We have been through a lot of Penn State policy now. We have looked at the 1940 IP policy, the 1991 IP policy, and the current IP policy. We have also looked at the 1992 IP Agreement and compared it with the 1991 IP policy, and the current IP Agreement in the context of current IP policy. Now we can look at how the IP Agreement has changed. We have touched on some of these changes already, but here we will work through the changes in detail.
Oh, I know, it likely doesn’t matter. No one at Penn State (I expect) will give a rat’s ass what the words say in their policy documents. It’s what administrators assert that matters, not what someone finds by carefully reading policy. Perhaps Penn State should have a policy requirement that personnel, if they read and understand policy, must not read carefully and must wait to have their understanding confirmed by administrators, who are expert in not reading carefully and so may guide others to an understanding that administrators will approve of.
In 1991 Penn State had an IP policy that required university personnel to sign an IP Agreement and stipulated that personnel had a “personal obligation” to disclose inventions (defined broadly and strangely to include non-inventions). The IP Agreement, in turn, required personnel to “abide by the terms” of the IP policy. Sort of stupid, but mostly harmless in its way. The part that wasn’t harmless, though, was that the IP Agreement produced the next year, in 1992, required personnel to assign all inventions to the university that were within the scope of what the IP policy required to be disclosed.
There are two scopes at work in an IP policy. One is the scope of disclosure–what should be reported; the other, the scope of assignment–what the university can properly claim. In practice, these scopes will be different, especially if there is room for administrative discretion in what the university may claim or ambiguity in what constitutes the formal scope of claim.
The scope of assignment concerns those specific inventions for which a university has contracted to obtain. Everything that is not so contracted is unclaimed by the university, as a matter of policy. A scope of disclosure may be drawn more broadly than the scope of assignment to cover cases where there might be some doubt as to whether an invention is subject to the university’s policy claim or where the university may decline to own an invention that has been reported. In those cases, the circumstances of development and the nature of the invention may be reviewed against the conditions stated in university policy to arrive at a determination of the university’s policy interest in a given invention. If an invention isn’t properly “commercial” or would not benefit from a patent monopoly, then the university would be wise not to accept the invention for patent management. Or, if the circumstances of development indicate that the university did not contribute expressly to the inventive work or did not specifically commission the inventive work, then again it would be equitable for the university to confirm that it has no ownership claim in the invention.
Of course, if the university’s policy conditions are a heap of parrot droppings, then it will be difficult for anyone to arrive at any objective interpretation of circumstances matched to policy. If university administrators anticipate this problem, then they might make the university’s claim to ownership as broad as possible (so that administrators may waive whatever they find they don’t want, rather than accept only what clearly meets the policy’s stated scope of ownership claim). In a grasping IP policy written by incompetents worried that they don’t know how to distinguish what they should legitimately have rights to from everything else, we often find a claim to most everything, with the ownership claims expanded to whatever is disclosed, as if there is no difference. Such policies run into significant problems. However, since the problems are buried in complexities of law and practice, it is not easy (or inexpensive) for any given inventor to identify the problems and resist overbroad administrative claims. Indeed, inventors risk not only huge legal expenses but also having their reputations damaged, if not their academic careers put at risk as administrators escalate charges from failure to comply with policy to unethical behavior in refusing to assign inventions to the university.
In early university patent policies, a standard practice was that patentable inventions were to be disclosed and reviewed for the circumstances of development, and based on those circumstances a faculty committee (sometimes with administrators) would recommend whatever (if any) equitable interest the university might have in any patent obtained. There was no outright claim by the university to own any invention prior to a review of circumstances. A number of policies stipulated that if someone was hired to invent (“official duties” and the like)–especially in the emerging research institutes focused on contract research–then that circumstance would create a university ownership claim. Over time, the review procedure became cumbersome and administrators did not like to have to produce evidence of university involvement, and so written agreements were dropped from patent policy, as were reviews for university interest. In their place were placed broad statements of university interest, often leaving intact earlier statements of limited university interest.
One can have disclosure and assignment scopes be the same in open IP policies–where when someone discloses an invention to the university, it is with a request that the university consider acquiring ownership of the invention. In this practice, an inventor decides that what is to be disclosed is also to be assigned. Even then, a university might decline to take ownership, and thus even in a voluntary disclosure/assignment patent practice, the disclosure scope can be expected to be broader than the assignment scope. The Penn State 1992 Patent Agreement, however, equates the IP policy obligation to disclose with an obligation to assign, even though there is nothing in the IP policy that authorizes such an obligation to assign.
By the time we get to the current Penn State IP policy and IP Agreement, things have gotten much worse for inventors, research enterprise, and innovation. Let’s work through the changes to form an idea of what has happened. The 1992 IP Agreement is here [see pages 57-58 of the linked report]. The current IP Agreement is here.
Here are the opening line of the current Penn State IP Agreement:
In consideration of my employment/appointment/ association, I agree to abide by the terms of the University Intellectual Property Policies and Procedures currently in effect, as well as any subsequent revisions thereto.
Compare the 1992 IP Agreement:
I agree, as a condition of my employment, to abide by the terms of the University’s Intellectual Property Policies and Procedures currently in effect, as well as any subsequent revisions thereto as approved by the Board of Trustees of the University.
This is sophisticated bombast. First, the IP agreement points back to the IP policy. So we have a loop. The IP policy opens with a requirement that as a condition of employment, people must sign the IP agreement and follows with a requirement that people disclose certain inventions and other new “creations.” The IP agreement in turn stipulates that people must “abide by” the IP policy. Big wow, given that faculty are already required to “abide by” university regulations elsewhere in policy.
The Effect of Making a Policy a Private Contract
What is added by creating a formal agreement on the matter? It turns out that the use of private contracts does not somehow strengthen the university’s control over the situation. Instead, the use of a private contract creates mutuality. Where a policy may be controlled entirely by the institution, both parties to a contract have an interest in the promises that have been made.
1) The university, by entering into a contract with each employee, gives up its ability to change unilaterally the requirements it places on those employees. The contract binds the university to its policy promises.
2) The university gives up its ability to waive the requirements of policy with respect to any other individual under a similar agreement–to do so would be to misrepresent the policy requirement that authorizes the signing of the IP Agreement;
3) The university cannot require anything of an individual having signed the IP Agreement that is not already in IP policy–so no rogue administrative demands, no making things up on the fly.
But these are all constraints on the university. What’s odd is that the IP policy and IP Agreement make it appear that their purpose is to place constraints on individuals–and in some ways, these documents do do that, but they do so by using an instrument that also places substantial restraints on the university itself.
The aim of the IP Agreement (not necessarily or expressly the IP policy) is to cause inventors to surrender their inventions (including non-inventions) to the university for its money-seeking technology transfer efforts. To do this, administrators had to get at inventors’ personal property–their inventions and works of authorship (the policy conflates inventors and authors). They had to find a way to force inventors as university employees to give over personal property that federal common law on inventions and federal copyright law vested in individuals, not in the institution. The IP policy (were it to require assignment of inventions) amounts to a demand that individuals acting in their “official” state capacities must also act privately to assign their personal property to the university. Most university IP policies that aim to be grasping have to navigate this problem–and usually do so by trying to obscure the situation.
Companies for the most part do not use this approach. Companies contract at employment for rights to inventions within the scope of the company’s business, present and future, and within the scope of the employee’s scope of employment. They handle works of authorship with work made for hire provisions. They combine control of inventions and works of authorship and information with trade secret and (in states that permit it) non-compete covenants. Company employees work at the direction and for the benefit of the company. These conditions create the basis for the company’s claim to employee work product and the IP that attaches to that work product. Continued employment is typically the primary consideration for compliance.
Universities differ substantially from the company circumstances. Faculty are not employed in the usual sense by a university for their research or scholarship. Even most sponsored research does not involve employment, though in folk usage we would insist that faculty working on a grant are employed. It’s just that the university does not assign or control the work and the work is not done for the benefit of the university, and the IP generally falls outside the scope of “business” for the university–except of course for the IP management program, which aims to assert that the “business” of the university is actually to make money from anything sellable (“that might be made available for commercial development in the public interest”) that anyone at the university might create. Thus, nothing can fall outside the scope of the university’s claim on IP, even as the university has, for almost all inventive work, no right to assign or control work. Part of the failure has to do with a misguided effort to impose company-style IP management on universities.
Part of the failure, though, has to do with a fundamental failure of competence to see that a university as social structure is not a for-profit company. Gosh, who’d have thought? Many (not all) university IP administrators have happily adopted the idea that they represent the vanguard of “culture change” within universities, that universities will be better if “run like companies” and if that is to be the case, then faculty are properly treated as mere “employees” who owe their work to their employer, to the IP administrators. Lost in all this is the fundamental reason for a federal subvention in support of faculty-proposed scientific research. But then who needs reasons once something like federal support for university research has become established as a given, as an entitlement, as so obviously necessary that reasons aren’t needed anymore, just the occasional reminder of how truly excellent it is to have university research?
An IP policy might authorize university administrators to enter into contracts with university personnel to manage ownership of future IP those employees create. The IP policy might also establish a scope for when the university should negotiate an ownership interest in that IP (such as, when an extramural contract requires university ownership or when the university expressly commissions work anticipating inventive results). An agreement, then, would state a clear scope and purpose and involve its own offer, acceptance, and consideration, independent of any employment agreement. It would be clear that the university was paying not only for the work performed but also for the resulting IP. The scope of such an agreement would never be “anything in one’s field of expertise” but rather only “these things specific to the purpose of the sponsored or commissioned work–that is, deliverables.”
By turning an IP policy into a private contract, a university gives up its unilateral right to interpret the policy or to change the contract by changing the policy. The contract requires interpretation of policy to follow the rules of contract interpretation–and where policy is concerned, the contract that binds that policy will take the form of an adhesion contract, since the individual has no right to negotiate the terms. Thus, for ambiguous provisions, the contract will be interpreted against the drafting party–the effect of making policy into an adhesion contract is to transfer the interpretation from the university to the contracted individual.
If IP policy controls, then a university official can control both its form and its interpretation, but cannot force inventors to assign rights, even if the policy insists that the university must obtain rights. (If a policy requires the university to obtain rights, that in itself is not a mandate to obtain rights by force or threat–it may mean simply that the university is authorized to offer whatever generous terms induce most inventors to willingly assign their inventions to the university.) If the university official then contracts to obtain rights, the official gives up control of IP policy and must rely on the contract that the university has entered into with each individual. The Shaw court found the University of California IP Agreement incorporated the university’s IP policy, that as a result, Shaw was equally entitled to the benefit of the contract that the university had formed with him, and thus (my emphasis):
The University prepared, and Shaw signed, a patent agreement whose references to the Patent Policy are so direct as to indicate the parties’ intent to incorporate the policy’s then-existing terms into the patent agreement, including the University’s promise to pay Shaw 50 percent of the net royalties of any patentable invention. Whether, absent the incorporation, the Patent Policy would constitute a mere statement of personnel policy is immaterial. Having made the Patent Policy a part of its written agreement with Shaw, the University may not unilaterally revise it as to him.
The Penn State IP policy/IP Agreement approach is remarkably similar to the University of California’s. Penn State’s IP Agreement expressly references the IP policy as the instrument that controls an inventor’s responsibilities.