There are no Bayh-Doles in Canada

In a recent hearing held by the Canadian House of Commons Standing Committee on Industry, Science and Technology, Michael Geist, a law professor at the University of Ottawa, made some valuable remarks, which he has published at his blog.

Geist argues that the emphasis on Bayh-Dole style patent commercialization is misplaced, and what is needed is a stronger emphasis on knowledge transfer and protections against what he calls “IP abuse.” These are worthwhile things to consider. Here’s Geist on Bayh-Dole:

Further, the potential emphasis on the U.S. Bayh-Dole approach is misplaced. As you heard from department officials, there is little evidence that the policies governing who owns IP rights have an overriding impact on the success of tech transfer as measured by the volume of patents and licenses.

There’s little evidence because Bayh-Dole makes the evidence subject to FOIA exclusions (well, not really–Bayh-Dole bungles the FOIA exclusion but since everyone “knows what was meant” the words don’t actually matter much). The evidence that’s out there–in bits and pieces–indicates that most university patented inventions are never licensed and of the few that are, most never become the basis for a commercial product. The effect of the patent, then, is to ensure that no casual user can practice an American university invention. A bureaucrat’s thumb must be in every innovation pie. That’s a handy way to put it (so to speak).

More Geist:

This should come as little surprise to anyone who has spent time on campuses with academic researchers.  The metrics of success in the academic environment – publications, grants, tenure, chairs, successful students – have little correlation with commercialization.  Even for those with commercial interests, those are often achieved through consulting arrangements or other mechanisms where the business expertise is left to business people.

We might expand the thought: the metrics of success current on university campuses have little to nothing to do with success. It doesn’t matter whether academic metrics correlate with commercialization metrics if the metrics in both cases are full of merd. What does it matter how many publications one has? I have 900+ articles up on Research Enterprise, and I’d give them all up for one that made a significant difference in the practice of university IP management. What does it matter how many patents or licenses? None of this has to do with the virtue of phronesis, of knowing how to get something done and getting it done–unless, of course, success is creating a pile of publications or patents.

As for “commercialization,” the turning of university-based inventions into commercial products is a narrow bit of the overall use of those inventions; using a patent monopoly in an effort to induce development that otherwise wouldn’t happen means that all DIY use is excluded in favor of the effort involving the transfer of a monopoly: the university issues its own private “patents” on research in the form of exclusive licenses. Think about it. There are inventions that indeed require private investment to create a product–but why should the effort to do so prevent anyone who wants to to make a DIY version of the invention for their own use?

But here is what’s worse–that there are inventions that anyone could make and could make into commercial products, but the university’s patent prevents them from doing so, so that a favorite company holds the sole right to practice the invention (and usually that is a class of things, perhaps hundreds of variations across dozens of claims and functional equivalents, applications, methods, and the like). That is, the patent has value to exclude practice *before* a commercial product is made, rather than to exclude practice as “free riders” *after* a commercial product has been made that required substantial private investment but is then easy to copy.

While there might be a federal policy that gives a limited benefit to companies willing to develop a research invention to product form for general use, and hold a brief monopoly after in order to recover their investment, why should that same federal policy allow also for companies to hold a monopoly for the life of a patent to exclude all other users, even when the required investment is nominal for the industry? That doesn’t make sense. Further, why should research uses and DIY uses necessarily be caught up in any effort to “commercialize” an invention? Why should those uses be suppressed, simply because a university administrator determines that a commercial product is desirable and that the means to obtain that product would be financially more favorable if the university offered a complete monopoly rather than a partial monopoly that permitted DIY use to compete with commercial product versions of the invention?

Geist makes a good point, too, about commercialization through consulting. Once a university demands to own all inventions, then someone who might otherwise pursue a line of inquiry within the university might instead choose to do that work–or enable others to do that work–outside the university. Consult, take a leave of absence, quit, send the idea to others, publish the idea generally. University administrators have a difficult time when they claim they can dictate where faculty do their work, who they can talk to, what they can disclose. The administrators are on thin ice when they define “invention” to include “non-patentable inventions” or “know-how” or “anything we can sell for value or hold long enough to make you feel bitter about us.” These sorts of things have no meaningful definition–anything an administrator wants, he or she merely calls an “invention.” If institutional ownership means an idea has gone to die, then those who make ideas that matter tend to prevent institutional ownership. If a university’s patent policy is then extended to attempt to keep ideas from “escaping”–for instance, by asserting non-compete covenants or making disclosure appear to be an ethics violation or conflict of interest or insubordination–then the university becomes oppressive, not open.

Geist then gets into the issue of the premise for licensing income:

I would argue that the emphasis on university-based patenting is misplaced. It can have a corrosive effect on universities, who forego important, publicly-funded research in favour of potential licensing or patenting opportunities.

It’s certainly true that university-based patenting has a corrosive effect, but Geist doesn’t quite stick the point. I don’t know of any university that foregoes public research for patenting. Administrators don’t have that control, and besides, they typically want as much research (money) as they can get. More often, they are willing to make industry-sponsored research more difficult. As one director of sponsored projects told me, “I wish industry-sponsored research would just go away.” All federal, all the time. That’s what all the “uniform patent policy” and “title certainty” hand-waving is about–allowing administrators not to care about the sources of funding, so that anything created ends up owned by the university and managed for monopoly value. Sell things back to sponsors or threaten to sell them exclusively to a sponsor’s competitors to drive up the value.

An exclusive license to one company turns all other companies into competitors to the university’s line of research. That’s corrosive. Public money can still flow in for research, but if that money is then funneled to the benefit of the exclusive licensee, it’s really a government subsidy for one company over the others. That in turn pits the government, through the actions of university administrators, into choosing favorites and competing with the rest of the industry. There may be reasons for a government to compete with an industry–but using a process of funding and licensing that’s made to look like it’s based on merit and public interest is deceptive.

Geist does have a point–in the history of American university patent licensing, an early justification was that patent licensing income provided more research support. But that argument largely became moot with the rise in federal funding in the 1960s and after. WARF created the actual engine for research funding–take some patent licensing income, invest it in common stocks, play the markets for big returns, and gift a share of those returns to whatever the university administration wants (within reason of course–WARF refused to fund social sciences and anything having to do with public service, hence the Bayh-Dole restriction on the use of licensing income to “scientific research or education.” No social sciences or public service funding allowed).

I doubt Geist could come up with any evidence that university faculty forgo any important publicly funded research in favor of patent licensing opportunities. But there is also no reason for universities to claim that they need the patent licensing income to support vital research in the public interest. Patent licensing income, when there is any, is typically administrative slush money. No one accounts for it. What it supports is largely meaningless in the big scheme of things. One might then argue that university patents should be licensed royalty free unless a company *wants to pay a royalty* and so turn a license into a binding contract. If the patent is to be managed in the public interest, then let the public pay for the costs of that management. If the public’s interest does not justify the $30K or so that go into obtaining and managing any given patent, then why patent at all?

With properly funded institutions, there is no need to chase licensing dollars. Instead, the cutting edge research ends up in the hands of businesses who can better leverage it for commercialization opportunities.  This should not be viewed as lost revenue for universities or their researchers, but rather as a better return on the public’s investment in post-secondary research.

Well, the funding for institutions is different from the funding of extramural research. As long as faculty have salaries, they can teach and consult and publish. If they want to do research, they come up with the sponsors who agree that the research is worth doing. That’s not a matter of institutional funding, but of making a line of research meaningful for someone (or some group) willing to support it. That’s not an entitlement.

Geist is spot on about “chasing licensing dollars.” Industry that might wish to support research rather than a university office dedicated to “extracting value” or “monetizing” a patent. That means patents get used for a purpose different from the one that a patent troll values–and most university patent licensing is built on a troll model as a fall-back to a monopoly trade. (Seek an exclusive license; if that fails, sue infringers; give each exclusive licensee the right to sue infringers; if the exclusive licensee fails to create product, then it can sue infringers to recover its costs, or the university can sue–that’s all based on a troll fall-back position). If commercialization did not mean “trade on a monopoly” but rather “rally support to create a product version readily available to all,” then university technology transfer would have a radically different practice, policy, and defaults. (I know, I spent nearly two decades battling the patent monopoly mindset in universities–and had 3 to 1 returns on investment, at that).

Geist then turns to publishing:

If the currency of academics is publishing – not patents – then the challenge is how to ensure that the published research ends up as broadly distributed as possible.

The premise doesn’t really matter–the challenge is how to make publishing effective. But it’s not a meaningful dichotomy between publishing and patents. A patent is a publication. That is an essential part of its public character–trading publication of what otherwise would be a trade secret or indifferently held in exchange for the right for a limited time to exclude others. The problem is why is it that anyone would want to exclude others from practicing what has been otherwise published, that people want to publish. If academics have chosen not to hold trade secrets and publish, then either a patent provides a form of impactful publication not otherwise available in academic journals, popular press, books, and the web, or there’s no point to it. The commitment to publish is already there. There’s no need for a public offer of a limited monopoly in exchange. Being appointed to a public university and paid a stipend to support research already represents the fundamental exchange with regard to publishing.

The more significant problem with patenting has to do with the exclusion of practice that may conflict with publishing. If a publication represents an academic product, then any publication that doesn’t carry the proper patent markings ought to work to limit the scope of enforcement for infringement that a university has in the patents it obtains. The truly horrible thing is that published research can be broadly distributed but no one can practice what the published research teaches unless they compete for the one license that the university administration is willing to grant. That’s horrible. What does it mean in practice? That companies advise their scientists and engineers not to consult the academic literature–treble damages for willful patent infringement if they do. Bad patent management undermines publication, works against knowledge transfer. It’s not enough that a publication circulate broadly–anyone who reads the publication should be free to practice what the publication teaches. That’s the problem with patent management that does not include broad, non-exclusive licensing–it undermines publication, turns publication into advertising. Knowledge transfer in such circumstances is stupid transfer: “I now know something that I must find a way to avoid or circumvent because I’ll never get the exclusive license and I don’t want an exclusive license anyway.”

When insulin was discovered and developed, the three inventors assigned their rights to the University of Toronto for $1 each to resolve a dispute among themselves over patenting. The university then licensed rights non-exclusively. The purpose of the patent was to control quality, to make sure that anything that called itself insulin was manufactured to a high standard and really was insulin and wasn’t contaminated or degraded. The patent had a purpose–but it wasn’t to induce speculation based on a monopoly. It may be that “knowledge” transfer is more important than “technology” transfer–but knowledge transfer without the right to technology transfer is next to worthless.

To add to Geist’s point about open access publication–it’s not simply a matter of publishing articles that are not behind paywalls. There’s also the matter of the data from which the article draws–raw, corrected, analyzed–and the data structure and formats, the tools for data acquisition and manipulation and presentation. If those are off-the-shelf products, fine. But otherwise open access publication also means public deposit of data and unique data management tools. If an analysis has been coded up in software, then that software code, too, is essential to determining the validity of claims–there may be coding errors, for instance, or the algorithm present in the code does not do what the publication claims. Knowledge transfer is not merely reading. It involves establishing new practice, not just abstractions and logic. That is, the ability to use and to do–we might say, then, that technology transfer is an essential form of knowledge transfer, wherever the knowledge depends on data and the tools of observation, collection, analysis, and presentation of data. Whatever the desire for commercialization, that desire cannot and should not preclude open access and public deposit of data and research tools necessary to assess that data.

For government-funded research, it does not appear to be necessary to mandate that all publications be “open access.” As long as the investigators provide the government with reports of the research, the government can make those reports “open access.” Whatever other expression of the findings a research wishes to make–in whatever forum–ought not matter so much. But if the claims presented are based on data (and not just, um, logic and reasoning), then there ought to be some way to prevent the government from taking action based on such publications without public (and government) access to the full data set and the means to manage that data (for the U.S. attempt, see 2 CFR 200.315(e)). Besides, we are finding that much of what is passed off as “new knowledge” in academic scientific publications turns out to be wrong and even when not wrong turns out not to be helpful. Perhaps we might think that making announcements of crappy research less accessible would be a public good. Why should companies be induced to “collaborate” with academics who publish stuff that doesn’t hold up? Consider the problems–and cost–that Amgen had when it found it couldn’t replicate most of the cancer studies that it investigated. See here and here.

Geist turns then to “anti-IP abuse law”:

Given that Canada already meets or exceeds international standards on IP, a key concern is to address the abuse of IP rights that may inhibit innovation. The Canadian government could address the issue through an anti-IP abuse law.

Geist’s interest is in patent trolling and forum shopping. Good stuff, of course, but that’s sort of like extending antitrust law so that it’s effective in matters of patent rights, and that requires people to consider a patent as a government grant in trust rather than as a private property entitlement, and there’s a ton of work there to overcome the rhetoric both of private property and of public interest.

I’d suggest a different direction for “anti-IP abuse law”–how about limiting the claims that non-profit employers can make on the IP of faculty? That is, create a “freedom to innovate” clause that limits what a nonprofit can call a “service” invention involving federal funding–how is it that inventions made with public money can be transformed by administrative demands into inventions made to serve the financial interests of the university employer? If the nonprofit overclaims, then it loses its nonprofit status–or further government research funding.

One might also consider a limitation on government and non-profit “re-issue” of patents using overscoped exclusive licensing, especially for inventions that have been made with federal support. This is what Bayh-Dole enables that’s most pernicious. The government funds research resulting in an invention. First the university that hosts the research claims the invention, so that the government issues a patent to the university. Then the university, as owner of the patent, in essence “re-issues” the patent to a favored company using an instrument labeled “exclusive license” but in effect the invention is assigned to the company for royalties. Skip all the rationales and focus on these transactions. The university operates as a non-federal patent office, re-issuing government-issued patents, but not to the inventors based on their merit but rather to favored companies. Ban private non-profit patent office trade based on government-supported inventions. Either the university assigns its entire right, title, and interest and gets out of the business or it grants non-exclusive licenses or true exclusive licenses (ones that do not involve a grant of all substantial rights in an invention, do not permit the licensee to enforce the patent, do not permit sublicensing or settlement of infringement). Better–the university gets out of the patent business entirely, unless it is asked (voluntarily) to become involved to manage a patent for a reason other than re-issuing it to a favored company.

Nonprofit patent re-issue of federally supported inventions is perhaps the worst bit of IP abuse enabled by Bayh-Dole. Universities run shadow patent offices. Where those universities are instruments of state government, essentially they function under the old regime of patents as government monopoly favors–what the U.S. patent system was designed to preclude. Maybe the Canadian patent system has a different foundation, but in the U.S., the idea was to exclude the government from using a monopoly right to generate money for itself or to play favorites among friends by granting them monopoly rights. The monopoly goes to the inventor based on merit and the inventor’s free choice to request a patent. If government contracting sets up to circumvent this basic exchange, placing ownership of inventions with research hosts rather than inventors, requiring inventors to apply for patents (even when they do not want to), and then giving the research hosts the right to re-issue those patents for profit and for favors, then we have regressed to the patent system that prevailed before the U.S. constitution put an end to it in favor of inventor merit only. Bayh-Dole then becomes an ugly, regressive statute that returns us to the petty favoritism of the pre-U.S. patent system, and its objective to “use the patent system to promote the utilization of inventions arising in federally supported research or development” turns into something of an ironic parody–in practice, universities use the patent system to circumvent the patent system. IP law abuse.

There are no gnomes in Sweden, and there should be no Bayh-Doles in Canada–or in the U.S., for that matter.

This entry was posted in Bayh-Dole, Commons, Innovation, Technology Transfer. Bookmark the permalink.