skip navigation

583 F. 3d 832, affirmed.

HTML version
PDF version
HTML version
PDF version
HTML version
PDF version
HTML version
PDF version

Breyer, J., dissenting



on writ of certiorari to the united states court of appeals for the federal circuit

[June 6, 2011]

     Justice Breyer , with whom Justice Ginsburg joins, dissenting.

     The question presented in this case is:

“Whether a federal contractor university’s statutory right under the Bayh-Dole Act, 35 U. S. C. §§200–212, in inventions arising from federally funded research can be terminated unilaterally by an individual in-ventor through a separate agreement purporting to as-sign the inventor’s rights to a third party.” Brief for Petitioner i.

In my view, the answer to this question is likely no. But because that answer turns on matters that have not been fully briefed (and are not resolved by the opinion of the Court), I would return this case to the Federal Circuit for further argument.


     The Bayh-Dole Act creates a three-tier system for patent rights ownership applicable to federally funded research conducted by nonprofit organizations, such as universities, and small businesses. It sets forth conditions that mean (1) the funded firm; (2) failing that, the United States Government; and (3) failing that, the employee who made the invention, will likely obtain (or retain) any result- ing patent rights (normally in that just-listed order). 35 U. S. C. §§202–203. The statute applies to “subject in-vention[s]” defined as “any invention of the contractor conceived or first actually reduced to practice in the performance of work under a funding agreement.” §201(e) (emphasis added). Since the “contractor” ( e.g., a university or small business) is unlikely to “conceiv[e]” of an idea or “reduc[e]” it “to practice” other than through its employees, the term “invention of the contractor” must refer to the work and ideas of those employees. We all agree that the term covers those employee inventions that the employee properly assigns to the contractor, i.e. , his or her employer. But does the term “subject invention” also include inventions that the employee fails to assign properly?


     Congress enacted this statute against a background norm that often, but not always, denies individual inventors patent rights growing out of research for which the public has already paid. This legal norm reflects the fact that patents themselves have both benefits and costs. Patents, for example, help to elicit useful inventions and research and to assure public disclosure of technological advances. See, e.g., Mazer v. Stein , 347 U. S. 201, 219 (1954) ; Bilski v. Kappos , 561 U. S. ___, ___ (2010) (slip op., at 4); id., at ___ (slip op., at 10) (Stevens, J., concurring in judgment). But patents sometimes mean unnecessarily high prices or restricted dissemination; and they sometimes discourage further innovation and competition by re-quiring costly searches for earlier, related patents or by tying up ideas, which, were they free, would more effectively spur research and development. See, e.g., Laboratory Corp. of America Holdings v. Metabolite Laboratories, Inc. , 548 U. S. 124, 128 (2006) ( Breyer , J., dissenting from dismissal of certiorari as improvidently granted); Heller & Eisenberg, Can Patents Deter Innovation? The Anticommons in Biomedical Research, 280 Science 698 (1998).

     Thus, Thomas Jefferson wrote of “the difficulty of drawing a line between the things which are worth to the public the embarrassment of an exclusive patent, and those which are not.” Letter to Isaac McPherson (Aug. 13, 1813), in 6 Writings of Thomas Jefferson 181 (H. Washington ed. 1854). And James Madison favored the patent monopoly because it amounted to “compensation for” a community “benefit.” Monopolies. Perpetuities. Corporations. Ecclesiastical Endowments., in J. Madison, Writings 756 (J. Rakove ed. 1999).

     The importance of assuring this community “benefit” is reflected in legal rules that may deny or limit the award of patent rights where the public has already paid to produce an invention, lest the public bear the potential costs of patent protection where there is no offsetting need for such protection to elicit that invention. Why should the public have to pay twice for the same invention?

     Legal rules of this kind include an Executive Order that ordinarily gives to the Government “the entire right, title and interest” to inventions made by Government employees who “conduct or perform research, development work, or both.” 37 CFR §501.6 (2010) (codifying, as amended, Exec. Order 10096, 3 CFR 292 (1949–1953 Comp.)). See also Heinemann v. United States , 796 F. 2d 451, 455–456 (CA Fed. 1986) (holding Executive Order constitutional and finding “no ‘taking’ because the invention was not the property of Heinemann”). They also include statutes, which, in specific research areas, give the Government title to inventions made pursuant to Government contracts. See Atomic Energy Act of 1954, §152, 68 Stat. 944 (codified as amended at 42 U. S. C. §2182); National Aeronautics and Space Act of 1958, §305, 72 Stat. 435 (codified at 42 U. S. C. §2457), repealed by §6, 124 Stat. 3444; Federal Nonnuclear Energy Research and Development Act of 1974, §9, 88 Stat. 1887 (codified as amended at 42 U. S. C. §5908(a)). And they have included Government regulations, established prior to the Bayh-Dole Act’s enactment, that work in roughly similar ways. See, e.g., 45 CFR §650.4(b) (1977) (National Science Foundation regulations providing that Foundation would “determine the disposition of the invention [made under the grant] and title to and rights under any patent application”); §§8.1(a), 8.2(d) (Department of Health, Education, and Welfare regulations providing that inventions made under department grants “shall be subject to determination” by the agency and that the department may “require that all domestic rights in the invention shall be assigned to the United States”).

     These legal rules provide the basic background against which Congress passed the Bayh-Dole Act. And the Act’s provisions reflect a related effort to assure that rights to inventions arising out of research for which the public has paid are distributed and used in ways that further specific important public interests. I agree with the majority that the Act does not simply take the individual inventors’ rights and grant them to the Government. Rather, it assumes that the federal funds’ recipient, say a university or small business, will possess those rights. The Act leaves those rights in the hands of that recipient, not because it seeks to make the public pay twice for the same invention, but for a special public policy reason. In doing so, it seeks to encourage those institutions to commercialize inventions that otherwise might not realize their potentially beneficial public use. 35 U. S. C. §200. The Act helps assure that commercialization (while “promot[ing] free competition” and “protect[ing] the public,” ibid. ) by imposing a set of conditions upon the federal funds recipient, by providing that sometimes the Government will take direct control of the patent rights, and by adding that on occasion the Government will permit the individual inventor to retain those rights. §§202–203.

     Given this basic statutory objective, I cannot so easily accept the majority’s conclusion—that the individual in-ventor can lawfully assign an invention (produced by public funds) to a third party, thereby taking that invention out from under the Bayh-Dole Act’s restrictions, conditions, and allocation rules. That conclusion, in my view, is inconsistent with the Act’s basic purposes. It may significantly undercut the Act’s ability to achieve its objectives. It allows individual inventors, for whose invention the public has paid, to avoid the Act’s corresponding restrictions and conditions. And it makes the commercialization and marketing of such an invention more difficult: A potential purchaser of rights from the contractor, say a university, will not know if the university itself possesses the patent right in question or whether, as here, the individual, inadvertently or deliberately, has previously assigned the title to a third party.

     Moreover, I do not agree that the language to which the majority points—the words “invention of the contractor” and “retain”—requires its result. As the majority concedes, Stanford’s alternative reading of the phrase “ ‘in-vention of the contractor’ ” is “plausible enough in the abstract.” Ante , at 10. Nor do I agree that the Act’s lack of an explicit provision for “an interested third party” to claim that an invention was not the result of federal funding “bolsters” the majority’s interpretation. Ante , at 13. In any event, universities and businesses have worked out ways to protect the various participants to research. See Brief for Association of American Universities et al. as Amici Curiae 22–24 (hereinafter AAU Brief); App. 118–124 (Materials Transfer Agreement between Cetus and Stanford University).

     Ultimately, the majority rejects Stanford’s reading (and the Government’s reading) of the Act because it believes that it is inconsistent with certain background norms of patent law, norms that ordinarily provide an individual inventor with full patent rights. Ante , at 10 . But in my view, the competing norms governing rights in inventions for which the public has already paid, along with the Bayh-Dole Act’s objectives, suggest a different result.


     There are two different legal routes to what I consider an interpretation more consistent with the statute’s objectives. First, we could set aside the Federal Circuit’s interpretation of the licensing agreements and its related licensing doctrine. That doctrine governs interpretation of licensing agreements made before an invention is conceived or reduced to practice. Here, there are two such agreements. In the earlier agreement—that between Dr. Holodniy and Stanford University—Dr. Holodniy said, “I agree to assign … to Stanford … that right, title and interest in and to … such inventions as required by Contracts and Grants.” App. to Pet. for Cert. 119a (emphasis added). In the later agreement—that between Dr. Holodniy and the private research firm Cetus—Dr. Holodniy said, “I will assign and do hereby assign to Cetus, my right, title, and interest in” here relevant “ideas” and “inventions.” Id., at 123a (emphasis added; capitalization omitted).

     The Federal Circuit held that the earlier Stanford agreement’s use of the words “agree to assign,” when compared with the later Cetus agreement’s use of the words “do hereby assign,” made all the difference. It concluded that, once the invention came into existence, the latter words meant that the Cetus agreement trumped the earlier, Stanford agreement. 583 F. 3d 832, 841–842 (CA Fed. 2009). That, in the Circuit’s view, is because the latter words operated upon the invention automatically, while the former did not. Quoting its 1991 opinion in FilmTec Corp. v. Allied-Signal, Inc. , 939 F. 2d 1568, 1572, the Circuit declared that “ ‘[o]nce the invention is made and [the] application for [a] patent is filed, … legal title to the rights accruing thereunder would be in the assignee [ i.e. , Cetus] … , and the assignor-inventor would have nothing remaining to assign.’ ” 583 F. 3d, at 842.

     Given what seem only slight linguistic differences in the contractual language, this reasoning seems to make too much of too little. Dr. Holodniy executed his agreement with Stanford in 1988. At that time, patent law appears to have long specified that a present assignment of future inventions (as in both contracts here) conveyed equitable, but not legal, title. See, e.g., G. Curtis, A Treatise on the Law of Patents for Useful Inventions §170, p. 155 (3d ed. 1867) (“A contract to convey a future invention … cannot alone authorize a patent to be taken by the party in whose favor such a contract was intended to operate”); Comment, Contract Rights as Commercial Security: Present and Future Intangibles, 67 Yale L. J. 847, 854, n. 27 (1958) (“The rule generally applicable grants equitable enforcement to an assignment of an expectancy but demands a further act, either reduction to possession or further assignment of the right when it comes into existence”).

     Under this rule, both the initial Stanford and later Cetus agreements would have given rise only to equitable interests in Dr. Holodniy’s invention. And as between these two claims in equity, the facts that Stanford’s contract came first and that Stanford subsequently obtained a postinvention assignment as well should have meant that Stanford, not Cetus, would receive the rights its contract conveyed.

     In 1991, however, the Federal Circuit, in FilmTec , adopted the new rule quoted above—a rule that distinguishes between these equitable claims and, in effect, says that Cetus must win. The Federal Circuit provided no explanation for what seems a significant change in the law. See 939 F. 2d, at 1572. Nor did it give any explanation for that change in its opinion in this case. See 583 F. 3d, at 841–842. The Federal Circuit’s FilmTec rule undercuts the objectives of the Bayh-Dole Act. While the cognoscenti may be able to meet the FilmTec rule in future contracts simply by copying the precise words blessed by the Federal Circuit, the rule nonetheless remains a technical drafting trap for the unwary. See AAU Brief 35–36. But cf. ante , at 15 (assuming ease of obtaining effective assignments). It is unclear to me why, where the Bayh-Dole Act is at issue, we should prefer the Federal Circuit’s FilmTec rule to the rule, of apparently much longer vintage, that would treat both agreements in this case as creating merely equitable rights.

     At the same time, the Federal Circuit’s reasoning brings about an interpretation contrary to the intention of the parties to the earlier, Stanford, contract. See App. to Pet. for Cert. 120a (provision in Stanford contract promising that Dr. Holodniy “will not enter into any agreement creating copyright or patent obligations in conflict with this agreement”). And it runs counter to what may well have been the drafters’ reasonable expectations of how courts would interpret the relevant language.

     Second, we could interpret the Bayh-Dole Act as ordinarily assuming, and thereby ordinarily requiring, an assignment of patent rights by the federally funded employee to the federally funded employer. I concede that this interpretation would treat federally funded employees of contractors (subject to the Act) differently than the law ordinarily treats private sector employees. The Court long ago described the latter, private sector principles. In United States v. Dubilier Condenser Corp. , 289 U. S. 178 (1933) , the Court explained that a “patent is property, and title to it can pass only by assignment.” Id., at 187. It then described two categories of private sector employee-to-employer assignments as follows: First, a person who is

“employed to make an invention, who succeeds, during his term of service, in accomplishing that task, is bound to assign to his employer any patent obtained.” Ibid.

But, second,

“if the employment be general, albeit it cover a field of labor and effort in the performance of which the employee conceived the invention for which he obtained a patent, the contract is not so broadly construed as to require an assignment of the patent.” Ibid.

The Court added that, because of “the peculiar nature of the act of invention,” courts are “reluctan[t] … to imply or infer an agreement by the employee to assign his patent.” Id., at 188. And it applied these same principles governing assignment to inventions made by employees of the United States. Id., at 189–190.

     Subsequently, however, the President promulgated Executive Order 10096. Courts have since found that this Executive Order, not Dubilier, governs Federal Government employee-to-employer patent right assignments. See, e.g., Kaplan v. Corcoran , 545 F. 2d 1073, 1076–1077 (CA7 1976); Heinemann , 796 F. 2d, at 455–456; Wright v. United States , 164 F. 3d 267, 269 (CA5 1999); Halas v. United States , 28 Fed. Cl. 354, 364 (1993). The Bayh-Dole Act seeks objectives roughly analogous to the objectives of the Executive Order. At least one agency has promulgated regulations that require Bayh-Dole contractors to insist upon similar assignments. See NIH Policies, Procedures, and Forms, A “20–20” View of Invention Reporting to the National Institutes of Health (Sept. 22, 1995) (available in the Clerk of Court’s case file) (requiring a Government contractor, such as Stanford University, to “have in place employee agreements requiring an inventor to ‘assign’ or give ownership of an invention to the organization upon acceptance of Federal funds,” as the Bayh-Dole Act “require[s]”). And an amicus brief, filed by major associations of universities, scientists, medical researchers, and others, argues that we should interpret the rules governing assignments of the employees at issue here (and consequently the Act’s reference to “inventions of the con-tractor”) in a similar way. AAU Brief 5–14.

     The District Court in this case adopted roughly this approach. 487 F. Supp. 2d 1099, 1118 (ND Cal. 2007) (“[A]lthough title still vests in the named inventor, the inventor remains under a legal obligation to assign his interest either to the government or the nonprofit contractor unless the inventor acts within the statutory framework to retain title”). And since a university often enters into a grant agreement with the Government for a researcher’s benefit and at his request, see J. Hall, Grant Management 205 (2010), implying such a presumption in favor of compliance with the grant agreement, and thus with the Bayh-Dole Act, would ordinarily be equitable.


     As I have suggested, these views are tentative. That is because the parties have not fully argued these matters (though one amicus brief raises the license interpretation question, see Brief for Alexander M. Shukh as Amicus Curiae 18–24, and at least one other can be read as supporting something like the equitable presumption I have described, see AAU Brief 5–14). Cf. ante , at 5, n. 2. While I do not understand the majority to have foreclosed a similarly situated party from raising these matters in a future case, see ibid., I believe them relevant to our efforts to answer the question presented here. Consequently, I would vacate the judgment of the Federal Circuit and remand this case to provide the parties with an opportunity to argue these, or related, matters more fully.

     Because the Court decides otherwise, with respect, I dissent.