We are working through the AAU fantasy about invention commercialization. In AAU usage, to “develop” an invention may mean to “suppress” the invention but also may include to design around an invention–as in, invent what is actually needed to create a meaningful product. The university’s patent gets used to suppress interest by others in any part of the licensed invention. But the university’s invention may not be used at all, or if used, only in a little bit. The patent serves to suppress all the unused parts of the invention. When a university assigns the invention to the company, it gives the company to litigate to suppress the use of all the unused bits of the patent, too. Or, if companies use some of those unused bits anyway, to litigate to extract payment from those companies–to troll them. If the company doesn’t, so the typical university license goes, then the university can do so. If the company builds its new product relying on “B” inventions, then it can terminate the licensing agreement with the university (especially if the university is paying attention and demands development), having spent only the upfront fees to buy time to develop its own version of the technology. No royalties paid. No technology from the university actually “transferred” despite the license.
AAU folks would like you to believe with them that such things never happen. My experience in licensing argues that this sort of thing happens all the time. Companies develop outside the scope of the licensed patent, companies use the fact of the license to raise money and then develop whatever they want, and often the companies sell out before any product reaches the market. Yes, there was a license, and even money paid or an equity position converted later to cash–but there was no technology transfer, no product based on the licensed invention. The patent gets used, the patent may even earn some money, but the transaction activity has nothing to do with the practical application of the licensed invention.
Even when a company stays within the licensed patents, as may happen with the special special case of pharmaceutical inventions, the patent typically claims hundreds to thousands of compounds as a class, and whatever gets developed into a commercial product represents a tiny, tiny “A” bit, with the rest of the compounds suppressed.
So now what do you make of AAU’s use of “develop and commercialize”? What are the “necessary rights”? Rather than think of “necessary” to mean “only those limited exclusive rights necessary to justify the use of substantial private risk capital,” we might think instead “the exclusive rights necessary to suppress most if not all of a university’s invention in favor of something else.” We might then ask, “Is this sort of development what Bayh-Dole enabled? Was this development what Congress intended?”
I know this is the long way round to responding to the AAU’s thought that companies needed exclusive licenses from university bureaucrats in order to develop and commercialize research inventions. There are shorter ways. For instance, that the AAU’s fantasy system has never been shown to operate as claimed. It’s fantasy until demonstrated, and it has never been demonstrated as a general case. No university publishes its data to show what inventions claimed by bureaucrats have actually ended up “developed and commercialized” into “practical application” and what inventions have not. Here’s the AAU attempt at demonstration. First, a broad claim:
The system of public-private technology transfer that was established under the Bayh-Dole Act has been extraordinarily successful in moving university discoveries from research laboratories to the marketplace. Technology transfer has provided a rich return on public funding for basic research in the form of countless innovative products that benefit consumers, create jobs, and contribute to U.S. economic competitiveness and technological leadership internationally.
The system we have was not “established under the Bayh-Dole Act.” It may have been established to take advantage of ignorance and indifference to Bayh-Dole. But it has not been “extraordinarily successful” or provided a “rich return” with “countless innovative products.” Look at the data AAU produces to support its claims:
The most recent survey by the Association of University Technology Managers (AUTM) shows that U.S. universities executed 5790 licenses agreements and options with companies and were issued 5163 U.S. patents in 2013. This represents 59 percent more license agreements and options executed and 55 percent more patents issued to universities than just 10 years earlier.
License agreements do not mean there are products in “the marketplace.” Notice that “options” are thrown in to the count, too, as if licenses do not rise to an adequate amount of extraordinariness. Keep in mind, too, that AUTM does not report subject inventions separately from other inventions (about half of patents issued to universities recite federal funding) and further, AUTM counts software licenses, material transfer agreements, most anything that grosses at least $1,000 in the transaction as a “license” agreement. There may be no invention, no patent, no development or commercialization for a license to get pushed into the AUTM count. One bit of software might be licensed 100 times, while most patents sit. Patents and licenses, too, are not related in the same year. The patents that issued in a year are generally not the ones licensed in that same year.
But look where the claim then goes–not to actual use or products, but rather to the growth of the activities over a decade. Every new patent means, absent further information, more technology excluded from general use. Every exclusive license for a patent, absent further information, means a technology excluded from all use but for the use the exclusive licensee makes of the patent, the license, or the invention, or bits of the invention. Nothing here demonstrates extraordinary success. The figures point to the opposite. But there’s more:
During the same period research performed at universities led to the formation of nearly 750 new start-up companies, more than doubling the number of university based start-ups created compared to 2003.
Same thing. No information on use of inventions in any sort. There’s not even a clear connection between research and startups. The University of Washington was faking its start up metrics in 2013. How many other universities were counting any new or recent company as a startup, regardless of whether that company started around a licensed patent arising in research? We don’t know. Further, AUTM takes reports from universities without auditing these reports for accuracy or redundancy. A start company may take licenses from two or more universities. Each reports the startup as “its” startup. Thus, one cannot simply add the number of startups reported by each university and arrive at a total number of startups. AUTM’s numbers are necessarily and conveniently overstated. But even at 750 startups as an upper limit, consider that 629,000 new businesses formed in the U.S. in 2013. Even as an unreliable, bloated figure, AAU claims as “extraordinary” the formation of 0.1% of the company startups in the country. I don’t know how this figure matters, economically. And remember, we have nothing here that says anything about licenses to the startups, use of inventions licensed by the startups, or commercial products introduced by these startups. Nada.
While these start-up companies provide economic benefits to the nation, they are especially important to the regions and states in which research universities are located; more than three-quarters of these new start-up companies had their primary place of business in the licensing institution’s home state.
Meaningless information. Nothing showing that any of the startups have provided any “economic benefits.” And whether a startup has its place of business in a state is meaningless as well. Startups move. Startups locate their manufacturing wherever. Startups sell where there’s market. AAU might have noted that most of the University of Utah’s startups over five years had their primary place of business as a university address. And that’s stretching “business.”
AUTM also reports that in 2013 there was over $22 billion in sales of products based on academic research with more than 700 new commercial products created.
And now we drift. AUTM doesn’t track product sales. Generally, the method used to estimate product sales is to multiply licensing income by an arbitrary royalty rate to estimate a royalty base and then assume that base is sales (rather than milestone payments, patent cost recovery, sublicensing fees, whatever). Even then, one cannot add the “estimated sales” from one university to that of another, because both universities may have licensed to the same company, and so both receive income that’s part of the same royalty base. And we have drifted from royalties paid on the sales of products made under a university patent license that called forth risk capital to “products based on academic research.” The figure claimed lies largely outside the system that has been so extraordinarily successful.
But even if we take the $22 billion as a fantastical upper bound, then what do we make of the report that in 2013, retail sales in the U.S. were $4.5 trillion. Yeah, sales of all products ever licensed by universities over the past 35 years amounts to 0.5% of the total retail sales in 2013. The economic impact of the claimed, overstated university licensing effort is approximately 0. As for the 700 new commercial products created. Despite the prospect of overstating the number by not checking for redundant counting of products and including products that have no patented significance, we are looking at 700 new products reported this year from the development of nearly 100,000 patented inventions issued to universities over the past 20 years. That’s a production rate of 0.7%, as a coarse upper bound to the idea of “extraordinary” success. Compare with the 25% claimed by universities prior to Bayh-Dole. Or the 23% by the Public Health Service before Bayh-Dole. Or even the 5% for universities licensing under the IPA program.
Worse, we are not even considering the economic impact of university bureaucrats withholding 99.5% of research inventions, that never get licensed (80% or so) or get licensed exclusively and don’t become products (19.5%). Nothing here that indicates that the university patent licensing system has been extraordinary in any positive, impressive way.
A recent Biotechnology Industry Organization (BIO) study estimates that between 1996 and 2010, patents commercialized from university and nonprofit organizations (supported mainly by federal R&D dollars) contributed as much as: $836 billion to the US gross domestic output; $388 billion to the US gross domestic product; and supported a cumulative total of 3 million person years of employment.
Here we have estimates based on an economic model that makes a bunch of unsupported assumptions and produces apparently big numbers. The model expands beyond to universities to include other “nonprofit organizations”–hospitals, especially. And we get the “as much as”–so, only the upper limit of an estimate without the lower limit. Look at the GDP figure of $388 billion. That’s over 15 years. The U.S. GDP over that time was $706.6 trillion. Do the math: 0.055%. That’s 55/1000s of a percent. That’s the contribution claimed by the upper bound of an estimate based on an iffy economic model. Maybe making a vanishing small contribution to the GDP is a worthy thing. And maybe research management has nothing much at all to do with the GDP. But apparently BIO, AUTM, and the AAU think people making policy decisions about Bayh-Dole and university patent practices are clueless. Or maybe their target is the general public–clearly with the same idea that the public is clueless and will be easily deceived by estimated numbers treated as fact and presented out of context.
To ensure that universities are good stewards of discoveries and intellectual property developed from research supported by federal funding, many universities have developed and implemented policies and procedures drawn from key recommendations made in 2010 by a National Research Council (NRC) committee, chaired by Washington University in St. Louis Chancellor Mark Wrighton, and principles articulated in a 2007 white paper developed by a small subset of research universities along with the Association of American Medical Colleges entitled, NinePoints to Consider in Licensing University Technology.
More nonsense. I don’t know of any universities that “developed and implemented policies” based on the NRC recommendations. Those NRC recommendations reprised the practices endorsed by the university participants on the committee. The NRC proposed keeping things as they were. As for the Nine Points document, I was there for that. No university to my knowledge has made any changes in its patent policies or licensing practices as a result of the Nine Points document. The few universities that were doing non-exclusive licensing as a viable alternative to a monopoly trade in patents didn’t need the Nine Points document, and the rest have ignored the Nine Points, other than to cite it with approval.
Instead, university administrators have doubled down on demands to own all inventions, broadened the definition of invention to include non-patentable inventions and stuff that’s not even inventive; they have become even more unselective in their patenting; and they still don’t report publicly the performance of their “system” of patent management. Again, the AAU presents an aspiration as fact. It’s just that if pressed, it is unlikely that they could produce any evidence to support their claim. What is true is the assumption implicit in the “good steward” claim–that universities will demand ownership of inventions, and that being “good” means trading in patent rights for a share of monopoly profits enabled by their trade.
And thus we arrive at the AAU task force’s recommendations. These have to do with wordsmithing mission statements for management of intellectual property. That is–no change in the underlying claim to own, just a better statement of the “vision” for such ownership. Affirm commitments, make procedures public, avoid “so-called patent trolls” (but it’s okay to be one or allow your failed startup to become one), and find a way to evaluate technology transfer that isn’t based on “measuring revenue generation.” And so we return to the problem raised by the unnamed critics–that universities focused too much on maximizing licensing revenue. This last bit is too funny. Here’s the 2015 AUTM annual licensing survey report comment on the matter:
AUTM disclaims the idea of “maximizing income” as “the core driver” but then turns around and uses income as proxy for the success of the prevailing university technology transfer “system.” Left unstated is the implied equation “licensing income is success.” That is, whatever the wordsmithed intentions for university IP management, the standard of evaluation, as far as AUTM is concerned, is still licensing income.
Thus AUTM fails AAU. That much is obvious. We can take it further. The idea that prevails in all of these materials is that the measure of the impact of university research is IP licensing income. This is a rotten idea. There’s nothing in it to account for the open dissemination of research findings, so that the impact would come by other measures than IP licensing income. There’s nothing to account for royalty-free patent licensing, standards licensing, cross-licensing, or any other method of exploiting inventions that does not involve payment to relieve one of the threat of patent litigation. There’s nothing, even, to account for income from other sources as a result of open IP management practices–from donations, from sponsored research, from consulting, from workshops, from tuition.
In the fantasy world articulated by the AAU task force, the aspiration to be a “good steward” of intellectual property means that universities must adopt the special special case as the general case. That nothing can be considered that would interfere with a university’s assertion of ownership of inventions, of a university’s determination to secure patents on these inventions, and then to trade on these patents for a share of money from the exploitation of the monopoly rights reserved for the “commercialization partner.” If critics complain that universities focus too much on revenue, then the thing to do is to change policy and vision statements to make it appear that revenue doesn’t matter so much–but keep the same practice. Make universities a bastion of opposition to ideas of open innovation, of contributing to the public domain, of using patents to ensure quality and safety of developed inventions.
It is strange to see the AAU exploited in such a manner. Perhaps no one at the AAU cares. Or perhaps they are clueless to what’s going on. The “technology transfer system” attributed to Bayh-Dole is a dismal failure over pre-Bayh-Dole practice. Universities are advised to make more statements to sound like “good stewards” even while they actively change their practices to work the other way, broadening claims to institutional ownership and insisting on a fixation with monopoly trade in patents as the only way by which to “advance the public interest” with research discoveries.
When will their scam end?