A couple of university front organizations, APLU and AAU, have published a two-page infographic that presents a glowing picture of Bayh-Dole and university patent-based commercialization. Titled “How Technology Transfer Transforms Society,” the infographic confidently presents arguments for university commercialization practices. Let’s have a look.
University research may indeed “drive” some innovation. But university research is not the only driver, by far. Furthermore, the light bulb is not one of the results of university research. Perhaps APLU could have picked something that really has been produced by university research. Tomato pickers, say. Invented at UC Davis, and “transformed society”:
There are many stories we could tell about the harvester and its impact on California agriculture, California eating habits, and California’s farm labor. One of them would explain how it displaced thousands of mostly Mexican-American farm laborers in the 1960s, and then became the subject of a major lawsuit against the University of California, ultimately resulting in a new ethos of worker-impact-centered agricultural research on our campuses. Another would illuminate the lightning-fast implementation of the machines and the rapid changes they brought to farming and consumer practices.
One might also tell the story of how UC Davis then created new varieties of tomatoes that could be picked mechanically, starting the trend toward the red, tasteless tomatoes now available in most any grocery store.
Quick, pick out which story APLU and AAU would pick for their infographic.
There’s another problem, here, though, that’s deeper. Research may be conducted at universities, but is not the university’s research. It is ideas that people come up with while at universities. That’s what the light bulb is an emblem for–ideas. Those ideas don’t arise necessarily because of research in the abstract (the domain of administrators counting university beans), but rather ideas arise in people’s heads in a context that includes research–but also includes conversation, making things, consulting, musing about stuff. The ideas that matter might well be ones that recognize that the research activities aren’t going anywhere, but something else might. So APLU and AAU present “university research” as the driver (and thus claim administrative possession of the activity and its results), when the driver, apparently, is people with good ideas, some of whom are doing research.
Two claims here. The first one has no support. There’s no evidence that Bayh-Dole has “increased and accelerated” the transfer of “discoveries” “to the marketplace.” Before Bayh-Dole, universities claimed their commercialization rate for non-federal inventions was around 25%. Under the IPA program, universities reported a commercialization rate of just under 5%. Bayh-Dole was cleverly amended in 1984 to make all reports of invention use a government secret, but the few accounts of university licensing we have peg the commercialization rate overall at about 0.5%. Perhaps federally supported inventions are doing way, way better–unlike in 1978, when they were doing 5x worse. But these figures suggest that Bayh-Dole has actually significantly decreased technology transfer and slowed the transfer of discoveries to “the marketplace.” Every patented and unlicensed invention is a barrier to use–and university administrations now patent early and often. Every exclusively licensed invention that does not result in a commercial product is worse–it’s a time bomb of litigation in addition to keeping research findings out of circulation.
What’s also odd is the focus on “the marketplace”–as if the primary (or only, or best) purpose for research findings is that they must be used to create products. Many research findings can be used without going through a product stage–a company can take a process in-house to improve production or quality assurance. Diagnostic tests for diseases can be developed by any competent lab medicine clinic without waiting around for a commercial version. Some findings may become research tools; others, standards; others, part of a developing platform of technology that gains its value as additional contributions are made to it. Commercial versions of such technology may come later, or may not be necessary at all. It’s just silly to make it out that the only measure of university research findings is that university administrators create patent monopolies in the hope of royalty income.
Consider this: the unselective patenting at universities, combined with default monopoly licensing, creates a huge incentive in industry to ignore university research (mostly), to undermine it (because one is unlikely to be the favored commercialization partner), to design around it and thus render it useless (because one doesn’t care to “commercialize” it), and to design it out of standards (because it’s proprietary with no hope of commitment to a standard). Of course, this isn’t something that anyone at APLU or AAU wants to think about and certainly not something for their infographic.
Now for the second claim. It may well be true that products and processes may “emerge” from technology transfer and benefit the public. The tag “in a number of ways” is silly–especially since the second page of the infographic is dedicated to one way only. What’s left out is that new products and processes may also harm the public. Consider the problems with the hepatitis B vaccine, say, or the losses and opportunity costs incurred by the many failed companies built around the premise that some university-licensed technology was really good, when it turned out it wasn’t. All that investment that could have gone to something useful, but didn’t. Meanwhile, entrepreneurs trying to do things on their own get squeezed for attention by university shell companies slurping up available government funding.
There is another implication in the second claim, and it has to do with “emerge.” Technology may move from lab to market despite university administrative, patent-based technology transfer operations. That is, a university’s administrative transfer procedures often have nothing positive to do with whether the discovery is adopted by industry or benefits the public. “Emerge” here is a good verb–as in “emerge from the clutches of bureaucrats.” The infographic makes it appear that administrative technology transfer is the necessary connection between discovery and use. Often, it’s not. Even with the Cohen-Boyer gene-splicing patents, Stanford’s licensing officer agreed that the patents weren’t necessary. Industry was going to use the methods regardless. The patents served to extract money from industry for Stanford and the University of California. The patents may have had other uses, as well, but these aren’t acknowledged. Thus, we are left with the open question whether university technology transfer is basically parasitic on research enterprise and industry uptake, or whether it performs a valuable service, if not a necessary one. An honest infographic from APLU and AAU would ask the question.
We won’t cover everything here, and it’s not possible to check the claims because the infographic doesn’t document them. But we can notice that the claim that’s made is different from the process that will be described later–smart phone, weather report, MRI–these are products of multiple bits of technology from all sorts of places, but the infographic makes it sound like these are the result of university patent licensing. Perhaps a feature is–in the case of MIT, a number of features. The first patent I find for doppler radar issued to the US Air Force in 1965. One of the inventors, David Atlas, went on to work for MIT. So MIT might have been involved, too. But none of Atlas’s other patents on improvements show MIT as the owner. In any case, the origin of the idea was not university research. Some improvements, sure. Again, is university research the driver, or is university research something that gravitates to trendy things, hangs out some new features, and by the use of patents extracts payment for them? That’s fair, but it’s not the same thing as being the driver.
There’s a second thing here, and that’s the passive “that’s been brought to the marketplace.” A technology may be used in a commercial product, and thus be “in the marketplace,” but it may well be that the university’s patent positions have nothing to do with it. Consider how Caltech, among others, is suing technology companies for using (so it claims) its patented technologies without the benefit of some one of them getting an exclusive license first. Even Stanford, in its famous (and losing) case against Roche, argued that Roche must take a license and pay, even though Roche had successful product (a kit to evaluate the efficacy of HIV treatments) “in the marketplace” without any need for university technology transfer. These companies fielded products despite university patent claims.
As for MRI, Raymond Damadian did work at SUNY, but none of his early patents, even well into the Bayh-Dole era, show any evidence of federal funding. Odd, but one might think that SUNY would include a federal funding statement. No matter, the patents don’t even show SUNY ownership. Maybe all that happened later, after the patents were issued. By way of context, since 1976, there are 3,275 U.S. patents that include “mri” in the abstract; of these, 614 include “university” or “institute” in the assignee field, 129 of which include a federal funding statement. As a gross measure, about 4% of the patent activity is federally funded university research. It’s technically true that there may be some federally funded, university-patented technology in today’s MRI equipment, but it’s pretty ballocky to make it appear that the bulk of the technology comes from universities.
[There’s a similar argument for touch screens, which were invented in Britain, not the U.S. While Hurst and a student at the University of Kentucky invented a resistive touch sensor in the early 1970s, it’s not at all clear that they used federal funds. A touch screen based on resistive touch was invented later, at Hurst’s company Elographics. It’s a great story–a serendipitous discovery in research to study something else entirely. The touch sensor was a tool to help conduct other research. Hurst moved to Oak Ridge and developed a resistive touch screen at Elographics in his spare time. Hurst didn’t invent the touch screen, and the university patent was licensed (apparently) back to Hurst, where it remained as an interesting idea, but it wasn’t a touch screen. The university’s foundation could have as easily provided funds for Hurst and Parks to obtain a patent, in exchange, say, for a share of any income from exploiting the patent. No university patent licensing step would have been involved. The APLU and AAU infographic, however, makes it appear that an institutional patent on university research was the driver. Rather, Hurst’s insights and initiative were the drivers.]
“Countless” here is a problem. But the weasel words come next–“other technologies and everyday products.” It’s just that most of what university patent licensing produces are not “everyday products.” Yes, there may be many “other technologies”–but linking the two counts with an “and” and tying them to “countless” is misleading. It’s technically true, but designed to deceive an audience into thinking “countless everyday objects are the result of federally funded university research managed for patents and licensed to industry.” To be generous, that’s a matter for some proof. How likely do you think that proof is?
Notice that there are no inventors here–it’s all passive when it comes to inventors, all active with regard to “universities” and the “process” of technology transfer. Inventors, for all APLU and AAU may think, are faceless slaves on plantations run by bureaucrats or hens in office coops laying an egg a day until they get burned out. One might think that arguing that plantations are good for universities would be bad enough–faculty, in this view, work for administrators, who look out after the university. But the implied APLU and AAU argument is that plantations are good for society; they transform society (always in good ways) with patent licensing as the driver. What’s good about slave labor is good for America. Neat.
Inventors may seek patents–that’s the Constitutional essence of the U.S. patent system–but here in the infographic “universities patent” things, and then, because there are patents, there must also be licenses. University administrators patent things because they demand to own everything that inventors make (even if not a patentable invention, even if not an invention, other than “traditional” scholarly things like articles and poems). So the first act of APLU and AAU-style technology transfer is to cut investigators and inventors out of the activity.
The infographic doesn’t bother to point out that the primary energy of university licensing efforts is to secure exclusive licenses, and that those licenses often constitute assignments of the inventions. University administrators demand inventions from faculty, student, and staff inventors (and sue them if they don’t comply), obtain patents, and then demand that a company license the rights exclusively. Universities never license exclusively to individuals (“entrepreneurs”)–only to companies.
Now for some nonsense. There’s nothing particularly distinctive about university discoveries that make them “early-stage” or “high-risk.” Many inventions are research tools that have immediate uses–cell lines, instruments, algorithms, processes. Others are features or improvements that build on existing technology and can be readily included in commercial products and are not “early stage”–it’s just a matter of economics, good design, and customer demand. What’s interesting is that many of the things that one might call “high risk” are actually “goofball.” They aren’t “high risk, high return” but rather “high risk, low or no or negative return.” These, if they are licensed as a result of university “marketing” are marketed to suckers or marketed deceptively. How many university patented inventions are indeed “early stage” and “high risk” and “meaningfully beneficial to the public” let alone “moderate return” or better?
Consider the claim–based on a faulty premise, logically anything may follow–that university inventions, because they “tend” to a characterization, “require the patent system” that “protects” them. This is nuts. Let’s start with the digital computer, developed in the basement of Fine Hall at Princeton. Deliberately not patented. How about the protocols that underlie the internet? Not patented–instead, became open standards. And then there’s SPICE, the gold standard of circuit design, developed at Berkeley–and released open source. No patents necessary. Patents were not necessary for Cohen-Boyer and many other research tools, disease assay kits, and the like. Anyone who could read the publications could practice the inventions. Patents necessarily and absolutely block early access. One does not need to collect evidence for this fact. What one has to do is find evidence that universities routinely make patented inventions available non-exclusively, if not royalty-free. But they don’t. I know. I worked in university licensing for over 15 years. The default is the exclusive license, broken only by private research sponsors, federal consortia, and patent trolling. Ugly.
The term “protect” requires a brief explanation. What could “protect” possibly mean here? In a commercial setting, a patent might “protect” one’s investment against copy-cat competition. But in a university setting, where the putative goal is to get technology out and in use, “protect” runs against the whole idea of “publish” and “teach” and “disseminate.” The university, so the infographic implies, “protects” an invention from immediate use so that someone can develop a monopoly position in industry and the university and that someone can share in the upside of that monopoly. The patent “protects” not the innovation but the financial value in the threat to exclude others. It’s an odd thing, to derive one’s value from a threat to exclude those that would adopt a technology, so that someone can exclusively make and sell a product instead. One might argue that most discoveries made at universities need such “protection.” It’s an administrative choice of a business proposition, not a necessity. “Corruption” comes to mind–even if it is “noble cause corruption.”
Now for the pack of lies. Before Bayh-Dole, there was the NIH and NSF IPA program. Universities could acquire title to inventions, patent them, and license them. The federal government could do the same. The federal government rate was 23% for biomedical inventions. The university rate for mostly biomedical inventions under the IPA program was under 5%. The debate did not have to do with patent protection. It had to do with whether a patent on federally funded inventions directed at public welfare should carry the full complement of rights, including the right to prevent all uses, the right to charge monopoly prices, the right to hold the monopoly for the entire term of the patent. That debate created the public covenant that ran with such inventions from the Kennedy patent policy (1963) until Bayh-Dole (1981).
The point was the monopoly right in a patent on a federally funded invention should be broken up as soon as it is clear that an invention is in a condition to be used with public benefit. If private “risk capital” comes forward to develop an invention faster and better for public use, then those doing the investing can enjoy a limited term for their monopoly to recover their investment, and then the patent monopoly is exhausted and access provided for everyone, so there may be “free competition and enterprise” (to use the language of Bayh-Dole). The need for a monopoly was infrequent; the term of the monopoly was short–three years from the date of patent issue (Kennedy); no more than eight years under an exclusive license (IPA, circumventing Kennedy)–even Bayh-Dole as passed limited nonprofit organization exclusive licenses to non-small companies to eight years.
So the claim here is simply not true. Universities had ample opportunity to acquire patents on federally funded inventions for a decade before Bayh-Dole, did so, and did 5x worse than the federal government for biomedical patents. What the infographic does not bother to point out is that while inventions often may have “languished” (ideas are cheap, even novel ones) for lack of commercial interest before Bayh-Dole, inventions now are “imprisoned” as it were by a patent claim without any easy mechanism by which the inventions made be used, short of the dreaded 35-page exclusive patent license agreement.
Here’s the rest of this silly paragraph. The “patent protections” are held by the universities, not by the businesses (and never by entrepreneurs). When Harbridge House did its study of federally funded inventions in the mid-1960s, it found that a licensed technology had 1/2 the likelihood of commercial use over one that was owned by the developing company. And the rate of success dropped to 1/4 if the organization holding the patent had no prior experience with the technology. Combine universities (no experience) and licensing (rather than ownership) and you have created the worst possible combination of outcomes for managing inventions. Yet the infographic makes this condition out to be a virtue.
It may well be true that an exclusive license will give a monopoly-minded investor a chance to develop a commercial product. But the infographic does not point out that in the current environment, that monopoly extends for two decades and maybe longer if follow-on improvements and applications are funded. There is no room for competitors. Look at the miserable result of MIT licensing powder-based 3d printing to Z-Corp (since acquired). Z-Corp went off and filed on a number of iffy “improvements” and extended its monopoly for decades, pricing its powders at $2 or more per cubic inch, when anyone could print with common materials–sugar, maltodextrin, whatever, for pennies per cubic inch. Nice for the royalties to MIT–a total crapfest for public benefit, competition, and development. As a result, powder-based printing has “languished” and other modes have come to the foreground.
Finally, the infographic claims this “process” (which will be displayed graphically on the next page) is “central” to US economic growth and innovation leadership. This is beyond silly. First, patents with government funding represent 1.1% of all patents issued presently. The university share of that is about 50%, so federally funded patents held by universities account for about 0.5% of patents. Of these, about 20% are ever licensed. So we are at 0.1%. Even fewer become commercial products. This sub 0.1% of all patents, the infographic claims, is “central” to the American economy. What do you think? I think there’s absolutely no evidence to support the claim, either for the U.S. economy or for U.S. as the “global innovation leader.”
As for “global innovation” leadership–that’s weird nationalist populism. Why should it matter if the U.S. is an “innovation leader”? What does that mean? Why should we not share our research technology with other countries, and expect that they will do so as well? Here’s the strangeness–most federally funded inventions, other than a handful in biomedical licensed/assigned to pharmaceutical companies, never work any foreign rights. Many are U.S. patents only, so the research that’s published, supported by federal funds, is made available internationally with no rights, or no enforceable rights, where anyone can use, develop, and sell products based on the U.S. invention without bothering with a license, a monopoly, or paying an American university. The effect of U.S. universities taking out U.S. patent positions is largely to gum up the U.S. market, leaving the rest of the world to fend for itself–freely. Of course, when it comes to prescription drugs, U.S. universities use exclusive licenses to finance world-wide patent protection, ensuring that any country that has the ability to pay must pay a monopoly price to gain access to the latest therapeutics. Nice, but is that what is meant by “global innovation leader”? that the U.S. through the agency of university patent positions, helps to force other countries to pay high prices for public health interventions? Seems a bit, um, nasty.
I participated in the meeting at Stanford that led to the Nine Points document. Let’s say that the purpose of the meeting was to get at bad licensing behaviors at WARF, among other organizations–and those behaviors were directed at exposing universities to patent infringement suits by WARF licensees. The Nine Points document, though passed around by AUTM as if it was something AUTM member universities should subscribe to, is routinely ignored in university licensing practice, which is largely focused on exclusive licenses, not non-exclusive, and does not routinely carve out rights for socially responsible licensing, as Carol Mimura has championed.
As for federally funded inventions acquired by universities, the controlling document is the standard patent rights clause authorized by Bayh-Dole (37 CFR 401.14(a)). And university patent brokers have worked to take out of that clause anything that would restrict the full use of a patent–including enforcing non-use, assigning monopoly rights to companies, charging monopoly prices, and allowing the monopoly to run the full term of the patent. These are not the things addressed in the Nine Points document. Here, the infographic (and APLU and AAU) simply spout something that sounds good to cover for actual practice, which runs in an entirely different direction, pursuing institutional self-interest, and in particular, money. If UC and MIT were following the Nine Points document’s “points,” would there be litigation between them over CRISPR?
Notice that now we have shifted from federally funded inventions to patents, regardless of whether they are the result of federal funding. In 2015, there were 298,407 issued U.S. utility patents. Of these only about 3,500 were issued to universities and related foundations and recited federal funding. We are talking about less than 0.3% of all patents issued in 2015. Numbers are impressive, as long as you don’t put them in context.
The infographic then discusses startups–950. But in context, consider that in 2015, according to the US Bureau of Labor Statistics, there were 679,072 startups. Let’s see, that puts the university output at 0.1% of the total. Please pause with me and contemplate the significance of the university contribution.
There’s more, of course. The university startups aren’t based on patents issued in 2015. They might not be based on patented technology at all. Heck, the University of Washington counted as startup companies board games, software running as a service, and even companies started at other universities with which the university did some sort of business (and which the other universities were reporting as their startups). So the infographic merrily slips its base from patents on inventions made in federally supported research (the basis for the Bayh-Dole claims that open the text) to the importance of patents (to “protect” the university’s financial interest in taxing industry) and now to startups, which are attributed to “university-based research” but now without the need, specifically, for federal funding or patented inventions. Graduate students can start companies, faculty can leave and start companies, entrepreneurs can start companies and sponsor research at a university (as with the SBIR/STTR programs), and heck, universities can create shell companies (like Utah and Washington did for years) with no operations and count those as “startups.” None of this is particularly impressive, as far as numbers go.
Any individual company, as a legitimate startup, is worthy of respect, regardless of the numbers. Google, a Stanford startup, doesn’t need to have any other companies recited to be impressive. One is enough. But consider: even those startups that deal with a university’s patented invention–how much stronger would they be if they weren’t burdened by the fact of 35-page license agreement with all sorts of payment, audit, diligence, and indemnification requirements? What would it be like if those startups owned the inventions outright? Or had a free license to them and could build their own patent portfolio around improvements–and focus on paying for those rather than for the university’s patenting work?
I don’t know what to make of the “700+ commercial products.” Are these new commercial products on the market in 2015 that “university research led to”? Or is this just a count of new licenses which meet AUTM’s definition of a “commercialization license”–that is, any licensing earning more than $1,000 in a year (and without regard for whether the license is based on a patent right, or based on an exclusive right–after all, the infographic just got done making the claim that without exclusive rights, companies won’t invest in research technologies, which is nonsense, of course). There’s a big difference between commercial products based on exclusive university licenses granted under patents on inventions made in federally funded research and any license transaction worth more than $1,000. There’s also a big difference between research that contributes to a product (through testing, say, such as a clinical trial) and research that results in patented inventions that are then licensed to create commercial products. The infographic provides no information about what the “700+ commercial products” means–other than that the word “alone” suggests that everyone is supposed to be impressed.
Anyway, in context, Forbes suggests that in 2010 there were “about 250,000” new products worldwide, with a failure rate on the order of 90%. Let’s see, that university research contributed “700+.” That’s about 0.3% of all new products. If university research “led” products have about the same failure rate, then we are looking for about 60 new products. Let’s see. From 2011 to 2014, the federal government provided about $27 to $28 billion per year to universities for research. And we might see 60 or 65 products that don’t fail immediately. And this, so says APLU and AAU, is what drives the American economy? Who are they kidding? There may be some helpful products in there–I’m fine with that. But for $27 billion? There must be a way better way to produce new products from federally sponsored research than providing the money to universities to produce patentable inventions to be licensed for profit to investors. That is, is the role of federal research funding to faculty at universities actually about producing inventions to create private monopolies? Maybe, despite what APLU and AAU suggest, that’s not the point of federally supported research. Maybe federally supported research does “drive innovation”–but not via institutionally held patents that monopolize research-derived inventions. Ever wonder about that?
The infographic finishes with a fine account of the “linear model” of innovation which no one appears to think is credible in science policy discussions but exists on in the backward podunk reaches of university administrative thought. Here’s the five steps, in one compound exhibit:
Well, actually universities don’t compete for funding. They release faculty to apply for funding, and the faculty in some way might be said to compete for funding. University administrators don’t compete any more than fantasy baseball participants compete. They merely bet against each other and find personal status in rankings against other bettors. Again, APLU and AAU seem to think that universities are abstractions that act and ignore the individuals who do the work. That’s thinking that effaces the individual nature of inquiry and makes it appear that the entire effort is an administrative system.
There may be “groundbreaking discoveries”–especially in soil science and geology. But most of the things that pass for “inventions” are variations on themes, clever research tools, and ideas that came along during a morning jog rather than “in the lab.” The idea that university personnel (ignoring the research staff–just faculty and students) work in laboratories where they make their discoveries is like bad Hollywood movies. Young Frankenstein comes to mind, though that’s a good bad Hollywood movie. Nutty Professor. Yeah, like that. Or Flubber.
Here’s where the fun starts. First, university technology transfers offices confiscate all inventions from the inventors. They don’t merely “patent” these discoveries; first, they take them. As for “copyright”–copyright happens, technology transfer offices don’t “copyright” anything, and certainly not discoveries. Copyright vests in works of original authorship fixed in tangible media of expression. One does not discover an original work–one authors it. But this is just like the ignorance that exists among the folks at APLU and AAU with regard to copyright. They think copyright is patent spelled in a strange language, like Welsh.
The worst of it is that the patenting work is largely unthinking. It does not occur that perhaps the greatest benefit (even financial) might come from not asserting an institutional ownership position and allowing inventors or sponsors to work out ownership–or not asserting any ownership position at all and shifting the focus to developing technology platforms or commons or standards. You know, like the internet. Oh, wait, the internet can’t be a “driver of innovation” because it wasn’t institutionally patented and licensed to a favorite company to become a commercial product, one of 700 or whatever.
Basically, whatever university bureaucrats patent is dead to use. Only 1 in 200 inventions makes it through the prison colony called the licensing office to become a commercial product. Technology transfer offices is where discoveries go to die.
Here, there’s plenty of confusion. Startups are businesses, too. And university technology transfer offices don’t license to entrepreneurs. They license to the startups that entrepreneurs start. Actually, mostly universities license rights back to startups started by faculty, staff, and student inventors as a way to get access to their inventions for use outside the university.
A patent creates a right to exclude use. The licensing offices don’t transfer any rights to “use” ideas. It transfers the right to exclude others from using claimed inventions–not just what was discovered, but everything that a clever patent attorney can think of that would be related, applied, or functionally equivalent to what was invented. Not one compound, but a hundred compounds, of which one might get worked on; the rest are off limits to everyone else.
And what do you make of the use of “help” here: technology transfer office “help transfer the rights.” This is nonsense. The technology transfer offices manage ownership of patents. They don’t “help” to transfer rights–they grant (or mostly don’t grant) rights via 35-page exclusive patent license agreements. The image created is licensing offices as helper folk, when in fact they are the executive authority, or, another way, the bureaucratic bottlenecks. The dam builders in Madagascar 2.
In any event, only about 20% of what’s claimed at the best university practices gets licensed. The rest rots–by design. Try it some time. Ask for a non-exclusive license to make and use an invention any patent that hasn’t been licensed in five years after issue. You’ll get a 35-page patent license with “exclusive” changed to “non-exclusive” with the same royalty structure and diligence.
It’s a nice idea. Every so often, this does happen, that a product is created, and even more rarely, the product creates new jobs or improves the quality of life. But often when there’s a product, it goes nowhere. And of the products that are created become little monopolies–the price of drugs is sky-high in America. Have you noticed? Xtandi–based on a compound patented by UCLA and created with federal funding–sells for $80 or so a pill in the US, but could be made and sold at a profit for less than $5 a pill. But because UCLA granted Medivation an exclusive license without any constraints, we pay a monopoly price for two decades. Nice!
In all, this five-step process ignores all the other pathways by which research informs commerce. Back in the early 1990s, one of the technology transfer officers at the University of Washington looked at all of our licenses and studied how these were achieved. He found that under 5% followed anything close to this five-step model.
Put another way, most deals didn’t follow this model at all. The company (or companies) that want the technology might already be known before an invention is made–they may even have supplied the ideas or protocol for the research. Or the idea that motivates the invention didn’t come from a “lab”–perhaps it came from a “shop” or a “library” or a “studio.” Or, the inventor or an investigator or a next-door neighbor made an introduction long before an invention was reported to the university. In one situation I worked with, an investigator had already distributed his inventive software to five or six companies. “I told them if I got around to licensing it to them, we would deal with that later.”
Anyone trying to force things down this 5-step process is nuts. Anyone who ascribes every result, or most results, or even more than a handful of results to this 5-step process is simply worshiping the political rhetoric that got Bayh-Dole passed, and that rhetoric is based on using Bayh-Dole to circumvent public policy that argued that publicly funded work ought to be available to all and not packaged up as monopoly rights that hold back each bit of inventiveness from all uses so that a favorite company can exploit the patent (or, as it turns out, often not exploit the patent but be sure that no one else can).
One might go so far, even if one wished to preserve this linear model, of eliminating step 3 entirely. If university technology transfer offices stop patenting things, then we can eliminate step 4, too. Now we have a three step process by which discoveries get to companies–federal funding, discoveries, and companies making products. That sounds even better. Back when Bayh-Dole was being debated, advocates were asked why not just vest government funded inventions in companies rather than have nonprofit patent brokers? There were no good answers other than that “Congress isn’t ready for that.” Perhaps. But other than to make a pink-ribboned parting gift for Sen. Bayh, why was Congress ready for Bayh-Dole?
As it is, this 5-step model is actually a prophetic emblem. It is a hope, a wish-lust of university administrators, that patents will become products. That process takes place just often enough that one can say, “Why, yes, sometimes that does happen.” What’s left out is Step 6, which is why the effort is made in the first place: to generate money to fund the technology transfer office (so it will get out of debt) and provide money for administrative slush funds.
As a matter of prophecy, this version of the linear model, sometimes with the dramatic addition of the heroic technology transfer professional taking up the role of the Lord my Shepherd to guide research inventions through the “valley of death” of the 23rd psalm, creates an environment where no evidence can disconfirm it. Like a mad sect waiting for UFOs from the Planet Clarion to come to rescue it, university bureaucrats like the ones behind this APLU/AAU infographic will double down and repeat the same myths regardless of the actual circumstances. I’m reminded of the black knight’s problem. Perhaps something else comes to mind for you. Maybe some passage from Atlas Shrugged that seemed goofy at the time, but now sounds strangely prescient.
An honest infographic might suggest that university patent licensing is an inconsequential drop in the bucket when it comes to the economic activity of America, but that drop now comes with the loss of academic freedom, loss of open interactions between universities and industry, and a lot of monopoly pricing for a few “lifesaving medicines” that benefits from university-created patent positions for twenty years before those medicines become “affordable” for the most of us.
There’s certainly value in university research–it’s just that so much of it isn’t useful, other than for career purposes. And there’s often value in helping people who do have good ideas to get those ideas into broader circulation–call that technology transfer if you want, but it’s also teaching and publication and making data available and creating forums in which people can exchange information and work. And there is, at times, value in obtaining patents on inventive work. It’s just a rather open question what value there is when university administrators obtain those patents, and how they then permit those patent rights to be exploited. The outcomes that are declared–“transforming” society, “lifesaving” medicines, “countless” everyday products–who could argue against such valuable things? But these outcomes are emblems for prophecy more than the reality.
The reality is that universities use exclusive licensing that maintains patent monopolies as a way to make as much money as possible in a few transactions over a span of two or three decades–and it’s not even evident that their approach is all that successful. To run a financially successful university patent licensing operation, all that’s needed is one big license agreement every twenty years or so. Spread across a hundred universities, there are enough “successes” to fill an infographic. But those “successes” do not mean that the methods that are associated with them–compulsory institutional ownership, exclusive patent licensing, willful breach of Bayh-Dole’s standard patent rights clause–are the cause even of those successes. What the present methods do provide is that whatever does happen to “emerge” has a bureaucrat’s thumb in it, so that the transfer claimed results in payments to the university under a patent license.
As it is, APLU and AAU have produced a fake infographic, but one that flatters their political positions. Understandably deceptive, and in that way useful to their purpose. But it’s a false narrative, a failed prophecy.
[updated 5/15/17 to include an account of touch screens]