Bayh-Dole asks inventors in faculty-led, government supported research to choose an invention management agent. If the inventors don’t choose, then the government gets to choose.
Unlike practice in industry, university faculty had a range of options available to them when Bayh-Dole became law: in many universities, they could choose their own path. In some, they were constrained to use a contracted agent, such as Research Corporation. In fewer yet, there was an affiliated research foundation at hand, which would act as an invention management agent. In a very few, there was an in-house invention licensing operation. It was understood that the university approach–this diversity of practice, of agents, of options–was in some sense successful in selecting inventions to manage and placing them for development in industry. Certainly the effort by Stanford with the Cohen-Boyer inventions for gene splicing were smart, thoughtful, and very effective. It was clear that federally supported inventions made by university faculty and those working with them might benefit from this same approach. In this way, Bayh-Dole invited university inventors to use the choices available to them for subject inventions, as they did for other inventions.
Bayh-Dole amounts to a request by government for university inventors to use their broker model of invention management in furthering the impact of federally supported research. The uniform policy required of agencies by Bayh-Dole allows inventors a great deal of discretion in decisions with regard to the inventions to manage, and if they are managed, the methods to be used, the objectives to be accomplished. Under Bayh-Dole, the objective is practical application writ large: that may be new practice, new standards, new products, or new companies. The standard patent rights clause authorized by Bayh-Dole does not care. There is no mandate in Bayh-Dole for commercialization, or for startups, or for “economic development”; certainly there is nothing that demands inventions made in a state stay in that state, or that profit from such inventions be returned to that state, or to the host university, or that there be any profit at all, or that payment be required by those that would use or develop such inventions. Bayh-Dole simply requires that those choices be first made by other than the funding agencies, and if no choice is made, then the funding agency may have its own chance.
It is in this way that Bayh-Dole creates opportunity, and at the same time leaves university inventors exposed in their choices to the desires of those who would seek power over them for what they possess: inventions made in the course of their investigations, their scholarship. Bayh-Dole appears to recognize the problem. University and nonprofit contractors (meaning, at the time, primarily the research foundations tied to universities, some of which also contracted for and managed sponsored research agreements as well as inventions) were singled out in Bayh-Dole to bear the most restrictions of all–more than small business contractors, and more than faculty inventors on their own. These restrictions are set forth at 35 USC 202(c)(7) and are repeated at 37 CFR 401.14(a)(k). The universities cannot assign inventions to organizations without invention management functions except with agency approval; the universities must share royalties with inventors, as part of their expenses (and not, apparently as a mere perk of employment); the universities should give preference to small businesses in licensing; and they must use any remaining funds after expenses for scientific research or education.
While these restrictions have been championed as a federal gift to universities, each carries with it the implication that more so than small business contractors, or university faculty as inventors, the public has reason to be concerned about allowing universities to manage inventions unchecked by case-by-case agency oversight. Universities might simply sell off inventions so that they would not be managed, but merely be handled for the sale value of the patent. Universities might not share royalties with inventors, making a case that their own funding is more important than any personal recognition (and in this way disrespect faculty inventors).
a requirement that the balance of any royalties or income earned by the contractor with respect to subject inventions, after payment of expenses (including payments to inventors) incidental to the administration of subject inventions,
More subtly, however, the language of the royalty-sharing provision suggests that universities will not bargain for invention rights with faculty inventors, and so become determined not to treat royalty sharing as an expense, something of value paid in exchange for obtaining rights to an invention, but rather as a form of institutional generosity, whereby the obligation to assign inventions becomes an assertion from the advantage the university holds by being designated to manage federal funds on behalf of university investigators. Government grants are directed at individuals, not their institutions. If the institutions come to believe otherwise, then they will treat the federal funding as their own, and will represent the royalty sharing as generosity rather than something bargained for.
The drafting here is yet more thoughtful: “any” to modify royalties or income earned; the addition of “income earned” beyond royalties. If a royalty is any consideration for the grant of a patent license, the income earned has to do with other exploitation of an invention not tied to licensing: the sale of products directly, services based on exploitation of an invention, realized equity from a startup company. And a limitation on expenses to those “incidental to the administration of subject inventions”–to prevent a university from charging expenses from the management of other inventions, or non-inventions, to the royalties and income from its management of federally supported inventions. So much wrapped up in the language of the statute, evidencing the concerns Congress had over the potential for bad behavior among university licensing officials and administrators who come to the scent of money and authority and value that scent over practical application and benefits. The idea in the royalty sharing clause is that if there is going to be money made, it will go to expenses and to the inventors; if there is anything more, it will go to research and education, not to administrators, not lost as slush funds, not funneled back to feed empire building, even in technology transfer operations.
Finally, preference for small business, because, it would appear, universities might otherwise simply suck up to power or go where the money on offer might appear to be the greatest, but the opportunity the less, or the motivation not quite as keen.
It is in within this context of advancing federally supported inventions into an already active program of invention management created largely by faculty and on behalf of faculty, with protections against a range of adverse institutional behaviors, but not nearly enough, as we come to recognize, that Bayh-Dole introduces a standard patent rights clause and steps aside to see how faculty, universities, and their invention management agents behave. It is clear that from the start, the universities and invention management agents felt a keen need to appear successful. For this, they developed a story of the growth of the “industry”–of the number of patent applications filed and patents issued, the number of technology transfer offices, the licenses granted, the income. It is an understandable effort. But it does not tell the whole story, and the story it doesn’t tell is a great deal of the whole story, and at some point, the partial story seeks to displace and suppress what is not being told, so that policy makers come to believe the part stands for the whole, and provide more funding, and ignore pleas from the field, and do not diligently ask for reports of utilization, invention by invention, and in getting reports, do not audit them for accuracy.
In setting up these arrangements, Bayh-Dole makes a clear path by which a broker can hold invention rights assigned to it by a university inventor. This is the title “certainty” of the legislative discussions: will or will not the federal agency allow title that has been assigned by the inventors to remain where it is assigned, or will there be a big fuss, a lot of delay, and the like? Bayh-Dole resolves this by requiring agencies to leave well enough alone and save their administrative breath and expense, when the assignee is the university host, or is an organization that “has as one of its primary functions the management of inventions.” (Interestingly, there is no restriction on the assignment of foreign rights in subject inventions). It is easy to see, then, how one could think it expedient to conclude that Bayh-Dole intended for brokers to have rights, for universities to become those brokers, and that it was the will of the federal government that inventions made with federal funds were for institutional money-making adventures, to take the burden off taxpayers and governments everywhere to provide support directly to the universities.
The intended implication of this presumption, as it has been called, appears to be that universities should work to find ways to efficiently take ownership of all such inventions. That is, this presumption argues that federal policy is that faculty and others should never own what they invent, despite federal law on the matter. The presumption comes to be viewed as a gift by the government to university administrators; the universities then construe their ownership as a benefit to the inventors rather than an imposition; and to the public, an implication that inventors are incapable of managing their affairs, and internally, a determined effort to make this true by adding regulations and withholding resources so that an inventor in a university is truly at a disadvantage at all points. There is another implication, however, in the argument that universities will prosper from the federal gift of faculty inventions made with government support, and this one is deeper, more damaging, and not readily withdrawn: that if universities are so successful in obtaining funds from licensing, then they do not need to rely on the subsidies of taxpayers and governments.
Every dollar earned in royalties as institutional profit is another argument for cutting university budgets. In this way, the presumption, self-serving for impressing policy makers with the rightfulness of Bayh-Dole, has played a role in turning the public away from the need to fund universities. In making the appearance of money and licensing success, the presumption of ownership has damaged universities deeply–in the relationship between faculty and administration, in the integrity of the administration to maintain the integrity of the institution, in the connection of the faculty and students with industry, in the relationship of the administration with industry.
Bayh-Dole does not require the broker rhetoric, and it would be impossible to expect a law to prevent it, or any other rhetoric. The effort here is to consider, however, how the Act sets up the situation, and what policy makers might do, recognizing what has taken place. As I have pointed out in another essay, the Standard Patent Rights Clause (or SPRC), creates one contractor for certain (the university), requires the establishment of a second contractor on the condition of an invention (the inventor, through the (f)(2) agreement), and permits the creation of others by assignment, substitution of parties, and subcontracting. Of these, the ones that matter for our discussion are invention management agents and research subcontractors. The SPRC also sets up the expectation that if a broker is going to handle an invention, it is going to do so that the invention can be licensed, under patent, to promote practical application. For the SPRC, licensing is an artifact of ownership. For Bayh-Dole, there is no mandate that the licensing demand commercialization (the creation of new products for sale) or the payment of royalties. There is no claim in Bayh-Dole that payment provides a return on taxpayer investment.
As far as the SPRC is concerned, the choice of invention management agents is with the inventors unless they cede that choice to the university. It is the inventors that own their inventions. It is the inventors that the government has agreed to fund to pursue their research. It is the inventors that, by default, have the first chance to choose an invention management agent. It is this choice that has been attacked by the university technology transfer industry through the construction of a misleading reading of Bayh-Dole that construes the law to vest title to inventions with the universities, or in any of the variations that show the untruth of the presumption, that federal law requires that the inventors assign, or law mandates that universities get assignment, or that title vests when universities elect to retain title, or that inventors are prevented by law from assigning to any entity other than the university, or that the university has a right of first refusal. The industry cannot agree on just how the presumption actually comes to operate, though there is widespread agreement that somehow it surely must operate. And, in a way, it does operate, exactly as a presumption, backed up by assertions of authority, but not by law or federal contract.
Thus, it has come about that universities have declared administrators, not inventors, to be the choosers of invention management agents. Thus AUTM could immediately mock the Kauffman Foundation for suggesting that university inventors might have some choice in the invention management agent, even if the university had insisted on becoming the owner of faculty inventions. Harvard Business Review thought the idea was one of the ten best for 2010. AUTM did its best to prevent the idea from being considered by the Department of Commerce or being discussed in a meaningful way in the field. The rise in the university “technology transfer” industry corresponds with the ejection of the invention management agent from the university approach. The number of “technology transfer” offices at universities is a measure of the effectiveness of this effort. The university response to Bayh-Dole has not merely been to use the opportunity to claim ownership of faculty inventions made with federally funds, and to claim the endorsement of federal policy indicates they should claim all other faculty IP assets as well, to the limit of “traditional academic works.” The deeper effect, and perhaps the primary unstated motivation for Bayh-Dole, was to drive the national invention management agents out of university business and establish a provincial model of invention brokering, one in which the university names itself as the invention management agent and does so that it might have a greater share of any royalties or income earned from faculty-created (and initially owned) intellectual property–inventions, and anything else that can be packed into a policy statement, including “know how” and “data” and “software” and “digital media”.
Thus, one of Bayh-Dole’s primary effects has been as a weapon to disrupt the use by faculty of a range of possible invention management agents. A second of these effects has been to concentrate the emphasis in university exploitation on the value of the patent rather than the value of the invention. A third has been to draw university invention management away from efforts to support each invention to one of a portfolio held for the financial success of some few patents. This emphasis on financial success in turns leads universities increasingly to seek arrangements with speculators, often in the form of investors in startup companies, but sometimes merely showing up as agents of arbitrage–obtaining patents from the university for less than they can get through their connections–to justify infringement litigation as a means to obtain income, and to threaten such litigation in an effort to drive up the speculative value of the patents it holds. The creation of a a portfolio in this way leads to accumulation of inventions and other such assets, with a refusal to grant licenses at no cost or low cost, as doing so would undermine efforts to plump the value of inventions and constitute a “race to the bottom.”
Turning to the other form of contractor, we have the subcontractor. Oddly, here, the tradition of faculty choice is unsullied. It is the investigators that choose the subcontractors, not the university, and in the choice of subcontractors, the faculty investigators also choose how to divide the work, and thus how to distribute that work relative to, among other things, where inventions might be made, and under the auspices of what other universities or other organizations. Each subcontractor becomes a contractor, and as a contractor has independent standing with regard to inventions made in the subcontract, and by implementing the (f)(2) agreement as it should, it passes a portion of that standing, conditionally, to any future inventors. For all the tight-fistedness by which university administrators might demand assignment of all inventions made under each federal grant, they have no control over the division of labor by which faculty might assign the valuable inventive work contemplated under a federal grant to investigators in other organizations. It is rather absurd to consider the implications. A contractor cannot claim an interest in any inventions made by a subcontractor as a condition of the subcontract:
The subcontractor will retain all rights provided for the contractor in this clause, and the contractor will not, as part of the consideration for awarding the subcontract, obtain rights in the subcontractor’s subject inventions.
This provision is at SPRC (g)(1)–another section that is in the regulations authorized by Bayh-Dole, and hence operates in funding agreements, but is not to be found in Bayh-Dole itself (beyond the definition of funding agreement), where attorneys who prefer reading law to reading contracts will look for it in vain, or not knowing what they lack, will reason unmindful of it, as they have with SPRC (f) as well. The upshot is that in addition to the choice of invention management agent, the investigators who may become inventors also may divide the work so as to move, whether intentionally or not, rights to inventions to other contractors, who may operate with their own policies and arrangements, and who may be better suited to, say, entrepreneurial work. Co-inventing across subcontracts then becomes a simple and effective way for inventors to gain a choice of invention management organization, even when each universities involved insists under policy of naming itself as the sole invention management agent, to the exclusion of others. University officials have asserted in print that they would never consider managing inventions made at other institutions, and yet co-inventing across subcontracts is just one of a number of ways that they do so routinely, though often with much administrative pain in fussing over the management of such joint inventions.
Given the diversity of possible contractors under a given SPRC, and the possibilities for invention management agents to be chosen by inventors or universities, as they work out invention ownership, it is rather shocking to find that in the course of thirty years all this opportunity has been reduced to a monoculture that largely excludes choice and agents and proceeds directly to holding patent rights to preferentially license for commercialization.
We can show these tensions created by the SPRC and the architecture of Bayh-Dole with a diagram. It is a heuristic only, but perhaps it helps to sort out the relationships in play.
This first diagram shows part of the flow of the SPRC from the government to the initial contractor, the university. From there it moves via (f)(2) to all investigators in the project, conditional upon their becoming inventors. Further, however, is university practice that permits investigators to choose the division of labor involving subcontracts. Often the subcontractors are identified in the grant proposal. It is rare that a university administrator would refuse to allow a proposal to be submitted that identifies other organizations as subcontractors unless it involves a conflict of interest (such as an investigator proposing to subcontract to his or her own company). Thus, the subcontract distribution of the SPRC is shown with a green arrow, as is the assignment of inventions from inventors to the university-as-employer. Universities have moved to block any other distribution of the SPRC, whether by the university directly, or upon recommendation of a faculty principal investigator seeking to manage the entire project, or as might be requested subsequently by an inventor. Note that the flow of invention rights back to the federal agencies from inventors has largely been blocked (policies that permit university inventors to place inventions in the public domain do not permit them to choose the federal government as their invention management agent).
The loss of independent agents has had an additional effect on university licensing practices. Just as it has been difficult for federal agencies to grant licenses, so also universities, especially public ones that operate as instruments of state government, have substantial overhead in granting licenses. Not only are there multiple requirements to address any possible risk–warranties, indemnification, insurance to cover the indemnification, documentation to prove the insurance, threats if the insurance lapses, demands that the insurance operates before the university’s own insurance–but also matters having to do with governing law, jurisdiction for disputes, arbitration before litigation, choice of rules for such arbitration, expenses of arbitration, whether the arbitration is binding, and the like. One advantage of invention management agents is that they do not have to become hung up on whether the governing law will be, say, Michigan or Hawaii (though when it comes to designating which courts will be used, Hawaii comes pleasantly to mind).
In this second diagram, I identify five common objectives in the management of inventions. These range from an emphasis on the information content at the top, to utility of the invention nearer the middle, to financial and leverage aspects of the patent rights in the invention towards the bottom. “Science” stands for the activities of scholarship, of exchange of information, deposit of records for testing and for use in research. “Practice” identifies various forms of use that may be made of invention by those reading published accounts and having the capacity to implement the invention–a typical situation, for instance, with method-based inventions. “Development” emphasizes efforts to advance an invention for applications to be used by others, often in the form of commercial products. Development may or may not benefit from exclusive patent positions, depending on the industry, the circumstances, those undertaking the effort, and their operating model. “Investment” considers a patent right for its expectation of returns on expenditures, and often takes the form of a startup company seeking investors as its primary initial mission. For investment, patent rights often figure prominently in planning, though in practice they may never come to anything. “Speculation” may take the form of arbitrage, in which a patent right is acquired with the expectation it may be transacted later for more than one has paid for it. Speculation also involves troll behaviors, in which a patent owner seeks to extract a toll from users of an invention, deriving some portion of the value from the perceived cost to fight infringement litigation. Speculation depends heavily on the patent right for its practice, and considers income as a primary objective, without particular regard for the effects on decisions regarding use or development of an invention.
For each of these areas, I have identified five potential agents that might deploy invention rights made in federally supported faculty research. Of these, three are associated, generally, with a range of practices. The federal government has tended to work the top portion of the stack, emphasizing research, use, and road maps for development shared within an industry. This federal tendency came in for heavy criticism by advocates for Bayh-Dole, who argued for the advantages of exclusive licensing, commercial interests, and the attraction of investment capital to “early stage” (as they are often called) inventions.
Agents and universities acting as their own agents tend to have different ranges. Some agents take the form of non-profit foundations and manage rights on behalf of a community of users. Examples include the Apache Foundation and the Ecopatent Commons, along with standard setting organizations. Other agents operate non-exclusive licensing programs to advance development of an invention, to spur competition, and enhance economic development. Agents also, of course, can focus on attracting investment to an invention. However, agents typically are tasked invention by invention to provide services. As with a sports agent, one does not take on a client merely to add to the list of clients–the agent has to land the client a roster spot with a good contract or the agent gets fired. Similarly with an agent taking on the task of promoting a federally sponsored invention. Under the SPRC, each invention that is retained must be promoted. There is no indication under Bayh-Dole that Congress intended the law to be exploited by agents to accumulate inventions so that some few might be successful and the rest can rot as a necessary collateral effect, so that a program can appear successful in part because it has no competitors for the assets it intends to take under management.
Universities acting as their own agents have a problem: they cannot fire themselves. Nor can they challenge their own practices, reveal their failures or weaknesses, or recover damages for their own misdeeds. Universities, by claiming ownership of inventions by compulsory arguments (such as making assignment of inventions a condition of employment, though frequently it is not, or a condition of use of resources, though the universities provide those resources anyway, and in the case of federal funding have already agreed with the government to provide the resources, for which the universities are compensated), also relieve themselves of pressure to act on every invention they claim. They claim everything, and release only what is worthless. If someone wants them to release an invention, however, so their reasoning goes, the invention must not be worthless, and so it is worth keeping and attempting to license for payment. In state university settings, release is made all the more difficult by arguments that an invention is an asset of the state, and must be offered for value to the public before allowing anyone, such as, say, its inventor, to get a special favor in the form of a release of the invention, especially at no charge. In some cases, universities will not consider the release of an invention unless the inventors agree to pay the university’s costs in processing the invention and filing patent applications, even if the university shows no indication that it has the ability or interest in pursuing licensing opportunities of any sort–from science to speculation.
Subcontractors may turn out, as well, to be agents of one sort or the other, or may contract with agents, and therefore less can be said about their preferred manner of management, other than that university subcontractors tend to behave as universities, and company subcontractors often have no interest in working an invention through an agent or a direct licensing program, preferring to hold patent rights as a company asset and developing the invention directly.
The great uncertainty in all of this is what a faculty inventor might do. In nearly all discussions of Bayh-Dole, no attention is paid to the role of the principal investigator, who may choose subcontractors and therefore indirectly move some aspects of invention management around. A faculty investigator also would have the opportunity, if a university administration allowed it, to choose both a pattern of emphasis for inventions made in a project, and the form of management agent suited to that pattern. If an investigator made such choices at the outset of a federally funded project, then all those working in the project would understand the situation from the outset, and the interests of inventors in the project would necessarily be mediated by the established ground rules for the work. Universities, however, generally do not allow investigators the standing to make those choices, though often there is a nod to informing them of what is being considered.
Most of the emphasis in discussions of Bayh-Dole are on what inventors might do, and how universities offer superior services over what inventors could do on their own. Such discussions are generally unhelpful, in part because the issue is not a comparison between an institutional invention licensing office and an individual inventor on the matter of doing deals for commercial development based on investment of funds, but rather on whether the inventor, having chosen a course of action, can find or create better resources matched to that course than those on offer from the university. The discussions are unhelpful in another way, as inventors and investigators alike are often at a substantial disadvantage as a result of university policies imposed by the technology transfer office, to enfranchise its chosen operating model and to suppress anything that might prove competitive with that model.
Thus, there is a question mark regarding where a faculty inventor might play across the spectrum of uses for an invention, from illustration of science to speculation for financial purposes. This question, however, is not a defect, but rather the most important of opportunities for diversity. Rather than imagining invention management as an assembly line of production processes leading from disclosures to patent applications to issued patents to licenses and income, as inventors make choices independent of institutional planning and demands, a great many things become possible across the spectrum. In particular, it can be expected that faculty inventors will spend a greater share of time encouraging use of inventions rather than withholding use in favor of financial returns, and indeed may prefer government ownership and management to that of their own university, especially if the university runs its invention activities as a financial portfolio that is successful with one or two lucrative licenses per decade, which is all it takes to look good on paper if one hides the condition of the portfolio inventions in the right public metrics.
The university architecture of Bayh-Dole has been set up to advantage the kind of patent licensing that has been favored in the pharmaceutical industry. That architecture is baked into Bayh-Dole, with the emphasis on the use of the patent system, the presumption (though not the fact) of university ownership, and the supplied mandate that the purpose of university patent management is commercialization, and the reason for that commercialization is so that the university might make money. While universities also recite such platitudes as “public benefit,” their licensing actions indicate that what they mean by public benefit is along the lines of “institutional benefit, because if the institution benefits, then the public should consider that a public benefit.”
What has been lost in the thirty years of Bayh-Dole has been the commons that was fed by faculty-led research. That commons formed the foundation for platform development, for standards, for the ready movement of ideas and technology among organizations and between universities and industry. That commons supported long-term, slow developing work that might take the form of hunches or “networked, non-market” collaborations. That commons allowed graduate students to practice what they learned in the university laboratories, and in their degree research, without having to ask permission of administrators, or negotiate complicated legal agreements. That commons permitted a reader of a university publication to expect to be able to practice what was learned without having to ask for permission from university administrators. That commons has been largely undermined by a combination of university comprehensive claims to ownership of faculty inventions, the insertion of those inventions for management into a licensing program built on a financial portfolio model, and the accumulation of inventions that therefore cannot be practiced without first accommodating a paywall and a preference for commercialization, either through attraction of speculative investors or the production of a commercial product, either preferentially under an exclusive license.
Any unworked patented invention is a barrier to practice. Any exclusive license is a barrier to practice for each claim that remains unworked. Any exclusive license to a patent that results in a financial return to the university but never results in the practical application of the underlying invention is, in the end, a failure. While there are indeed times when a patent is important for investment (to justify a substantial investment) or development (to provide for cross-licensing or standards) or even for practice (to control quality, say), the present university approach of patent accumulation without selectivity or investigator and inventor choice is contributing to research stagnation and actually works against pre-conditions for innovation–collaboration, serendipity, freedom to act, freedom to move among organizations, ready formation of networks, and access without substantial overhead, delays, and cost.
In the basic formulation of Bayh-Dole, the discussion is between the government and the inventor. The inventor can opt out of that discussion by choosing an agent. University administrators have largely disabled these fundamentals by demanding ownership and operating their own institutionally self-interested patent licensing programs, blocking the use and development of agents focused on industry segments, or specializing in particular technologies, or working in a particular region. The growth of university licensing programs, called evidence of the success of Bayh-Dole, can also be viewed as the suppression of the fundamental bargains that Bayh-Dole offered faculty inventors, in the hope that these inventors might do for federally supported inventions what they were doing, selectively, across a range of choices, for inventions supported in other ways. The job for policy-makers these days, then, is to consider how to restore that basic bargain before the infrastructure that it anticipates has been completely dismantled.