Should universities manage patents like corporations do?

We have discussed how university-owned patents are not like other patents. A university owner of patents is not free to do just anything with a patent. A university-held patent is not simply a property right–there are limitations (though many university patent administrators are happy to forget them).

The reasons for these limitations include federal law, regulations, and funding agreements; a university’s own public commitments and policies; and of course its funding agreements. There may also be state laws. Not every limitation applies to every possible patent, but these limitations cover the majority of cases.

There is a debate that underlies the situation. Should university-owned patents be different from other patents? My answer is “yes”–and I give reasons that back the answer. Certainly there are those who disagree. It would be interesting to see their reasoning.

A different version of the question is this: Should universities manage their patents differently from how corporations, in particular, manage patents? But this question is misleading. You might think that the answer I would give also would be “yes.” But here’s the problem–corporations, for the most part, manage their patents *better* than universities do. Corporations, with an eye on profits arising from selling goods and services, have reasons to trade patents, to let go of patents, to dedicate patents to a commons or standard, and to avoid taking an ownership interest in the first place. Universities have no such guidance. They rarely let go, they rarely trade, they rarely use their patents to augment their core business. For universities, patents are a separate concern. Inventions are forcibly removed from the lab (“disclosed and assigned”) and placed for management with administrators, to make money from licensing. Research, somehow, is viewed as an activity separate from “commercialization,” and so university administrators find it hard to believe that they should spend $15K on a patent only to give rights away to enhance research collaboration.

Patents and university mission run in separate channels, though they don’t have to. Patent policies claim that money from patents will support more research and people paying money for patents are more likely to create products that will benefit the public and that the money from investors going to develop such products will stimulate the economy. Thus, university administrators find themselves with a rationale that is based almost entirely on making money from patents–if not a very, very, very, very, very large amount of cash, then at least enough money to pay off the expense of filing a lot of patent applications. The answer, then, to the second version of the question is more like “no–universities should manage patents more like corporations do, at least those corporations that use patents to increase their development and sales of their primary products.” But an even better answer to this second version of the question is “universities have an opportunity, if not a responsibility, to manage their patents to be complementary both to industry, to artisans, and to the public at large. That is, universities should not manage their patents in the same way as corporations, and they most certainly should not manage their patents the way most universities do now!

A university patent is not like a regular patent. A university patent is often a subject invention–made with federal support. Subject inventions form a special class of inventions under federal patent law. Universities are limited in what they can do with subject inventions. In particular, they must manage patents on subject inventions to meet the objectives stated in federal law for such inventions. Even when an invention has been developed without federal support, universities are limited by their own formal documents, including statements of policy and agreements they make within the context of these policies. A policy statement displaces institutional common law. Once a policy statement is in place, university options for the management of patents are limited to the purposes set forth in the policy. Otherwise, what’s the point of the policy statement, but to deceive and cause harm?

Of course, virtually no one appeals to statements of objectives, whether in federal law or university patent policies (or research policies, or academic freedom policies). To do so might sound naive–doesn’t everyone know that these statements are not intended to carry any meaning? That they are fluff, surplusage, happy words to deflect any critique of how university administrators actually do their work?

Yet if we look at the working claims for university administration of patents–money, benefit from products, and economic stimulation from investment in products–there’s not much to show that for all the complex, grasping, poorly conceived and executed university patent policies; for all the pain and bitterness and delay and goofball drafting of university patent management practice; for all the expense, disruption, threats, and litigation–for all this, there is hardly anything to show for the effort. Most inventions don’t get licensed at all. Of those that are licensed, most never result in a commercial product. The vast majority of inventions never recover the cost to manage them. A few “winners” survive the bureaucratic fondling and make money–but that money disappears into another bureaucratic hole, to be splayed around on petty research, slush funds, and administrative “overhead.”

Every so often, there’s a huge bolus of licensing cash and perhaps a building gets built–to be maintained ever after by a diversion of funds from elsewhere. Or the money is put in an investment pool and the dividends on that money is spread around wherever these’s some bit of need (ignoring, as far as I can see, the Bayh-Dole requirement that the money be used for scientific research and education–not for speculation such as buying up shares of companies and then using some (if that) of the money from speculation for scientific research and education. But then, until the federal government bothers to audit universities for the use of money from licensing patents on subject inventions, the law is just a silly gesture).

AUTM doesn’t publish meaningful metrics of what universities do with the inventions they demand to own. AUTM collects metrics that show how fat technology licensing programs have got–how much they can spend on claiming ownership, how many patents they get, the number of licenses, the number of startups, the gross licensing income. But these are numbers floating in the air without anchors to reality, presented to invite people to make assumptions and draw conclusions of the form of “wow–look at all that activity” and “Bayh-Dole must be a wild success.”

But AUTM’s metrics are largely sophisticated crap. AUTM does not ask for or report what Bayh-Dole indicates as the significant events: for each subject invention acquired by a university, what’s the status of development, the date of first commercial sale or use, and how much money has been received from licenses? Not in aggregate, but for each invention. AUTM further conflates subject inventions with all other “inventions” (including non-inventions), conflates licensing with sales (and clever sales in the form of leases called bailments), and conflates licensing of patents with licensing of most anything else. The number of non-exclusive licenses is a measure of workload, not impact. One “technology”–maybe a piece of software–could have 1,000 licenses, renewed each year (and so counted anew, as well, each year), while the entire portfolio of patents languishes. Similarly, one license can generate 90% of the total income, but AUTM’s presentation invites people to divide the total income by the number of licenses (or patents) and create an absurd “average income” figure–per license, or per patent. And yet the behavior in the portfolio has nothing to do with such averages.

The AUTM metrics are useless for evaluating whether university licensing of patents results in beneficial products or whether the investment in developing such products actually results in any economic benefits. That people lose their investments chasing ill-conceived and poorly managed university inventions is not a particularly exciting metric. The same goes for university money-making. AUTM does not report net income–how much money over costs does a university have each year to distribute? And how does it spend that money, if there is any? Universities don’t report their net, and rarely report how they use that net. It’s secret–and mostly inconsequential, and for all the expense and bother and complexity, fails in its stated purpose to make money (or “to provide resources” or “to show a return on the public’s investment”).

So we don’t have the data we should have. It’s clear that what’s happening is largely ineffective. But university administrators are not about to show anyone what is going on. And academic researchers are, for the most part, too poorly informed to know what to ask for, and so spend their time making spreadsheets of AUTM data and posing questions of the form of “if these data are accurate, then…” when there’s no indication whatsoever that the data repeated by AUTM are accurate, relevant, or worth anyone’s time. Just that most academics generally wouldn’t know. Sigh.

We have seen that subject inventions are different in character than other inventions, and so must be managed differently from those other inventions. The preambles in university patent policies point to some of these differences, and as policy statements, these preambles carry more than just fluff–they set the boundaries for university management. It’s just that university administrators don’t read the preambles as anything more than rationalizations for the policy, rather than as the foundations for deciding what patenting and licensing practices are authorized by the university. University licensing officers are fine with keeping inventions out of use if doing so enhances their licensing leverage; licensing officers are fine with trolling industry for infringement, rather than treating industry use of university inventions as success; university licensing officers are fine licensing inventions to startups that raise money to build products not based on the licensed inventions–so long as the university makes money from its equity in the startups. In any of these circumstances, university preambles (and federal regulations and funding agreements) are violated. People ought to care. Auditors ought to care. Federal agencies ought to care. Faculty ought to care. Journalists ought to care. Folks in STI policy ought to care.

But the reality is–most folks don’t care. Perhaps, then, the reality also is that university technology transfer, for the most part, is so inconsequential, so trivial, so useless, that no one has a reason to care, other than to try to find ways to make more money from the assets bound up in university ownership. If that’s the case, then there’s a follow-up: university research, too, is for the most part also inconsequential, trivial, and useless. Not, of course, for those using university research as a career, and not for the administrators that make their living from the indirect cost droppings–but for everyone else, in terms of impact, we would not notice even a blip if 90% of federal funding for university research disappeared overnight. If no one cares how key results of university research are managed–patentable inventions (and non-patentable non-inventions, too)–then it’s not the patents that are the only problem. The problem is also the research that is producing the patents.

If the research mattered, then people would be riled up about the practices of university patent administrators. But people aren’t riled up. Apparently the research doesn’t matter. University research is more like a stage play. One federal grants officer told me that when they see in the first year that research supported by a three-year grant is going nowhere, they don’t take any action, because they are too busy getting ready for the next year’s awards. The literature on bad science continues to grow (see work by John Ionnidis and Ben Goldacre, among others)–fraud, errors in analysis, cherry-picking data, confirmation biases, withholding data–stuff that might be at best uncertain, provisional hints is turned into publishable claims of fact and potential for which no one is accountable (except, for a time, a few geologists in Italy). It may as well be treated as expensive fiction. Mildly entertaining, mostly for show. As Hesse put it (in The Glass Bead Game), an age of the feuilletons:

but what astonishes us far more is that authors of repute and of decent education should have helped to “service” this gigantic consumption of empty whimsies.

If the preambles to university patent policies are just fluff, and whatever happens in licensing doesn’t matter, then the research is treated as fluff, too. (People might say the research is oh so important, but they treat it as fluff.) It’s not that people should read this little article and think to care–it’s worse than that. They should already be riled about the situation, and most clearly aren’t. So, perhaps there is no situation to worry, and most university research is fluff. At least that’s the implication of present university patent practice. Even if university research isn’t fluff to start with, university patent practice, by failing to respect policy statements and regulatory guidance, turns research into irrelevance by creating Fragmented Ownership Institutionally Licensed (FOIL) technology, delaying or preventing use, disrupting collaborations, and subverting the university’s research role in society.

For my part, I think university research should matter, and shouldn’t be fluff. For that, how those around the research deal with it–examine it, challenge it, respect it–says a great deal about whether the research matters and will matter. Universities should offer ways to manage patents that respect the research pathway more than they are willing to be bought out. For that, university administrators must learn ways to manage patents that aren’t superficial copies of the least interesting corporate behaviors or the happy lives of patent trolls.


This entry was posted in IP, Litigation, Metrics, Policy. Bookmark the permalink.