Penn State gets innovative

[updated 5/31/16 to repair/replace broken links; PSU has removed the committee report proposing the change in industry contracting requirements]

Penn State has for years been one of the leaders in industry sponsored research.  In the past few months, they have also changed their IP policies on ownership.  Penn State has announced that it will no longer claim ownership of all inventions sponsored by industry grants.  Here is the Core Council’s recommendation to explore a “new model.” The PSU tech transfer office was the subject of familiar concerns:


The idea proposed by the Core Council, then, is to remove university ownership claims to inventions in industry-sponsored research agreements and instead to raise the indirect cost rate to increase revenue (bold in original):


While it is not at all clear that the delays in negotiating ISRAs are the result of contesting ownership of inventions, removing an ownership claim certainly would change the dynamics of negotiation. Raising the indirect cost rate, however, might make negotiations shorter, if the rate no longer appeared competitive with other research universities offering similar research opportunities.

Hank Foley, the Penn State Vice President for Research, explains the backing thinking on the policy change in a story in [story removed from web]:

“With this decision, the university will see increased revenues, because successful start-ups will repatriate the funds to a much greater extent than with the old ironclad licensing agreements that Penn State used to employ,” he said.

The Penn State thinking is, the university does not have to own and then license to get inventions into the hands of companies.  A good background discussion is at Keystone Edge:

The real value is not in IP ownership, Foley contends, but rather in the contact students and faculty have with real problems in the world.

“We’re moving to the position where if a corporation sponsors research with us, they own it. We prefer it,” says Foley. “We’re looking to get the interactions, the relationships and the ability to work on more pressing problems.”

Here is thoughtful take on it by a seed fund specialist (I differ on the analysis of the problems with tech transfer and especially the “conceptual funnel”–which is, to be sure, conceptual, but misses practice–but it’s a nice discussion anyway).

In the April edition of BVR’s new journal IP Management & Valuation (free Feb version available–h/t Chuck), Hank Foley lays out four rules for industry engagement (gosh, I like that word).

  1. The value is in the research itself, not in IP ownership
  2. Use the simplest agreement indicated
  3. In federal subcontracts from an industry prime where university ownership is required, offer a license option
  4. Explain clearly any exceptions to the above and find a workable solution

This is all good progress.  Rule 3, however, could use a little more work, as it appears to rely on Bayh-Dole requiring ownership, which of course it doesn’t.  Even where the standard patent rights clause gets modified, it is unlikely that a federal agency will require university ownership–no ownership or government ownership, perhaps, but not university ownership.   When a company is sub-contracting federal dollars, if it is under Bayh-Dole protocols, then the company cannot require title or a license to inventions made by any subcontractor as a condition of the subcontract (see the standard patent rights clause, section (g)(1): “the contractor will not, as part of the consideration for awarding the subcontract, obtain rights in the subcontractor’s subject inventions.”

The way around this is to create a university policy that allows the university and investigators involved in the subcontract work to allow industry ownership of inventions under a separate agreement–not part of the subcontract, not a condition of awarding the subcontract, but rather a pre-existing condition, one that is already available to industry in choosing its subcontractors:  “Why, if we choose Penn State, they already have a mechanism by which we can obtain title.”

Note that under the SPRC, section (k)(1), universities may assign to any organization that has “as one of its primary functions the management of inventions”.   The company then is subject to the same provisions as the university contractor–which means, sharing royalties with university inventors (even if not with the university).

Of course, the university could also offer a license, provided the consideration for the license is something other than the funding under the subcontract.  Setting something up for that makes sense.  A company then has a choice (always good):  take assignment, along with the added requirements in (k) (the company already has the other Bayh-Dole obligations–since it is taking federal dollars), or take a license with consideration.  If a company chooses the latter, well, then, it has decided to pay.  No one is making them do it.

Penn State calls its option the Penn State Two-Step licensing process.   The university worries the problem of setting a value on a license in advance.  This is the fear of the Tax Reform Act of 1986 and Revenue Procedure 2007-47, which concerns the determination of private use in the case of facilities built with tax-free bonds.   I have commented on this Rev Proc before, since it makes assumptions about research technology that are a) unfounded and b) inhibit collaboration and trade–making the bond issue more important–tax treatment for investors–than the thing the bond issue was to support–research and innovation.  And that is the kind of administrative arrogance that causes stagnation.

To benefit from 6.03 of Revenue Procedure 2007-47, one has to meet four criteria–1) basic research that’s 2) controlled by the university where 3) IP is owned by the university and 4) the sponsor is entitled to no more than a non-exclusive royalty free license (a “NERF”).

Note that this is not a federal law requiring university ownership of inventions.  It is an IRS statement of a safe harbor.  It is only a protection relative to IRS interpretations of the tax status of bond issues.   The core elements are that the university take title, and offer nothing more than a non-exclusive license under the research agreement.  Well, if the money is a flow down of federal funds, this last item cannot be part of the research agreement!  That is, there is no way to meet condition 4) within the context of a Bayh-Dole compliant standard patent rights clause.

Luckily, there is an amazingly simple way to deal with the nonsense of Revenue Procedure 2007-47:  the university cannot set a price in advance for a technology, but it can allow the sponsor of the research to set the price later–which is exactly what it might do with any company showing up at some later date seeking acquisition of a technology in the university’s back portfolio.

The two-step deal would then look like this:

Step 1:  The university has a policy protocol that allows any sponsor of research, prior to entering into a funding agreement with the university, obtain an option to license any invention that might be made under such funding agreement and can legally be made available for license by the university.

Step 2:  The protocol for such option allows the company sponsor to choose either a non-exclusive royalty-free license, or to establish a reasonable price at its sole discretion when the technology is “available for use”, provided that the reasonable price will include sufficient funds to cover the university’s out of pocket expenditures to acquire patent rights.

The first step removes the transaction from the funding agreement.  This meets the requirements of Bayh-Dole–it is not a condition of the subcontract–and addresses the safe harbor of Rev Proc 2007-47–there is no price set in advance, and no license is offered within the funding agreement–which is exceeds the threshold of “no more than” a NERF.  The second step allows the company to obtain access to research results outside the conditions of its role as a sponsor of research.  That is, the company becomes what 6.02 (the prior section) calls “any non-sponsoring party for those same rights”.  The rights are negotiated prior to the sponsoring.  The company is just such a “non-sponsoring party”.

The university can pack this all down, then, by offering other companies any non-exclusive licenses on the same terms that the non-sponsoring first company (who later sponsors) has established as a reasonable value.  The higher that value, the more everyone else will have to pony up to acquire.  Market rates.   A really high initial reasonable value becomes, effectively, a sole license.

Thus, rule 3 could be changed to:

3.  Independently of subcontracts of federal funds coming through a company prime contractor to the university, the university offers a protocol by which companies, including sponsors of research, may obtain an option to future inventions that complies with the standard patent rights clause of Bayh-Dole and with the Tax Reform Act of 1986.

Note that with such a protocol, companies that don’t have federal funding agreements could reserve a spot to propose a reasonable value for a subject invention made via a subcontract with a different company prime.  The prime would then offer a reasonable value that would have to meet that of other companies seeking a non-exclusive.  If the prime wanted an exclusive, it would have to offer more than the sum of what the companies seeking to reserve a non-exclusive have offered.  Interesting way to establish market value–actually create a market!

In all, the Penn State effort is a good one.  As commentator Michael Carrier puts it, Penn State is placing its academic values ahead of commercial interests, and in doing so, it believes that it can spark collaboration and innovation, and that in the long run, it will also come out ahead in terms of financial support for its programs.  Forward thinking all the way.

Update 5/27/16. Here is Penn State’s current policy on ownership of inventions made in industry-sponsored research. The policy is broader than just patentable inventions and extends to “Intellectual Property”–a complication. In the “Project Participation Intellectual Property Agreement” (called PIPA), however, the wording skips back and forth between “any patents resulting from the above referenced project” and “intellectual property that has potential commercial value.” (Why the pdf document posted on the web at PSU is titled “University of Georgia” is beyond me!) These are mixed up propositions. Intellectual property could refer to patents and copyrights, or it could include the stuff that is the subject matter for patents and copyrights, and it could include as well technical information (under a trade secret theory).

Here is the Penn State definition of “research intellectual property” (distinguished from “instructional intellectual property” and “scholarly intellectual property”–clearly, research is not scholarship at Penn State, at least not in the strange world of IP administration):

Research intellectual property is the term used to describe the discoveries, inventions and creations with potential commercial value that result from research activities.  Most research intellectual property developed at the University can be protected by patents, but some University research intellectual property (i.e. software) is more appropriately protected by copyright.  Although some research intellectual property may be protected by trademark or trade secret, it is rare for the University to utilize these methods of protection.

In short, there is no definition! All the “definition” says is that the phrase “research intellectual property” is a term used by administrators. The term then substitutes for stuff (“discoveries, inventions, creations”) with “potential commercial value that result from research activities.”

Something with relatively stable usage–inventions which are or may be patentable–is replaced by this definition with something that’s purely a matter of opinion–“potential commercial value.” How would anyone be expected to know whether something has “potential commercial value”? Anything could have “potential.” The point of working with an industry research sponsor is that the sponsor sees what’s been done and recognizes value. Value is not the researcher’s responsibility. In industry settings, even whether something is patentable doesn’t matter. Just report the results, and let the sponsor decide on value (the first thing) and then bring in the patent attorneys to consider whether the valuable thing is also patentable. There’s then a third step in industry–deciding whether to pursue a patent, even if something is valuable. Trade secret and open publication are perfectly viable alternatives to patenting.

Because Penn State’s definition of  intellectual property expressly references patent, copyright, trademark, and trade secret, it is reasonable to think that when Penn State discusses ownership of intellectual property with research investigators and industry sponsors, the discussion includes trade secret. If an industry sponsor is granted the option to own “intellectual property arising from the research,” then, it could choose to own as trade secret–meaning, no publication of the results. Likely, Penn State would object. But if so, then Penn State does not mean what it has put in its definition of “research intellectual property.” It is just writing administrative goofiness here.

The PIPA goes one step further and notes with emphasis the following:


But if Penn State is charging a higher indirect cost rate in order to give up an ownership claim, then that added amount is, in effect, the consideration for the deal–a royalty (with the meaning “any consideration for a patent license”). Inventors should, under policy, have the right to a share of any income involving the disposition of their inventions, regardless of what university administrators choose to call the fees. That was one of the issues in the Singer case at the University of California. One would think administrators would learn the lesson that it’s not what they call something, but rather what that something does.

The upshot of this warning in the PIPA, then, is that inventors won’t benefit “financially” from their inventions, but Penn State will. And by “commercialization” here one can only infer that administrators mean “assigning ownership to the research sponsor” and not something along the lines of “development of new product offered for sale.”

But all the worser then is the complication that the Penn State policy cannot bring itself to have inventors assign IP directly to the company sponsor upon request (which is how the standard patent rights clause authorized by Bayh-Dole works it). Instead, there’s a convoluted path of disclosure to the university, assignment to the university’s research foundation, communication to the sponsor, and then assignment from the research foundation to the sponsor. Why have just one assignment and disclosure when you can have two of each? Beyond that, Penn State appears to “reserve” for itself a non-exclusive license to “practice the intellectual property for noncommercial research and educational purposes.” Essentially, the university insists on a grant-back that degrades the idea that a sponsor may own the inventions that its funding (and statement of work, perhaps) has supported. “Practice” is a broad term–in the context of patents, it could mean “practice the invention” or “use.” But in the PIPA the university uses “practice the intellectual property”–which suggests “practice the patent” or any of the rights of a patent owner that aren’t overtly “commercial”–make, have made, use, import. It’s not clear. One has to make a guess at what the administrators mean here, as it does not appear that they are capable of getting anything out directly.

One final note of strangeness. Penn State’s patent policy requires an intellectual property agreement to be signed by all administrators, faculty, and staff. That agreement includes a present assignment (“hereby assign”)–as most university policies and patent agreements do, thinking that “hereby assign” are the magic words, and not the scope of the assignment. If the scope is nebulous, then how can anything be said to be assigned by the magic of “hereby assign”? Think, oh administrators. If you have to fuss over what is included in the scope–especially for faculty–“to the expect specified in University policy” (that is, the scope is however policy gets interpreted, not something definite) and further is restricted to PSU-scoperest

Perhaps it is too hard to think about these things. An invention may get made using facilities or resources, or while one is using facilities or resources one might get an idea about something entirely different. A new way to do things, not the way one is doing things. A new application that’s rather very different from what one is aiming to do in using facilities or resources. That new idea is not “with the use of University facilities or resources” until the person who has the idea decides to act on it “with the use of University facilities or resources.” That is, inventive ideas do not necessarily come about with the use of facilities or resources. One may read or hear something quite apart from fussing in a lab or spending university money or using university photocopy machines and come to an inventive idea. For prong (a), such inventive ideas are outside the scope as stated.

No doubt, following established university IP administrator practice, “with the use of University facilities or resources” perhaps is intended to include “without the use of University facilities or resources” and as well “at the same time that one is also using University facilities or resources.” But the agreement does not state these things. A true intention, if not expressed, is irrelevant, to paraphrase the Shaw court.

Now look at prong (b). There are multiple claims here, packed into administrative slices and made ambiguous by a terminal modifier “covered by my [list] with the University” that may apply to the last element (the grammatical thing) or to all elements (the apparent thing).

We have the following combinations:

in the field of expertise covered by my employment with the University

within the scope of responsibilities covered by my employment with…

Nothing here says “within the scope of my employment.” There is “scope of responsibilities”–but what makes something a “responsibility”? And how is it “covered” by “employment”? Furthermore, how are responsibilities “covered” by “appointment” or by “association”? And then there’s the idea of “field” of “expertise.” What could that include? If a faculty member is appointed as an associate professor in the electrical engineering department, is the “field of expertise” then “electrical engineering” or is it “electromechanical systems” or is it “nanowire assemblies”?

When one gets hired as an electrical engineer at a company, the scope of claim on inventions is not the field of electrical engineering (that would be a non-compete claim) but rather the company’s business purposes–look at what the company does and plans to do, then consider what it is one is directed to do, and one has a good idea regarding scope. If a company merely says, “we own everything that you do in the general field of electrical engineering because we declare that in addition to everything else, our company intends to profit from whatever you do, regardless of what our company otherwise does, then the company is merely abusing its position. Some states outlaw such behavior as a matter of public policy. But university administrators don’t care about such sensitivities as public policy–they make up their own public policy, as it suits them.

It is true that “we own everything” has a definite feel to it, that there’s no contesting the scope. But it is an open question whether an employment agreement (or appointment) can extend to everything in one’s “field of expertise” as a matter of contracting. That is, administrators can put anything they want on a paper, but is that demand enforceable? In a world of slavery–perhaps–where the master owns the slave, and therefore can require most anything. But in a world of employment, the employer’s bargain is not the expertise of the employee, but the use of that expertise to those things that the employer directs. Anything else is an abuse of the employment relationship.

There might be an agreement between an employer and employee outside the employment agreement–with its own scope and carrying its own consideration–but that’s not what is going on here. The Penn State intellectual property agreement makes it expressly a matter of “in consideration for my employment/appointment/association.”

One might think “continued employment” (as that is often acceptable consideration for a change in an employment or non-compete agreement) but no, the idea stated here is that the employee provides “consideration” to the university in the form of IP in exchange for employment (or appointment, or association). One would think that the employer would provide consideration (“pay”) for the provision of services. That would be conventional employment. Here the employee provides consideration (“all my IP are belong to you”) for the employer’s provision of the service of employing the employee.

It’s all very convoluted. It is painful to try to make sense of the language. One might come to think that there is no intent to make the language clear and direct, to find proper limits to claims of ownership. There is no meeting of minds, other than that everyone must “abide by the terms of the University’s Intellectual Property Policies and Procedures currently in effect, as well as any subsequent revisions thereto.”

In a policy document, one might require employees to follow the policy, at the risk of sanctions or loss of employment. But this is not a policy document. It is an intellectual property agreement with an embedded assignment clause. It says, “you agree to whatever we ask, or may in the future change our policy to ask for, and you assign all of that upfront, now, before there’s anything at all to assign.” The scope is indefinite. This is an agreement to agree–unenforceable. It is a policy that bites its own bum:

IP Policy: you must sign an IP agreement

IP agreement: you must obey IP policy

The IP agreement should document the scope of employment, list the responsibilities and the expected areas of invention in which the employer has a claim. Instead, the IP agreement waxes nebulous and recites the abstracted claims of the IP policy! Perhaps the Penn State IP complex is merely a form of psychomagnotheric slime. Whatever it is you think you have created that might mean something to you opens up a portal for administrative ghosts to enter your world and slime whatever it is that you have.

The language used by Penn State has all the appearances of a “legal document.” It is intended, no doubt, to be “enforceable.” But not in the way you might think–not with clarity, not with a meeting of the minds, but rather with such a tangled mess of definitions that aren’t and restrictions that are expansive that it would take a court a long time to figure out just what the wording reasonably might mean. Of course, as policy, the wording can mean whatever administrators want it to mean. The problem is, as a contract it cannot mean whatever administrators want it to mean. It has to have a meaning that any reasonable person can come to, and that meaning has to be stable, and within public policy.

This then is a fundamental mental flaw in the management of university IP policies. Administrators start with policy statements. Policy statements distribute organizational authority and provide a standard for allocation of organizational resources and sanctions. If someone does not follow policy, they may be disciplined or exiled from the organization–fired. Contracts, by contrast, are an exchange of mutual promises framed by offer, acceptance, and consideration. Both parties are bound to a contract by a common understanding of its terms consideration and reliance. One party cannot change the terms unilaterally. If one can, then it’s not a contract. It’s just a wish, or a threat, or pscyhomanotheric slime.

The problem for faculty (and anyone else that gets close to a university haunted by Slimers) is that it costs about $200K to go into court and show a judge that the university’s policy and agreement documents don’t mean whatever administrators say they mean and run against public policy to claim to own anything that a faculty member might do because the policy states a claim to “field of expertise” rather than states, for any given employee, a scope of employment.

So these policies are written not for clarity but to run up the cost in court of contesting them. The “burden” of showing how screwed up the policy and mandated agreements are is transferred (as much as possible) to the employee. In the Penn State IP agreement, this transfer is made express:


That is, here, in what appears to be a contract (“I intend to be legally bound by this IPA”), the faculty member is made to have the obligation to understand university IP policies, but does not have any right to an interpretation of those IP policies. The right of knowing is reserved to the Office of Technology Management.

Worse, the IPA includes a reverse incorporation of terms by reference:


Skipping the question of how one can have “terms of association” and the like, the language here asserts that the IPA (though signed separately, and likely after accepting employment or a faculty appointment) is somehow part of that employment or appointment agreement. It comes in later, through a portal, to claim that whatever the meeting of minds may have been, it also had to include this IPA, too. More so, the agreement claims that any future agreement will incorporate this IPA agreement by non-reference! That is, if any future agreement does not disclaim this IPA, then the IPA becomes part of that agreement. So very rich. Instead of the university expressly incorporating the IPA into any new agreement, it now can simply claim that the IPA is part of every other agreement. It does not matter even what that agreement might be–it could be an agreement regarding a release from the IP policy, but no matter, the IPA, if not referenced would be part of that agreement. The IPA is an agreement that intends to be a part of every agreement that does not exclude it. How utterly strange. An agreement incorporated by non-reference into every other agreement.

This IPA is no contract, though it is made to look like one. It ought not to be enforceable, lacking definite scope, making unreasonable and inequitable claims to own private property, constituting an agreement to agree, and making outrageous claims that violate public policy. It is an administrative instrument designed to abuse power, a creature of policy, a slimer. In a reasonable world, unlike the one we inhabit, this kind of agreement would be malpractice and get a coven of administrators and their legal advisors fired and sanctioned. As it is, it’s just another day in the haunted house of university IP management practice. Folks will get away with this stuff until they are called to account for it–for its legal monstrosities, for its damage to the mission and values of the university, for its disruption of publication, for its hapless approach to innovation, for its failure to generate promised millions in income, for its undermining of goodwill.

This entry was posted in Policy, Sponsored Research, Technology Transfer. Bookmark the permalink.