There’s a new study out at Future Internet that looks at how Wikipedia’s norms have developed over the years. In “The Evolution of Wikipedia’s Norm Network,” Bradi Heaberlin and Simon DeDeo examine Wikipedia’s form of governance and find it to be “highly conservative”–pages involving norms were created early in the life of Wikipedia. As they put it, “Policy growth precedes population growth.” The organization of work fell into “five central clusters” of interest (and influence): Article Quality, Collaboration, Administrators, Formatting, and Content Policy. These neighborhoods over time developed separate norms and patterns of control. Looking at Wikipedia policy pages, Heaberlin and DeDeo find that “all of these core norms were created early in the system’s history….The earliest members of the community first defined and articulated its core norms.” The authors ask, “Why are there no normative revolutions on Wikipedia?” They sketch then a provisional answer (my emphasis):
Recent work [60] has suggested that early users later form an oligarchy that monopolizes power, subverts democratic control, and comes into increasing conflict with the larger collective. If this is true, the enduring centrality of their own interests in the norm network may be a source of power.
…
Much of Wikipedia’s network simply coordinates technical practices, such as article naming conventions. The most important norms, however, attempt to rationalize the system around universal concepts, such as neutrality, verifiability, consensus, and civility. An important insight comes from a theory of bureaucracy and institutionalized organization developed by Meyer and Rowan (1977 [41]). They propose that norms such as these can function as institutional myths that make the system appear legitimate and less ad hoc, by connecting it to a rational framework.
A growing community then tends to prefer early statements of norms, which then become a source of power for those that have created these norms. The norms become the common narratives, the “institutional myths” that turn ad hoc–make it up as you go–behaviors into reasoned conventions. “What I did spontaneously you should do as a matter of rule.” “You should follow my example of what I did, not my example of doing things without regard for precedents.” The authors argue that as a community ages, matures, grows, norming “myths” decouple into separate spheres of influence rather than engaging to deal with inconsistencies between the “myths.” In other words, rather than engaging in debate regarding claims made by rationalizing stories, members of the community withdraw to focus on practices that advance the particular rationalizing stories that they are invested in, that give them their status.
We should consider these findings in light of the development of university “technology transfer.” Early norming statements, especially those surrounding the formation of SUPA (later, AUTM) and the Bayh-Dole Act represented a fundamental shift in the norms that came along with university patent policies.
In the pre-Bayh-Dole era, patent administration was largely ad hoc–have a committee look at a reported invention, send it to Research Corporation or another invention management agent for review, and if the agent will take it on, have the inventors work it out with the agent. But one of the early norming statements associated with Bayh-Dole, but not in the law as passed, was that federal law mandated university ownership of faculty inventions. This was utterly untrue by the law. Bayh-Dole is remarkably silent on private ownership of inventions made with federal support. Bayh-Dole dictates what federal agencies can ask for, by default, with regard to patent rights, not who among private parties involved in any given research project should own rights to patentable inventions. But as a founding myth, the idea that federal law gave universities ownership of inventions made with federal support was attractive to university patent administrators. The law, then, fueled a revolution in norms, from ad hoc to (apparently) statutory practice.
The norming myths were rationalized in a variety of ways–that the federal government was ineffective in managing its own patent portfolio (shown by Eisenberg to be a flawed claim); that exclusive licenses are necessary to develop inventions into products (untrue in the general case and clearly not true for the early biotech inventions such as Cohen-Boyer, Axel, and Hall); that universities are better positioned than inventors to manage patents (true only in the comparison–universities are not necessarily better positioned than other patent management organizations to manage inventions); that the public has paid for these inventions and must be protected from greedy self-interest of inventors (Bayh-Dole actually requires the reverse logic, or the federal government should own all inventions made with federal support); that economic development happens primarily through patent licensing to speculative investors (history shows otherwise, for which start with Steven Johnson’s Where Good Ideas Come From and Michael Heller’s The Gridlock Economy).
But no matter–the norming myths of technology transfer are not intended to reason from observation to practice, but from established power to practice. That is, practice is constrained to follow the norms, rather than to follow opportunity that might disregard the norms. Thus, university administrators get into pissing matches when faculty inventors challenge their norming myths and take it as a direct challenge to their institutional authority. It is better to spend a million dollars on lawyers to preserve a status-making myth than to enter the uncertain world of all the other ways one might deal with a discovery, in the public interest. As one university tech transfer professional aptly put it (I paraphrase)–“you may be right that there are other ways to do things, but my career depends on me staying within the conventions of ‘best practices.'” The norming myths become more important than the practices possible. Or, another way, the norming myths dictate how any new discovery might be managed, even if there are other ways available for such management. One doesn’t readily stray from the norming myths, especially if doing so involves potential liability and calls into question one’s professional judgment.
Senior administrators like norming myths for similar reasons. If a norming myth about “technology transfer” is in place, then they, too, do not have to develop a personal judgment with regard to opportunities except in exceptional cases (ones involving a great deal of money or a powerful individual or company). Instead, use a system. Write procedures. Require people to follow the procedures, except be more forgiving if administrators fail to follow procedures than when inventors fail to follow procedures. For administrators, the procedures are more like guidelines (another norming myth) but for inventors procedures are laws, contracts, ethical commitments that are enforced to the tiniest technical point, pound-of-flesh style.
Like Wikipedia, university technology transfer appears to be “highly conservative”–that is, it conserves the norming myths that rationalize its legitimacy. It is *not* that university technology transfer is illegitimate (that’s a separate question), but that the norming myths are used to rationalize that legitimacy, and therefore also constrain practice, even when the result of the practice appears to work against the claimed intentions of the norming myth (things like encouraging innovation, contributing to the regional economy, developing federally supported inventions for public benefit). The norms that protect the dual monopoly system (university ownership, exclusive licensing) appear to work against the claimed intentions. As these norming myths show up in university patent policies and university presidents’ speeches, one can start to see the contradictions.
The seed that started the process of turning an ad hoc, freedom-premised, open-to-opportunity approach to research innovation into a norm-bound, status-preserving, force- opportunity-to-fit-policy approach was the Bayh-Dole Act. Or, more accurately, it was the misrepresentation that the Bayh-Dole Act had overturned existing ad hoc norms for invention management and imposed by federal law a policy of administration-first, rather than one of inventor-first or opportunity-first. Of these three approaches, administration first appears to be the least capable of supporting innovation. Sure, there are still “successes”–pointing to the fact that even in the most oppressive regimes, delight is still possible. Early on, universities followed inventor-first approaches. They had no patent policies. If a faculty member invented, that was a private matter, along with buying a car or farting. Frederick Cottrell created Research Corporation to help academic inventors with their patents. If a faculty member invented and assigned to a research sponsor, no matter–it was a private deal, like selling one’s car or stamp collection.University adminstrators were out of the picture. No costs, no overhead, no distraction, no liability.
Later, universities moved toward opportunity-first approaches. If a legitimate opportunity came up, then the university ought to provide resources and pay attention to management. In this approach, patentable inventions that an inventor chose to pursue were reported to a committee that looked at the equities of the situation and decided the extent to which the supply of university resources created a claim for reimbursement or a share of income or an ownership share. Now there was overhead, but it was focused only on those few inventions that appeared to matter. If a faculty member invented and blew the rights, no matter. If a faculty member chose to patent–then it was worth taking a look. The review was still ad hoc, even if the policy was uniform.
But the faux version of Bayh-Dole drove out the opportunity-first approach. According to the new norm, the federal government demanded that all inventions be documented, assigned, and disposed of by a university administration. Thus, the remarkable transformation of university “technology transfer” offices (meaning, transfer of invention information from inventors to external management agents, and especially Research Corporation) into “technology licensing offices” (owning and licensing patent and other rights directly from the university), and the crazy growth of these offices across a host of metrics–staff (AUTM membership growth), budget (10x to 100x growth), patents (from highly selective to indiscriminate), and licenses (redefined to be any transaction of $1,000 or more). These metrics mark not the success of technology change driven by university discovery, but rather the success of a revolution in norms that enfranchises a professional class of administrators to deal speculatively in inventions (and non-inventions, and whatever) made by university faculty (and students, and staff, and visitors, and collaborators, and contractors, and whomever).
We know there are other ways to work things. The University of Waterloo, in Canada, takes pride in not making claims on faculty inventions, and appears to be remarkably successful with its research and economic development programs. At the University of Washington, for over a decade we ran a software and digital media management program based on norms arising in copyright, academic publication, and open source software, and we generated 3x more income on 1/3 the total budget compared to the norm-conventional patent licensing group. Even early “big hit” transactions such as Cohen-Boyer and Google appear to have been the result of ad hoc opportunities rather than some systematic administrative attempt to make money on volume.
The University of Washington has just completed such an effort to systematically produce “unicorn” companies of the likes of Google, and blew $100M over six years, without apparently accomplishing any of its stated goals other than perhaps advertising its efforts broadly (and expensively), and certainly not transforming the financial model of the university with the riches gained from selling off equity in startup companies. The University of Washington administrators bought off on a new set of norms in burning their wad–that second-rate venture capitalists could create vast sums of wealth where the conventional norm of owning everything and shopping patents to industry on exclusive licenses failed. And yes, owning everything and shopping for industry monopoly partners does poorly. It’s just that the UW administrators couldn’t let go of the norm of owning everything–instead, they were attracted by a different way of handling patent rights. Rather than deal with industry directly, deal with speculators in startups. Make money when industry buys the startup, or when later stupider investors buy out the initial clever investors. Except this model does even worse, and as was show at Utah and Washington both, to be based largely on a “culture of lies and deception.” If the SEC monitored forward-looking statements at universities holding equity in startups, there would be fraud charges everywhere. Isn’t it odd that universities get a free pass to make all sorts of unfounded, unguarded claims about their patented inventions and startup companies, but Martha Stewart gets jail time for apparently acting on a bit of inside information? Perhaps that is the power of the controlling norming myths.
Competing with the prevailing conventions of university technology transfer is a huge challenge. The myths of technology transfer–that universities should own all faculty work, that universities in trying to make money for themselves by patent licensing as if by an invisible hand also will create prosperity for others, that making an administrative system out of invention management is the great work of society–control the decision-making narratives within universities. It’s not a matter of reason (one does not reason with myths). It’s not a matter of evidence (the universities collect and report only information that confirms the myths, and academics lacking access to other information merely rework what the universities report, to little effect). As with failed prophecies, when faced with adversity, university licensing administrators double down to evangelize their norming myths, transmogrifying practice as necessary to keep everything in a favorable light.
This, then, is a problem with freedom, even with freedom to innovate. As DeDeo puts it in an interview with Gizmodo, “But what happens when a tiny Thomas Jefferson Libertarian fantasy has to grow up?” Apparently, it creates norms, abstracts them, rationalizes the abstractions, creates clusters of interest, decouples the clusters, and, well, sort of nests in the clusters. What starts as ad hoc and free ends up formalized and controlled to preserve the status of the oligarchy that controls the norms. We see much the same thing in the growth of standards in technology, from opportunistically ad hoc to oppressively formal.
Similarly, in business regulation. Various studies (see here, for instance) argue that the more red tape there is for businesses, the less developed the country, the less favorable the business environment. A common aphorism is that to start a business means to break the law–even in “developed” countries, the regulations are so fast and furious that no one can keep up with them all, and in some cases only big companies have the resources to put on a serious effort to comply (or appear to comply) and thus regulations themselves become a convenient barrier to entry for new companies and practices. Regulations themselves become a barrier to innovation except innovation undertaken by those companies that already dominate their markets. A representative for the World Bank told the Wall Street Journal that the World Bank’s index for business climate was “a very good proxy for corruption“–the less favorable the business environment, the greater the corruption.
What may come off as corruption might also be characterized as adherence to norm-myths that focus first on maintaining one’s power, advantage, status within a community. Corruption in this view is simply a gulf from one set of norms–say, the free association of people to form a company–and another set of norms–such as following the rules by which various administrators of company formation gain their standing.
In discussing the history of literary style in Mimesis, Eric Auerbach argued for a cycle of classical simplicity that then built in elaboration and sophistication to a point of “mixed” or muddled complexity and style, to be swept aside by a new wave of classical clarity. Perhaps in this, too, there is Schumpeter’s great hope for capitalism, of “creative destruction”–that capitalism might so impair reigning capitalists through technology change that there will ever be new opportunities for new capitalists-in-the-making without having to suffer an actual peasant’s revolt that burns through all those sweet assets rather than live a life of continual humiliation, proverty, and disenfranchisement.
A prevailing myth-norm of university innovation is that any peasant faculty member might become, through participation in a formal program of administration, an entrepreneur and make millions (sharing most of those millions, of course, per policy with the university administration). What an upside down world. It used to be that university faculty might be thought of as knights errant, seeking adventure on behalf of all citizens. Now faculty are peasants, and it is the entrepreneur who is the knight errant. Technology transfer’s contribution is that administrators by owning patent rights and seeking to create speculative betting parlors for wealthy investors can turn faculty peasants into entrepreneurial knights. It’s a goofy myth, but there it is.
For freedom to innovate, the challenge is not so much to make something free of all regulation, to grow up inevitably once again into another oligarchy of norms; rather, the challenge is to limit to some extent the control by the state, and by the university-institution, over those people who take on the role of faculty. Bayh-Dole, in its own way, was to do this relative to the relation between the government and universities and their foundations hosting research and managing patent rights.
In the US, we have never had a federal invent-for-hire statute. Instead, it’s been a matter of common law, of agreements regarding invention that are not implied by employment itself. An employer does not have a right to the inventions of an employee even if the employer’s resources have been used, unless there’s an agreement about such things. Similarly, when the federal government contracts for research, it can as a condition of the deal, require patent rights as a deliverable. That is, before Bayh-Dole, which expressly limits what a government agency can ask for in a funding agreement with a nonprofit or a US small business.
But that funding relationship is only half the deal, since the federal government provides funding to universities, at least, on behalf of individual investigators, for which the university provides (and is paid for) the research infrastructure (buildings, power, janitors, coffee) that supports these individuals. These agreements, then, are ones involving special performance. This second relationship, then, is the one between the host university, which is paid by the federal government to provide support to an investigator and is also paid to release the investigator from duties to the university so the investigator may work on the government-supported project (and be paid by the government for doing so).
If it made sense for Bayh-Dole to restrict the claims of the federal government with regard to university interest in inventive work, why is it that university administrators do not also see that it makes sense–at least in parallel–to limit the claims made by university administrators on the work done by the faculty. If the genius of Bayh-Dole is that federal agencies cannot as a condition of funding demand ownership of inventions made with federal support, why should university administrators then come to enjoy just this same status, to demand ownership of inventions made with federal support?
If the norming myth is that university administrators can sell off public property to monopolist speculators with less worries than federal agency administrators can, then it all makes sense. But if the federal government never cared to sell off inventions to monopolists, then the comparison fails. Do university adminstrators do a better job than the federal agencies in creating technology commons? standards? interoperability? open competition? It does not appear so. Even if technology change only came about because there were monopolies (Peter Thiel likes ’em), it does not follow that by creating hundreds of petty monopolies around research discoveries, there will be even more technology change. These monopolies, apparently, are not the monopolies that we are looking for.
It is not merely the fact of monopoly but the context in which the monopoly comes about. A competing market of monopolies on discovery is gridlock, as the nascent aircraft industry in the United States discovered in the 1920s, and which hit nanotech (thanks in large part to university patenting) in the 1990s. The fusses over cell phone patents is largely one of dealing with patented technology incorporated into standards, with cross-licensing to create defacto standards, and with trolls, who simply game the system. It’s a kind of gridlock using the courts to decide who gets a share of the profits.
There are other norms for university involvement in research and discovery. One starting point to explore these norms is history. That shows that opportunity-first approaches were more effective than administration-first approaches. Another starting point is a bill of rights to limit the grasp of institutions. Bayh-Dole does this for federal agencies. States should do the same for universities, at least with regard to federal funding. The principle is the same: you pay to support inventive work, not to acquire the patent on it. Or, if you pay to acquire the patent, too, then you expect to make the invention available broadly to all, or demand that whoever licenses the patent exclusively makes the invention available broadly to all (and on reasonable terms, including price), without excluding direct practice of the invention by those capable of doing so. Thus, for disease assays, perhaps a monopoly could be granted on making a commercial version of the assay, but that monopoly should never prevent any lab medicine department or doctor from performing the same assay directly, without the need for purchasing either the commercial product or a patent license.
Giving inventors complete freedom is a useful–call it classical–first step. But like Wikipedia, libertarian freedom may unwind over time into a conservative hierarchy. Perhaps we should plan for this state of affairs and simply sunset university patent policies every fifty years or so, and see what new things develop, freed of bureaucrats and their self-love. For universities, making a change to return to freedom in patent policy is simple–though, given the norming myths, not apparently easy. All it takes is to remove the wording–usually one poorly drafted, badly interpreted sentence–that claims that all inventions (as defined by policy) are owned by the university. Replace that with a claim only on patentable inventions that the university is required to own (by contract, say) or as may be mutually decided by inventors and administrators. (Or, just re-adopt a patent policy from, say, 1960.) That is, the university is authorized to own patents, but does not require inventors to assign unless they have entered into funding or other agreements under which they have agreed to assign.
Then let’s see what new norming myths get created, and what happens in the free space opened up before the next oligarchy of administrative norms comes to dominate practice.