In 1999, the Council on Governmental Relations, a non-profit organization that fronts for over 100 research universities on lobbying and advocacy issues, produced an influential summary of the Bayh-Dole Act, “The Bayh-Dole Act: A Guide to the Law and Implementing Regulations.” The COGR summary has made its way to a number of university technology transfer office web sites and may well be the primary authority for many in understanding the Act. Unfortunately, the COGR summary has got Bayh-Dole wrong in critical places and contributes to the misapprehension of the Act, the Standard Patent Rights Clause, and university invention policies.
I wrote to COGR at one point a few years ago asking for a clarification on their interpretation of Bayh-Dole and they didn’t even bother to reply. Later, I saw a public copy of some fun email where COGR folks worried to some other folks about my examination of Bayh-Dole and appear to fret in passing that maybe I had a point. So folks at COGR, perhaps, can take this as an open letter about where they might revise their Guide to Bayh-Dole, and perhaps in doing so take the time to point out their changes to the broader community, for everyone’s benefit. Since the Supreme Court’s ruling in Stanford v Roche, which made it clear that Bayh-Dole is not a statute that vests ownership of inventions made with federal support with the university that hosts the research, COGR really does need to reconsider the status of its published documents pertaining to Bayh-Dole. Of course, if they want assistance, I’m available to help.
We will go through the brochure section by section. So open up a second browser window, and follow along. Here is are links to Bayh-Dole and to the implementing regulations, if you want to check text for yourself.
The 1999 COGR Guide to the Law:
This brochure is intended to inform the public about the Patent and Trademark Law Amendment Act of l980 [sic], more commonly known as the Bayh-Dole Act, which is the legal framework for transfer of university generated, federally funded inventions to the commercial market place.
Before we even get to the introduction COGR takes things off the rails. The Bayh-Dole Act is not a legal framework for the transfer of inventions. The Bayh-Dole Act sets out when the federal government can contract to obtain title to inventions made with federal support, and minimum standard requirements on invention management agents when they obtain title to federally supported inventions. How these invention management agents promote, or interfere, with the transfer of technology is entirely outside the scope of the Bayh-Dole Act. At best, the Bayh-Dole Act is a law limiting federal agency contracting for the transfer to the federal government of the right to patent inventions made with federal support.
One might say that the Bayh-Dole Act resolved, at least for a time, the debate over whether a government preference for using the patent system to provide non-exclusive access to federally supported inventions is better than university administration desires to grant monopoly-creating exclusive licenses in the hope of windfall royalties. One might also say that the Bayh-Dole Act resolved a competition between invention management agents affiliated with a single university and those offering services nationally or to multiple institutions. While the Bayh-Dole Act does not prevent the use of national invention management agents, the interpretation placed on the Act by university administrators aimed to ensure that university administrators were required to decide on the choice of an invention management agent, not faculty and not inventors.
If the Guide were forthright, it would use “university faculty and staff generated” rather than “university generated.” The “university” is neither the commissioning agent, nor the supervisor of the work, nor the party for whom the work is undertaken. The university is merely the steward of funds and provider of facilities and other support, for which it is compensated through indirect costs payments. By using “university generated” the Guide cleverly seeks to introduce language that gives the appearance of legitimacy to the broad claim that the Guide will make: that since the “university” generates inventions, it is only fair that the university be granted by federal law the right to own those inventions and seek to profit from them.
The Bayh-Dole Act’s focus is not on the “commercial market place.” Throughout, the guiding definition is “practical application”:
(f) The term “practical application” means to manufacture in the case of a composition or product, to practice in the case of a process or method, or to operate in the case of a machine or system; and, in each case, under such conditions as to establish that the invention is being utilized and that its benefits are to the extent permitted by law or Government regulations available to the public on reasonable terms. (35 USC 201(f))
There is no mention of “commercialization” or “sale” here. Selling stuff is one way that the benefits of an invention might be made available to the public, and if the pricing is done appropriately, even on reasonable terms. The “reasonable terms” part is there to remind everyone that university invention management agents had been caught out in anti-competitive behaviors based on their monopoly patent positions, and Congress was not going to stand for more of that. The emphasis of the definition, however, is that the benefits of using each invention are available. There is no requirement or mandate or even hint that an invention must be sold, or that an invention be incorporated into a product that is sold: the emphasis is on use of the invention. Certainly there is nothing that approves of seeking income from dealing with patent rights among speculators whose interests range from plumping an opportunity to sell it to even more gullible speculators to waiting for industry to stumble over a thicket of patents and then assert rights for profit.
The stated objectives of the Bayh-Dole Act also emphasis use and warn invention management agents multiple times against antitrust and other bad behaviors:
It is the policy and objective of the Congress to use the patent system to promote the utilization of inventions arising from federally supported research or development; to encourage maximum participation of small business firms in federally supported research and development efforts; to promote collaboration between commercial concerns and nonprofit organizations, including universities; to ensure that inventions made by nonprofit organizations and small business firms are used in a manner to promote free competition and enterprise without unduly encumbering future research and discovery; to promote the commercialization and public availability of inventions made in the United States by United States industry and labor; to ensure that the Government obtains sufficient rights in federally supported inventions to meet the needs of the Government and protect the public against nonuse or unreasonable use of inventions; and to minimize the costs of administering policies in this area.
Yes, the word “commercialization” does appear, but only joined to “public availability” and only when linked to “industry” as the agent that undertakes either, or both. Wherever in the Act and the implementing regulations there is any hint of selling, there one will find as well a parallel reference to use. It is clear that the Act was not intended to enfranchise university “commercialization”–exclusive licensing to create local monopolies–as the primary focus of all federal support for faculty-led research. It is entirely possible, however, that the university administrators and their supporters involved in drafting the Act had to include “use” as an alternative, to settle down anyone who might oppose a federal law that would appear to approve of university administrators withholding inventions from use until someone arrived willing to attempt to make commercial products and also willing to pay for the exclusive right to do so.
It is no coincidence that the University of California was a big supporter of the Bayh-Dole Act, and later of the mistaken but administratively attractive idea that Bayh-Dole vests ownership of inventions with universities, and most recently of the idea that all faculty must agree to assign all future inventions to the university before those inventions have even been made. If one was at the University of California’s dedicated centralized licensing office, and did not want faculty considering any alternatives, then a federal law that appeared to drop a mantel of law around the licensing office would be just the thing. COGR calls out the director the University of California licensing office for assistance in drafting the Guide. One might then put one and one and one together.
Introduction
The current national emphasis on this activity, however, can be dated to the 1980 enactment of P.L. 96-517
The Bayh-Dole Act is the beginning of a national emphasis by university administrators on university ownership of faculty inventions made with federal support. Bayh-Dole indeed was a pivot in technology transfer that turned an already robust university technology transfer environment from a faculty-led activity with many inventions going to the public domain or to be managed by the federal government for broad access to an administrator-led activity focused on seeking profits from patent positions, which has scaled in cost, delay, and accumulation of patents while narrowing the focus, methods, and resources involved. See David Mowery’s work for a well documented history.
If COGR wishes to present an editorial giving the perspective of university administrators, that is of course fine. But then present it as such. In that case COGR could retitle its work as a “Guide to the Opinions of University Administrators about Bayh-Dole.”
Background
The Guide presents an odd history of university technology transfer. The Guide mentions Vannevar Bush but does not point out that present university practice abjures the approach advocated by Bush and substitutes instead a compulsory institutional ownership model that overwhelmed the existing faculty-mediated invention equity approach. The irony is that the compulsory institutional ownership model is what was discredited at the level of the federal agency, the primary difference being that the federal agencies were not prepared to offer monopoly licenses to whomever showed up, and all the more so were not interested in offering such monopoly positions to speculators on the premise that the public would benefit when the speculators made money from such patent rights.
These activities continue to accelerate nationally as the importance of the Bayh-Dole Act becomes fully appreciated. Evidence of this is reflected in the fact that the membership of the Association of University Technology Managers (AUTM) increased from 691 in 1989 to 2,178 in 1999. In 1979, the year before passage of the Bayh-Dole Act, the Association counted only 113 members.
The metrics reciting the growth of university licensing officers is one that tracks a shift in management of inventions from invention management agents, the federal government, and the public domain using modes resources to university administrations devoting substantial resources to those same activities, apparently using less discretion, with less financial discipline, and adopting more expensive methods. The Guide makes it appear that activity was springing new from the earth rather than that activity was shifting from other venues to be concentrated in university administrations.
Growing numbers of universities have demonstrated that their newly formed technology transfer programs are effective in licensing inventions made with federal support to commercial partners. As a result, many new technologies have been diligently and successfully introduced into public use.
The Guide makes it appear that the growth of administration was responsible for the introduction into public use of new technologies. The growth in the number of university licensing officers does not indicate that faculty were increasing their rate of discovery as a result of Bayh-Dole, or that more inventions were getting used because Bayh-Dole allowed universities to insist on owning such inventions. Changes in patent law in the early 1980s allowed software and biomaterials to be patented. Patent law was changing, creating a lot more work for university administrators who wanted to own faculty inventions.
The introduction of provisional patent applications in 1996 allowed patent applications to be filed as bare manuscripts without claims, further increasing the number of patent applications filed–not only because it was inexpensive to file a provisional application, but also because one got to count a second patent application if the provisional was converted to a full utility patent application. Add a third patent application if a PCT application for foreign rights was filed with a little reformatting at the same time. The steady increase in federal funding supporting faculty-led projects at universities also expanded opportunities for invention. None of this change in activity surrounding the patenting of research results has much of anything to do with the Bayh-Dole Act. If it does, someone will have to do a lot more work than this hand-waving Guide to demonstrate it.
The increase in technology transfer offices certainly does represent an increase in university administrative expenditure devoted to pursuing claims of ownership and profits from patent positions. That is, university administrators chose to divert funds that would have been spent elsewhere to the support of patent ownership and licensing. One might reasonably surmise that such growth reflects the conversion of university administrators to a philosophy of spending additional resources on the institutional ownership of patents, induced by an attractive misreading of the Bayh-Dole Act under which the institution was given the right to take faculty inventions, as a matter of federal law.
There was no government-wide policy regarding ownership of inventions made by government contractors and grantees under federal funding.
Yet there was–it was a policy embedded in executive orders from Roosevelt to Carter. That policy at the time of Bayh-Dole, emphasized flexibility among the agencies to ensure that government interests were served and inventions got used. See Rebecca Eisenberg’s well documented study.
Until the Bayh-Dole Act became effective on July 1, 1981, the federal agencies kept tight control over intellectual property rights resulting from funded research, premised largely on traditional expectations rooted in the procurement process.
The Guide makes it appears that there was a uniform federal policy with regard to ownership of inventions made with federal support–that agencies kept “tight control.” But earlier the Guide argues there was no government-wide policy. Further, Bayh-Dole pertains only to inventions and plant varieties, not to intellectual property rights generally. But the truth of it is, some federal agencies did not require assignment of inventions to the government–notably the NSF and Department of Defense. Some did–especially HEW/PHS/NIH and the Department of Energy. For HEW/PHS/NIH in particular, there was a long-standing legal opinion that inventions were to be assigned to the government.
The problem for the university-affiliated folks determined to pursue income from patent positions was that agencies that required assignment did not always subsequently agree to allow universities or their dedicated invention management agents manage inventions made with federal support. Universities and their agents often did not have any capability to develop inventions themselves; at least one leading invention management agent, WARF, had a history of antitrust run-ins with the government; and a federal agency may not want a university granting a monopoly position to one company over other companies (and universities)–which easily can be regarded as blocking rapid adoption of inventions rather than a step necessary to the adoption of inventions. Thus, when a university or invention management agent went out and obtained assignment anyway, there was no “certainty” that the federal agency would respect that assignment, and could demand that the agent comply with the agency’s procurement requirements. One could see how much bother this might be to aggressive patenting folks who jumped the gun, and for federal agencies that then had to unwind the deal with such folks.
The provisions apply to all inventions conceived or first actually reduced to practice in the performance of a federal grant, contract, or cooperative agreement.
The Guide misrepresents the Bayh-Dole as applying to inventions when the law clearly applies to federal funding agreements offered by federal agencies, and establishes how federal agencies may procure inventions made under federal funding agreements. The law applies to federal agencies. 37 CFR 401.1(b): “This part inplements [sic] 35 U.S.C. 202 through 204 and is applicable to all Federal agencies.”
The Standard Patent Rights Clause authorized by the law and as may be modified by federal agencies sets the requirements for contractors, including universities, subcontractors, and employees. These requirements are conditions of federal contracts. How obligations pertaining to inventions made with federal support arrive at universities is critical to an understanding of Bayh-Dole. Only by hiding the process does the Guide set up the argument that Bayh-Dole operates to give university administrators control of ownership of faculty inventions.
The Guide misrepresents, further, that the SPRC applies to all inventions made with federal support, when the SPRC only applies to those inventions that become subject inventions, and subject inventions are those inventions “of the contractor”–that is, that have come to be owned by the contractor. The SPRC does not apply to inventions made by nontechnical workers, by volunteers or informal collaborators, or even to employees of the contractor who have not made the written agreement required by the SPRC. Again, the details matter. The Guide makes it appear that all inventions must be therefore be managed by university administrators, when the SPRC makes it clear the government has a limited, carefully qualified interest in inventions.
The university is obligated to have written agreements with its faculty and technical staff requiring disclosure and assignment of inventions.
Here the Guide misrepresents the nature of the written agreements required by the Standard Patent Rights Clause. COGR makes it appear that these written agreements are between a university and its employees. The actual agreement is one that certain employees make to protect the government’s interest and by doing so those employees become parties to the funding agreement. See 37 CFR 401.14(a)(f)(2) and the heading for section (f). The agreement is for the benefit of the government, not the university.
Upon election of title, the university must file a patent application within one year, or prior to the end of any statutory period in which valid patent protection can be obtained in the United States.
When a university elects title to an invention, it assumes responsibility for taking certain actions to properly manage the invention and provide certain reports to the government regarding the invention as outlined in the section on Current Regulations above.
Throughout, the Guide waffles from “retain title” to “elect to retain title” to “elect title”. Bayh-Dole and the SPRC use these same expressions, but the meaning there is that a contractor obtaining title to an invention is to notify the government if the contractor will retain that title. The Guide fails to point out that the SPRC (or Bayh-Dole, as the Guide would have it) does not require inventors to assign inventions to the university that employs them; nor that such assignment is necessary before a university may elect to retain title; nor that “elect title” does not mean “take title” but rather “retain title that one has got through a normal process of negotiating assignment of title.
If the university does not elect to retain title, the agency may take title to the invention.
Here the Guide misleads readers with regard to the outcomes possible if a university does not elect to retain title. If the university has not obtained title, then the disposition of title is a matter between the inventors and the federal funding agency, per 35 USC 202(d) and 37 CFR 401.9. That is, inventors own rights to their inventions until they assign those rights. Inventors could assign title to an invention management agent other than the university, especially if the university does not want to manage the invention. Indeed, this is how things were done at the University of Wisconsin and WARF, where inventors assigned directly to WARF, not to Wisconsin, which then would have had to assign to WARF.
A university has no claim to take invention rights by virtue of Bayh-Dole.There is absolutely nothing in Bayh-Dole that gives a university a right to the personal property of inventors working with federal funds. This, in fact, is one of the hallmarks of Bayh-Dole. These arguments apply as well to federal agencies. Bayh-Dole gives agencies no right to take title from inventors. Agencies must contract with the inventors’ employers so that those employers require the inventors to commit to establish the government’s rights–namely, to assign or license to the government, when the government as represented by a federal agency party to a funding agreement is authorized by Bayh-Dole that it may make such a request.
In their marketing of an invention, universities must give preference to small business firms (fewer than 500 employees), provided such firms have the resources and capability for bringing the invention to practical application.
The Guide misrepresents the requirements regarding university licensing efforts for small businesses, omitting the requirement that universities must make efforts to attract licensees that are small businesses, not merely give a preference to small business firms. Assumes that all such licenses will be exclusive. See 37 CFR 401.14(a)(k)(4).
Universities may not assign their ownership of inventions to third parties, except to patent management organizations.
Here the Guide fails to explain or qualify university (and other contractor) rights to assign both inventions as well as obligations under the Standard Patent Rights Clause. These two matters–the SPRC and inventions made subject to the SPRC, are distinct. Any organization which has as one of its primary functions the management of inventions (not merely patents) may receive assignment without the need for federal funding agency approval. See 37 CFR 401.14(a)(k)(1). Other assignments may be made with agency approval. By way of contrast, obligations of the SPRC may be transacted by assignments, substitutions of party, and subcontracts, and change who is a party to the funding agreement, and therefore, also, who is a contractor–one of the key definitions in play in the SPRC.
Agencies may decide, for compelling reasons, that title should be vested in the federal government.
Here the Guide continues to misrepresents the process by which title to a subject invention is obtained by the federal government, making it appear that such title “vests” when in fact that title is conveyed upon request, in reliance upon the SPRC in the funding agreement. A “vested” right is a “right belonging completely and unconditionally to a person as a property interest which cannot be impaired or taken away (as through retroactive legislation) without the consent of the owner.” Title to subject inventions vests in the inventor, following federal patent law and the U.S. Constitution.
University patenting and licensing efforts under the Bayh-Dole Act have fostered the commercialization of many new technological advances that impact the lives of millions of people across the nation….Most of the inventions involved were the result of federal funding. While it would be impossible to list all such inventions, a few examples of technologies and products originating from federally funded university discoveries include:
The Guide claims that the Bayh-Dole Act has fostered the commercialization of “many new technological advances” yet makes a quick switch so that it may include in its list inventions that were not made under Bayh-Dole, such as the recombinant DNA technology.
These examples of successful new technologies demonstrate that a strong national infrastructure to support technology transfer has been established at academic institutions across the nation since passage of the Bayh-Dole Act. In 1980 there were approximately 25-30 universities actively engaged in the patenting and licensing of inventions. It is estimated that there has been close to a ten-fold increase in institutional involvement since then.
The Guide misleads the reader with regard to the condition of “national infrastructure to support technology transfer” by reciting only the universities that were active in patent licensing prior to 1980 rather than pointing out that many universities chose not to be active in such matters and encouraged faculty to deal directly with external research foundations that took on these responsibilities so university administrations need not fund, nor be conflicted by, and could therefore independently monitor issues pertaining to invention management.
Academic institutions were granted more than 8,000 U.S. patents between 1993 and 1997 for technologies discovered by their researchers.
Here the Guide fails to distinguish subject inventions from other inventions managed by universities, leaving the impression that all such inventions have been federally funded, when federal funding represents no more than about 60% of present university research expenditures. It may be that the Guide is making the case that Bayh-Dole provided the leverage for university administrators to claim to own inventions other than those that were federally supported. If so, this would be an excellent point at which to make that case expressly.
Across the metrics presented in this list of points on page 9, the Guide fails to provide context to show how many inventions that previously may have been managed by an external agent such as Research Corporation or Battelle Development Corporation have been sequestered in university administrative hands. Furthermore, the Guide changes time scales from one point to the next to make it difficult to ascertain the relationship between ownership of inventions, patents, and resulting activities. Here are inventions from 1993 to 1997. But here is company formation since 1980 (even though Bayh-Dole did not go into effect until July 1, 1981). Here are company startups in FY 1997–but whose fiscal year? and why not a calendar year?). Then stuff that appears not to vary by year–economic activity and jobs, and a statement about the number of products on the market without any reference to whether these are products created since Bayh-Dole went into effect, specifically the result of university management under Bayh-Dole, and without any indication why there are 2,200 new companies attributed to university licensing since 1980 but only 1000 products on the market from all university licensing across all years.
The Guide then turns to “technologies licensed from academia,” which the Guide claims have done all sorts of wonderful things. Perhaps so. But this is not the point. If academia (what does “academia” mean, relative to faculty? to external licensing agents? to inventions dedicated to the public domain, such as those that were foundational to the internet?) has been so generally productive, what has been the particular effect of the Bayh-Dole Act? The Guide’s claim, as stated, can be true even if the Bayh-Dole Act has led universities to disrupt the transfer of technology through their licensing requirements.
It is clear that the Guide has no intention of presenting information that is specific to university licensing practices directly attributed to the benefits provided by the SPRC under the Bayh-Dole Act. The Guide is clearly praising compulsory institutional ownership of faculty inventions, not the restriction of federal agency requirements to assign of inventions made with federal support that are subsequently assigned to an invention management agent. If you want to see even one critical difference between the Guide and reality, this is the one to see.
Over 2,200 new companies have been formed since 1980 that were based on the licensing of an invention from an academic institution, including over 330 companies formed in FY 1997 alone.
Approximately $30 billion of economic activity each year, supporting 250,000 jobs can be attributed to the commercialization of new technologies from academic institutions.
The Guide provides no documentation for the claims of company formation or commercial sale of products, nor how many of those claimed arise from Bayh-Dole. The Guide fails to disclose that the figures for economic activity and jobs are arrived at, apparently, by simply multiplying university reported royalty income by the reciprocal of an assumed royalty rate (2% would mean a factor of 50), and the jobs estimate arrived at dividing this larger figure by an assumed typical salary plus benefits of a well paid technology industry worker (such as, say, $120,000). Such calculations have never been checked against actual data and do not take into account financial losses incurred by companies taking licenses to university patents, job losses created by new technology, or profit-taking by licensees that do not expand their employment base but rather use a monopoly position to increase their margins.
Certainty of title to inventions made under federal funding is perhaps the most important incentive for commercialization.
The Guide leaves unqualified that “certainty of title” means that an agency cannot contest a properly assigned invention made with federal support, and does not mean that a university is assured by the law of the right to demand title from its employee inventors. If certainty of title is indeed “the most important incentive for commercialization” then that applies equally well to inventors using federal funds. Nothing in Bayh-Dole sets forth either a mandate or incentive for university administrators to become predatory on faculty (and other) inventors.
The Guide fails to provide documentation to substantiate the claim that universities operating under the Bayh-Dole Act have increased the “technology transfer from universities to industry, and ultimately to the public.” No such publicly reviewable documentation exists. Such reports are protected under Bayh-Dole as exempt from public disclosure and universities do not volunteer this information. Folks will have to pry it out of public universities by making public disclosure requests under state law.
Certainty of title to inventions made under Federal funding has one other significant benefit—it protects the right of scientists to continue to use and to build on a specific line of inquiry.
Here the Guide steps over the line from suppressing information to the absurd. Certainty of title has to do with whether a federal agency might claim the contractual right to require assignment from university inventors. There was no concern that the federal government might sue university faculty for continuing to work on their inventions. Further, under Bayh-Dole, even if the government obtains ownership, the university is assured of a right to practice the invention, which then has nothing to do with ownership. Under the government’s non-exclusive license, the government can authorize scientists to continue work on inventions made with federal funding, quite apart from university-held title.
Finally, and most importantly, the single greatest threat to the right of scientists to continue to use inventions made with federal support is the monopoly licensing practices of universities. Under such exclusive licenses, the universities routinely do not reserve rights for all scientists–not just those employed by universities–to continue to practice inventions. Indeed, at the summit that produced the Nine Points to Consider document, many university administrators spoke out strongly against my proposal that such rights should be routinely reserved in exclusive licenses. Recently I talked with a long-time lab director at a research university. His university had licensed exclusively his fundamental discovery (in material science) to a startup company and had failed to reserve broad rights for the practice community. The university’s licensee had since shifted directions, wasn’t developing his invention very quickly, and he was now, years later, having to say no to company after company offering to collaborate with his lab. That is wanting “certainty of title” has done, the way the Guide uses it.
The misrepresentation goes even deeper. Compulsory institutional ownership does not allow inventors to continue to use their inventions when they separate from university employment, or even do work in the community outside the scope of their “official duties,” such as consulting or public service. In those cases, university inventors have no right to practice their inventions, and cannot even reserve those rights for themselves by negotiating with university administrators. If “certainty of title” is what the Guide presents it as, then it is the single worst antagonist of academic freedom created by federal law. Fortunately, the Guide at this point has gone raving mad and Bayh-Dole actually does nothing of the sort. But a reader would never know it without having a little bit of context and a willingness to read the law and implementing regulations to check COGR’s work.
The retention of title to inventions by the institution is the only way of ensuring that the institution will be able to accept funding from interested research partners in the future. This is a critically important benefit of the Bayh-Dole Act that is not widely understood.
Finally, the Guide claims wrongly and absurdly that compulsory institutional ownership of federally funded inventions is “the only way” that the institution can accept research funding in the future. An institution merely needs to have permission to practice–which is typically available by means of a “shop right,” a license based in equity. Compulsory institutional ownership has absolutely nothing to do with it.
The Guide ends with another reference to what Vannevar Bush foresaw. Dr. Bush would not have felt particularly pleased with what COGR and university administrators have done to Bayh-Dole in their efforts to impose bureaucratic controls on faculty inventions made with federal support. It would be most wonderful if COGR would withdraw their Guide, and republish it with a better documented, reasoned, and qualified statement of what Bayh-Dole does, and what it has done.