Here is a short form of the argument that nothing is a better way than the approach to technology transfer, IP, licensing that universities have at present.
The present approach universities take to IP management/technology transfer does not work, has not worked.
The approach is wasteful, ineffective, damaging, even corrupt. It’s not working. It’s very. bad.
Despite what you read, the whole effort has been a huge clusterfuck of misrepresentations, bungling, pettiness, and half-truth reporting. Big picture, over the past forty-odd years, universities (and their research foundations) have obtained around 150,000 U.S, patents on maybe 300,000 inventions reported (and many more claimed). They are presently patenting away at about 7,000 U.S. patents per year. If obtaining a patent costs say in round numbers $10,000 then we are talking about a spend of $1.5 billion spend over the past forty years just on patenting. And that does not count maintenance fees, or all the failed patent applications, or the foreign filings. So let’s say a big nice round $2B. What have we got for that $2B spend–and what of that would we have got anyway?
Pretty much nothing. Most of these patents have not been licensed, and of those relatively few that have, most have not resulted in “practical application.” The University of California once estimated that only 1 in 2,000 inventions in their program achieved commercial use. Stanford a decade ago reported only 53 of 6,400 inventions over a span of 36 years earned sufficient income to be considered even modestly commercially significant. And those are top programs. Not much has changed. It’s still a bad approach.
On any measure–technology transferred, practical application, making money for university research and instruction, economic development–the university approach to IP simply stinks. Most technology claimed by university IP policy dies behind bureaucratic paywalls. Just when there might be the greatest interest in something new, university policy folks swoop in with sweaty thumbs to take it away and hold it for some hoped for exclusive commercial deal that never comes, or if it comes doesn’t result in anything but some payments, or if it results in more than that, it likely disserves the public with unreasonable pricing and suppression of development of other aspects and applications and improvements of the claimed invention.
Most university technology these days is never licensed and universities demand a license in order to permit transfer, and even when rights are conveyed, there’s no indication that the technology covered by those rights is transferred, or adopted, by the licensee companies. It is not that technology transfer is hard–perhaps it is–but that university IP policies and practices make it a whole lot worse. So much worse that no policy would be way, way better than what universities are doing now.
Universities don’t even bother to track post-license use. Nor do they track the consequences of relying on exclusive licenses, which are hard to get and introduce all sorts of additional problems for an industry and for research relationships with universities. Practical application, for instance, means that products based on the invention are available to the public on reasonable–as if competitive–terms. Universities don’t bother to track that either, even though it is at the heart of Bayh-Dole, which they say they love, but you know they are cheating on it every day.
And most university licensing programs lose money. The financial model is to try to get one big hit per decade to float the office financially and give the appearance of success for all the other inventions claimed by the university and sent away to die. That’s the 1 in 2,000 metric again, in which 1,999 inventions get killed by bureaucracy in the hope that some lucky one might make huge money. Keep looting until you break into the right store and snatch something of real value. Or, blandly to hide the truth–“winners pay for losers.” I use “lucky” with intention. Even when a university gets a big hit license, it is typically not their management process that is the driving element. Those deals are more a result of luck, and happen despite the university’s management of IP, rather than because of it.
The path to improving things does not lie in fixing the present system. It is unfixable. It must be abandoned.
You don’t help an alcoholic by trying to improve the choice of liquor.
Universities have tried to fix their approach since Bayh-Dole was passed. They have demanded more ownership, they have skipped talking to industry before filing patent applications, they have started suing the companies that sponsor their research–over patent rights–, they have limited royalties shared with inventors, they have expanded what they claim, they have made university personnel sign away all rights to future inventions before those inventions have even been imagined. They have fixated on exclusive licensing. They have resisted allowing companies to have continuing access to improvements and background rights and alternative inventions made in research once a company has taken an initial license. They have added “poison pills” in licenses to prevent companies from cross-licensing or contributing rights to a working standard. They have tried to make Bayh-Dole appear to endorse their practices with ludicrous claims about how the law works, and disingenuous claims about how well the law is working. Gosh, nothing seems to have worked.
It’s not just that the folks involved are the wrong folks to fix what they have created–that would involve admitting they have clusterfucked up, and they have no record of being able to do that. It’s that the approach itself is antagonistic to innovation, to opportunity, to the role of a university, to the role of public research.
Folks have to recognize that no approach at all is way, way better than the present approach.
No approach relies on the defaults of Dubilier and its dependent concepts of shop rights and equitable title. No way is a better way. There may be things to do beyond the defaults that improve particular situations–say, disclaiming equitable title even when it might be indicated, in favor of keeping title with inventors or with their assignees (when not the university). But in general, the starting point, the thing that is immediately do-able, is for a university to renounce its IP policy, go back to the defaults, and quit wasting everyone’s time. Sure, keep a licensing program/tech transfer office/startup program. But make it voluntary. Don’t need IP policy for that. If you are any good, then some people will choose to work with you. And even if you are really good people will still choose to work with others anyway–and for good reasons. In all, people like choice. Indeed giving people choice (or not taking choice from them) creates better working relationships with better chances of transferring technology. Even–people in companies will pay (gladly) for good service and treatment. They don’t pay more with threats, and a university does itself no good at all when it makes most of the value of an invention lie in preventing others from practicing it.
No way does not require Bayh-Dole to be repealed or fixed. If a university does not take ownership of an invention made in federally supported work, and the university does not make the inventors parties to the federal funding agreement, then any inventions that are made are never subject inventions. The university does not have to own them, file patent applications on them, or even report them to the government. There’s a huge savings of worthless paperwork that no one reads, no one cares about, never helps innovation.
Once we are back to no way is a better way, then we can restart the discussion about how to do things better with no way, or to add even better ways. But technology transfer may be unsuited to the application of management processes. Consider that–that technology transfer may be a thing of opportunism, personal judgment, chance relationships, and the ability to move quickly. Technology transfer may depend on articulate, committed technologists, on generosity, on reciprocity, and on an ability to carry many common relationships rather than on a single exclusive license with a patent monopolist willing to share with a grasping micromanaging slightly loopy university licensing office.
You won’t find personal judgment in university IP policies. Nor generosity. University IP policies don’t enable a university to play the role of rainmaker, or of curator for a common good, or to raise funding or visibility for something it does not own and would ruin if it did demand to own. So, make those policies go away. Start afresh. See what happens. Really, truly, it cannot be worse than what universities are spending tens of millions of dollars per year to screw up. If they tell you they are doing excellent–look at how many inventions are reported to them, how many patents they’ve got, how much licensing income they have, how many startups–it’s all a scam unless they also tell you how many of those inventions they have placed into public use. And there, the answer is dismal.
A university that rescinds its IP policy takes the first step toward significantly improving its profile for technology transfer, economic development, and even for making money.
But universities are IP drunks. They are not going to change easily. A change would involve a conversion experience–confession of being wrong, on the wrong path, on the wrong side of innovation history, a “my god, what have I done” Water Flowing Underground moment. Universities never apologize. Yale will never apologize to the late Professor Fenn. Perhaps change in federal law would work a change, but repeal of Bayh-Dole won’t do it. Bayh-Dole was an agent used to establish the present approach. It is not needed to keep that approach. So any change in federal law will have to give incentives for universities to de-bureaucratize technology transfer and claims on inventions made in federal funded work.
Think about it Imagine. Imagine that good ideas might come out into broader use without the compulsive interference of a university middle-office. Imagine an approach to innovation or economic development that does not require university administrators to have a thumb in every opportunity pie. While there are all sorts of expensive programs for IP management that might be tried, there’s a really simple program that costs next to nothing, saves millions of dollars a year, pulls the university back from bitter, threatening, blundering, bureaucratic relationships with its inventors and with companies. It is simply not that hard or expensive to repeal IP policy and unwind a compulsory IP management program antagonistic to academic freedom and freedom to publish and that claims everything, produces next to nothing, fixates on exclusive patent licenses, uses the threat of litigation more often the opportunity to provide access, makes its money on the value of exclusion more than the value of use.
Start with the Dubilier defaults. Make technology transfer voluntary. Focus only on inventions for which the inventors want to obtain patents, and then only to consider helping them out with patent costs, as a kind of assistance grant, along with other kinds of assistance grant. Stop the institutional envying of folks that make something with potential value. Embrace the rainmaker role. Release university claims. Consider getting involved only when asked. Get out of the patenting business. Get out of the exclusive licensing business. Get out of the patent enforcement business. Stop trying to run the value of industry relationships through licensing policy. Stop, even, thinking about “managing” innovation–such a suppressing, bureaucratic thought. Stop being Moloch.