Howard Forman, a long-time patent attorney in the chemical industry turned federal employee, introduced the 28,000 unused federal patents meme into Bayh-Dole rhetoric in his congressional testimony in 1976. Senator Bayh includes Forman’s meme in his introduction of S. 414, which eventually, ahem, is passed and becomes Bayh-Dole.
Bayh has argued that the monopoly meme controls the use of federally supported inventions. If a company has no patent monopoly, then it won’t invest to develop any invention, no matter how useful the invention might be. According to the monopoly meme, inventions available to all will be used by none. Bayh exploits the monopoly meme to argue that, necessarily, if the federal government makes inventions available to all, they aren’t used and the public fails to see a “return” on the federal government’s research expenditures. History shows Bayh was wrong, but no matter. The essence of Bayh’s use of the monopoly meme is that any invention made with federal support must end up as a patent monopoly in the hands of a company.
But Bayh does not propose simply changing federal patent law to vest ownership of patentable inventions with companies. There’s problems with that. What happens when the federal government funds research at a university? Bayh wants faculty inventors stripped of their rights–he makes that clear in his amicus brief in Stanford v Roche many years later. But who should choose which company should receive the patent monopoly franchise? That’s the pivot point for Bayh-Dole–contractors, not the federal government–should have the first right to choose their favorite company.
One can see that Bayh dismantles the core of the US patent system–that the federal government is authorized to reserve to inventors for limited times exclusive rights to their inventions. The inventor gains the patent right as a result of (1) an application for the right and (2) a review that grants the right based on the objective merit of the invention. Bayh undoes these things by insisting that inventors must be forced to file patent applications and then must be forced to give up their exclusive rights to a federal contractor–regardless of that contractor’s capability to manage the invention–and that contractor then must make an effort to convey the patent monopoly to a single company.
If the university breaks up the patent monopoly and grants non-exclusive licenses or fails to enforce the patent, then the monopoly meme fails and there’s no reason for the university to have been “granted” the patent right in the first place. The federal government could do the same. To protect the monopoly meme, then, university officials deeply enmeshed in preserving their franchise to deal in patents, prevent any alternatives to the monopoly meme from flourishing–they balk at consortium non-exclusive licensing; they fight free and open source software; they reject dedication to the public; they reject participation in standards.
What is most odd in all of this is that if one were to enforce Bayh-Dole all the way through the standard patent rights clause with its variation for inventors, Bayh-Dole would not support universities as privileged to exploit the monopoly meme. University-based inventors would own their inventions and universities would be precluded from making any claim on that ownership unless the university had assigned the inventor to the task for the purpose of inventing and had therefore held equitable title. But the neat thing about Bayh-Dole is that it is a law that isn’t really a law–it’s whatever university officials say it is. Another reason Bayh-Dole is a big turd.
Having argued that biomedical inventions are especially not getting used because the federal government makes them available to all, Bayh makes a great leap in his introduction of S. 414:
Unless universities and small businesses receive the right to retain the patent on these inventions,
The problem at the time, as identified by the Harbridge House report in 1968, was pharmaceutical company participation in federally supported research programs, for which Bayh claims that companies require patent monopoly positions. Instead, however, Bayh claims that universities must have ownership of federally supported inventions rather than that inventors at universities should be given incentives to assign their inventions to the pharmaceutical companies of their choice (if that’s really what we the public want them to do). Bayh argues that universities (and small businesses–but that is political baggage here) must “receive the right” to choose the company instead of the faculty inventors.
There is a bunch packed into “receive the right.” Universities at the time had the right to request to retain title. For that, the university had to acquire title from the inventor and request from the agency a determination of principal rights. With that request, the university would have to provide information regarding its plan for developing the invention to the point of practical application and making the invention available to the public. How would university exclusive control of an invention made in federal work better serve the public than open access? A crappy plan would mean a delay in the agency determination. One might complain, then, about delays without bothering to raise the issue of whether universities were presenting well considered plans for development and access. If a university plan amounted to “we will find a wealthy speculative company and turn over our patent monopoly to it but we really have no idea how this might be done and we have a poor track record of ever doing so,” one might see reason to delay the determination.
Bayh then ends with a fine non-sequitur:
valuable discoveries wind up wasting away on the funding agency’s shelves, benefiting no one.
The discoveries aren’t necessarily “valuable”–just because something is technically “patentable” does not mean that it is valuable. One could patent improvements to steam locomotives, but the improvements are not necessarily valuable because we don’t value steam locomotives like we once did. The “discoveries” don’t “waste away” on “shelves”–they are published, if no where else, then through the patent system. But faculty publish their discoveries. Priority in discovery matters. Publication matters. The “shelf” that a given discovery sits on in such a case is that of the public domain, the commons.
Bayh here is not arguing merely that universities ought to intercept inventions to save them from being made generally available by federal agencies, but rather, without quite saying it, that universities should be allowed to convey exclusive rights in each biomedical invention to their favorite company. Otherwise, stuff “wastes away.”
The general case is not true. The biomedical case also is not true, and wasn’t true when Bayh made these remarks. As Rebecca Eisenberg has show, the federal government’s licensing efforts with regard to biomedical inventions resulted in a 23% licensing rate–comparable to the top licensing rates that universities and their patent licensing agents were reporting for non-federally supported inventions (and across all fields). For inventions managed by universities and their agents under the NIH’s IPA program, the licensing rate was about 10% and the commercialization rate was no more than 5%. Quite apart from these numbers, Bayh ignores the success of medicines that were developed and made available to all. For instance, the Salk polio vaccine. The monopoly meme simply does not account for development work. Insulin is another example of a medicine released to multiple manufacturers. Same for Vitamin D.
Of course, it is true that companies do obtain exclusive control over inventions and do at times develop some aspect of such inventions into commercial products. It just does not follow that because such things happen, that is the only way that they may happen. Steven Johnson, in Where Good Ideas Come From, includes an appendix that argues most major new technology since 1800 has arisen through anything but a market-directed single person (or entity) activity–that is, outside the requirement of a single company holding exclusive market rights as a basis to develop a new commercial product. It happens. It just has not been frequent. And nothing from the data for the IPA program shows that forcing inventors to give up their inventions for management by university-designated patent brokers to be licensed exclusively makes exclusive market development any more effective. The IPA program was shut down as ineffective and contrary to public policy. Bayh-Dole took its place almost immediately. And Bayh-Dole’s numbers are so awful that no university will report them, nor will the federal government.
Bayh’s remarks are also all bluff because the NIH IPA program had been in effect from 1968 to 1978. This program was directed to universities and nonprofits and targeted biomedical inventions. Whatever the 28,000 patent meme had to say about federally owned inventions, it had next to nothing to do with the 300-odd biomedical inventions owned by the federal government and the not quite 100 biomedical inventions made with federal support and owned by the 70-odd universities and their patent licensing agents involved with the IPA program. There were no delays in “granting patent waivers” for the universities in the IPA program, since the IPA program didn’t involve waivers–universities could take ownership of any invention they chose to file a patent application on, within the scope of the IPA program, of course.
Admiral Rickover, an opponent of Bayh-Dole, dismantles the fallacies of Forman’s testimony:
In rationalizing their claim for title or exclusive rights to Government financed inventions, contractors often use the age old arguments of the patent lobby;
That’s interesting–that Bayh-Dole might be the product of a patent lobby, not universities and not disinterested policy analysts.
they claim that the Government is stifling technology by retaining title to approximately 25,000 patents; that these patents reflect worthwhile ideas that are not being used; that without patent protection companies will not commercialize these inventions; and that the public therefore does not get the benefit of the Government’s R&D expenditures.
That’s a pretty fair account of Forman’s argument, no?
Generally, these are the arguments of patent lawyers, contractors, and those unable to find sponsors for their inventions.
Have we heard anyone advocate for Bayh-Dole that wasn’t in one of these classes? If one has the money to develop an invention, does one need the incentive of a patent monopoly owned by someone else?
Truly good ideas tend to be used.
That might be the best anti-Bayh-Dole koan ever stated. Rickover then critiques the uses made of the patent system:
The reason so many Government-owned and privately-owned patents are not used stems from considerations other than the need for monopoly patent rights. A vast majority of patents are of little or no significance. Many companies seem to file patents defensively; meaning that they file numerous patents for minor details primarily to keep someone else from getting a patent in that area or to discourage potential competitors. Some people file patents as status symbols; others simply misjudge the attractiveness of their ideas. The Patent Office itself, when in doubt, tends to patent questionable items on the assumption that, if the patent becomes important, the validity of the patent can be tested in court.
Rickover’s point is that the federal government has little need for any of these uses of the patent system. Thus, it also has little need to start a trade in patent monopolies it acquires so that private speculators can do any of these things.
Then Rickover goes after the claim–made without documentation–that valuable discoveries just sit on the shelf if there’s no patent monopoly to be exploited to suppress all competing development or use:
Finally, it is almost impossible to tell the extent to which patented inventions are being used, particularly in the case of Government-owned patents. Government agencies do not have a reason to search for patent infringement. The Government, unlike private parties, generally has no desire to prevent others from using its inventions.
Rickover turns to reasons for the federal government issuing patents to itself:
The reasons the Government should take title to these inventions are primarily to ensure the Government is not subsequently barred by someone else’s patent from using the idea;
The government is not barred from the use of any patented idea–its problem is that it may be exposed to a claim for compensation for that use.
to preclude the establishment of a private monopoly for a publicly financed invention;
That is, the government has a responsibility to determine the scope of the patent system with regard to government-supported activities. This is at the core of any policy on federally supported inventions. Advocates for Bayh-Dole argue that the federal government should be indiscriminate in permitting private monopolies–other than that inventors should be disenfranchised from any ownership in their work.
and to ensure the public has equal access to these inventions.
Equal public access is precisely what Forman argues suppresses the use of inventions, and that if an invention is patentable, and so can be prevented from equal public access, then somehow those inventions will be developed and used.
It’s clear. Some inventions, having been patented and a private monopoly asserted, get used. And some inventions in that same class do not get used, despite the private monopoly. Some inventions, not patented or otherwise freely available, get used (think: standards). And some inventions in that same class do not get used, even though freely available. It’s not the patent that makes the difference.
The Harbridge House study found that when an invention is owned by the company that made the invention, and that company has expertise in the area of the invention, the likelihood is greatest that the invention will be used. Patenting doesn’t have a heck of a lot to do with it. One would think, then, that federal policy to maximize invention use would be to award contracts only to companies that have an established capability in the area of the contract. Universities and small businesses would be left out.
If participation is what matters–everyone gets a trophy stuff–rather than use, then we can go off into feel-good land policies, but then we have produced the problem of use by contracting with organizations that don’t have any use and don’t have any idea how to recognize use or do anything about it. Yes, exclusive patent licensing is a bad response to a cruddy policy about federal contracting.
If, however, a portion of federal research contracting is for “basic” science–the science Vannevar Bush envisioned would open up new frontiers–then use of inventions is not the primary goal. The primary goal is public access to new science. That goal is not enabled by packaging each new advance as a patent monopoly held for two decades in the hope that one company will take over the monopoly, suppress all use but its own, and make an effort (likely to be abandoned) to develop one tiny part of that advance as a commercial product.
It is also foolish (other than from the perspective of the patent lobby) to argue that this packaging frontier research advances as patent monopolies is just the thing for all other areas of federal research contracting, creating a “uniform” policy. Because after all, a “uniform” policy must be a fundamentally better policy. How is that?