered (and a reasonable bit more, so trail building is as good if not a little bit better than keeping the money in a bank), the tolls go away and everyone has access to the trail (we skip over who ought to maintain the trail then–sigh–I’m just trying to illustrate the thinking). If you follow the thinking, you will see things that otherwise are lost to you.
From Kennedy to Bayh-Dole
Thus we arrive at the Kennedy patent policy of 1963, which lays out executive branch policy along just these lines. The government will procure patent rights in its research contracts except in a couple of narrow instances. The government will use its patent rights to provide access to all and only use its self-issued monopoly powers when access for all depends on private initiative, so long as that monopoly is carefully managed to serve its purpose and not become a monster. Thus, the public covenant–reserve an unlimited government license, promote use, require reasonable (not monopoly) pricing, break the monopoly when it has served its limited public purpose (non-exclusive licensing, time limit on exclusivity).
It is in this context that the IPA program was revived by the NIH in 1968, on the heels of the Harbridge House report, which documented the spat the PHS had been having over medicinal chemistry research, which in turn had been incited by the Wisconsin Alumni Research Foundation licensing a drug candidate discovered (in part) with federal support but exclusively to a Swiss pharmaceutical without any government license (or any other public covenant). The PHS had intervened (litigation!) and got a share of the patent, licensed that share non-exclusively and broke the pharma’s monopoly. In protest, pharma boycotted PHS research in medicinal chemistry and refused to screen compounds. A blockade of science in the service of monopoly.
The IPA program was designed to break the PHS policy, end-run the Kennedy patent policy, end the pharma boycott, and provide a pathway for the creation of new pharma monopolies on compounds discovered with federal support. In this effort, universities and their affiliated foundations were the primary agents, starting with WARF, which had incited the problem and which would continue to be the nemesis of the idea of a patent public covenant through Bayh-Dole to the present. (Or, rather, WARF created a competing version of public covenant that focused on the public benefit arising from profits on the exploitation of patent rights rather than on access to the underlying inventions–access was a side effect, good for advertising but not a consideration in licensing.)
Here is how the IPA program dealt with patent rights. Outside of any federal funding agreement, the NIH and a university could agree up front regarding the disposition of inventions. We have to be clear on the sequence and details here, because these were clever, careful folks navigating a tangle of policies and regulations to get what they wanted without creating a ruckus about it. If the university (or other nonprofit) decided to file a patent application on an NIH-funded invention, and accepted the public covenant with regard to the use of any patent that issued, then the university was released from the obligation of assigning the invention to the federal government under the standard regulatory terms of any NIH funding agreement. To accomplish this result, the IPA required that the university require its employee-inventors to assign their inventions to the university whenever the university made the decision to patent.
Keep this sequence in mind. IPA requires a patent agreement between the university and its potential inventors. That patent agreement has a conditional element: the university must decide to file a patent application on an invention, and then has standing under the IPA to require under its patent agreement with inventors that they assign their rights. The IPA does not allow the university to require assignment upfront of all such inventions, such as at the time of reporting the invention or (with the clever use of magical present assignment language) even at the time the invention is made, even before being reported. The IPA contemplates that the inventor will assign inventions when a patent application has been prepared–and thus at the time that the application meets the inventor’s approval for specification and claims–not before. The patent application defines the invention; the assignment is directed to exactly those things that the university has chosen to claim, and not to an invention generally, and not to an invention that “may not be patentable.” It’s a specific, defined invention and time for assignment.
Under the IPA, the government gets the benefit of its public covenant–a broad government license to practice and have practiced, a default apparatus of non-exclusive licensing, a limitation on the term of any exclusive license (with possibility of extension by petition), and the opportunity to require licensing for any government purpose for which the university’s own licensing practices (or failure to practice) might interfere. All of these things are established by the IPA, and the IPA in turn depends on the executive branch policy that authorizes such agreements as part of government contracting (even if the IPA works purposefully outside that authorization). With the IPA in place, the government still has a deal with every investigator, by regulation, to assign inventions to the government, but the IPA allows a university to step in when it wants to patent an invention and require the inventor to assign the invention, as represented by a patent application, to the university (or its designee). The IPA, in effect, gives the university the right to pre-empt the government’s own federal contract requirement that inventors assign to the government, subject to the public covenant.
The cleverness in all this–or the swindle, depending on your frame of mind–was that the IPA depended not only on the university’s compliance with the public covenant but also with the willingness of the NIH patent officers to enforce the IPA. It turns out that these officers didn’t have any interest in enforcing the IPA’s public covenant. Talk about “deep state.” Instead, they largely ignored the non-exclusive apparatus and universities licensed subject inventions routinely exclusively, and when it came down to it, for the life of the patents involved. The IPA enabled patents on federally supported research to create private monopolies with all the benefits that a patent owner might find in them to make money. Practical application became a matter of diligence in an exclusive license agreement, not a commitment based on federal contract. And as a matter of diligence, practical application was just one possible outcome that provided satisfactory value in the contract. If the company provided equity as consideration, and was later acquired for millions of dollars, then the compensation represented by the value of the university’s share of equity was more than enough to compensate for a lack of practical application in any licensed invention. What mattered was value from the patent right, not necessarily public benefit from the use of any variant of the underlying inventions.
The IPA was a rabbit path through the thorny briers of an apparatus of public covenant. What you see from the outside is a bush of protections for the public interest, and from the practice side a sweet, straight path to private monopolies, so long as you stay on the path and don’t shake the bushes to draw attention to what you are doing. The IPA worked for a decade and then it was caught out when Latker tried to make it government wide. A Congressional committee blocked it, and in reviewing practices found that all the licenses were exclusive, that the public covenant wasn’t operating as expected, and worse, the exclusive licensees were getting such good deals one couldn’t even justify them in terms of royalty payments to support the universities. It was patent brokers run amok in the service of pharmaceutical monopolies. Yes, drugs were on the market. But no, they weren’t reasonably priced, there wasn’t competition, and research was disrupted. So the public covenant apparatus had failed, and public money for research was just a big happy subsidy to the pharmaceutical industry, for which research foundations (mostly) got enough of the action to make a good professional livelihood of it and to satisfy university administrators that there was good money to be made eventually, but it was “high risk” stuff and might take a few years, or decades.
Bayh-Dole was the result of shutting down the IPA program. Bayh-Dole took a different approach, to avoid PHS policies (and the authority of the Surgeon General or anyone else), and even to prevent any executive branch policy from interfering. Thus Bayh-Dole was designed as a statute that dictated executive branch policy on inventions made in federal funding agreements for research with nonprofits (including universities). Small business was tossed in as well, for political expediency. Further, Bayh-Dole was placed not with federal laws on research funding, where it should have been put, but rather in federal patent law, where it made no sense but it turns out made significant changes to patent law.