G. Kenneth Smith’s 1997 article “Faculty And Graduate Student Generated Inventions: Is University Ownership A Legal Certainty?” on university patent policies is worth the read, even nearly 20 years after it was published. The Bayh-Dole part might be redone, now that we have Stanford v Roche, but much of the rest has stood the test of time remarkably well. Smith has gone on to a career as a top IP officer in some leading biotech companies.
The article covers key aspects of the situation at universities–common law, policy statements, and statutes. Smith ends by suggesting that common law approaches appear to be the best way to deal with ownership, but noting that when present, pre-employment agreements, laws, and policy documents “are controlling”–but with the various practice implementation limitations that Smith points out in his article.
Some points of interest in Smith’s article are worth highlighting, if for no other reason than to create curiosity that leads one to read the entire piece.
1. University patent policies are likely adhesion contracts. That does not make them unenforceable, however. But unconscionable terms in them could change that.
–Smith does not consider whether a university patent policy can be treated as a contract entirely separate from other policies at the university. If one policy is a condition of employment, then what about others, such as academic freedom?
–The interpretation of an adhesion contract runs against the drafting party. This makes a difference in how a contract (or patent policy) is interpreted. Universities try to make it also a requirement of policy that only the university has final authority to interpret its policy. This may be an unconscionable term–it appears to demand an agreement to agree, invites sloppy drafting (since slop means administrators can make a policy later mean whatever they want, and so many university patent policies are slop from start to end), and conflates an administrative process with a contractual one. Often, the starting point for determining what a sloppily written patent policy requires is an act of interpretation. In an adhesion contract patent policy, a university does not have the sole right to claim what is required. “True intent, if unexpressed, is irrelevant”–to paraphrase the Shaw court.
2. Pay may be adequate consideration for a pre-employment agreement to assign patent rights, but subsequent changes to a policy may lack consideration and thus not be binding. Where is the exchange of promises for the new requirements?
–The Shaw court ruled that the employee has an equal right to such a contract, and distinguished changing the policy from altering the terms of the contract the employee has been required to agree to.
–The University of California argued (and lost, twice) that its policies were a kind of administrative law, not “private” contracts. Courts disagreed, but UC persists in making it appear that faculty are acknowledging an edict rather than entering into a set of mutual promises. Lawyers can be stubborn, especially when they are clearly wrong.
3. In the case of UPI v Kligman, the court declined to rule that “policies cannot constitute implied contracts against employees.” But the court also then went on to consider what must happen if the implied contract is enforceable.
–If a patent policy constitutes a contract, then the university, too, has obligations–often going well beyond pay. Further, if the patent policy forms a contract, then the university is not free to change the policy, and is exposed to liability arising from breaching its side of the bargain.
–One might start with considering whether a university, having argued that it has a contractual right to obtain an inventor’s assignment, also is faced with an obligation to secure licenses and generate revenue on behalf of the inventor.
–Isn’t this the fundamental bargain–that inventions will be licensed to generate revenue while producing public benefit from their use and sale? Even if a patent policy claims it does not have to generate revenue, given that nearly all university patent policies state that public benefit is a purpose of the policy, the public in such policies is still a named beneficiary of the contract, expecting “practical application” through licensing. If the patent policy is read “the university demands ownership of inventions to fulfill its mandate to benefit the public through patent licensing, but the university does not have to license any patents it thus obtains nor create any benefit from the underlying invention”–then one sees what is actually happening. If that part is not enforceable, then how can the obligation to assign be enforceable?
–If a university gets to pick and choose what parts of policy it will follow, then the policy fails to form a contract. It’s not that the policy forms a contract, subject to multiple actions for breach–it’s the other way. There never was an intent to be bound by the university. No contract. No obligation to assign. University administrators want the obligation to assign to persist, even if the university fails in material ways on its end of the bargain. “We want title to your car, even if we never pay you for it–you can go to court and have a judge decide what we owe you, if anything, but you can never have the car back.”
–One might gather from this pattern of analysis that ownership of inventions–not innovation, public benefit, economic development, not even making money or collaboration with industry–is now the consuming material term in university patent policies. Otherwise, the focus of attention would be on compliance with these other areas. Ownership would not be necessary.
4. In Kligman, the court found that the handbook (with its patent policy) failed to create a contract because (Smith’s wording here): “1) the handbook was not communicated as a definite offer of employment and 2) the Handbook states that it is a guide rather than a legal document.” Smith continues: “In a footnote the court implies that the use of handbooks for assignment of patent rights is probably inadvisable.”
–The idea that mere statement of an obligation to assign in a patent policy creates an “implied” assignment is magical thinking. Adding present assignment language might make it appear that the implied assignment is now express, but if there is nothing that connects the policy statement with the offer of employment, nothing in writing, nothing signed, then it is not clear at all that the assignment self-executes, or there is any obligation at all.
–Note that when a policy statement is made, it may limit the options available under common law for the university. It may well be that a university’s position is better (on a number of criteria) if it relies on common law rather than a badly written policy. A badly written policy, even if it does not rise to a formal contract binding employees, may well bind the university to certain actions, with exposure to liability if it fails to perform those actions.
5. States that have attempted to limit the reach of employers to their employees’ inventions may have unintentionally replaced common law principles that provide employers limited access to employee invention with expanded opportunities for employers to claim those inventions.
–The actual scope of an employer’s claim still depends on what an employer does claim. The law does not require the broadest possible scope.
–If the employer can simply state that the employer’s future business includes whatever employees invent, then the law has virtually no effect other than technical paperwork (creating the workaround policy, informing employees of the law).
6. Smith points to two law cases that indicate that inventors and universities have no private cause of action under Bayh-Dole for failure of an inventor to assign patent rights in a subject invention:
Interestingly, the statute provides no private right of action for disposition of patent rights made with federal assistance. That is, it appears that only the U.S. government can sue contractors or inventors who have used federal funding for the invention under 35 U.S. C. § 202 to retain rights it has preserved under this section. In Gen-Probe Inc. v. Center for Neurological Study, the United States District Court for Southern California determined that private corporations have no standing under 35 U.S.C. § 202 to sue an inventor employed by the corporation for refusing to assign patent rights to the corporation from inventions discovered utilizing federal funding.[72]The court also determined that an inventor is not bound by the agreement between the corporation and the U.S. government in the course of obtaining federal funding.[73] A corporation’s employee/inventor is not bound by the government-corporation agreement simply because he is employed by the corporation that receives federal funding. To be bound by such an agreement he must acquiesce to the arrangement as evidenced by his signed statement of his intent to be bound.[74]
–We will come back around to this one. Think (f)(2).
7. The case law (at least as of 1997) was not settled on the matter of whether an employee handbook can create an employment contract. It appears that situations depend on circumstances. Local promises can restrict otherwise at-will assertions and override claims that a policy is “more like guidelines.”
–It is fascinating how university administrators have chosen such convoluted, unsettled practices–using a patent policy to try to obtain assignment, adopting corporate terminology pertaining to “employees” and “at-will” employment rather than acknowledging that faculty are appointed (in addition to whatever status they have as “employees”) and are promised academic freedom and, often, tenure. If they insist on working this way, at least they could rise to the occasion and draft clear, consistent, statements and implement clear, consistent practices.
–A big part of the problem with university patent policies is that their core architecture was for elucidating practices that address common law principles of patent ownership and designating external agents to take care of patenting and licensing matters. Rather than build new policies that announce the suppression of academic freedom and direct university participation in patenting, administrators merely added more language (and mostly badly) and tried to change the interpretation of the language they retained. Probably such direct announcement of a change would be met with stiff opposition; thus, the subterfuges of claiming federal law required them, or that they were just “clarifying” what the policy surely must have meant all along, or that they were just making their policy consistent with the policies at peer institutions.
–The upshot is that the *immediate effect* of most American university patent policies is to create an appearance of certain obligation wrapped in a bundle of uncertainty–uncertain wording, uncertain practices, uncertain statutes, uncertain case law. The obligation that matters to university administrators in their policies is that of assignment. But the place for that is in the employment offer, not buried in an after-thought policy twisted to undermine primary conditions of faculty appointments–academic freedom and tenure among them.