From provider to predator: University of Texas patent policy, Part 1

I have previously looked at the University of Texas System intellectual property policy (“Texas wants you anyway“; “The most wonderful thing in the world“). I find myself drawn back to it again and again. In its 2012 version, the Texas policy is full of contradictions and inconsistencies, all written up in the finest legalese. Everything is bigger in Texas, including patent policy bombast. Broadly, Section 2 of the policy states a general claim of ownership, except for Section 5, which is about stuff not claimed, but Section 2 does not notice that Section 7, having to do with copyrights, also restricts the claim in Section 2, but is itself subject to Section 12, which allows for contracts to alter policy and so perhaps to override Section 7 and all the other Sections, too. “We compel you to follow policy, but we are not bound to follow the same policy, and if we change our minds, you have to change, too.” Follow that?

If a policy is not mutually enforceable, it’s not a contract, and so cannot be enforced in court. And if it’s not law, then it does not have the backing of the power of the state to prosecute it. It’s more like “guidelines.” If it is not a contract and not law, then it must operate as a delegation of authorities within an institutional framework, and not as an instrument that can extract private ownership of anything from anyone unless they choose to comply (or, in the abstracted world of administrative politics, are deceived or coerced into complying). A delegation of authority within an organization does not carry with it the power to compel–all it can do is confer standing to act on behalf of the organization (fictional but helpless and conscience-less thing that an organization is) and offer organizational benefits and sanctions. That is, “if you refuse to assign your personal property to us, we can fire you” but not “we can sue you to obtain ownership of your personal property” (which would appear to require a theory of contract or law somewhere to sustain).

The typical primary benefit in companies for following a policy guideline is continued employment. In a world of academic tenure, however, ignoring guidelines is not necessarily a sufficient cause for termination. Of course, university administrators packing lawyers with nearly unlimited legal budgets can attempt to impose their will on most anyone. A good university patent policy’s first aim is to place limits on what administrators–in positions of power but not with comparable status–can demand. But that does not make their policy any more “right” or “effective” even if they “win” (unless one buys into Thrasymachus’s pragmatic argument that justice is whatever those in power say it is, or believes in the mandate of heaven, which some think left the world along with the divine right of kingship, but no, apparently, it lives on in the dark recesses of the administrative brain).

A good university patent policy’s second aim is to make sure that the power for important decisions, such as institutional ownership claims on inventions, is placed with someone with comparable status within the university. Often, this person is the president or chancellor. While making such decisions may be challenging, and even seem to be a bother, that’s the proper level at which such power should be employed, and delegating the decision on ownership, or worse, making the decision an abstraction of policy to operate on its own, as mere bureaucrat-crafted words, creates the opportunity for low-status people–underlings, henchlings, assistant vice provostlings, bozos–to be petty, mean, stupid, and self-interested. It’s not that all assistant vice provostlings are these things–but the potential gets created, the policy makes the odds greater. Give a halfling a ring of power and in the face of administrative terrors, it is an easy move to put it on.

The Board of Regents defines “intellectual property” to be most anything, whether intellectual property or not, whether property or not. The scope of interest for such intellectual property is anything made by anyone employed by the System or who uses System “facilities or resources” or with System facilities or The Board then claims that it owns all such “intellectual property. For these people, the policy claims that the Board owns their “intellectual property” automatically. This, of course, is nonsense, or put another way, administrative fantasy, or another way, an unconstitutional taking of private property for public purposes. Anyone with half a wit knows that to own patent rights in an invention, one has have a written instrument signed by the inventors. Same for copyright, where it is other than work made for hire (in which case, one does not need a policy to state what federal law already establishes–but at a university one might want a policy to make clear that scholarly work is not within the scope of employment, that the university does not have a right as employer of faculty to control or direct or suppress the disclosure of their scholarly work).

Later, the policy asserts that everyone subject to the policy will “hereby assign” all that “intellectual property” to the Board, forgetting apparently that the policy has already asserted that the Board “automatically” owns their work. Here the policy reads as if it is pretending to be a contract of some sort rather than a statement of policy that authorizes administrators to act in some way (such as, to require a patent agreement with each employee or person using System facilities or resources). The “hereby” gets added in the 2012 version of the Rules and Regulations–presumably following the lousy advice of lawyers who decided that the Stanford v Roche decision hinged on Stanford’s policy use of “promise to assign” only, while Cetus used “hereby assign.”

[Sigh. In case you missed it, I will put this in tiny print. The problem was, of course, that the courts never got to that issue. All we have is a minority opinion musing on whether the words “hereby assign” should make so much difference compared to “promise to assign.” But that’s an irrelevant musing for the Stanford case, because the promise to assign to Stanford was not prior to the assignment to Cetus. First, there was the Cetus assignment. Then the post doc came back from nine months at Cetus, and almost immediately he invented at Stanford in just the area he had worked on while at Cetus–precisely the kind of bad behavior that a startup company with hot new technology guards against. Stanford filed a patent application and only much later got around to asking for an assignment of patent rights. The appeals court found that the assignment to Cetus was valid, and therefore the inventor had no rights to assign to Stanford when Stanford got around to requiring assignment.
A present assignment stuck in a policy statement would not have changed this outcome–Stanford’s policy at the time allowed inventors to own their inventions whenever possible, and Stanford had approved the nine-month leave to work at Cetus, and as the court ruled, Stanford knew what the conditions were for having a post-doc work at Cetus, so a present assignment would not have changed things. The invention was out of scope of Stanford’s policy claim, and the addition of federal funding later to the development of the invention would not have magically voided the assignment to Cetus, even with Bayh-Dole requirements. So Stanford was reduced to arguing that Bayh-Dole vested ownership of inventions with Stanford, voiding any contracts or assignments to the contrary. The Supreme Court rejected that argument, pointing out that Bayh-Dole only provides that if a university has got ownership of an invention made with federal support, then it can choose to keep that ownership, provided it behaves.
Further, and this is the kicker, putting “hereby assign” in a patent policy would not have made any difference. A patent policy is not a contract, and there’s no signature. It’s fantasy, or, an attempt to abuse administrative power. But apparently in Texas, no one writing policy has a clue and so throws in a “hereby assign” because, well, it sounds so good.]

Yet later, the policy excludes “intellectual property” created “outside the course and scope of employment” and “on his/her own time” and “without the support of the U. T. System or any U. T. System institution or use of U. T. System facilities or resources”–leaving open the question of what exactly such mysteriously abstract U. T. System “support” might be, if not use of facilities and resources.

Later still, the policy asserts that the Board’s “interest” in “intellectual property” is subject to a process of “determination,” started with a disclosure of the “intellectual property owned by the Board of Regents.” The Board, according to the policy, might choose not to “assert” its “ownership interest.” And if this is the case, then the Board will “release” the “intellectual property” to the “creator,” with various caveats. But of course, if the Board already owns “automatically” and by means of the magical “hereby assign” wording (as if “automatically” wasn’t already sufficiently magical), then the Board has to do more than “release” the “intellectual property” to “creators”–it has to assign it to the creators. The Board, here, is just plain confused. The policy already asserts the Board’s ownership interest. The Board can’t later in the policy reserve the right to pretend that the Board hasn’t decided whether to assert its ownership interest.

Within a few paragraphs, the Texas policy asserts automatic ownership, demands a present assignment, treats the policy as if it is a written instrument signed by the “creator,” (and so, a weird form of adhesion contract or a non-legislative enactment of law), and then makes it appear that the Board doesn’t actually own, but only has a policy claim to ownership, under which it may decide to “assert” or to “release” its claim to own. The policy takes, then, three different approaches to ownership–(i) automatic, (ii) by way of required assignment, or (iii) after review for interest. Worse, it appears that those drafting the policy don’t recognize the differences in these approaches, or the contradictions, or the–for lack of a better word–stupidity of putting out such a strange statement.

There’s enough here to argue that the policy is so confused, so conflicted and contradictory, that no reasonable person could possibly know what the Board intends, and therefore the policy here is simply malformed garble and binds no-one to anything regarding ownership. Not that one can say that to a Board who has loaded attorneys pre-paid and ready to fire.

The 1946 Texas Patent Policy

Let’s look at some history to see if there’s a story for how this situation comes about. A good start is with the 1946 policy, which was in effect largely in the same form until the 1970s. This policy is remarkably simple. Here’s the outline:

Except when a contract specifies otherwise, employees own their patentable inventions, but they owe the university 10% of net income over $1,000 [over $10,000 in 2016 dollars] and 20% when net income is over $5,000. Organizations supporting research at the university should expect a FRAND (fair, reasonable, and non-discriminatory non-exclusive) license.

Here’s the text in full, reprinted from Archie Palmer’s compendium:


In 1960, essentially the same patent policy shows up in the “Personnel” section of The Rules and Regulations of the Board of Regents for the Governance of the University of Texas:


It’s worth seeing the table of contents to see just where patent issues fit into the Board of Regent’s thinking–until recently, the Board considered patents a part of “General Personnel” matters. Then, for some reason, “Intellectual Property” was given its own section at the end–as if the whole idea of ownership of what people do is no longer about how the university provides an environment for personnel, but in the abstract makes ownership demands on whomever comes within its sphere of power, regardless of who they might be.

The 1946 patent policy makes good sense. Inventors own their inventions, even if they are university employees, unless there’s a contract that provides otherwise. That is, the policy restates common law. Contracts that provide otherwise could be ones with an external sponsor of research, or could involve a special deal with “general personnel”–outside of an appointment or employment agreement–under which the university commissions specific work for its own benefit or provides substantial resources with an expectation of reimbursement or a return on investment if the result proves valuable.

If an employee-inventor wants to file a patent application, that’s up to the employee-inventor, at the employee-inventor’s expense. If the employee-inventor wants to give away licenses, that’s fine, too. The university claims a share of licensing income only when the net income from licensing exceeds a threshold (roughly, the cost of applying for a patent using a patent attorney). Not only is the policy simple to draft but also it is free of many administrative complications: there is no need for the university to own anything, stuff does not have to be disclosed so university administrators can sniff at it, and inventors can continue to use university facilities to develop their work. The condition for working on stuff is not who owns it but whether the work is consistent with the mission of the university–are students getting trained? is something of value for Texas citizens being developed? is good science and engineering being published? Only stuff that an inventor chooses to patent and exploit for profit comes within scope of the policy, and there we find only two conditions: 1) if there’s a profit, then the university gets a share; and 2) if there are sponsors, they should expect a non-exclusive license.

The 1946 patent policy is simple, clear, and smart. One can’t ask for much more. But the Board, apparently, was not satisfied with simple, clear, and smart, and so has sought to improve on that with convoluted, garbled, and grasping. Only a bozonet could love the changes. And only a determined bozonet could draft such crap, sell it to itself as necessary and advantageous, approve it, and stick it to the people who work at the universities in the Texas System of Higher Education.

The 1977 Texas Patent Policy

In 1977, the Board adopted a new patent policy. It adds a preamble that disclaims inventing as a “primary objective of the System” (leaving aside why the System–abstraction (“system”) of abstraction (“university”) of abstraction (“college”) of abstraction (“faculty”)–should have objectives at all, rather than serve as the repository for governance of competing interests). The effect of the disclaimer is to reduce the apparent importance of patents. This denigration will become important in later policy revisions. The policy states two objectives:

  1. encourage the development of inventions
  2. permit timely disclosure of inventions

Each of these objectives includes a qualification, worth noting. Administrators have a habit of dropping qualifications as if they were merely illustrative rationales and not restrictions on a general case. Thus, “we own everything made in research work assigned by the university”–the claim is limited to the stated condition–transmogrifies into “we own everything, including but not necessarily limited to stuff made in research work assigned by the university”–where now the stated case is read not as a limitation but as an illustration, a member of a set, and becomes a figure, a metonymy, for a class of ownership created by the administrator (“everything” not “everything within the stated, limited case”). If you follow this pattern, then you begin to see what the administrative drafting habit is.

The first policy objective is qualified in this way:

encourage the development of such inventions for the best interest of the public, the inventor, and the sponsor whenever an invention occurs

So it’s not merely the development of inventions–meaning, one would think, moving from whatever was first realized to something practical for others to use through testing, prototype variations, applications, and productization. The objective of the System is that the development must somehow be “for the best interest” of three groups–“the public,” “the inventor,” and “the sponsor.” That sponsors show up indicates the sea change that took place after the second world war, as the National Science Foundation ramped up funding for university-hosted research and other federal agencies followed with their own funding programs. That the System itself does not show up in this list is consistent with the idea that the System (and its university components) are trustees in this process (cf. 2 CFR 215.37) and not active, self-interested participants.

The tension in this list conventionally is between the inventor and the public. This tension has been in university patent policy debates from the very outset. On the one hand, the constitutional basis for the federal authority to create patents is progress in the useful arts by creating a personal property right for individual inventors. The inventor, seeking his or her self-interest, uses the patent system as he or she sees fit, and Adam Smith’s invisible hand will ensure that there’s “progress”–meaning, “to spread” or “diffuse” the invention. The patent system does this spreading in two ways. First, a patent publishes the invention and teaches how to practice the invention so that anyone with “ordinary skill” can do it. Second, the patent gives the inventor the right to exclude others from practicing the invention (making, using, selling, importing) for a limited time. During this limited time, the inventor and those working with the inventor’s authorization get to do the making, using, and selling, and importing, and so gets a head start at benefiting from his or her inventiveness.

That’s all pretty good, with the usual intellectual fussing that comes about when academics get to talking. But universities had a different take. If an inventor happens to be working in a university, on a university faculty, doesn’t the inventor have obligations as well to the faculty, to the mission of the university, to those that oversee that mission? If so, then the university should have some say in how an invention is patented, and if patented, how the patent is used “for public purposes” rather than merely for the self-interest of the inventor. Self-interest, in the context of a university, easily becomes “greed” or “selfishness,” and such vices in turn suggest administrative controls to contain that greed and channel it into a thirst for righteousness. It is then another easy step for administrators (and even humanities faculty) to desire a policy that formalizes this position as a formal policy aimed to protect the university and the public from personal patent greed.

One university view of patenting, then–and the dominant view, Lehigh University and MIT notwithstanding–was that the patent was a creature of monopolies, and monopolies were antithetical to the collaborative, public-spirited mission of the university. Patents were cast as a threat to academic research, denied the public broad access to the benefits of university-hosted inventions and discoveries, and in some areas–particularly the practice of medicine–were viewed as fundamentally unethical. No one should be denied a treatment that might improve their health or save their life.

As a result of this tension, things developed along two paths. On one path was laid out by Cottrell, the Research Corporation, the Wisconsin Alumni Research Foundation, and the many other university-affiliated research foundations. Their premise is that it is the faculty inventor’s choice whether to use patenting, and if so, the patenting ought to be done outside the university, and in exchange for providing funding and expertise, the invention management agent would after costs and whatever share the inventor wanted dedicate the remainder to support more research. Whatever the particulars of the deployment of a patent, any net income would go to a good cause.

University administrators responded to this pathway with a policy apparatus based on financial interest. Committees would review circumstances under which an invention was made and further developed. If the university had commissioned the work, or had provided extra resources, then there might be a claim for equitable treatment–its funding should be reimbursed, it should get a share of income, it should be part owner as an outcome of the agreements that had been made and common law. Issues then revolved around how committees were constituted, what was meant by “significant use” of resources, and the precedence of contracting obligations over defaults (inventor ownership, for instance) in policy.

The second path went through university ownership of inventions. Here there were two competing policy claims. One was that a university should own inventions so it could preclude monopoly dealing by corporations. A university would own patents as a trustee, to ensure ready access to the inventions by all, and to prevent abuse of the patenting system by companies aiming to patent improvements and applications that then cut out everyone else. Patents, in this way of thinking, would be licensed non-exclusively and royalty-free to all behaving, qualified applicants. The university’s interest was the “progress” of the useful arts–that is, the spread and diffusion of inventions. There was little–at least in policy–of the idea that the public was most benefited when the university sought to license exclusively and to generate a financial “return” on the “public’s investment” in the university or the research.

The second competing policy claim was that a university as a publicly spirited organization would do a better job managing patents for income than inventors and their agents might, and money was a valuable thing for a publicly spirited organization to have, so it could have, well, more public spirit. Karl Compton, the president of MIT, reflects this approach, instituting a patent policy at MIT that claimed ownership of inventions in 1932.

As things developed, the idea of a financial return was borrowed from the research foundation pathway–that the university sought royalties from its patents to fund its own research, and it could do better job itself than if inventors were left on their own to decide what to do, if anything. A University of California committee in the early 1960s made this pitch–but without the claim that the university must own all inventions– back when federal funding for research was still not particularly significant (and still fraught with ethical questions, at least for some of the academics of the time). The idea was that if a university inventor took his invention to a national agent, such as the Research Corporation, then the share of royalties devoted to research would be spread all over the country and only some of the money would come back to the University of California. If the university acquired ownership of patents, then (so the reasoning appears to have gone), then a greater share of licensing income would come back to the university. Essentially, this was an argument based on provincial wants rather than on national opportunities, but to its credit, it was and still is a compelling argument among university administrators and faculty (not that I find it persuasive or even sound).

This emphasis on income shifted, then, the idea of “public benefit” from “progress” (spread, diffusion) to “the university as beneficiary to lighten the burden of public support for the university”–that is, university administrators and faculty shifted the focus from providing a service (progress) to self-service (if we make money for ourselves, then we won’t have to ask you for for as much money). Later, this argument transmogrified into something less desirable (since you appear to be making so much frickin’ money through licensing your patents, we should put our money into areas of society with fewer resources to work with). “Excellence” became a public rhetoric against the public funding of universities (other for federal grants, where “excellence” means “successfully demonstrated better potential capabilities than others”)–much to the consternation of university administrators who don’t appear to know where to take the rhetoric and so keep repeating, black knight style, that despite the cuts in public funding, their universities are still “excellent.”

So the reasoning appears to have gone: if a university exists to serve the public, then when it makes money from patent licensing, the public benefits because the university benefits. In some technical, narrow sense, this might even be true. A license might lead to a beneficial product, and more money for the university might be a good thing. But in the larger sense, these are two radically different ideas of benefit. In one, administrators and faculty work for a good beyond the institution. In the other, administrators and faculty work for their own good. The more money they bring in from licensing, the more they have to work with. The university, in this line of reasoning, gets better when it has more money that it makes for itself.

Patent licensing then represents a “new source of revenue” and if a university is to “run more like a business,” then administrators would do well to systematize and scale the operation to maximize royalty income. For that, why wait for an invention to be used when one can get substantial fees from licensing patents exclusively, especially into speculative schemes that aim to sell up or go public or litigate? Most such schemes depend for their success not on the practical application of an invention, but rather on the change in perceived value of the patents that cover the invention. One need not do when value also changes dramatically as one appears well positioned to do. One can make better licensing income faster–so the argument goes–by speculating on patent rights with exclusive, monopoly-producing transactions than in promoting the progress (spread, diffusion) of the useful arts.

Whereas in the first pathway, inventors worked out with their invention management agents how they would deploy an invention and its patent rights–that is, by negotiation outside the institution, outside any employment relationship–in the Compton scheme, the university takes ownership of inventions and then decides, administratively, what to do to exploit the patent right as an institutional asset rather than as a personal one. Those universities that adopted the Compton approach also followed MIT’s lead in contracting for invention management services with Research Corporation, and later with Battelle Memorial Institute, University Patents, and Competitive Technologies, and with their affiliated research foundations. In a number of cases, university policy directed ownership of inventions to an affiliated foundation, which then contracted with a national invention management agent. The common ground in these variations is that the university administration controls what to patent, who manages the patents, and who gets the money from any licensing of patents.

Returning to the second of the two objectives announced by the 1977 Texas policy–timely disclosure of inventions–the objective is qualified in an interesting way:

permit the timely disclosure of any patentable discoveries, whether by patent, publication or both.

That is, “disclosure” here does not mean “disclosure to the administration” but rather “disclosure to the public.” The policy aims to ensure that inventions are made visible to the public. “Submitting ideas” regarding patentable inventions to a patent committee is only one step in the overall process.

There is an additional sentence in the preamble that then sets the objectives into context:

The policy is further intended to protect the respective interests of all concerned by ensuring that the benefits of patents accrue to the public, to the inventor, to the System and to sponsors of specific research in varying degrees of protection, monetary return and recognition, as circumstances justify or require.

To decide among the competing interests–which now includes the System itself–across three areas of concern (in short, ownership, money, and fame), the policy expects to be guided by “circumstances.” That is, any claim ought to be justified (or required) by the conditions that present, the circumstances, and not for-ordained by policy. The 1977 policy is committed to an empirical approach to invention management. Review the conditions of invention and development to decide the equities due each party involved. This is the essence of governance–adjudicate competing interests–as distinct from management, which involves dictating behaviors for production efficiency.

The policy then states that it applies to everyone employed, using resources, or in doctoral programs or just completed a doctorate (“post-docs”). This requirement makes sense. It makes sense to require that everyone involved will accept administrative resolution of competing interests based on the circumstances of each case. Casting a broad scope restates a basic principle of governance: accept arbitration where there are competing interests.

Less evident is the location of the “public” and “System” side of the tension involving disposition of patent rights. An inventor is an individual and enjoys the prospect of patent rights by operation of the Constitution, common law, and federal law. The “public” is an abstraction. “The public” neither thinks (no brain) nor has a Constitutional right to inventions. Same for the “System.” Individuals acting in the name of the System, or taking on the mantle as a representative of “the public” must be making a case for something other than individual ownership of inventions. That’s what the 1977 patent policy has to work out.

The starting point for this effort to deal with the tension between the public and the individual is to translate it into a tension between the System and the individual:

It is the intent of this policy to permit staff members maximum freedom in respect to their findings, consistent with their obligations to the System.

Here, the drafting is of the form, “you inventors have everything but for what you have obligated to the System.” It’s remarkably like the Stanford University policy that came into dispute in the Stanford v Roche case–inventors own everything except what the university is required to own (as a result of, say, law or contract or court order). In such an approach, it is up to the university and its potential inventors to state clearly what it is inventors have promised the System. Anything not obligated, isn’t. Timid and uncertain administrators often take a different approach and aim to “shift the burden” by drafting something along the lines of “The System owns everything except what it expressly agrees to release.” This latter approach is the way of most American university patent policies these days–no longer called patent policies but rather “intellectual property” policies, with “intellectual property” defined to be both intellectual property and non-intellectual property (stuff that is not patents, copyrights, or trademarks; stuff that is chattels or material property; stuff that is not property and cannot be said to be owned; stuff–information, practices–that is said to be owned only because of promises by others not to use or disclose the stuff).

The 1977 Texas policy then distinguishes between making an invention public and property rights in an invention:

Any person affected by this policy who, as a result of his or her researches makes a discovery, other than on certain government or other sponsored research projects, where individual grant agreements provide otherwise, should retain the ultimate right to decide how it is to be made public–by publication, by patenting, or both.

Other than when a funding agreement dictates publication and ownership terms–and for faculty investigators, they choose those terms and are not assigned to do work on any given grant–the inventors decide whether something should be patented or not. This is important–to decide whether to patent an invention is to decide whether the invention will be owned. If an inventor decides to publish and not patent, there’s no further discussion about ownership, since there’s nothing to own (other than by redefining invention in strange ways–not beyond the imagination of administrators, if still beyond their skills).

One can see how this approach dramatically reduces the volume of inventions that need to be considered by the institution. If the inventor decides on ownership, only then is there a further review for institutional interest:

Property rights in inventions will be based on the degree of System support, as hereinafter specified.

This is utterly unlike the present approach, dominant in most American universities, in which ownership of inventions (broadly defined–inventions and non-inventions, etc) is claimed outright, and patenting is a decision made by administrators. That is, in present policies (including the current policy at the University of Texas), ownership is a matter of administrative fiat–it is not even a matter of judgment. The only matter for judgment is whether the System will keep the ownership it has claimed.

Because a patent is a publication–that is one of its essential properties in promoting the “progress” (spread, diffusion) of the useful arts–claiming institutional ownership of an invention and claiming the right to file a patent application means that the institution may compel a faculty member to publish and dictate where that publication will take place. That’s quite an assertion of administrative power (or, in the case of public universities, state power). Certainly, it is a challenge to–if not a blatant disregard for–the policy commitment to academic freedom found at nearly all universities, that the university (and the state) should not and will not mess with when and how a member of the faculty conducts research or publishes research. Except, when something is an invention, and invention is anything administrators say it is, and ownership is whatever administrators say they own. In which case academic freedom is laughable, like wearing a top hat and tails to work in the university’s patent licensing office.

Even if one decides on patenting–that is ownership–and there’s an institutional interest arising from the circumstances of institutional support or contracts–there is a further issue of how patent rights should be deployed. One tradition considers non-exclusive licensing, to make technology available; another tradition favors exclusive licensing to enhance financial returns by placing technology into a stream of speculative investment. The 1977 Texas policy opts for the former tradition as the default:

It is a basic policy of the System that patents be developed primarily to serve the public interest. This objective usually will require patent development by non-exclusive licensing but the public interest may best be promoted by the granting of a limited exclusive license or even an exclusive license for the period of the patent.

This is in line with the university’s role as trustee rather than self-interested party: non-exclusive first, limited exclusive next, and only in some cases, an exclusive license for the life of the patent. Such a priority stack is in line with the expectations of the NIH’s old Institutional Patent Agreements, which expected non-exclusive licensing unless one had a meaningful reasons otherwise, and limited exclusive licenses to three years from the date of first sale or eight years from the date of license, whichever was shorter. At one point Stanford incorporated these limits into its formal statement of policy on licensing.

After discussing the purpose and duties of an Institutional Patent Committee to make determinations of System interest in inventions that inventors have chosen to patent, the policy distinguishes inventions based on circumstances under which “the patentable idea has developed” (I paraphrase and add header tags for readability):

  1. Independent: on one’s own time, with no direct System support or use of System facilities. No System ownership or financial interest. An inventor can offer an invention to the System, and it’s up to the patent committee whether to take it.
  2. State support: from research performed on System time, with support by State funds, or using System facilities. Inventor’s choice whether to seek a patent. If so, then the patent committee recommends the System’s interest, with approvals from the System Patent Officer and System President. If the System asserts an interest, then it will share royalties with the inventor on a sliding scale of more at small amounts of income and less at larger amounts. Otherwise, the inventor is for the most part back to Independent status.
  3. Grant support: supported by a grant or contract with a Federal agency, or with a nonprofit foundation, or by a private gift to the System. The value of getting the grant outweighs any exception to the patent policy, so the terms of the grant control.
  4. Industry: resulting from research supported by commercial concerns or industry. Industry contracts must be negotiated based on the circumstances, with the most important consideration being “the interests of the State and its citizens who provide basic fiscal support.” Any contract here that alters the patent policy requires advance approval from the patent committee, the chief administrative officer, the System President, and the Board.

Yeah, don’t mess with changes to policy in industry contracts unless it’s a pretty darned important contract.

Income after costs to manage the “System Patent Office” is to be used “for research purposes at the component institutions where the invention originated.” This last provision reflects two additional tensions set up by the policy itself. The first is the extent to which the System Patent Office is concerned largely with recovering its own expenses (and growing those expenses to gobble up whatever income there might be) and allocating income for research (in which case, the System Patent Office would be selective in what it takes on and limit its costs to maximize the net available for research). There is nothing in most university patent policies that addresses the size of the patent and licensing operation. They get to eat what they harvest, and they are limited in their eating only by the self-discipline of their administrative leaders.

The other tension is between public benefit writ large–university as trustee, getting stuff out in a form that prevents monopoly control–and public benefit write tiny in the form of more money for research at the university. This is a challenging tension, especially for administrators whose status depends on how much money they bring in. “Socially responsible licensing” only works in places where administrators (and faculty who think they have an inside track to get access to research dollars) respect the public benefit of making stuff available, even without a demand for royalties, or only nominal amounts–FRAND licensing, for instance, to create standards and ensure everyone who is capable has access.

If patent policies and practices are any indication, most American universities are stuffed with administrators who don’t give a rat’s ass about public benefit writ large–that’s idealism, misplaced altruism, and a failure to grasp that survival depends on getting one’s due rather than creating opportunities for others. If all universities pursue their own self-interest, then, why, an invisible hand will guide all of them to the most equitable society possible–this, the myth of the provincial administrator. What actually happens is that discoveries get tied up in institutional ownership claims, institutions can’t license without stuffing the deal with poisonous terms, and end up licensing mostly to snakes, not all of them poisonous. 99.5% of what universities claim under this approach never sees commercial product. Tell all the stories one wants about the “successes”–the reality is that 199 out of 200 discoveries that a university claims to own sit for two decades or more.

In Part 2, we will look in some detail at how Texas embraced the dark side of policy, starting with its 1988 policy.

This entry was posted in Bozonet, Policy and tagged , , , , . Bookmark the permalink.