It may be useful to map out four approaches to university IP management:
- Personal
- Entrepreneurial
- Institutional
- Open
The discussion below does not advocate for one approach over the others, though compulsory institutional IP management seems not to have worked all that well for American universities. That does not mean there is not a role for institutional management, but as Jane Jacobs has one of her characters argue, “Whatever the justifications for monopolies to start with, they all end up dumbed-down, elderly, and hard to get rid of…. Worse, think about the struggles of conquered people to throw off monopolies on salt, matches, and everything else that their conquerors wanted to buy or sell” (The Nature of Economies, 36). In this case, university administrators have established a monopoly over faculty research results, in essence dominating faculty personal interests–and faculty choices about their professional and public engagement involving their work.
The primary weapon the university administrators used was the Bayh-Dole Act. Administrators–starting with administrators responsible for patents–made it appear that federal law required institutional ownership and institutional efforts to commercialize. The law did no such thing–it was purely a university administrative decision to demand institutional ownership and dedicate that ownership to monopoly licensing in pursuit of profit. Faculty accepted the change because (i) they were told the law required it and (ii) because they were led to believe that institutional licensing would make money for more research and (iii) that institutional ownership of inventions and money-making (or even money-seeking) was somehow better than either inventor ownership and money-making or inventors letting anyone use or make or sell and no-one at all profiting from the threat of exclusion built into every patent.
Administrators also used a secondary strategy to get faculty approval. They held out the prospect of substantial personal wealth by means of the operation of an administrative patenting and licensing process that claims to relieve inventors of most of the work involved but offering them a share of licensing income. This prospect for personal wealth, too, has proven to be unfounded in the general case. Only one or two inventions a decade at even the luckiest universities end up making life-changing millions for inventors. Inventors end up being involved more than they would expect, the process itself often breaks down after much time and expense, and good deals appear more at random than in response to an orchestrated effort to identify companies (or to start one).
And even after all that, this secondary approach meant that administrators worked to turn faculty motivation to do research from scholarship, curiosity, independent messing around, and working on pressing community issues to wealth-seeking. The administrators had the moxie to top off this secondary argument with the idea that faculty were too lazy or indifferent to developing their discoveries in practice and the incentive of substantial personal gain–properly managed through the university so it would washed ethically–would be just the thing to do what? Yes, to divert faculty from doing more research as scholarship to money-seeking development of commercial products protected by patent positions. This was called a “culture change” and the rise of “entrepreneurial faculty” or the “entrepreneurial university.”
Places like the University of Washington raised money to recruit in such “entrepreneurial” faculty because apparently the ones already appointed just weren’t good enough. Implicit in this argument was that inventors had to do more than assist in the preparation of patent applications–they had to become the principals in commercialization efforts. And in that effort, some few have thrived–though not generally by routing all such opportunities through the university patenting office. This split, then, gives rise to a yet broader claim of conflict of interest if faculty so much as choose to place an opportunity for technology development or exploitation or commercialization beyond the reach of university management. If a faculty member hands an idea off to a company through consulting rather than “developing” it with grants at the university, that’s conflict of interest. If a faculty member starts a company to pursue an idea that has no connection with the faculty member’s university research, that’s a conflict of interest. If a faculty member releases new technology for open access with the potential to consult with companies adopting that new technology, that’s a conflict of interest.
(I dealt with a major research university that threatened to sue one of its software programmers for releasing code open source rather than submitting that code to a university invention-patent-and-license scheme on just this claim–open source means you, programmer, could (in theory) make money consulting on how to use this code without causing to that income come first to the university as a licensing deal and therefore you have denied the university its rightful share of money derived from your work in favor of your own (in theory) personal financial interest–and that’s egregious personal conflict of interest for which you will be lucky if you escape with your job or your reputation or your career).
As it becomes apparent that universities adopting a compulsory, comprehensive institutional intellectual property program are doing worse on commercialization metrics than was the federal government prior to Bayh-Dole, and at greater damage to their constituencies and generally for little benefit, folks will eventually turn to alternatives, and those that do so first will have substantial advantages over the others–they will save money, build goodwill, develop substantial networks of resources, allow specialized agents to become involved, and reduce their own liability, complexity, and conflicts of interest.
Sounds good, doesn’t it?
Personal
When someone has an idea, writes a song, invents, aside from the artifacts of activity, the spark of insight is in the mind. The someone having the idea possesses that idea. There’s not much more to it. So long as the person does not express the idea–the song, the invention–to others, the possession is absolute. One may as well say that the person owns the idea–but only in an odd private sense of ownership. Rather, until the person discloses the idea, there’s not much anyone else can do to extract it, or even know it’s there.
United States IP law defaults to the individual. Inventions are owned by inventors, not employers. Works of authorship are owned by authors, not employers. An employer or commissioning party can be an author, but only if there’s a written agreement or the work is made within the scope of employment.
Individuals can then decide what to do with their work. They can disclose and publish, they can assign for money, they can hold everything secret, they can renounce any claim to ownership of intellectual property. If individuals manage their own work, they decide what matters, and how intellectual property figures, or doesn’t, in their plans. No one forces authors and inventors to seek money, to pander to venture capitalists, to threaten industry with lawsuits if companies practice what the inventor or author knows, or to participate in “technology transfer.” No one claims to know more than the inventors and authors about what might make for personal advantage, or be in the public interest, or benefit industry, or a university that happened to have hosted research.
Once upon a time, fifty years ago, university faculty, in inventing and authoring, were treated as individuals. If they chose the path of commercial engagement, by filing patents, university policies might require them to report their activity, and might limit the involvement of university resources in commercial ventures. If they had used university resources, and then became successful in their commercial ventures, the university might ask for reimbursement, or might offer to help in exchange for a share of the upside. But if the inventors don’t pursue commercial opportunities, the university does not have to be involved.
Public release does not mean that “commercialization” does not take place. By publishing inventions, by sharing software and data, by not taking positions that would exclude others, or would appear to require payment, an inventor or author may well promote use of new work in industry, and encourage development of commercial products.
It is simply untrue that no one will use an invention unless there is a patent on it, and no one will develop an invention unless there is a monopoly on it. It is undoubtedly true that there are some people who demand a patent and a monopoly before they will put their own resources into developing an invention. The number of such people is small. I doubt, however, that even these people refuse to use open software, such as, say, WordPress or gmail. In the US, surgical techniques are not patentable, though they may be at times inventive. Is it the case that a surgeon will refuse to use a new technique because there’s no patent on it? Is it the case that companies will not make new tools and instruments to support the technique, even though there is no monopoly position available on it?
The university administrator fear with the personal approach is two-fold: first, that individuals, if they own their own intellectual property, will simply ignore intellectual property opportunities and give away their “potentially” valuable work. Second, that individuals will not give away their work, but rather try to exploit it and in doing so, show themselves to be greedy, distracted, and inept, leading to all sorts of terrible liability for the university. University administrators have a third, darker, secret fear, which they do not tell to just anyone: that individuals will make money without the involvement of the university, and without sharing back with the university, and that would make the university’s own technology licensing program look greedy, distracted, and inept.
The greatest attack on the research commons, on open collaboration, on the steady flow of practicable new ideas into the public domain for industry, investors, entrepreneurs, and the general public alike, has been the efforts by university administrators to prevent personal management of the work of scholarship and the associated intellectual property. The reasons for the attack are rooted in the desire for a professional class to attach themselves to education and so find a source of stable funding for their work. This they have accomplished with marvelous success, as might, say, a tick on the back of a rhino. The AUTM licensing survey tracks, not the development of innovation but the number of ticks on the rhino–that is, the expansion of the professional activity as it displaces the personal in university intellectual property management.
Entrepreneurial
A second approach we may call the “entrepreneurial” approach. An entrepreneur seeks start a company to develop a product, or engage existing companies to do so. For that, the entrepreneur often needs money and talent and resources–a garage, say, or a laboratory, or a test site. The inventors recognized in American lore have generally been also entrepreneurs–Whitney, Bell, Edison, Ford, Westinghouse, Sarnoff. But as Andrew Hargadon makes clear in How Breakthroughs Happen, “entrepreneurs and inventors are no smarter, no more courageous, tenacious, or rebellious than the rest of use–they are simply better connected” (11). Hargadon builds an argument for the entrepreneur’s network. Again, not the number of “connections” on LinkedIn, but rather a network that provides the resources needed to make a new initiative successful.
Scott Shane (The Illusions of Entrepreneurship) has done some work to characterize the typical entrepreneur–one core finding is that the entrepreneur “doesn’t start a business because of a desire to make money, for the thrill of starting businesses, to support their families, or to become well known; the typical entrepreneur starts a business because he doesn’t like working for someone else” (62).
Considering just these two properties–good connections and the desire to avoid working for someone else–one might find in a faculty member’s decision to pursue a commercial opportunity something very different from what is put forward by the typical university technology transfer office, which assumes that faculty are poorly connected and afraid of stepping out into the big, wide world.
The entrepreneurial approach is about using one’s connections and developing new ones if the existing ones won’t do. The end point is not to be working for someone else, as if on a job hunt, but rather to lead the development of a new company, with or without something innovative, just to be free of an overlord employer. For the entrepreneur to prosper, he or she needs a network that will be productive at need. If a university wanted more faculty entrepreneurs, it might start by creating an environment in which faculty meet a wider range of people, and have time and opportunity to form working relationships with some of these people, so that later, when something interesting does arise, a faculty member recognizes that the connections are there to start an entrepreneurial opportunity. This is what the UCSD Connect program has done, for instance. Just give diverse sets of people the opportunity to get together, like a big, continuous cocktail party or rock concert, except will less drugs and alcohol.
If the entrepreneurial motive is also about not working for The Man, then universities also would be wise to reconsider demanding that faculty accept the claim that they are employees working at the beck and call of administrators. Turning faculty into indentured servants does not lead them to be more inventive, or more entrepreneurial within their scope of employment–just the opposite: there is more incentive to leave the university, or work through the “backdoor,” outside of the claims of policy. Making entrepreneurial activity part of the scope of employment, as a recent university advisory committee has recommended, would appear to run exactly opposite of the claimed objective, of creating a more entrepreneurial culture.
Institutional
The institutional approach to intellectual property management started out as a voluntary activity using external invention management agents. A university would refer inventors to an external agent, which would then represent both the inventor’s and university’s interests in seeking patents and licensees. These days, the institutional approach has taken invention management in-house, and now often starts with a requirement that all intellectual property that the institution wants will be owned and managed by administrators hired for the purpose. The general idea, however, is that an institution can do a better job promoting an invention than can an inventor. That’s an idea worth considering, but for our purposes here it is enough to point it out.
In a company, the intellectual property management folks tend to be in the legal department, and their business is to identify, secure, and protect intellectual property. How intellectual property is deployed–licensed, abandoned, asserted, swapped, sold, contributed to a standard–is the job of management. In companies, there’s a good tool to use to reason about how to manage intellectual property, and that is the business that the company is in. It’s not just the competitive advantage to be had from locking out everyone else, but also the need to gain access to technology of others, or to develop standards, or to be sure the company’s stuff plays nice with other companies’ stuff. A university, as an institution, generally lacks this tool. For any given invention, a university cannot reason about how to situate itself in a competitive landscape where it also produces goods and services. It either lands as a disrupter of the industry (by trolling it or by picking a single company within the industry to do the trolling or to exclude all others from a line of improvement or development) or must think about the public interest with regard to the health of the industry and its interaction with the public–and that’s the more difficult thing to consider. No wonder, then, that most university administrators opt for making a sophisticated go at how they disrupt an industry for profit.
In Europe, inventions made in employment are separated into “service” inventions and “free” inventions. Service inventions are those made within the scope of employment, and an employer has a reasonable claim to them. Free inventions lie outside that claim, and are the employees’ private property. In the United States, inventions are personal property of the inventors until assigned in writing. Employment is not sufficient to make the employer the owner of an invention (unlike, say, in the UK). There must be some other agreement, a patent agreement, that addresses the obligation to assign. Typically, in a company, the patent agreement is made at the time of employment, alongside the employment agreement. Absent a patent agreement, a company might expect a shop right in an invention made within the scope of employment or using company equipment–the right to make and use the invention under an implied license.
In a university, by contrast, institutional control came about slowly at first, and then more rapidly after the passage of the Bayh-Dole Act. In university practice, a single office, the Technology Licensing Office or the Technology Transfer Office (or with more exotic names) decides what to own, what to seek patent protection for, who to license to, and on what terms.
Institutional management of intellectual property carries attractions. An institution tends to have more resources than individuals and therefore can act directly to file patent applications. An institution can keep people on staff who know how to work with patent attorneys or negotiate licenses, and that can provide guidance for those who happen to invent. An institution has visibility and reputation, and that can attract collaborators who recognize the name of the institution but do not usually know the inventors (or authors). An institution can create relationships that then can be used by any number of faculty (and staff, and students) in their efforts to develop inventions (and other works). An institution, because it often has significant financial resources, can protect its intellectual property and sue infringers that refuse to take a proper license, or forego practice of an invention that has been exclusively licensed to a competitor. An institution can bring order and process to creative activity, so faculty are not out repeatedly pitching their latest ideas to the same external sources of venture funding and to the same big businesses that have shown an interest in the university. An institution can mediate between faculty and commercial concerns, so faculty do not mix academics and money interests or risk being snookered by cleverly malicious investors and business people.
These attractions are just the ones that a university administrator might long for. They are not, however, quite the thing that faculty, entrepreneurial or otherwise, might desire. Institutional control of inventions is not strongly aligned with innovation. One engineer told me that at Xerox PARC, for instance, PARC classified inventions into “strategic” and “non-strategic.” Strategic inventions were owned by Xerox, who then assigned managers to start an operation to commercialize. These attempts usually failed. Non-strategic inventions were “free” and individuals could deploy them as they pleased. Engineers hoped their inventions would be judged “non-strategic.”
It’s clear that in the hands of a company, an invention that’s “strategic” can be put to good use–it’s just that it takes expertise, diligence, and good fortune to do so. Adoption by a major customer can be the tipping point, such as when Wal-Mart adopted RFID. Even here, the monopoly position provided by intellectual property does not necessarily determine the outcome. Many companies forego monopoly in order to gain access to complementary technology owned by others. They would rather cross-license, or develop a standard, than fight tooth and nail over bits and pieces of what only works if it is a platform. In such cases, the distinguishing competitive features will be add-ons, build quality, availability, versions, customer service, price, and especially, brand.
Universities, however, lack most of a company’s resources and mandate to make and sell product, at least technology products in a commercial marketplace. Universities do make technology products in the form of research tools–software, custom instruments, data processing protocols, assays, and the like. These tools can be made available to others, or held tightly for competitive advantage in winning research grants. Occasionally, the market in research tools can be lucrative, as in the case of the Cohen-Boyer method of splicing genes. Thus, universities rely on licensing rights as their pathway to income. And that’s where the problems start. Those attractive features of institutional control turn out to have a dark side.
Universities have found it is easier to accumulate IP than it is to move it into use. Universities have found that the processes they put in place to identify inventions and related assets, log them into a database, secure IP protection, and try to find licensees fail in the purpose and serve only the middling invention, the most common invention by volume. Universities are saddled by liabilities–as “deep pockets” they have to be wary; as visible institutions, they have a reputation to preserve; to be consistent, they have to do the next deal on the same basis as the previous deal, or they will have to rewrite policy or explain why they are playing favorites; when they make a mistake, they impose restrictions that affect all later activity, to avoid making that mistake again, but at the cost of cutting out opportunities that would otherwise arise but for the restrictions. An institution requires layers of review; contracts with expensive legal counsel; does everything according to policy requirements; sorts activities by priorities, so what may be the most important thing for an inventor is only, say, 19th on a list of things to do today for an overworked university invention management administrator who will be lucky to get to item 12.
The general result is that university ownership of IP as a matter of policy stifles innovation, suppresses collaborative networks, introduces substantial additional overhead, and over time builds up a legacy of policy and practices that are increasingly constrained by policy and rule out of fear of mistakes, a desire to be proper, and a demand to be consistent.
Institutional control does have its place, despite the dark side. An institution may act as a trustee for IP rather than as an agent seeking to a self-interested return. Thus, if a number of inventors and companies desire a neutral organization to hold a collection of rights, a university might be just the thing. But then the university is operating to support work directed by others, rather than trying to “pick a winner” and make money. If the service of holding IP on behalf of others is valuable, then a university can expect to receive support for its efforts–through membership programs, sponsored research, subscriptions, and service contracts–as well as, for public universities, political advocacy for state support for its programs.
Given there are many, many more companies than there are venture capitalists, the moment a university administration turns from service to industry to service to speculative investors, it has reduced its political advocacy pool from 10,000 to 10. When a university licenses a patent exclusively, it alienates every other company in the same field, including potential sponsors of research or those who would hire university graduates or assist in internship programs.
Thus, institutional management of intellectual property is best done when done voluntarily and non-exclusively, for the benefit of industries and to provide access to the public, rather than in the service of, and acting to subsidize the interests of, speculative investors. If the only way a research finding will become a lucrative product is through a monopoly managed by a speculative investor, a smart university will find a different channel for that deal to move through: the deal should not happen from the university, because the university cannot afford the downside. That’s were individual initiative and external invention management agents unconnected to the university play a valuable role. The background rights involved are only those of the inventors involved; the licensing decisions do not involve university administrators; the monopoly that’s established is done on terms and conditions that the university does not control and is not responsible for enforcing, and the deal as a whole is not subject to public disclosure law as it does not involve the university at any point.
As it is, most US research universities have implemented comprehensive, compulsory institutional ownership programs. The backing policies claim more than patentable inventions–including knowledge, software, data–and insist on “commercialization,” which generally starts with a default effort to find an exclusive licensee. I know of no US university that has a default licensing position that is non-exclusive, and I have read a great many university IP policies. Institutional licensing operations are not able to close many exclusive licenses per year, so the hit to their universities’ reputations and goodwill degrades slowly over time. Their IP portfolios, by contrast, continue to grow. In effect, unlicensed patents keep research inventions out of use. No one can afford to use these inventions until the university has exhausted its efforts to create a monopoly commercial position–before then, a university licensing office will not consider a non-exclusive, royalty-free (even if for a one-time fee) arrangement. AUTM advertises the increase in patents as a sign of vitality and success; in reality, these patents are wasting assets, and ones that mark out where university administrators have stopped the flow of significant findings into the broader research and development communities.
Open
A fourth approach is the open approach. Open may be as simple as disinterest and indifference, which takes little effort and despite objections from those committed to trying to make money from paying patent licenses, is remarkably effective as a way of getting technology into use. Without any ownership claim, a research finding may be used by others with only the overhead of learning of it. No delays, no negotiation, no payment, no publicity, no threat of audit, no litigation for infringement. These are all highly attractive attributes of indifference.
Open, however, is generally aligned with an intellectual property position that is then used to limit the effect of other intellectual property positions. An important development in open was “open source licensing” of software. Not only is the source code made available, but recipients are allowed to make modifications, make copies, and redistribute both the original and modified versions of the software, often with little or no restrictions. Similar efforts have been undertaken for hardware (Arduino, for instance), for synthetic biology, and for environmental patents.
Of course, open has a much longer history. Richard White, in Railroaded, tells of locomotive engineers customizing their engines and trading technology. The US government had to intervene early in the history of aerospace and broker (0r enforce, as you wish) a stand-down of patent claims on the various bits and pieces of airplane technology, to prevent the US from falling behind the developments in the rest of the world. If only the US had done the same for nanotechnology in the past decade!
Industry as well, in a kind of lex mercatoria, develops standards and cross-licensing arrangements that basically say–no matter who has developed these bits and pieces, it is to our advantage as merchants to compete on something other than intellectual property positions. Standards, and their looser cousins platforms and libraries, enable broad access to technology, allow for interoperability across competing products, allow complex products to be manufactured with lower overhead and uncertainty, and provide opportunities for experimentation and versioning that serve minority markets that otherwise would not receive attention (for minority markets see, for instance, Matt Ridley’s account in The Rational Optimist, 107).
Open, then, has a range of flavors. Open can be free or controlled, can be a basis for payment or not, can impose restrictions on others or be liberal (or indifferent). The challenge in making open a management approach is knowing what flavor to choose. For university administrators, open is regarded generally as a threat, a form of defiance, a way to circumvent the technology transfer program that seeks first to find a way to profit from an exclusive patent deal. For other administrators, open is acceptable if it is controlled in some way, that administrators decide on the flavor, not those directly involved, preserving the semblance of administrative control. Finally, there is the problem in open that some people wish to impose open on the rest, a kind of technology communism, which takes the form of a non-institutional coercion that attempts to constrain personal initiative just as fully as institutional management. Here, a university dedicated to academic freedom has an interesting role to play: to limit, as well, coercion to participate in some flavor of open over others, as if each were a competing religion or political ideology.
Open approaches have their role, but some things do not thrive in an arena with no owner. Some initiatives peter out without income, operated by volunteers who discover after a few years that they have lives, or could have, or ought to try at least. Many of the open source projects on-line rapidly go moribund, but perhaps that is for the best–try something, see if it catches on, move on if it doesn’t, leave what’s been done for whoever finds it next, like restocking a hunter’s cabin in a wilderness and leaving the door unlocked to help out whoever comes next. Other stuff needs an owner, someone vested with an interest in looking out for the asset. That’s different than an institutional owner warehousing IP, hoping to ensure that among the boxes stored in the dark, there will be one or two a decade containing treasure. The owners of IP need to have a personal interest, a primary focus, a sense of diligence not based in a bureaucratic process but in attention to opportunity and detail. That’s the sort of thing that starts with an individual who realizes that he or she has connections, and decides this time not to be indifferent. The appropriate movement of the IP, then, is to those among the connections who are willing and able to respond in a like manner to the proposed course of action.
If an institution intrudes to demand that the individual drop his or her connections in favor of the institution’s connections, that often disrupts the activity and drains its energy. The institution will have to supply an administrator with comparable energy and knowledge of the invention to move the activity forward. That’s unlikely, not because technology administrators cannot be sharp and committed, but rather because they cannot be as sharp and committed to each and every opportunity that shows up and is assigned to their portfolio. As Hayek might have phrased it, they cannot possibly anticipate and act in time for the benefit of each individual over whose IP they have asserted ownership and control. In the end, they ask individuals to support their institutional program, to make it successful, rather than acknowledge their inherent limitations. If asking fails, the next step, or temptation, is to demand support, and when that happens we get a compulsory program and start the slow slide from vitality to monopoly dullness.