Research Enterprise Policy Issues: fragmentation of noisy research

We have looked at noisy research and quiet research. Policy folks don’t much care, but it appears to make a difference whether research is conducted quietly or noisily. In quiet research, variations are explored, applications considered, data assembled, evidence checked before anything is announced. In noisy research, topics and areas are announced in advance, with proposals submitted as a kind of competitive boast in the hope of obtaining funding, with many other researchers turning their attention to whatever is announced as available for funding. The result of noisy research is that it attracts participants–if that is where the funding is, then that is where proposals are directed, awards are granted, and new IP created.

If we add in that the organizations that host research insist on claiming ownership of results and furthermore intend to hold ownership as an intellectual property monopoly in the hope of finding an exclusive business partner to “develop” the results into commercial products (and pay for the right to do so), then the necessary result is that the area of research fragments into the ownership positions of the organizations whose researchers have all been attracted to the noisy research.

The upshot is that when ten or fifty or two hundred researchers and their organizations each think this way about approximately the same area of inquiry, the work itself fragments into many pieces, each controlled exclusively and independently of the others–tens, to fifties, to hundreds of little IP portals. If each IP position is held for exclusive licensing, then it rapidly becomes impossible for anyone to acquire sufficient licenses to exploit the research results being created–whether to use those results (as in research or professional activity) or to make and sell products based on those results.

Significant allocations of federal funding for research–noisy funding–increases the number of researchers working in a given area of inquiry, increases the number of organizations involved, and spreads the geographic distribution of projects. If IP in research results is held by the organizations that host this research, and those organizations insist on seeking exclusive licenses, then federal funding, as a matter of practice, contributes to the further fragmentation of IP. Rather than speeding up “commercialization” of results, federal funding delays use of results for twenty years wherever multiple pieces of IP are necessary to practice or apply a discovery and those pieces of IP are not readily available. No amount of effort to “commercialize” any one piece of IP by exclusive license will, on its own, bring about the circumstances under which all the necessary IP has been assembled and made available for use.

One can see, then, that in the case of multiple IP required for use, either nonprofit organizations obtaining ownership of pieces of that collective IP must make their pieces available non-exclusively or one company must take control of all of their pieces–in which case each piece of IP becomes worth about 1/(total pieces of IP needed for practice). If the total is 50 patents, then each nonprofit might expect a royalty of 1/50 of, say, 5% of sales–too small to bother with, in many cases. Even to achieve this result, ten or thirty or fifty organizations would all have to agree to license exclusively to this one company, would have to settle on mutually compatible licensing terms, and would have to get their deals done promptly. And then the company, loaded with all these patents, would have to execute on the development and deployment of product–for which it may need permission from other IP rights holders (such as IP rights in standards), further degrading the royalty positions of the nonprofit “commercializers.” There is still a chance–maybe one in 1,000–that the company could succeed in creating viable commercial product. Perhaps the more telling probability, however, is that it is an almost certainty that no one else, during this one in 1,000 effort, will have the legal right to use or develop this same package of inventions covered by the collective, fragmented IP positions held by the one exclusive licensee.

This result in practice comes about by the combination of multiple policies, each oblivious to the others:

Noisy research is held to be in the public interest, much preferred over research conducted in secret

Federal research money is to be spread around, so many organizations and many geographical regions are involved in work in any given area designated as important

Each organization involved is given the opportunity to acquire and hold exclusive rights in any research invention (or any other asset, once one steps outside Bayh-Dole)

Each organization announces as its goal to make money for itself and benefit the public through exclusive commercialization licenses

By contrast:

Quiet research might allow fewer researchers working longer and out of the spotlight to explore further and gather together results with a single point of control, limiting fragmentation of IP.

Federal research money could be allocated to concentrate work with a very few organizations, so even if the research were noisy, only those few organizations would gain the benefit of the federal money and could move more quickly than most other, less well funded organizations, to accomplish the research goals and consolidate IP positions.

Federal research money could come with the stipulation that all results must be placed in the public domain or if secured by IP positions must be licensed on FRAND (fair, reasonable, and non-discriminatory terms)–something that the Department of Education is attempting to do for copyright works created with its funding and something the NIH and NSF have tried to do for inventions by circumventing Bayh-Dole’s requirements as much as possible, respectively, for research tools and results produced by cooperative research centers. This also was a default position for the Public Health Service research under the Kennedy and Nixon patent policies, and is the default position for open source software licensing and open innovation initiatives.

Each organization involved in the noisy research could require non-exclusive licensing of its IP positions–if the research is noisy (such as, publicly funded and not classified), then licensing should necessarily be also noisy (such as FRAND).

Each organization involved in noisy research could set as its immediate goal making the use and development of the results of the research as noisy as the proposals and the reporting of results by teaching their new results and aiming to develop libraries, platforms, cumulative technologies, interoperable technologies, commons, and standards–in short, changing the desired intangible asset from the exclusive patent license to an audience of companies benefiting from instruction, delivery of data, easy access to rights, and the externalities of common usage.

Even one of these policy flips would address the problem of fragmentation in noisy research. Even one of these policy flips might improve the uptake of otherwise fragmented research results. Even one of these policy flips might create new intangible assets of much greater general value than the desired exclusive commercialization patent license–intangible assets such as a standard or a hundred companies all desiring to work with an organization’s research team.

The monopoly meme claims that without a patent monopoly, no one will use the results of research discovery. It’s a nonsense claim, on the face of it. But even if the monopoly meme were true in the case of a single invention covered by a single patent, the idea simply does not work when ten patents held by ten organizations must all be licensed before anyone can apply the research to a particular area of application, and worse yet if 50 patents are needed (as with, say, an ink-jet printer) or 300 hundred (as with, say, an automobile).

Bayh-Dole, which makes it difficult for federal agencies to take ownership of inventions for the purpose of making the inventions available to all, also therefore makes it much more likely that research funded by the federal government will be fragmented into tiny bits of IP controlled by a number of organizations, each attempting to make the monopoly meme work for itself and collectively making it all but impossible for anyone to make, use, or sell anything that requires access to multiple bits of related IP that has been withheld in the hope of finding an exclusive licensee.

If it were true that the purpose of Bayh-Dole is to give nonprofit administrators incentives to establish monopoly IP positions and trade these monopoly positions as exclusive licenses to commercialization partners, then Bayh-Dole in practice greatly enhances the fragmentation of areas of noisy research, makes commercialization–however much desired–all the more difficult, and effectively delays broad use for two decades or more.

One might almost think that the intended policy result of Bayh-Dole has been to use nonprofit dullwittedness about making money from patent licensing to create a strategic barrier to entry for research, professional, and commercial use of noisy research results, allowing the leaders of private sector commercial practice to decide when and how such results are to be taken up. In some few cases where only a handful of patents might define a new product–notably the discovery of classes of potential drugs–the industry leaders might agree to divide up their interests with nonprofits as chosen exclusive “partners.” For the rest, however, Bayh-Dole when combined with the claimed incentive of nonprofit commercialization, creates a tremendous fragmented IP barrier to uptake, use, development, and commercialization of noisy research results.

One might almost think that Bayh-Dole was intended to make it difficult for American companies to use federally supported research while making that research openly available to companies and governments in the rest of the world. If nonprofits patent their fragments of federally supported invention only in the United States, then those fragments are not freely available in the United States but are free to the rest of the world. Only the most well endowed (or reckless) nonprofits have sufficient funds to undertake global patenting–most have to first have an exclusive commercialization partner willing to pay the costs of international patenting, and for that we are back to a tiny, tiny subset of nonprofit-acquired inventions–those that are biomedical or are of crucial to venture capital-backed startups.

In the context of cumulative technology–forming a research platform (needed for continuing research in a given area of inquiry) or forming a library of functions or materials (useful in conducting all sorts of research, development, or production of product) or creating a standard (to ensure interoperability, common function, reliable testing and consistent training)–each exclusive license and each startup company with an exclusive license represents a barrier. As a university announces another startup–with the claim that the startup represents potential for new products and economic growth–the reality is that the startup represents the deepening fragmentation of a collective technology that now is much less likely to be used until the exclusively licensed patent expires. The university granting such an exclusive license may imagine the prospect for making money from speculation trading on the future value of its licensed patent, but any technology to which the patented invention is an important part likewise suffers serious damage.

Fragmentation is a condition of noisy research. If the federal government takes ownership of inventions made in noisy research and makes those inventions available without charge or formalities, then it may mitigate some effects of noisy research. If the federal government creates reciprocal commons (FRAND licenses extended to include a range of improvements), then it may mitigate additional effects of noisy research. Bayh-Dole, however, works against mitigating the effects of noisy research. Universities and small companies don’t have to invoke the monopoly meme; they don’t have to exploit patent positions to secure exclusive “commercial” partners to “develop” research discoveries.

Nonprofits could make things available FRAND or dedicate their research results to the public domain (even with Bayh-Dole, there are legal ways to do this). But if, in any given area of inquiry, even one organization defects on the overall effort, then other organizations are also induced to defect (in defense of their investments in IP, or in retaliation), and we end up back with unresolvable twenty-year fragmentation, with each organization clinging to its bit of IP, now almost impossible to license. The folks who would yet take an exclusive license are (i) speculators; (ii) companies with such market power that they can acquire or suppress other IP or withstand an infringement suit; and (iii) companies that want to prevent the IP from ending up exclusively with a competitor.

It doesn’t have to be this way. It is, however. And that’s the result of multiple policy positions being pursued as if there’s only one invention, one patent, one company, and one product to be produced, where the critical factor is that noisy research inventions get turned into patent monopolies. The practice reality is fragmentation, barrier, delay, nonuse, and irrelevancy. Bayh-Dole is not the cause–though it is a catalyst, an enabler, an opioid.

This entry was posted in Commons, Innovation, Policy and tagged , , , . Bookmark the permalink.