Before their efforts turned to Bayh-Dole, the folks advocating invention mining at universities aimed to make the IPA a standard across all government agencies. Yes, “uniform” policy was what they argued for, but that was the politics. What they wanted was to improve their ability as patent brokers to gain access to inventions made with federal support–and which generally flowed to the public domain or to a federal patent commons. In the federal patent commons, 4% to 5% of patents ended up being “commercialized” in some way. We don’t have figures, however, on how many of the underlying inventions were being used. The 4% figure represents about the same rate of “commercial” use as patents from all sources have generally. That is, without much trying, the federal government was doing about as good as independent and corporate inventors. Contemplate that.
Meanwhile, under the HEW IPA, which over 70 non-profit institutions had signed off on (and another 18 apparently held in an early, customized form), the commercialization rate was, ahem, also about 4% to 5% (4 products from 96 inventions). The head of Research Corporation, Willard Marcy, testified that only “50 to 67” of 2,000 university discoveries per year (in 1978) “might get to the marketplace.” Or, 5 to 7 or so in 200 patentable inventions, or 3%.
When Norman Latker, patent counsel at the NIH, made an effort to have the General Services Administration adopt a template IPA for all federal agencies to use, the Senate Select Committee on Small Business stepped in to question the move. Not only was the GSA not the proper authority for making the decision–that would be the Office of Management and Budget–but also the regulation was being adopted without public notice or an opportunity for other than universities and related parties to comment on the matter. Senator Gaylord Nelson (Wisconsin) chaired the Senate committee and wrote to Lester Fettig in the OMB with his concerns after the committee had held hearings on the issues.
Having pointed out problems with the IPA template as well as the procedural problem, Senator Nelson then went after the logic and the data. The university folks, he pointed out, kept changing their position, dancing about as it were:
The universities claimed the government did a lousy job with commercializing patents, but then could not show that their own activity under the IPAs was doing any better. To cover that, they advocated for “technology transfer” and the idea that the proper and most important thing was that discoveries made in the laboratory become commercialized–that is, secured by patents and licensed exclusively to companies. In play was a convenient conflation of whether the business of patent brokering was satisfied by licensing–the broker made enough money to break even or a little better, and speculated that it could make more–or by practical application–the company worked the invention so that the benefits of such use were reasonably accessible to the public (to paraphrase the Kennedy patent policy definition of “developed to the point of practical application).
Pause here. The patent broker operating model was to cover its costs in the initial licensing fees, and make its upside on downstream fees and royalties. In this model, to get a big upfront fee requires (often) an exclusive license. That is–to get sufficient income to cover the costs of filing patent applications, the broker prefers to create a private monopoly, regardless of whether such a monopoly is needed to promote use or to “call forth private risk funds” that otherwise would not listen. The monopoly provides the rationale for sufficient upfront payment to cover the marginal costs of patenting the invention in the first place. Once the monopoly has been created and paid for, the patent broker has met its primary goal–breaking even. The patent broker pays itself, and the inventor and university administrators and the general public have to wait until the exclusive licensee creates a product–which often is never.
That is, the operating model that has been adopted–using exclusive licensing–pays off for the patent broker ahead of the public. No wonder that university patent administrators were attracted to the idea of bringing such work in-house. As long as the licensing office broke even, there would be more royalties for the university (since it would not have to share with an outside foundation–often that was a 40-60 foundation-university split). The argument could slip even further to a rationale to subsidize an internal patent mining operation on the prospect of a big hit–even one every decade was more than plenty to recover the subsidy (or so the thinking went). That is, one invention of 1,000 or 2,000 licensed to become a product that paid substantial royalties made this business model successful. No matter that 999 or 1,999 inventions were appropriated, sequestered, delayed, or licensed exclusively into oblivion. All that matters are the success stories.
If the operating model of the patent broker, by contrast, required payment only when a licensing organization used an invention productively–whether in research, internal operations, or to create a product for sale–then the patent broker would have to focus on payments for use rather than payments for monopoly followed by (maybe) payments for use (with the hope these payments will be higher because they arise in a monopoly that can dictate price without competition). If the model requires payments only with use, then non-exclusive licensing or staggered sole licensing becomes much more attractive. (In a staggered sole licensing strategy, licenses are awarded one at a time, so that the first adopter gains some time advantage over later adopters and has competitive reasons therefore to reach the market first.) Most of the great early inventions–electrostatic precipitator, creating vitamin D in milk (and God knows what else) via radiation to end-run of federal regulations, warfarin, gene-splicing, expression of polypeptides in yeast–all were licensed non-exclusively. Funny how that seemed to work out.
The IPA model, dictated by requirements of the Kennedy patent policy, provided for a short time to achieve practical application under an exclusive license–about six years, maybe a bit more–and furthermore limited the length of any exclusive license to eight years–less if practical application came about sooner (three years from date of first commercial sale). The result was that whatever a company paid for a monopoly position, it did not have the entire term of the patent to futz about “developing” the invention. Think of it from the happy monopolist point of view. One pays the university to gain exclusive access to an invention–no one else may make or use the invention for any reason, not for research, not for their own use, and certainly not to create product. The poor sots cannot even make or use in anticipation of the invention coming off patent (witness the fuss over gearing up to produce generic drugs). And now that one has acquired the monopoly, development can take place as whatever pace is “commercially reasonable.”
Given that the exclusive licensee will pay a royalty on sales, once there are sales, and perhaps big milestone cash payments as well, then what’s “reasonable” includes how the new product will affect existing products, and how long any other company will have to work to develop a competing product once the new product is on the market. Why undercut sales that don’t involve a royalty (even a 1% or 2% royalty) with a new product that does pay out a royalty? The effect of a full-term patent monopoly is that a company has a few years to get things staged for maximum company profit–not, say, maximum patent broker revenue.
The reasonable business focus of the exclusive licensee is to limit the exposure to the patent royalty in whatever ways the licensing agreement does not forbid. Whatever good faith the exclusive licensee has to make the deal work for the patent broker is bounded by what’s “commercially reasonable.” Is such “commercially reasonable” thinking in the public interest? That’s one of the questions that folks considering reasonable pricing might contemplate. Another is whether exclusive licenses, even if they do result from time to time in new products also delay the introduction of those products for a matter of years over what would happen with shorter terms of exclusivity. The exclusive license for the full term of a patent delays commercial use, precludes most research use, and prevents competition. A patent broker makes money upfront (payment for the right of monopoly) and makes better money if the monopoly runs for the full term of the patent, and makes best money when the monopoly covers all fields and all uses. Payment for use is always a “success” for the broker, even if the use is delayed, the market stifled, and the public pays a higher price than otherwise to gain access to the product.
Perhaps you see the difference now between a license based on a monopoly and a license based on use. The patent brokers worked out that they could make a living with patent mining so long as they could license enough patents to recover their costs. The best way to do that (to their thinking) is an exclusive license. But such exclusive licenses shift the risk from the patent broker to the rest of the university and to the public. The primary beneficiary of the monopoly is the company, which can then adjust its development to maximize its own profits across all products rather than rush something new into the market at the expense of its present investments. Perhaps you can see why university “success stories” never indicate the time to market. The exclusive license doesn’t get things to market sooner–it prevents others from competing to do so.
In any case, Senator Nelson found the university patent brokers’ claims about technology transfer unconvincing:
That is, just because the government’s commercial licensing rate was 4% or so doesn’t mean that universities, even trying, can do better than that. What if the problem is not whether people are trying to create monopolies or not–what if the problem is inherent in what kind of work the government funds, and what sort of inventions are “called forth” by requirements for compliance with invention reporting? What if the problem is a matter of what industry is prepared to take up under the terms of patent agreements compared to everything else that industry might adopt as preferred practice? Bring more material within the sphere of monopoly-making does not do anything to address the underlying conditions that lead to adoption of discoveries for use, let alone turning those discoveries into products–and especially doing so in a manner that excludes all other uses while the attempt is made.
It turns out that the university licensing rate is now more like 0.1%–10x worse than the government rate, 10x worse than the IPA rate, 100x worse than the claims made about the private patent broker mediated rate. Why the differences–the private patent brokers were highly selective, worked generally on a mutually negotiated basis, and focused on industries in which they were known–and often started with a “big hit” invention that got everyone’s attention. Trying to turn things into a “system” and then making the system compulsory, and then turning the system to monopoly licensing all have contributed to making the approach expensive, wasteful, and inefficient.
We get, then, to the “principle of technology transfer” and the need for “greater cooperation between Government, academia and industry to move discoveries out of the laboratory into the marketplace.” In 1978, “technology transfer offices” were ones that largely did invention mining as a front-end for off-site patent brokers. The broker could not be on-campus every day, so the first angle on getting universities to subsidize the patent mining operation was to argue that they had a responsibility to find inventions and get those inventions reported to a patent broker. Of course, if federal funding was involved, the patent broker could have a time trying to get a federal agency to release its claim on the invention–unless the invention was funded by the HEW or NSF under an IPA.
The one area of patent mining practice largely unaffected by the Senate committee delaying and ultimately blocking GSA adoption of a uniform, government-wide IPA template–and therefore, later, largely unaffected three years later by Bayh-Dole–was biomedical. Whatever else one might want to claim for Bayh-Dole, almost nothing changed for biomedical inventions made with federal support when the Bayh-Dole Act went into effect. Inventions funded by HEW/NIH still could pass to university ownership and from there to monopoly licenses–and even the time limit on exclusive licenses was later lifted by amendment. Anyone looking for the effects of Bayh-Dole in the area of therapeutics is a lost, misdirected soul. The effects of Bayh-Dole are to be found in the other areas of university research–engineering, physics, information technology, semiconductors, and the like.
The “principle of technology transfer” itself is an iffy concept. The foundational use in university practice involves invention mining for the purposes of feeding patent brokers. The “technology” that gets “transferred” is a right to patent–a right to create private monopolies. There is no “transfer” of the underlying “technology” that surrounds an invention–that gets “thrown in,” often without saying. Further, that other technology–often created by others, not the “named inventors”–is given no value, and may also be within the scope of exclusivity (as a matter of drafting the scope of the exclusive license) even if not expressly covered by a claim of the licensed patent. And for all this, the transfer is not to the public or even to industry–it is to the patent broker. Thus, the need for greater cooperation between Government, universities, and industry comes down to a need for the Government to give up claims to inventions so the patent broker can obtain ownership rights from the university or its inventors and turn these into commodities that allow the patent broker to stay in business making private monopolies. One a decade is enough. This is a good deal for the patent broker–administrators and patent attorneys always make their money. This is the first premise of a business model that is not a public interest model. If the trustee makes money at the expense of the beneficiaries, what sort of trustee have we got? A self-interested one, to be sure.
The idea that discoveries in a lab ought to “get out of the lab” is a fine one. Publication, teaching, and collaboration often see to that. These days, what’s more difficult is to get data and tools released as well. Often these are suppressed to prevent “competition” for federal grants or to challenge published findings–much to the consternation of companies aiming to rely on those publications and those same data and tools. There is even a good argument–going back over a century to Frederick Cottrell–that from time to time, the pathway out can be mediated by a patent, and then negotiating with a broker to handle the strategy makes some sense. But the idea that all inventions must first pass through the hands of a broker, and that the broker must be dedicated to create monopolies, and that success is a monopoly commercial product–that’s not at all obvious or supported.
That is, the idea that the primary role of university labs is to provide a supply of inventions to patent brokers makes no sense. Or, it only makes sense if you are already aligned with the difficult life of the patent broker. If the patent broker is pitched as good-hearted, thinking only of the public welfare, and that commercial products best serve that welfare, why, then patenting the heck out of university discoveries is the way to go. Any change in this idea–that the patent broker is not so well-meaning, that commercialization is not the only or even the best or even a good path, that monopolies for the full term of a patent work against early adoption and use–and the whole thing collapses.
Bayh-Dole was created as a welfare law for private patent brokers. In that regard, the law has been wildly successful. The details of how the move was made from the IPA (which had its own sneakiness) to Bayh-Dole bear out why patent brokers fight so hard to keep Bayh-Dole. The change in the burden from the monopolist to the agency to demonstrate public interest has not been met. The horrible complications in march-in procedures introduced by universities and their patent brokers. The limitations on march-in procedures, so they address only “availability” in limited circumstances (and failure to require U.S. manufacture) and not “free competition and enterprise” or “unreasonable use” or “collaboration.” The removal of oversight for exclusive licenses. The secrecy on utilization and status reports. The convoluted wording that made it plausible that Bayh-Dole handed ownership of inventions to university administrators. The conflating of small business contracting with university contracting, to make university research appear to be common with commercial contracting. All of these things (and more) are the work of a patent broker lobby building a nest in the largesse created by the federal government commitment to university research. They were clever enough to pitch the idea as a response to government inadequacy and a vision for “technology transfer”–both of which failed the university mandate to take on federal research in the first place, but sure sounded good.
There are great reasons for university folks to teach what they discover–to put things in a condition to be used and assessed by others. There are also great reasons for professionals to be around to assist in this effort, even professionals who know intellectual property (and also NIPIA–non-IP intangible assets). But there is no good reason why the first, primary, only pathway is from lab to monopoly. Lab to lab, lab to lab in another area of study (as what happened when the biologists made the rounds begging computer scientists to help. As one presentation I attended put it, “It takes only three months for you to learn the biology, and it will take us ten years to learn the computer science”). University lab to industry lab. University lab to standard or platform or library or tool set. University lab to graduate students leaving for industry jobs. Every so often, university lab to product. Every so often within that scenario, university lab to time-limited monopoly. And sometimes, even, just a monopoly when that’s the only alternative left.
Consider this. There are two massive traditions that have arisen from federally supported university research. The first is the digital computer, the internet, and related information technology. Starting in the 1940s and gaining speed in the 1970s, the computer and internet were based on open standards that created a platform that has created massive opportunities for wealth and what most would take to be public benefits (I’m still on the fence about Facebook). Competition takes place within a set of collective standards that define the opportunity. Build a device that’s faster and more reliable, and you will get business. Build one that breaks a standard, creates a monopoly, and isn’t interoperable, and you are screwed.
The second is the health care marketplace shaped by a focus on therapeutics, also starting in the 1940s and gaining speed in the 1970s. This market was based on private monopolies and when the federal government resisted, it was beaten back, leading to a vision of patents as the mechanism to route all university inventions for inspection by monopoly-minded firms. The competition is to find a broker willing to cut a monopoly deal. The result has been the creation of huge wealth for pharmaceutical companies, for venture capitalists selling to them startup companies loaded with monopoly rights, and for the shareholders in back of all this. The result for the public is high drug prices and a focus on drugs rather than prevention, cures, and advanced science of biology as the state of the art. And, of course, drugs that manage widespread chronic conditions are the new “wonder” drugs–the sweet spot for profits. The drug that makes a fatal condition chronic is the best of all.
If the pharmaceutical industry wishes to play at monopolies on therapeutics, that’s a private decision. That the federal government goes along with the idea is a matter for public discussion and does not depend at all on the way monopolists go about funding their opportunities. But it is another entirely to cover one’s tracks in all this and force the same patent brokering on all other areas of research activity and public use, as if what is good for big pharma and its patent brokers is good for engineering, nanotechnology, environmental technology, energy, 3d printing, software, computing, or anything else. The idea of a “uniform” policy that gives patent brokers first access to the results of university research is pernicious. The primary purpose of federal patent policy with regard to university research cannot possibly be to provide convenience to patent brokers. This is not to argue that federal agency practice was well formed across all agencies. It’s just that the technology transfer that happens as a result of patent rights is hardly the most compelling form of transfer that we can imagine–or document. And certainly when the default operating model of such patent-based transfer is compulsory taking of invention ownership combined with the creation of private monopolies at the expense of all other use and competition, we have a huge problem on our hands.
The problem arises with the influx of federal funds to universities following Vannevar Bush’s call for the federal government to support basic research and training at the frontiers of knowledge. That sort of work required special handling, not the typical handling of procurement nor even of research conducted for the benefit of any particular agency mission. The mission of the National Research Foundation would be to create the conditions for the rapid advance of scientific frontiers. How the knowledge from such advances moved out into other areas of practice–that was a matter for the engagement of creative minds with an established order. The response of the targeted established orders–starting immediately with the military and the biomedical industry–was to use procurement instruments for their research and try to sort out what would work within such contracts. Well, it didn’t much work well, was inconsistent even within an agency, and tended to keep the government in control and industry dependent on government funding. Bush created a problem, and universities by not supporting his proposed solution–to keep university frontier research away from federal procurement–ensured that the problem would get worse, leading first to the IPA approach and when that failed politically, to the Bayh-Dole Act that imposed procurement and patents on every agency and every university funding agreement.
Bush turned his attention to information technology–writing “As We May Think” and “For Man to Know” in The Atlantic. These articles focused on how the outpouring of new knowledge might be recorded, transmitted, and made available for use. We work within that vision now. I write and you read using technology that came about because federal research built open systems and standards that industry then adopted and developed, competing all the while, fighting over IP positions, but only after the “essential claims” had been addressed in standards negotiations.
There are, perhaps, two great traditions in the funding of science as there are in politics (See Kagan’s “The End of Dreams and the Return of History.”) One values freedom, democracy, and is willing to accept popular thinking as a route to the future of community–try it and see, leave it if it fails, improve on it by choice. The other values autocracy, philosopher kings, elites with power creating order and efficiency and social just for all by degree. Markets and monopolies. Federal invention policy is not immune to these forces. The patent brokers showed up appearing to be in support of private initiative, non-exclusive licensing (that’s what they advocated publicly), and public benefit; in practice they desired compulsory access to inventions, monopoly creation, and a cut of the action they created. Even so, there appear to be roles for monopolies, whether statutory or defacto or patent-created. Even a standard can have the effect of monopoly, as can popular consensus, as can a dominant religious belief or a madness of crowds. The issue, then, is not the elimination of market or monopoly, but a steering among these choices, interrupting consistency with change, and beating back change with consistency.
A “uniform” policy of federal patent management does not have to require that patent brokers enjoy a monopoly business. The Kennedy patent policy was “uniform”–just in a way that did not give patent brokers easy access to inventions they wished to deal in as monopolies. And even the Kennedy’ patent policy’s uniformity conflated the special case of university frontier research with any other sort of services contract.What we have lost is the special attention to that university frontier research in the effort to achieve administrative convenience (and dominant power) for patent brokers and a disengagement of federal grants officers with the disaster of trying to “contract by regulation.” What we have lost is the attention we should have been giving to university frontier research–exploration on the edges, the skunk works led by people with deep, rich, playful insight, for which no group of peers nor committee of reviewers of proposals nor federal agency program officers (bless their hearts) could possibly have a think to say with regard to relevance or merit. This was what Bush was after, the problem he saw, the reason he wanted an agency outside the scope of all other agencies but still accountable to Congress and reporting to the President, and agency for exploration research, funded by the federal government, overseen by the best of citizen scientists, run by a director that reported to them, and with a patent policy that did not dictate ownership at all, but rather provided only that the government wouldn’t be sued for using what might be discovered.
That’s what we have lost in all this. That’s the discussion we ought to have. That’s where freedom ought to have more say than monopoly. That’s where the patent broker ought to leave well enough alone until asked to become involved and only then on the terms of those doing the asking. That’s were compulsory assignment policies are nuts, where laying out the course of study in a contract is foolish, and where review of merit tied to objective measures of what is trendy is worse than no review at all. I want to get it back–I don’t care about the hum-drum that university faculty might take on. I don’t even care that drug companies have worked an end-run about public policy that favors openness over monopolies, with universities getting 1% and the drug companies getting 99%, monopoly pricing and monopoly delays. I do care that uniform federal law forces everything else into this model, into the hands of patent brokers fully committed to creating private monopolies, disdaining everything else, and unwilling (and incapable) of creating university policies that would give expression to the open, democratic, competitive side of discoveries moving from lab to lab, from lab to use, from use to standard or product–or both–as the opportunity presents.