Shaping supreme court policy through appointments: The impact of a new justice

Minnesota Law Review

Volumn 93, Issue 5, 2009, Pages 1820-1870

  Cameront, Charles ^a   Park, Jee Kwang ^b   Beim, Deborah ^a  

^a PRINCETON UNIVERSITY   (United States)

^b PRINCETON UNIVERSITY   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 68349111819     PISSN: 00265535     EISSN: None     Source Type: Journal    
DOI: None     Document Type: Review

References (126)

1. See DAVID E. LEWIS, THE POLITICS OF PRESIDENTIAL APPOINTMENTS 6-8, 55-58 (2008).
2. See Erwin Chemerinsky, Ideology and the Selection of Federal Judges, 36 U.C. DAVIS L. REV. 619, 620 (2003) ("Every President in American history, to a greater or lesser extent, has chosen federal judges, in part, based on their ideology."); see also Richard L. Revesz, Environmental Regulation, Ideology, and the D.C. Circuit, 83 VA. L. REV. 1717, 1717-18 (1997) (noting the politici-zation of decision making in environmental cases before the D.C. Circuit and how participating judges generally reflect the views held by the party of the appointing President).
3. See Thomas S. Clark, The Separation of Powers, Court-Curbing and Judicial Legitimacy 14-16 (Nov. 5, 2008) (unpublished manuscript), available at http://userwww.service.emory.edu/̃tclark7/constitutional.pdf (discussing various impetuses behind court-curbing legislation).
4. See THOMAS R. MARSHALL, PUBLIC OPINION AND THE SUPREME COURT 78-79, 97-98, 117 (1989) (noting that a majority of Supreme Court decisions reflect public opinion and that the Supreme Court will seldom defy nationwide public opinion).
5. See sources cited supra notes 1-4 (discussing the effectiveness of these tools in varying circumstances).
6. See David A. Strauss, Memo to the President (and His Opponents): Ideology Still Counts, 102 Nw. U. L. REV. COLLOQUY 49, 50 (2007), http://www.law.northwestern.edu/lawreview/ Colloquy/2007/22/LRCol12007n22Strauss.pdf (discussing how President Franklin Roosevelt proposed legislation to reshape the Court, hoping to assure a majority in favor of his legislative programs, after the Supreme Court declared key features of the New Deal unconstitution-al).
7. Charles M. Cameron et al., Strategic Auditing in a Political Hierarchy: An Informational Model of the Supreme Court's Certiorari Decisions, 94 AM. POL. SCI. REV. 101, 113-14 (2000) (noting that the Court can use the certiorari process to enforce its doctrinal preferences throughout the judicial hierarchy).
8. See THE FEDERALIST NO. 70, at 396-97 (Alexander Hamilton) (Clinton Rossiter ed., 1961) (discussing the value of a unitary executive where every magistrate would be personally responsible for his behavior, thereby allowing the chief magistrate to take swift and decisive action).
9. Youngstown Sheet & Tube Co. v. Sawyer (Steel Seizure), 343 U.S. 579 (1952).
10. See SHELDON GOLDMAN, PICKING FEDERAL JUDGES 68-76 (1997) ("Truman and his administration primarily treated judgeships as patronage, in the service of the partisan agenda."); DAVID ALISTAIR YALOF, PURSUIT OF JUSTICES 20, 38-40 (1999) (discussing how President Truman wanted to feel a sense of "personal comfort" with and loyalty from his Supreme Court nominees).
11. See LEWIS, supra note 1, at 20-21 ("Today the federal government employs 2.5 million civilians in full-time positions ....").
12. See, e.g., Department of Homeland Security, Department Subcomponents & Agencies, http://www.dhs.gov/xabout/structure/ (last visited Apr. 13, 2009) ("More than 87, 000 different governmental jurisdictions at the federal, state, and local level have homeland security responsibilities. ").
13. Cf. Emerson H. Tiller & Frank B. Cross, What Is Legal Doctrine?, 100 NW. U. L. REV. 517, 532 (2006) (noting that judges may be less inclined to adhere to strict legal doctrine in cases where the political saliency of the issue is high).
14. See, e.g., Charles M. Cameron et al., Senate Voting on Supreme Court Nominees: A Neoinstitutional Model, 84 AM. POL. SCI. REV. 525, 532 (1990) [hereinafter Cameron et al., Senate Voting] (noting that a strong President can successfully nominate a highly qualified and ideologically congruent Justice); Lee Epstein et al., The Increasing Importance of Ideology in the Nomination and Confirmation of Supreme Court Justices, 56 DRAKE L. REV. 609, 631-35 (2008) (noting that contemporary Justices "may be entrenching their President's views into the law-and may be doing so long after he leaves office"). Collectively and consistently, these papers show that presidents can almost always place a nominee on the Court who is seen as well qualified. See, e.g., Charles M. Cameron & Jee-Kwang Park, Voting for Justices: Change and Continuity in Confirmation Voting 1937-2006, at 9-10 (2007) (unpublished manuscript, on file with authors). More specifically, when the President's co-partisans control the Senate, he can place almost any reasonable nominee on the Court. See id. at 7. Even if the President's copartisans do not control the Senate, a nominee will likely prove successful if her record is void of scandalous behavior, partisan political involvement, or controversial activities and if she is manifestly well qualified for the job. See id. at 10; Epstein et al., supra at 628. And, even when the Senate rejects a nominee, the President can return with an ideologically similar person and achieve success. See Cameron et al., Senate Voting, supra at 532.
15. See STEVEN TELES, THE RISE OF THE CONSERVATIVE LEGAL MOVE-MENT 1-2 (2008) (identifying the fact that conservatives have made a concerted effort to find and place reliable agents on the Court); Ward Farnsworth, The Use and Limits of Martin-Quinn Scores to Assess Supreme Court Justices, with Special Attention to the Problem of Ideological Drift, 101 Nw. U. L. REV. COLLOQUY 1891, 1901-02 (2007), http://www.law.northwestern.edu/ lawreview/v101/n4/1891/LR101n4Farnsworth.pdf (noting that presidents can act to reduce the chances of future erratic ideological behavior while on the Court by choosing their agents more carefully); see also Charles M. Cameron & Jee-Kwang Park, How Will they Vote? Predicting the Future Behavior of Supreme Court Nominees, 1937-2006, at 4, 14-15, 22-24 (2007) (unpublished manuscript, on file with authors), available at http://www.princeton.edu/ ̃ccameron/HowWillTheyVote.pdf (describing how improved measures predict future behavior better while also noting that the predictability of nominees has increased substantially over time). But see Lee Epstein et al., Ideological Drift Among Supreme Court Justices: Who, When, and How Important?, 101 NW. U. L. REV. 1483, 1493, 1497, 1519, 1526 (2007) (commenting that some Justices seem to change their voting behavior on case dispositions over time, thus calling into question the ability of presidents to place faithful agents on the Court).
16. See, e.g., Bernard Grofman & Timothy J. Brazill, Identifying the Median Justice on the Supreme Court Through Multidimensional Scaling: Analysis of "Natural Courts, " 112 PUB. CHOICE 55, 57, 63 (2002); Keith Krehbiel, Supreme Court Appointments as a Move-the-Median Game, 51 AM. J. POL. SCI. 231, 235, 238 (2007) (discussing the move-the-median model and its implication for Supreme Court appointments to shift policy objectives of the Court).
17. See, e.g., Guy-Uriel E. Charles, Colored Speech: Cross Burnings, Epis-temics, and the Triumph of the Crits?, 93 GEO. L.J. 575, 624 (2005) ("For the majority of Justice Marshall's time on the Court, he was one of the most liberal members .. . ").
18. See, e.g., Lee Epstein & Tonja Jacobi, Super Medians, 61 STAN. L. REV. 37, 82 n.183 (2008) (noting commentary suggesting that Justice Thomas is "consistently the most conservative voter on the Court").
19. A "natural court" refers to a Supreme Court whose membership is constant. The departure of a Justice marks the end of one natural court; the arrival of his or her replacement marks the beginning of the next. LEE EPSTEIN ET AL., THE SUPREME COURT COMPENDIUM 313 (1994). Because the bargaining dynamics on a multimembered court reflect its membership, natural courts (rather than annual terms, for instance) form the logical unit for studying bargaining and coalition formation on the Supreme Court. See id. at 303-13 for a listing of Supreme Court natural courts through 1993.
20. Throughout this Article, we label natural courts according to the classifications provided in Harold Spaeth's database. See HAROLD J. SPAETH, THE ORIGINAL UNITED STATES SUPREME COURT JUDICIAL DATABASE (2009), http://www.cas.sc.edu/poli/juri/sctdata.htm. The Spaeth database begins a natural court each time a Justice leaves or joins the Court. As a result, some natural courts identified by the Spaeth database have open seats. Professor Epstein et al. provide an alternative classification that begins a natural court each time a new Justice joins the Court. See EPSTEIN ET AL., supra note 19, at 303-13.
21. To these one might add a case selection effect-the difference in type of cases selected following the turnover in membership. For instance: Did the new Court take more or fewer extremely consequential cases than did the pre-vious Court? We will not consider case selection effects in this Article.
22. The Monopoly Author theory can be seen as the game-theoretic version of the earlier behavioralist "attitudinal model, " which asserted that Justices vote on the merits in accord with their personal ideologies. See JEFFREY A. SEGAL & HAROLD J. SPAETH, THE SUPREME COURT AND THE ATTITUDINAL MODEL REVISITED 86-97 (2002) (discussing the attitudinal approach in detail); see also infra Part IΠ.C (discussing the Monopoly Author approach).
23. See SEGAL & SPAETH, supra note 22, at 86.
24. See THOMAS H. HAMMOND ET AL., STRATEGIC BEHAVIOR AND POLICY CHOICE ON THE U.S. SUPREME COURT 52 (2005) (noting that Segal and Spaeth's attitudinal model "did not provide any kind of carefully specified model" to predict "whether the justice's earlier behavior is influenced by strategic considerations").
25. See, e.g., id. at 79-215 (applying the Median Voter theorem to develop a formal model of strategic decision making on the Supreme Court); Epstein & Jacobi, supra note 18, at 44-49 (same).
26. See, e.g., Krehbiel, supra note 16, at 232-36; Bryon J. Moraski & Charles R. Shipan, The Politics of Supreme Court Nominations: A Theory of Institutional Constraints and Choices, 43 AM. J. POL. SCI. 1069, 1072-73 (1999); David W. Rohde & Kenneth A. Shepsle, Advising and Consenting on the 60-Vote Senate: Strategic Appointments to the Supreme Court, 69 J. POL. 664, 666-67 (2007).
27. See infra Part IΠ.A (explaining the role of the Median Voter approach in Supreme Court decision making).
28. See, e.g., Ward Farnsworth, Dissents Against Type, 93 MINN. L. REV. 1535, 1536-37 (2009) (noting how most legal realists assume that judges vote according to their policy preferences); Lewis A. Kornhauser, Appeal and Supreme Courts, in VII ENCYCLOPEDIA LAW AND ECONOMICS, CIVIL AND CRIMINAL PROCEDURE 45, 54, 57 (2000), available at http://encyclo. findlaw.com/7200book.pdf (noting in a survey of economic literature on judicial appeals and collegίality of courts that various authors assume judges have an ideal or most-preferred policy).
29. See Kornhauser, supra note 28, at 54, 57 (noting that a majority of Supreme Court decision-making theories assume that judges have policy preferences).
30. See Daniel R. Pinello, Linking Party to Judicial Ideology in American Courts: A Meta-Analysis, 20 JUST. SYS. J. 219, 221, 240-43 (1999) (analyzing numerous empirical studies that have sought to connect party identification with judicial ideology and concluding that political party affiliation "is a dependable measure of ideology in modern American courts").
31. SPAETH, supra note 20.
32. Id. at 53-55.
33. See, e.g., Frank B. Cross, Political Science and the New Legal Realism: A Case of Unfortunate Interdisciplinary Ignorance, 92 Nw. U. L. REV. 251, 252, 303 (1997) (discussing Segal and Spaeth's use of the Judicial Database in designing their attitudinal model).
34. See Joseph Bafumi et al., Practical Issues In Implementing and Understanding Bayesian Ideal Point Estimation, 13 POL. ANALYSIS 171, 171-87 (2005); Michael A. Bailey, Comparable Preference Estimates Across Time and Institutions for the Court, Congress, and Presidency, 51 AM. J. POL. SCI. 433 (2007); Grofman & Brazill, supra note 16, at 55-73; Andrew D. Martin & Kevin M. Quinn, Dynamic Ideal Point Estimation via Markov Chain Monte Carlo for the U.S. Supreme Court, 1953-1999, 10 POL. ANALYSIS 134 (2002). In fact, the scaling methods were originally developed by psychologists to infer latent variables like respondent ability (for example, IQ), based on test performance. See, e.g, R. Darrell Bock, A Brief History of Item Response Theory, 16 EDUC. MEASUREMENT: ISSUES AND PRACTICE 21, 21-33 (1997). From the perspective of so-called "item response theory, " each case is equivalent to a test question, a dispositional vote is equivalent to an answer to the question, and the inferred Justice ideology is equivalent to the respondent's ability. See id. A given answer is not assumed to be correct or incorrect, but rather it is assumed that discriminating questions can differentiate liberal and conservative answers. See id.
35. See Bailey, supra note 34, at 434; Martin & Quinn, supra note 34, at 145-47.
36. See Farnsworth, supra note 15, passim.
37. See id. at 1892-94.
38. See Bailey, supra note 34, at 435 fig.2.
39. In similar fashion, the latent respondent ability measures recovered by item response theory typically correlate at .95 or better with the simple percentage of correct answers on the test.
40. See Farnsworth, supra note 15, at 1892-94 ("The Martin-Quinn method keeps track of only one thing: whether a Justice voted to affirm or reverse a case."); Martin & Quinn, supra note 34, at 137. One must make a few other assumptions to tie down the direction of the scale, for example, by assuming that a well-known liberal like William Douglas identifies the liberal end of the scale.
41. See Martin & Quinn, supra note 34, at 137.
42. See Bailey, supra note 34, at 434. These assumptions include: no strategic voting on case dispositions, an unchanging mix of cases on the docket, no response by the Justices to outside pressure such as court-curbing legislation, and invariance of the scale to the entry and exit of Justices, among others.
43. Martin and Quinn have calculated yearly scores. See Martin & Quinn, supra note 34, at 146 tbl.l (reporting posterior density of ideal points of U.S. Supreme Court Justices, 1953-1999, for the constant ideal point model). But these scores are extremely unstable, and we doubt they provide much useful information beyond the lifetime scores. Michael Bailey has estimated similar scores. See Bailey, supra note 34. The Bailey scores are stabilized through "bridging techniques" used to "link actors across time and institutions." Id. at 433-34 (explaining data observations and measures used in the methodology). When Bailey's scores are averaged over the two natural courts, they present a picture similar to the lifetime MQ scores and the court-specific percent liberal voting scores. These averages can be calculated from data at http://www.georgetown.edu/ faculty/baileyma/Data.htm.
44. Cardinal scores for extreme Justices (i.e., Rehnquist and Marshall) can be particularly problematic since by definition there is no more extreme Justice with which to contrast the extremists' voting behavior. As a result, the scaling methods have a hard time pinning down these Justices' scores very precisely.
45. See generally LEE EPSTEIN & JACK KNIGHT, THE CHOICES JUSTICES MAKE (1997) (providing a strategic model of judicial behavior premised on the theory that Justices' votes are shaped by external factors as well as policy preferences); SEGAL & SPAETH, supra note 22 (developing a behavioral model based on the presupposition that Justices' votes are the product of their policy preferences).
46. Theories of Supreme Court decision making can be tested indirectly, for example, by examining predicted patterns of opinion assignment, see Jeffrey R. Lax & Charles M. Cameron, Bargaining and Opinion Assignment on the Supreme Court, 23 J.L. ECON. & ORG. 276 (2007), or by examining patterns in vote fluidity, Jeffrey R. Lax & Kelly T. Rader, Legal Constraints on Supreme Court Decision Making: Do Jurisprudential Regimes Exist?, 71 J. POL. (forthcoming 2009) (manuscript at 2-4), available at http://www.columbia.edu/̃jrl2124/Random%20Regimes.pdf. We, however, emphasize opinion content because our focus is on policy content.
47. See, e.g., JEFFREY ALLAN SEGAL, PREDICTING SUPREME COURT CASES PROBABILISTICALLY: THE SEARCH AND SEIZURE CASES, 1962-1981, at 3-15 (1983).
48. Id.; Lax & Rader, supra note 46 (manuscript at 4, 20-21).
49. Mark J. Richards & Herbert M. Kritzer, Jurisprudential Regimes in Supreme Court Decision Making, 96 AM. POL. SCI. REV. 305, 305-06 (2002) (positing that changes in the law form different "jurisprudential regimes" that influence relevant decision-making processes).
50. See Lax & Rader, supra note 46 (manuscript at 3-5).
51. Jonathan P. Kastellec, The Structure of Legal Rules and the Analysis of Judicial Decisions 4-13 (Jan. 30, 2007) (unpublished manuscript), available at http://papers.ssrn.com/sol3/papers.cfm?abstract-id=960190#.
52. Tom S. Clark & Benjamin Lauderdale, Locating Supreme Court Opinions in Doctrine Space (Mar. 23, 2009) (unpublished manuscript), available at http://userwww.service.emory.edu/̃tclark7/doctrine.pdf.
53. Id. at 6-11.
54. Id. at 11-15; see also infra Part III.
55. Clark & Lauderdale, supra note 52, at 18-21.
56. Id.
57. Id. ("[The] results represent considerable evidence in support of the coalition median model, at least relative to the median justice and author monopoly models.").
58. See infra Part III.
59. Precursor ideas can be found in Chad Westerland, Who Owns the Majority Opinion? An Examination of Policy Making on the U.S. Supreme Court (Aug. 29, 2003) (unpublished manuscript), available at http://www.allacademic.com/ meta/p62042-index.html, and Cliff Carrubba et al., Does the Median Justice Control the Content of Supreme Court Opinions? (unpublished manuscript), available at http://www.law.nyu.edu/ecm-dlv/groups/public/nyu-law- website-academics-colloquia-law-economics-and-politics/documents/documents/ ecm-pro-059079.pdf. We do not fully pursue these precursor ideas in this Article.
60. Charles M. Cameron & Lewis A. Kornhauser, Modeling Collegial Courts (3): Judicial Objectives, Opinion Content, Voting and Adjudication Equilibria (L. & Econ. Res. Paper Series, Working Paper No. 08-54, 2008), available at http://papers.ssrn.com/sol3/papers.cfm?abstract-id= 1296071.
61. But see, for example, Bush v. Vera, 517 U.S. 952, 956, 990 (1996), where Justice O'Connor announced the judgment of the Court, wrote a first opinion joined by Chief Justice Rehnquist and Justice Kennedy, and then wrote a second, unjoined concurring opinion.
62. The simple version of the join technology used here also assumes similar underlying join behavior by the Justices. If a particular Justice displayed idiosyncratic join tendencies-high or low-it would make detection of the most likely opinion location more difficult. Sophisticated statistical methods offer hope of a "fix-up" to this problem, but we do not pursue this point any further here.
63. See generally EPSTEIN ET AL., supra note 19 (calculating many voting indices based on dispositional votes).
64. An important precursor to contemporary theory is Edward P. Schwartz, Policy, Precedent, and Power: A Positive Theory of Supreme Court Decision-Making, 8 J.L. ECON. & ORG. 219 (1992), which first attempted to adapt median voter models to collegial courts, id. at 219-21. Another important precursor is Lewis A. Kornhauser, Modeling Collegial Courts. II. Legal Doctrine, 8 J.L. ECON. & ORG. 441 (1992), which created a framework, adopted in much recent work, that allows one to distinguish case locations, case dispositions, and opinion content. See id. at 443-59.
65. See HAMMOND ET AL., supra note 24; Lax & Cameron, supra note 46. In essence, the theory argues that the Supreme Court is like the floor of Congress (open-rule). Id. at 276.
66. See Carrubba et al., supra note 59. This theory argues that voting on dispositions is always sincere, and that the majority is something like an open-rule Congress, where there is free debate and amendments can be proposed from the floor. Id. at 15-18.
67. See HAMMOND ET AL., supra note 24; Cameron & Kornhauser, supra note 60; Lax & Cameron, supra note 46. In this model, the Supreme Court is like the closed-rule floor of Congress, where time limits are set on debate and no amendments may be made from the floor. Id. at 276-77.
68. See Lax & Cameron, supra note 46. Here, writing costs afford the author some monopoly power. Id. at 276-80.
69. See Cameron & Kornhauser, supra note 60. In this model, authors like proximate opinions but seek joins. Id. at 13.
70. See PETER C. OKDESHOOK, A POLITICAL THEORY PRIMER 80-86 (1992) (providing a description of the conditions under which voting rules select a Condorcet winner).
71. See HAMMOND ET AL., supra note 24 (applying the MV model directly to the study of judicial decision making). The MV approach also emerges as a special case in several other models of the Supreme Court. See Carrubba et al., supra note 59, at 26-28 (explaining that the MV model emerges when the case disposition is unanimous); Lax & Cameron, supra note 46, at 279-83 (recognizing that the MV model emerges when authoring costs are zero).
72. See Carrubba et al., supra note 59, at 3 ("[C]onsiderable empirical evidence reveals the influential position of the median in legislative settings."). The MV theorem does not require an actual dynamic process, as the players may be able to identify the Condorcet proposal and offer it immediately. See David P. Baron, A Dynamic Theory of Collective Goods Programs, 90 AM. POL. SCI. REV. 316, 316-30 (1996) (detailing a dynamic version of the MV model).
73. Indeed, the MV approach laid out by Hammond et al. does not distinguish between joins and concurrences; case location likewise plays no role in the model. See HAMMOND ET AL., supra note 24, at 260. This is not surprising as the standard MV model was originally developed to study settings like committees or legislatures, in which policymaking does not involve dispute resolution. See id. at 75 (recognizing that much of the original research "in-volve[d] research on committee decision-making by majority rule"); Carrubba et al., supra note 59, at 1 ("[C]onsiderable empirical evidence reveals the influential position of the median voter in legislative settings."). The MV approach emerges as a special case in the Entry Blocking model, which does generate case dispositions. See Lax & Cameron, supra note 46, at 279-83. But that model portrays dispositions as wholly incidental to opinion location; opinion location is taken as central. Id.
74. See Carrubba et al., supra note 59, at 15-20 ("If judicial decisions . . . involve a single, dominant dimension, a straightforward application of the median-voter theorem suggests that the median justice should exercise decisive influence over the content of opinions.").
75. See Grofman & Brazill, supra note 16, at 58-63 (stating that the "identity of the median justice is frequently shifting").
76. See id.
77. See Krehbiel, supra note 16, at 233 (describing the effect of a new member on the location of the median voter).
78. See id. at 233-34 (explaining that assumption of a MV theory of the Court is at the heart of the nominations theories).
79. See Grofman & Brazill, supra note 16, at 58-63.
80. See id.
81. See Krehbiel, supra note 16, at 234 (noting that a new Justice on the Court will often not change the Court's median).
82. See Carrubba et al., supra note 59, at 2, 15-20 (laying out the foundational analysis for the MM approach); Westerland, supra note 59, at 8 (speculating prior to Carrubba et al.'s work that such a theory could exist and might be useful in studying the Court).
83. See Carrubba et al., supra note 59, at 9 (recognizing that the MM model "assume[s] that the justices care about the disposition of the case").
84. See, e.g., id. at 5 ("[A] justice will not be indifferent as to a rule that is a bit more conservative .. . [or] liberal than she would like but yields the opposite outcome in that particular case.").
85. See id. at 15 ("[B]argaining over the majority opinion proceeds only among the justices who prefer the disposition adopted by the majority.").
86. See id. at 12 ("Switching sides would require the justices to support a disposition with which they disagree, and they are not willing to do so if they care enough about the disposition." (citation omitted)).
87. See id. at 7 ("[0]ur model predicts that median members of the majority coalition will have disproportional influence over the content of opinions."); see also Westerland, supra note 59, at 14 ("Disagreements about the outcome of a case preclude agreement on the majority opinion, which means the majority coalition could be an autonomous bargaining unit.").
88. See Carrubba et al., supra note 59, at 1 ("|T]he median justice typically does not determine the content of Supreme Court decisions."); Grofman & Brazill, supra note 16, at 58-63 (noting that the "identity of the median justice is frequently shifting").
89. See Westerland, supra note 59, at 16 ("The only way an opinion author can protect against an opinion from within the majority coalition that is more preferable to a majority of justices in the coalition is to write the opinion at the median of the majority coalition.").
90. The author-opinion diagram for any other "connected" disposition coalition (say, 7-2, composed of the first seven Justices in the majority and Burger and Rehnquist in the minority) would be a horizontal line, as in Figure 3, but located at the predicted location for that coalition shown in Figure 4.
91. The data in the table do not distinguish between a 7-2 and 2-7 coalition, and concern only those cases in which all nine Justices participated.
92. See HAMMOND ET AL., supra note 24, at 110.
93. See Cameron & Kornhauser, supra note 60, at 23. The MA approach has a historic connection with the so-called attitudinal model of Segal and Spaeth, which asserts that Justices are free to vote or write opinions as they please. See generally SEGAL & SPAETH, supra note 22. Unlike the attitudinal model, however, the MA approach is a theory of Supreme Court decision making. Consequently, the individual behavior asserted by Segal and Spaeth must emerge endogenously in a bargaining game played by the Justices. Hammond et al. root one version of the MA approach in a well-known model of take-it-or-leave-it bargaining. See HAMMOND ET AL., supra note 24, at 111; see also Tho-mas Romer & Howard Rosenthal, Political Resource Allocation, Controlled Agendas, and the Status Quo, 33 PUB. CHOICE 27, 27-29 (1978). The Entry Blocking model contains the MA approach as a special case, when entry costs for authors other than the designated author are very high. See infra Part III.D. The Gravitational Attraction model also yields an MA result as a special case, when an opinion author is centrally located within a dense cluster of ideal points. See infra Part III.E.
94. Cf. Cameron & Kornhauser, supra note 60, at 23.
95. See Lax & Cameron, supra note 46, at 292 ("If the Chief Justice is a member of the initial majority coalition (empirically, this is by far the most common occurrence), he/she assigns the opinion; if not, the senior justice in the majority does so.").
96. See HAMMOND ET AL., supra note 24, at 66-67; Lax & Cameron, supra note 46, at 278-80.
97. See Lax & Cameron, supra note 46, at 279 (defining legal quality as "clarity, persuasiveness, completeness, or craftsmanship").
98. See id. at 295 ("[E]ach justice must decide how to allocate effort across all the cases in his/her current portfolio of cases, and the Chief or other justice who assigns an opinion must consider the consequences of a heavier workload not only for the resolution of the instant case but also for all the others in the assignee's portfolio.").
99. See id. at 276. But see Carrubba et al., supra note 59, at 5 (arguing that the requirement that the Court produce not only a judgment affirming or rejecting the lower court's position but also a new judicial rule, will "drive a wedge between the content of opinions and the preferred position of the median justice").
100. See Lax & Cameron, supra note 46, at 276.
101. See id. at 288.
102. The EB model makes particularly clear predictions about the quality of opinions, as a function of the author's ideal point and other variables. See id. at 282. We cannot address these predictions here, however.
103. For example, Segal and Spaeth demonstrated how the appointment of Thurgood Marshall solidified the liberal cohesion of the Warren Court. See Harold J. Spaeth & Jeffrey A. Segal, Decisional Trends on the Warren and Burger Courts: Results from the Supreme Court Data Base Project, 73 JUDICATURE 103, 107 (1989). Conversely, "the substitution of Clarence Thomas for Thurgood Marshall moved the Rehnquist Court to the right-but not enough to make it a coherent Court." Neal Devins, Ideological Cohesion and Precedent (or Why the Court Only Cares About Precedent When Most Justices Agree with Each Other), 86 N.C. L. REV. 1399, 1404 (2008).
104. See Lax & Cameron, supra note 46, at 287 (postulating that since "[e]very nonmedian opinion is vulnerable to some counteroffer, " an opinion writer must be as concerned with the stability of his or her opinion-that is, its proximity to the median-as with quality and precedent).
105. See id. at 277.
106. See Cameron & Kornhauser, supra note 60, at 11.
107. See id. at 30.
108. The Mathematica computer code used to generate these estimates is available on request from the authors. By "moderate" values we mean nonzero but not huge writing costs, and non-zero but not overwhelming importance of case clarity. In the example, case disposition value is set to zero, as would hold not only for extreme case locations resulting in a unanimous disposition, but presumably for many or even most cases heard by the Supreme Court.
109. See Steven R. Weisman, Stewart Will Quit High Court July 3; Reasons Not Given, N.Y. TIMES, June 19, 1981, at Al.
110. See Terrance Sandalow, Potter Stewart, 95 HARV. L. REV. 6, 6 (1981).
111. See id. at 8-9.
112. See also supra tbl.l.
113. See 453 U.S. III (1981); fig.8.
114. SEE HENRY J. ABRAHAM, JUSTICES, PRESIDENTS, AND SENATORS: A HISTORY OF THE U.S. SUPREME COURT APPOINTMENTS FROM WASHINGTON TO CLINTON 256 (1999).
115. See id. at 255.
116. See supra fig.8.
117. See infra tbl.3.
118. See supra Part IΠ.C and fig.5.
119. We score liberals as those with MQ scores less than-.33 (Marshall, Brennan, and Stevens); moderates as those with MQ scores between-.33 and .33 (Blackmun and White); and conservatives as those with scores greater than .33 (Powell, Burger, and Rehnquist).
120. See Cameron & Kornhauser, supra note 60, at 30.
121. See generally Lax & Cameron, supra note 46 (hypothesizing that subtle variables such as policy content, writing costs, legal quality, and agenda-setting are crucial considerations in the bargaining and opinion assignment process).
122. We score Justice Blackmun's locations according to the discussion in Part II.
123. For example, in the upper right-hand panel, three points are shown for Marshall's opinions. These correspond to the three highest points on his empirical join function, as the second and third highest points gained almost as many joins as the highest one. If no other locations gained as many joins as the maximum-gaining location, the maximum is triply weighed, to assure equal observations for each Justice.
124. This is a "loess" regression with spans of one-half and two-thirds for Burger 6 and Burger 7, respectively. These spans somewhat smooth the data but allow their local shapes to remain. For an explanation of nonparametric regression, loess regression, and weighted regression, see DAVID HAND ET AL., PRINCIPLES OF DATA MINING 175-77 (2001).
125. See Lax & Cameron, supra note 46, at 292.
126. We write following the election of Democratic President Barack Ob-ama but before he has had an opportunity to nominate any Justices.