Harvard Law Review Harvard Law Review Harvard Law Review

Harvard Law Review Forum

Back to “Business” at the Supreme Court: The “Administrative Side” of Chief Justice Roberts

The Appendix to this Article may be viewed at this link.


This November, like every November for decades, the Harvard Law Review published its annual review of the U.S. Supreme Court’s most recently completed Term — October Term 2014.1×1. The Supreme Court, 2014 Term, 129 Harv. L. Rev., at i (2015). The November issue’s main event is the Foreword written by a distinguished legal scholar reviewing the Court’s decisions that Term,2×2. David A. Strauss, The Supreme Court, 2014 Term — Foreword: Does the Constitution Mean What It Says?, 129 Harv. L. Rev. 1 (2015). followed by a series of faculty case comments3×3. Abbe R. Gluck, The Supreme Court, 2014 Term — Comment: Imperfect Statutes, Imperfect Courts: Understanding Congress’s Plan in the Era of Unorthodox Lawmaking, 129 Harv. L. Rev. 62 (2015); Jack Goldsmith, The Supreme Court, 2014 Term — Comment: Zivotofsky II as Precedent in the Executive Branch, 129 Harv. L. Rev. 112 (2015); Kenji Yoshino, The Supreme Court, 2014 Term — Comment: A New Birth of Freedom?: Obergefell v. Hodges, 129 Harv. L. Rev. 147 (2015). and student essays discussing the “leading cases” of the Term.4×4. The Supreme Court, 2014 Term — Leading Cases, 129 Harv. L. Rev. 181 (2015). As in past Novembers, there is a section buried at the very end of the issue and largely overlooked by most readers called The Statistics, consisting of a series of tables setting forth statistical facts about the Term.5×5. The Supreme Court, 2014 Term — The Statistics, 129 Harv. L. Rev. 381 (2015). There is no accompanying textual analysis.

What readers today fail to appreciate is how much the Harvard Law Review’s current approach to reviewing the Supreme Court’s Term differs from its original approach, which began in the early twentieth century. Indeed, it has gone completely topsy-turvy. When the Harvard Law Review first began to publish articles regularly reviewing the Supreme Court’s work, the faculty-authored articles focused primarily on the statistics of the Court and mostly eschewed any substantive discussion of the rulings themselves. What was once the headliner, authored by the most famous law professors and legal scholars of the day, has become an incidental, mostly forgotten sideshow compiled by anonymous law-student editors.

Harvard Law School Professor Felix Frankfurter championed the original approach. Beginning ninety years ago and continuing until he left the Harvard Law faculty to join the Court in 1939, Justice Frankfurter regularly published in the Harvard Law Review statistical studies of the Supreme Court’s work entitled The Business of the Supreme Court. Justice Frankfurter had a series of coauthors: He began in 1925 with his Harvard faculty colleague, former student,6×6. Jamal Greene, The Anticanon, 125 Harv. L. Rev. 379, 451 (2011) (describing Landis as a former student of Frankfurter’s). and future Harvard Law School Dean James Landis,7×7. Frankfurter and Landis began The Business of the Supreme Court series with eight separate publications in the Harvard Law Review that sought, in effect, to survey the work of the Court from its early years to the present. See Felix Frankfurter & James M. Landis, The Business of the Supreme Court of the United States — A Study in the Federal Judicial System, 38 Harv. L. Rev. 1005 (1925); Felix Frankfurter, The Business of the Supreme Court of the United States — A Study in the Federal Judicial System, 39 Harv. L. Rev. 35 (1925) [hereinafter Frankfurter, Study Part II]; Felix Frankfurter, The Business of the Supreme Court of the United States — A Study in the Federal Judicial System, 39 Harv. L. Rev. 325 (1926); Felix Frankfurter, The Business of the Supreme Court of the United States — A Study in the Federal Judicial System, 39 Harv. L. Rev. 587 (1926); Felix Frankfurter, The Business of the Supreme Court of the United States — A Study in the Federal Judicial System, 39 Harv. L. Rev. 1046 (1926); Felix Frankfurter & James M. Landis, The Business of the Supreme Court of the United States — A Study in the Federal Judicial System, 40 Harv. L. Rev. 431 (1927) [hereinafter Frankfurter & Landis, Study Part VI]; Felix Frankfurter & James M. Landis, The Business of the Supreme Court of the United States — A Study in the Federal Judicial System, 40 Harv. L. Rev. 834 (1927); Felix Frankfurter & James M. Landis, The Business of the Supreme Court of the United States — A Study in the Federal Judicial System, 40 Harv. L. Rev. 1110 (1927). Frankfurter and Landis followed up with five more articles annually reviewing the Court’s statistics. Felix Frankfurter & James M. Landis, The Supreme Court Under the Judiciary Act of 1925, 42 Harv. L. Rev. 1 (1928) [hereinafter Frankfurter & Landis, Judiciary Act of 1925]; Felix Frankfurter & James M. Landis, The Business of the Supreme Court at October Term, 1928, 43 Harv. L. Rev. 33 (1929); Felix Frankfurter & James M. Landis, The Business of the Supreme Court at October Term, 1929, 44 Harv. L. Rev. 1 (1930); Felix Frankfurter & James M. Landis, The Business of the Supreme Court at October Term, 1930, 45 Harv. L. Rev. 271 (1931); Felix Frankfurter & James M. Landis, The Business of the Supreme Court at October Term, 1931, 46 Harv. L. Rev. 226 (1932). with whom Justice Frankfurter published in 1927 an identically named book.8×8. Felix Frankfurter & James M. Landis, The Business of the Supreme Court (1927). When Landis moved in 1933 to the Federal Trade Commission,9×9. Humphrey Ousted from Trade Board, N.Y. Times, Oct. 8, 1933, at 24 (describing Landis’s appointment to the Federal Trade Commission). Justice Frankfurter’s coauthor for several years became another former student10×10. William N. Eskridge, Jr. & Philip P. Frickey, Commentary, The Making of the Legal Process, 107 Harv. L. Rev. 2031, 2034 (1994) (describing Henry M. Hart, Jr., as one of Frankfurter’s former students). who had just joined the Harvard Law faculty, Henry M. Hart, Jr.11×11. Felix Frankfurter & Henry M. Hart, Jr., The Business of the Supreme Court at October Term, 1932, 47 Harv. L. Rev. 245 (1933) [hereinafter Frankfurter & Hart, Business at October Term 1932]; Felix Frankfurter & Henry M. Hart, Jr., The Business of the Supreme Court at October Term, 1933, 48 Harv. L. Rev. 238 (1934); Felix Frankfurter & Henry M. Hart, Jr., The Business of the Supreme Court at October Term, 1934, 49 Harv. L. Rev. 68 (1935). And then finally, Justice Frankfurter was joined by a then-current law student and future judicial clerk, Adrian Fisher,12×12. See Obituary, Adrian S. Fisher, 69, Arms Treaty Negotiator, N.Y. Times, Mar. 19, 1983, at 28 (describing Fisher as a former law clerk of Justice Frankfurter’s and a 1937 graduate of Harvard Law School). in analyzing the statistics for the Supreme Court’s October Terms 1935 and 1936.13×13. Felix Frankfurter & Adrian S. Fisher, The Business of the Supreme Court at the October Terms, 1935 and 1936, 51 Harv. L. Rev. 577 (1938) (reviewing October Terms 1935 and 1936). In 1940, the year after Justice Frankfurter joined the bench, Professor Hart published the last in the faculty-authored series of The Business of the Supreme Court articles, analyzing the statistics of October Terms 1937 and 1938.14×14. Henry M. Hart, Jr., The Business of the Supreme Court at the October Terms, 1937 and 1938, 53 Harv. L. Rev. 579 (1940).

The Business of the Supreme Court deliberately focused on the statistics of the Court’s work, rather than on the substance of the Court’s rulings. Justice Frankfurter believed that the Court’s “judicial statistics tell a deal of the tale” in understanding the Court and its operations.15×15. Frankfurter, Study Part II, supra note 7, at 39. Justice Frankfurter was a staunch advocate for such statistical analysis, contending that “an adequate system of judicial statistics . . . will, through the critical interpretation of the figures, steadily make for a vigorous and scientific approach to the problems of the administration of justice.”16×16. Frankfurter & Landis, Study Part VI, supra note 7, at 468 (footnote omitted). “A survey of the Court’s work makes abundantly clear that opinions only in part tell the story of its labors,” Justice Frankfurter explained.17×17. Frankfurter & Landis, Judiciary Act of 1925, supra note 7, at 15. That is why Justice Frankfurter’s and then Hart’s annual reviews made plain that the “dramatic issues” within the Court’s individual opinions were “not the immediate concern of the papers in this series. . . . This is a study not of product, but of form and function.”18×18. Frankfurter & Fisher, supra note 13, at 578. “[I]ntensive analysis . . . of the substantive issues before the Court . . . is no part of the concern of this series of papers.”19×19. Hart, supra note 14, at 579–80.

In 1949, soon after the Harvard Law Review celebrated Justice Frankfurter’s tenth year on the Court,20×20. See Augustus N. Hand, Dedication, Mr. Justice Frankfurter, 62 Harv. L. Rev. 353 (1949). the Review’s student editors first began regularly to dedicate the November issue to a systematic review of the Court’s prior Term.21×21. See Felix Frankfurter, “The Administrative Side” of Chief Justice Hughes, 63 Harv. L. Rev. 1 (1949); Edwin McElwain, The Business of the Supreme Court as Conducted by Chief Justice Hughes, 63 Harv. L. Rev. 5 (1949); The Supreme Court, 1948 Term, 63 Harv. L. Rev. 119 (1949). Even then, Justice Frankfurter’s continuing influence on the Review was evident. The November 1949 issue began with an article by Justice Frankfurter titled “The Administrative Side” of Chief Justice Hughes,22×22. Frankfurter, supra note 21. followed immediately by an article by Edwin McElwain, another former Frankfurter student,23×23. McElwain graduated from Harvard Law School in 1934, clerked for Chief Justice Hughes, and lived in the Washington, D.C., area with a group of former Supreme Court clerks and Harvard Law students with ties to Justice Frankfurter, including Adrian Fisher. See supra note 12 and accompanying text; see also Katharine Graham, Personal History 106 (1997); Todd C. Peppers, Courtiers of the Marble Palace: The Rise and Influence of the Supreme Court Law Clerk 94–95 (2006) (describing McElwain’s experience clerking for Chief Justice Hughes). titled The Business of the Supreme Court as Conducted by Chief Justice Hughes.24×24. McElwain, supra note 21. The student editors also included for the first time a student-authored statistical analysis of the Court’s most recently completed Term, October Term 1948; they titled the analysis The Business of the Court in keeping with Justice Frankfurter’s original 1925 title of his own statistical analysis of the Court’s work.25×25. The Supreme Court, 1948 Term — The Business of the Court, 63 Harv. L. Rev. 119 (1949). Unlike Justice Frankfurter, Landis, and Hart, however, the Review did not limit its analysis of the Court’s Term in the November issue of the Harvard Law Review to the Court’s “Business,” but also included for the first time a series of student-written “notes” on the most important cases of the Term.26×26. The Supreme Court, 1948 Term — The Notes, 63 Harv. L. Rev. 125 (1949). However, consistent with Justice Frankfurter’s emphasis on the statistics, the student editors placed The Business of the Court’s statistical presentation before the case notes.27×27. See id.; The Supreme Court, 1948 Term — The Business of the Court, supra note 25.

With one exception,28×28. See The Supreme Court, 1965 Term — The Statistics, 80 Harv. L. Rev. 141 (1966). the Review editors retained the heading of “The Business of the Court” in the annual November issue reviewing the Court’s prior Term through 1970, changing the title to simply “The Statistics” in 1971.29×29. See The Supreme Court, 1970 Term — The Statistics, 85 Harv. L. Rev. 344 (1971). Also, through 1967, the student editors emulated Justice Frankfurter’s heightened emphasis on statistics by placing the statistics up front in the volume before their case notes,30×30. The November 1950 issue was the only exception to this practice. That year, the student editors placed The Business of the Court at the end of the November issue. See The Supreme Court, 1949 Term — The Business of the Court, 64 Harv. L. Rev. 157 (1950). and they did not move the statistical presentation to its current location — the very end of the November issue — until 1968.31×31. See The Supreme Court, 1967 Term — Business of the Court, 82 Harv. L. Rev. 296 (1968). Since the November 1971 issue of the Harvard Law Review, the renamed “The Statistics” section has, just as it appears in the November issue published this month,32×32. See The Supreme Court, 2014 Term — The Statistics, supra note 5. routinely been placed at the very end — where its title, location, and lack of textual analysis render it a distant relative to the Harvard Law Review’s original, primary focus on judicial statistics.

The purpose of this article is to take a turn back toward Justice Frankfurter’s original vision of the significance of the Court’s “Business.” The article accordingly adds some analytical gloss to the current Review’s un-Frankfurter-like practice of providing bare statistical tables without also gleaning “a deal of the tale”33×33. Frankfurter, Study Part II, supra note 7, at 39. that a plumbing of those numbers could tell about the Court. There are many possible stories to be discerned from those numbers. But, because for Justice Frankfurter the distribution of the “opinions of the Court”34×34. The Court did not always issue an “opinion of the Court” as part of its ruling. In its earliest years, the Justices issued seriatim opinions, followed by a judgment “by the Court,” and it was not until the tenures of Chief Justices Oliver Ellsworth and John Marshall that the seriatim practice receded and the issuance of an “opinion of the Court” became the norm. See G. Edward White, The Internal Powers of the Chief Justice: The Nineteenth-Century Legacy, 154 U. Pa. L. Rev. 1463, 1466–69 (2006). amongst the Justices was always one of the most telling of statistics, this Article will focus on that same statistic. As Justice Frankfurter explained: “Perhaps no aspect of the ‘administrative side’ that is vested in the Chief Justice is more important than the duty to assign the writing of the Court’s opinion.”35×35. Frankfurter, supra note 21, at 3. “[I]f the duty is wisely discharged,” it is “perhaps the most delicate judgment demanded of the Chief Justice.”36×36. Id. at 4.

This is also an especially opportune moment to assess the “administrative side” of Chief Justice John G. Roberts, Jr., by taking a close look at how he has been exercising his opinion assignment authority. Chief Justice Roberts completed this past October his tenth Term as Chief Justice. Roberts is the nation’s seventeenth Chief Justice and is currently the tenth-longest-serving Chief in the nation’s history.37×37. See Members of the Supreme Court of the United States, Supreme Court of the United States, http://www.supremecourt.gov/about/members.aspx (last updated Oct. 6, 2015) [http://perma.cc/NX6A-B2AN]; Biographies of Current Justices of the Supreme Court, Supreme Court of the United States, http://www.supremecourt.gov/about/biographies.aspx (last updated Oct. 5, 2015) [http://perma.cc/Z6PE-LPTP]. Because Roberts, at age fifty, was the youngest Chief Justice to join the Court since John Marshall in 1801, who was then forty-five, precisely how the current Chief exercises such authority is of more than mere historical interest.38×38. See John G. Roberts, Jr., Oyez, https://www.oyez.org/justices/john_g_roberts_jr [https://perma.cc/3DHM-GJ7L]. Marshall famously went on to serve as Chief Justice for more than thirty-four years, longer than any other Chief.39×39. See Frequently Asked Questions on Justices, Supreme Court of the United States, http://www.supremecourt.gov/faq_justices.aspx [http://perma.cc/C8Z2-TGJ7].

This Article is divided into three parts. Part I reviews the history of the Chief Justice’s exercise of opinion assignment authority. Part II examines closely what the statistics reveal about the current Chief Justice’s use of such authority. And Part III offers some concluding remarks.

I. The Chief Justice’s Authority to Assign “Opinions of the Court”

A Chief Justice’s authority to assign opinions is highly consequential. Which of the nine Justices drafts the opinion of the Court in any specific case can determine the substance of the Court’s ruling and its precedential impact.40×40. The Chief Justice assigns the majority opinion when the Chief is in the majority at conference. For that reason, a Chief who wished to maximize his influence over the substance of the Court’s rulings could in theory manipulate his vote at conference to ensure that he was in the majority and therefore in control of the opinion assignment in a particular case. The Chief would not, of course, need to do that in a case where he was himself the fifth vote, but could do so in any case in which the Chief would otherwise be in dissent and the vote was not five to four. There have been suggestions that Chief Justice Earl Warren did not shy away from this practice. See G. Edward White, Earl Warren: A Public Life (1982), as reprinted in Inside the Supreme Court 706, 708–11 (Susan Low Bloch et al. eds., 2d ed. 2008) (discussing the assignment strategy of Chief Justice Warren). But it is Chief Justice Burger who was apparently the most notorious for engaging in this practice. At conference, the Chief Justice votes first, but Burger would reportedly pass on his vote to see how the majority was shaping up, and then cast his vote only after he knew the majority outcome, thus preserving his authority to assign the opinion. See Forrest Maltzman, James F. Spriggs II & Paul J. Wahlbeck, Crafting Law on the Supreme Court 32 (2000) (“Inevitably, one reason Burger is likely to have deferred his vote was to see what position would prevail so he might cast a vote that would allow him to assign the majority opinion.”); G. Bernard Schwartz, The Ascent of Pragmatism: The Burger Court in Action (1990), as reprinted in Inside the Supreme Court, supra, at 715, 722–23 (discussing assignment strategy of Chief Justice Burger). See generally Paul J. Wahlbeck, Strategy and Constraints on Supreme Court Opinion Assignment, 154 U. Pa. L. Rev. 1729, 1731–34 (2006). A more recent study concluded that when Chief Justice Burger passed on his initial vote, he would make opinion assignments to Justices that diverged more strongly from the ideological composition of the majority than when he did not pass. See Kaitlyn L. Sill et al., Strategic Passing and Opinion Assignment on the Burger Court, 31 Just. Sys. J. 164, 176–77 (2010). At conference, Justices vote to affirm or reverse a lower court judgment, but there are invariably many different possible analytical pathways that a majority opinion could pursue consistent with that bottom line. A Justice could draft the majority opinion extremely narrowly, creating little precedent, or just the opposite — draft the opinion in a very broad fashion, for the purpose of establishing a more sweeping precedent. As Justice Abe Fortas once described, “[i]f the Chief Justice assigns the writing of the opinion of the Court to Mr. Justice A, a statement of profound consequence may emerge. If he assigns it to Mr. Justice B, the opinion of the Court may be of limited consequence.”41×41. Abe Fortas, Dedication, Chief Justice Warren: The Enigma of Leadership, 84 Yale L.J. 405, 405 (1975). Justice Fortas’s exclusive use of the masculine title “Mr.” plainly dates his statement.

The current Chief Justice’s authority to assign the opinions of the Court is not, however, as sweeping as it once was. An informal product of evolving tradition and practices within the Court,42×42. Neither the Constitution nor any federal statute assigns the Chief Justice any heightened role in the Court’s decisionmaking process, including the authority to decide which Justice will draft the opinion of the Court or even that there will be an opinion of the Court. Nor is there any internal Supreme Court manual that sets forth precisely or even generally what administrative or other decisions the Chief can make unilaterally and which can be made only based upon consultation with other Justices or upon a formal vote of all the Justices. But it is common ground that the Court, like most any institution, has a compelling practical need for a leader who can make decisions for the institution as a whole and that leader should, in keeping with the Constitution’s creation of the distinct Chief Justice position within the Court, be the Chief Justice. As a matter of historical practice, since the Court began issuing “opinions of the Court” rather than merely opinions by individual Justices seriatim followed by a “judgment” by the Court, all members of the Court have apparently informally embraced the common-sense notion that because of the Chief’s status as the most “senior” member of the Court, the Chief should be responsible for assigning opinions when the Chief is in the majority. See White, supra note 34, at 1473, 1476–77, 1490–91. the Chief’s opinion assignment authority today is limited to those cases in which he is in the majority; otherwise the most senior Justice within the majority has the assignment power. By contrast, Chief Justice Marshall not only personally announced all of the opinions of the Court, even if “contrary to his own judgment and vote,”43×43. Id. at 1474; see id. at 1472–74. he also assigned himself the responsibility of drafting the opinion of the Court in the vast majority of cases.44×44. Id. at 1476.

A tradition of “silent acquiescence” also reigned under Chief Justice Marshall, which discouraged public dissent by Justices from the majority view.45×45. Id. at 1470–76. As a result, most of the opinions of the Court handed down by the Marshall Court were both delivered and written by the Chief Justice himself,46×46. For instance, during his 34 years as Chief, Chief Justice Marshall authored 547 opinions, while Justice Gabriel Duvall, who served on the Court for 23 of those years, wrote only 15 opinions, and Justice Thomas Todd, who served for 18 of Marshall’s years, wrote only 14. Id. at 1476. making the longstanding practice of referring to the Supreme Court by the name of its Chief Justice more apt then than it is today. It is doubtful a Chief today could duplicate that feat without triggering a rebellion amongst his colleagues and perhaps even a challenge within the Court to the Chief’s authority to assign opinions.

No less significantly, in Chief Justice Marshall’s time, it was not mandatory that the Justice responsible for authoring the opinion of the Court always circulate the draft opinion to the other Justices’ chambers for their review.47×47. Id. at 1471–73. The other Justices who voted in favor of the Court’s judgment, and whose votes were accordingly the reason why the opinion constituted an “opinion of the Court” with legal force, would not necessarily have the opportunity to review and approve the final opinion prior to its publication.48×48. Id. Absent such a natural check, the importance of the identity of the Justice responsible for drafting that opinion, and for that reason the significance of the authority to choose that person, was that much greater still. This general practice did not change until 1947, under Chief Justice Vinson.49×49. Id. at 1503–04.

The other chambers now carefully review draft opinions and other Justices in the majority can bargain for changes in draft opinions and ultimately even threaten to withdraw their votes if the opinion goes too far astray from reasoning they are willing to join. But, in most cases, the opinion writer has a great deal of discretion in opinion drafting without risking the loss of so many votes as to lose the majority. And, even when such an outcome is at risk, that possibility is what makes the Chief’s assignment of an opinion to a particular Justice all the more important. It can require an especially skilled opinion writer to identify the line of reasoning capable of maintaining (that is, not losing) the majority in a closely divided case.

Although no one has previously undertaken an in-depth statistical review of Chief Justice Roberts’s exercise of his opinion assignment authority,50×50. The former Supreme Court reporter for The New York Times, Linda Greenhouse, published a brief and casual assessment of the current Chief’s practices during his first few terms — in which she asserted that he had overassigned opinions to himself and then speculated why — but no one has undertaken an in-depth and searching inquiry. See Linda Greenhouse, Chief Justice Roberts in His Own Voice: The Chief Justice’s Self-Assignment of Majority Opinions, 97 Judicature 90 (2013). the record of prior Chiefs has not escaped close scholarly scrutiny, at least since the late nineteenth century. Political scientists rather than legal academics have done most of this work.51×51. See, e.g., Maltzman, Spriggs & Wahlbeck, supra note 40, at 29–56; Sara C. Benesh, Reginald S. Sheehan & Harold J. Spaeth, Equity in Supreme Court Opinion Assignment, 39 Jurimetrics 377 (1999); Saul Brenner, Strategic Choice and Opinion Assignment on the U.S. Supreme Court: A Reexamination, 35 W. Pol. Q. 204 (1982); Saul Brenner, Timothy Hagle & Harold J. Spaeth, Increasing the Size of the Minimum Winning Original Coalitions on the Warren Court, 23 Polity 309 (1990); Saul Brenner & Harold J. Spaeth, Majority Opinion Assignments and the Maintenance of the Original Coalition on the Warren Court, 32 Am. J. Pol. Sci. 72 (1988); Sue Davis, Power on the Court: Chief Justice Rehnquist’s Opinion Assignments, 74 Judicature 66 (1990); Jeffrey R. Lax & Charles M. Cameron, Bargaining and Opinion Assignment on the US Supreme Court, 23 J.L. Econ. & Org. 276 (2007); Forrest Maltzman & Paul J. Wahlbeck, May It Please the Chief? Opinion Assignments in the Rehnquist Court, 40 Am. J. Pol. Sci. 421 (1996); William P. McLauchlan, Research Note: Ideology and Conflict in Supreme Court Opinion Assignment, 1946–1962, 25 W. Pol. Q. 16 (1972); Gregory James Rathjen, Policy Goals, Strategic Choice, and Majority Opinion Assignments in the U.S. Supreme Court: A Replication, 18 Am. J. Pol. Sci. 713 (1974); David W. Rohde, Policy Goals, Strategic Choice, and Majority Opinion Assignments in the U.S. Supreme Court, 16 Midwest J. Pol. Sci. 652 (1972); Elliot E. Slotnick, Who Speaks for the Court? Majority Opinion Assignment from Taft to Burger, 23 Am. J. Pol. Sci. 60 (1979); S. Sidney Ulmer, The Use of Power in the Supreme Court: The Opinion Assignments of Earl Warren, 1953–1960, 19 J. Pub. L. 49 (1970); Sandra L. Wood et al., Opinion Assignment and the Chief Justice: 1888–1940, 81 Soc. Sci. Q. 798 (2000). A few generalizations are possible from that literature surveying the practices of past Chiefs.

First, for the past sixty years, beginning with Chief Justice Fred Vinson’s later years as Chief,52×52. Chief Justice Vinson evidenced little interest in ensuring numeric equality during his first years on the Court, but then shifted in that direction in his later years at the helm. See Saul Brenner & Jan Palmer, The Time Taken to Write Opinions as a Determinant of Opinion Assignments, 72 Judicature 179, 180–84 (1988); Elliot E. Slotnick, The Equality Principle & Majority Opinion Assignment on the United States Supreme Court, 12 Polity 318, 327–28 (1979). Chiefs have increasingly sought to assign opinions in a manner that promotes greater parity in the number of opinion assignments that each Justice receives.53×53. See Saul Brenner, Majority Opinion Assignment on the U.S. Supreme Court: A Bibliographic Overview of the Social Science Studies, 83 Law Libr. J. 763, 764 (1991) (citing Elliot E. Slotnick, Who Speaks for the Court? The Chief Justice and the Assignment of Majority Opinions (1977) (unpublished Ph.D. dissertation, University of Minnesota)); Slotnick, supra note 52, at 327–28; Harold J. Spaeth, Distributive Justice: Majority Opinion Assignments in the Burger Court, 67 Judicature 299, 301–02 (1984). Although Chief Justice Burger’s assignments were more equitable in distribution than Chiefs prior to Vinson, he did not adhere as strictly to numeric parity as other Chiefs since Vinson: Burger wrote an average of 10% more opinions than the other Justices during his tenure as Chief. David M. O’Brien, Storm Center: The Supreme Court in American Politics 270 (10th ed. 2014). But see Spaeth, supra, at 302–03 (describing Chief Justice Burger’s assignments as being far more numerically equal than suggested by other commentators). And they do so largely unaffected by how often a particular Justice is or is not in the majority and therefore eligible for the assignment. The Chiefs do, however, take into account assignments made in cases in which the Chief is not in the majority and therefore the assignment is made by another Justice — the most senior joining that opinion. Under this approach, a Justice who is in the majority only half the time would receive roughly the same number of assignments to write an “opinion of the Court” as a Justice in the majority 90% of the time and it would not matter how often the Chief is himself part of that majority.54×54. See Slotnick, supra note 52, at 320.

The same cannot be said of all Chief Justices prior to Chief Justice Vinson, for whom equality in numbers was not nearly as weighty a concern in opinion assignment.55×55. Wood et al., supra note 51, at 807 (concluding that numeric equality considerations were less important in Chief Justice opinion assignments in the late nineteenth and first half of the twentieth centuries). As previously described, Chief Justice John Marshall assigned most of the opinions to himself.56×56. See supra p. 39. And Frankfurter’s own annual reviews of the Court’s statistics, including opinion assignment distribution, do not leave the impression that numerical equality was nearly as strong a norm in the first half of the twentieth century as it has since become. For instance, in the Harvard Law Review article The Business of the Supreme Court at October Term, 1932 co-authored by Frankfurter and Hart in 1933, they provided a table showing the distribution of opinions of the Court during the prior ten Terms (1923–1932).57×57. Frankfurter & Hart, Business at October Term 1932, supra note 11, at 264–67 tbl.VII. According to that Table, during October Term 1925, the median number of opinions of the Court for each Justice was 21, yet Chief Justice William Howard Taft wrote 37 opinions of the Court and Justices Willis Van Devanter and George Sutherland each wrote only 17.58×58. Id. at 264 tbl.VII. During October Term 1927, the median was 22 opinions, Chief Justice Taft wrote 23, and Justices Van Devanter and Sutherland wrote 8 and 9, respectively.59×59. Id. Between 1930 and 1938, Chief Justice Hughes wrote approximately 21 opinions each Term, while the Associate Justices authored an average of 16 opinions each.60×60. O’Brien, supra note 53, at 269.

Second, past Chief Justices’ assignment practices also suggest that Chiefs have regularly taken into account certain strategic considerations apart from numeric equality. Some scholars have concluded based on statistical analysis that some Chiefs overassign to Justices who share the Chief’s own ideology in important cases.61×61. Ulmer, supra note 51, at 53, 57 (discussing Chief Justice Warren’s use of opinion assignment authority); see also Rohde, supra note 51, at 679–80 (discussing Chief Justice Warren’s use of opinion assignment authority in “important constitutional cases”). Others have concluded that in closely divided cases, some Chiefs (but not all) favored assignments to the so-called “marginal” Justice, meaning the Justice whose views were closest to those of the dissenting Justices,62×62. Compare Brenner, supra note 51, at 210 (discussing Chief Justice Warren’s tendency to assign closely divided cases to the “marginal” Justice), and Forrest Maltzman & Paul J. Wahlbeck, Opinion Assignment on the Rehnquist Court, 89 Judicature 121, 128 (2005) (concluding that Chief Justice Rehnquist “favored justices furthest from him” in close cases to retain majorities), with Davis, supra note 51, at 72 (concluding that Chief Justice Rehnquist did not in his first few years as Chief tend to assign closely divided cases to the “marginal” Justice). because that can be the best way to ensure that the resulting (and likely narrower) draft opinion is one that will keep the majority necessary for it to become an opinion of the Court.63×63. Reportedly, in divided cases, Chief Justice Hughes “would assign the case to the Justice nearest the center for the purpose of preventing any extreme opinions.” McElwain, supra note 21, at 18.

Scholars have also concluded that some Chiefs assign more opinions to Justices with specialized knowledge of a particular area of law, such as civil rights and civil liberties.64×64. Saul Brenner, Issue Specialization as a Variable in Opinion Assignment on the U.S. Supreme Court, 46 J. Pol. 1217, 1220–21 (1984) (discussing Chief Justice Warren); Saul Brenner & Harold J. Spaeth, Issue Specialization in Majority Opinion Assignment on the Burger Court, 39 W. Pol. Q. 520, 524 (1986) (discussing Chief Justice Burger). There is also statistical evidence suggesting that Chiefs have given fewer assignments to Justices who are perceived to be less competent or less efficient in opinion production than others;65×65. See Brenner & Palmer, supra note 52, at 184 (concluding that Chief Justice Vinson considered each Justice’s writing speed when making opinion assignments); White, supra note 34, at 1500 (describing how Chief Justices Morrison Waite and Melville Fuller sought to “ensure[] that dilatory or less competent Justices received fewer assignments”); id. at 1494 (describing how Chief Justice Waite would “bypass[] Justices . . . who were either disinclined to write opinions or whom Waite thought inept”). Chiefs also appear to have singled out for favorable treatment Justices who are especially able writers.66×66. See White, supra note 34, at 1494 (describing how Chief Justice Morrison Waite would assign significant opinions “to his more gifted colleagues”). Justice Frankfurter lauded Chief Justice Hughes’s “resourcefulness” in exercising his opinion assignment authority by allowing “that the Court should not be denied the persuasiveness of a particular Justice, though himself procedurally in dissent, in speaking for the Court on the merits.” See Frankfurter, supra note 21, at 4. In support of this unusual suggestion that Chief Justice Hughes assigned the opinion to a Justice who was not in the majority for all aspects of the Court’s ruling, Justice Frankfurter cited the Court’s then-recent decision in Helvering v. Davis, 301 U.S. 619 (1937), in which Justice Benjamin Cardozo wrote the Court’s opinion upholding the constitutionality of the Social Security Act of 1935, id. at 634, 645–46. Frankfurter, supra note 21, at 4 n.2. Frankfurter helpfully cited the page number of the Helvering opinion that makes clear that Justice Cardozo did not agree with the majority that the Court needed to reach the constitutional issue; but given that he had been outvoted on that threshold issue, he was authoring the Court’s opinion, including its discussion of those merits. See Helvering, 301 U.S. at 640 (“Under the compulsion of that ruling, the merits are now here.”). One can fairly speculate that Chief Justice Hughes thought Justice Cardozo best for the opinion assignment, notwithstanding his dissent on the threshold procedural matter, because, in addition to being a Justice celebrated for his brilliance, Cardozo was already writing the Court’s opinion in Steward Machine Co. v. Davis, 301 U.S. 548 (1937), which involved the constitutionality of a different part of the Social Security Act, id. at 573, and which had been argued earlier in the Term. Oral argument in Helvering had also taken place especially late in the Term (May 5th), see Helvering, 301 U.S. at 619, thereby placing a premium on an author who could both produce the opinion quickly and coordinate its writing with the drafting of the Court opinion in Steward Machine Co. Cardozo was clearly the best situated Justice for all those tasks. The Court announced its rulings in both cases only 19 days after Helvering was argued. See id. Scholars have contended further that their statistical analysis demonstrates that both time constraints toward the end of a Term and a particular Justice’s existing workload have influenced a Chief’s assignment practices.67×67. Maltzman, Spriggs & Wahlbeck, supra note 40, at 52 (concluding that Chief Justice Burger considered a Justice’s existing workload and the time left before the Term’s summer recess in making opinion assignments). Chief Justice Rehnquist made explicit the relevancy of both of these considerations to his opinion assignments in a memorandum to his colleagues.68×68. Chief Justice William Rehnquist advised his colleagues that in order “to avoid the annual ‘June Crunch,’” he was going to depart from the “principal rule” that he had “followed” in assigning opinions, which was “to give everyone approximately the same number of assignments of opinions for the Court during any one term,” and “to give some preference to those who are ‘current’ with respect to past work.” See id. at 30–31 (quoting Memorandum from Chief Justice William Rehnquist to the Conference on Policy Regarding Assignments (Nov. 24, 1989) (on file with Papers of Justice Thurgood Marshall, Manuscript Division, Library of Congress, Washington, D.C.)).

Finally, political science scholars have examined whether Chiefs self-assign disproportionately. In terms of the sheer number of opinion assignments, before numerical equality became a more settled norm, some Chiefs clearly assigned themselves an outsized share of the Court’s opinions. Certainly John Marshall did, by leaps and bounds,69×69. See supra note 44 and accompanying text. and, as previously described, the distribution of opinion assignments in October Term 1932 leaves little doubt that Chief Justice Taft did not shy away from the practice.70×70. See supra notes 57–59 and accompanying text. Since Chief Justice Vinson’s time, however, the principle of numeric equality has become sufficiently weighty that one does not see Chiefs self-assigning in such an outsized way.

Scholars do perceive, however, a practice of Chief self-assignment in the assignment of opinions in the more high-profile, salient, and arguably more important subset of cases on the Court’s docket.71×71. See Elliot E. Slotnick, The Chief Justices and Self-Assignment of Majority Opinions: A Research Note, 31 W. Pol. Q. 219, 219–22 (1978). There are, without question, cases on the Court’s docket that are more important in terms of their legal significance, their public profile, or, relatedly, their associated political controversy. And there are also cases on the docket that are true head-scratchers, in the sense that it is hard to fathom how such a seemingly mundane, technical, and downright uninteresting legal issue made it all the way to the Supreme Court. For the former, scholars have concluded that some Chiefs have not shied away from assigning themselves disproportionately the more important cases,72×72. See Slotnick, supra note 52, at 331 (Chief Justices Taft and Hughes); Ulmer, supra note 51, at 57 (Chief Justice Warren); Spaeth, supra note 53, at 304 (Chief Justice Burger); Maltzman & Wahlbeck, supra note 62, at 126 (Chief Justice Rehnquist). which of course may be perfectly sensible given the potential positive symbolic value of the Chief’s authorship.73×73. See Frankfurter, supra note 21, at 3–4. For the latter, however, scholars have not found similarly that the Chiefs have assigned more of the less complex cases to the most junior Justice on the bench.74×74. Maltzman, Spriggs & Wahlbeck, supra note 40, at 52.

II. The Opinion Assignments of Chief Justice Roberts

What about the current Chief Justice, John Roberts? What stories do the statistics tell?75×75. The statistics relevant to the Chief Justice’s exercise of his opinion assignment authority are the product of a series of tables appended to this Article (see Tables 1–17, infra), and the research methods used to produce those tables, including their statistical significance, are described at the outset of that appendix, infra at pp. 72–73. Though diminished since Chief Justice Marshall’s day, the current Chief’s opinion assignment authority remains among his most important administrative responsibilities. Since joining the Court in 2005, Chief Justice Roberts has been in the majority an average of 86.3% of the time, with a single-Term high of 91.8% and a low of 77.3%.76×76. See Tables 1 & 2, infra. Based on the statistics in those tables, the Chief was in the majority 88.4%, 83.6%, 88.1%, 81.1%, 91.8%, 89.3%, 87.5%, 84.9%, 91.0%, and 77.3% of the time, commencing in October Term 2005 and ending with the most recently completed Term, October Term 2014. The current Chief’s percentages are roughly the same or a bit higher than other Chiefs during the past seventy years and lower than Chiefs earlier in the nation’s history.77×77. Assuming these percentages roughly correspond to the percentage of cases in which the Chief Justice was in the majority at the time of the conference vote and therefore had opinion assignment authority (see infra notes 78–80 and accompanying text), the current Chief Justice’s percentages would be essentially the same as Chief Justices Burger and Stone (both at about 85%), higher than Chief Justices Rehnquist (81%) for October Terms 1986 to 1993 and Warren (80%) for October Terms 1953 to 1960, and lower than Chief Justices Hughes and Taft (averaging 95%). See Maltzman & Wahlbeck, supra note 62, at 123–24; Spaeth, supra note 53, at 301; Ulmer, supra note 51, at 53.

Of course, the fact that the Chief was in the majority at the time an opinion was published does not necessarily mean the Chief was similarly in the majority when the initial opinion assignment was made. And it is the most senior Justice in the majority at the time of the initial vote at conference who is authorized to assign the opinion of the Court to a particular Justice. The Chief, like any other Justice, could have dissented at the time of the conference vote and then decided to change his vote in favor of the majority view once he reviewed the draft majority opinion.78×78. O’Brien, supra note 53, at 265–66. Much bargaining over wording can result between the chambers. See William H. Rehnquist, The Supreme Court 264–65 (2001). Based on my own review in the Library of Congress Manuscript Division of the papers of many past Justices, including their recordings of the votes at conference, it is not uncommon for votes to shift once the majority and any concurring or dissenting opinions are circulated. Nor should that be surprising. One would hope that the individual Justices would be open to shifting their views (in either direction) once they have had an opportunity to see in writing the competing opinions addressing the legal issues in depth. But, for that same reason, if he did switch his vote from a dissent to the majority in a case, an assumption that he assigned the opinion of the Court in that case would in fact be mistaken.

Because, however, there are no public records for the Roberts Court of the votes cast by each Justice at conference, my only option is to acknowledge this limitation, just as prior scholars have done when faced with this same information gap in studying the exercise of opinion assignment authority by previous Chief Justices.79×79. Political scientists who are fortunate enough to study past Courts for which those records are available are not similarly limited, but many scholarly reviews are based on the same assumption made necessary here. See Brenner, supra note 53, at 766. Fortunately, for the purposes of this Article’s statistical analysis over ten Terms, it is fair to assume that the two votes — the Chief’s conference vote and final vote — are the same with sufficient frequency in the hundreds of cases covered by the database for the analysis to maintain its analytical value.80×80. Id. One final limitation to my analysis is that I am considering a vote in favor of the majority’s judgment in a case as sufficient to place the Chief in the majority at conference for the purposes of opinion-assignment authority. Apart from the fact that the Chief’s joining only the judgment — and not the majority opinion — might well signal he was not with the majority in conference, it is possible that the Chief decided at conference that his rationale for the result departed so significantly from a majority that he did not exercise opinion assignment authority. For the purposes of this analysis, I have resisted the temptation to go through all the cases individually and try to determine when that was more and less likely, in favor of treating all the cases the same.
For example, the Supreme Court Database supports the characterization of League of Latin American Citizens v. Perry, 548 U.S. 399 (2006), as a case in which the Chief had opinion assignment authority. But that is certainly debatable. Justice Kennedy’s opinion was divided into seven parts, two parts of which amounted to an opinion of the Court with five votes (joined by Justices Stevens, Souter, Ginsburg, and Breyer), for three parts of which it was a plurality of three (joined by Justices Souter and Ginsburg), and for two other parts of which it was a plurality of a different three (joined by Chief Justice Roberts and Justice Alito). Besides Justice Kennedy, everyone else concurred in part and dissented in part. See id. at 406-07. There was no single “good” answer on how to characterize the case for my statistical purposes. The case could perhaps be most easily labeled as an opinion assignment by Justice Kennedy (to himself) with no one else eligible. Based on the Supreme Court Database, I ultimately chose another “bad” answer, which was to treat it as a Chief assignment and a closely divided case. Certainly not the only possible classification or even irrefutably the best. But arguably within bounds given that the Chief likely had the institutional responsibility to sort out and interpret the voting mess in the first instance. And the good news is that these kinds of cases that resist clear classification are rare enough over ten Terms so as not to affect the overall analysis.

With this necessary threshold caveat, what tales do the statistics of his past ten Terms as Chief tell about how Chief Justice Roberts has exercised what Justice Frankfurter described as this most important administrative authority?81×81. See supra notes 35–36 and accompanying text. There are several.

First, Chief Justice Roberts has achieved maximum numeric equality in the number of assignments that each Justice receives to write opinions of the Court each Term to an extent unmatched by any prior Chief Justice.82×82. See infra section II.A. Indeed, his commitment to numeric equality is so keen that it allows observers to identify instances when a Justice who originally had the opinion of the Court subsequently lost the majority because of later changes in voting.83×83. See, e.g., infra notes 119–20 and accompanying text.

Second, Chief Justice Roberts’s unmatched desire for numerical equality has not prevented him from being strategic in deciding which cases to assign to particular Justices based on a case’s relative importance (or the lack thereof), or its difficulty rooted in the closely divided nature of the vote at conference. Like past Chiefs, the current Chief seems to assign the more high-profile and the most closely divided cases disproportionately to certain favored Justices, including himself, with little regard to seniority.84×84. See infra section II.B. For the closely divided cases in particular, the Chief appears to place a premium on opinion writers who can write more narrowly and therefore can be more trusted to maintain the majority established at conference. His assignment patterns also suggest a possible practice of assigning the “dogs” (that is, the less interesting cases) of the docket disproportionately to other, less favored Justices.85×85. See infra notes 176–83 and accompanying text.

Third, the Chief uses his opinion assignment authority to promote other institutional objectives important to him. Like his predecessors, Chief Justice Roberts seeks to promote the stature of the Court and public acceptance of its rulings by assigning cases in a way that challenges the notion that the Court’s decisions merely express the partisan political preferences of the individual Justices rather than their application of neutral legal principles.86×86. See infra section II.C. And, unlike prior Chiefs, Chief Justice Roberts does not assign himself more opinions than all other Justices on the Court, does not avoid writing opinions in closely divided cases, and assigns himself a proportionate share of the duller cases — all perhaps as a symbolic expression of his stated preference for judicial modesty.87×87. See infra notes 228–31 and accompanying text.

A. Numeric Equality

First, like his immediate predecessors and unlike Chiefs in earlier times, Chief Justice Roberts has clearly sought to achieve parity in the number of opinions each Justice writes for the Court. The statistics strongly suggest that he has generally strived for what could fairly be characterized as maximum numeric equality, meaning that any differences in the resulting number of opinions were required for the simple reason that the total number of opinions of the Court for that particular Term could not be divided evenly by the number of Justices (nine). For instance, if the Court decided seventy-two cases in a Term, maximum equality would be achieved if each Justice received eight assignments to draft the opinion of the Court. But, if the Court decided seventy-one cases, maximum equality would instead require that all Justices have eight opinion assignments, except one Justice, who would have only seven.

Achieving maximum numeric equality plainly can be challenging. Because the Chief assigns the vast majority but not all of the opinions in a Term — the most senior Justice in the majority assigns in those cases in which the Chief is in dissent — the Chief must coordinate his assignments with those made by others with assignment authority in individual cases. And the decisions made by another Justice with such authority can make it hard for the Chief to achieve numeric equality.

The Court’s own argument timetable and overall Term calendar also pose obvious hurdles to any effort to achieve strict numeric equality in opinion assignments. The Court hears oral arguments in seven monthly argument sessions lasting two weeks, and the Chief makes assignments within a few days after the completion of each session.88×88. See Conversations with Bill Kristol, Justice Samuel Alito on the Supreme Court, Recent Court Decisions, and His Education, YouTube (July 19, 2015), https://www.youtube.com/watch?v=Gd2VzBmr6YM (interview with Justice Alito describing when Chief Justice Roberts assigns opinions); O’Brien, supra note 53, at 271. Chief Justice Rehnquist reportedly made his assignments on a similar timetable. See id.; Rehnquist, supra note 78, at 259. Assignments in prior months necessarily limit assignment discretion in subsequent months because a Justice who has received more assignments than others in an earlier monthly argument session — which will invariably happen because the Court frequently hears more or fewer than precisely nine cases in each argument session when prior assignments are made89×89. This clearly happens frequently. For instance, the Court in October Term 2013 heard oral argument in more than nine cases in five of the seven monthly sessions and fewer than nine in the two remaining sessions. See Table 2, infra. For each of the sessions with more than nine cases, every Justice received at least one opinion assignment and one or more of the Justices received two. As the Term progressed, however, the maximum numeric equality standard required assigning fewer opinions in later argument sessions to any Justice who had received two rather than only one opinion assignments in a prior session. — must receive fewer (or more) later on if strict numeric equality is to be maintained over the entire Term. As described by Chief Justice Burger in commenting on the related challenge of achieving numeric parity in opinion assignments, “[e]ven one change in [opinion assignments] has a domino impact on all other assignments.”90×90. Maltzman & Wahlbeck, supra note 62, at 124 & n.23 (quoting Memorandum from Chief Justice Warren Burger to the Conference (Apr. 28, 1978)) (second alternation in original).

Statistical assessment of the Chief’s “assignment opportunities” — when both the Chief and a specific Justice are together in the majority — reveals that he nearly always had multiple assignment opportunities for any one Justice in any one argument session.91×91. See Table 4, infra. It was only on a few isolated occasions that the Chief had only one case that he could possibly assign to a particular Justice.92×92. For example, for the March argument session of October Term 2006, Justice Stevens was in the majority with Chief Justice Roberts in only one of the seven cases argued that month, Credit Suisse Securities (USA) LLC v. Billing, 551 U.S. 264 (2007). See Table 4, infra. No single Justice during his tenure has so frequently and systematically been in dissent as to make it essentially impossible to assign a proportionate share of opinions to that Justice. What nonetheless did make maintaining numeric equality more difficult was when there was a group of Justices in the majority less frequently, which had the same practical effect — sharply reducing the cases that the Chief could possibly assign to any one of them.93×93. For example, for the March argument session of October Term 2006, in which the Court heard seven cases, the Chief Justice had only one opportunity to assign to Justice Stevens, two for Justice Ginsburg, and three each for Justices Breyer and Souter, but these assignment opportunities consisted of the same three cases, Wilkie v. Robbins, 551 U.S. 537 (2007); Tellabs, Inc. v. Makor Issues & Rights, Ltd., 551 U.S. 308 (2007); Credit Suisse, 551 U.S. 264. See Tables 2 & 4, infra.

An additional factor can make it even harder for a Chief Justice to achieve maximum numeric equality. Although the Chief’s ability to promote numerical equality is largely limited to the making of the initial opinion assignments at the conference following oral argument, it is the final votes of Justices that determine which opinions are ultimately those “of the Court.” Voting shifts by individual Justices in response to the circulation of draft opinions can result in the loss of a majority, such that the majority becomes the dissent and the dissent the majority.94×94. See supra note 78 and accompanying text. When that happens, the Justice who “lost” that majority is likely to have fewer opinions of the Court at the close of the Term. As explained by Justice Ginsburg, “[t]he vote at conference isn’t fixed in stone;”95×95. Ruth Bader Ginsburg with Linda Greenhouse, A Conversation with Justice Ginsburg, 122 Yale L.J. Online 283, 300 (2013). those subsequent shifts in votes and sometimes in outcomes are why “[e]very year, a Justice or two will announce one case less than the others.”96×96. Id. Indeed, for this same reason, clear departures from numerical equality can be the basis of reasoned speculation that vote shifts changed the result of a Supreme Court ruling between the time of the initial conference vote and release of the final Court opinion several months later. See, e.g., infra note 120. When, moreover, there are only a few decisions remaining to be decided before the summer recess, one can often accurately predict their authors based on how many opinions for the Court the various Justices have (and have not yet) written. See, e.g., Jonathan H. Adler, Reading the SCOTUS Tea Leaves on EPA’s Mercury Rule, Wash. Post: Volokh Conspiracy (June 10, 2015), https://www.washingtonpost.com/news/volokh-conspiracy/wp/2015/06/10/reading-the-scotus-tea-leaves-on-epas-mercury-rule [https://perma.cc/EKV7-3ADX].

Notwithstanding these several hurdles, Chief Justice Roberts precisely adhered to that maximum equality standard in four of his ten Terms as Chief; for three more he can be fairly characterized as at least roughly achieving the maximum equality standard; and for the final three he missed that high mark more than incidentally. The current Chief’s record is even stronger in this regard than both Chief Justice Burger’s, which is credited with being far more equitable than all of his predecessors, and Chief Justice Rehnquist’s, which was stronger than Burger’s, especially during Rehnquist’s final years on the Court.97×97. See infra section II.A.4.

1. Maximum Numeric Equality. — The Chief precisely achieved maximum numeric equality in four Terms: October Terms 2006, 2007, 2012, and 2013. For each of those four Terms, he did so by accounting for the assignments made in cases in which the Chief himself was not in the majority and assignments were consequently made by the most senior Associate Justice in the majority in those cases.

In each of October Terms 2006, 2007, and 2013, the Court decided 67 cases on the merits, which meant that maximum numeric equality required that five Justices receive seven assignments and four receive eight assignments, which is what happened in all three Terms.98×98. See Table 2, infra. Only Justice Kennedy received eight opinions all three times.99×99. Table 2, infra. In October Term 2012, the Court decided 73 cases, and maximum numeric equality required that eight Justices receive eight opinion assignments and one receive nine assignments.100×100. Table 2, infra. The Chief achieved maximum numeric equality with each Justice receiving eight opinions except for Justice Ginsburg, who had nine.101×101. Table 2, infra.

2. Rough Maximum Numeric Equality. — For three of the six Terms in which the Chief did not achieve maximum numeric equality, he can be fairly characterized as at least roughly achieving the maximum numeric equality standard, missing it only by the thinnest of margins and for reasons likely outside his control.

(a). In October Term 2005, maximum numeric equality required that all nine members of the Court write either seven or eight opinions.102×102. See Table 2, infra. One Term-specific twist was that ten Justices served that Term because Justice O’Connor was on the bench until January 31, 2006, the same day that Justice Alito joined the Court.103×103. See Members of the Supreme Court of the United States, supra note 37. Although neither Justice O’Connor nor Justice Alito wrote seven or eight opinions for the Court, their combined total was seven, which is consistent with the maximum numeric equality standard.104×104. Table 2, infra. The only remaining statistic inconsistent with maximum numeric equality for the Term is Justice Scalia’s having written nine opinions of the Court.105×105. Table 2, infra. But that number may be misleading because it includes Hudson v. Michigan,106×106. 547 U.S. 586 (2006). a case that was reargued in May, after the end of the regular argument sessions.107×107. See id. at 586. Justice Scalia’s opinion in Hudson was also for the Court only in part, and a plurality opinion in part because Justice Kennedy did not join Justice Scalia’s opinion in full and four Justices dissented.108×108. Id. at 588–99 (opinion of the Court); id. at 599–602 (plurality opinion); id. at 602 (Kennedy, J., concurring in part and concurring in the judgment); id. at 604 (Breyer, J., dissenting, joined by Stevens, Souter, and Ginsburg, JJ.). Those two additional facts suggest that late-breaking postconference disagreements about the opinion rationale, possibly extending to changes in the votes of the Justices and even the majority outcome, may have hampered the Chief’s ability to achieve parity in numbers of opinions.

(b). In October Term 2009, maximum numeric equality required that all nine members of the Court write either eight or nine opinions.109×109. See Table 2, infra. They all did except for Justice Stevens, who wrote only six,110×110. Table 2, infra. which is a remarkably low number especially considering Justice Stevens’s status as the most senior Associate Justice. Here again, there is a ready explanation based on a closer review of that Term’s statistics that explains this discrepancy in a manner consistent with the Chief’s practice of seeking maximum numeric equality.

Justice Stevens’s low number results from his having written no opinions for cases argued that October Term during the November and April argument sessions.111×111. Table 2, infra. For November, however, there are two cases for which Stevens likely had the original opinion assignment. The first, Pottawattamie County v. McGhee,112×112. 558 U.S. 1103 (2010). was argued on November 4, 2009,113×113. See Transcript of Oral Argument, Pottawattamie Cty., 558 U.S. 1103 (No. 08-1065). but the writ was dismissed as improvidently granted two months later on January 4, 2010, after the parties settled the case.114×114. See Pottawattamie Cty., 558 U.S. at 1103 (mem.) (dismissing case pursuant to Rule 46). Pottawattamie County concerned the scope of prosecutorial immunity for fabricating evidence — the kind of high-stakes case in which Justice Stevens would likely have been very interested, especially given that he had written the Court’s leading relevant precedent.115×115. See Buckley v. Fitzsimmons, 509 U.S. 259 (1993). Also argued that same month was a significant patent case, Bilski v. Kappos,116×116. 130 S. Ct. 3218 (2010). for which it seems likely that Stevens received the original opinion assignment in November but then lost the majority to Justice Kennedy before the decision was handed down on June 28, 2010, the last opinion day before the summer recess. Justice Kennedy’s opinion is for the Court only in part, and for a plurality of four in other parts. And it is far shorter (only 16 slip opinion pages) than Justice Stevens’s much longer opinion concurring in the judgment (47 slip opinion pages), which was joined by three other Justices.117×117. See Bilski v. Kappos, No. 08-964, slip op. (U.S. June 28, 2010). There is also an easy explanation for Justice Stevens’s receiving no opinions for that Term’s April argument session. He announced his resignation from the Court on the 9th of that month, and he likely preferred to reduce his workload as he readied himself for his departure only a few weeks later, “after the Court rises for the summer recess.”118×118. Letter from Justice John Paul Stevens to the President of the United States (Apr. 9, 2010), http://www.supremecourt.gov/publicinfo/press/JPSLetter.pdf [http://perma.cc/P35S-Q8YC]; see also Sheryl Gay Stolberg & Charlie Savage, Justice Stevens Retiring, Giving Obama a 2nd Pick, N.Y. Times, Apr. 10, 2010, at A1.

(c). A close examination of the opinion assignments for October Term 2014, just completed, is also consistent with the maximum numeric equality standard. For that Term, all the Justices wrote either seven or eight opinions, except for Justice Kennedy who wrote six and Justice Scalia who is credited with nine.119×119. See Table 2, infra. Here, the numeric discrepancy seems likely to have resulted from Justice Kennedy’s having lost the opinion of the Court in a case argued during the February session and Justice Scalia’s having gained the plurality opinion in that same case as a result.120×120. It is reasonable to speculate that in Kerry v. Din, 135 S. Ct. 2128 (2015), argued during the February argument session of October Term 2014, Justice Kennedy received the original opinion assignment after fragmented voting at conference with the expectation that he was best positioned to produce an opinion of the Court, but he was subsequently unable to secure the necessary majority for his draft opinion. That would readily explain why he wrote no opinions for the Court that month and Justice Scalia wrote both an opinion for the Court in another case, EEOC v. Abercrombie & Fitch Stores, Inc., 135 S. Ct. 2028 (2015), and the plurality opinion in Din, with Justice Kennedy filing an opinion in that case concurring in the judgment. If one added that case to Justice Kennedy’s total, he would have seven, and subtracted it from Justice Scalia’s — which would seem sensible in any event because Justice Scalia’s “ninth” was a plurality and not a true “opinion of the Court” at all — he would have eight. Either way, maximum numeric equality would be readily restored to the opinion assignments for October Term 2014 as well.

3. No Maximum Numeric Equality. — Finally, there are three remaining Terms where the opinion distribution departs far more significantly from the maximum numeric equality standard.121×121. See Table 2, infra. But, even for these Terms, the lack of strict parity in opinion distribution appears to be largely the result of a significant number of postconference vote shifts by the Justices and therefore the kind of numeric inequality that a Chief would be hard-pressed to prevent. In October Term 2008, Justice Scalia wrote eleven opinions, more than anyone else, with the other Justices writing nine (Thomas and Stevens), eight (Roberts, Souter, and Breyer), or only seven (Kennedy, Ginsburg, and Alito).122×122. Table 2, infra. In October Term 2010, there was an equally wide spread: Justice Kennedy had eleven opinions; Justice Scalia had ten; Justices Thomas and Ginsburg each had nine; the Chief Justice had eight; and Justices Breyer, Alito, Sotomayor, and Kagan each had seven.123×123. Table 2, infra. The spread in October Term 2011 is less stark but not incidental: Justices Kennedy and Scalia again led the pack with nine and eight opinions, respectively; the Chief Justice and Justices Ginsburg, Breyer, Alito, and Kagan each had seven opinions; and Justices Thomas and Sotomayor both had six opinions.124×124. Table 2, infra.

As before, one can explain these discrepancies by identifying specific cases in which majorities (and accordingly opinion assignments) appear to have been lost and won between the time of the original assignment and the final opinion announcement months later,125×125. For example, Justice Scalia’s authorship of three opinions in cases argued in the January argument session of October Term 2008 and three again in cases argued in the November session of October Term 2010 strongly suggests that he was not the original assignee in all of those cases, especially because in the latter case, both Justices Alito and Sotomayor wrote no opinions for any of the eleven cases argued that month. In the January argument session of October Term 2010, Justice Kennedy wrote two opinions for cases argued that session and Justice Breyer wrote none. Table 2, infra. For one of Justice Kennedy’s two opinions, J. McIntyre Machinery, Ltd. v. Nicastro, 131 S. Ct. 2780 (2011), it seems quite possible that Justice Breyer, who ended up writing an opinion concurring in the judgment, see id. at 2791 (Breyer, J., concurring in the judgment), had the original opinion assignment, especially since the Court did not hand down the decision until the last opinion day of the Term. Such significant time delay is a traditional marker for a possible opinion assignment shift. or just lost because of a subsequent decision to dismiss the writ of certiorari altogether following oral argument.126×126. In First American Financial Corp. v. Edwards, 132 S. Ct. 2536 (2012), the Court dismissed the writ as improvidently granted in the last order list before the summer recess, see id. (per curiam), after having heard oral argument in the case in the December argument session, see Transcript of Oral Argument, First American, 132 S. Ct. 2536 (No. 10-708). Because Justice Thomas was the only Justice to write no opinions for the Court in a case argued that month, Table 2, infra, he most likely had the original opinion assignment in the case, see Ginsburg with Greenhouse, supra note 95, at 300 (suggesting as much). But, for these three Terms, the numerical spread is sufficiently great that the number and corresponding weight of the necessary explanations quickly become too heavily compounded by layers of speculation for those explanations to be very meaningful. The better analytical pathway is instead to recognize that notwithstanding the current Chief’s obvious strong commitment to achieving maximum numeric equality, there are practical limits on his ability to adhere to that standard, especially with a closely divided Court.

4. In Comparison to Prior Chief Justices. — Chief Justice Roberts’s record in achieving numeric equality in opinion assignments is better than that of any prior Chief Justice. There is not even a remote contest until Chief Justice Warren because the Chiefs before Chief Justice Vinson did not even make a pretense of seeking numeric equality and Vinson himself did not make equality an important factor in opinion assignments until his final years as Chief.127×127. See supra notes 52–60 and accompanying text.

Chief Justice Burger’s record of achieving numeric equality was, at the time, far better than that of any prior Chief, including his immediate predecessor Chief Justice Warren. To compare the records of different Chiefs, one can begin with the standard deviations of their assignments for each Term, but the best statistical measure for comparing the records is to calculate, from the standard deviation, the coefficient of relative variation (CRV), which compares the size of the standard deviation relative to the size of the mean. This measure allows for fair comparison of Supreme Court Terms across which the numbers of total opinion assignments differs. The lower the CRV, the greater the extent to which numeric equality has been achieved in opinion assignment among the Justices.128×128. Terms with maximum numeric equality may nonetheless have nonzero CRVs. Unless the number of cases heard in a given Term is divisible evenly by nine, some Justices will necessarily author more opinions than other Justices. See supra pp. 47–48. As a result, these Terms will have nonzero standard deviations, and therefore nonzero CRVs.
Therefore, while the CRV measure normalizes numeric equality for variation in the number of cases decided in a given Term, it does not do so entirely. Nonetheless, a comparison of CRV measures will be more fair than a comparison of standard deviations alone.
Chief Justice Burger’s CRV was 0.179, which is much lower than Chief Justice Warren’s, which was 0.240, and far lower still than those of Chief Justices Stone and Vinson, which were both more than twice as high as Burger’s.129×129. Spaeth, supra note 53, at 302 (calculating Chief Justice Burger’s CRV for October Terms 1969 through 1980).

During his first several years as Chief Justice, Rehnquist apparently paid some attention to numeric equality, but “d[id] not appear to ensure that every [J]ustice ha[d] approximately the same number of assignments.”130×130. Maltzman & Wahlbeck, supra note 62, at 124. Between October Terms 1986 and 1993, according to one study, there was at least a four-opinion difference between those Justices receiving the highest and lowest number of Court opinion assignments, or as much as a 33% differential.131×131. Id. By contrast, during Chief Justice Rehnquist’s final ten Terms as Chief, my own statistical analysis of opinion assignments makes clear that he paid far closer attention to numeric equality and that he adhered to that standard even more than Chief Justice Burger had and therefore more than any prior Chief.132×132. See Table 17, infra. The opinion assignment data for ten of the Terms when Rehnquist was Chief Justice was that reported in the The Statistics of the November issue of the Harvard Law Review for each of those ten years. See The Supreme Court, 1994 Term — The Statistics, 109 Harv. L. Rev. 340 (1995); The Supreme Court, 1995 Term — The Statistics, 110 Harv. L. Rev. 367 (1996); The Supreme Court, 1996 Term — The Statistics, 111 Harv. L. Rev. 431 (1997); The Supreme Court, 1997 Term — The Statistics, 112 Harv. L. Rev. 366 (1998); The Supreme Court, 1998 Term — The Statistics, 113 Harv. L. Rev. 400 (1999); The Supreme Court, 1999 Term — The Statistics, 114 Harv. L. Rev. 390 (2000); The Supreme Court, 2000 Term — The Statistics, 115 Harv. L. Rev. 539 (2001); The Supreme Court, 2001 Term — The Statistics, 116 Harv. L. Rev. 453 (2002); The Supreme Court, 2002 Term — The Statistics, 117 Harv. L. Rev. 480 (2003); The Supreme Court, 2003 Term — The Statistics, 118 Harv. L. Rev. 497 (2004). Also, in reviewing Chief Justice Rehnquist’s “final” ten Terms as Chief, I reviewed the statistics for October Terms 1993 through 2003 and did not include what was in fact the Chief’s final Term, October Term 2004, because Chief Justice Rehnquist was ill during much of that Term and physically absent from the Court. See Lawrence K. Altman & Linda Greenhouse, Rumors Fly Over Rehnquist’s Plans, N.Y. Times (July 9, 2005), http://www.nytimes.com/2005/07/09/politics/rumors-fly-over-rehnquists-plans.html. Chief Justice Rehnquist’s CRV from October Terms 1994 through 2003 was 0.109, or 39% lower still than Chief Justice Burger’s CRV of 0.179 for October Terms 1969 through 1980.133×133. Compare Table 17, infra, with Spaeth, supra note 53, at 302.

Chief Justice Roberts’s record is even stronger. His average CRV for the past ten Terms was 0.105,134×134. See Table 3, infra. which is (slightly) lower than even Chief Justice Rehnquist’s final ten Terms (which are, in turn, likely to be much lower than his earlier terms as Chief Justice). Chief Justice Roberts also achieved maximum numeric equality for four Terms of his ten.135×135. See supra section II.A.1. Chief Justice Rehnquist did not reach that maximum level of equality even once.136×136. See Table 17, infra. There could of course be many possible explanations for the difference. One possibility is that the discrepancy is rooted in Chief Justice Rehnquist’s stated preference to assign more cases to Justices who, as the Term progressed, had completed past opinion assignments in a timely fashion. See supra note 68 and accompanying text.

Chief Justice Roberts’s opinion assignments also reflect greater numeric equality than his predecessors along another dimension: the relative absence of a high number of self-assignments. Even Chief Justices like Burger and Rehnquist, who plainly sought to achieve numeric equality far more than Chiefs before them, nonetheless regularly wrote more majority opinions than any other Justice. For instance, during October Terms 1969 through 1980, Chief Justice Burger authored the most opinions in October Terms 1972, 1977, and 1978.137×137. Spaeth, supra note 53, at 302 tbl.1. During October Terms 1994 through 2003, Chief Justice Rehnquist wrote more opinions than any other Justice for five of those ten Terms.138×138. See Table 17, infra. By contrast, in nine of his ten Terms, Chief Justice Roberts wrote fewer opinions of the Court than at least one other Justice and more often multiple Justices.139×139. See Table 3, infra. In only one Term (October Term 2007) was there no Justice who wrote more than the Chief; the Chief tied with three colleagues that Term — each writing eight opinions and the other five Justices each writing seven opinions.140×140. See Table 3, infra. Nor do the identities of those on the bench receiving the slightly higher number that Term or any other Term suggest a tendency to favor the more senior members. For instance, in October Term 2007, the four Justices writing eight opinions were the Chief Justice and Justices Kennedy, Ginsburg, and Breyer. See Table 3, infra. Nor does there appear to be significant numerical skewing based on seniority or shared ideology occurring in the other nine Terms. See infra pp. 57–58; Table 3, infra.

As discussed next, however, the current Chief’s strong desire to maintain maximum numeric equality does not mean that the Chief’s assignments are otherwise random or neutral in nature. A Chief may simultaneously pursue numeric parity along with other strategic and institutional objectives. Chief Justice Roberts has done just that.

B. Assigning the Biggest and Smallest Cases

All Supreme Court cases are, of course, not equal. Some are a lot more interesting than others. Some are far more important than others. Some, including those in which the vote is closely divided, require opinions that are much more difficult to write than others. And, some are stultifyingly dull. Nor, for those same reasons, are the Justices benignly neutral about a Chief’s opinion assignments. Justices naturally desire the opportunity to write the opinion of the Court in the “big,” more high-profile, salient cases.141×141. See infra pp. 63–64. And, conversely, they relish far less the prospect of writing the opinion of the Court in the “small,” mundane, highly technical, and seemingly less important cases.

Studies of prior Chief Justices have all found Chiefs being exceedingly strategic in deciding which Justices write the opinion of the Court in the biggest cases. In assigning the biggest cases, they have frequently favored themselves.142×142. See, e.g., Maltzman & Wahlbeck, supra note 62, at 126 (Chief Justice Rehnquist); Slotnick, supra note 51, at 222–23 (covering multiple Chief Justices); Slotnick, supra note 52, at 331 (Chief Justices Taft and Hughes); Spaeth, supra note 53, at 303–04 (Chief Justice Burger); Ulmer, supra note 51, at 57 (Chief Justice Warren). They have also not shied away from favoring particular Justices, including those closest to themselves ideologically143×143. See, e.g., Brenner, supra note 53, at 764 (summarizing research related to Chief Justices Taft to Burger); Rohde, supra note 51, at 681–82 (Chief Justice Warren in significant constitutional cases); Spaeth, supra note 53, at 304 (Chief Justice Burger); Wahlbeck, supra note 40, at 1753 (Chief Justice Rehnquist). or even those most ideologically distant if the Chief believes that such an assignment is the best way to maintain a fragile majority in a controversial case.144×144. See Maltzman & Wahlbeck, supra note 62, at 126, 181 (Chief Justice Rehnquist).

What do the statistics tell us about Chief Justice Roberts? What we learn is that the current Chief has been similarly strategic, albeit with a somewhat different emphasis. Looking just at the Court’s “big” — highest profile — cases, the numeric equality evident in the total number of opinion assignments each Justice received quickly dissipates. More than one-eighth (14%), or 85, of the 600 total cases assigned by the Chief over the past ten Terms fell within my high-profile “salient” category.145×145. See Table 5, infra. Salient Court decisions identified for this purpose are those cases identified as the most important cases of the Term by the New York Times for each of the ten Terms. At the close of every Term, the New York Times publishes a separate listing of those important rulings for its readers, typically numbering around ten cases in total but once as high as fifteen (October Term 2014). See The End of the Supreme Court Term, N.Y. Times (July 2, 2006), http://www.nytimes.com/2006/07/02/washington/20060702_SCOTUS_GRAPHIC.html (October Term 2005); Notable Cases of the 2006–7 Term So Far, N.Y. Times (June 29, 2007), http://www.nytimes.com/ref/washington/scotuscases_AFFIRMATIVEACTION.html (October Term 2006); Major Rulings of the 2007–8 Term, N.Y. Times (June 29, 2008), http://www.nytimes.com/imagepages/2008/06/29/washington/29scotus.web.2.html (October Term 2007); Major Rulings of the 2008–9 Term, N.Y. Times (July 1, 2009), http://www.nytimes.com/imagepages/2009/07/01/us/01scotus.graphic1.html (October Term 2008); Major Rulings of the 2009–2010 Term, N.Y. Times (June 30, 2010), http://www.nytimes.com/imagepages/2010/06/30/us/politics/30scotus-graphic.html (October Term 2009); Major Rulings of the 2010–11 Term, N.Y. Times (June 28, 2011), http://www.nytimes.com/interactive/2011/06/28/us/scotus-graphic.html (October Term 2010); Major Rulings of the 2011–12 Term, N.Y. Times (July 1, 2012) http://www.nytimes.com/imagepages/2012/07/01/us/01scotus-graphic.html (October Term 2011); Major Rulings of the 2012–13 Term, N.Y. Times (June 27, 2013), http://www.nytimes.com/interactive/2013/06/28/us/major-rulings-of-the-2012-13-term.html (October Term 2012); Key Supreme Court Decisions in 2014, N.Y. Times (June 30, 2014), http://www.nytimes.com/interactive/2014/06/19/us/major-supreme-court-decisions-in-2014.html (October Term 2013); Adam Liptak & Alicia Parlapiano, Major Supreme Court Cases in 2015, N.Y. Times (July 1, 2015), http://www.nytimes.com/interactive/2015/us/major-supreme-court-cases-in-2015.html (October Term 2014). While certainly far from a strictly scientific measure, the judgment of the New York Times Supreme Court correspondent strikes me as a consistently reliable grounds for identifying the most salient cases of the past ten Terms. For this subset of cases, the Chief Justice did assign a larger share of those cases to several Justices than what strict numeric equality would have provided, and the Chief favored himself above all others. In addition, his assignments favored certain individual Justices more than would have been suggested based on pure notions of either seniority or “shared ideology” with the Chief, although both these factors were clearly relevant.146×146. See Table 4, infra. By “shared ideology,” I mean nothing more than the relative frequency with which a particular Justice votes with the Chief. On the current Court, the more senior Associate Justices — Justices Scalia, Kennedy, and Thomas — are also among those who most frequently vote with the Chief in divided cases.

Because 14% of all the cases assigned by the Chief were within the salient category, numeric equality would be established if the salient cases constituted that same fraction of the cases assigned to each of the Justices. However, both the Chief himself and Justice Kennedy received salient assignments with twice the frequency of anyone else on the bench. The Chief was highest with 34%, followed by Justice Kennedy at 31%.147×147. Table 5, infra. The next highest were Justices Alito, Scalia, and Breyer at 16%, 15%, and 11% respectively, and the remaining Justices were lower: Justice Souter at 8%, Justice Thomas at 5%, Justice Stevens at 4%, Justices Ginsburg and Kagan at 3%, and Justice Sotomayor (alone) at 0%.148×148. Table 5, infra.

The distribution also changes a bit, but not dramatically, once one takes into account the obvious possibility that the Chief had more opportunities to assign salient cases for some of the Justices than for others.149×149. See Table 6, infra. The number of salient assignment opportunities naturally depends on how often the Chief and a particular Justice were in the majority together in salient cases, which are more likely to be closely divided than the average case on the Court’s docket.150×150. For those cases in which the Chief was in the majority, the average number of Justices in the majority for salient and non-salient cases was 6.5 and 7.6, respectively. See Tables 4 & 6, infra. These averages were calculated using our dataset by comparing the number of Justices in the majority across salient and non-salient cases assigned by the Chief Justice. For instance, the Chief assigned Justice Souter twenty-four total cases, only two of which were salient, but the Chief had only fifteen total opportunities to assign Justice Souter a salient case, which means he assigned Justice Souter a salient case 13% of the times that he could.151×151. See Table 6, infra. As it happens, however, Justice Souter is the only Justice for whom there is a marked percentage increase.152×152. See Tables 5 & 6, infra. And Justice Kennedy’s percentage is the only one to decrease significantly. Although 31% of his assignments were in salient cases, that amounted to 23% of his assignment opportunities because he was aligned with the Chief in all but three of the 85 salient cases. The percentages for the others all increased (except for the Chief and Justice Alito), but none to the extent Justice Souter’s percentage increased.153×153. See Tables 5 & 6, infra.

Three interesting inferences are suggested by these statistics. The first is that putting aside the two most junior Justices (Sotomayor and Kagan), Justices Thomas and Ginsburg are the most disfavored for the Chief’s assignments of opinions in salient cases and Justices Alito and Breyer are relatively favored. Justices Alito and Breyer are both junior to Justices Thomas and Ginsburg, yet each received far more salient assignments than either Justice Thomas or Justice Ginsburg. That is evident both in terms of the percentage of salient assignments based on all of the Chief’s opinion assignments to each of those Justices and in terms of the percentage of opportunities that the Chief had to assign those Justices a salient case. On the former measure, Justices Alito and Breyer bested Justice Thomas by multiples of almost three and greater than two respectively, and they bested Justice Ginsburg by multiples of greater than five and three respectively.154×154. See Table 5, infra. On the latter measure, Justice Alito and Justice Breyer both received salient opinion assignments approximately two-and-a-half times more frequently than Justice Thomas and three times more frequently than Justice Ginsburg.155×155. See Table 6, infra.

To be sure, Justices Sotomayor’s and Kagan’s numbers and rates of salient opinion assignments are similarly low. But that is less telling because they are also the most junior Justices, and on that basis can fairly be expected to receive fewer salient assignments than far more senior Justices like Thomas and Ginsburg, who have served on the Court for 25 and 22 years, respectively.156×156. See Members of the Supreme Court of the United States, supra note 37. It nonetheless may be significant that Justice Sotomayor is the only Justice to have received no salient opinion assignments from the Chief during her six Terms on the Court.157×157. See Tables 5 & 6, infra. Zero is inherently unique and in this particular context could be a bit portentous.

The second inference suggested is that Justice Scalia is not favored for salient opinion assignments notwithstanding both his seniority and his shared ideology with the Chief. Justice Alito is significantly junior to Justice Scalia, who has served on the Court twenty years longer than Justice Alito,158×158. See Members of the Supreme Court of the United States, supra note 37. yet the two Justices have received precisely the same number of salient opinion assignments — eleven. Their percentages of salient opinion assignments compared to non-salient assignments are essentially the same (15% for Justice Scalia and 16% for Justice Alito) and their percentages of salient assignments compared to assigning opportunities are also the same (15%). Especially given how much seniority plays a systematically important role within the Court — from the seating in the Courtroom to the order of voting at conference159×159. The seating in the Courtroom of the Justices is strictly hierarchical based on seniority. The Chief sits in the middle of the bench and the most senior Associate Justice and second-most senior Associate Justice sit to his immediate right and left, respectively. The same right/left pattern continues, with the most junior Justice, currently Justice Kagan, sitting to the Chief’s far left. See Visitor’s Guide to Oral Argument, Supreme Court of the United States, http://www.supremecourt.gov/visiting/visitorsguidetooralargument.aspx (last updated Oct. 14, 2015) [http://perma.cc/SNE4-JZDU]. Seniority similarly governs the seating and the order of voting at the private conference of the Justices in which the Justices discuss the cases and vote after the oral arguments. The Chief sits at the head of the conference table and the most senior Associate Justice at the other end, and the seating along the sides is also by seniority. The Chief votes first, followed by the most senior Associate Justice until they reach the most junior Justice, who votes last. The most junior Justice is also responsible for note-taking and for answering the door if anyone knocks. See Rehnquist, supra note 78, at 252–54. — this is a striking result. Scholars reviewing Chief Justice Rehnquist’s assignment practices similarly found no preference for Justice Scalia and a marked preference for assigning the important cases to Justices Byron White and Lewis Powell instead.160×160. See Maltzman & Wahlbeck, supra note 62, at 126; see also Davis, supra note 51, at 69 (concluding that Justice White received a disproportionate share of Chief Justice Rehnquist’s assignments in important cases during Rehnquist’s first few years as Chief).

There are many possible explanations for this discrepancy. The most obvious is that the Chief has greater confidence that Justice Alito will write the kind of opinion that the Chief prefers in the most important cases, either in substance or in style. The Chief, for instance, has expressed a general preference for narrower rather than broader rulings,161×161. See Chief Justice John Roberts, Georgetown University Law Center Commencement Address (May 20, 2006), quoted in Chief Justice Says His Goal Is More Consensus on Court, N.Y. Times (May 22, 2006), http://www.nytimes.com/2006/05/22/washington/22justice.html (“If it is not necessary to decide more to a case, then in my view it is necessary not to decide more to a case.”). The kinds of unintended consequences that may be created when a Court ruling is written more broadly than necessary are illustrated by a recent opinion for the Court assigned by the Chief to Justice Thomas, Reed v. Town of Gilbert, 135 S. Ct. 2218 (2015); see also Adam Liptak, Court’s Free-Speech Expansion Has Far-Reaching Consequences, N.Y. Times (Aug. 17, 2015), http://www.nytimes.com/2015/08/18/us/politics/courts-free-speech-expansion-has-far-reaching-consequences.html (“In an otherwise minor decision about a municipal sign ordinance, [Justice Thomas’s majority opinion in Reed] transformed the First Amendment.”). while Justice Scalia has in his own opinions ridiculed the Chief for what Justice Scalia describes as “faux judicial restraint.”162×162. FEC v. Wis. Right to Life, Inc., 551 U.S. 449, 499 n.7 (2007) (Scalia, J., concurring in part and concurring in the judgment); Linda Greenhouse, Even in Agreement, Scalia Puts Roberts to Lash, N.Y. Times (June 28, 2007), http://www.nytimes.com/2007/06/28/washington/28memo.html. Justice Alito has also displayed a willingness, unlike Justice Scalia, to carve the more middle-ground, less absolute positions that the Chief favors.163×163. For example, the Chief assigned Justice Alito the plurality opinion in Hein v. Freedom from Religion Foundation, Inc., 551 U.S. 587 (2007), a significant taxpayer-standing opinion that Justice Scalia harshly criticized for exhibiting judicial “minimalism.” See id. at 633 (Scalia, J., concurring in the judgment) (“Minimalism is an admirable judicial trait, but not when it comes at the cost of meaningless and disingenuous distinctions . . . .”). The Chief’s own writing style is closer to Alito’s than to Scalia’s, especially to the extent that the latter is well known for his barbed and frequently sarcastic wit.164×164. See Greenhouse, supra note 162. Compare, e.g., Obergefell v. Hodges, 135 S. Ct. 2584, 2626 (2015) (Roberts, C.J., dissenting) (“If you are among the many Americans — of whatever sexual orientation — who favor expanding same-sex marriage, by all means celebrate today’s decision. . . . Celebrate the opportunity for a new expression of commitment to a partner. . . . But do not celebrate the Constitution. It had nothing to do with it.”), with id. at 2630 n.22 (Scalia, J., dissenting) (describing the majority opinion as being written in “the mystical aphorisms of the fortune cookie”).

For that same reason, perhaps the Chief believes that Justice Alito is more likely to produce an opinion that can be more quickly joined by others, without the need for pushback by other Justices unhappy with some of Scalia’s rhetorical flourishes. For a Chief Justice, there can be great institutional value in such speedier and less contentious opinion production. As discussed further below,165×165. See infra pp. 64–65. perhaps the Chief also believes that Justice Alito is less likely to lose the majority than either Justice Scalia or Justice Thomas in closely divided cases, which are more likely to be salient.

The third inference that can be drawn is that the Chief feels a heightened need to assign salient cases to Justice Kennedy. The Chief assigns cases to Justice Kennedy at a rate far higher than to anyone but himself, even though Justice Kennedy does not need the Chief to assign him salient cases to ensure that he receives such high-profile assignments: Justice Kennedy is the member of the Court who is most frequently the most senior Justice in the majority in salient cases in which the Chief is in dissent,166×166. The Chief’s assigning to Justice Stevens only one salient case out of nineteen opportunities in the four Terms that they served together can certainly be explained on that ground. Justice Stevens had plenty of opportunities to assign himself salient cases, which he did. See, e.g., Massachusetts v. EPA, 549 U.S. 497 (2007); Hamdan v. Rumsfeld, 548 U.S. 557 (2006). and therefore Justice Kennedy, more than any other Associate Justice on the bench, can (and does) self-assign the writing of the opinions in those cases.167×167. See, e.g., Obergefell, 135 S. Ct. 2584; Tex. Dep’t of Hous. & Cmty. Affairs v. Inclusive Cmtys. Project, Inc., 135 S. Ct. 2507 (2015); Hall v. Florida, 134 S. Ct. 1986 (2014); United States v. Windsor, 133 S. Ct. 2675 (2013). The most likely explanation is that many of the salient cases in which the Chief and Justice Kennedy are together in the majority are closely divided cases in which Justice Kennedy is the “swing” vote.168×168. See, e.g., Town of Greece v. Galloway, 134 S. Ct. 1811 (2014); United States v. Alvarez, 132 S. Ct. 2537 (2012); Arizona v. United States, 132 S. Ct. 2492 (2012); Salazar v. Buono, 559 U.S. 700 (2010); Citizens United v. FEC, 558 U.S. 310 (2010); Gonzales v. Carhart, 550 U.S. 124 (2007). One of the easiest ways to reduce the risk of the swing Justice swinging the other way is to assign the opinion to that Justice, thereby ensuring that the opinion is one he or she will be willing to join, even if the Court’s holding may be far narrower as a result.169×169. Justice Stevens appears frequently to have done the same in closely divided cases in which Justice Kennedy can be fairly described as having been the “swing” Justice, assigning closely divided cases to Justice Kennedy when Justice Stevens was the most senior Justice in the majority. See, e.g., Caperton v. A.T. Massey Coal Co., 556 U.S. 868 (2009); Kennedy v. Louisiana, 554 U.S. 407 (2008); Boumediene v. Bush, 553 U.S. 723 (2008). In his assignments to Justice Kennedy, the Chief Justice appears to have embraced that tactic.

Finally, what do the statistics tell us about the Chief’s practice in assigning the less interesting cases on the Court’s docket? Are they distributed evenly, or do more junior Justices receive a disproportionately high number of such opinion assignments? Or, if a disparity does exist, do the statistics even more tellingly suggest that the Chief might have other reasons to assign more of these less interesting cases to some Justices rather than to others?

One possible marker of a less interesting case is a unanimous decision, although there are obvious, celebrated counterexamples (Brown v. Board of Education170×170. 347 U.S. 483 (1954). and United States v. Nixon,171×171. 418 U.S. 683 (1974). for example) in which the Justices strove for unanimity precisely because of, and to underscore, their ruling’s importance. Considering nonetheless just the unanimous cases, the most junior Justices, Sotomayor and Kagan, do receive a higher percentage of those assignments. During the past ten Terms, 41% of the Chief’s opinion assignments have been in unanimous cases.172×172. Tables 10–12, infra, in the Appendix to this Article set forth the breakdown of the Chief’s assignments in unanimous cases. For the purposes of this analysis, a unanimous opinion includes any cases decided by a vote of 9–0. It does not include cases that were decided by votes of 8–0 or 7–0, although those rulings could also fairly be described as unanimous. The reason for sticking strictly to 9–0 is that the 8–0 and 7–0 cases were disproportionately cases in which a new Justice had just joined the Court and did not participate in the Court’s consideration of the case. There were 10 such cases in Justice Alito’s first Term, October Term 2005; 2 in Justice Sotomayor’s first Term, October Term 2009; and 15 in Justice Kagan’s first Term, October Term 2010. Including those cases, as well as others in these Justices’ subsequent terms, would accordingly have added a lot of cases in which the most junior Justice was, because of a formal recusal or not yet being on the Court at the time of the oral argument, not available for the opinion assignment. The inevitable upshot of their inclusion would be a significant skewing of the opinion assignment distribution in a large subset of the cases to Justices other than the most junior Justice. Because the primary purpose of this Article’s analysis of the Chief’s assignment practices includes his willingness to assign different kinds of cases to the most junior Justices, it seemed best to eliminate the unanimous cases in which the junior Justices were disproportionately unavailable for assignment. Both Justices Sotomayor’s and Kagan’s percentages are substantially higher than that average: 63% of Justice Sotomayor’s assignments from the Chief have been in unanimous cases, and 60% of Justice Kagan’s have been.173×173. Table 10, infra. Shared ideology also appears, however, to have played a role. 51% of Justice Ginsburg’s assignments have been in unanimous cases, while both Justices Breyer and Alito, who are both junior to Justice Ginsburg, are at 38%.174×174. Table 10, infra. Notably, Justice Thomas’s percentage at 49% is just shy of Justice Ginsburg’s, and substantially higher than Justice Breyer’s.175×175. Table 10, infra. The latter discrepancy might seem odd, except that the Chief has actually been more frequently in full agreement with Justice Breyer in recent Terms than he has been with Justice Thomas, casting doubts on popular assumptions about the reach of the Chief Justice and Justice Thomas’s shared ideology.176×176. See Kedar S. Bhatia, Stat Pack for October Term 2012, SCOTUSblog 19 (June 27, 2013), http://sblog.s3.amazonaws.com/wp-content/uploads/2013/09/SCOTUSblog_StatPack_OT121.pdf [http://perma.cc/PJ8R-M6CX]; Kedar S. Bhatia, Stat Pack for October Term 2013, SCOTUSblog 26 (July 3, 2014), http://sblog.s3.amazonaws.com/wp-content/uploads/2014/07/SCOTUSblog_Stat_Pack_for_OT13.pdf [http://perma.cc/F3MK-T3XJ]; Kedar S. Bhatia, Stat Pack for October Term 2014, SCOTUSblog 28 (June 30, 2015), http://sblog.s3.amazonaws.com/wp-content/uploads/2015/07/SB_Stat_Pack_OT14.pdf [http://perma.cc/4LV9-H68A]. This divergence is also in part a result of Justice Thomas’s clear willingness to depart from an otherwise unanimous Court. See, e.g., Mata v. Lynch, 135 S. Ct. 2150 (2015). The higher percentages of unanimous-opinion assignments for both Justice Thomas and Justice Ginsburg are also consistent with the earlier conclusion that they were disfavored Justices for salient assignments.

A more precise measure of the significantly less interesting cases on the Court’s docket filters the unanimous cases even further by isolating those that were not only unanimous, but also decided both very quickly and with relatively short opinions.177×177. The statistical criteria used to identify the “dogs” of the docket for the purposes of this Article are fourfold: the case must be (1) unanimous; (2) nonsalient; (3) short (12 or fewer slip opinion pages, which places the opinion in the lowest quartile in terms of opinion length); and (4) quickly decided (announced in 75 or fewer days from the last conference date for the session in which it was argued, or 50 or fewer days for cases argued during the April argument session). One obvious limitation of this measure is that a Justice can avoid the “dog” designation simply by writing a longer opinion or by taking longer to issue it. I declined to have the “dog” designation consider whether the Court reversed, affirmed, or vacated and remanded the lower court ruling. The corresponding statistical table (reproduced in the Appendix) does, however, reveal some differentiation among the Justices on this variable, with Justices Ginsburg, Kennedy, Breyer, and Kagan writing opinions that affirm the least often, and the Chief Justice and Justice Sotomayor writing opinions that affirm the most often. See Table 14, infra. This smaller subset of cases likely comes closer to capturing those that the current Chief once described as the “dogs” on the Court’s docket, while commenting “I can’t take all the good ones for myself, . . . I have to take my share of the dogs.”178×178. O’Brien, supra note 53, at 270.

The Chief does in fact assign himself a fair share of the dogs. Just over 12% of all the cases that he assigns meet my statistical criteria for dogs, and 11% of his self-assignments are such cases.179×179. See Table 15, infra. Justice Breyer has received more than his proportionate share of the dogs — thirteen cases, amounting to 21% of his assignments — followed by Justices Ginsburg (18%), Sotomayor (17%), and Thomas (15%).180×180. Table 15, infra. Consistent with the Chief’s heavier reliance on Justices Kennedy and Alito for the more salient cases, each had a much lighter share of the dogs (3% and 7%, respectively).181×181. Table 15, infra. Interestingly, Justice Scalia’s dog assignments have trended upwards over time,182×182. See Table 16, infra. The Chief assigned Justice Scalia just one dog during his first five Terms as Chief and eight dogs during the last five Terms. Id. While the numbers are too small to indicate any kind of definite trend, the difference is notable. and Justices Thomas, Ginsburg, and Sotomayor have had distinct Terms during which they have received an especially high number of dog assignments. The Chief’s assigning Justice Ginsburg three dogs out of four total assignments from the Chief during October Term 2008 is the highest percentage for a single Term, and Justice Sotomayor’s receiving four dog assignments (out of eight total Chief assignments) during October Term 2013 is the highest absolute number of such assignments received by any Justice for one Term.183×183. See Table 16, infra. In contrast to Justice Sotomayor’s relatively low profile in terms of her opinion assignments, the Justice has frequently had the highest profile outside the Court, underscored by her “book tour” in 2013 and 2014, which reportedly even prompted her to reschedule the swearing-in of the Vice President of the United States several hours earlier than the constitutionally assumed time of noon so that the Justice could make an appearance at Barnes & Noble in New York later that same day. See Jodi Kantor, On Book-Tour Circuit, Sotomayor Sees a New Niche for a Justice, N.Y. Times (Feb. 3, 2013), http://www.nytimes.com/2013/02/04/us/politics/book-tour-rock-star-sotomayor-sees-an-even-higher-calling.html.

C. Closely Divided Cases, Majority Preservation, and Institutional Messaging

Likewise revealing is the Chief’s practice of assigning opinions in the more closely divided cases.184×184. A “closely divided” case for these purposes means a case decided by either a 5–4 vote or a 5–3 vote. Although 19% of all the Chief’s case assignments occurred in closely divided cases, 34% of the Chief’s assignments to Justices Kennedy and Alito were in closely divided cases — far greater than for others, including the Chief himself at 25%.185×185. See Table 7, infra. Those high percentages likely reflect the need to keep Justice Kennedy in the majority fold as well as the confidence that the Chief has in both Justice Kennedy’s and Justice Alito’s abilities to write an opinion for the Court that will maintain the tentative majority established by the initial conference vote.

Unlike some past Chiefs, however, Chief Justice Roberts has shown no disinclination to assign himself closely divided cases. Reportedly, some Chiefs have generally disfavored themselves for such assignments so that they could “remain at a distance from judicial divisiveness” and the related suggestion that they were displaying an absence of effective leadership.186×186. Slotnick, supra note 53, at 76; see also Brenner, supra note 53, at 764. With 25% of his self-assignments being in closely divided cases, the current Chief has the third-highest percentage of all the Justices187×187. See Table 7, infra. and therefore apparently harbors no such concerns about messaging. Alternatively, even if he harbors any such concerns, he may believe himself more qualified than others to walk the tightrope sometimes required to hold together such fragile majorities — perhaps because he views himself as the very Justice at the “margin” whom scholars have theorized should receive the assignment in order to maximize the likelihood of maintaining the fragile majority coalition established by the conference vote.188×188. Maltzman & Wahlbeck, supra note 51, at 423.

No less revealingly, the closely divided case statistics show that when Chief Justice Roberts has the opportunity, he is counterintuitively more likely to assign closely divided cases to some of the Justices considered “liberals” than to some who are characterized as “conservatives.” Chief Justice Rehnquist engaged in the same practice.189×189. See Maltzman & Wahlbeck, supra note 62, at 126; Wahlbeck, supra note 40, at 1753.

For instance, Chief Justice Roberts and Justice Ginsburg have been in the majority together in only thirteen closely divided cases, but the Chief has assigned Justice Ginsburg the majority opinion in four of those cases, or 31% of the total number of possible assignments.190×190. See Table 8, infra. His comparable percentage for Justice Stevens was also 29% and for Justice Kagan was 40%.191×191. Table 8, infra. Those percentages are markedly higher than the frequency with which the Chief has assigned an opinion in a closely divided case to two “conservative” members of the Court, Justices Scalia and Thomas;192×192. Table 8, infra. when the Chief had the opportunity, he assigned Justices Scalia and Thomas opinions in closely divided cases only 20% and 15% of the time, respectively.193×193. Table 8, infra. The lowest percentage of the Chief’s closely divided assignments has been to Justice Sotomayor, who has received only one out of nine opportunities and out of thirty-five assignments overall.194×194. Tables 7 & 8, infra. That case was Hall v. United States, 132 S. Ct. 1882 (2012). Justice Sotomayor’s low number of closely divided assignments is consistent with her receiving no salient assignments and a larger number of the unanimous and less interesting assignments.195×195. See supra pp. 59–60, 62, 63–64.

Regarding Justices Scalia and Thomas, here again, the Chief’s possible concern with their ability to hold on to a majority might explain their lower percentages.196×196. O’Brien, supra note 53, at 274–75. The need to consider a Justice’s ability to write an opinion capable of maintaining the initial majority established at the conference vote naturally became more important once the Justices routinely circulated draft opinions to the other chambers for their review and approval prior to their final publication. See White, supra note 34, at 1506. Other Chiefs have done the same; Justice Louis Brandeis may have received fewer opinion assignments in important cases not because of any plausible lack of ability, but rather based on the analogous concern that he was “a radical, ‘whose judicial craftsmanship would be unorthodox and unpredictable.’”197×197. Saul Brenner, Is Competence Related to Majority Opinion Assignment on the United States Supreme Court?, 15 Cap. U. L. Rev. 35, 39 (1985) (quoting Slotnick, supra note 53, at 178).

The Chief’s relative reluctance to assign closely divided cases to Justice Scalia may have earlier origins and perhaps has dissipated over time. During the Chief’s very first Term on the Court, October Term 2005, the Chief assigned Justice Scalia the opinion for the Court in Rapanos v. United States,198×198. 547 U.S. 715 (2006). but Justice Kennedy ultimately declined to join Justice Scalia’s opinion and wrote separately199×199. Id. at 759 (Kennedy, J., concurring in the judgment). — this split had the effect of denying Justice Scalia the vote necessary for an opinion of the Court, rendering Scalia’s a plurality opinion instead. Rapanos was a very significant environmental law case, and the lack of a majority ruling has caused much subsequent confusion in the lower courts.200×200. See Matthew A. Macdonald, Case Comment, Rapanos v. United States and Carabell v. United States Army Corps of Engineers, 31 Harv. Envtl. L. Rev. 321, 327–28 (2007) (“The lack of a majority opinion has created doubt and confusion about which opinion(s) to apply.” Id. at 327.); Editorial, Cleaning up the Clean Water Act, N.Y. Times (May 27, 2007), http://www.nytimes.com/2007/05/27/opinion/27sun3.html (“A series of murky Supreme Court decisions have left the agencies responsible for enforcing the Clean Water Act in a state of confused paralysis . . . .”). The reason for Justice Kennedy’s departure is also clear: Justice Scalia’s plurality was much further reaching in its precedential impact than Justice Kennedy’s opinion.201×201. Compare Rapanos, 547 U.S. at 730–39 (adopting a narrow construction of Army Corps of Engineers’ authority under the Clean Water Act), with id. at 776–83 (Kennedy, J., concurring in the judgment) (arguing for a more flexible interpretation of the scope of Corps authority). What is striking is how the Chief shied away from assigning Justice Scalia any closely divided cases immediately afterwards. The next Term, October Term 2006, the Chief had fifteen opportunities to assign Justice Scalia an opinion for the Court in a closely divided case.202×202. See Leegin Creative Leather Prods., Inc. v. PSKS, Inc., 551 U.S. 877 (2007); Parents Involved in Cmty. Sch. v. Seattle Sch. Dist. No. 1, 551 U.S. 701 (2007); Nat’l Ass’n of Home Builders v. Defs. of Wildlife, 551 U.S. 644 (2007); Hein v. Freedom from Religion Found., Inc., 551 U.S. 587 (2007); FEC v. Wis. Right to Life, Inc., 551 U.S. 449 (2007); Morse v. Frederick, 551 U.S. 393 (2007); Bowles v. Russell, 551 U.S. 205 (2007); Fry v. Pliler, 551 U.S. 112 (2007); Uttecht v. Brown, 551 U.S. 1 (2007); Ledbetter v. Goodyear Tire & Rubber Co., 550 U.S. 618 (2007); Schriro v. Landrigan, 550 U.S. 465 (2007); Gonzales v. Carhart, 550 U.S. 124 (2007); Limtiaco v. Camacho, 549 U.S. 483 (2007); Lawrence v. Florida, 549 U.S. 327 (2007); Ayers v. Belmontes, 549 U.S. 7 (2006); see also Table 9, infra. He assigned him only one and that was in a case unanimous in all respects but one203×203. Fry, 551 U.S. 112. In Fry, the dissent was limited to the question of whether the Court should address an additional legal issue, id. at 122–25 (Stevens, J., concurring in part and dissenting in part), which the majority declined to do on the ground that the question was not properly presented, id. at 120–21 (majority opinion). — despite Justice Scalia’s stature as the most senior Associate Justice in the majority. Justice Kennedy received four of those assignments204×204. Leegin, 551 U.S. 877; Uttecht, 551 U.S. 1; Carhart, 550 U.S. 124; Ayers, 549 U.S. 7. and Justice Alito, a Justice far junior to Justice Scalia, received three.205×205. Nat’l Ass’n of Home Builders, 551 U.S. 664; Hein, 551 U.S. 587; Ledbetter, 550 U.S. 618. Notably, Justice Thomas received four opinion assignments in closely divided cases that Term, strongly suggesting the Chief then felt very comfortable assigning Thomas closely divided cases. Table 9, infra. This allocation sharply contrasts to the Chief’s relative reluctance to assign closely divided cases to Thomas since that Term. See id. The Chief has instead assigned to Justice Thomas more unanimous opinions, 36, than to any other Justice, see Table 10, infra; Thomas has been assigned the greatest percentage of the unanimous opinions (15%) of the current Roberts Court, see Table 12, infra. In later Terms, however, the Chief appeared less reluctant to assign closely divided cases to Justice Scalia, assigning him three such cases in October Term 2014.206×206. See Table 9, infra.

Another reason for the Chief’s tendency to assign closely divided cases to Justices with whom he votes less frequently could be to contradict the popular notion that the Justices cast their votes based merely on their own partisan political preferences rather than on the application of neutral legal principles.207×207. Cf. McElwain, supra note 21, at 18 (describing how Chief Justice Hughes would assign “certain Southern civil rights cases” to Justice Hugo Black “with an eye to the unfortunate controversy which had enveloped Justice Black at the time of his appointment to the bench”). The Chief stressed in his opening statement during his Senate confirmation hearings that “[j]udges are not politicians”208×208. Confirmation Hearing on the Nomination of John G. Roberts, Jr., to Be Chief Justice of the United States, Hearing Before the S. Comm. on the Judiciary, 109th Cong. 56 (2005). and has since echoed that theme in public speeches in describing how the Justices decide cases: “We are not Democrats and Republicans in how we go about it.”209×209. Joe Duggan, Chief Justice: Court Detached from Partisanship of Congress, Omaha World-Herald (Sept. 20, 2014, 4:45 AM), http://www.omaha.com/eedition/iowa/articles/chief-justice-court-detached-from-partisanship-of-congress/article_eded7f40-8141-57d9-8fc1-f45e7b1f95ae.html [http://perma.cc/2BA4-L2S8]. Accordingly, when a particular Justice votes in a manner that might surprise people because it seems contrary to what the public assumes are that Justice’s own political preferences, the Chief can highlight that publicly by assigning the opinion of the Court to that Justice. Studies have also shown that members of the public are more willing to agree with a Court opinion when an “ideologically congruent [J]ustice” writes the opinion, and therefore such assignments can promote public acceptance and mute controversy.210×210. Scott S. Boddery & Jeff Yates, Do Policy Messengers Matter? Majority Opinion Writers as Policy Cues in Public Agreement with Supreme Court Decisions, 67 Pol. Res. Q. 851, 853–54 (2014).

Chief Justice Hughes reportedly assigned opinions to particular Justices for this same reason.211×211. See supra note 66. So too did Chief Justice Warren, in assigning conservative Justice Tom Clark the Court’s opinion in liberal cases.212×212. Ulmer, supra note 51, at 67. The most famous example, however, is Chief Justice Stone’s decision to reassign a major civil rights ruling from Justice Frankfurter to Justice Stanley Reed after other members of the Court expressed concern about a pro–civil rights ruling being authored by a Justice from the Northeast rather than a Justice from the South.213×213. Del Dickson, The Supreme Court in Conference (1940–1985), at 197 n.1 (2001). (Justice Thomas may have harbored analogous concerns in deciding last Term not to self-assign the Court’s opinion in Walker v. Texas Division, Sons of Confederate Veterans, Inc.,214×214. 135 S. Ct. 2239 (2015). in which the Court rejected a First Amendment challenge to a state agency’s refusal to approve a motor-vehicle license plate design featuring a Confederate battle flag.215×215. Id. at 2243–44. Justice Thomas was the most senior Justice in the majority in Walker. Instead of writing Walker, a high-profile, important First Amendment case, Justice Thomas authored the opinion for the Court in Bank of America, N.A. v. Caulkett, 135 S. Ct. 1995 (2015), a seven-page unanimous bankruptcy opinion (with two and one half pages of legal analysis) assigned by the Chief Justice. And because Justice Thomas assigned Walker to Justice Breyer, Justice Breyer ended up writing eight opinions for the Term and Justice Thomas only seven. Justice Thomas’s decision not to assign himself Walker is both surprising and revealing. )

The judicial statistics suggest that Chief Justice Roberts has the same proclivity toward institutional messaging through opinion assignments. Not just in closely divided cases,216×216. Table 13, reproduced in the Appendix below, infra, lists the closely divided cases in which the majority included the Chief along with an unusual and therefore unlikely line-up of Justices, defined for this purpose as any closely divided case in which the Chief was joined in a five-Justice majority with at least one of the following Justices: Stevens, Souter, Ginsburg, Breyer, Sotomayor, or Kagan. but more broadly, the Chief regularly assigns opinions to “liberal” Justices in cases in which the public might consider the outcome “conservative” and to “conservative” Justices in cases in which the public might consider the outcome “liberal.” For example, Justice Stevens wrote the Court opinions upholding voter identification laws217×217. Crawford v. Marion Cty. Election Bd., 553 U.S. 181 (2008). and rejecting altogether a claim for damages against Shell for hazardous waste contamination;218×218. Burlington N. & Santa Fe Ry. Co. v. United States, 556 U.S. 599 (2009). Justice Souter authored the opinion reducing the plaintiffs’ right to recover punitive damages for the Exxon Valdez oil spill;219×219. Exxon Shipping Co. v. Baker, 554 U.S. 471 (2008). Justice Ginsburg wrote the opinion overturning a lower court ruling in favor of environmentalists suing power plants for contributing to climate change;220×220. Am. Elec. Power Co. v. Connecticut, 131 S. Ct. 2527 (2011). and Justice Breyer authored the opinion limiting a tobacco company’s liability for punitive damages to a smoker who died of lung cancer.221×221. Philip Morris USA v. Williams, 549 U.S. 346 (2007).

On the other side of the ledger, the Chief Justice assigned himself the opinions upholding state campaign finance limitations on judicial candidates,222×222. Williams-Yulee v. Fla. Bar, 135 S. Ct. 1656 (2015). rejecting prosecutors’ claim that police may search the digital information on a cell phone without a warrant,223×223. Riley v. California, 134 S. Ct. 2473 (2014). and, of course, preserving the viability of President Obama’s Affordable Care Act by allowing federal subsidies to extend to states with federal insurance exchanges;224×224. King v. Burwell, 135 S. Ct. 2480 (2015). Justice Scalia wrote the opinions for the Court overturning a conviction based on the admission of evidence obtained from a GPS device without a warrant,225×225. United States v. Jones, 132 S. Ct. 945 (2012). and in favor of an employee’s claim of discrimination based on disparate treatment;226×226. EEOC v. Abercrombie & Fitch Stores, Inc., 135 S. Ct. 2028 (2015). and Justice Alito wrote the Court’s opinion in favor of a Muslim inmate’s right to grow a beard in accordance with his religious beliefs.227×227. Holt v. Hobbs, 135 S. Ct. 853 (2015). A better example for Justice Alito is Justice Stevens’s assignment to Justice Alito of the majority opinion in Gomez-Perez v. Potter, 553 U.S. 474 (2008), ruling that federal law bars employers from retaliating against a federal employee who is complaining of age discrimination, id. at 477.

III. Conclusion

Mark Twain notwithstanding,228×228. The celebrated humorist did not originate but did popularize the saying that “there are three kinds of lies: lies, damned lies, and statistics,” which has been attributed to Benjamin Disraeli. See Michael Wheeler, Lies, Damned Lies, and Statistics: The Manipulation of Public Opinion in America ix (1976); Proctor & Gamble Mfg. Co. v. Fisher, 449 U.S. 1115, 1118 (1981) (Rehnquist, J., dissenting from denial of certiorari). Professor, then Justice, Felix Frankfurter was plainly on to something. Judicial statistics do have a story to tell, including about the too often overlooked “administrative side” of a Chief Justice’s responsibilities. Chief Justice Roberts is no exception in this regard.

The judicial statistics demonstrate that during the past ten Terms, Chief Justice Roberts has achieved a significantly higher degree of numeric equality in opinion assignments than any of the sixteen Chiefs who preceded him. The Chief’s assignments also reflect his stated preference for judicial modesty. Unlike many past Chief Justices, the current Chief does not assign himself most of the cases or most of the highest-profile cases, nor does he display a reluctance to take on the closely divided cases that highlight the degree of division within the Court.

To be sure, the Chief’s assignments reflect his view that there are certain high-profile cases that warrant the extra symbolic value of being authored by the Chief Justice. He has assigned himself the salient cases more often than he has any other Justice. But, in that respect, his self-assignments are consistent with the Court’s tradition, and the public’s expectation about the responsibility of the Chief to take ownership of such cases.229×229. See Frankfurter, supra note 21, at 4 (“[T]here are occasions when an opinion should carry the extra weight which pronouncement by the Chief Justice gives.”).

The Chief’s assignments, however, are not merely a product of a desire for numeric parity or an expression of judicial modesty. In the assignment of both the salient, higher-profile cases and the closely divided cases, the Chief’s assignments favor Justices Kennedy and Alito and disfavor Justice Ginsburg some and Justice Sotomayor more so. Justices Kennedy and Alito’s shared ideology with the Chief explains this favoritism in part, but not completely. The Chief also disfavors Justice Thomas, does not favor Justice Scalia, and assigns in a manner that relatively favors Justice Breyer and shows an increasing potential to favor Justice Kagan. In these respects, the Chief seems highly motivated to make assignments to those in the majority who he believes can write more narrowly and are better able to maintain (and not lose) the majority established at conference with less rather than more hassle, which is why he likely disfavors assignments of higher-profile and closely divided cases on a relative basis to other conservative Justices. The Chief is also likely constrained by his heightened need to ensure that Justice Kennedy does not wander from the majority view established at conference, which places a special premium on assignments to Justice Kennedy.

Finally, just as some commentators have contended that the Chief’s votes in certain highly contentious cases reflect his strong commitment to maintaining the Court’s institutional stature and a nonpartisan view of judging,230×230. See, e.g., Adam Liptak, Angering Conservatives and Liberals, Chief Justice Roberts Defends Steady Restraint, N.Y. Times (June 26, 2015), http://www.nytimes.com/2015/06/27/us/chief-justice-john-roberts-defends-steady-restraint.html; Ruth Marcus, Opinion, The Obamacare and Marriage Rulings Prove Justice Roberts Is No Partisan, Wash. Post (June 26, 2015), https://www.washingtonpost.com/opinions/john-roberts-is-no-partisan-jurist/2015/06/26/788e420a-1c22-11e5-93b7-5eddc056ad8a_story.html [https://perma.cc/T7R3-MGDZ]; Jeffrey Rosen, Opinion, John Roberts, the Umpire in Chief, N.Y. Times (June 27, 2015), http://www.nytimes.com/2015/06/28/opinion/john-roberts-the-umpire-in-chief.html. his opinion assignments promote that same objective. It is evident in the Chief’s self-assignment of certain high-profile opinions in which the ruling he is authoring departs sharply from the policy agenda of the political party whose President championed his nomination. And the same is true in the Chief’s assignments of cases to other Justices, in which he regularly seeks to assign opinions with seemingly conservative results to Justices considered “liberal,” and vice versa.

Today, as in Justice Frankfurter’s time, a Chief Justice’s “administrative side” offers an important story for the Harvard Law Review to tell (at least online), especially with regard to the Chief’s exercise of opinion assignment authority. The current Chief’s exercise of that authority during the past ten Terms tells us a lot about him: his commitment to numeric equality combined with his quieter but no less significant attention to strategic factors that advance his preferred legal rulings, maintain fragile coalitions, and promote the image of the Court as a nonpartisan decisionmaker. Striking such a balance requires, as Justice Frankfurter also made clear, “the most delicate judgment.”231×231. Frankfurter, supra note 21, at 4.

* Howard J. & Katherine W. Aibel Professor of Law, Harvard Law School. I would like to thank my colleague Mark Tushnet for commenting on an earlier draft of this article; Dr. Arevik Avedian, Applied Research Statistician, of the Harvard Law School Law Library for her invaluable assistance in creating the database upon which the statistical tables in this Article are based, and determining their relative statistical significance; Robert Niles, Harvard Law School and Harvard Business School Class of 2016, for his skillful assistance in preparing this Article, especially in working with me to produce the statistical tables from that initial database and reviewing repeated drafts; and my assistant, Matt Rose, for his terrific editorial assistance. Finally, I would like to express my appreciation to a platoon of my Class of 2017 Torts students who, with little notice and on a very short time frame, calculated the number of slip opinion pages in all the cases decided during the past ten Terms as needed to assist in my identification of the “dogs” of the docket. See infra note 177 and accompanying text; Tables 15 & 16, infra.