Do Not Delete 1/4/2013 6:39 PM ARTICLE THE END OF AN EPITHET? AN EXPLORATION OF THE USE OF LEGAL SCHOLARSHIP IN INTELLECTUAL PROPERTY DECISIONS Lee Petherbridge∗ & David L. Schwartz∗∗ ABSTRACT The argument that the Federal Circuit is an insular court lacking in intellectual curiosity has invaded academic discourse about the court and its jurisprudence. It is rarely used as a compliment. It is rather used as an epithetic description intended to have broad explanatory power for nearly any of the many and glaring limitations that speakers and commentators contend exist in the court’s body of law. The empirical basis for the argument comes from a single finding: that between 1996 and 2000 the Second and Ninth Circuit Courts of Appeals referenced legal scholarship in their decisions in trademark and copyright cases nearly four times as often as the Federal Circuit referenced legal scholarship in its decisions in patent cases. While scholars and others have relied upon this finding to assert that the Federal Circuit’s performance is intellectually deficient, there are reasons to question whether that inference is proper. For instance, it stands in contrast to other observations that suggest ∗ Professor of Law and Richard A. Vachon, S.J. Fellow, Loyola Law School Los Angeles. ∗∗ Associate Professor of Law, Chicago-Kent College of Law. For helpful comments and suggestions, we thank Colleen Chien, Kevin Collins, Mark Patterson, Kathy Strandburg, Oskar Liivak, Kristin Osenga, David Olson, Sapna Kumar, Greg Vetter, Paul Janicke, Michael Olivas, Jason Rantanen, and Derek Simpson. We thank the Institute for Intellectual Property & Information Law (IPIL) at the University of Houston Law Center, Meg Boulware, and the other sponsors of the Institute’s National Conference in Santa Fe. We also thank Paul Cohen, Crystal Coveney, Laura Foody, Anderson S. Jonas, Lauren Locke, Matt Robinson, Gagik Stepanyan, Jacek Wnuk, Rose Ohanesian, and Richard Tuminello for excellent research assistance. 523 Do Not Delete 524 1/4/2013 6:39 PM HOUSTON LAW REVIEW [50:2 the Federal Circuit behaves a lot like the regional circuits when it comes to its relationship to legal scholarship. In this study, we explore the U.S. Supreme Court’s use of legal scholarship in its patent, trademark, and copyright jurisprudence. We report evidence of striking differences in the use of scholarship that depend on whether a decision concerns patent, trademark, or copyright law. The Court uses scholarship in patent decisions at a rate of 28%, but uses it around twice as often in its copyright (67%) and trademark (53%) decisions. While, moreover, the rates at which the Court uses legal scholarship in its copyright and trademark decisions are some of the highest observed in the Court’s entire jurisprudence, the rate at which the Court uses legal scholarship in its patent decisions is among the lowest observed. This Article discusses the significance of these observations to the empirical basis for the epithetic argument that the Federal Circuit is more insular and lacking in intellectual curiosity than other circuit courts of appeals. TABLE OF CONTENTS I. INTRODUCTION & BACKGROUND ........................................... 525 II. DESIGN.................................................................................. 534 III. METHODOLOGY ..................................................................... 537 IV. RESULTS & DISCUSSION ....................................................... 539 A. The Court’s Use of Scholarship in IP Jurisprudence ............................................................... 539 B. IP Jurisprudence Compared to Other Decision Types ............................................................... 541 C. Cross-Sectional Concerns .............................................. 542 D. Implications .................................................................. 548 1. Differential Use of Legal Scholarship May Not Determine a Court’s Intellectual Range or Openness ................................................. 548 2. Other Reasons Patent Decisions Might Draw Less Judicial Use of Scholarship than Trademark and Copyright Decisions..................... 549 3. Why do Trademark and Copyright Decisions Appear to Be Outliers? ........................................... 556 4. Other Explanations & Concerns ............................ 558 V. THE END OF AN EPITHET, FOR NOW? .................................... 560 VI. CONCLUSION ......................................................................... 561 Do Not Delete 1/4/2013 6:39 PM 2012] THE END OF AN EPITHET? I. 525 INTRODUCTION & BACKGROUND Ten years ago, Craig Allen Nard presented a paper to an audience at the University of Houston’s Institute for Intellectual Property and Information Law symposium in Santa Fe, New Mexico, that when published would be called Toward a Cautious Approach to Obeisance: The Role of Scholarship in Federal Circuit 1 Patent Law Jurisprudence. (“Nard” or “Toward a Cautious Approach to Obeisance”). The paper went on to become one of the most influential studies of the Federal Circuit published over the last decade. At its core, the paper is perhaps a call for Federal Circuit judges to use more empirical literature in the legal decisional process.2 However, while the paper contains an informative discussion about the value that empirical scholarship and economic scholarship might bring to the legal decisional process in patent cases, and an equally informative discussion about the limitations of applying such legal scholarship to the process, it is rarely relied upon for this content. Instead it is relied upon for a single observation: that between 1996 and 2000, “the Second and Ninth Circuits [in trademark and copyright decisions] cite scholarship roughly four times as often as the Federal Circuit [does in its patent decisions].”3 Professor Nard offered several explanations for this finding. One is that the comparatively high volume of patent cases decided by the Federal Circuit might mean that it “is more familiar and comfortable with patent law than the Ninth and Second Circuits are with trademark and copyright law.”4 Perhaps, therefore, these regional circuits’ lack of familiarity with trademark and copyright law “may lead to greater reliance on secondary authority.”5 He also speculated that perhaps Second and Ninth Circuit judges are more often drawn from the academy and thus “come from a culture that is more receptive to academic legal scholarship.”6 Another explanation suggested was that there is something different about trademark and copyright scholarship, viz. perhaps there was more of it from which to choose, that might explain the observed differences between the Federal Circuit and the Second and Ninth Circuits.7 1. Craig Allen Nard, Toward a Cautious Approach to Obeisance: The Role of Scholarship in Federal Circuit Patent Law Jurisprudence, 39 HOUS. L. REV. 667 (2002). 2. Id. at 684. 3. Id. at 682–83. 4. Id. at 683. 5. Id. 6. Id. 7. Id. Do Not Delete 526 1/4/2013 6:39 PM HOUSTON LAW REVIEW [50:2 Although no further empirical research was conducted, Professor Nard’s finding has been often cited by other works, many of which stretch and amplify its significance—often brushing aside Professor Nard’s cautious description of his 8 results in the process. Perhaps the most important amplification of the finding can be found in the National Research Council of the National Academies publication A Patent System for the 21st Century. Relying on Toward a Cautious Approach to Obeisance, it asserts: “[S]pecialized institutions have insular tendencies. . . . [T]he Federal Circuit appears to rely less on independent scholarly analysis, even legal scholarship, than the regional generalist appeals courts.”9 Although it mentions the concern that the observation in Toward a Cautious Approach to Obeisance might be susceptible to a different interpretation, it gives the concern no weight, going on to recommend that the Federal Circuit seek briefings that “draw upon insights from . . . legal scholarship on the patent system, and the growing body of patent-related economics literature.”10 Others have repeated and reinforced this view. For example, a recently published essay expresses the concern that “empirical research has show[n] that the Federal Circuit is less likely than other courts to cite scholarship in its opinions.”11 Professor Nard’s finding has been used to support a wide variety of claims about the Federal Circuit and its jurisprudence. Below we set forth some examples. Our purpose here is to provide some illustration of how Nard’s finding is employed as a rhetorical tool—to show how the work has been modalized—not to name and blame the works we use as examples, nor to evaluate the broader arguments presented in those works. For present purposes we are agnostic on that topic. In the first example, the finding is used to explain why the Federal Circuit did not articulate an analytical step in an opinion: 8. Id. at 684–85 (expressing concern that the benefits of using empirical research not be overstated); id. at 687–89 (describing some of the difficulties the court might encounter in deciding which research to treat as persuasive). 9. COMM. ON INTELLECTUAL PROP. RIGHTS IN THE KNOWLEDGE-BASED ECON., NAT’L RESEARCH COUNCIL OF THE NAT’L ACADEMIES, A PATENT SYSTEM FOR THE 21ST CENTURY 86 (Stephen A. Merrill et al. eds., 2004) (footnote omitted), available at http://www.nap.edu/html/patentsystem/0309089107.pdf. 10. Id. (applying Nard’s finding as a general recommendation for briefs before the court). 11. John R. Thomas, Academics and the Federal Circuit: Is There a Gulf and How Do We Bridge It?, 15 FED. CIR. B.J. 315, 315–16 (2005) (citing Nard’s 2002 finding). Do Not Delete 2012] 1/4/2013 6:39 PM THE END OF AN EPITHET? 527 Although it may be implied . . . the court does not identify the particular patent theory it is trying to effectuate. Such a failure to identify the preferred patent theory is not surprising, considering the Federal Circuit historically has shown a lack of interest in patent law scholarship. See Craig Allen Nard, Toward a Cautious Approach to Obeisance: The Role of Scholarship in Federal Circuit Patent Law Jurisprudence, 39 HOUS. L. REV. 667, 678–81 (2002) (demonstrating that scholarship is rarely cited by 12 the Federal Circuit). In another example, the finding is used to support the following argument: Some evidence of the court’s reluctance to take on the mantle of policy leadership can be found in Craig Nard’s recent study demonstrating that the Federal Circuit pays far less attention to legal and economic scholarship than 13 other circuit courts. In another example, the finding is used to support the argument: The end result is a growing skepticism about the court’s ability to experiment successfully, to adapt its jurisprudence to changing scientific norms, and to develop a common law that accurately reflects the patent system’s varied role in fostering technological innovation. This constraint becomes more pronounced in light of the court’s singularness, and also in light of its reluctance to engage the empirical and social science literature on patent law as a way to offset its relative institutional disconnectedness from the various technological communities its decisions 14 affect. Nor are patent scholars the only ones who have picked up on the power of the finding that between 1996 and 2000 the Second 12. Christopher A. Cotropia, Patent Claim Interpretation Methodologies and Their Claim Scope Paradigms, 47 WM. & MARY L. REV. 49, 116 n.353 (2005). 13. Dan L. Burk & Mark A. Lemley, Policy Levers in Patent Law, 89 VA. L. REV. 1575, 1671 (2003). “Nard’s study shows that the Federal Circuit cites less scholarship than other courts with significant responsibility for intellectual property matters.” Id. at n.365. 14. Craig Allen Nard & John F. Duffy, Rethinking Patent Law’s Uniformity Principle, 101 NW. U. L. REV. 1619, 1647–48 (2007) (footnotes omitted). Note 102, appended to the word “affect,” states: See Craig Allen Nard, Toward a Cautious Approach to Obeisance: The Role of Scholarship in Federal Circuit Patent Law Jurisprudence, 39 HOUS. L. REV. 667, 676–83 (2002) (setting forth empirical research demonstrating that, in their intellectual property cases, the other circuits cite scholarship roughly four times more frequently than does the Federal Circuit in its patent cases). Id. at 1648 n.102. Do Not Delete 528 1/4/2013 6:39 PM HOUSTON LAW REVIEW [50:2 and Ninth Circuits cited more scholarship in their trademark and copyright decisions than the Federal Circuit cited in its patent decisions. The Computer & Communications Industry Association had this to say in its Brief to the Supreme Court in eBay Inc. v. MercExchange, LLC: Through ritual invocation of the patent incentive while ignoring external precedent, scholarly literature and empirical economic evidence, the Federal Circuit has proved not only a booster of patents but a guardian of a lawyer’s patent system, narrowly focused on the individual case but blind to the effects of patents on particular fields and 15 indifferent to aggregate economic impact. Note 3, from the quote, states: The Federal Circuit’s insularity is confirmed by research showing its reluctance to cite scholarship or even judicial decisions other than its own. Craig A. Nard, Toward a Cautious Approach to Obeisance: The Role of Scholarship in Federal Circuit Patent Law Jurisprudence, 39 HOUS. L. 16 REV. 667, 678–81 (2002). Two years later, the Computer & Communications Industry Association offered the following argument in its Brief to the Federal Circuit in In re Bilski: This problem [a lack of a high regard for the public interest in the patent system’s institutions] is aggravated by the fact that the Court has also developed a reputation for insularity, which compounds the limitations of an adversarial process 17 focused on the competing interests of the two parties. Note 18 states: See Craig A. Nard, The Future of Patent Law: Toward a Cautious Approach to Obeisance: The Role of Scholarship in Federal Circuit Patent Law Jurisprudence, 39 HOUS. L. REV. 667, 692 (2002). See also John R. Thomas, Academics and the Federal Circuit: Is There a Gulf and How Do We Bridge 18 It? 15 FED. CIR. B.J. 315 (2005-2006). At the time of the writings mentioned above, Toward a Cautious Approach to Obeisance represented the only reported 15. Brief for Amicus Curiae Computer & Commc’ns Indus. Ass’n in Support of Petitioners at 5, eBay Inc. v. MercExchange, LLC, 547 U.S. 388 (2006) (No. 05-130) (footnote omitted). 16. Id. at n.3. 17. Brief Amicus Curiae of Computer & Commc’ns Indus. Ass’n in Support of Appellee Director of the U.S. Patent and Trademark Office and Urging Affirmance at 18, In re Bilski, 545 F.3d 943 (Fed. Cir. 2008) (No. 2007-1130). 18. Id. at n.18. The second article cited relies on Nard. See Thomas, supra note 11, at 315–16. Do Not Delete 2012] 1/4/2013 6:39 PM THE END OF AN EPITHET? 529 observations of the Federal Circuit’s use of scholarship. Professor Nard’s novel, yet limited, research had come to be treated as uncontested proof of a broader epithetic19 argument capable of general application, namely, that the performance of Federal Circuit judges is intellectually deficient.20 Nor can the pervasiveness of the argument be underestimated. Today it can often be observed at meetings of patent academics, and many who read this Article will be aware of the frequency with which it is raised to explain this or that dissatisfaction with the Federal Circuit’s jurisprudence.21 The hegemony of this argument is concerning. The reasons are pretty obvious. To begin with, it encourages disrespect for the court and for the judges personally. While legal scholars have an obligation to be skeptical and critical of the governing paradigms and institutions that define the patent system, the kind of hostility the epithetic argument engenders is unnecessary. Legal academics do not need to proceed from a posture that judges are intellectually uninformed to fulfill their obligations as critical observers.22 A debate over the performance of the Federal Circuit can proceed without postulating that the judges are not really sharp enough to be in the game. 19. Epithetic is used in its technical sense; in this case, pertaining to a phrase applied to a thing to describe an actual or attributed quality. Epithets may, but need not be used to derogate or diminish. Describing the argument as epithetic therefore seems appropriate. One may earnestly believe that the Federal Circuit is “insular and lacking in intellectual curiosity” and wield the argument in a purely descriptive sense. One might also use the argument with the purpose to derogate or diminish, or perhaps even more invectively, or somewhere in between. In any case, the term is appropriate. This Article is agnostic on the absolute truth of the argument that the Federal Circuit is “insular and lacking in intellectual curiosity.” Its purpose is much more modest, an examination of the argument’s published empirical support. 20. See Nard & Duffy, supra note 14, at 1647–48 (suggesting that the Federal Circuit’s tendency to forego legal scholarship contributes to its “relative institutional disconnectedness from the various technological communities its decisions affect”). 21. Rochelle Cooper Dreyfuss, The Federal Circuit: A Continuing Experiment in Specialization, 54 CASE W. RES. L. REV. 769, 772 & n.11 (2004) (reporting that practitioners and academics raised concerns over the Federal Circuit’s use of scholarship when she observed them in connection with a reexamination of her seminal work about the Federal Circuit). 22. Nor is this sort of argument easily cabined to just the Federal Circuit’s patent jurisprudence. Since most circuit courts of appeals rarely cite legal scholarship (and the district courts likely even less), this argument can be easily ported to vilify or disparage any court or judge. See, e.g., David L. Schwartz & Lee Petherbridge, Legal Scholarship and the United States Court of Appeals for the Federal Circuit: An Empirical Study of a National Circuit, 26 BERKELEY TECH. L.J. 1561, 1571, 1578 (2011) [hereinafter Schwartz & Petherbridge, Federal Circuit] (reporting an in-depth study of the use of legal scholarship by the Federal Circuit and the regional circuits); David L. Schwartz & Lee Petherbridge, The Use of Legal Scholarship by the Federal Courts of Appeals: An Empirical Study, 96 CORNELL L. REV. 1345, 1351–52, 1357 (2011) [hereinafter Schwartz & Petherbridge, Federal Courts of Appeals] (showing the use of scholarship across all the circuits over a 50-year period). Do Not Delete 530 1/4/2013 6:39 PM HOUSTON LAW REVIEW [50:2 The epithetic argument also fosters public disrespect for the patent system, which presents a similar set of concerns. Legal scholars should be critical of courts, of the patent office, and of the whole idea of the patent system. But when the criticism consists of characterizing the individuals charged with overseeing the administration of the patent system—at least the parts of the system that judges can oversee—as lacking in intellectual firepower, it might encourage private actors to perceive the patent system in ways that are socially harmful. For example, it might encourage private actors to be less willing to respect patents and innovation. Genuine innovators might be harmed because some set of users of patented technology that should license become marginally more willing to ignore a patent and risk litigation, rather than license. Moreover, when the patent system is characterized as poorly managed and out of control, that sort of characterization might encourage the perception that nearly any behavior is acceptable, including, for example, very speculative and aggressive theories of patent enforcement. None of the preceding, of course, means that problems in the patent system should not be ruthlessly discovered and illuminated. They should. If the patent system is poorly managed and out of control, the argument for why that is so should be made. The problem of the epithetic argument is that it suppresses useful arguments and observations about limitations in the patent system and about the Federal Circuit because it encourages intellectual laziness on the part of observers. It is much easier to argue that “I’m right because the institutions I’m criticizing are intellectually limited,” than it is to actually think deeply about what the real issues are and how the performance of the institutions at issue have impacted those issues. In this vein, it seems mostly to be the case that the epithetic argument is an add-on—it is extra weight on the scale. It is not, so far as we have seen, ever the entire basis for the argument criticizing the patent system and its institutions. This is still, however, somewhat troublesome because the tactic can encourage the belief in a reader (or hearer) that the argument to which the epithetic argument has been appended is stronger than it really is.23 A final concern to raise here about the epithetic argument is that it may be sufficiently hegemonic to chill the expression of young legal scholars. This point mostly makes itself. Ask 23. In fact, the hegemonic quality of the epithetic argument may be such that arguments made about the patent system are given its benefit even when the writer or speaker makes no mention of it. Do Not Delete 2012] 1/4/2013 6:39 PM THE END OF AN EPITHET? 531 yourself when the last time was that you heard an untenured legal scholar say anything even ambivalent about the Federal Circuit, much less anything positive. Have you heard young scholars at meetings argue that the court has worked a “good” development? Or even an “interesting” one? If the answer is, “It’s been a long time,” then the next question for a patent system-knowledgeable observer would be, “Is it really the case that there have been no developments affecting the patent law or the patent system, which might be laid at the doorstep of the Federal Circuit, that are worthy of being made a topic of study by a young legal academic interested in the patent system, and which could be so characterized?” One might have thought the epithetic argument already adequately confronted. In late 2010, Schwartz & Petherbridge reported an in-depth study of the use of legal scholarship by the Federal Circuit and the regional circuits.24 The study used a novel and large dataset that extends from 1990 until 200825 across all circuit level appellate courts in the federal system. Considering various measures, such as proportion of reported opinions citing legal scholarship, median citation per circuit court judge, and mean citation per circuit court judge, it observed that the Federal Circuit is not an outlier among circuit courts in its use of scholarship.26 A line of argument emerged in response to these observations, however, and to the interpretation given to them by Schwartz & Petherbridge, which was that because the Federal Circuit uses legal scholarship like the other circuit courts of appeals it was not appropriate to infer based on use of legal scholarship that the Federal Circuit was any more intellectually limited than any other circuit court.27 Some maintained that the Nard finding was still appropriate for the inference because of the way Schwartz & Petherbridge directly compared the circuit courts of appeals.28 In particular, it was argued that Nard showed it is the Federal Circuit’s use of legal scholarship in patent cases that is the problem, so demonstrating—as Schwartz and Petherbridge did—that the Federal Circuit uses legal scholarship across its entire body of 24. Schwartz & Petherbridge, Federal Circuit, supra note 22, at 1571. 25. Id. at 1576–77. The Nard finding examined a period between 1996 and 2000. Nard, supra note 1, at 678–81. 26. Schwartz & Petherbridge, Federal Circuit, supra note 22, at 1578–82. 27. Id. at 1582. 28. These arguments were given in response to presentations of Schwartz & Petherbridge at various meetings. Do Not Delete 532 1/4/2013 6:39 PM HOUSTON LAW REVIEW [50:2 jurisprudence in similar amounts to regional courts does not address that specific concern.29 In short, the argument stated that the low use of legal scholarship happens in patent cases; having not carved out patent cases, the use of legal scholarship in such cases might still be low and the inference might still be justified. To address this, Schwartz & Petherbridge collected more data categorizing patent cases at the Federal Circuit as well as civil cases (non-criminal cases) at the circuit courts. This approach again revealed that the Federal Circuit did not seem to be an outlying circuit court (it also indicated that the Federal Circuit uses scholarship somewhat more in its patent jurisprudence than it does in the remainder of its jurisprudence).30 Even when patent cases were compared to regional circuit courts’ civil–only jurisprudence, the Federal Circuit’s use of legal scholarship seems to be within a range defined by the other circuit courts of appeals.31 In response, some maintained that the comparison of Federal Circuit patent cases to regional circuit civil cases was only modestly informative, due to the intuition that regional circuits presumably had many civil cases that were very easy and would thus not provoke the use of legal scholarship.32 Taken together, Nard, and Schwartz & Petherbridge put the meaning of the finding: “[T]hat the Second and Ninth Circuits [in trademark and copyright decisions] cite scholarship roughly four times as often as the Federal Circuit [does in its patent decisions]” 33 into question. Is it properly modalized as the “Federal Circuit pays far less attention to legal and economic scholarship than other 34 circuit courts,” or is that sort of qualification not quite on target? Might the finding mean little more than the Federal Circuit does not have direct statutory jurisdiction over copyright and trademark infringement claims; e.g., is it possible “that there is something about trademark and copyright cases (and relevant scholarship) . . . that argues for differential use of scholarship in these areas than is normative for other areas of law”?35 29. Id. at 1586–89. 30. Id. at 1587–88. 31. Id. at 1588. 32. Id. at n.74. 33. Nard, supra note 1, at 683. 34. E.g., Burk & Lemley, supra note 13, at 1671. 35. Schwartz & Petherbridge, Federal Circuit, supra note 22, at 1570 (noting that the results “might suggest that trademark and copyright cases in the regional circuits, rather than patent cases at the Federal Circuit, are the outliers in terms of use of scholarship”). Do Not Delete 2012] 1/4/2013 6:39 PM THE END OF AN EPITHET? 533 The central normative concern of this Article is whether Nard’s empirical finding is good evidence for the epithetic argument that the Federal Circuit is an insular court that lacks intellectual curiosity, and by extension, whether it is good evidence for all of the various arguments and claims that rely on the epithetic argument either entirely or in part for their validity. We note that we could have approached this normative concern without an empirical study. A more purely analytical approach could expose many of the limitations of the epithetic argument. The logical connection between differential citation to scholarship and intellectual disconnectedness is hardly a straight line. But such work has already been done, starting with Nard himself. It is our sense that too few people have listened, or at least too few have been encouraged, to critically evaluate the connection between the empirical evidence and the epithetic argument. We hope that by adding some observational evidence to the discussion we will encourage the patent community to view the significance of Nard’s finding—and our own—more critically than it might have before. It should also be clear then that we did not undertake this project to definitively determine the validity of the argument that the Federal Circuit is an insular court lacking in intellectual curiosity. Nor did we undertake this project to evaluate directly any of the arguments criticizing the court that rely on that argument. Our purpose is much more modest; we undertook this project in part to evaluate the published empirical support for a specific chain of reasoning:36 (1) the differential use of scholarship between the Federal Circuit and the regional circuits means that the Federal Circuit is intellectually deficient; which (2) helps to explain (and sometimes might explain completely) why many criticisms of the court’s jurisprudence are correct. If the empirical support for this chain of reasoning is weak, that does not mean the court cannot still be intellectually limited. It just means the proposition is not empirically well-supported. There may be other reasons or approaches, or empirics yet to come that could still lend support to such a proposition. We thus remain agnostic on the degree of truth in the broad argument that the court’s performance is intellectually deficient, and any arguments that rely on that argument. What we hope to encourage is a neutral reevaluation of an increasingly unquestioned belief in intellectual disconnectedness, beginning with the observational evidence used to support it. The remainder of this Article proceeds as follows: Part II develops the thinking underlying the design of the data 36. In part also because we are curious about the U.S. Supreme Court’s relationship to legal scholarship in intellectual property decisions more generally. Do Not Delete 534 1/4/2013 6:39 PM HOUSTON LAW REVIEW [50:2 collection and analysis. Part III explains the methods used to collect the data and establish the dataset. Part IV presents the results of the observations and discusses some implications. Part V finishes the Article with a brief conclusion. II. DESIGN What if the Second and Ninth Circuits also had the Federal 37 Circuit’s patent jurisdiction? Would the use of legal scholarship in Second and Ninth Circuit patent cases look the same as Nard observed in trademark and copyright cases? What if the Federal Circuit had statutory jurisdiction over copyright and trademark infringement claims, as well as the many subjects currently within its jurisdiction? Would its use of scholarship in trademark and copyright opinions mirror that in its patent opinions or would it be different? These questions are difficult to approach empirically; such jurisdictions do not exist. There is, however, one American court that presently has jurisdiction over appeals in not only cases arising under the patent laws, but also in cases arising under the trademark and copyright laws. That court is the U.S. Supreme Court. Studying the Supreme Court’s use of legal scholarship in “intellectual property” decisions might help resolve some ambiguities in the meaning of the finding in Toward a Cautious Approach to Obeisance. To begin with, it is a single court, which has several advantages for this study. The design choice of studying a single court avoids the differences in jurisdiction that exist between the Federal Circuit and the regional circuits. Next, because it is a single court the Supreme Court presents a single set of judges (usually called justices). This fact helps to avoid the impact of cultural differences that might exist between circuit courts of appeals, differences that might be explained by the many locales in which, e.g., Ninth Circuit, judges might reside, e.g., California, Idaho, Alaska, Oregon, etc., versus those where Federal Circuit judges might reside, e.g., Washington, D.C. as well as differences like what judges had for breakfast, whether they are getting divorced, and whether one or more of them is working on a book deal and thus has less time to devote to opinion writing. The Supreme Court also has broad federal subject matter jurisdiction in the sense that it hears appeals from the broad set of 37. Long ago they used to have part of it, shared of course with other regional circuits. Nard & Duffy, supra note 14, at 1644. Do Not Delete 2012] 1/4/2013 6:39 PM THE END OF AN EPITHET? 535 issues that can be heard in federal courts. In that respect, the Supreme Court may be fairly practically comparable to the Second and Ninth Circuits, and may be as close to anything we can observe that is comparable to giving the Second and Ninth Circuits the Federal Circuit’s patent jurisdiction.38 The average case, moreover, that reaches the Supreme Court might be more likely to be broadly meaningful and uncertain in outcome than appeals that are heard by the circuit courts. The reason for this is simple: unlike the circuits, the Supreme Court’s jurisdiction, for the most part, is discretionary. The Court chooses which appeals to hear; part of its job is to broadly monitor the shape of federal law, and to resolve problematic differences that may 39 arise. It might thus often base a choice to hear an appeal on a prediction about the socio-legal significance of the issues presented, viz. the greater the predicted significance might often mean a greater likelihood that the Court takes the appeal. If so, then use of legal scholarship might be more critical to an “engaged jurisprudence”—assuming it ever is—for Supreme Court decisions than it is for the average circuit level decision. Importantly, if one believes—as the epithetic argument requires—that the judicial citation to legal scholarship represents a good proxy for the quality of a court’s decisional law, and more specifically for judicial intellectual engagement and connectedness to important issues of law and policy, the Supreme Court uses legal 38. A difference that remains and which might be of analytical significance is that the Federal Circuit hears vastly more patent appeals than the regional circuits hear copyright and trademark appeals. We cannot rule out that the frequency with which a court hears a category of appeal has some bearing on the frequency with which it uses legal scholarship when it writes decisions deciding an appeal. In our data, the Supreme Court decided more patent cases (46) during the period studied than it did trademark (17) and copyright cases (24). Forty-six is more than 17 and more than 24, although not much more than their sum (41). In any event, the differences in frequency with which the Court decided cases falling into the different IP categories seems trivial when one considers that the Federal Circuit decided something like 11 times as many patent appeals as the Second and Ninth Circuits decided trademark and copyright appeals during the period Nard studied. Moreover, if anything, hearing many more appeals in one issue area might be expected to reduce the frequency on a per-decision basis with which a court uses legal scholarship. Even though over a defined period of time it might encourage more actual use. Accord Nard, supra note 1, at 681–82 (seeming to observe that the Federal Circuit used about twice as much law review type scholarship in patent decisions than the Second and Ninth Circuits used in trademark and copyright decisions between 1996–2000). 39. See Lee Petherbridge & David L. Schwartz, An Empirical Assessment of the Supreme Court’s Use of Legal Scholarship, Nw. U. L. REV. (forthcoming 2012) (manuscript at 17–18), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_ id=1884462 (presenting an empirical study examining the use of legal scholarship in Supreme Court decisions). Do Not Delete 536 1/4/2013 6:39 PM HOUSTON LAW REVIEW [50:2 scholarship in spades.40 If, therefore, one is willing to dismiss the Supreme Court as insular and intellectually disconnected, like the Federal Circuit is supposed to be, then one verges on rejecting scholarship usage in opinions as a proxy for intellectual engagement. On the one hand, if the Supreme Court is an insular and intellectually disconnected institution, then the use of legal scholarship by justices (and by analogy judges) would seem to have little evidentiary meaning to the proposition. On the other hand, if one accepts that the Supreme Court is not an insular and intellectually disconnected institution, then examining its use of legal scholarship in intellectual property, viz. patent, copyright, and trademark cases, should shine a helpful comparative light on the meaning of the differences in use of legal scholarship observed by Nard between the Second and Ninth Circuits and the Federal Circuit. Finally, the Supreme Court has acknowledged the theoretical, doctrinal, and policy kinship of at least patent law and copyright law;41 and has borrowed doctrines from patent law and grafted them into copyright law on several occasions.42 43 These similarities suggest—all else being equal—that if one examined the use of scholarship in Supreme Court patent and copyright decisions one might expect to observe equality of use. Trademark differs in what might be substantial aspects, and so we were unsure what to predict. In approaching the issue, we recognized that an assumption of the epithetic 40. Id. at 11–12 (showing the level of use of legal scholarship for all of Supreme Court jurisprudence). 41. See, e.g., Sony Corp. of Am. v. Universal City Studios, Inc., 464 U.S. 417, 439 (1984) (“The closest analogy is provided by the patent law cases to which it is appropriate to refer because of the historic kinship between patent and copyright law.”). 42. See, e.g., eBay Inc. v. MercExchange, L.L.C., 547 U.S. 388, 392 (2006) (reviewing standard for injunctions in copyright cases as part of analysis of standard for injunctions in patent cases); Metro-Goldwyn-Mayer Studios Inc. v. Grokster, Ltd., 545 U.S. 913, 936– 37 (2005) (importing patent law doctrine on inducement liability into copyright law); Sony Corp. of Am., 464 U.S. at 439–42 (importing patent law doctrine on contributory infringement into copyright law); see also Edward Lee, Freedom of the Press 2.0, 42 GA. L. REV. 309, 393–94 (2008) (noting the Supreme Court’s importation of patent law doctrines into copyright law). 43. And others: Both the patent and copyright laws are authorized by the same clause of the U.S. Constitution, which notes that the laws are intended to “promote the Progress of Science and useful Arts.” U.S. CONST. art. I, § 8, cl. 8. Both bodies of law are similar in that they are exceptions to the general policy of disfavoring monopolies, and both share the same normative underpinnings (e.g., optimizing the production of a class of public goods). HERBERT HOVENKAMP, MARK D. JANIS & MARK A. LEMLEY, IP AND ANTITRUST: AN ANALYSIS OF ANTITRUST PRINCIPLES APPLIED TO INTELLECTUAL PROPERTY LAW § 1.3, at 1–13 (Supp. 2009) (“Intellectual property, while it does not generally create a monopoly, may in some cases permit or even encourage monopoly in order to give incentives for invention.”). Do Not Delete 2012] 1/4/2013 6:39 PM THE END OF AN EPITHET? 537 argument is that patent jurisprudence should use legal scholarship at least as often as copyright and trademark jurisprudence. A complementary—and possibly animating of the epithetic argument—assumption is that “intellectual property” is a unitary body of subject matter which includes all three areas of law.44 We took the conservative route, hewing to these assumptions notwithstanding our uncertainties. Accordingly, we assume that trademark cases may have attributes, all else equal, similar to those present in patent and copyright cases that encourage a court to cite legal scholarship in a decision. Our research hypothesis, for the abovementioned reasons, was that the Supreme Court would use legal scholarship at similar levels in patent, trademark, and copyright decisions. III. METHODOLOGY We observed the rate of use of legal scholarship in Supreme Court “intellectual property” decisions, viz. decisions concerning patent, trademark, and copyright law. We also observed the Court’s use of scholarship in all other areas of its jurisprudence. The methods are substantially as described in 45 Petherbridge & Schwartz. Briefly, the dataset includes information from all Supreme Court decisions issued between 1949 until 2009. The cases were located using The Supreme Court Database,46 which contains substantial information about each Supreme Court decision.47 Decisions were human coded for use of legal scholarship, and intercoder agreement tested using Cohen’s Kappa.48 For our coding of legal 44. See, e.g., ROBERT P. MERGES, PETER S. MENELL & MARK A. LEMLEY, INTELLECTUAL PROPERTY IN THE NEW TECHNOLOGICAL AGE ix (5th ed. 2010) (showing the casebook divided into chapters for patent law, copyright law, and trademark law). 45. See generally Petherbridge & Schwartz, supra note 39, at 6–9. 46. Current Dataset, SUPREME COURT DATABASE, http://scdb.wustl.edu/ data.php (last visited Nov. 28, 2012). It is considered the “gold standard” of Supreme Court databases, and is reputed to have highly reliable coding. Several leading law and political science professors have noted that “in the law reviews, virtually no empirical study of the U.S. Supreme Court produced by political scientists fails to draw on [the database].” See, e.g., Lee Epstein, Jack Knight & Andrew D. Martin, The Political (Science) Context of Judging, 47 ST. LOUIS U. L.J. 783, 812 (2003). 47. An online codebook for each of the fields in The Supreme Court Database can be found on the website. Online Code Book, THE SUPREME COURT DATABASE, http://scdb.wustl.edu/documentation.php?s=1 (last visited Nov. 28, 2012). The documentation indicates that the coding of the variables is of high reliability. See Harold J. Spaeth, Online Code Book: Introduction, THE SUPREME COURT DATABASE, http://scdb.wustl.edu/documentation.php?s=1 (last visited Nov. 28, 2012) (showing the depth and flexibility the Supreme Court Database provides for researchers). 48. See Jacob Cohen, A Coefficient of Agreement for Nominal Scales, 20 EDUC. & Do Not Delete 538 1/4/2013 6:39 PM HOUSTON LAW REVIEW [50:2 scholarship we observed Kappas across eight coders that ranged from 1.000 (perfect) to .868 (almost perfect/excellent). Overall, the dataset includes 7,730 Supreme Court decisions, about a third of which use legal scholarship.49 Patent decisions were identified in The Supreme Court Database using issue codes 80180 and 80210. Copyright decisions were identified with issue code 80190. Trademark decisions were identified by assigning four separate coders independently the task of identifying all Supreme Court trademark decisions between 1949 and 2009. The results from all four coders were synthesized, disparities investigated, and a final list of trademark decisions determined. We did not consider a trade secret category of decisions. 50 The methods applied are a form of content analysis, and therefore implicate concerns of selection bias, unobserved reasoning, and the like.51 We define “legal scholarship” as articles published in law reviews or law journals. We have not PSYCHOL. MEASUREMENT 37, 38–40 (1960) (describing its logic of the statistical argument and defending its merits). The closer the kappa statistic is to 1.00, the greater the level of agreement. Id. at 41. There is no set k-value that signifies “good enough” agreement, but magnitude guidelines have been suggested. Richard Landis and Gary Koch suggest that k-values of 0.00–0.20 reflect “slight” agreement; 0.21–0.40 reflect “fair” agreement; 0.41– 0.60 reflect “moderate” agreement; 0.61–0.80 reflect “substantial” agreement; and 0.81– 1.00 reflect “almost perfect” agreement. J. Richard Landis & Gary G. Koch, The Measurement of Observer Agreement for Categorical Data, 33 BIOMETRICS 159, 164–65 (1977); see also JOSEPH L. FLEISS, BRUCE LEVIN, & MYUNGHEE CHO PAIK, STATISTICAL METHODS FOR RATES AND PROPORTIONS 604 (3d ed. 2003) (describing 0.40–0.75 as fair to good and over 0.75 as excellent). In this study, approximately 30% of the measurements were coded by multiple coders to establish reliability. See Lee Petherbridge & R. Polk Wagner, The Federal Circuit and Patentability: An Empirical Assessment of the Law of Obviousness, 85 TEX. L. REV. 2051, 2074 & n.118 (2007) (using a 25% sample and noting literature that suggests using at least a 10% sample). 49. Petherbridge & Schwartz, supra note 39, at 12 (reporting this in Table 1). 50. See Mark A. Hall & Ronald F. Wright, Systematic Content Analysis of Judicial Opinions, 96 CALIF. L. REV. 63, 64 (2008) (describing content analysis as a method of searching judicial opinions for consistencies in order to make inferences about a specific subject). The technique has been applied often in the study of patent law. See John R. Allison & Mark A. Lemley, The (Unnoticed) Demise of the Doctrine of Equivalents, 59 STAN. L. REV. 955, 966–76 (2007) (explaining the dataset used of district court and Federal Circuit opinions and drawing inferences from the consistencies within); Petherbridge & Wagner, supra note 48, at 2076–102 (examining cases where “the Federal Circuit affirmed, reversed, or vacated the decision of the lower tribunal on the question of obviousness” and exploring implications of those findings); David L. Schwartz, Practice Makes Perfect? An Empirical Study of Claim Construction Reversal Rates in Patent Cases, 107 MICH. L. REV. 223, 245–58 (2008) (making inferences concerning reversal rates from 485 district court and magistrate judges). It has been applied in other areas of intellectual property law as well. See Barton Beebe, An Empirical Study of U.S. Copyright Fair Use Opinions, 1978–2005, 156 U. PA. L. REV. 549, 557–58, 564–65 (2008) (analyzing a dataset of reported federal opinions employing the four-factor test from section 107 of the Copyright Act starting in 1978 up through 2005). 51. See, e.g., Hall & Wright, supra note 50, at 102–03 (discussing content analysis and its potential disadvantages). Do Not Delete 2012] 1/4/2013 6:39 PM THE END OF AN EPITHET? 539 discriminated in our definition of law reviews and law journals. In contrast to other studies of legal scholarship, we have included non-elite and specialty journals. We have excluded treatises, hornbooks, and academic books from our definition.52 We define “use” of legal scholarship to mean the Supreme Court uses legal scholarship when it cites to it in a decision. We thus employ the terms use and citation synonymously. As we have explained elsewhere, this is at once a too narrow, and perhaps, a too broad definition of the meaning of the idea of 53 use of scholarship. It seems indisputable that measuring use in this way can only provide a partial picture of what it means to say that a court uses scholarship. This definition of use, nonetheless, is ubiquitous in the field of study of legal scholarship and it has the merit of being amendable to fairly clean and objective measurement. In addition, when we speak of a Supreme Court “decision,” we mean all of the writings— the opinions—attending the decision, so opinions for the Court, concurrences, dissents, and mixed opinions. IV. RESULTS & DISCUSSION In this Part we set forth our observations, and discuss them. A. The Court’s Use of Scholarship in IP Jurisprudence Table 1 shows the rate of use of scholarship in Supreme Court patent, trademark, and copyright decisions. It shows striking differences in the use of scholarship that depend on the type of decision. Patent decisions are overwhelmingly the least likely to use scholarship, using it only 28.26% of the time.54 By contrast, the Court uses scholarship nearly twice as frequently in its 52. We exclude treatises, hornbooks, and academic books from our definition because the literature criticizing legal scholarship typically focuses on law review articles and excludes other forms of legal scholarship. Nard reported data for sources beyond just law reviews and law journals. See Nard, supra note 1, at 678 n.40. Nard’s report, however, allows for some differentiation between law reviews and journals, and other forms of scholarship. Id. at 680–81 & nn.43–45 (evaluating treatises, bar journals, and law reviews separately). As the introduction indicates, the Federal Circuit is not typically criticized for failing to cite the sort of scholarship embodied in patent treatises. Id. at 674 n.25 (explaining that the court cites treatises frequently). 53. Petherbridge & Schwartz, supra note 39, at 6 (explaining that the idea that “citation equals use” is imperfect); Schwartz & Petherbridge, Federal Courts of Appeals, supra note 22, at 1354 n.32. 54. See infra Table 1. Do Not Delete 540 1/4/2013 6:39 PM HOUSTON LAW REVIEW [50:2 trademark jurisprudence (52.94%) and more than twice as frequently in its copyright jurisprudence (66.67%).55 Our data includes the entire population of Supreme Court decisions during the period studied, and the results are, arguably, therefore statistically significant by definition. To develop a sense of the importance of the differences we observed, we assumed that our data was a sample of a larger population and considered whether the differences observed would have been significant in that scenario. Our null hypothesis was that there was no difference in use of scholarship between patent, trademark, and copyright decisions, and the statistical argument we used is the chisquare test. The chi-square calculated, df(2), is 10.2; p<.01, and is strong evidence against the null hypothesis. If our data were from a sample, instead of the entire population, the evidence suggests that the differences observed are not due to chance. Decision Type Decisions Not Citing Decisions Citing Total Decisions % of Decisions Citing Patent 33 13 46 28.26% Trademark 8 9 17 52.94% Copyright 8 16 24 66.67% Table 1. The Rate of Use of Scholarship in Supreme Court IP Decisions 1949–2009. The rate is measured in binary. If an opinion cites legal scholarship it is included; if it does not, it is excluded. The Supreme Court thus uses legal scholarship significantly more frequently in its trademark and copyright jurisprudence than 56 in its patent jurisprudence. The measured difference is, roughly, 57 twice as often. This finding, we think, adds some perspectives from which to view the finding “that the Second and Ninth Circuits [in trademark and copyright decisions] cite scholarship roughly four times as often as the Federal Circuit [does in its patent decisions].”58 We discuss some of them in the Implications Part. 55. 56. 57. 58. Infra Table 1. Supra Table 1. Supra Table 1. Nard, supra note 1, at 683. Do Not Delete 2012] 1/4/2013 6:39 PM THE END OF AN EPITHET? 541 B. IP Jurisprudence Compared to Other Decision Types The Federal Circuit has been observed to use legal scholarship somewhat more in its patent decisions than in its other decisions. We wondered if the Supreme Court was similar. The observations reported in Table 1 do not reveal whether patent decisions might merely be at the low end of a set of decisions that together evince a relatively high rate of citation to scholarship. In other words, while it is clearly the case that the Court more frequently uses legal scholarship in its trademark and copyright decisions than in its patent decisions, it was still unclear whether the Court’s use of scholarship in patent decisions might nonetheless be fairly high when compared to other areas of jurisprudence. Figure 1 describes the Court’s use of scholarship in IP decisions to its use of scholarship in 13 other decision types.59 The overall average rate of use of scholarship is 32.21%, so 60 just between the rate for Unions and Federalism decisions. The rate of use of scholarship in trademark and copyright 61 decisions are among the highest in the Court’s law. The rate at which the Court uses legal scholarship in patent decisions, however, is below the average rate of use for the entire dataset 62 and among the lowest of the decisional areas examined. Accordingly, it seems that patent decisions are not at the low end of a set of decisions that evince a relatively high rate of citation. In something of a contrast to the Federal Circuit— which uses legal scholarship a little bit more in patent cases than in the rest of its jurisprudence—the use of legal scholarship in Supreme Court patent decisions seems modestly disfavored.63 Copyright and trademark cases, by contrast, appear to be high flying outliers.64 59. See infra Figure 1 (showing the use of legal scholarship in Supreme Court decisions). 60. Petherbridge & Schwartz, supra note 39, at 5; see infra Figure 1. 61. See infra Figure 1. 62. See infra Figure 1 (showing that legal scholarship is used at a 28.3% rate for patent cases). 63. Nard, supra note 1, at 678; see infra Figure 1. 64. See infra Figure 1. Do Not Delete 1/4/2013 6:39 PM 542 HOUSTON LAW REVIEW / [50:2 Z : W & d W Z & h & W W d D Figure 1. The Rate of Use of Scholarship in Supreme Court Decisions 1949–2009. The Decision types correspond to “Issue Area” categorizations in the Spaeth dataset, http://scdb.wustl.edu/documentation.php?var=issueArea. Counting is again binary. Misc n=17; Copyright n=24; Trademark n=17; Privacy n=100; Due Process n=303; First Amendment n=629; Attorneys n=91; Economic Activity n=1477; Unions n=317; Federalism n=360; Crimpro n=1788; CivRights n=1300; Patent n=46; FedTax n=278; Judicial Power n=1098; Interstate Relations n=85. C. Cross-Sectional Concerns The dataset represents a cross section that includes sixtyone years’ worth of observations. One concern, therefore, is that our observations might be affected by, inter alia, the Court’s general attitude toward scholarship across time and the temporal distribution of decisions. For example, perhaps the Court decided most of the patent decisions at a time when its jurisprudence generally did not reflect much use of scholarship, but decided most of its copyright and trademark cases at a time when its jurisprudence reflected a much healthier appetite. Do Not Delete 2012] 1/4/2013 6:39 PM THE END OF AN EPITHET? 543 To examine this issue we employed regression, using indicator variables for term, and an indicator variable defining observations taken before and after the creation of the Federal Circuit. The purpose is to statistically account for variation in use of scholarship that is linked to term or that is linked to all terms in the dataset that succeeded the creation of the Federal Circuit. If our observations are affected by the Court’s general attitude toward scholarship across time, and the fact that patent, trademark, and copyright decisions are presumably not perfectly uniformly distributed across terms, we hope to observe some impact on the magnitude of an intellectual property category’s ability to affect the odds of a citing decision. % Impact on Odds of Use of Legal Scholarship Variable Model 1 Model 2 Model 3 Patent -17 -16 -15 Trademark 122 121 96 Copyright 311 387 288 term indicators na X na federal_circuit na na 41 indicator Table 2. Logistic regression models accounting for variation in time (term) and whether an observation was made before or after the creation of the Federal Circuit (federal_circuit). The reference term is the 1982 term, selected essentially arbitrarily. The reference category for federal_circuit is decisions issued before the 1983 term. The response variable is bincite, which identifies decisions that cite at least one piece of legal scholarship. Table 2 suggests that once we take into account the variation in citation that is linked to term, copyright decisions become a somewhat more impressive predictor of the Court’s use 65 of scholarship. In addition, it appears that observations made from the 1983 to 2009 terms (federal_circuit) do explain some of the variation in the model—these observations have a distinctly positive impact on the likelihood that a decision references legal scholarship.66 A Supreme Court decision that corresponds with 65. See supra Table 2 (displaying logistic regression models accounting for variation in time (term) and whether an observation was made before or after the creation of the Federal Circuit (federal_circuit)). 66. This is consistent with recent studies of the Supreme Court and the circuit courts of appeals. See Petherbridge & Schwartz, supra note 39, at 14; Schwartz & Do Not Delete 544 1/4/2013 6:39 PM HOUSTON LAW REVIEW [50:2 the Federal Circuit era seems somewhat more likely to use legal scholarship,67 and so being in a more recent term might help somewhat to explain why a patent, trademark, or copyright decision cited legal scholarship. As the relative relationship in the impact on the odds of a citing decision for the various IP categories stayed fairly constant, and none changed direction of effect, viz. patent decisions predict lower odds of use while trademark and copyright decisions predict higher odds of use, Table 2 may present some evidence that our observations are not explained by the Court’s general attitude toward scholarship across time and the temporal distribution of decisions.68 Regressions are fine. But IP cases make up a small part of the Court’s jurisprudence, and if a pattern exists it is nice to view it graphically. Figure 2 presents a graph of the rate of use of scholarship in the Supreme Court’s patent and copyright decisions across the last 61 years. It suggests that the Supreme Court has pretty much always been more likely to reference legal 69 scholarship in a copyright decision than in a patent decision. At least two additional things about Figure 2 are worth mentioning at this point. The first is that while the big picture presented by Figure 2 is a fairly big-looking difference that has persisted for a long time, the rate of use of scholarship in patent and copyright jurisprudence does get close at one point, the 70 decade of 1961 to 1970. It is not clear what this means. The differences observed both before and after the decade of 1961 to 1970 seem to us to be pretty pronounced. For that reason, we suspect that the best interpretation of the closeness in rate of use at this one point does not really mean very much in terms of the broader picture of comparing the Supreme Court’s use of legal scholarship across patent and copyright jurisprudence. That said, a very plausible explanation for this observation was suggested to us by the learned participants at the Institute for Intellectual Property & Information Law (IPIL) at the University of Houston Law Center’s meeting in Santa Fe. In particular, it was suggested that the period between 1961 and 1970 might show the Petherbridge, Federal Courts of Appeals, supra note 22, at 1360 (examining the use of scholarship at the circuit courts of appeals); supra Table 2 (examining the trend in use of scholarship at the Supreme Court); accord Robert J. Hume, Strategic-Instrument Theory and the Use of Non-Authoritative Sources by Federal Judges: Explaining References to Law Review Articles, 31 JUST. SYS. J. 291, 299 (2010). 67. See supra Table 2 (indicating that the creation of the Federal Circuit has positively impacted the likelihood of legal scholarship use). 68. See supra Table 2. 69. See infra Figure 2. 70. See infra Figure 2. Do Not Delete 2012] 1/4/2013 6:39 PM THE END OF AN EPITHET? 545 level of scholarship that it does because it represents the Supreme Court coming to grips with the changes in patent law imposed by the 1952 Act.71 The informed reader will know that in the decade in question, the Court decided several cases concerning a statutory requirement for patentability known as nonobviousness that had not before been part of the statutory 72 law. The implication of this observation is that perhaps one factor influencing the Court’s use of legal scholarship in these cases is new statutory enactments. A mechanism for this turns out to be fairly intuitive: statutory changes can create ambiguities in otherwise mostly settled case law. When the Court confronts these ambiguities for the first time, it is likely to be writing on more of a blank slate in terms of its precedents and other formal, legally acceptable sources of decision. If, as Richard Posner has suggested, judges and justices “feel naked before society when the positions they take on novel cases . . . are seen to reflect 73 unstructured intuition,” then perhaps justices would be more willing to resort to legal scholarship, an informal but still rigorous appearing form of authority, to strategically mask the uncomfortableness they feel in such instances. Compatible with this is the fact that the Copyright Act experienced a major revision in 1976,74 and so if new statutory enactments stimulate the use of scholarship by the Court, an expectation might be that copyright law should show higher levels of use of legal scholarship in the 70s and 80s. There have been other changes to the copyright law in 75 recent years, for example: The Audio Home Recording Act (1992), 76 the Digital Performance Right in Sound Recordings Act (1995), the No Electronic Theft Act (1997),77 not to mention, of course, the 78 Digital Millennium Copyright Act (1998), some of which could be 71. Sapna Kumar deserves special credit for pressing this point with us, which is not meant to diminish the fact that others also seemed to independently reach a similar conclusion. 72. See Graham v. John Deere Co. of Kan. City, 383 U.S. 1, 12 (1966) (including consolidated cases). A companion case is United States v. Adams, 383 U.S. 39, 48 (1966) (deciding a question of nonobviousness). 73. See, e.g., Richard A. Posner, Pragmatic Adjudication, 18 CARDOZO L. REV. 1, 11 (1996). 74. Copyright Act of 1976, Pub. L. No. 94-553, § 101, 90 Stat. 2541 (codified as amended at 17 U.S.C. §§ 101–810 (2006)). 75. Audio Home Recording Act of 1992, Pub. L. No. 102–563, 106 Stat. 4237 (1992) (codified as amended in scattered sections of 17 and 19 U.S.C.). 76. Digital Performance Right in Sound Recordings Act of 1995, Pub. L. No. 104–39, 109 Stat. 336 (1995) (codified as amended in scattered sections of 17 U.S.C.). 77. No Electronic Theft (NET) Act, Pub. L. No. 105–147, 111 Stat. 2678 (1997) (codified as amended in scattered sections of 17, 18, and 28 U.S.C.). 78. Digital Millennium Copyright Act, Pub. L. No. 105–304, 112 Stat. 2860 (1998) Do Not Delete 1/4/2013 6:39 PM 546 HOU USTON LAW REVIEW [50:2 Rate of Decisions Using Legal Scholarship helping to maintain a rrelatively high level of use of legal scholarship in Supreme Court copy yright jurisprudence. 100% 75% 50% 25% 0% Figure 2. The R Rate of Use of Scholarship in Supreme Court Patent and C Copyright Decisions by Decade, 1949– 2009. Decade referss to “terms,” viz. 1961–1970 is ten terms, the 1961 term, 1962 term, etc. Patent n=46; Copyright n=24. Table 3. The Rate o of Use of Scholarship in Supreme Court Patent and Copyrigh ht Decisions by Decade, 1949–2009. Data corresponds to that p plotted in Figure 2. Term Patent Patent Copyright Copyright Count Rate Count Rate 1949–1960 0/6 0 2/5 .40 1961–1970 6/14 .43 1/2 .50 1971–1980 4/11 .36 3/3 1 1981–1990 1/3 .33 6/6 1 1991–2000 1/4 .25 3/5 .60 2001–2009 1/8 .12 2/3 .67 (codified as amended in scatteered sections of 5, 17, 28, and 35 U.S.C.). Do Not Delete 2012] 1/4/2013 6:39 PM THE END OF AN EPITHET? 547 The second thing about Figure 2 worth mentioning at this point is that the data presented tends to make less likely the idea that perhaps the Federal Circuit somehow broke the Supreme Court’s patent jurisprudence. The underlying idea here is illustrated by the old adage: “garbage in, garbage out.” Perhaps the use of scholarship in Supreme Court decisions reflects either the use of scholarship in circuit court decisions, or perhaps the general quality of circuit court work product. Consider two hypothetical examples: If the circuit court does a wonderful job presenting and analyzing the issues in a decision on appeal, then perhaps the Supreme Court’s decision addressing the appeal will provide a more robust analysis that will include reference to legal scholarship. In the contrasting example, an intellectually limited circuit court might produce a decision on appeal that does a poor job revealing the nuances of an issue. If so, then perhaps the Supreme Court’s decision addressing the appeal will appear to be more abstract and intellectually limited because it has a poorer substrate from which to work. If this explanation is correct then our observations of the Supreme Court’s jurisprudence might not reveal much about what, if anything, differential use of legal scholarship has to say about an institution’s intellectual competence. For much of our data, the argument would go: The Federal Circuit’s deficiencies have handicapped the Supreme Court’s patent jurisprudence. Our results in this subpart provide some evidence that seems inconsistent with such a theory. In particular, the evidence suggests that the timing of decisions does not fully explain the observations. And Figure 2, in particular, suggests that the Supreme Court’s tendency to reference scholarship more frequently in its copyright decisions over its patent decisions is a tendency that long predates the creation of the Federal Circuit.79 Our results thus seem to provide little support for the notion that the Supreme Court’s performance has been hampered by the Federal Circuit.80 As a final exploration of whether the distribution of Supreme Court decisions might explain our observations, we visually examined the distribution of patent, trademark, and copyright 79. The description of Figure 2 is even more forceful if one accepts that it took some time to get the Federal Circuit up and running. For example, if one postulates that, say, between 1983 and 1990, the court might have been finding its legs, and thus, might not have cited much scholarship; an examination of Figure 2 shows an impressive-appearing gap extending back at least as far as 1970. 80. See supra Figure 2 (demonstrating that the Supreme Court’s tendency to cite legal scholarship more often when deciding copyright cases than patent cases long predated the creation of the Federal Circuit). Do Not Delete 548 1/4/2013 6:39 PM HOUSTON LAW REVIEW [50:2 decisions by term. The graph is presented as an Appendix but generally shows that patent and copyright decisions are distributed relatively evenly, while trademark decisions become more frequent starting around 1980. D. Implications What are the implications of the observations reported in Parts IV.A, IV.B, and IV.C for the meaning of the finding made by Toward a Cautious Approach to Obeisance?; more importantly perhaps, what should we think about the utility of the finding to support the epithetic argument that the performance of Federal Circuit judges is more intellectually circumscribed than other judges? In addition, the observations are puzzling. Why does the Supreme Court use legal scholarship so much more consistently in its trademark and copyright jurisprudence as compared to its patent jurisprudence? Why does the use of legal scholarship in patent decisions seem to be somewhat disfavored relative to other issue areas? In this Part, we offer some preliminary answers to these questions that may be of use to future research into patent jurisprudence. 1. Differential Use of Legal Scholarship May Not Determine a Court’s Intellectual Range or Openness. In Part II (Design) we postulated that the Supreme Court by virtue of its jurisdiction and supervision responsibilities was unlikely to be an insular and disconnected institution, like the Federal Circuit has been claimed by some to be. If that is true, then the observations in Part IV indicate that “healthy” courts can exhibit differential use of legal scholarship across patent, trademark, and copyright jurisprudence. It follows that differential citation in this set of cases may not determine whether a court is adequately intellectually connected and curious about the law, policy, and social consequences of its jurisprudence. Analogizing this logic to the circuit court level, the observation that the Federal Circuit cites legal scholarship less often in its patent jurisprudence than the Second and Ninth Circuits cite legal scholarship in their trademark and copyright jurisprudence does not mean the Federal Circuit is necessarily an intellectually circumscribed court. As a matter of purely abstract logic, it might be, and an observation of differential citation could be consonant with that determination, but the observation is inconclusive because differential citation can exist in the jurisprudence of a court that lacks the characteristic of intellectual deficiency. This interpretation might be strengthened if there are explanations for why patent decisions might less frequently draw Do Not Delete 2012] 1/4/2013 6:39 PM THE END OF AN EPITHET? 549 judicial reference to legal scholarship than trademark and copyright decisions, especially if those explanations themselves do not require the conclusion that the performance of the judges deciding patent appeals is intellectually deficient. It is to these possibilities we turn next. 2. Other Reasons Patent Decisions Might Draw Less Judicial Use of Scholarship than Trademark and Copyright Decisions. A fairly direct interpretation of these results is that there might be something different about patent decisions that set them apart from trademark and copyright decisions when it comes to the use of scholarship. What might it be? Below we suggest some possible explanations. i. Patent Law is Easier for Judicial Decisionmakers. One is that patent decisions are different because patent law is easier for judicial decisionmakers to work than trademark and copyright law. There is an assumption at work here, namely that law review-type legal scholarship may be looked to by judges to help understand complex areas of law; it must be acknowledged that this may not at all be the case. Perhaps a judge (or a clerk) is more likely to crack open a treatise when a case presents a serious challenge than to search a database for law review articles directed to a particular topic. In any event, the point to be made is that perhaps patent law is more intellectually accessible than trademark and copyright law. Perhaps it has less entropy: the symbols and concepts of patent law organize information fairly effectively, making it easier to access and easier to apply. Jurists are less inclined to look to scholarship for a patent law issue because they look at the issue and say, “Oh, I get it,” and move on to other work in other cases. Perhaps this is less true of trademark and copyright law, the symbols and concepts of which might leave more disorder and uncertainty in meaning. ii. Patent Law is Better Developed than Trademark and Copyright Law. A similar, but different, point is that patent decisions may be different because patent law is better developed. It is usually agreed that because the Federal Circuit hears so many more patent cases it is developing patent law much faster than the regional circuits are developing trademark and copyright law.81 If that is true, it could mean that there is so 81. Depending on one’s point of view, the Federal Circuit might have quickly developed it to stagnation. See Nard & Duffy, supra note 14, at 1641 (characterizing the Federal Circuit’s jurisdiction over patent law as monopolistic); Randall R. Rader, The Do Not Delete 550 1/4/2013 6:39 PM HOUSTON LAW REVIEW [50:2 much precedent, varied in content, that parties and jurists tend to reach the conclusion that the issues presented by patent cases can be resolved using traditional forms of legal analysis that focus on conventional theoretic logic and the interpretation of legally binding precedent. Perhaps much of the doctrinal nuance and leeway for policy manipulation is already built into patent law, and available in its precedents;82 judges and practitioners may find less need to look elsewhere for guidance or suggestions on how to argue or analyze an issue. Trademark and copyright law might present a contrast. As noted earlier, statutory developments might play a more pronounced role in the development of jurisprudence, which could impede judicial development of the law. Judicial development of the law in these areas might also proceed at a slower pace because jurisdiction is distributed across so many circuits. Judges have a harder time getting the repetitions they need to be highly effective at comprehending and applying the relevant law. Perhaps only a few become very highly knowledgeable, and perhaps the body of cases less often provide the blazemarks needed for judges to use conventional forms of case analysis. If so, trademark and copyright appeals might more often than patent appeals appear to be hard to resolve using traditional legal analysis, viz. applying formal legally binding precedents using analogical reasoning in the context of statutes, and might, therefore, more often encourage a resort to scholarship. iii. Patent Appeals are Less Important. Our observations could reflect that patent appeals are deemed less important by the Supreme Court (or its clerks) than copyright and trademark appeals. One theory holds that courts use legal scholarship strategically; they manipulate the content of opinions, including by manipulating the use of authority, on the hope that doing so will cause an opinion to have a desired effect on a relevant audience.83 The desired effect perhaps most likely to be in play is United States Court of Appeals for the Federal Circuit: The Promise and Perils of a Court of Limited Jurisdiction, 5 MARQ. INTELL. PROP. L. REV. 1, 3–4 (2001) (concluding that the Federal Circuit develops patent law at twenty-five times the pace of the regional circuits’ development of copyright law); Nard, supra note 1, at 683 (characterizing the Federal Circuit’s quick development of patent law as “unique”). 82. See generally Lee Petherbridge, Patent Law Uniformity?, 22 HARV. J.L. TECH. 421 (2009). 83. See, e.g., Hume, supra note 66, at 295–96 (proposing that strategic judges may utilize legal scholarship in order to persuade readers); Petherbridge & Schwartz, supra note 39, at 19 (suggesting that judges may use legal scholarship “to encourage the perception that precedent-altering change is orderly”); Kimberly West-Faulcon & Lee Do Not Delete 2012] 1/4/2013 6:39 PM THE END OF AN EPITHET? 551 the goal of winning support for the judgment reflected in the opinion. The relevant audience might include, e.g., “other current or future justices, former colleagues, legal academics, legislators, the constituency of constitutional litigators, etc.”84 It has been observed that courts are more interested in using scholarship when cases are more important or difficult to decide,85 when disagreement with the outcome evinced in a majority opinion is more vigorous,86 when ideological differences exist,87 when the court upsets settled expectations,88 and when cases involve 89 constitutional issues. Patent cases in the hands of the Supreme Court might not have always fit into these categories. It is common for us today to think of patent cases as being very important, but it might not always have been the case that they enjoyed such significance. If perhaps legal scholars or law professors failed to impress on the law students who went on to become justices that patent law is significant to the national economy, the Court might have long believed that patent cases really did not matter much, or that the perceptions of the people reading patent opinions are largely irrelevant to the Court’s dayto-day and legacy concerns.90 In addition, even though patent cases might be more important to justices today—sheer Petherbridge, Legal Scholarship in Supreme Court Constitutional Decisionmaking: An Empirical Examination 7 (forthcoming), available at http://papers.ssrn.com/ sol3/papers.cfm?abstract_id=2110347 (noting that judges may cite legal scholarship in order to bolster the perceived validity of their opinions). 84. West-Faulcon & Petherbridge, supra note 83, at 7. 85. Petherbridge & Schwartz, supra note 39, at 17, 20–21; see also Hume, supra note 66, at 308–09. 86. Hume, supra note 66, at 307–08; Petherbridge & Schwartz, supra note 39, at 17–18; West-Faulcon & Petherbridge, supra note 83, at 18–19. 87. Schwartz & Petherbridge, Federal Courts of Appeals, supra note 22, at 1366–68 (finding that citation to legal scholarship is affected by judicial ideology); West-Faulcon & Petherbridge, supra note 83, at 32–33 (observing that “the use of legal scholarship is more vigorous in opinions that promote liberal case outcomes”); accord Hume, supra note 66, at 309. 88. Petherbridge & Schwartz, supra note 39, at 20; West-Faulcon & Petherbridge, supra note 83, at 27–29. 89. Schwartz & Petherbridge, Federal Courts of Appeals, supra note 22, at 1366–67; West-Faulcon & Petherbridge, supra note 83, at 16–17; but see Hume, supra note 66, at 308–09 (finding that the presence of a constitutional issue had no significant effect on circuit courts’ citation of legal scholarship). 90. In this vein, an interesting vignette: Woodward and Armstrong’s famous book, The Brethren: Inside the Supreme Court, discusses the decisionmaking process in the notorious Sakraida v. Ag Pro, Inc. patent case, which involved the central patent doctrine of obviousness. Sakraida v. Ag Pro, Inc., 425 U.S. 273, 280–82 (1976). Woodward and Armstrong note that the Court found the case “of no significance” and the opinion was assigned as punishment to Justice Brennan. According to the book, “Brennan was insulted, but refused to pass along the humiliation to his clerks. He did all the work on the five-page opinion himself.” BOB WOODWARD & SCOTT ARMSTRONG, THE BRETHREN: INSIDE THE SUPREME COURT 418–19 (1st ed. 1979). Do Not Delete 552 1/4/2013 6:39 PM HOUSTON LAW REVIEW [50:2 speculation of course—patent cases may not be important in ways that encourage the use of legal scholarship. The Court may not be particularly divided on the outcome, and the role of ideology in Supreme Court intellectual property decisions is still being explored.91 It might also be the case that patent cases are less likely than other sorts of intellectual property cases to be decided in ways that upset settled expectations. Indeed, the Court has emphasized in its patent decisions the importance of maintaining settled expectations in patent decisions, and such an emphasis implies that the Court seeks to maintain something close to the status quo, rather than move patent law in any particular direction.92 The implications of this theory apply with additional force when generalized to the circuit court level. The Federal Circuit decided something like eleven times as many patent appeals as the Second and Ninth Circuits decided trademark and copyright appeals during the period Nard studied.93 This difference could suggest that patent cases at the Federal Circuit rarely present the sorts of forceful differences of opinion that stimulate amici, encourage dissenting votes, or upset expectations, so as to encourage judges to strategically use legal scholarship. Observers of the Federal Circuit, in fact, might reasonably conclude that most of the decisions the court issues are narrow in scope and directed mostly at the parties.94 At the circuits, by contrast, trademark and copyright cases might be more exotic. A feature, that flavored with other considerations—such as greater statutory manipulation; significance of issue; breadth of stakes involved (for example, for content owners as a group)—might disproportionately encourage the strategic use of legal scholarship. iv. Patent Law is Scary. Those who teach patent law are aware that to this day there exist the remnants of a culture that 91. See Matthew Sag, Tonja Jacobi & Maxim Sytch, The Effect of Judicial Ideology in Intellectual Property Cases 42–46 (Jan. 2007), available at http://works.bepress.com/cgi/viewcontent.cgi?article=1003&context=matthew_sag (finding that ideology matters in Supreme Court intellectual property cases, but works differently than in other areas of law). 92. See Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki Co., 535 U.S. 722, 739 (2002) (“[C]ourts must be cautious before adopting changes that disrupt the settled expectations of the inventing community.”). 93. Nard, supra note 1, at 682. 94. Which is not to say the court never addresses patent issues of broad import or issues that provoke powerful support on both sides. As the court clearly uses legal scholarship in its decisions, it may well be the case that it behaves like other courts (including the Supreme Court); opinion writers use it strategically to win support in those cases where it is necessary to try harder to win support. Do Not Delete 2012] 1/4/2013 6:39 PM THE END OF AN EPITHET? 553 preferred attorneys with technical backgrounds to other attorneys. The attitude was that if you didn’t have a technical background, you couldn’t understand patent law well, and so you might as well just stay out of it. Today that view seems archaic, except perhaps when applied to that set of young lawyers who want to devote their careers to patent prosecution. Nonetheless, if there is a general perception that patent law is inaccessible because it deals with technological issues, such a perception could discourage jurists from delving into scholarship. For similar reasons, jurists confronted with patent decisions might be more concerned about sounding unartful or being embarrassed by a mistake.95 Patent law might also be scary to jurists because it is of significance to the national economy. Jurists might be concerned that the innovation policy questions presented by patent law are at the edge of—if not beyond—a court’s institutional competence. So not only might jurists be embarrassed by freelancing too much with a decision, they might also be worried that by deciding one case too erroneously, they might provoke a parade of unintended consequences that could harm the nation’s economic competitiveness. Taken together, these concerns could encourage jurists to decide most patent cases in a fairly conservative manner, which might predict a more modest use of scholarship in decisions. By contrast, trademark and copyright law have mostly not developed the same level of exclusivity as patent law.96 Attention to secondary sources may thus be more appealing, and may produce less fear of embarrassment or failure in the minds of jurists. If jurists are less worried about making a mistake or being humiliated in deciding a case, they might be more willing to be creative in the decision, which might encourage more use, of legal scholarship. 95. Almost invariably when we give a talk about the judicial use of legal scholarship we are confronted with the claim: “It’s the clerks.” Assuming so, perhaps our observations might be explained by the Supreme Court not doing a good job of hiring clerks who have excelled in patent law. If clerks are intimidated by patent law, perhaps it affects the way they prepare materials for the justices, including the way they draft decisions. 96. See Petherbridge, supra note 82, at 472 (concluding that the Federal Circuit has developed a “small-law” framework for patent cases that emphasizes decisions of modest precedential value); see also William K. Ford, Judging Expertise in Copyright Law, 14 J. INTELL. PROP. L. 1, 3 (2006) (noting that generalist judges decide copyright and most trademark cases, while patent law is governed by a specialized panel of judges); Nard, supra note 1, at 683 (suggesting that, while the regional circuits’ diverse dockets may increase their reliance on secondary authority, the Federal Circuit may be more comfortable with patent law and thus use legal scholarship less frequently). Do Not Delete 554 1/4/2013 6:39 PM HOUSTON LAW REVIEW [50:2 v. Patent Scholarship is Not Very Interesting to Judicial Decisionmakers. Another possibility is that patent decisions are different in their use of scholarship because patent scholarship is different from trademark and copyright scholarship. Our observations might reflect the fact that legal scholarship directed to patent issues is a lot less interesting, compelling, and helpful to judges than scholarship directed to trademark and copyright issues. Perhaps, for example, when it comes to patent scholarship it is more difficult for judicial decisionmakers to separate the wheat from the chaff, and so they avoid it altogether. Perhaps patent scholarship, more than other areas, does a poor job of explaining the legal principles and arguments on which it relies. Similarly, perhaps the underlying biases, empirical limitations, economic assumptions, or other priors that inform at least some patent scholarship are detectable by judges, which causes them to avoid expressly citing work in a decision. Yet another explanation is that patent scholarship might be more than other areas of scholarship populated by “law and” and other forms of legal/social sciences scholarship that at least some federal judges claim to find useless and poorly done.97 It might also be that some patent scholarship verges sufficiently toward the abstract98 as to be unhelpful to jurists that expect to address more specific legal and policy questions within the constraints imposed by existing precedent, and in a systemic context that features substantial economic reliance on settled expectations. vi. The Patent Bar Has Dropped the Ball. Patent decisions might be different in their use of scholarship because the patent bar has done a poor job absorbing the teachings of patent 97. See, e.g., Harry T. Edwards, The Growing Disjunction Between Legal Education and the Legal Profession, 91 MICH. L. REV. 34, 35 (1992) (“I see no reason why law professors should write mediocre economics, or philosophy, or literary criticism, when arts and sciences professors could be doing a better job . . . .”). 98. See, e.g., Thomas L. Ambro, Citing Legal Articles in Judicial Opinions: A Sympathetic Antipathy, 80 AM. BANKR. L.J. 547, 549 (2006) (“When we [judges] do read the occasional article, we find it often not only unpersuasive, but even at times at odds with accepted means of analysis.”); Judith S. Kaye, One Judge’s View of Academic Law Review Writing, 39 J. LEGAL EDUC. 313, 319–20 (1989) (“Prominent law reviews are increasingly dedicated to abstract, theoretical subjects . . . and less and less to practice and professional issues . . . . I am disappointed not to find more in the law reviews that is of value and pertinence to our cases.”); David Segal, What They Don’t Teach Law Students: Lawyering, N.Y. TIMES, Nov. 20, 2011, at 22–23 (quoting Supreme Court Justice Stephen Breyer: “There is evidence that law review articles have left terra firma to soar into outer space,” and quoting Chief Justice John Roberts: “Pick up a copy of any law review that you see . . . and the first article is likely to be, you know, the influence of Immanuel Kant on evidentiary approaches in 18th-century Bulgaria, or something, which I’m sure was of great interest to the academic that wrote it but isn’t of much help to the bar.”). Do Not Delete 2012] 1/4/2013 6:39 PM THE END OF AN EPITHET? 555 scholarship. By definition, the information central to a decision in a case is presented to a court by the parties, at least in most cases. Perhaps the patent bar does a worse job than the trademark and copyright bars when it comes to taking in the teachings of legal academics. It might follow that, compared to their counterparts in the trademark and copyright bars, they are less effective at translating those teachings into legal arguments. If so, it may be that when courts address patent appeals they are less often presented with arguments expressly derived from legal scholarship, and so are less likely to discuss scholarship when writing a decision. The opposite might also be true. Perhaps the patent bar is much more knowledgeable and engaged with scholarship than the trademark and copyright bars. In such a case, it might be possible that the bar is so good at absorbing and translating the teachings of legal scholars that it has no need to refer to the articles from which it draws ideas. In this case, the fact that an argument was taken from scholarship might not be made express; jurists will be unaware that scholarship is the source and therefore not refer to it. Finally, one reason the patent bar might have “dropped the ball” can be found in the preceding section’s explanation. Perhaps legal scholarship directed to patent issues is a lot less interesting, compelling, and helpful to practitioners than is scholarship directed to trademark and copyright issues. This might be more the fault of legal academics than it is of an insular and disconnected patent bar; but the result is the same. Lack of practitioner interest in what legal academics have to teach about the patent system and its interaction with innovation and public policy leads to no reference in briefs, which might lower the probability that express references are made to such teachings in a decision. vii. There is Less Patent Law Scholarship. A very simple explanation for what we observe is that there might be less scholarship directed to patent law than there is directed to trademark and copyright law. The basic hypothesis is this: there might be a functional relationship between the amount of scholarship available for an issue area and the probability that courts will reference scholarship in a decision addressing the issue area; and the function defining the relationship is at least somewhat direct so that more scholarship available for an issue area produces more scholarship referenced in judicial decisions. If so, smaller amounts of patent scholarship should produce smaller numbers of references in judicial decisions. Do Not Delete 556 1/4/2013 6:39 PM HOUSTON LAW REVIEW [50:2 In this vein, it may have been (and may still be) the case that patent scholarship was the province of a relatively small number of scholars. If the field was highly esoteric, then scholarly production may have been low (and perhaps difficult to 99 grasp) relative to the trademark and copyright areas; assuming, as may be reasonable, that historically a broader body of law professors felt they could competently write into those areas. 3. Why do Trademark and Copyright Decisions Appear to Be Outliers? As noted earlier in this part, the average rate of use 100 of scholarship in a Supreme Court decision is 32.21%. In its patent jurisprudence, the Court uses legal scholarship at a rate 101 of 28.26%. This represents a difference of 3.95 points, which, while notable, is comparatively modest to what is observed in trademark and copyright decisions. The reference to scholarship in those decisions is 20.73 points and 34.46 points, respectively, above the overall average for the Court.102 From this perspective, the Supreme Court might be treating legal scholarship in patent decisions as it would treat legal scholarship in (close to) an “average” decision. Trademark and copyright decisions are then not just more likely to reference scholarship than patent decisions, they are far more likely to reference scholarship than the Court’s decisions addressing other issues.103 The judicial response, in terms of reference to legal scholarship, to trademark and copyright decisions is an outlier—different not just from patent decisions, but far from the center of the Supreme Court’s behavior when authoring decisions generally. This finding is of interest. It is not the point of the study, however, and so we will not devote much space to it and offer our preliminary thoughts in only general terms. Why might trademark and copyright decisions be outliers? The short answer is that we are not really sure. Most of the explanations, 99. Indeed, one of the reactions to the presentation of this Article, when it was presented to the participants at the IPIL’s conference, that most stood out was the view that much patent scholarship remains difficult to grasp. As there is indisputably a cost to a judge or justice who uses legal scholarship, it stands to reason that scholarship that is hard to figure out or understand is not likely to get much use in judicial opinions. Nor is cost the only reason why. If, as ample evidence suggests, judges and justices use scholarship strategically, it stands to reason that the strategic value of using a piece of legal scholarship decreases if the audience who is supposed to be (a little more) won over by the reference cannot understand it, or worse yet, might draw a contrary conclusion from it. 100. Petherbridge & Schwartz, supra note 39, at 12. 101. See supra Table 1. 102. See Petherbridge & Schwartz, supra note 39, at 12; supra Table 1. 103. See supra Figure 1. Do Not Delete 2012] 1/4/2013 6:39 PM THE END OF AN EPITHET? 557 moreover, that we came up with are essentially the inverse of the explanations we offered for why patent decisions show so much less use of scholarship than trademark and copyright decisions. A brief summary thus includes explanations such as: statutory manipulation might disrupt the utility of cases to settle the most prominent statutory ambiguities, and so the judges and justices are more often deciding in spaces where they have less ability to rely on formal, legally binding authority. As legal scholarship can offer rigorous-seeming authority (although not binding), it might be helpful to judges and justices who feel uncomfortable revealing the more intuitive or pragmatic reasoning that genuinely animated a decision. Likewise, perhaps the multiple factors that may lead to strategic use of scholarship are for some reason more prominent in trademark and copyright opinions. A couple of other things that might be plausible come to mind. Trademark and copyright might be subjects that are somehow more broadly intellectually stimulating than other areas of law. If so, perhaps when a copyright issue is raised, jurists and their clerks tend to be stimulated to read and investigate the issues more than they do when the issue is something else. They might also be less afraid that they will make a mistake if they bring independent research to bear on a decision. Another possibility is that trademark and copyright scholarship stands out for both being easy to understand, and being especially on point to the issues that end up arising in litigation. As was noted earlier, legal academics have been crucified over the last ten to fifteen years for producing scholarship that judges, the bar, and a set of legal scholars as well, insist is useless and unhelpful to deciding cases.104 When jurists speak of this issue they usually finger as the culprit “law and” and other forms of law and social sciences, e.g., economics, scholarship, as especially unhelpful. Perhaps the body of trademark and copyright scholarship is characterized by less of this form of legal scholarship than other areas of study, and is more comprised of the sort of legal scholarship the jurists and practitioners might find valuable in the execution of their dayto-day responsibilities. 104. SECTION OF LEGAL EDUC. & ADMISSIONS TO THE BAR, AM. BAR ASS’N, LEGAL EDUCATION AND PROFESSIONAL DEVELOPMENT—AN EDUCATIONAL CONTINUUM: REPORT OF THE TASK FORCE ON LAW SCHOOLS AND THE PROFESSION: NARROWING THE GAP 5 (1992); David Hricik & Victoria S. Salzmann, Why There Should Be Fewer Articles Like This One: Law Professors Should Write More for Legal Decision-Makers and Less for Themselves, 38 SUFFOLK U. L. REV. 761, 778 (2005) (arguing that law review articles are becoming not only “unhelpful,” but “useless”). Do Not Delete 558 1/4/2013 6:39 PM HOUSTON LAW REVIEW [50:2 4. Other Explanations & Concerns. In Part III, we referenced a number of concerns that should be considered when interpreting observations made using the methods employed in this study. Here, we raise a few more. i. The Supreme Court is Insular and Lacking in Intellectual Curiosity. As noted earlier, we postulated that the Supreme Court by virtue of its jurisdiction and supervision responsibilities was unlikely to be an insular and disconnected institution. Those unpersuaded by the reasoning we provided may be drawn to conclude from our observations that the Supreme Court, when it comes to its patent jurisprudence is, like the Federal Circuit is contended to be, an insular and disconnected institution. The same logic presumably leads to the conclusion when it comes to trademark and copyright decisions that the Court may be well-adjusted, engaged, and appropriately intellectually curious. We cannot rule this out. Our observations, however, indicate that trademark and copyright decisions are the outliers. Patent cases are pretty close to the central measure of the Supreme Court’s use of scholarship in decisions. So when it comes to trademark and copyright decisions, the implication might be that the Court is hyper well-adjusted, hyper-engaged, and hyperintellectually curious. Unless those decisions have a special need to cite scholarship that surpasses that of nearly every other issue area that the Court handles, in which case the heightened use of scholarship could just be a natural response to the subject matter. If patent law presented the same special need, then the observations might indicate that the Supreme Court is operating poorly when it comes to patent decisions. The cited scholarship might have nothing to do with intellectual property or the other issue areas of the decisions in which it was observed. We have not examined our coding for the identity of the scholarship cited in the Supreme Court’s patent, trademark, and copyright jurisprudence. As a consequence, we don’t know for sure that the scholarship cited is directed to those topics. It is not beyond the realm of possibility that some (perhaps much) of the scholarship cited in the Supreme Court’s decisions addressing these matters is, e.g., scholarship concerning civil procedure, remedies, or questions of administrative law. If the scholarship used in the decisions does not concern the substance of the various IP subject matter categories, it could affect a number of the explanations offered. For example, such a finding might favor an explanation along the lines of: strategic use, patent law being Do Not Delete 2012] 1/4/2013 6:39 PM THE END OF AN EPITHET? 559 clearer and easier to jurists, less important to jurists, or alternatively, harder, scarier, and more important to jurists than trademark and copyright appeals. ii. Can Looking at Judicial Use of Scholarship This Way Really Tell Us Anything? Another concern is that the study is observing characteristics of judicial opinions that may not be that informative on the question of a court’s intellectual engagement. This is illustrated not only by the myriad reasons that we have raised that could explain the use of legal scholarship vel non independent of intellectual engagement by judges or justices, but also by the epithetic argument’s implication that if the Federal Circuit cited more legal scholarship in its decisions, observers should expect that the court is doing a better job deciding cases in view of the policies that animate patent law (or an equally good job as the other circuits that are supposedly better at least in part because they cite legal scholarship somewhat more often). How plausible is that? Does it verge a little bit on the absurd? For at least some readers, we think it might. Neither of us is inclined to give the Federal Circuit, or any court for that matter, a pass on critical review because it cites a lot of law review articles. We might be happy they are doing it; surely it is the case that being a law professor can be isolating—that no one is paying attention, that propositions we raise are not often enough getting turned into “fact” through the process of citation by courts (and other law professors for that matter). But in the grand scheme of things, it is not obvious that case-by-case contributions from law review-style legal scholarship even should be a particularly important source of authority for judges. It may well be the case that judges, justices, reporters, and some law professors all overly emphasize the significance of citation to law review and law journal articles in judicial opinions. iii. Maybe the Supreme Court is Really Different. Our design emphasizes an analysis of the Supreme Court’s use of scholarship in its patent, trademark and copyright decisions, and the idea that the analysis can be somewhat analogized to the circuit court context. Perhaps this is incorrect. Perhaps for some reason the Supreme Court is just so different from the circuits in ways important to the comparison we engage that our observations do not provide insight into factors that might be at work in the circuit context. iv. Other Error. As always, there is the possibility that we have erred in some way in designing or executing this study. Do Not Delete 560 1/4/2013 6:39 PM HOUSTON LAW REVIEW [50:2 V. THE END OF AN EPITHET, FOR NOW? Recall that the central normative concern of this paper is the utility of the empirical observation: “[T]he Second and Ninth Circuits [in trademark and copyright decisions] cite scholarship roughly four times as often as the Federal Circuit [does in its 105 patent decisions],” as evidence for the epithetic argument that the performance of Federal Circuit judges is intellectually deficient. Recall also that Professor Nard cautioned of ambiguity in the meaning of this finding, and that we are concerned that others have brushed that ambiguity aside and relied on the finding to create a portable argument used to give explanatory weight to nearly any criticism aimed at the court and its work product. The risk we are concerned about is that the epithetic argument has the potential to interfere with the meaningful academic analysis of the court’s work product because it provides a shortcut to an easy, and always negative, answer. A review of the results and analysis drawn from Part IV leads us to the conclusion that empirical support for the chain of reasoning linking differential citation to intellectual deficiency is weak, if not extremely weak. Differential citation to legal scholarship in a court’s jurisprudence might correlate with a court that is intellectually limited, but it doesn’t have to. Not only is the correlation not mandatory, there are so many other possible explanations for observing differential use of legal scholarship between patent, and copyright and trademark decisions that the level of complexity presses the conscience into some degree of caution in picking one as “correct” over all the others. Our findings might, instead, offer some evidence for the view that differential use of legal scholarship is a more natural phenomenon that may be unconnected, at least in many instances, from a court’s intellectual connectedness to ongoing law and policy discourse. Perhaps, then, this Article will put an end to an epithet, for now. At least until better evidence, some of which at least should be relatively formally empirically founded, can be gathered to support the idea that Federal Circuit judges are insular and lacking in intellectual curiosity. In the meantime, work can focus on evaluating the court’s work product for its effect on the patent system, the innovation community, and the public interest without relying on the analytical shortcut that the Federal Circuit is intellectually circumscribed and so its work product is necessarily subpar. 105. Nard, supra note 1, at 683. Do Not Delete 2012] 1/4/2013 6:39 PM THE END OF AN EPITHET? 561 VI. CONCLUSION We hope that in presenting this Article we have encouraged the patent community to consider the significance of Nard’s finding—and our own—more critically than it might have before, and thereby encouraged some fresh possibilities in approach to the study of patent jurisprudence, and the patent system and its attendant institutions. APPENDIX [Distribution of IP Decisions by Term]