University of California, Davis, School of Law
In this talk, I would like to address three topics. First, I would like to present some ongoing research on the use of “private ordering” mechanisms to broaden access to the fruits of publicly-sponsored science. I would then like to apply some lessons from this practice to the challenge of designing a microbial research commons. Finally, I would like to explore some additional principles and considerations for constructing such a commons.
Let me begin by providing some context. Much of my recent research focuses on the role of “public institutions” in creating a noncommercial research commons in biomedicine. In particular, it focuses on efforts by government, academic, and nonprofit entities to enhance access to patented biomedical research tools. Obviously, enhancing access to intellectual property is very different from enhancing access to physical resources, such as microbes, that are ordinarily subject to material transfer agreements (MTAs). However, there are some striking similarities between these two scenarios and some lessons to be learned. With that in mind, allow me to turn in greater detail to my first topic, the use of public norms and private ordering to create a biomedical research commons.
The problem that these efforts address is familiar to many, and it has to do with proprietary claims on scientific inputs operating to inhibit valuable research. I focus on patented biomedical research tools, which can encompass anything from extracted and purified human embryonic stem cells to genetically modified organisms, including genetically modified microbes. The relevant theory here—which, of course, must be empirically verified—is that patents on research tools constrain access to these resources, thus inhibiting basic scientific research and the development of valuable technologies and industrial applications.
The present tendency of parties to patent research tools gives rise to the challenge of how to provide appropriate access to such resources. In exploring this challenge, let us start with some first principles. Wide access to research tools enables more parties to conduct scientific research, thus generating significant positive externalities. In general, society is better off when scientists have ready access to resources like human embryonic stem cells and microbes that are critical to conducting basic research.
This state of affairs suggests a particular set of policy responses. Given the desirability of maintaining wide access to research tools, perhaps we should simply commit them to the public domain by prohibiting parties from patenting them. This, however, would be problematic for several reasons. First, as commentators have noted, many research tools are actually dual-status inventions: research tools, such as extracted and purified human embryonic stem cells, are often precursors to commercial products in addition to being valuable enablers of basic research. Policy considerations thus weigh in
26 Presentation slides available at http://sites.nationalacademies.org/xpedio/idcplg?IdcService=GET_FILE&dDocName=PGA_053668&RevisionSelectionMethod=Latest.
favor of granting exclusive rights on these research tools, even publicly-funded research tools, to encourage their further development and commercialization. Indeed, this was in large part the intuition underlying the enactment of the Bayh-Dole Act, which allows federal grant recipients to take title to patents arising from taxpayer-funded research.
Consequently, we do not want to categorically commit research tools to the public domain. Rather, it would be preferable to create a mediated semicommons where such resources are subject to context-specific access and exclusivity. Unfortunately, traditional policy levers such as legislation, judge-made common law, and regulation have proven inadequate in this regard. In large part, these instruments are too blunt and cumbersome to facilitate mediated access to research resources on a case-by-case basis.
However, where the law fails to provide for optimal resource management, interested parties often resort to private ordering. We see this in the real property context where communal management of environmental resources, such as fisheries, has proven highly efficient. We also see this in the intellectual property context. In the copyright sphere, for example, collective-rights organizations such as ASCAP and BMI have created pools of licenses that allow interested parties to access performance rights for proprietary musical works.
In the technology sector, private ordering has taken the form of massive cross-licensing and patent pools in patent-intensive industries, such as software and telecommunications. In addition, we see similar attempts to use private law mechanisms to enhance access to protected content in the emergence of open-source licensing, such as the General Public License for software and Creative Commons licenses for a variety of copyrighted works. In all of these contexts, norms and contracts are driving enhanced access to proprietary material. Furthermore, “private ordering” is not simply the domain of private institutions; public institutions are also fruitfully engaged in private regulation. This is particularly apparent in the life sciences sector, a field in which public institutions enjoy enormous leverage.
Take, for example, the National Institutes of Health (NIH), which provides about $30 billion per year in funds for biomedical research. As noted, under the Bayh-Dole Act, federal grantees can take title to patents arising from taxpayer-financed research. Some have criticized this law as providing a “double subsidy” to grantees, who receive both taxpayer funds as well as patent rights. That being said, these public funds come with certain strings and expectations attached.
Along these lines, in 1999, NIH issued principles and guidelines for obtaining and disseminating NIH-funded biological resources. There are two principles in particular that I think are very relevant. First, the NIH guidelines advocated the wide availability of NIH-funded, grantee-patented research tools for noncommercial uses. At the same time, however, these guidelines allowed for targeted exclusivity of such resources for commercial development. While the Bayh-Dole Act complicates and arguably prohibits direct enforcement of these guidelines, NIH considers compliance with its principles and guidelines in reviewing grant proposals and awarding research funds.
The model that arises is one where NIH provides some sort of consideration, in this case money, to a downstream resource developer, and in return that downstream developer is expected to provide qualified access to proprietary resources for research purposes. That access applies not only to NIH scientists, but extends on behalf of NIH to the wider research community as well.
Of course, NIH is not the only game in town. I teach in California, and there the California Institute for Regenerative Medicine (CIRM) will provide about $3 billion over
ten years for human embryonic stem cell research. And here again, public funds are embedded in a quid pro quo. As with the federal Bayh-Dole Act, recipients of CIRM funds can patent the results of publicly-financed research. However, CIRM imposes certain access requirements for state-funded research resources. Grant recipients must agree to make their patented, CIRM-funded inventions readily accessible on reasonable terms to other California organizations for use in noncommercial research. In addition, grantees must make CIRM-funded materials described in scientific publications widely available for research purposes. These access requirements are embedded in the terms of funding agreements and are legally enforceable.
CIRM places strings on public money in another way as well. Unlike NIH, CIRM collects royalties on inventions arising from public funding. These royalties are deposited into the state’s general fund and can be used for a wide variety of public expenditures. If multiple funding sources contributed to an invention, the state is entitled to a share that is “proportionate to the support provided by CIRM for the discovery of the invention.”27
In addition to the federal government and state governments, universities are also major contributors to biomedical research. Increasingly, universities are using contracts—specifically, technology transfer licenses—to enhance access to patented research tools. For example, here is some boilerplate language from an exclusive license at Harvard University: “Harvard will retain the right, for itself and other not-for-profit research organizations, to practice the subject matter of the patent rights for internal research, teaching and other educational purposes.”28 So, when Harvard exclusively licenses out some patented invention, it retains the right to use that invention for research purposes and to allow other nonprofit institutions to engage in similar activities.
In addition to demonstrating how licenses can enhance access to proprietary resources, university practice can inform the design of a microbial commons in other ways as well. As a general matter, resource owners tend to systematically overvalue their assets. We see this in the physical property realm, and we also see such overvaluation in the intellectual property context. Based in part on the expectation of generating significant revenues, universities have dramatically increased their patenting activities over the past several decades. However, university intellectual property actually generates relatively little income. Income from licensing is largely hit or miss, and it is overwhelmingly miss. In fact, a large proportion of university technology transfer offices lose money. It appears that universities systematically overvalue their patents, a phenomenon that is likely to apply to other owners of research resources as well.
From the perspective of designing a microbial commons, universities are also informative in that they demonstrate a high degree of normative plurality. Norms can diverge widely among seemingly similar institutions, as illustrated in various universities’ approaches to technology transfer. For example, Columbia University and the University of California have been very aggressive in patenting and licensing, while other leading research universities, such as Johns Hopkins, have been much less active in this realm. In addition to differences between institutions, we also see important normative plurality within institutions. In general, research faculties and individual scientists tend to be quite committed to the norms of open science. However, this
27 CAL. CODE REGS. tit. 17, § 100308(c).
28 Harvard University Office of Technology Development, Licensing Harvard Patent Rights: A Guideline to the Essentials of Harvard’s License Agreements, available at http://www.techtransfer.harvard.edu/resources/guidelines/license.
normative orientation may diverge sharply from that of senior university executives and technology transfer offices, which may place greater emphasis on asserting patent rights and maximizing licensing income.
Returning to our survey of “public institutions,” nonprofit funding organizations are also significant sources of scientific venture capital. One very important player is the Howard Hughes Medical Institute (HHMI), which provides approximately $600 million per year in research funding. Consistent with the theme of placing strings on research funding, HHMI “expects all HHMI research tools to be made available to the scientific research community on reasonable terms and in a manner that enhances their widespread availability.”29 This policy is consistent with NIH’s principles and guidelines on sharing biomedical research resources.
In these examples from government, academia, and the nonprofit sector, we see attempts to formalize the informal norm of open science. Public institutions are using quid pro quos and contracts to leverage their provision of “upstream” scientific capital to ensure access to “downstream” proprietary research assets. In this manner, public institutions are utilizing private ordering to create a biomedical research commons. Drawing from these case studies, there are several lessons that can be applied to designing a commons for microbial resources. First, it may be useful to consider an integrated approach that provides access to both intellectual and physical property. The first generation of the microbial research commons will likely focus on enhancing access to physical resources, namely microbes themselves. However, in future iterations, it may be useful to include patents related to these resources within the commons. After all, scientists often have to clear patent rights in addition to obtaining physical materials in order to conduct valuable research.
The next lesson for designing a microbial commons relates to how parties go about formalizing informal norms. Here, issues of authority and process are critical. For instance, who speaks for the scientific community, and how does a community arrive at normative consensus? Here, the scientific community can learn valuable lessons from NIH’s development of its principles and guidelines. These principles and guidelines arose through a consultative process that included formal notice and comment, which may serve as a helpful model for designing policies governing a microbial research commons.
Within this endeavor, institutional buy-in will obviously be important. In achieving such buy-in, it is crucial to consider normative plurality, that is, the fact that different institutions, such as various culture collections, may have different norms. And there might be normative plurality even within particular institutions, as we see in various universities.
So, how can we encourage participation in the microbial research commons? As noted earlier, the experience of universities suggests that parties systematically overvalue their assets, which may discourage them from contributing their resources to a commons. One way to discipline this tendency is to leverage upstream, scientific capital to mandate or encourage compliance with access objectives. Many public institutions enjoy significant leverage in this field, such as government agencies, universities, private foundations, and even scientific journals. One promising model would involve public institutions providing some sort of material support (or, in the case of journals, publishing
opportunities) to downstream parties, who must in return provide access to proprietary resources to the research community
Instructive in this regard, there is an existing track history of calibrating property regimes to accommodate dual purpose assets—assets that are both directly used in noncommercial research and have high potential for commercial application. Past experience shows that a single framework can accommodate these different uses. The potential for commercial exploitation does, however, give rise to difficulties regarding apportioning upstream versus downstream contributions in calculating appropriate compensation. Consider the following hypothetical: party A transfers some material to party B, who finds a way to commercialize it. What is the obligation of B to compensate A, and how is this complicated when parties C and D also contributed to commercializing this product? As we have seen, CIRM has already addressed this problem when determining royalties owed to the state of California where CIRM is only one of several funding sources for some commercialized product. I suggest that the proportionality principle used by CIRM can be a useful guide to allocating compensation when one party commercializes assets contributed by another.
In designing a microbial commons, institutional considerations are of course paramount. In any type of initiative encompassing a broad array of institutions and practices, centralized coordination and standardization can play a valuable role in lowering information and transaction costs. In the patenting and licensing sphere, NIH has played a central role in standardizing the ways in which various entities promote access to patented biomedical research tools. Is there an analogous body in the microbial research commons area, and to what extent are these challenges exacerbated because of the global dimensions of this initiative?
Finally, I want to explore some additional design principles and considerations for constructing a microbial research commons. In doing so, I have loosely modeled my observations on the objectives outlined in the draft monograph by Reichman, Dedeurwaerdere, and Uhlir. First, how can we shore up incentives to share proprietary resources? In terms of benefits, it is important to emphasize communal norms as well as the language of self-interest. Generally speaking, nobody likes giving up his property for free. In some cases, however, such behavior can actually advance one’s self-interest. In this respect, patent pools in the software industry represent an influential model, as firms are voluntarily donating their patents to a common pool because it enhances innovation for everyone.
Second, it is useful for parties to assess realistically the value of their microbial assets. As we have seen, parties systematically overvalue their assets. However, rational expectations concerning the profitability of these resources may encourage greater participation in the sharing system.
Within any attempt to build a microbial commons, distributing duplicates and derivative resources will be very important. In many ways, this need to broaden access to “downstream” iterations of microbes distinguishes this initiative from traditional MTA-mediated exchanges. Here, open source and viral licenses that ensure access to derivative works can provide a useful model.
Maintaining high-quality standards for microbial resources will require dealing with a variety of governance issues, which I understand will be the subject of a later panel. At this time, however, let me just say that there is a strong need for institutional leadership and community consensus to define standards, assign monitoring
responsibilities, and delineate sanctions. There may be organizations that are already well-suited to provide these functions.
Reputational benefits are critical to the normative economy and, in some sense, provide a solution to the incentives question. Why should parties voluntarily contribute proprietary assets to a commons? One answer is that participation makes it possible to establish priority, increase citations, and obtain communal recognition.
Along these lines, it is interesting to consider reputation as both a carrot and a stick. In many ways, the effort to build a microbial research commons reflects an attempt to create a community of strangers. Creating such a community gives rise to the challenge of cultivating trust and identifying bad actors. Here, the literature on peer production can offer a useful guide. User-generated ratings and “distributed accreditation” offer a new twist on peer review. Consider, for example, eBay, another community of strangers. How do I know that I can trust another party on eBay? The answer is that everybody on eBay is rated by everybody else.
Finally, I would like to address the challenge of securing equitable compensation in a commons. If I contribute some asset to a commons that another party eventually commercializes, how much compensation, if any, should I receive? I am a big advocate of liability rules, and I have argued in favor of them in other contexts. Such regimes, however, give rise to a number of challenges related to valuation and apportionment. Who determines equitable compensation, and how exactly is a reasonable royalty calculated? While considering these questions, it is also important to consider eliminating or mitigating the threat of exit. Is it going to be possible for an entity to simply opt out of the system once a product becomes highly profitable? Perhaps the party will think it can do better than equitable compensation and choose to leave rather than commit that resource to a commons.
Along these lines, it is useful to look forward, beyond merely sharing physical materials, to consider commercialization and patenting. In this regard, it may be helpful to have structured rules whereby parties who obtain resources from the commons must inform those who provided them of plans to patent any related inventions. Among other functions, such notice might initiate discussions regarding co-inventorship and co-assignment of inventions.
To summarize, I believe the microbial research commons is eminently feasible, and I have suggested that it can benefit from related experiences to create a biomedical research commons in the intellectual property context. We already have relevant experience dealing with multiple-purpose assets that are useful in both basic research and commercial applications. In designing the microbial commons, it would be helpful to integrate access to all relevant intellectual, physical, and informational resources. Within this project, we can leverage public norms and private ordering to formalize an informal ethos of scientific sharing. Finally, there are, of course, a host of institutional challenges regarding standardization, coordination, and determination of equitable compensation. If these institutional hurdles can be overcome, the creation of a microbial commons promises to greatly streamline and enhance microbial research.
Question and Answer Session
PARTICIPANT: I would like a better understanding of your metaphor of the new public license. We are all trying to deal with the difficulty of intellectual property.
PROF. LEE: The basic idea is a contractual understanding that “derivative works” based on some asset will also be disseminated in an open-source manner. The analogy to software is not perfect, but essentially the concept is to have an embedded obligation that all future iterations of some asset obtained in the commons will also be widely available to others, particularly for noncommercial research use. To the extent that we want to encourage openness, I think that we can use viral mechanisms to facilitate that outcome.
PARTICIPANT: I also am curious about what CIRM’s experience has been with this obligation to negotiate reasonable terms ex post for access to materials. I am wondering with CIRM (a) if there is much history of how successful they have been with this process, and (b) whether they can be held confidential by the party.
PROF. LEE: There are actually two components to CIRM’s royalty hierarchy. One involves ex ante definitions. Here, there is an established regulatory framework for determining what the royalties will be, and they define certain blockbuster categories. For example, if a grant recipient has $500,000 or more in sales, that triggers a certain percentage of royalties, and then there are graduated categories going up from there. So, those royalties are based on an established, ex ante schedule. The difficult part is apportioning royalties when a CIRM grant recipient combines state money with funding from other sources. CIRM recognizes that there are often multiple inputs to a commercial product, and CIRM’s claim relative to other upstream funders tends to be determined on an ex post fashion based on proportionality analysis.
Regarding the other component of your question, there is not much track history here in apportioning royalties. CIRM was established relatively recently. It is doing a lot of grant work, but not a lot of commercial products have been produced yet to provide test cases for these rules.