In July, Setareh Samii published “The Importance of the Bayh-Dole Act” at The Catalyst, a web site operated by PhRMA. Samii argues that Bayh-Dole “created a framework for technology transfer that helped rejuvenate the American economy.” Samii then proceeds to retell–without any credible evidence–a story about how before Bayh-Dole, the problem that worried the federal government was commercialization of federally supported research at universities. There apparently wasn’t any similar worry about work done in the federal government’s own labs. But actually, while those advocating for Bayh-Dole made lack of innovation a premise of the proposing the law, they never were able to make a connection between innovation and their law. What’s more, the “commercialization rate” under the IPA program was just under 5%–about what the advocates claimed the federal government rate was. And the government wasn’t trying to “commercialize” inventions made with public funds, at least not by creating private monopolies.
The working premise of federal invention management was that there were multiple ways by which discoveries might come to be used by the American public. The expectation was that inventions made with federal support, like other findings, when published were dedicated to the public domain. Such inventions, if patented, were published as well in the patent literature and could be made available under non-exclusive license when additional restrictions might be required, such as, say, not blocking future research with patents on improvements or conformance to a standard or maintaining a consistent quality. Given that much HEW funding for university research involved matters of public health, the working premise of federal invention management also argued that the government should not be in the business of setting up private monopolies to exploit public health needs. There ought not to be a government-enabled market for patent monopolies directed at profiting from relieving suffering.
Oddly, however, this premise is nearly exactly opposed to what the pharma industry is all about. That’s a basic disagreement. Patents on health care inventions was where the argument about federal patent policy has played out. PhRMA support for the practices under Bayh-Dole is really an argument that profit-making–monopoly-enabled profit-making–is essential for the U.S. to lead the world in exploiting human suffering as an industry. The PhRMA argument is that if it weren’t for monopolies, no industry for new drugs would exist. That’s perhaps truer than PhRMA intends. It is not the case that no new drugs would exist, but that there would not be a viable way for distant shareholders, speculating on the fortunes of companies seeking to exploit human suffering, to profit. There would be no betting pool for the treatment of disease and accident and misfortune. Especially, there would be no betting pool for finding new drugs that transformed acute suffering into chronic suffering–the sweet spot for profits, where everything else is “uneconomical” or “orphaned.” Feeling cynical? Ah, it’s nothing new. Reread Middlemarch.
Thus, pharma needs Bayh-Dole because it needs to prop up the idea that monopoly is the only way by which new discoveries are developed into beneficial products. Pharma happily ignores the idea that there have been and still are other approaches. Without Bayh-Dole, there might not be an industry backed by speculators, but there would still be drug discovery and development, just as there was before Bayh-Dole, before IPAs, and even before the massive ramp up of federal funding.
Consider these alternatives.
- We would develop drugs using different ways, as we have done before
- We would focus on different problems, not selected by whether speculators profit
- We would not seek to profit so much from human suffering
- We would form consortia and commons on which to build shared technology
- Speculators would find other things to bet on against each other
- More health information would be openly available and less held as trade secrets
- Monopolies that advance public benefit would happen later in the process
This is how the semiconductor industry has operated, and the energy industry, and internet, and telecommunications. There’s no reason that drugs cannot be developed in a similar way, spreading information and risk and expense and sharing the opportunities.
Without monopolies to make betting pools for speculators, attention might shift from drugs that turn acute conditions into chronic ones to prevention, to cure, to a knowledge of how a system works. That is, federal funding might focus on overturning or transforming consensus views of physiology and genetics and environmental effects rather than trying to find new compounds to throw at health conditions, rather as a stop-gap–though a lucrative one if the right conditions are picked as targets. Put another way, when a bio-active compound has been identified, its life need not be dedicated to serving a monopoly betting pool. It could fall into other hands, where it would be investigated for what it teaches about how biological systems (if they are even systems) operate. Rather than being one compound bought up with 10,000 others–and really just one of intractably many variants–being tested for safety and efficacy, a compound would be considered for what we learn from it.
We then might not make such a big deal over the statistical outcome of a pseudo-random “clinical trial” (about as reasoned as as a “trial,” for all its apparent sophistication as a trial about whether witches float). Evidence can be manipulated, as at any trial, and when the outcome of the clinical trial is more important to the fortunes of the betting pool than it is to public health, the odds favor manipulation in the design of the trial protocol, in the analysis of data, and in the suppression of problems. The manipulation cannot be overt–it has to be buried in the details, so one skips over the problem areas. Sort of how Bayh-Dole was drafted to look like but not be much at all like the IPA, which in turn was drafted to look like but not be much at all like the Kennedy statement of government patent policy. Slip-sliding away–by design. If we did not have an industry exploiting human suffering that was more important to its betting pools than to its effects on human health, then clinical testing might take on an entirely different character.
The working premise of federal funding for research was that public money should result in results being available to all. This was the opinion of the Attorney General in 1947 and the operating default for years–during which all sorts of worthwhile things resulted from federally supported research, without an initial monopoly. It’s not that monopolies don’t have a role, but that their role generally is later than that of discovery and early characterization. If the core discoveries are shared, then any number of people can study them, use them (such as for more research), and vary them.
These characterizations create a common platform that defines the discovery. Applications then can be made of the platform, and specific new inventions–pathway dependent–can “protect” the investment of any one company in its implementation of the discovery, all without creating a monopoly over the initial discovery or making a demand that only one company can characterize the discovery to find variations or build a platform. This was how the digital computer was developed, how the internet has been developed, how most anything involving use before products has developed. How technology can, in general, develop. There is a role for monopoly–call it curatorial or call it ownership that inspires attention or call it a right to one’s own efforts–but there is little to indicate that extending that monopoly to discovery of scientific knowledge contributes to the development of practical applications of the discovery.
There is a common expression–often made without thinking–that university research represents “early stage” technology, as if all technology is in need of monopoly-based development. While it is possible to characterize every discovery as “early stage,” doing so assumes an approach that isn’t necessary. One might also argue that university research represents opportunities to develop various sorts of commons–of shared, examined, tested, varied technology–and that the monopoly parts of the effort come once there is a commons, a platform. Such a monopoly does not control the commons or platform, but represents a narrow development of the platform. Others then also can use the same platform, and develop it in different directions to create products. And anyone who does not need a product version can access the platform at need. In this case, “early stage” technology means technology that is forming a platform, not something that demands a monopoly in order to be developed at all.
From a standards point of view, the platform constitutes the “essential claims” of any patent–those claims that must be shared so that the platform is available to all. The “non-essential claims” are those that a monopolist can exploit without interfering in the use of the platform or its common development–variations and characterizations. Once a platform has a critical mass of contributors, then its future direction depends on negotiation. If some company wants to keep its monopoly on a non-essential claim that could become part of a new standard, then the standard might move in another direction, isolating that non-essential claim and leaving the company making the claim as an outlier, having to fund all its own development.
The claim, then, that Bayh-Dole is necessary for “early stage” discovery then ends up being a claim that there should be no commons between a discovery made in public interest research and commercial monopolies purporting to create useful products. The monopolists are in competition with the commoners, or to be blunt, the speculators compete with those that serve. The speculators claim that making immediate monopolies is the only or best way for the public to benefit. What Bayh-Dole did was allow speculators to move from betting on the fortunes of companies–that is, acting as shareholders–to moving ahead of companies and making bets on whether some companies would want to acquire a monopoly right rather than just use a discovery or contribute to a developing commons.
It is easy to see that a betting pool before any company has access can speculate on the value of access rather than on the value of any particular downstream product. Bayh-Dole allows speculators to exploit not only the value of an invention to product development but the value of an invention to a developing commons–denying access to a commons, preventing the formation of a commons as a way to increase the speculative value of patent rights in any given initial invention. No wonder, then, that the cost of developing any invention denied access to a commons becomes huge for the monopolist. It’s not that the cost to develop an invention is so great that only a monopolist can do it–just the opposite. The high cost comes about because a monopolist has arrived too early in the process, so that multiple variations have not been made, contributions have not been made, ideas have not been exchanged.
The monopoly raises the costs; the monopoly comes first, then the high costs. Not the other way around, not that monopolies are necessary to cover the high costs. Monopolies create the high costs, concentrate the high costs in a single speculative entity. Such monopolies are the work product of patent speculators–the flip side of the patent troll, who waits until people are using an invention to sue them for payment. Let’s call them patent llorts. A patent llort creates monopolies and speculates that someone, anyone will purchase the monopoly for more than the llort has paid to acquire it. Of course, then, getting the patent monopoly for absolutely free (or, just the cost of filing a patent application) is a huge windfall for the llort industry. And that’s what Bayh-Dole did–it allowed llorts to take ownership of inventions that otherwise would have gone to the public or to the federal patent commons or would have been held by inventors.
Here’s how Samii puts it–the PhRMA spin–
The Act created a viable route by which new insights and valuable research results from universities and other institutions can make their way efficiently to start-up and established firms. These companies then assume the full risk of development and cost for commercializing the few technologies that eventually prove to be technically and economically viable products.
The IPA was already, so it was claimed, a viable route. So Bayh-Dole didn’t create any new viable route. Further, Bayh-Dole has nothing to do with “new insights and valuable research results” nor with “technologies”–it deals with patent rights in inventions. And insights, results, and inventions all can make their way to companies through publication, teaching, and consulting–those were and still are perfectly viable routes. The problem–turned to a virtue–as that whoever obtains a private monopoly then must “assume the full risk of development” for whatever inventions fall into this new “route” created by Bayh-Dole. And note the last sentence, hidden there: “… the few technologies that eventually prove to be technically and economically viable products.” That is, Bayh-Dole is horrifically wasteful. There are no data on Bayh-Dole’s “success” rate–so it is difficult to know, but the overall “success” rate from a few published sources appears to be around 0.1%–so, call it a disaster rate of 99.9%. The llort model is to capture everything possible as private monopolies and speculates on them, with only a tiny portion making money and the rest withheld from use to prevent any other viable route to become evident.
All the llorts had to do was to persuade everyone that the only way that the public would ever benefit from federal research was if a llort was involved, was if a llort first created a speculative monopoly before any commons could form, was if a company took assignment (but call it an exclusive license) of the monopoly from a llort. Of course, if llort could not do anything with the monopoly, it could always wait and become a troll. Or it could create a shell company (as universities such as Utah and Washington did) and have that shell company acquire the patent (call it an exclusive license) to make it appear that the llort program of “technology transfer” was working. Then, when the shall company failed–as most do–it could serve as the troll, the front to sue industry for adopting an invention anyway, without the need for any llort-created initial monopoly. The threat of trolling is essential to the operating model of the llort. Take assignment from the llort or expect to get trolled later.
Bayh-Dole is an attack on the research commons–not only the federal patent commons but all forms of discovery access and exchange. Bayh-Dole enables llorting and trolling. Bayh-Dole makes llorting an industry dedicated to preventing the formation of research and development commons and instead creating speculative monopolies. Not every research discovery benefits from a development commons, but nearly all research discoveries are harmed by an initial monopoly operated by llorts–that is an initial monopoly on a research discovery that must be assigned only to a single monopoly interest for development and “commercialization.” It is not even clear whether drug development is best served by the business model of the pharma industry–end-to-end monopoly. That may be the preferred betting position of the shareholders aiming to make money from the fortunes of pharma companies, but there’s no reason for the federal government or for university administrators to give a rat’s ass about what betting shareholders want. It is as if federal research patent policy has been dictated by people betting on pro football. Sure, there’s a lot of money at stake. But why should that matter to anyone doing research or managing research for discovery and public benefit.
The real question is not whether Bayh-Dole is a success or is even necessary–it is clearly neither–but what is necessary for a robust commons to develop prior to efforts to “commercially” apply what has been discovered. The role of federal patent policy should be to prevent the early formation of private monopolies in new discoveries. That is, delay the intrusion of speculators into a fundamental research process–delay corporate speculation, delay troll speculation, delay university administrator speculation. For this, Bayh-Dole has been a total, unqualified disaster–not on its own, but in combination with unthinking, self-interested university patent administrators who repeat the claims of the pharma industry, that monopoly is the only way that research discovery
Back to Samii’s piece for PhRMA:
Seeing an opportunity for reform, leaders from both parties came together to enact legislation that would facilitate orderly and efficient technology transfer from universities and other institutions receiving government research funding to the private sector.
Except that pharma already had access to “orderly and efficient technology transfer” (if that’s what it was) under HEW’s IPA program, which had enrolled over 70 university and nonprofit research institutions. There was essentially no change in practice for the pharmaceutical industry from the IPA to Bayh-Dole–other than that Bayh-Dole eliminated most public oversight, eliminated most conditions on exclusive licenses (i.e., assignment of monopolies), and forced all other federal agencies to adopt the same approach, regardless of their missions, the nature of their contracting, the needs of research or the market, or how innovation comes about.
The approach that Bayh-Dole destroyed allowed the creation of commons and platforms before commercialization. That approach was highly efficient–in most cases, all that was required was for someone to read a publication and use the invention. No licensing, no negotiation, no delays, no demand for payment, no requirement to “commercialize,” and no license conditioned on such “commercialization.” Bayh-Dole created an industry of llorts speculating on deriving value from preventing any such commons or platform to come about. What an awful piece of legislation!
And what does Samii show for Bayh-Dole’s success?
The results have been staggering:
Since the passage of the law, commercialization of federally-funded research has increased dramatically — between 1980 and 2002 alone, U.S. universities generated a tenfold increase in patents.
Let’s see, a staggering result of Bayh-Dole should be that many inventions made with federal support are being used. Instead, Samii quotes an increase in the number of patents. The number of patents has nothing to do with an increase in commercialization. Patents exclude all use but licensed use. An increase in the number of patents, on its own, is an increase in the number of excluded uses. A 10-fold increase in patents means a 100-fold increase in exclusion, if only 10 companies on average might have desired access to any one invention. Even the number of exclusive licenses granted has nothing to do with commercialization. Commercialization happens when a product is available on the market and people buy it. All the number of exclusive licenses indicates is the relative success of the llort model in speculating on patent rights ahead of platform development and company interest in use.
Technology transfer activity has a significant impact on the U.S. economy, with one study finding that between 1996 and 2013, academia-private sector patent licensing across all industries bolstered U.S. GDP by up to $518 billion and supported up to 3,824,000 U.S. jobs.
Now the metric skips from subject inventions under Bayh-Dole to “technology transfer activity” in general–no distinction between subject inventions and any other inventions. Technology transfer generally is no proxy for Bayh-Dole–one has to look at practical application arising from the use of the patent system for federally supported inventions. What does Bayh-Dole itself indicate the metric should be? The standard patent rights clause authorized by Bayh-Dole sets forth the metrics at 37 CFR 401.14(a)(h):
(i) the status of development
(ii) the date of first commercial sale or use
(iii) gross royalties received by the contractor.
None of these figures are ever reported to the public for any subject invention, and Bayh-Dole exempts such reports from federal public disclosure law. Of these metrics, the date of first commercial sale or use–the date on which practical application (as defined by Bayh-Dole) has been achieved–tells us two things: first, that there has been commercialization for a given subject invention, and second, the duration of the effort from the date a patent has issued on a subject invention. From just these two figures we would learn a great deal about how well the llorts have done in bringing inventions to public use while suppressing the operation of any commons. Or, as appears to the case, how terribly.
But Samii powers right through the utter lack of evidence for any sort of Bayh-Dole “success” and asserts what ought to have been demonstrated by any data at all:
Despite the law’s track record of proven success, today Bayh-Dole is under attack.
Well, no wonder. There is no proven success. There are no data to support Bayh-Dole’s success. There are data to indicate that Bayh-Dole has been a disaster. And for pharma, Bayh-Dole changed virtually nothing from the IPA system that was already in place at HEW. The IPA program was canceled in 1978 because the director of HEW determined that the program was not operating in the public interest. Despite an apparatus that made a gesture to non-exclusive licensing, licenses under the IPA program were mostly exclusive and were mostly unproductive. Over 95% of claimed–monopolized–subject inventions were simply that–monopolized, not developed to the point of practical application by monopolists. Just monopolized. Kept from the public, kept from the federal patent commons.
The far-reaching success of Bayh-Dole should not be put at risk. Today, the law’s technology transfer policy is fundamental to the current, highly successful U.S. biomedical R&D ecosystem and has helped foster the development of innovative therapies that have revolutionized medicine and patients’ lives in cancer and many other disease areas.
The only far-reaching success of Bayh-Dole is the destruction of the federal patent commons and IP policies at universities that encouraged collaboration and collective action in all industries but for those operating under the IPA program at HEW (and, briefly, for a few institutions, at the NSF). Bayh-Dole created a world for monopolist speculators–for llorts and trolls–but lousy for diverse innovation dynamics, for commons, for collective action. If anything, Bayh-Dole is the federal law that has done the most to shut off innovation enterprise arising from university research.
It won’t be enough to repeal Bayh-Dole. The liver doesn’t heal instantly because one stops the binge drinking. There will have to be protections for inventors, for platform development before anyone claims monopoly positions, for commons and collective action. There will have to be strong federal agency and public oversight, at least until things stabilize in productive directions.
If Congress determines that we should subsidize a betting pool for speculators hoping to profit from public suffering, then give pharma its own special program, but keep everyone else out of that program. There is absolutely no public interest served in adopting a “uniform” federal policy on patents that amounts to a welfare system for wealthy speculative monopolists and the university patent administrators who carry their water for them.