PhRMA, a pharmaceutical industry lobbying group, has published a white paper championing Bayh-Dole. What they are after is to prevent the Bayh-Dole march-in provisions from ever operating. To do this, they make a variety of assertions about Bayh-Dole that can’t be supported or which involve data that has nothing to do with Bayh-Dole. The broad issue is whether, without limitations on monopoly pricing and term, we pay way more than we ought to for prescription drugs created based on inventions made with public funding. If pharma companies get the benefit of federal research funding and can do pretty much as they please with the results, aren’t they enjoying a public subsidy that bolsters the income of shareholders? Doesn’t the argument then reduce to “public-subsidized private monopolies that exploit the public in matters of healthcare are in the public’s interest”? Perhaps it is no wonder that pharmaceutical companies like the idea of university bureaucrats handing patent rights to them.
Let’s take a look at their claims for the success of Bayh-Dole, and then decide what evidence there is for actual success–if that’s the right word. Sadly, almost every sentence in the PhRMA document is nonsense. Unsupported, misstated, distorted, suppressed. Deceive for the sake of the cause. If you like the idea of university bureaucrats in collusion with big company bureaucrats to make sure that the federal government does not have oversight over monopoly deals, where the big companies get 99% and the university bureaucrats get 1% and the government gets 0% and the public gets screwed with monopolies, nonuse, and price-gouging for 20 years, read no further.
Still with me? Let’s take some of the PhRMA executive summary sentence by sentence. It’s dense with spin and misinformation. Read and enjoy.
The University and Small Business Patent Procedures Act of 1980 (commonly referred to as “Bayh-Dole”) created the uniform framework that facilitates orderly and efficient technology transfer from universities and other institutions receiving government research funding to the private sector.
Actually, Bayh-Dole doesn’t have anything to say about how licensing gets done or even if it does get done, other than a preference for small businesses and that exclusive licenses for US must require (with exceptions) substantial manufacture in the US. Nothing in Bayh-Dole addresses a “uniform framework” for technology transfer from universities. Bayh-Dole creates an arbitrary conditional default with regard to federal ownership claims to inventions made with federal support and provides a procedure for federal agencies to vary from that default.
Certainly there’s nothing in Bayh-Dole that facilitates “orderly and efficient technology transfer.” Even if a federal agency did require ownership of an invention made with federal support, or did delay in agreeing to allow a contractor or assignee retain invention rights, that requirement or delay does not affect technology transfer so much as the ability to close an exclusive license deal. If the federal government acquired ownership of the invention, regardless of any government patenting, the invention would generally be made freely available–so technology transfer would happen even more readily. It is the university operating under Bayh-Dole that creates barriers to technology transfer to protect a patent position for which it expects compensation. If “technology transfer” means providing access to new technology and assistance in its use, then patents are not generally needed. If what PhRMA means, rather, is “exclusive patent licensing by nonprofits to gain a share of the upside of private investment and profits, to the exclusion of any other form of support for development or use of the invention,” then PhRMA ought to write what it really means.
There’s no evidence that university licensing is any more orderly and efficient for Bayh-Dole subject inventions than it was for IPA subject inventions before Bayh-Dole or for inventions handled by invention management organizations rather than by universities, or for any inventions that were not federally supported.
Bayh-Dole allows universities and other institutions to own title to the patents arising directly from their research activities.
Silly and not true. Silly part: patents don’t arise in research. Patents are a result of a decision to exclude others. Inventions may arise in research. Patents never do. Now the not true part: Bayh-Dole deals only with inventions arising from federally supported research, not all inventions made at universities. It’s not “their research activities” but only those research activities supported by federal funds. The impression given is that Bayh-Dole changes patent law on ownership. The Supreme Court rejected that idea. Further, to be picky about it, Bayh-Dole restricts the claims a federal agency may make on subject inventions–inventions made with federal support *and* which a university or other federal contractor has gone out its way to acquire. Wherever an agency claim is so restricted, someone else who owns the invention is allowed to continue to own the invention.
It is therefore equally true then that Bayh-Dole allows university inventors to own their inventions. See 35 USC 202(d) and 37 CFR 401.9. A true, complete, statement would read: “Bayh-Dole conditionally prevents federal agencies from claiming ownership of inventions made with federal support when the inventors own their inventions or assign their rights to the universities that have hosted their work or to other approved organizations.”
With these clear patent rights, universities are then free to license the right to use the most promising technologies to private sector partners in order to commercialize them.
The “clear” in patent rights must refer, lazily, to the idea that before Bayh-Dole, other than with HEW and NSF (which used IPAs) and DoD (which allowed contractor ownership), if a university took ownership of an invention made with federal support without bothering to ask the federal agency that funded the work, the agency, under its funding agreement and the Federal Procurement Regulation patent rights clause, could require assignment to the federal government.
Universities are not “free” to license the “most promising” technologies under Bayh-Dole–they are obligated to promote the utilization for each subject invention they claim. Not just the “most promising ones.”
As such, Bayh-Dole—which passed with strong bipartisan support–created a viable route by which new insights and valuable research results from universities and other institutions could make their way efficiently to start-up and established firms, who then assume the full risk of development and cost for commercializing the few technologies that eventually prove to be technically and economically viable products.
Um, patentable inventions, not just any insights or results. But the next part is funny–the government has undertaken the cost of research that has led to the invention and that, too, is part of the cost of development. For disease assays and other method inventions, most any capable organization can implement use without any company assuming risk or costs. So that part of PhRMA’s account about assuming the full risk is only good where no-one can practice without substantial investment in development of the invention, and that additional investment for some reason is not made by the government and not shared among a variety of companies. PhMRA here writes into Bayh-Dole a defense of the pharma business model–giants competing for monopoly rights in multi-billion-dollar products.
And we get the apology for the lousy performance record. Bayh-Dole requires each subject invention that’s claimed to achieve practical application or the federal government can “march-in” and compel licensing of the invention. Commercial products are not the primary or necessary goal of the law, even if that’s the goal of PhRMA member organizations. Bayh-Dole does not endorse a portfolio model that lets universities accumulate a bunch of patents from which investors and companies can pick and choose. Bayh-Dole does not provide that if a subject invention is patented it can be used to exclude others even if licensed to a company but the company has not developed the invention to the point of practical application. Such exclusion makes sure that no one else can try to use the invention or develop it while a company futzes around with trying to make a commercial product of it or an investor holds the invention hoping that it will block others’ efforts to develop products and so in that way turn out to be valuable.
This paper focuses specifically on the contributions of Bayh-Dole in fostering technology transfer in the life sciences and current threats to this robust framework.
Except that life sciences tech transfer was doing fine–apparently– under the IPA regime, and in the private network of invention management agents. The problem was that HEW, in owning patents on inventions that did not move through an IPA to a university or invention management organization, preferred non-exclusive licensing, and the pharma industry mounted a boycott. The work-around then was to have universities obtain the rights to inventions, and let industry go back to its monopoly-loving ways by paying a 1% or such royalty to whatever university happened to obtain patent title. Before Bayh-Dole, university tech transfer was already well-fostered. See Mowrey’s work. It remains to be seen if Bayh-Dole is a “robust” framework. On the face of it, Bayh-Dole is less robust than the IPA system–built by the same folks–that Bayh-Dole replaced.
Ill-informed proposals to eliminate fundamental aspects of the cooperative academia-industry framework which developed as a result of Bayh-Dole and has been operating successfully for 36 years, or to use this framework to regulate drug prices, reflect a fundamental lack of understanding of the research and development (R&D) process and the benefits that accrue to patients, society, and the economy through the transfer of intellectual property (IP) and the development of innovative treatments.
Anchoring text here. There’s no evidence that any proposals are ill-informed or that they have a “fundamental lack of understanding.” This is the old, foolish idea that if everyone just knew the facts, then everyone would come around into agreement. The problem here is that well-informed people, with a clear understanding, disagree with PhRMA on the role of publicly funded research. The issue is not about stupid people bothering the bright and talented ones, but about whether the pharma industry should have such fine grazing for profits at the public expense. That, despite the pharma industry’s protestations, is a matter for civic discourse, not bowing to industry’s assertions.
And, of course, Bayh-Dole only deals in patent rights (and for some inexplicable reason, plant varieties) on subject inventions, not intellectual property generally. Expanding the discourse here obfuscates, making Bayh-Dole appear to be the creator of order and efficiency for all research results, not merely patent rights extracted from those results.
In the specific case of biopharmaceuticals, together with other factors such as the development of advanced scientific tools and techniques and the emergence of the modern risk-based venture capital market, Bayh-Dole helped lay the foundation for today’s robust biomedical R&D ecosystem and it’s spirit of entrepreneurship which has helped propel U.S. global leadership in the life sciences.
No question that Bayh-Dole has had a role in the development of the “R&D ecosystem.” The government pours billions into life sciences research. Bayh-Dole affects the portion of that research that is contracted out. It doesn’t have anything to do with the portion of research conducted by government scientists in federal labs. That part is covered by Stevenson-Wydler, among other things. But notice the move to qualifiers–first it was Bayh-Dole, and now it’s Bayh-Dole among “other factors” that weren’t ever Bayh-Dole, and Bayh-Dole just “helped”–it didn’t cause.
Recall if you will all the fuss around patents on scientific tools, and the efforts by the NIH to try to push non-exclusive if not open availability for research tools, to be resisted by the universities. Oh, was that Bayh-Dole “helping”? Perhaps PhRMA uses an idiosyncratic meaning for “help.” If one wants to get at it, perhaps we could argue that Research Corporation and the IPA system laid the foundations for connecting university research with industry development. Bayh-Dole’s role was to push federally supported inventions toward this same infrastructure, only to cause the infrastructure to be destroyed and replaced with a university bureaucracy. It’s difficult to see how that did anything for the “spirit of entrepreneurship.” A bureaucrat’s thumb in every innovation pie. Ah, just what the enterpreneurs’ lobby was pitching for.
And better, university bureaucrats are happy to do monopoly deals with big companies, as long as they get 1% or maybe 0.5% of the gross. The startups + VC funding came in when it became clear that the big companies were not snapping up patents from universities and the VCs could do better by trading on the apparent future potential of a new company interposed between university administrators and pharma than they could on the future revenue from any particular product.
And while arguments for Bayh-Dole did worry about the US falling behind in technology, the law itself has nothing in that but a bit of protectionism. The policy and objective is to use the patent system when doing so will promote the practical application of subject inventions. Whether the industry has global leadership is different. But the implication is that somehow, if Bayh-Dole were to be reformed to better reflect public expectations of federally supported research, we would lose global leadership. And here I thought that the goal was better health care–that’s what this discussion is all about, not the well being of the present industry leaders that find their profits in bad health. To be blunt, if the U.S. succeeded in preventing or curing most diseases and other bad healthiness, then the U.S. would not be the global leader in pharma–we wouldn’t need disease maintenance drugs. We would be way down the list, and happy for it. To delight in global leadership in remedies for bad health is to delight in the prospect that we will always have a robust market in bad health–the more chronically bad, the better. Just saying.
The clear and consistent approach to U.S. licensing policy and IP rights established by Bayh-Dole create a predictable mechanism by which early-stage research that is supported in whole or in part by the federal government can attract the subsequent private sector investment necessary to enable successful commercialization for the benefit of patients, society, and the economy.
The licensing policy is (other than reading 35 USC 200 to mean something) that there is no policy or public oversight or accountability. The matter of IP rights is only that of patents on subject inventions. There is nothing predictable about it other than that inventors (would be entrepreneurs) are butted out of the way immediately and university administrators deal in patents for money positions rather than for public service–oh, yes, their mantra is that by chasing money positions, an invisible hand will bring about the public service in time.
What is the evidence that we are dealing with “early-stage” research? That’s assuming the conclusion in the premise. Much university research is applied and involves development to the point of use–for research tools, for instruments, for sensors, for biomaterials, for disease assays. The idea that all or even most university research is “early-stage” is not supported. It is true, for commercial drug development, that there is plenty of work to be done. No question. But that does not make university research in general or even biomedical early-stage.
Much rides on this word “necessary.” Under Bayh-Dole’s statement of policy and objective, patents are to be used when doing so promotes practical application. So if it is indeed “necessary” for there to be significant private sector “investment” before there can be practical application, then patent licensing is authorized. Otherwise, Bayh-Dole doesn’t authorize the use of the patent system. But no university I know of worries whether a patent is necessary–just whether a patent can make money.
And now the punch line: this all is for the benefit of “patients, society, and the economy.” Holy cow! I was ill-informed. I thought the pharma companies were in it for profits, but company profits don’t even figure. And that’s what folks are all upset about. I can see now how everyone is ill-informed. These pharma companies develop drugs as a public service, at the lowest possible cost, and their shareholders are all good with that, holding stock as a public trust to keep the companies in business. Sweet. Perhaps the “benefit for society” is intended to mean “benefit for the pharma companies and their shareholders, who if they remain wealthy, won’t cause problems for everyone else and that will be a benefit to all.” I know, there’s not much room in executive summaries for nuance like this.
It is not at all clear, however, how high drug prices benefit the economy. Explain that. I could see how a company might make back its investment (provided that investment is properly accounted), and have a bit of profit to go along, and perhaps a bunch of profit to be invested in further development of drugs–but how much profit should go to the salaries of the top execs? how much to the shareholders? how much to be spent on global headquarters and glossy white papers? If we were dealing in luxury goods, we might say, as much as you-all want. But we are talking disease and injury, we are talking health. It’s like water, air, and food. At some point–and this is the point of the civic discussion–that’s enough profit. We might go so far as to say it would be nice to chase out the folks who are parasitic on our health problems and deal with those who put service over speculation. Again, it’s a civic discussion, not one that gives a rat’s ass for how much profit any given company might make or might share with university bureaucrats and a few lucky inventors (but not everyone else who may have worked the problem with them).
Assessments of Bayh-Dole have found it be a vast improvement over the previous state of affairs:
There have been no assessments of Bayh-Dole based on its stated policy and objectives. There is no evidence that Bayh-Dole has improved things. Just the opposite. The IPA system operated at 4%. The private system operated at 25% to 30%. Commercialization under Bayh-Dole is under 0.5%. How is that a vast improvement? Oh, yes, an idiosyncratic use of “vast improvement.”
Prior to Bayh-Dole, commercialization rates of federally-funded research were estimated to be less than 5%. Since the passage of the law, however, commercialization of federally-funded research has increased dramatically– between 1980 and 2002 alone, U.S. universities generated a tenfold increase in patents.
Here’s nonsense in a nutshell. Federally owned patents were commercialized at 5%, according to one estimate. But the federal government did not seek to commercialize–it held patents to create a commons. That is, its research output created new technology platforms, and only from time to time created actual commercial products. The internet, for instance, is the work of a commons. It is possible because many inventors and organizations contributed inventions to standards rather than held out for a monopoly play. To add to the nonsense, 5% is about the expected average rate for any patent to result in something commercial. It’s the rate the IPAs were doing. It’s 10x the rate that university licensing is presently doing. And the government wasn’t even trying to commercialize. The private network of invention management agents was doing way better than this because they were doubly selective–inventors chose them, and they chose their inventions carefully.
And now the switcheroo. We were talking commercialization i the first sentence. But now we are talking an increase in patenting. That has nothing to do with commercialization. Commercialization can take place without initial patents altogether–even in pharma. All those natural products, for instance. All that’s needed is to create a set of subsequent blocking patents to create a monopoly position. One doesn’t have to have a patent on the original compound. That’s just nonsense. But the deeper, deceptive thing is the switch from product to patent. A tenfold increase in patents is a measure of the non-selectivity of university patenting, a blocking of the commons, and a shift of money from other forms of support to paying patent attorneys to write expansive claims all the better to create monopolies and profit from them. A tenfold increase in patents has nothing whatsoever to do with practical application, with benefit to patients, with benefit to society (other than the society of technology managers and patent attorneys–they, of course, make money regardless).
Collaborations between universities and government-funded researchers and the private sector have proven to be a successful model to leverage complementary roles in basic and applied research to support the development ofmedical innovations and to address unmet patient needs. Without clear patent rights and protections and the economic incentive of exclusive licensing established under Bayh-Dole, private firms would not devote scarce resources to the highly uncertain development efforts needed to advance research from laboratories receiving public sector funding to the market or the bedside, in the case of medical therapies.
Remember, this is in a bullet list on what the assessments of Bayh-Dole have found over the previous state of affairs. There were collaborations before Bayh-Dole. Recall the 77 research universities–nearly anyone that mattered–had IPAs specific to life sciences research that permitted university administration of inventions made with federal support. Companies were working with researchers then, too. What did Bayh-Dole do to vastly improve this situation? Apparently, it induces vast swatches of abstract bombast. What does “leverage complementary roles in basic and applied research to support the development of medical innovations and to address unmet patient needs” actually mean? That university scientists work with industry scientists? If that’s the case, then is industry really taking on the full cost of development? Oh, well, it depends on an idiosyncratic use of “full.”
The argument here is that companies, without a monopoly, don’t have any motive to assist people in need. Well, then, what good are the companies? Are they, as Joel Bakan argues, psychopathic engines designed only for the profit of their shareholders? And why must monopoly be the only route to profits? Why not consortia, as the semiconductor and energy industries use to spread risk and cost and develop common roadmaps and standards? In the civic discussion, the very real question is whether we are better off with companies involved in the relentless pursuit of profits in leading our health care.
Now let’s look at that “economic incentive of exclusive licensing” established by Bayh-Dole. Bayh-Dole doesn’t establish anything of the sort. It merely (and significantly) prohibits federal agencies from interfering in whatever economic incentives the owners of patents on subject inventions put themselves up to. The economic incentives to a patent owner arise from the ability to practice an invention while excluding others. In the situation in which university middlemen intervene–never to practice anything–the economic incentive is not to grant exclusive licenses but to promote practical application–that is, expressly in the law, to use a subject invention
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.
If there is a use of patents established by Bayh-Dole, this is it, and any economic incentive has to arise as a subsidiary interest of the university broker. Anything else is outside the authority of Bayh-Dole. Furthermore, the great early examples of university biotech licensing–Cohen-Boyer, Axel, and Hall–were all non-exclusive licensing programs. There’s nothing in exclusive licensing that lends itself to “economic incentives” for a patent broker. And if there are economic incentives for exclusive licensing, they are even greater for limiting the term of exclusivity and licensing everyone in the market rather than creating the conditions for folks to work around the patent and thus avoid paying anything. While company monopolies may be great for investors and for pharma execs, they aren’t so great for patent brokers, who take huge risks putting all their patent rights with a single company.
Now for Bayh-Dole efficiencies. Bayh-Dole creates an apparatus that has been used for universities to intervene to take ownership of subject inventions and then licensing them. That licensing is a nasty piece of work–pages and pages of legalese; months and months of negotiation. At the time Bayh-Dole was being debated, the question was raised–by Rep. Kastenmeier, “why not just have companies acquire title directly?” Bremer had no answer other than politically, they couldn’t get such a bill passed immediately. The better bill would by-pass the middlemen altogether–especially when the middleman is a state university with just about as much regulatory uptightedness as the federal government. Let inventors assign directly to big pharma. Give them a standard license to use. But, no, Bayh-Dole is a Judy Collins kind of law. This is what PhRMA is advocating for–the less efficient law written by the middlemen for the middlemen.
Such collaborations and licensing models have been a critical building block of the biomedical R&D ecosystem and the significant contributions it has made to the U.S. economy—and have contributed to the Nation’s competitive advantage in biomedical innovation globally.
The collaboration and licensing models have nothing to do with Bayh-Dole. The federal government commercialization rate before Bayh-Dole was 10x higher than the present university commercialization rate. Those models were around before Bayh-Dole. All Bayh-Dole has done is to make them non-selective, expensive, complicated, and further removed from industry than before.
I know this is just an executive summary, but we’ll have to see just how Bayh-Dole has contributed to models that have become blocks of ecosystem that have made significant contributions to the economy. I’d say, rather, that Bayh-Dole has been a “block” on innovation, not a contributor to it. And at any rate, the purpose of Bayh-Dole is not to make the US dominate global markets–it is to get inventions into use. The purpose–in Bayh-Dole at least–is not to be a welfare subsidy for the investors in big pharma. But that’s the implicit claim here. Without Bayh-Dole, investors don’t make as much money. Understandable, but still bizarre in the context of a civic discussion about who ought to be involved in creating new medicines and whether profit-seeking mega-corporations should be those folks. Just a discussion, of course, but why not have it?
While collaborations and licensing between academia and the private sector are particularly important to the biomedical R&D ecosystem, they are vital to driving innovation in other industries as well, particularly high-technology industries such as semiconductors. As a result, 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.
We have now shifted from Bayh-Dole to how collaborations and licensing are “vital” to “driving innovation” in other industries. But we haven’t established that Bayh-Dole has created collaborations let alone vastly improved them. And before Bayh-Dole, there wasn’t licensing between academia and industry because for the most part academics were smart enough to stay out of licensing. Licensing happened when academics decided to patent their inventions and took their inventions to invention management agents external to academia. Those agents–especially Research Corporation–were already closely involved with industry and that’s where any licensing happened. Bayh-Dole created the conditions to destroy that set of relationships, so universities ended up dealing directly with industry–ahem, licensing–creating all sorts of barriers to collaboration by non-selectively claiming ownership of all inventions–preventing things from entering the public domain or being openly shared. The universities went from being supporters of collaboration to disrupters of it, creating all sorts of tension between the development office, the sponsored projects office, the technology licensing office, and the faculty doing the research and making the attempt to collaborate.
But the real argument here–totally nuts–is that what’s good for the pharma industry is good for all other industries. That’s so untrue. While there’s certainly collaboration between industry and academia in semiconductor research and other areas of “high tech,” the operating models are anything but mega-company monopolies. No one could survive that approach. Instead, there are all sorts of standards, commons, consortia, open licenses, and cross-licensing arrangements to diminish the role of patents and increase the role of platforms and interoperability.
As for the economic impact study, note the “up to” weasel. And for all that, we are far, far afield from Bayh-Dole, from subject inventions, from patent licensing, from exclusive patent licensing, and talking “technology transfer.” In any event, the US GDP is about $18 trillion. So where does that $518 billion fall? Let’s see 0.5/18 is 3% max of the economy–and we’re talking about all the innovation that moves through university research to any sort of product.
A National Academy of Sciences study found “no reason to believe that either governmental retention of title or routine retention of title by individual inventors would yield more commercial applications or achieve a better balance of the public’s stakes” than Bayh-Dole.
Of course, that study ignores the system that was in place before Bayh-Dole. They compare a caricature of Bayh-Dole (the faux version) to a caricature of history (a made-up version) and find Bayh-Dole presents a better face. Well, now. The Academy study assumes that faculty need a monetary, monopoly incentive to continue their research. It can’t be that they would pursue research and teaching because they have already chosen to do so, and that in publishing, making data available, consulting, and instruction faculty do their role to make their findings available and useful. On top of that, faculty have available (even before they were compelled to use them) resources to assist in patenting and licensing, when patenting and licensing are needed to encourage the use of new findings.
But the Academy committee sets this all aside and decides that money is the most important incentive, and not just money, but money from monopoly positions shared with industry and with capitalists looking to carve out a share of an industry for an upside. It’s a nice apology for the importance of the state helping capitalists find new things to speculate in, but it’s pretty damning of the thought leadership regarding what should motivate people choosing academic research as a profession. The claim that the only or best motivator is money is itself a bit of idealism. It’s not a matter of avoiding idealism; it’s a matter of what ideals one chooses to serve, and the consequences of doing so. Perhaps the idea is that self-interest means money (doesn’t, but let it go), and thus, when people pursue money, an invisible hand makes all things right with an economy. So strange.
The Academy study presumes the version of faux Bayh-Dole invention ownership that the U.S. Supreme Court expressly rejected. The Academy study ignores the private network of patent management and the IPA system that were in place before Bayh-Dole. It’s as if these things didn’t exist. It’s easy to favor the thing you have over the thing you don’t have any clue used to exist. The Academy committee does not consider any evidence for Bayh-Dole’s actual commercialization success rate, nor show any awareness of the IPA system that Bayh-Dole displaced. They even point out such data are lacking. Lack of evidence doesn’t mean there isn’t any. The committee goes out of its way to argue that academic authors (“qualified ones, ones not of the ribble-rabble that might upset Bayh-Dole) should have access to government-held secret data on Bayh-Dole’s rate of practical application.
The Academy study concludes that it doesn’t have good evidence for Bayh-Dole’s success:
Empirical research on universities’ patenting and licensing activity has relied almost entirely on what we consider to be a seriously deficient set of criteria represented by the metrics used in the annual survey of technology transfer offices by the Association of University Technology Managers . . . rather than any direct measure of public availability of the patented technology and benefit from it, economic, social, or otherwise. Studies have also paid little attention to different roles of patents in different fields of technology.
In other words, the Academy committee listened to folks tell revisionist stories about Bayh-Dole that make Bayh-Dole sound better than nothing, couldn’t find evidence that Bayh-Dole has had a bad effect, and admits they don’t have good information regarding Bayh-Dole.
The Academy committee was asked to consider whether faculty ownership of their inventions was better than what they took to be Bayh-Dole’s shifting that title to institutions. That, too, was a false dichotomy. Bayh-Dole doesn’t shift title. And faculty ownership of inventions does not mean that they have to do everything themselves. The Academy committee worked in a cocoon of confirmation bias and selective history and concluded that they should stay there.
At least sometimes, some university bureaucrats will be better than some faculty inventors acting on their own. The question is whether university bureaucrats are in general better than any other possible brokers, or better than direct relationships between faculty and companies–collaborations, you know–no brokers needed. Bayh-Dole boils down to “For every love story, there must be a Pandarus.” And even if we are restricted to university bureaucrats, why should a faculty member at university A be restricted to using the services of bureaucrats at university A? Why not university C or university Z? If the goal is transfer of technology, why must things be so horribly fragmented and provincial?
The Academy ends up making a defense of university technology licensing offices. That is, Bayh-Dole is important because it has induced university administrations to spend money to create licensing offices–where before licensing moved more efficiently and at less cost through external agents. Bayh-Dole is a law for middleman welfare and subsidies for wealthy investors. No wonder the Academy wrote up a report to keep the bloat in place, even lacking any evidence that Bayh-Dole had delivered on its claims. And it makes sense that PhRMA would cite the Academy report.
To ensure timely and effective commercialization of federally funded research, Congress built in safeguards through a provision of Bayh-Dole that grants the federal agency funding the research a limited right to “march-in” and require the owner of a patent developed through federal funding to grant additional licenses to the technology.
This is too funny, written so matter-of-factually. Keep in mind, Bayh-Dole was drafted Norman Latker, a former university patent manager ensconsed as patent counsel for NIH, along with Howard Bremer the head of what would become AUTM. Latker had to have someone else type up the draft of the law for fear that HEW officials would trace the copy to his government typewriter. Once Bayh-Dole was passed, Latker moved over to the Office of Federal Procurement Policy to write the implementing regulations (for which he was only partly successful, as it would appear). Some of the bio is here. But what Latker and Bremer did was to gut the public oversight that they had previously built into the IPA program, which was constrained to follow the authority and limitations of the Kennedy statement on government patent policy. Instead, with Bayh-Dole, they replaced an executive order with a Congressional policy (35 USC 200). Then they set about removing all forms of public oversight–usage reports were made secret, exclusive licenses did not have to be justified, practical application did not have to be achieved timely or ever, march-in criteria were changed from reasonable terms to reasonable availability, and federal agencies were excluded from marching in just because they had determined the public welfare would be better served without a monopoly–that is, without abuse of the patent system, monopoly pricing, and the like. Bremer later took credit for gutting the march-in procedures, noting that he was “heavily involved in all facets of the process” by which Bayh-Dole was drafted, passed, and implemented:
As a consequence, much greater attention was given to the regulations by a university group which built into the regulations protection against both arbitrary exemptions to the law at agency discretion and to the exercise of march-in rights.
It is, then, somewhat funny to find PhRMA claiming that the march-in procedures act at all as safeguards of the public interest–certainly they are safeguards of private, monopoly-thriving interests. Under the IPA regime, a university had to achieve practical application or lose its status as administrator of the patents on a subject invention. Under Bayh-Dole, all that’s required is that there is an expectation of “effective steps” toward practical application. Not even the demonstration of steps and a showing that they are effective. Some march-in criterion.
The executive summary then turns to spin on these march-in procedures, arguing that efforts to use march-in to manage monopoly price setting for drugs based on subject inventions is “misguided.” I will deal with that claim in a separate article. It turns out, given how the march-in procedures are set up in Bayh-Dole, that PhMRA is right on this point. But they are wrong in thinking that’s the only point of attack to their unconstrained use of patents to subject inventions.
To mop up, here are a few additional passages from the body of the PhRMA report, just to show how deeply the faux Bayh-Dole narrative runs in the AUTM-industry sphere of influence.
University research and start-up companies, which rely on Bayh-Dole’s incentives and a partnership model between academia and the private sector, have become an engine for regional economic performance and growth.
This is an unsupported assertion made by an AUTM official. And it’s more switcheroo. University research does not rely on Bayh-Dole. Bayh-Dole forces federal agencies to impose requirements on the research. And the startup companies in the health sciences and basic sciences could just as easily have relied on the IPA model in effect before Bayh-Dole.
For all that, the primary incentive in Bayh-Dole for faculty inventors–again, eliminated from PhRMA’s account–is that both government and university leave them alone, in which case they become small business contractors–with fewer obligations than other small business contractors. The best deal in Bayh-Dole, where the major incentive is, is to leave inventors alone until they don’t want to be involved. Then folks can intervene. But PhRMA is dedicated to cutting the inventors out, just as the Academy committee was. A startup company formed by the direct assignment of inventions from the inventors to the company without passing through the ownership of an often balky, power-loving, risk-averse institution is necessarily more efficient. Why would PhRMA then advocate for the faux version of Bayh-Dole, in which university administrators must always intervene?
While AUTM folk may claim that startups are an engine for regional economic growth, where are the data to back the claim? How many of those startups are linked to subject inventions? How does exclusive licensing to a regional biotech startup in Ann Arbor matter for drug companies in New Jersey? To lab medicine departments in Seattle or Palo Alto? No information. Just assertions.
The importance of start-up firms to regional and national job creation is substantial—it has been estimated elsewhere that start-up businesses are a key driver of job growth, accounting for 70% of gross job creation.
More switcheroo. Where is the connection between Bayh-Dole as a cause of startups and the importance of startups generally? Ah, an estimate. And what good is an estimate? What did Bayh-Dole do for Google that it didn’t do for Facebook–both student-started, university-based enterprises. And even if startups account for lots of job creation, how much job creation comes from university-related startups? (and when does a startup cease being a startup–does AUTM count Google jobs forever?). And of those university startups, how many involve a subject invention? And of those startups, how many ever achieve practical application–meeting the five points of the definition in Bayh-Dole? We don’t know, because no one bothers to find out. All we have are wishful assertions made as fact, without a regard for the truth–i.e., bullshit. Worse, because it is made with such an air of authority, where we might call it corruption of a sort, or just guardian syndrome–deceive for the sake of the cause. Or bluffing–anything goes when the truth is not expected.
As AUTM’s President, David Winwood has noted, “when academic research yields a new idea, that idea often leads to a new startup company and then to new products in the marketplace. These ideas have the capacity to save lives, improve the way we work and play, and boost local economies—from seed varietals for our farmers to improved treatments for obesity and diabetes. Time and again these companies blossom, grow and stay in our local communities enhancing economic development.”
Lots of weasel amid the hooey. Is there any support for the “often” or the “time and again”? No. How often do university startups “grow and stay”–other than, say, in Silicon Valley and around Boston? And of those companies–and this is the Bayh-Dole question–how many are based on a subject invention and achieve practical application? We don’t know. Based on my experience, about half of university startups drop the licensed stuff for their own stuff; about half of university startups that are acquired drop work on their licensed stuff; and of the startups that are modestly successful or better, most move away from isolated communities to tech centers and markets suited to their business. They don’t enhance the local communities in which they started nearly so much as they do wherever they happen to land. But again, where’s the data that looks at the actual impact, and how in any way Bayh-Dole matters, or that a monopoly right matters, or that practical application of subject inventions is much ever achieved?
Bayh-Dole’s impact on start-up activity across all industries is substantial:
There’s no evidence given for the bold-faced claim. What follows is entirely skew–we get no information about the distribution of startup companies, and no account of how Bayh-Dole contributes to these startup companies, let alone “impacts” them in a “substantial” way.
In 2014, 909 startup companies were formed as a result of Bayh-Dole and technology transfer activities, 702 of them having their primary place of business in the licensing institution’s home state.
This mistates the stats–AUTM’s survey does not differentiate subject inventions from other inventions. The startup statistics do not show any indication that Bayh-Dole played any role. Transfer to startups does not even have to involve patented inventions to count in AUTM’s definition of a startup–and there does not even have to be a license. The startup just must have been formed in anticipate of a license. And of course, where universities manufacture startups, they will have a home address near the university. The University of Utah produced 20 startups a year–most with an address at the university’s tech transfer office. The University of Washington for multiple years simply faked the startups, counting any small company it had some dealing with as a startup, even when started years earlier, or in another place, or around other technology. And AUTM has no information whether any of these 909 startups for 2014 have achieved practical application–the goal of Bayh-Dole. It’s just quantitative rhetorical bluster. Or, corruption.
11,210 startup companies were reported as having been formed between 1980 and 2014 as a result of technology transfer activities–in 2014 alone, these firms along with other Bayh-Dole licensees introduced over 960 products across a range of technologies.
More AUTM spin quoted as fact, this time through Wendy Schacht, who writes whatever presents as plausible. There’s no connection between Bayh-Dole and the startups. No connection between subject inventions and the products on offer. And even then, if we run the numbers, we have an 8.5% commercialization rate, if all the products are based on subject inventions and each product is based on one subject invention alone. That’s the outer bound. In any event, the activity of commercialization described here is speculation. The output of commercialization is products. When those products provide a benefit to the public on reasonable terms, then there’s practical application and Bayh-Dole’s objective has been met, at least for those subject inventions.
Again, and worse, consider a study that argues that U.S. startup rates are falling. Why? Here’s a Washington Post article (including citations to research by folks who have been associated with the Kauffman Foundation, as I was for three years of grant-funded research, just to be upfront about it):
But ever-multiplying regulation does hamper business activity. The Economist magazine argues that the U.S. economy has grown less competitive in the past 20 years. After a wave of deregulation in the 1980s, red tape has proliferated, licensing requirements have expanded and legal costs have risen dramatically.
Perhaps Bayh-Dole is part of that wave of deregulation. But the result has been institutional capture of opportunity, and with that has come bureaucracy. The winners? The big companies. You know, PhRMA members. Bayh-Dole is a government-imposed anti-competitive regulation that preserves the interests of the incumbents, even as they cite statistics about startups, without mentioning that startups are falling. And to put the numbers cited from AUTM in perspective, there have been about 500,000 new businesses started every year in the U.S. since 1980. That’s somewhere over 15 million startups in the 1980-2014 time frame–and 11,000 of those were related to universities, and some portion of those involving subject inventions. Let’s round to the nearest hundredth: Bayh-Dole contributed 0% of the U.S. startups during the period, by number, and that’s assuming Bayh-Dole accounts for every last one of the university-reported startups. Yes, the federal government did sponsor research at Stanford leading to the page rank metric used by Google–so that’s something. But as a number, the university patent-licensed contribution to national startups is 0%. In its way, that’s an impressive number, too. Just not one PhRMA wants you to contemplate.
Universities create an average of more than two start-up companies each day, and these university-based start-ups have longer life spans and raise more capital than non-university-affiliated start-ups, meaning they support job creation and sustained economic benefits to local economies.
This comes from a paper by Valerie McDevitt (appointed AUTM executive director in 2014) and a list of mostly AUTM officers or former officers. The citation that backs these figures? Another article by an AUTM officer. And the backing for that assertion? A 2008 AUTM survey on startups–2 per each working day–and nothing for how long they last or how they get funding. Meaning–it’s just spin–in another article aimed at bashing the idea that faculty should have any role in how their research is exploited. University administrators, it is agreed by PhRMA, the National Academy of Science, and AUTM, are much better at cutting monopoly deals than are faculty. Universities created maybe 500 startups of the 500,000 or so that year, 0.1%. It is no wonder that the McDevitt et al. paper tells the same revisionist history–that university tech transfer started with Bayh-Dole. Research Corporation and the IPA system don’t exist. The economic impact is great if you mistake hand-waving for fact, prefer fallacious reasoning to substance, and like to view things out of context, for clarity, you know.
More specific to the life sciences industry, Bayh-Dole has become a critical element in the rise of “biotech clusters” (i.e. geographic concentrations of biotech firms actively exchanging expertise, human capital and infrastructure, often located near or including universities) and other mechanisms that help pave the way for technology transfer from academia to industry.
Again, there’s no evidence that Bayh-Dole has led to the formation of biotech clusters. There are many other ways that expertise and entrepreneurs are attracted to universities doing research in biotech–including obtaining a share of subcontracts on federally funded research, selling research-related tools to universities engaged in research, and building on published findings that don’t involve patents at all. For all that, there’s no indication that licensees of subject inventions are local to a university at all, other than university startups, which are mostly applied research repackaged as companies to gain the advantage of siphoning away state early-stage investment capital and SBIR funding from the small companies also in the area.
One more bit follows, with a long quote, but just to show the degree of distortion offered up by PhRMA:
As summarized by one researcher, “In recent years, there has been a substantial rise in the rate of commercialization of university-based technologies—through patenting, licensing, research joint ventures, and the formation of startup companies. We have also witnessed an increase in investment in science parks and other property-based institutions that facilitate the transfer of technology from universities to firms . . . most commentators attribute a substantial portion of this activity to the Bayh-Dole Act of 1980, which dramatically changed the incentives of U.S. universities to commercialize their intellectual property. Bayh-Dole instituted a uniform patent policy across federal agencies, removed many restrictions on licensing, and most importantly, allowed universities, rather than the federal government, to own patents arising from federal research grants.
The researcher? Actually, three, and they are editors providing an overview of university tech transfer globally, not reporting their original work. The quote is the opening sentences of their introduction, with no citations to back the claim. It’s like talking the weather. The book includes articles covering situations in other countries. The omitted stuff (where I’ve boldfaced the ellipses)? Ah, that’s the part where the editors cite David Mowery’s work to show that the claims made for the faux history of Bayh-Dole starting it all are wrong. The introduction restates the context in which the book’s various articles operate–it is not establishing a conclusion drawn from that research!
The editors then are content to repeat the common misstatements of what Bayh-Dole did and covers–that’s appropriate for an introduction to the setting in which the papers are presented. Bayh-Dole concerns only patents on subject inventions, not all intellectual property, and it’s not “their” intellectual property anyway, until the universities acquire it. Bayh-Dole, in its faux version, claimed to strip invention ownership from researchers and place it with the universities. Perhaps making a virtue of looting personal property is an incentive. I hadn’t thought of it that way.
Before Bayh-Dole, federal agencies permitted inventors or contractors to own inventions, sometimes with a public covenant. They did so when a contractor had a commercial position, when the government wasn’t funding to produce a product, when doing so did not give a contractor an unfair advantage in competing for federal funds, and when, under an IPA, a university undertook to administrate an invention with a public covenant that favored non-exclusive licensing, short terms for exclusive licenses, and the opportunity for the government to intervene and take over administration if the university could not achieve practical application in a reasonable time–such as three years from patent issue or seven to eight years overall. The federal government did allow contractors, including universities, to own inventions made with federal support.
It’s just that the government also attempted to mitigate the exploitation of its support, to prevent contractors and their business partners from holding patents to obtain unfair advantages, to disrupt the research of others, and to create monopolies that gouged the public. It is true–PhRMA does get this part correct–that Bayh-Dole did remove many restrictions on licensing, and along with that removal, almost everything associated with federal oversight or public accountability. It was the federal oversight and public accountability that Bayh-Dole took down and replaced with an approach having none to speak of. No wonder PhMRA likes the present situation.
Perhaps that’s enough. You get the idea. The illusion of facts. Nothing to back claims up but statements by folks happy to be corrupted in the service of their jobs. PhRMA presents no evidence that Bayh-Dole has been successful according to Bayh-Dole’s own statement of policy and objective. It’s clear that PhRMA likes the present state of affairs anyway. Why? Because the pharma industry benefits from federal subsidies, paying at best a 1% royalty on sales in return. Look through the text–PhRMA never distinguishes Bayh-Dole subject invention activity from other activity. They attribute all university licensing activity to Bayh-Dole. This, too, is nonsense.
There was collaboration between academics and industry before Bayh-Dole. There was technology transfer before Bayh-Dole, and it operated with better success rates whether done by invention management organizations or the federal government. What pharma did not like was the federal government funding pharma-like research, taking ownership of inventions made in that research, and making those results available to all. Pharma wanted monopolies, with monopoly pricing. So they cut a deal with universities to work around the government patent commons. That’s bad enough, but to keep their happy scheme, they have to keep Bayh-Dole in place and so must impose its awful practices on everyone else in every other industry. The key thing is to prevent public oversight–and for that Bayh-Dole’s hatchet job on march-in rights suits just fine. Nice stuff.