Administrative patent policy swarming

I have been intrigued by a story David Byrne tells in How Music Works.   Byrne needed three dancers as part of an ensemble, so they held an audition that had fifty dancers.  The choreographer had the dancers do an exercise with these rules:

1.  Improvise moving to the music and come up with an eight-count phrase.
2.  When you find a phrase you like, loop it.
3.  When you see someone else with a stronger phrase, copy it.
4.  When everyone is doing the same phrase the exercise is over.

Byrne then describes what happened:

It was like watching evolution on fast-forward, or an emergent lifeform coming into being.  At first the room was chaos, writhing bodies everywhere.  Then one could see that folks had chosen their phrases, and almost immediately one could see a pocket of dancers who had all adopted the same phrase.  The copying had begun already, albeit just in one area.  This pocket of copying began to expand, to go viral, while yet another one now emerged on the other side of the room.  One clump grew faster than the other, and within four minutes the whole room was filled with dancers moving in perfect unison.  Unbelievable!  It took only four minutes for this evolutionary process to kick in, and for the “strongest” (unfortunate word, maybe) to dominate. (67-68)

Now consider what Archie Palmer was doing–gathering university patent policies, publishing them, and encouraging universities to adopt such policies.  Without formalizing the rules, it sure looks like the process has a lot of similarities to Byrne’s dance audition.  In his 1962 compendium of policies, Palmer describes the process of revision:

In addition to the new policies adopted since the 1952 report, many of the earlier policies have been revised, as have the generally accepted practices, to meet changing situations and attitudes in the institutions concerned, particularly with respect to sponsored research and the disposition of patent rights in new discoveries and inventions emanating from such research. (9)

Palmer later elaborates on the process of policy adoption and revision:

Some of the more recently adopted policies are patterned after those already in effect at other institutions, particularly the earlier policies of the University of Illinois, Lehigh University, the Massachusetts Institute of Technology and the Pennsylvania State University.  Several institutions have embodied their policies in their employment agreements, following the example of Michigan College of Mining and Technology.  Other recently adopted policies show evidence of recommendations and suggestions made by the author in an advisory or consultative capacity, as well as through his earlier publications, particularly the Supplement to University Patent Policies and Practices, published in 1955 by the National Academy of Sciences–National Research Council. (14)

Seems a lot like the dance choreographer’s audition, doesn’t it?  Establish a policy, check around for other policies, adopt what looks “stronger” until clusters form, practices become uniform, and without central control “a simple algorithm” generates “unity out of chaos” (to quote Byrne again).  In the case of a dance, it’s pretty neat.  In the case of cultural practice generally, perhaps Byrne’s concern with “dominate” matters–variety is good, and monoculture is, well, monoculture.  Variety, while being apparently inefficient, inconsistent, inequitable, and administratively inconvenient, also turns out to be robust, anti-fragile, responsive to a range of uncertainties, and promotes freedom of choice matched to opportunities.  Variety is part of that theme of Stewart Kauffman’s–being in the ordered domain on the edge of chaos.

It would appear that Palmer, by giving visibility and ready access to university patent policies, the many schools that lacked these policies began to write them, and as they did so, others began to revise theirs, and in comparing notes, the policy drafters tended to adopt the “stronger” moves–perhaps stronger meaning, from the “established” programs, or moves that appeared “more authoritative.”  In any event, the first order of business, as far as Palmer was concerned, was that a university *have* a formalized patent policy.  The second order of business was that the policy was not case-by-case, for such things as the “equities” of the inventors and university.  “All too frequently these determinations are made on an ad hoc basis after the invention has been disclosed,” complains Palmer.  One wonders what an institution is going to do otherwise, if the policy arguing for an institutional “equity” in inventions bases the claim on unusual use of funds and resources beyond the normal “academic environment” and “regular salary”?  How would one know in advance what those unusual circumstances might be, and how they might change a determination of what’s fair for a university to claim, on its own behalf, or on behalf of an abstract public?

It would appear that the pressure to make uniform policy, indicated by Palmer’s commentary on practices, also leads policy makers to create conditions that justify such uniformity:  so rather than consider special assistance to faculty inventors, and when such assistance should be freely provided, when it should be done with an expectation of reimbursement, and when that reimbursement should be a share of proceeds, even if more than what was provided, and when a claim of ownership should be established–rather than considering these things, a policy-maker would be attracted to the idea of basing the ownership claim on employment or use of any resources, and then making the royalty-sharing schedule a firm deal, a matter of policy rather than of judgment.   One trades an empirical determination of what is equitable for an administratively expedient one.

It is odd.  By establishing a royalty-sharing schedule–how much the institution will “share” of patent licensing income with the inventors–a university makes uniform its treatment of inventors.  The royalty sharing deal is then as published.  In essence, the policy provides that there are no private deals, no special favors for the politically powerful faculty inventors, for instance.  Everyone knows what everyone else is getting.  However, to get to such a deal, one has to first reach to ownership.   Without ownership, then one is talking free or reimbursement, not the imposition of a uniform sharing of royalties.  Palmer worries this issue:

At most higher educational institutions the compulsory assignment of patent rights is not considered desirable, except when it is necessary in connection with cooperative or sponsored research and is required, even when not specified in the research contract, in order that the institution may fulfill its contractual obligations.  However, voluntary assignment is preferred and, in many institutions, is encouraged and facilitated either through prescribed procedures or through special machinery provided within the institutions for handling patent matters. (10-11)

A key driver of institutional ownership appears to have been that of sponsored research.  As the National Science Foundation takes off and government research funding to universities grows significantly, 1962 represents something of a pivot point for university invention ownership.   Suddenly, government funding becomes greater than foundation and industry funding.  Suddenly, there is a need to comply with grant contract obligations.  Ownership appears to be one way to do that.  There are other ways to comply, such as making researchers party to the contracts or requiring them to enter into specific agreements responsive to the sponsors’ requirements regarding ownership, licenses, publication review, deliverables, and the like. And there are other drivers, including the analogy that if the federal government can demand ownership of inventions as a condition of its funding, then so too can a state government demand ownership of inventions as a condition of its funding.  Monkey see, monkey do.

When Bayh-Dole came along, it was one more effort to make uniform a federal policy on ownership that had been based, expressly, on flexibility.  The advocates of Bayh-Dole believed that what they were doing was forcing universities to take an ownership position in federally supported inventions, as a matter of compliance with federal law.   They were forcing a policy change both by claiming that the federal government endorsed institutional ownership of faculty inventions, and that such ownership was required by law.  Like dancers in an audition, university administrators looked at how other universities were implementing Bayh-Dole, they looked at the guidance written up by COGR and AUTM, and they believed what they read–that universities owned federally funded inventions (or, had the first right of refusal, or that inventors had to assign, or that invention ownership changed when the university elected [to retain] title, or that little fairies carried title in their tiny hands at night, by moonlight, over dew-dripped clover to the desks of patent administrators–it really didn’t matter how).  The administrators adopted the “stronger” position, the “couldn’t hurt” or “just in case” or “to be on the safe side” position, and claimed ownership of inventions.  They danced their way to an expedient invention management monoculture based on a claim of institutional compliance with federal law regarding sponsored research.

Uniformity is, in university invention policy business, implicitly equated with fairness.  But nothing could be further from the truth.  There is nothing “fair” about using a university claiming ownership of an invention and 70% of proceeds after costs because an inventor used a university lab for a half an hour to evaluate an invention, as if the institution had put in $100K over three years along with specialized laboratory space in order to pull off a neat bit of work.  The only way such situations are comparable is in the expedient mindset of administrators who have already decided that ownership is their pathway to avoid having to make judgment.  It is judgment, however, and not arbitrary imposition of an outcome, that maps to equitable treatment.  Especially this is the case with discoveries, and the circumstances surrounding them, and even more so in universities, with expectations of independent faculty and open laboratories and freedom to publish.

The problem for any discussion of university ownership positions is how to break from the attraction of “stronger” positions, ones that favor institutional self-interest, administrative convenience, and avoid dealing with any special circumstances–or any circumstances whatsoever.  One can see in this development a movement from freedom to innovate to policy-based procedures, to claims of ownership, to making those claims uniform for special cases, and away from case-by-case consideration, to making the special cases the norm, and calling the old norm inequitable (“we have to screw over all inventors equally”).  The next move in administrative expediency is to make ownership compulsory.   That’s what the present assignment movement is doing.  It is going way beyond the claim that Bayh-Dole requires (enables, vests, enfairies, etc.) ownership and argues that universities should own outright, upfront, perfunctorily everything–not just patentable inventions, but copyrights (except for scholarly publications, and art), know how, data, whatever and however.

In making this expansion, however, invention management offices also create additional exposure to liability:  they are taking more than what their institutions have contracted for, they are running a portfolio model (a few lucrative licenses a decade create the appearance of success) but advertising an agent model (every client invention is marketed to industry for public benefit).   They are responsible for value creation across everything they have claimed, but clearly are capable of doing little with 99% of it.  Thus, the next stage is to make the decision-making of the institution a matter of autocracy.   Gone are the faculty patent committees.  Gone are the arbitration panels to decide the equities of the inventors and institution.  Gone are processes of review for management.  In their place are the final often non-public decisions of administrators, with mandates to interpret policies however they wish, and always for the benefit of the institution.  There’s really nothing equitable about it, other than perhaps in the appearances to those who aren’t involved in the practice of invention management, like policy makers and academics studying “models” of innovation.

Ed Yong has a piece in a recent Wired about “the science of swarms”.  There, he gives accounts of researchers who are studying self-organizing behaviors.  He describes the software modeling efforts of Craig Reynolds:

Sixteen years later, a computer animator named Craig Reynolds set out to find a way to automate the animated movements of large groups—a more efficient algorithm would save processing time and money. Reynolds’ software, Boids, created virtual agents that mimicked a flock of birds. It included behaviors like obstacle avoidance and the physics of flight, but at the heart of Boids were three simple rules: Move toward the average position of your neighbors, keep some distance from them, and align with their average heading (alignment is a measure of how close an individual’s direction of movement is to that of other individuals). That’s it.

It is not that university administrators want monoculture, or to control faculty scholarship, or to disrupt innovation pathways, or to lose out on financial gains because they are so frickin’ difficult to work with.  Hardly.  But they do fear being out of step with convention (as it puts their own judgment front and center), they do want the advantages they envy in the policies of others, and they do want to appear to be going in the current trendy direction, the direction which makes them appear to be “leaders” and “progressive” and “up with the times” rather than appear to be dopes and luddites.  If there is a whiff of present assignment in the air, why then fly towards it, observe your mates, keep some distance, swarm.

In essence, in university patent policies, we are seeing administrators swarm.  It takes decades for this to happen, since university policy changes typically take something like five years to be put in place and won’t be revised for some years more after that.  Driving forces are publications like those of Archie Palmer and the AUTM Practice Manual, as well as the huge national effects caused by the growth of federal funding and the uniform demands (and misinterpretation) of Bayh-Dole.  Add to that the collapse of the public domain (especially patenting life, software, and business methods) and the research exception (Madey v Duke) in patent practice and one has a cocktail for an administrative swarm, a bozonet of moving to the common mean based on whatever happened to be the institutional move that developed the strongest local cluster, the strongest meme.  No matter if that meme was the “strongest” overall, or was or is best matched to the local needs of other organizations or their researchers.

As Geoffrey Moore has pointed out in regards to technology adoption, institutional acquirers are pragmatists, not visionaries.  They adopt when they see others adopting.  They adopt what they think will be the strongest product from the strongest provider, and they do this especially when they believe they have a choice.  They create a dominant player, like a Microsoft or Dell, because in doing so they create a safe harbor for their choices:  they went with the market leader–that’s (to their minds) the most defensible position.  University patent policy writers are pragmatists through and through.  They won’t adopt a new, untested model.  As Clark Kerr has quipped, university administrators never adopt something for the first time.  It’s always pre-chewed somewhere else.

The emergent properties of this administrative swarm are not at all the stated intentions or objectives of the swarming behavior.  The stated objectives are things like public benefit, innovation, economic vitality, money for more publicly spirited research, equitable treatment of all “stakeholders”, compliance with law and contract, avoidance of liability, and protection of research, and even academic freedom.  But the consequences of the swarming behavior turns out to be very different, and in some cases almost the opposite of the claims.  Innovation happens despite the effects of the administrative swarm, despite the fragile expedients of a monoculture of invention management, despite the technology licensing offices. How much more innovation might happen without them, at least in their present policy-cemented positions?

Stagnation does not have to look like a lack of activity:  there can be plenty of activity, as the ever-misleading and unhelpful AUTM licensing survey reports–invention disclosures, patent applications, patents issuing, licenses, startups, money and money multiplied by modeling factors to create the appearance of even more money as “economic impact”.  But stagnation can look like everyone doing the same thing, without judgment, without even a need to consider circumstances.  The only changes a status quo can contemplate are ones that improve the condition of the status quo:  efficiencies, mitigations, incremental improvements–hey, “express licenses”! “present assignments”!  It is policy and process driven change–and that means, change that deepens entrenchment of a model, change that makes change harder to come by, and that means, next to no change at all.

Such change seeks to control the behaviors of others to conform to its expectations.  The purpose of participation in technology transfer is to make the program successful.  The starting point is university policies:  make everyone conform, and without alternatives.  Then create research contracts and licensing agreements that force conformity on industry.  Negotiating a sponsored research agreement with a university is not so much a give and take on possibilities as it is being informed of one’s non-compliance with university policies.  Policies have replaced judgment.  Fear of personal responsibility has replaced delegation of responsibility.  Being like everyone else is more important than doing what ought to be done.  I have even heard invention management administrators argue that straying from common practices creates for them professional liability.  No one has to justify their choice to follow convention or policy.  No one can be held accountable for regressing to the norm, or so it would seem.

Thus, if things are stagnating as a result of an administrative swarming around invention policies, the thing that is necessary is to document what the consequences of that swarming have been, and find ways to break up the policy monoculture so that there are more degrees of freedom, variety, if not a little chaos, which appears to be necessary both for life and for innovation.   One could start by imagining a university with *no patent policy whatsoever*.  It is easy if you try.

This entry was posted in History, Policy. Bookmark the permalink.