The AUTM CEOs Speech, Fitt 2

We are working through a speech that the CEO of the Association of University Technology Managers gave at a recent symposium sponsored by NIST with the dubious title “Unleashing American Innovation.” Our CEO has made pompous claims about his organization and recited a fake history of federal ownership of inventions made in extramural research to back up those claims, along with a presumption that inventions made in research hosted by universities of necessity must be university inventions. We reach the CEO’s pivotal question about pompous claims, fake history about the government, and presumptions about university ownership expectations:

So the question is, “Why is that important?”

The “that” here is so full of ambiguity. We might ask, Why is this fake history and chronic administrative delusion important? The fakery and lies are important because without them, people would see rather clearly the nature of the bargain that Bayh-Dole proposes, which is nothing like what the AUTM CEO depicts.

We read on:

Well, AUTM has been collecting technology transfer statistics since 1991.

And before 1991, AUTM also collected technology transfer statistics, just in a different form. But “technology transfer statistics” is not the same as statistics specific to performance under the Bayh-Dole Act. Technology transfer is not even specific to patentable inventions–it may include software, biomaterials, and know-how. Technology transfer may involve research-based inventions and inventions that have not been made in research settings. Technology transfer can involve inventions in research funded by industry sponsors where the sponsors hold a license in any results upfront, without any need for there to be patents or marketing or commercialization. And for AUTM, keep in mind that a “commercialization license” is any deal resulting in $1,000 or more, regardless of whether anything “licensed” is ever used, developed, or sold as a product. So take a breath.

And of the almost sixty-seven billion dollars of research for public sector research institutes.

That’s not the federal commitment total–it’s apparently the total expenditures–but for what time period? Since 1991? Surely not–the figure is way low. We get then a context-free number that sounds bigly. And that money coming to universities was not for “public sector research institutes” but rather was for faculty investigators proposing  meritorious projects hosted at universities. Thus, another administrative lie about the money.

In 2016 alone approximately sixty percent of that research funding was provided by the federal government. That’s you, me, other taxpayers going forward.

Yup. But the money isn’t entirely from taxes. The government gets its money other ways as well. Just sayin. The 60% figure is a big average. At some universities, the federal funding share is more, and at others, much less. One might consider whether there’s a threshold of federal funding after which a university administration changes its behavior to optimize for federal funding. As one director of sponsored projects once told me, “I wish the industry funding would just go away.” Yup–screw companies trying to collaborate with university faculty, such an administrative inconvenience. Consider the effect of a federal policy that federal funding could never amount to more than 30% of a university’s total research funding. That might be the most invigorating thing for university faculty–their administrations would have to configure for engaging foundations and companies and industry consortia rather than latching on to a federal teat and playing the baby cow forever. But it is just this sort of observation that AUTM aims to crush.

Not surprisingly, it turns out there’s a linear relationship, almost a linear relationship between research funding and the number of inventions.

Silliness. University licensing offices set goals for disclosures. For a long time, the word on the street was that one should aim for one invention disclosure for every $2m in extramural funds. There’s your “linear” relationship. “Technology transfer” as a term applied to university patent management was popularized by Research Corporation in the 1970s. It referred to a university hiring people to “walk the halls” to rustle up inventions that would then be reported to Research Corporation and possibly also “transfered” to Research Corporation. Yes, “technology transfer” goosed inventions out of the woodwork and transfered them out of the university, but not to industry–to Research Corporation. Again, the number of inventions is a function of the administrative effort to surface invention reports, without regard to the quality of the inventions nor even if those inventions are patentable. A number of bureaucratically inclined universities have changed their patent policies to claim inventions whether patentable or not–or even whether inventions or not. Thus, a count of disclosed inventions, at least as tracked by AUTM, is meaningless.

I could go get fifty invention disclosures any time I wanted at the public universities I worked at. But I didn’t *want* fifty random sucky invention disclosures just to rack up the bean count of inventions. If one wanted a measure of quality, one might expect that as research funding increased, the inventions managed by the university would go down in proportion. If there’s a linear relationship between the number of inventions and the amount of research funding, then a university has been indiscriminate in its management of invention reporting. The linear relationship is an administrative artifact, not a natural fact.

So my first take home message is this. As Marcia mentioned earlier, the more investment in research leads to more ground-breaking research discoveries –inventions.

Weaselosity. The federal government does not “invest” in research. It provides grants in aid to faculty at universities. Functionally, a federal grant is a a subvention, not an investment. Let us not speak falsely now/The hour is getting late. And it’s a huge jump from an invention report to a “ground-breaking research discovery.” An engineering lab can report thirty inventions in a year and they are all variations on the same design concept, and the next year that same engineering lab can have designed something new that obsolesces last year’s inventions. (I once inherited a situation in which a university  had licensed fifty inventions in information technology to a faculty startup, with the remedy for nonpayment the revocation of the license. The startup promptly developed another fifty inventions that obsolesced the first fifty, refused to pay for the patenting of those first fifty–and they had a point! Those first fifty patents–a hefty university legal investment–weren’t ground-breaking discoveries. Just a serviceable start.

Let’s actually put some hard numbers to that. From the approximately 67 billion dollars of research funding in 2016 alone, again for public sector research institutes, over 25 thousand inventions were discovered,

“inventions were discovered” is simply boggling confustion. Inventions are made, not “discovered.” Here, what the CEO surely must mean is inventions were “reported” or “disclosed.” But worse. The research funding in 2016 could not possibly have resulted in 25,000 inventions that same year. It’s nonsense. It’s boggling nonsense. The 25,000 invention reports result from a random sample of research (and non-research) from past years of activity–way more than $67b is in play. And look at the big picture. If there are 25,000 invention reports from university inventors across the country, then those inventions are a dime a dozen. Lots of tiny bits of least patentable randomness along with a few significant inventions.

Put it this way. Once every year, at a research university, there might be two or three invention reports that if acted on opportunistically in some often unique way might result in something important.  Once every five years, something might be packed down in a meaningful license, and once every decade or two, one invention might come into significant use so that we might say it was “ground-breaking.” The worst thing one could do with those two or three invention reports a year would be to bury them under another hundred or two hundred invention reports and then insist that one administrative process should dictate how inventions are managed. Check out most any university technology licensing office for the depiction of such a process. The message to bureaucrats is “we are organized and follow a logical process.” The message to any university inventor is “your invention is just one more crumb to be swallowed by our bureaucratic borg.” You’d have to be desperately ignorant or ignorantly desperate to buy into such a “process” of technology transfer.

. . . an average of approximately 2.6 million dollars in research funding per invention.

An utterly meaningless “average.” Averages throw out the structure of the data. Averages are useful only where the data are roughly consistent–“mediocrastan,” to use Taleb’s fine term for it. If the data are exponentially distributed, an average makes nonsense of it. If the data is cyclical, the average throws out the cycles. If people expect that to use an average means that the data is roughly consistent, then the use of an average makes an implicit claim about the data–that it is sound logic to use an average. Here, however, using an average is meaningless. Inventions are made without any research funding–indeed, some of the best inventions I’ve worked with had no research funding–and especially no federal funding. In fact, in my fifteen plus years of public research invention management, federal funding counter-indictated success in technology licensing. Industry money was better. No money and a voluntary agreement to work with the university, the best.

But inventions aren’t the goal; benefiting society through products and services is.

It’s difficult to find a way to unpack this assertion. The goal of university patent policies is in fact the report of inventions, the assignment of inventions to the university, the patenting of inventions, the commercialization of inventions. It’s just counter-reality for the AUTM CEO to claim inventions “aren’t the goal.” The invention is the essential starting point for AUTM. Inventions shouldn’t be the goal, but AUTM can’t possibly help itself about that.

But our AUTM CEO makes a different claim–that the goal of something is “benefiting society through products and services.” We may ask, the goal of what? “Investment in research”? Federal funding specifically? University “technology transfer”? And who is it who has this goal? Faculty inventors? Licensing officers? The speculative investors that university licensing officers seem to like to attract? Big pharma CEOs? Federal grants officers? Perhaps the AUTM CEO here is just making a random assertion that sounds good, and that’s enough, even if it is mostly reference-free.

Even if the goal were to benefit “society” through “products and services,” there’s no reason at all that patenting research inventions is the way to provide that benefit, no reason that institutions  have to assert the right to do the patenting, no reason that any resulting patents must be licensed exclusively (i.e., the inventions assigned). Nothing about AUTM’s model for dealing with inventions is required if the goal is to benefit society through products and services.

And we haven’t even got to the ground yet in the AUTM CEO’s assertion. Why must research have to be constrained to benefit society “through products and services”? Why can’t research benefit society in other ways–through prevention and cures, say, in the case of health-related research, or better public policies, or even in better research? What the AUTM CEO won’t say is that “through products and services” means, in practice, “through products and services covered by a patent and licensed exclusively for a royalty while all other uses are suppressed, including research and professional uses and the competitive development of alternative products and services claimed by licensed patents and not chosen for development by the exclusive licensee.” That’s a mouthful, granted, but perhaps reality–and not lying to oneself and one’s audience–takes more words that don’t sound quite as good and would make one think twice before aligning with them. We might conclude that the AUTM CEO doesn’t want anyone to think twice. Please, don’t think twice–it’s all right. 

So, although these public sector research institutions like universities develop thousands of inventions annually, they, as you well know, do not sell or fully develop those products that are on the market.

First inventions were “discovered.” Now, those inventions are “developed.” Perhaps this is the high school writing advice not to use the same word repeatedly–use synonyms and stuff, as if words don’t have to have any definite meaning. The AUTM CEO continues the administrative lie that institutions “develop inventions” when it is clear that individuals, often faculty members, do the inventing, and often they do that inventing despite the presence of administrators. University administrators don’t conceive of research to be done, don’t assign faculty to conduct research, don’t review that research, don’t decide who to collaborate with, don’t decide what to publish. It’s just a university bureaucrat’s self-delusion to imagine that a university is a corporation and faculty are just hired help to achieve the corporation’s business objectives. Again, that’s delusion.

As for sale–universities routinely sell research materials–mice, antibodies, software–and many methods (which may be claimed as inventive) may be practiced directly from published accounts, such as is the case with diagnostic disease assays. Finally, where an invention might be used directly by others (as in research or professional services or software), there’s simply no need for there to be a “market” for the invention to be “on.” The invention is practiced; the invention may become a standard; the invention may be taught or implementations of the invention may be contributed or exchanged. No market. No “commercial” product. That’s not what AUTM advocates, but it is the reality for many, if not most, of the inventions arising in university-hosted research.

Instead, that intellectual property underlying those inventions is licensed to companies that invest substantial resources, as was mentioned on the past panel, to develop those early stage inventions into products and services that serve the public.

The intellectual property does not “underlie” the invention. We are talking about patent rights in inventions owned by universities. That’s what Bayh-Dole is about. A patent “overlays” an invention. A company may use an invention without “investing substantial resources.” The claim that to create a commercial product requires huge sums of money and a monopoly position is only weirdly true. Big company officials may think this way. Venture capitalists in health care may also think this way. But funny, I didn’t know that the $67b in research funding placed at universities was there to ensure that big companies and venture capital firms were subsidized. The argument that products come about only because of patents and huge private investments is like the campfire story about swamp creatures. It has an effect and so gets repeated for its effect. But we don’t believe the story later, when the camping trip is over, or we become neurotic.

Reality: if a new product requires millions of dollars to develop, then generally there will be plenty of patentable inventions made in the process of doing so. An initial invention is not so important, then, if there will be plenty of additional patents to be had. The term of an exclusive license then might be three years, in which time a company had better come up with new inventions that track its substantive development work. And if a new product doesn’t require millions to develop, then all that patent talk is just blather. And if a new product doesn’t require millions to develop because the development is done in a small, capable company rather than in a big, bug-eyed company that can’t think in dollar amounts less than $300 million, then again the millions of dollars argument is simply bullshit. For that matter, a product might be readily developed in China for much less than a comparable product developed in the United States–with its regulatory overhead, taxes, labor costs, bloated budgets, and profit-taking.

AUTM is on a camping trip that it hopes will last forever. In the real world, many products and services have come into use without patents, without monopolies, without a single company having to foot the bill for “commercialization,” and without an “investment” of “substantial resources.” It’s just that AUTM’s operating model is that such products and services should be delayed or suppressed so that some monopoly patent position can be exploited to create a product or service. It’s bizarre, even neurotic, but that’s the camping trip AUTM is on, and the camping trip AUTM wants you to stay on forever.

And let’s be clear, since the AUTM CEO has worm-tongue: a university seeks to license an invention under an exclusive license. If an invention is ever licensed (rare enough), the invention goes to one favored company, not to “companies.” Those inventions are not managed to “serve the public” (even in a soylent way)–they are managed for their profit potential. Otherwise, we would see a heck of a lot of non-exclusive, royalty-free licenses, where the purpose of the license was to control quality, ensure compliance with a standard, and limit the opportunity for any single company (or interloper) to use patent positions to disable the accumulation of a new technology platform from which many companies might draw to create products and services.

Thus, the AUTM CEO’s tack on of “that serve the public” to the end of “products and services” is empty. We might observe that products and services that don’t serve the public are unfit for commerce. The point of “public service” in university research is almost but not quite exactly the opposite of the AUTM CEO’s bluster–that a university manages inventions not for monopoly exploitation but so that everyone has access, even if by doing so speculative investors with a mania for monopolies don’t bother. Case in point: the digital computer was not patented, and yet we have digital computers. The internet was not patented, and we have the internet. The Salk vaccine was not patented, and yet we got the Salk vaccine anyway. These are not exceptions that prove the rule. They establish the rule for public service.

This entry was posted in Bayh-Dole, Bozonet and tagged , , , , . Bookmark the permalink.