Moving to a new platform that’s really what Vannever Bush first proposed
If you see this difference between an approach that transfers the government’s right and the Bayh-Dole approach, which attempts to transfer ownership of patentable inventions directly to institutions, then you can see how stunk up the IPA and Bayh-Dole both are. We’ve had nearly sixty years of such stunkupidness. Perhaps, however, you know someone with delicate feelings about the government just designating organizations to take over patenting of inventions made with public support. Let’s help sort out their feelings. The outcome is roughly the same as in present practice, but the funding relationship with universities and inventors is much more convoluted with Bayh-Dole. In the approach I’ve outlined, the conditions run with the option on the government’s rights. Instead of issuing a patent to itself, the government designates a non-federal entity to which it will issue the patent. Then it’s up to the non-federal entity to do something with the patent that meets with public approval, however that standard is defined in the option–make a lot of money, scare the bejeezus out of industry on behalf of the government, develop the invention into a commercial product while excluding all other uses, license non-exclusively to create a standard or commons, or release the invention with butterflies and rainbows for everyone to use.
In present practice, Bayh-Dole leaves the exploitation of a patent on a subject invention up to the patent owner. Under Bayh-Dole, a patent owner can do pretty much WTF it wants, including assigning the invention to someone else (call it an exclusive license and everyone ignores the transaction), and the assignee then decides how to manage the invention, including sublicensing the invention to yet someone else or selling off a part of the company including the exclusive license agreement. The patent monopoly can thus be traded multiple times, so “technology transfer” can involve patent speculation for a number of transactions without any actual use of the invention or sale of product based on the invention. That wouldn’t necessarily change under a simpler federal option approach–it’s just that the transaction would be much more out in the open. The research host gets rights because it is offered access to the federal government’s rights, not because it is given out the appearance that it is entitled to those rights because it is also an employer or is in the business of profiting from research inventions, or is a public entity that cannot permit anyone to benefit personally from the use of public resources. It would hold rights because the federal government designated it to hold and manage the federal government’s rights.
If the worry is simply that a convoluted, incompetent, unenforced approach is preferred over a direct one that makes clear that the government gets to choose its favorites, then we are talking about tone not substance. And we are also talking about keeping something away from public notice, rather than being up front about it. Oh, I know, the public are brutes and wouldn’t understand, except that APLU and AAU want everyone to understand that the present approach is wildly successful and look around at all the transformation of society created by that 0.1% of patented inventions and 0.1% of startup companies and those 99.5% of federally supported inventions that have been fortunately kept from further development–spurring progress in the useful arts necessary to design around them or do without. How glad we can all be. Do I need a /sarc tag here or are you with me?
Of course, there are good reasons not to disenfranchise academic inventors or to create for university administrators unmanageable institutional conflicts of interest. And there are good reasons not to let any random university licensing office manage the next society-transforming invention. There are good reasons why neither UC nor MIT nor Harvard should manage CRISPR patents–and good reasons why they should not be spending a pile in legal fees to fight it out. It’s just that if we try to have a discussion of alternatives, special interest organizations like APLU, AAU, AUTM, COGR, BIO, and PhRMA are ready to kill it before it grows.
There are still independent investigators around, even if universities have largely forced them underground or into non-institutional settings. The open source/open hardware movements still operate, largely in defiance of university patent policies and patent administrators. There are still investigators that value open publication and don’t consider their work well served by patents, especially patents that allow owners the full set of monopoly rights, including excluding all uses or scaring the bejeezus out of industry by sitting on an invention until it’s widely used anyway and then trolling for fun and profit.
There were consequences for the federal government to get into the research business. One was the transformation of universities into administrative reliance on federal funding, so that even a small change in funding results in panic attacks and alarm that all future good things will be lost and society won’t be transformed further. Another was the repudiation of the role of the inventor in developing an invention for use. Rather than forcing inventors to give up rights to bureaucrats, the federal government might have considered requiring inventors who filed patent applications to give up further federal research funding while they pursued development of their work–or offering inventors federal development funding for the effort rather than research funding.
There isn’t an obvious solution to what I call the Feyerabend dilemma. Paul Feyerabend observed that in science, “anything goes.” Despite claims about scientific method, discovery appears to happen in all sorts of ways, with the backing of all sorts of ideas that are often not even close to the consensus scientific position. Consider the cause of stomach ulcers–not stress or spicy foods, as it turns out, but a bacterium. But how can a federal government fund just anything to advance the frontiers of science? If it did that, why, it would fund the nut cases–you know, the folks who think that continents drift or that four molecules of CO2 in 10,000 won’t cause catastrophic global warming or washing hands before assisting with childbirth prevents infections. Despite the idea that “anything goes” in science, Feyerabend also expects public accountability. There might be the “free play of free intellects” (as Vannever Bush has it), but those intellects, if they take public money, ought to be accountable for what they do, and do something mighty useful and practical and not worry about dark matter and string theory or proper career science.
Thus, the dilemma–how to advance the frontiers of science with public money, justify funding the outliers, and still make the funding accountable to the public? Vannever Bush tried and failed. He got as far as public funding for science, but that funding was quickly institutionalized and brought back into the service of establishment interests, if not what government agencies wanted then what panels of scientists agreed was proper to fund. It’s not that agency administrators (who also may have great scientific experience) and panels of scientists can’t see useful things to do–it’s just that as things get toward the frontiers, most everyone doing something unknown appears to be a nut case. Or, as Feynman had it, “we are just too stupid” to understand what is going on–that is, nutcases who turn out to be right make all of feel like chronic nutcases. In that case, perhaps it is just as well to let inventors decide what they are to do, without any public covenant for patents beyond that provided by antitrust law.
Perhaps, as well, it would be best not to let any institution exploit access to federal funding to force independent investigators to give up their interest in inventions. Institutional capture of public resources is a constant practice of the Moloch state. Institution-building can be a good thing, but in the exploration of the boundaries of science (and the boundaries of observation, explanation, and what’s possible in the physical world), institutions can easily dictate the bounds of curiosity and sanction what ought to be encouraged. It’s just that institutions are not very good at knowing what, as a matter of uniform practice, should be encouraged.
Thus, we come back around to Vannevar Bush’s thoughts–that confidence is necessary rather than fear; that the choices in what to fund ought to be made as much as possible outside establishment science, which has its own role and benefits; that those working to explore should be free of demands on their work, even as those that fund them are held accountable for what gets done. And in all that, perhaps the federal government does not need a public covenant for private use of inventions made with federal support, so long as it is the independent inventors that have made the choices and take the actions. Let the patent system have its day.
There might be reasons to invoke a public covenant for the institutions that provide assistance–the research resources, the funding–and there might also be a review of what inventors have done, when they do obtain patents on their work, before they receive more federal funding. In this, there may be ways to address the Feyerabend dilemma and for at least a portion of federal funding, create an environment that starts with the confidence of the “free play of free intellects” provided with institutional-scale resources, with both those that provide the funding and those who obtain patents on their work (and thus have a right to exclude all others and not use the invention themselves) accountable for what they do with those patents before they get further federal support.
Vannevar Bush had experience with three areas of scientific development. In the first, establishments managed science. This is the world of procurement of research services for a purpose. Contract research runs in this arena. This is the arena of military development and medical studies. People ask for what they can imagine is possible within their budget, time frame, and performance review. In the second, a skunk works manage development. This is what Bush led during World War 2, providing technology to the military derived from new science, clever science, an awareness of industrial practices, and the ability to make new things. This was organized research, outside the specifications of establishment research, but done with directed intent to contribute to establishment interests–by making current problems and practices obsolete.
But there’s a third area. And that’s independent exploration to expand the boundaries of science. If there are boundaries–if you accept a geographical metaphor for things of the mind–then what Bush realized is that by taking in new phenomena, learning to recognize it and characterize it, we would have more resources available for both establishment science and directed skunk works technology development.
It was the independent exploration that Bush was concerned with, the new science that made the skunk works engine go. Roosevelt wanted Bush to adapt his military skunk works to civilian uses, starting with health care. Work outside of the medical establishment, but on behalf of better medicine. What do you need to do that? Bush needed money free of establishment controls and researchers independent of government and the medical industry and the professional medical establishment to do the exploration. These, he thought, might be found at universities, where they had made a commitment to exploration, to a sense of truth about the natural world, an integrity with regard to discovery, with purpose in mind–whatever that individual purpose might be.
Combine those discoveries–and the people who understand those discoveries–with people who know industrial practice and with people who have a talent for making prototypes (then, gadgeteers, mechanics; now, makers and programmers), and you might have a chance of building new things that–hey–transform society. Isn’t this what Elon Musk is doing with cars and space exploration? Isn’t this what Jeff Bezos is doing with on-line commerce and backend web services? Don’t they operate as it were well funded skunk works, drawing on new science, industrial expertise, and a happy desire to obsolete much of what marches along to establishment expectations? Look at the great recent entrepreneurs–Gates, Jobs, and the like. They make some of the science they need, and they draw on the available science at need, and they make things that the status quo couldn’t bring itself to make, even if it could imagine it.
It’s strange to argue that federal patent policy is the crucial piece in all the expansion of federal funding for independent research, but perhaps even more than the choice of who and what to fund, it is the patent policy that expresses, if not dictates, who will participate in that research and what they are willing to propose in order to receive federal support. We might think of Carl Sagan’s allegory on the matter in Contact, pitting establishment science against the nutcase looking for messages from little green men in space. Sagan’s message, if I read it right, is that the nut case can be not only interesting but necessary for expanding the frontiers of science. Perhaps that’s an illusion. Perhaps only the orderly management of money for the best and safest of all proposals for research is the only way we learn anything new about the world. And perhaps pigs have wings–one day, with GMO techniques, I’m sure they will. But if it is illusions, then it is illusions all the way down and we are left having to choose among illusions. I say that there’s room in those illusions for the illusion of confidence in the integrity of at least a few scientists to explore independent of institutional claims on their work.
A good place to start is a federal research patent policy that ensures that investigators are responsible for their work and that any institutions involved have no right to that work. Bayh-Dole is nowhere close to that standard, and in fact goes a long way in the opposite direction, arbitrarily (and in a bungling way) handing control of inventions to institutions and dismissing the inventors with a financial interest without any expectation of institutional diligence. A few inventors get lucky and make a fortune. A few more work the system and push the rights to their startups and with hard work and some luck also make a fortune. But most inventors–99.5% of inventors–see nothing. Their involvement in institutional control of their work is a net loss for them–in lost time, lost opportunities, lost collaborators, lost referrals of interest. That’s the effect of Bayh-Dole, something that it’s difficult for academics to study or federal accountants to count and present in statistical tables. We see 950 startups for 2015 and not the failures of startups for 2010 or the accumulated institutionally owned inventions not available to anyone–now perhaps on the order of 30,000–and every day the paper boy brings more.
The simple way to manage federal patent rights in research contracts is for the government to require assignment of all inventions to the government and then offer an option on those government-side right to others. But that way disenfranchises inventors, has the government issuing patents to itself only to then market that right so that the government issues patents to any other favorite but the inventor–something quite alien to the U.S. patent system, an end-run around the patent system, as it were. No “public covenant” can really make up for the architecture of this exploit. It would be better that nothing developed with federal support constituted patentable subject matter. But even that means that federal funding for research is something of a scorched earth approach, eliminating the interest of investigators in their inventive work, merely because they have asked for and received federal support. Perhaps if there were no patents at all available for an area, such as the practice of medicine, we might be better off. Then federal funding would not bias ownership but only, perhaps, whether findings made with federal support could be withheld as private trade secrets.
The effort to expand the frontiers of science has been formalized and institutionalized into a system that has forgotten its compass. University patent management can still advertise the idea–that research leads to public benefit, so be sure to fund more and more research–but the administrators behind things like the APLU and AAU fakographic have no idea what actually happens or what could happen or what would be really important to happen. Their interest is in preserving the status quo, the formalized and institutionalized federally funded research “system.” “Research as an industry.” Research as geographical pork. Research where the results matter so little that people count publication citations to determine importance and status, count patents as proxy to innovation, and think that it’s all good if there’s money to be made by nonprofits trading in monopolies.
If the federal government wanted to do something that would restore the role of the independent investigator, it would start by making the (f)(2) written agreement expressly exclude institutional claims on inventions made with federal support. The inventions would still be subject inventions by operation of the (f)(2) agreement and the definitions of funding agreement and contractor, but the matter of invention ownership would be between the inventors and the federal government. Only if inventors decided to file patent applications, and place those applications with institutional managers would the transfer be allowed. In any case, inventors would then be considered for further federal funding only with an accounting of what they did with the inventions for which they sought patents and what the institutions and companies did with the associated patent rights if and when they acquired those rights.
If the federal government is to take ownership, then it cannot be in the business of issuing patents to itself only to re-issue them as private favors to companies with the idea that commercial development requires a monopoly that runs for the life of the patent. In this, the most pernicious part of Bayh-Dole is the part that permits federal agencies to license exclusively (and expressly, to give up standing to enforce patents to private interests–see 35 USC 207(a)(2)–that is, essentially assignment or patent re-issue). Further, given a design element of the U.S. patent system is that the federal government does not take a financial interest in the exploitation of patents, then that must also be the case in the management of any patents the government issues to itself or allows to be issued to anyone other than the inventor or someone to whom the inventor has voluntarily transferred ownership (and not as a condition of federal funding or use of resources paid for with federal funds). Thus, whatever the federal government does with patents it owns, it must do it royalty-free. Payment is not an option. Requiring actions responsive to public purposes may well be–manufacturing products and selling them at reasonable prices (not monopoly prices) might be acceptable actions under a limited exclusive license.
We could get into the dynamics of cumulative innovation, of how to use patents to create commons while still allowing the advantage to go to those that work for that advantage. But it’s enough to see the outline of the policy: focus on the investigators, not the institutions; hold the inventors accountable for their exploitation of patent rights with regard to any further federal funding; hold federal program officers accountable for how inventors have exploited their patent positions; hold any institutions that agree to take ownership of these inventions accountable for what they do with the patent rights they obtain. If folks want to have the right to exclude others, then let’s see how they get on with that. If, however, the government issues patents to itself, then those patents should be managed without a financial motive–royalty-free, but with a purposeful working requirement.