In “Overplaying the innovation card: The stronger intellectual property rights and competition law,” Steven Anderman makes a distinction between “initial inventor rights” and the rights of “cumulative” innovation. Anderman argues that intellectual property laws must balance these two sets of rights. Here’s a key passage:
There’s quite a bit going on in this passage. First we see the distinction between initial invention and cumulative invention. This distinction frames differences in incentives. If a first inventor finds an incentive in the patent system to publish an invention in exchange for exclusive rights, we have to acknowledge that these exclusive rights may in turn prevent others, for some time, from either using the invention or improving on it. The initial inventor might make a product, even, and sell that product, and still prevent others from using the invention in other ways, for other applications, even ones that do not compete with the product.
If we consider invention to be a sparsely populated set–so that inventions are few and far between–then initial inventions are pretty much all there is. Invent, have a couple of decades to exploit, and release to the winds. But if invention is densely populated, so that when there’s one invention, there may be many others right beside–variations on the theme, alternative ways to do, applications that are similar but distinct, improvements and extensions, follow-on developments, and supporting developments. That is, the potential for cumulative invention is great. But an initial invention–some random dart in the middle of all this potential–can stink up the entire set.
One might consider the idea of claiming land in an age of exploration. If one sets foot on some new shore and plants a flag, just how much of the land connected to that flagpole is encumbered by the claim? What’s the explorer’s contribution? Does the explorer claim an unknown continent just for having washed up first? Or does the explorer claim the exclusive right to explore for the next twenty years and get all the land that’s explored during that time? Or does the explorer have no more right to adjacent lands than any other explorer, and gets only what he or she is able to survey and describe, as for a fort or town or patent application? Finders may be keepers, but does finding extend to keeping everything else that might be found, and to the exclusive right to keep finding and keeping? We might think that the explorer’s contribution is to open up a new land for exploration and settlement, and if the explorer then may bar all others from exploring or settling, the explorer has overstepped the contribution and now becomes the despot of the land, the sun king. Le terre c’est moi.
We might then consider not only an initial invention but also the activity that this initial invention evokes or, having considered the possibilities, what activity a patent on the invention induces beyond that of the invention itself, and what activity a patent excludes. Beyond the fact that a patent publishes an invention, a patent owner might do nothing with the patent–not bother with it. In such a case, the invention with its publication evokes what it will. Deciding to use the patent, however, is where management choices come into play.
Patent law lays some groundwork for management. In the U.S., unlike many countries, there is no working requirement. A patent owner does not have to use the underlying invention or license it to others for use. That’s a strange thing, but there it is. Copyright is the same way. An owner of a copyright does not have read the protected manuscript or publish it or make copies of it or distribute the work in copies. (Trademark is different. Failure to use a trademark can cause the trademark to lapse.)
The challenge for federal funding of research, whether that research is conducted by federal employees or by contractors to the federal government, is what further management guidance is needed for patents on inventions made with federal support. We can distinguish two situations–procurement and subvention. In procurement, the government funds research to obtain the results as deliverables, whether from its own employees or from the employees of contractors. In subvention, however, the rationale for government funding is that whatever is being done is on its own good for the public.
There’s much, then, packed into “on its own.” That “on its own” tends to include work proposed by university faculty, to take place at universities established with a public mission, with a purpose to serve the public good by publishing new knowledge and by training people to do research and evaluate research claims. Subvention research funding has not generally been seen as appropriate for commercial concerns, where there is a profit motive, even when the results of that profit-generated work might also have a public benefit. Subvention funding isn’t a subsidy for profit-seeking companies and their shareholders. “On its own,” then carries important intents. Patent management guidance for subvention funding follows the expectations of research that is proposed “on its own” to be in the public interest.
This is where there’s a tension between the defaults of patent law focused on an initial inventor and the expectations of follow-on or cumulative development. If the initial inventor is embedded in a federally supported subvention research project, then the “on its own” applies just as well to any invention made in that project. Federal grant regulations–which form the funding agreement with universities for subvention support–make this responsibility express:
Real property, equipment, and intangible property, that are acquired or improved with a Federal award must be held in trust by the non-Federal entity as trustee for the beneficiaries of the project or program under which the property was acquired or improved.
That’s 2 CFR 200.316, where intangible property includes patents and patent applications. Whatever a university acquires by way of patents and patent applications from such subvention work is held in trust. To be a trustee of patents for the beneficiaries of a project is not to be any ordinary patent owner. Over the years, the federal government has sought to articulate in specific terms what it means to be a trustee of patent rights rather than to be the mere owner of patent rights. The general drift has been that patent rights should not prevent the use of an invention, ought to be used to make an invention available to all, and should be used selectively to attract investment capital when no use or development will take place without investment and no one is ready to invest without some chance of recovering that investment–so, where the cost of investment is high, the ease of copying once an invention has been developed is low, and the profitability of the developed invention makes it attractive to copy.
In such circumstances, federal policy has recognized that exclusive patent positions in subvention funded inventions is a reasonable practice, but with restrictions. But even here, federal guidance has been to allow exclusive positions only for limited times and for contractors without commercial positions or capabilities, only after non-exclusive efforts have failed and an invention still isn’t being used or developed. For exclusive positions, commercial patent owners had three years from the date of patent issue, or about six years overall, to get an invention into use. Nonprofits, under the IPA program, could license exclusively for three years from the date of first commercial sale or eight years overall. In the original Bayh-Dole, exclusive licenses to non-small companies could run only five years from the date of first commercial sale or eight years overall (with extensions for delays in regulatory approvals). Bayh-Dole a few years after passage was amended to eliminate this term restriction on exclusive licenses. Even so, Bayh-Dole’s statement of policy and objective, combined with various requirements specific to nonprofits, make it clear that federal policy expects that the patent system on its own is inadequate to the management of inventions made in subvention research, especially at nonprofits.
The idea, in a nutshell, is that subvention inventions should be available to all, and only in that context might exclusive rights be reserved, where there is otherwise no uptake, to deal with commercial investment–to use or to sell for commercial purposes–as a means to stimulate development. Such exclusive positions come after availability, not before availability and not instead of availability. Where the term of an exclusive license is limited, the idea is–if an investment cannot be realized within the term–six to eight years or so–then the use of the patent system is not worth it and the patent monopoly should be broken apart through non-exclusive license (preferably royalty-free) or by dedicating the invention to the public domain (if a patent hasn’t issued). These gestures, if done first and regularly and consistently, open up the possibility for cumulative development. Present practice under Bayh-Dole is to block cumulative development–in part because the other institutions involved in such cumulative development also are prepared to block cumulative development and attempt to capture all proceeds from exclusive positions for themselves.
That’s the case in the present dispute between the University of California and the Broad Institute and MIT over CRISPR. “We wouldn’t have to be jerks if you weren’t being jerks”–except the reality is “We both intend to be jerks and this is what jerks have to do when there are other jerks involved.” By “jerk” here I mean, technically, institutionally ready to establish an exclusive position that precludes cumulative development and fails to recognize the development on which the position is based–too bad for those patsies. By “patsy” I mean, technically, those institutions and individuals that choose to dedicate their work to the public and to science and so encourage cumulative development.
We might then consider whether a missing element in Bayh-Dole is recognition of cumulative development. That most subvention research requires access to prior work, and the outputs of that research must be readily available to others working in the same area, and for applications of the work, and for lateral shifts to other areas of work. These dynamics cannot happen if every institution stakes out patent claims on every possible invention made with federal subvention support and intends to make these patents available only for exclusive positions. The effort fails even if institutions delay immediately making inventions available openly–if on day 0 an invention is not available for cumulative development, then the damage is already mostly done.
Anderman points out in his article that the expansion of IP rights for both patents and copyrights also works against cumulative development–what Steven Johnson calls “networked, non-market” dynamics. Anderman calls for legislation to restore the balance between initial inventor rights and follow-on activity. Competition, then, is given renewed emphasis. If institutions can own patents on almost any bit of research finding that’s new, useful, and non-obvious, and hold those patents for “monetization”–especially by preserving the monopoly position in the patent against all possible uses–then one creates a FOIL environment–fragmented ownership, institutionally licensed technologies. Unless something is done to prevent fragmentation and at the same time to limit the ability of posi owners to extend their patent monopolies, it’s impossible to have competition form around new inventions and discoveries. The only competition that forms is one of speculative acquisition–tying up even more cumulative developments in exclusive positions offered for “monetization” (often referred to as “commercialization”–but meaning, in effect, dealing in speculative exchanges involving monopoly rights).
Unlike the general case, in which the patent system might be construed as an incentive to invent and to publish the invention in exchange for exclusive rights, in subvention research, the controlling purpose is neither. One invents or discovers as a consequence of a commitment to contribute to public welfare, and one publishes as a consequence of the commitment made to work in a university setting, where publication is one of the expectations of the appointment. Thus neither profit nor publication is a necessary component of the bargain involving posi. Federal efforts to manage patent property rights and activities reflects the different bargain presented by posi.