The partisan price of justice: An empirical analysis of campaign contributions and judicial decisions

New York University Law Review

Volumn 86, Issue 1, 2011, Pages 69-130

  Kang, Michael S ^a   Shepherd, Joanna M ^a  

^a EMORY UNIVERSITY SCHOOL OF LAW   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 79955505092     PISSN: 00287881     EISSN: None     Source Type: Journal    
DOI: None     Document Type: Article

References (251)

1. 129 S. Ct. 2252 (2009).
2. Id. at 2263.
3. We refer to the highest appellate court in a state as the "state supreme court."
4. The Caperton v. Massey Case: Not for Sale, ECONOMIST, June 13, 2009, at 36.
5. Chris W. Bonneau, The Dynamics of Campaign Spending in State Supreme Court Elections, in RUNNING FOR JUDGE: THE RISING POLITICAL, FINANCIAL, AND LEGAL STAKES OF JUDICIAL ELECTIONS 59, 63 fig.4.1 (Matthew J. Streb ed., 2007).
6. See DEBORAH GOLDBERG & SAMANTHA SANCHEZ, THE NEW POLITICS OF JUDICIAL ELECTIONS 2002, at 15 (Bert Brandenburg ed., 2002), available at http://brennan.3cdn.net/3e06222f06bc229762_yom6bgubs.pdf (With few exceptions, money means victory.).
7. Shirley S. Abrahamson, The Ballot and the Bench, 76 N.Y.U. L. REV. 973, 976 (2001);
8. see also Roy A. Schotland, Elective Judges' Campaign Financing: Are State Judges' Robes the Emperor's Clothes of American Democracy?, 2 J.L. & POL. 57, 77 (1985) (positing that state courts handle 98% of judicial business in United States).
9. Roy A. Schotland, New Challenges to States' Judicial Selection, 95 GEO. L.J. 1077 app. 2, at 1105 (2007).
10. See infra note 109 and accompanying text (describing data).
11. JAMES SAMPLE, LAUREN JONES & RACHEL WEISS, THE NEW POLITICS OF JUDICIAL ELECTIONS 2006, at 18 (Jesse Rutledge ed., 2006), available at http://www.justiceatstake.org/media/cms/NewPoliticsofJudicialElections2006_D2A2449B77CDA.pdf.
12. Id. at 3.
13. Id. at 7.
14. See CHRIS W. BONNEAU & MELINDA GANN HALL, IN DEFENSE OF JUDICIAL ELECTIONS 132 (2009) (arguing that adoption of nonpartisan system does not minimize role of politics and money in judicial elections because nonpartisan elections involve higher levels of spending).
15. See, e.g., Anthony Champagne, Political Parties and Judicial Elections, 34 LOY. L.A. L. REV. 1411, 1418-20 (2001) (describing partisan activity in nonpartisan judicial elections).
16. See BONNEAU & HALL, supra note 13, at 132.
17. Champagne, supra note 14, at 1418
18. Quoting J. CLARK KELSO, JUDICIAL ELECTIONS: PRACTICES AND TRENDS 18 (1999);
19. see also BONNEAU & HALL, supra note 13, at 127 ("[S]tates with highly competitive nonpartisan elections continue to face issues with campaign activities and negative advertising even without partisan labels on the ballot.");
20. Lawrence H. Averill, Jr., Observations on the Wyoming Experience with Merit Selection of Judges: A Model for Arkansas, 17 U. ARK. LITTLE ROCK L.J. 281, 322-23 (1995) (expressing similar skepticism about value of nonpartisan election systems because of persistent partisan endorsements, and urging state not to adopt nonpartisan elections).
21. Joanna M. Shepherd, Money, Politics, and Impartial Justice, 58 DUKE L.J. 623, 629 (2009). In this previous study, a judge's political party affiliation was defined by the political preferences of the voters or governor responsible for their retention (i.e., a judge seeking retention by Republican voters or a Republican governor was defined as a Republican judge).
22. Id. at 655-56.
23. Eighty-seven percent of state appellate court judges must be retained through either partisan elections, nonpartisan elections, or retention elections. Schotland, supra note 8, app. 2, at 1105. In contrast, as explained earlier, 89% of all state judges (appellate and trial) face voters at some point, either in the initial election or when seeking retention.
24. See supra note 8 and accompanying text.
25. See, e.g., Vicki C. Jackson, Packages of Judicial Independence: The Selection and Tenure of Article III Judges, 95 GEO. L.J. 965, 1007 (2007) ("The United States is unusual. in the degree to which it relies on popular elections for the selection or retention of its state court judges.");
26. Hans A. Linde, Elective Judges: Some Comparative Comments, 61 S. CAL. L. REV. 1995, 1996 (1988) (To the rest of the world, the American adherence to judicial elections is as incomprehensible as our rejection of the metric system.).
27. ARTHUR T. VANDERBILT, THE CHALLENGE OF LAW REFORM 14-15 (1955).
28. LARRY C. BERKSON & RACHEL CAUFIELD, AM. JUDICATURE SOC'Y, JUDICIAL SELECTION IN THE UNITED STATES: A SPECIAL REPORT 1 (2004), available at http://www.ajs.org/selection/docs/Berkson.pdf.
29. See Caleb Nelson, A Re-evaluation of Scholarly Explanations for the Rise of the Elective Judiciary in Antebellum America, 37 AM. J. LEGAL HIST. 190, 222-23 (1993).
30. Id. at 205.
31. 2 OFFICIAL REPORT OF THE DEBATES AND PROCEEDINGS IN THE STATE CONVENTION, ASSEMBLED MAY 4TH, 1853, TO REVISE AND AMEND THE CONSTITUTION OF THE COMMONWEALTH OF MASSACHUSETTS 773 (Boston, White & Potter 1853) (statement of Edward L. Keyes).
32. REPORT OF THE DEBATES AND PROCEEDINGS OF THE CONVENTION FOR THE REVISION OF THE CONSTITUTION OF THE STATE OF KENTUCKY 1849, at 273 (Frankfort, A.G. Hodges & Co. 1849) (statement of Francis M. Bristow).
33. See Nelson, supra note 22, at 192 (describing contentious debates in state constitutional conventions adopting elected judiciaries).
34. Id. at 224.
35. See Schotland, supra note 8, at 1094 (noting that constitutional devices were adopted "to limit the potentially disruptive consequences of popular election);
36. cf. Nelson, supra note 22, at 218 (recognizing that some delegates were concerned about overly excited voters but noting that even those delegates were more concerned about unchecked government officials).
37. Schotland, supra note 8, at 1094.
38. Nelson, supra note 22, at 218.
39. Schotland, supra note 8, at 1094.
40. Nelson, supra note 22, at 196 (quoting 2 THE CONSTITUTIONAL DEBATES OF 1847, at 462 (Arthur Charles Cole ed., 1919) (statement of David Davis)).
41. See Schotland, supra note 8, at 1093 (noting increasing adoption of elected judiciaries).
42. BERKSON & CAUFIELD, supra note 21, at 2.
43. Rachel Paine Caufield, In the Wake of White: How States Are Responding to Republican Party of Minnesota v. White and How Judicial Elections Are Changing, 38 AKRON L. REV. 625, 627-28 (2005).
44. Michael R. Dimino, The Futile Quest for a System of Judicial "Merit" Selection, 67 ALB. L. REV. 803, 804 (2004).
45. Schotland, supra note 8, at 1085 (summarizing judicial selection methods nationwide).
46. DAVID B. ROTTMAN ET AL., U.S. DEP'T OF JUSTICE, BUREAU OF JUSTICE STATISTICS, NO. NCJ 178932, STATE COURT ORGANIZATION 1998, at 21 tbl.4 (2000), available at http://bjs.ojp.usdoj.gov/content/pub/pdf/sco98.pdf.
47. Am. Judicature Soc'y, Methods of Judicial Selection, JUD. SELECTION STS., http://www.judicialselection.us/judicial_selection/methods/selection_of_judges.cfm (last visited Feb. 17, 2010).
48. ROTTMAN ET AL., supra note 40, at 21 tbl.4. Illinois, New Mexico, and Pennsylvania hold partisan elections to appoint judges initially to the bench, but they use unopposed retention elections to determine whether incumbent judges keep their positions beyond the initial term of appointment. Am. Judicature Soc'y, supra note 41.
49. ROTTMAN ET AL., supra note 40, at 21 tbl.4.
50. Judges in Massachusetts and New Hampshire are appointed for permanent tenure to age seventy, and Rhode Island grants its judges life tenure. Id. at 26 tbl.5.
51. Id. at 34 tbl.7.
52. Although there are other differences between the selection and retention methods of each state, they can be grouped into these primary categories.
53. The data in this table were collected from ROTTMAN ET AL., supra note 40, at 21 tbl.4, and the American Judicature Society, supra note 41. G = gubernatorial appointment or reappointment, P = partisan election or reelection, N = nonpartisan election or reelection, LA = legislative appointment or reappointment, LE = legislative election or reelection, M = merit plan, R = retention election, and J = reappointment by a judicial nominating commission.
54. In New Hampshire, judges serve until age seventy. ROTTMAN ET AL., supra note 40, at 28 tbl.4.
55. In New Jersey, after an initial gubernatorial reappointment, judges serve until age seventy. N.J. CONST. art. VI, § 6, ¶ 3.
56. In Connecticut, the governor nominates, and the legislature appoints. ROTTMAN ET AL., supra note 40, at 21 tbl.4, 25 n.2.
57. In Ohio, political parties nominate candidates to run in nonpartisan elections. Am. Judicature Soc'y, supra note 41.
58. In Rhode Island, judges have life tenure. ROTTMAN ET AL., supra note 40, at 28 tbl.5.
59. In Louisiana, candidates compete in a blanket primary with party labels on the ballot. The top two primary candidates go on to the general election. Am. Judicature Soc'y, supra note 41.
60. In Massachusetts, judges serve until age seventy. ROTTMAN ET AL., supra note 40, at 27 tbl.5.
61. In Michigan, political parties nominate candidates to run in nonpartisan elections. Am. Judicature Soc'y, supra note 41.
62. Peter D. Webster, Selection and Retention of Judges: Is There One "Best" Method?, 23 FLA. ST. U. L. REV. 1, 19 (1995).
63. Schotland, supra note 8, at 1079 (alteration in original) (quoting William C. Bayne, Lynchard's Candidacy, Ads Putting Spice into Justice Race, COM. APPEAL (Memphis), Oct. 29, 2000, at DS.1).
64. Id. at 1080.
65. Chris W. Bonneau, Patterns of Campaign Spending and Electoral Competition in State Supreme Court Elections, 25 JUST. SYS. J. 21, 27 tbl.6 (2004).
66. Melinda Gann Hall, Competition as Accountability in State Supreme Court Elections, in RUNNING FOR JUDGE: THE RISING POLITICAL, FINANCIAL, AND LEGAL STAKES OF JUDICIAL ELECTIONS 165, 177 (Matthew J. Streb ed., 2007) [hereinafter RUNNING FOR JUDGE].
67. From 1990 to 2000, the average reelection rates for state legislators, U.S. Senators, and U.S. Representatives were 85.1%, 89.3%, and 94.1%, respectively, while state supreme court justices were reelected at a lower average rate of 84.1%.
68. Melinda Gann Hall & Chris W. Bonneau, Does Quality Matter? Challengers in State Supreme Court Elections, 50 AM. J. POL. SCI. 20, 21 (2006);
69. see also Melinda Gann Hall, State Supreme Courts in American Democracy: Probing the Myths of Judicial Reform, 95 AM. POL. SCI. REV. 315, 319 & tbl.5 (2001) (finding that incumbent state supreme court judges have failed to be reelected at rates equivalent to, or greater than, members of U.S. House of Representatives)
70. Bonneau, supra note 5, at 63 fig.4.1.
71. See GOLDBERG & SANCHEZ, supra note 6, at 15 (With few exceptions, money means victory.).
72. MODEL CODE OF JUDICIAL CONDUCT Canon 7B(1)(c) (1989);
73. see also, e.g., MINN. CODE OF JUDICIAL CONDUCT Canon 5(A)(3)(d)(i) (1996) (retaining prohibition on announcement of disputed legal or political views).
74. MODEL CODE OF JUDICIAL CONDUCT Canon 5A(3)(d) annot. at 355-58 (2004) (prohibiting judges only from committing themselves on issues likely to come before court).
75. Schotland, supra note 8, at 1095 n.77.
76. 536 U.S. 765 (2002).
77. E.g., Republican Party of Minn. v. White, 416 F.3d 738, 754, 765-66 (8th Cir. 2005) (en banc) (striking down limits on judges' partisan conduct and personal solicitation of campaign contributions);
78. Weaver v. Bonner, 309 F.3d 1312, 1322 (11th Cir. 2002) (striking down canon prohibiting personal solicitation by judges). But cf. Schotland, supra note 8, at 1095-97 (discussing limited impact of White's permissiveness on judicial campaign behavior).
79. For a thorough explanation of the differences in campaign spending among the states,
80. see Bonneau, supra note 5, at 66-68.
81. The information in this table can be found in id. at 67 tbl.4.2.
82. Note that Texas has two courts of last resort: one for criminal trials and one for civil trials.
83. Because of issues like tort reform and the accompanying interest group involvement, Texas Supreme Court elections are significantly more expensive than elections to the Texas Court of Criminal Appeals.
84. Id. at 67.
85. See Anthony Champagne, Tort Reform and Judicial Selection, 38 LOY. L.A. L. REV. 1483, 1488 (2005) ("It is state judge selection that produces the major battles between economic interests that are concerned with a state's tort law.").
86. SAMPLE ET AL., supra note 10, at 16 fig.9.
87. Id. at 18.
88. Id. Researchers from the National Institute on Money in State Politics categorize donors into interest groups. The categorization of major labor and industry groups into interest groups is straightforward; the researchers use information on individuals' occupation, employer, or from professional directories to code contributions from individuals. About Our Data, NAT'L INST. ON MONEY ST. POL., http://www.followthemoney.org/Institute/about_data.phtml (last visited Feb. 17, 2010). There have been a few election years where lawyers' groups donated more direct campaign funds than business groups, but business groups routinely dominate independent expenditures. SAMPLE ET AL., supra note 10, at 19 fig.12.
89. RACHEL WEISS, INST. ON MONEY IN STATE POLITICS, FRINGE TACTICS: SPECIAL INTEREST GROUPS TARGET JUDICIAL RACES 13 (Aug. 25, 2005), available at http://www.followthemoney.org/press/Reports/200508251.pdf.
90. Id. at 14.
91. Id. at 18.
92. See RONALD D. ROTUNDA & JOHN S. DZIENKOWSKI, LEGAL ETHICS: THE LAWYER'S DESKBOOK ON PROFESSIONAL RESPONSIBILITY § 10.4-4.4(g) (2010) (noting that only campaign contributions, and not independent expenditures, are regulated by Model Code of Judicial Conduct);
93. Elizabeth Garrett & Daniel A. Smith, Veiled Political Actors and Campaign Disclosure Laws in Direct Democracy, 4 ELECTION L.J. 295, 322 (2005) (describing how business groups avoid disclosure requirements under state campaign finance law).
94. See Anthony Champagne, Television Ads in Judicial Campaigns, 35 IND. L. REV. 669, 670 (2002).
95. SAMPLE ET AL., supra note 10, at 1, 3.
96. Id. at 7.
97. DEBORAH GOLDBERG ET AL., THE NEW POLITICS OF JUDICIAL ELECTIONS 2004, at 1 (Jesse Rutledge ed., 2005), available at http://www.justiceatstake.org/media/cms/NewPoliticsReport2004_83BBFBD7C43A3.pdf (presenting data for 2004 elections);
98. GOLDBERG & SANCHEZ, supra note 6, at 9 (presenting data for 2002 elections);
99. SAMPLE ET AL., supra note 10, at 11 (presenting data for 2004 and 2006 elections).
100. Between 2000 and 2003, 21 of the 24 pro-business judges that the Chamber of Commerce supported were elected. Champagne, supra note 80, at 1502-03. In 2004, 15 of the 16 pro-business judges that the Chamber supported were elected. Press Release, Inst. for Legal Reform, Chamber Highlights Successful Pro-Business Election Effort (Nov. 3, 2004), available at http://www.instituteforlegalreform.com/component/ilr_media/30/pressrelease/2004/283.html
101. Champagne, supra note 80, at 1502.
102. GREENBERG QUINLAN ROSNER RESEARCH INC., JUSTICE AT STAKE-STATE JUDGES FREQUENCY QUESTIONNAIRE 5, 11 (2002), available at http://www.justiceatstake.org/media/cms/JASJudgesSurveyResults_EA8838C0504A5.pdf.
103. 552 U.S. 196, 212 (2008) (Kennedy, J., concurring in judgment).
104. GREENBERG QUINLAN ROSNER RESEARCH INC., JUSTICE AT STAKE: FREQUENCY QUESTIONNAIRE 4 (2001), available at http://www.justiceatstake.org/media/cms/JASNationalSurveyResults_6F537F99272D4.pdf.
105. Shepherd, supra note 17, at 669-72 & tbls.7 & 8. In that article, Shepherd reported results suggesting that elected judges' decisions, especially in partisan systems, were influenced by the political preferences of both voters and campaign contributors.
106. Id. at 684.
107. That article primarily focused on the relationship between the political party of the majority of voters-Republican or Democrat-and judges' votes for litigants that were typically aligned with the interests of each political party. It included one estimation that showed, when controlling for the political party of the majority of voters, that judges in partisan systems who received campaign contributions from business groups were more likely to favor the business litigant in cases between a business and an individual person. That article emphasized the importance of political parties to judicial voting, while this article investigates whether, regardless of political preferences, the type of election has a significant influence on the relationship between campaign contributions and judicial voting.
108. Stephen J. Ware, Money, Politics, and Judicial Decisions: A Case Study of Arbitration Law in Alabama, 15 J.L. & POL. 645, 661 (1999).
109. Eric N. Waltenburg & Charles S. Lopeman, Tort Decisions and Campaign Dollars, 28 SOUTHEASTERN POL. REV. 241, 256 (2000).
110. 100 Madhavi McCall, The Politics of Judicial Elections: The Influence of Campaign Contributions on the Voting Patterns of Texas Supreme Court Justices, 1994-1997, 31 POL. & POL'Y 314, 315, 330 (2003) (showing that when two litigants contribute to justices' campaigns, Texas Supreme Court decisions tend to favor litigant that contributed more money).
111. Damon M. Cann, Justice for Sale? Campaign Contributions and Judicial Decision Making 11, 16 (Aug. 10, 2006) (unpublished manuscript), available at http://ssrn.com/abstract=991364.
112. 129 S. Ct. 2252 (2009).
113. Id. at 2263-64.
114. Id. at 2264.
115. Id. at 2257-58.
116. Id. at 2264-65.
117. H. Thomas Wells Jr., President, Am. Bar Ass'n, Statement Regarding the Ruling of the Supreme Court of the United States in Caperton et al. v. A.T. Massey Coal Co., Inc., et al. (June 8, 2009), available at http://www.abanet.org/abanet/media/statement/statement.cfm?releaseid=671.
118. Adam Liptak, Justices Issue a Rule of Recusal in Cases of Judges' Big Donors, N.Y. TIMES, June 9, 2009, at A1.
119. Paul Brace & Melinda Gann Hall, Project Overview, ST. SUPREME CT. DATA PROJECT, http://www.ruf.rice.edu/~pbrace/statecourt/index.html (last visited Feb. 17, 2010). In the rare instance when state dockets exceed 200 cases in a single year, the researchers select a random sample of 200 cases.
120. See Shepherd, supra note 17, at 652.
121. See supra notes 98-101 and accompanying text.
122. 536 U.S. 765 (2002);
123. see supra text accompanying note 74.
124. See, e.g., Rachel P. Caufield, The Changing Tone of Judicial Election Campaigns as a Result of White, in RUNNING FOR JUDGE, supra note 63, at 36 (observing increased campaign spending and overtly political tenor of judicial campaigns since White).
125. See, e.g., Anthony Champagne, Interest Groups and Judicial Elections, 34 LOY. L.A. L. REV. 1391, 1404-05 (2001) (describing independent expenditures and issue advocacy by interest groups for judicial elections).
126. Champagne, supra note 80, at 1483-86 (providing examples of campaign spending and issue advocacy regarding tort reform in state judicial campaigns).
127. In Caperton, Blankenship supported then-attorney Benjamin in his campaign to unseat Justice McGraw.
128. See Caperton v. A.T. Massey Coal Co., 129 S. Ct. 2252, 2257 (2009).
129. The variable Contributions is coded as 0 for judges that received no money from business groups.
130. Mission & History, NAT'L INST. ON MONEY ST. POL., http://www.followthemoney.org/Institute/index.phtml (last visited Feb. 17, 2010).
131. About Our Data, supra note 83.
132. See WEISS, supra note 84, at 4 (discussing primary actors in tort reform debate).
133. See supra notes 61-62 and accompanying text.
134. See supra note 38 and accompanying text.
135. See supra notes 63-65 and accompanying text.
136. Cf. William K. Hall & Larry T. Aspin, What Twenty Years of Judicial Retention Elections Have Taught Us, 70 JUDICATURE 340, 343 (1987) (noting that, between 1964 and 1984, judges facing retention elections were retained over 98% of the time).
137. See supra notes 66-67 and accompanying text.
138. James J. Alfini & Terrence J. Brooks, Ethical Constraints on Judicial Election Campaigns: A Review and Critique of Canon 7, 77 KY. L.J. 671, 684-85 (1989) (discussing ethical codes prohibiting campaigning until opposition forms).
139. See Melinda Gann Hall, Electoral Politics and Strategic Voting in State Supreme Courts, 54 J. POL. 427 (1992) (demonstrating that judges facing reelection will not distinguish themselves by voicing disagreement with perceived popular opinion through dissents);
140. Melinda Gann Hall, Constituent Influence in State Supreme Courts: Conceptual Notes and a Case Study, 49 J. POL. 1117 (1987) (indicating same through case study).
141. See Paul Brace & Melinda Gann Hall, Studying Courts Comparatively: The View from the American States, 48 POL. RES. Q. 5, 24 (1995) ([W]here judges must face voters to retain their positions, state partisan competition exerts a positive influence on support for the death penalty.").
142. That is, if a third, omitted variable has significant influence on voting, and that omitted variable is strongly correlated with retention politics, our analysis may erroneously attribute to the retention agent variable the relationship between voting and the omitted third variable.
143. See generally Paul Brace, Laura Langer & Melinda Gann Hall, Measuring the Preferences of State Supreme Court Judges, 62 J. POL. 387 (2000) (developing PAJID score).
144. Id. at 388;
145. see also id. at 400-04 (showing validity of PAJID scoring and fact that it outperforms other scoring mechanisms).
146. This variable is actually the reverse of the years to retention. As the longest number of years to retention during our sample is 12, the inverse years to retention is 13 minus the years to retention. 133 Thirty-nine states have intermediate appellate courts, and those states' supreme courts have discretionary review. COURT STATISTICS PROJECT, NAT'L CTR. FOR STATE COURTS, STATE COURT CASELOAD STATISTICS, 2006: SUPPLEMENT TO EXAMINING THE WORK OF STATE COURTS 16-67 (2007), available at http://www.ncsconline.org/D_Research/csp/2006_files/StateCourtCaseloadStatistics2006.pdf
147. Conceivably, litigants could decide to settle after review of their case has been granted; the granting of review may be a signal that the court plans to vote ideologically. However, in a study of civil appeals in 46 counties between 2001 and 2005, no litigants withdrew their cases after the courts of last resort granted review, so litigants do not appear to anticipate an ideological outcome.
148. See THOMAS H. COHEN, BUREAU OF JUSTICE STATISTICS, U.S. DEP'T OF JUSTICE, NO. NCJ 212979, APPEALS FROM GENERAL CIVIL TRIALS IN 46 LARGE COUNTIES, 2001-2005, at 9 (2006), available at http://bjs.ojp.usdoj.gov/content/pub/pdf/agctlc05.pdf.
149. COURT STATISTICS PROJECT, supra note 133, at 16-67. In Connecticut, the court sits en banc only by order of the chief justice. Id. at 22. The District of Columbia's highestcourt also sits in panels and en banc. Id. at 24.
150. Id. at 16-67.
151. See, e.g., WILLIAM H. GREENE, ECONOMETRIC ANALYSIS 116-18 (5th ed. 2003) (explaining use of dummy variables in regression analysis).
152. We are unable to include state-level and judge-level fixed effects because most are perfectly collinear with the retention variables, many of which do not change during the four-year sample period.
153. For a general discussion of the probit model
154. see GREENE, supra note 137, at 814-15.
155. We also estimate a series of mixed-effects (hierarchical) logit models that can account for nested errors. The mixed-effects logit models account for two levels of nested groups; the case-specific random effects are nested within state-specific random effects. Although we do not report the results in the interest of brevity, they are consistent with the probit results, confirming that our results are robust to alternative specifications.
156. See generally JEFFREY M. WOOLDRIDGE, ECONOMETRIC ANALYSIS OF CROSS SECTION AND PANEL DATA 405-07 (2002).
157. For each regression, the table also reports the log likelihood that is used to perform a likelihood ratio test. This test determines whether the difference between the log likelihoods of two models is statistically significant; if it is, then the less restrictive model (the one with more variables) is said to fit the data significantly better than the more restrictive model. For more on likelihood ratio tests, see GREENE, supra note 137, at 151-53.
158. See SAMPLE ET AL., supra note 10, at 18.
159. The table reports the marginal effects of all variables on the probability of a judge voting for the business litigant, based on probit estimates. All control variables are included in the estimation but are not reported in the interest of brevity. T-statistics computed from standard errors clustered by retention method are reported underneath each marginal effect. The symbols "*," "+," and "a" represent significance at the 1%, 5%, and 10% levels, respectively.
160. See SAMPLE ET AL., supra note 10, at 3, 8-9 (describing business groups' spending on television advertising and attack ads).
161. The table reports the marginal effects of all variables on the probability of a judge voting for the business litigant, based on probit estimates. All control variables are included in the estimation but are not reported in the interest of brevity. T-statistics computed from standard errors clustered by retention method are reported underneath each marginal effect. The symbols "*," "+," and "a" represent significance at the 1%, 5%, and 10% levels, respectively.
162. The table reports the marginal effects of all variables on the probability of a judge voting for the business litigant, based on probit estimates. All control variables are included in the estimation but are not reported in the interest of brevity. T-statistics computed from standard errors clustered by state are reported underneath each marginal effect. The symbols "*," "+," and "a" represent significance at the 1%, 5%, and 10% levels, respectively.
163. The table reports the marginal effects of all variables on the probability of a judge voting for the business litigant, based on probit estimates. All control variables are included in the estimation but are not reported in the interest of brevity. T-statistics computed from standard errors clustered by retention method are reported underneath each marginal effect. The symbols "*," "+," and "a" represent significance at the 1%, 5%, and 10% levels, respectively.
164. Cf. Richard L. Hall & Frank W. Wayman, Buying Time: Moneyed Interests and the Mobilization of Bias in Congressional Committees, 84 AM. POL. SCI. REV. 797, 798 (1990) (noting that empirical work has been unable to show that campaign contributions directly influence legislators to vote differently than they otherwise would have).
165. Schotland, supra note 8, at 1105.
166. Cf. Daryl J. Levinson & Richard H. Pildes, Separation of Parties, Not Powers, 119 HARV. L. REV. 2311, 2365 (2006) (observing that separation of powers ensures that "when there is no legislative-executive consensus on policy, judicial review can be significantly countermajoritarian).
167. See id. at 2329 (Intraparty cooperation. smoothes over branch boundaries and suppresses the central dynamic assumed in the Madisonian model.).
168. See generally JOHN H. ALDRICH, WHY PARTIES? THE ORIGIN AND TRANSFORMATION OF POLITICAL PARTIES IN AMERICA (1995) (arguing that vast modern political parties are necessary to solve fundamental problems of democracy);
169. Michael S. Kang, The Hydraulics and Politics of Party Regulation, 91 IOWA L. REV. 131 (2005) (defining parties as bundles of political relationships, best understood in political, supralegal terms).
170. See, e.g., Kyle D. Cheek & Anthony Champagne, Partisan Judicial Elections: Lessons from a Bellwether State, 39 WILLAMETTE L. REV. 1357, 1362 (2003) (describing heavy party involvement in partisan judicial elections).
171. Cf. Champagne, supra note 14, at 1421 (Political parties jockeying for power in the selection of state court judges is an ancient political rite.).
172. Fed. Election Comm'n v. Colo. Republican Fed. Campaign Comm., 533 U.S. 431, 469 (2001) (Thomas, J., dissenting).
173. See generally ALDRICH, supra note 153, at 290 (explaining that parties cultivate and protect their "brand name," which connects candidates with voters);
174. GARY W. COX & MATHEW D. MCCUBBINS, LEGISLATIVE LEVIATHAN: PARTY GOVERNMENT IN THE HOUSE 2 (1993) (arguing that parties are "legislative cartel[s]" that monitor and seek to advance their own public image by monitoring and disciplining candidates).
175. See Stephen J. Choi, G. Mitu Gulati & Eric A. Posner, Professionals or Politicians: The Uncertain Empirical Case for an Elected Rather than Appointed Judiciary, 26 J.L. ECON. & ORG. 290 (2010) (finding that while appointed judges write higher quality opinions, elected judges write more opinions, focusing on service to voters).
176. See Colo. Republican Fed. Campaign Comm. v. Fed. Election Comm'n, 518 U.S. 604, 646 (1996) (Thomas, J., concurring in judgment and dissenting in part) ("The very aim of a political party is to influence its candidate's stance on issues and, if the candidate takes office or is reelected, his votes.").
177. Although the process by which judicial candidates are selected by parties to run in partisan elections differs from state to state, a lack of party leadership support can be effectively fatal to one's judicial candidacy.
178. See Ellen D. Katz, Barack Obama, Margarita Lopez Torres, and the Path to Nomination, 8 ELECTION L.J. 369, 371, 374-76 (2009) (illustrating point with case study from New York's judicial nomination process).
179. See SETH E. MASKET, NO MIDDLE GROUND: HOW INFORMAL PARTY ORGANIZATIONS CONTROL NOMINATIONS AND POLARIZE LEGISLATURES 43-53 (2009) (describing cooperation among party activists and networks of campaign donors to control party nominations);
180. Paul S. Herrnson, The Importance of Party Campaigning, 20 POLITY 714, 717 (1988) (describing, among other things, parties' "brokerage function" between candidates and campaign professionals and fundraisers);
181. Gregory Koger, Seth Masket & Hans Noel, Cooperative Party Factions in American Politics, 38 AM. POL. RES. 33, 35 (2010) (describing modern parties as expansive social networks that include campaign consultants and donors).
182. See MARTY COHEN, DAVID KAROL, HANS NOEL & JOHN ZALLER, THE PARTY DECIDES: PRESIDENTIAL NOMINATIONS BEFORE AND AFTER REFORM 3-7 (2008) (describing party-based networks of campaign donors and interest groups that regularly support party candidates).
183. See Cheek & Champagne, supra note 154, at 1379-80 (illustrating parties' significant role by finding that affiliation with party that wins in other races is strongest predictor of electoral success in partisan judicial elections);
184. L. Douglas Kiel, Carole Funk & Anthony Champagne, Two-Party Competition and Trial Court Elections in Texas, 77 JUDICATURE 290, 292 (1994) (finding similarly that fortunes of partisan judicial candidates were closely tied to their party's success at top of ticket).
185. BONNEAU & HALL, supra note 13, at 41-45 (noting less "ballot roll-off"-greater voter participation-in partisan judicial elections than in nonpartisan judicial elections);
186. Melinda Gann Hall, Voting in State Supreme Court Elections: Competition and Context as Democratic Incentives, 69 J. POL. 1147 (2007) (same).
187. See MASKET, supra note 161, at 53 ("[Parties] maintain leverage over officeholders by threatening to withhold. resources or to channel them toward a more faithful candidate.").
188. See George W. Soule, The Threats of Partisanship to Minnesota's Judicial Elections, 34 WM. MITCHELL L. REV. 701, 723 (2008) ("Many lawyers who would be good judges have little political background and are wary of running a high-profile election campaign.").
189. Daniel R. Pinello, Linking Party to Judicial Ideology in American Courts: A Meta-Analysis, 20 JUST. SYS. J. 219, 243 (1999).
190. See supra Part II.A.3.
191. The table reports the marginal effects of all variables on the probability of a judge voting for the business litigant, based on probit estimates. All control variables are included in the estimation but are not reported in the interest of brevity. T-statistics computed from standard errors clustered by retention method are reported underneath each marginal effect. The symbols "*," "+," and "a" represent significance at the 1%, 5%, and 10% levels, respectively.
192. 536 U.S. 765 (2002).
193. The table reports the marginal effects of all variables on the probability of a judge voting for the business litigant, based on probit estimates. All control variables are included in the estimation but are not reported in the interest of brevity. T-statistics computed from standard errors clustered by retention method are reported underneath each marginal effect. The symbols "*," "+," and "a" represent significance at the 1%, 5%, and 10% levels, respectively.
194. See supra notes 127-28 (collecting sources).
195. The variables are actually the interaction between each retention method and the inverse of the years to retention. Because the longest number of years to retention during our sample is 12, the inverse of the years to retention is 13 minus the years to retention. Moreover, the sample is limited to judges that do not have life tenure, so as to compare the impact of approaching retention on judges under different retention methods.
196. The table reports the marginal effects of all variables on the probability of a judge voting for the business litigant, based on probit estimates. All control variables are included in the estimation but are not reported in the interest of brevity. T-statistics computed from standard errors clustered by retention method are reported underneath each marginal effect. The symbols "*," "+," and "a" represent significance at the 1%, 5%, and 10% levels, respectively.
197. See Shepherd, supra note 17.
198. According to the data in our sample, the average contribution to Republican judges in partisan elections is $90,000, and the average contribution to Democratic judges in partisan elections is $40,000.
199. The table reports the marginal effects of all variables on the probability of a judge voting for the business litigant, based on probit estimates. All control variables are included in the estimation but are not reported in the interest of brevity. T-statistics computed from standard errors clustered by retention method are reported underneath each marginal effect. The symbols "*," "+," and "a" represent significance at the 1%, 5%, and 10% levels, respectively.
200. The total effect of the partisan reelection indicator on pro-business voting is the coefficient on the partisan reelection indicator plus the coefficient on the interaction variable multiplied by the mean percentage of Republican judges.
201. See Matthew J. Streb, Partisan Involvement in Partisan and Nonpartisan Trial Court Elections, in RUNNING FOR JUDGE, supra note 63, at 96, 104-10 (describing support provided by political parties to judicial campaigns).
202. Caperton v. A.T. Massey Coal Co., 129 S. Ct. 2252, 2267 (2009) (quoting Republican Party of Minn. v. White, 536 U.S. 765, 794 (2002)).
203. BRENNAN CTR. FOR JUSTICE, RECUSAL REFORM IN THE STATES: 2009 TRENDS AND INITIATIVES (July 28, 2009), available at http://www.brennancenter.org/page/-/2009DISQUALIFICATIONREFORMINSTATES-3.doc.
204. James Sample & Michael Young, Invigorating Judicial Disqualification: Ten Potential Reforms, 92 JUDICATURE 26, 31 (2008).
205. BRENNAN CTR. FOR JUSTICE, supra note 181.
206. Skeptics challenge the virtues of nonpartisan judicial elections on various grounds.
207. See, e.g., BONNEAU & HALL, supra note 13, at 135-37 (arguing that nonpartisan elections do not produce better quality judges than partisan elections and that partisan elections may actually produce better judges);
208. Averill, supra note 16, at 322 (listing undesirable features of nonpartisan elections, including likelihood of corruption, ability to be politically influenced, and low and uninformed voter turnout);
209. Brandice Canes-Wrone & Tom S. Clark, Judicial Independence and Nonpartisan Elections, 2009 WIS. L. REV. 21, 64 (concluding that judges are more responsive to public opinion in nonpartisan elections than in partisan elections because of absence of identifying party label);
210. Melinda Gann Hall, Voluntary Retirements from State Supreme Courts: Assessing Democratic Pressures To Relinquish the Bench, 63 J. POL. 1112 (2001) (finding that electoral considerations do not influence voluntary retirements in nonpartisan elections, indicating failure of democratic accountability mechanism);
211. Donald W. Jackson & James W. Riddlesperger, Jr., Money and Politics in Judicial Elections: The 1988 Election of the Chief Justice of the Texas Supreme Court, 74 JUDICATURE 184, 189 (1991) (concluding that nonpartisan elections offer no advantages over partisan elections and remove useful cues for voters). For skepticism about the very problem of judicial bias,
212. see Eric A. Posner, Does Political Bias in the Judiciary Matter? Implications of Judicial Bias Studies for Legal Reform, 75 U. CHI. L. REV. 853, 853-58 (2008), which explains that judicial bias is not harmful in a broad range of circumstances and that reforms may in fact impose costs on other values. 187 Arkansas, Georgia, Kentucky, Mississippi, and North Carolina have adopted nonpartisan elections for at least some judicial positions. BONNEAU & HALL, supra note 13, at 3.
213. For instance, Pennsylvania has long considered proposals to adopt nonpartisan elections or merit selection in place of partisan elections for judicial positions.
214. See Webster, supra note 57, at 22-24 (detailing longstanding consideration of issue in Pennsylvania and Texas).
215. See Schotland, supra note 8, at 1082-83 (describing judicial reforms under consideration, including returning to partisan elections, in variety of states).
216. See supra note 39 and accompanying text. Trial court judges are elected through partisan elections in 22% of jurisdictions, and trial court judges face partisan elections for retention in 14% of jurisdictions.
217. ROTTMAN ET AL., supra note 40, at 34 tbl.7. Twenty-one states used partisan elections for at least some judgeships compared to twenty states that use nonpartisan elections for at least some judgeships. Id.
218. 191 See, e.g., Roy A. Schotland, To the Endangered Species List, Add: Nonpartisan Judicial Elections, 39 WILLAMETTE L. REV. 1397, 1416 (2003) (explaining that state provisions barring candidate endorsement by political parties are surely unconstitutional and citing cases);
219. cf. Republican Party of Minn. v. White, 536 U.S. 765 (2002) (recognizing that First Amendment permits judicial candidates to announce views on how cases should be decided).
220. See generally Kathleen M. Sullivan, Republican Party of Minnesota v. White: What Are the Alternatives?, 21 GEO. J. LEGAL ETHICS 1327, 1328-33 (2008) (discussing White).
221. Averill, supra note 16, at 322.
222. Cheek & Champagne, supra note 154, at 1362
223. (quoting Richard Woodbury, Is Texas Justice for Sale?, TIME, Jan. 11, 1988, at 74).
224. Jonathan Ringel, Nonpartisan Voting Put in Jeopardy, Panel Reports, FULTON COUNTY DAILY REP. (Atlanta), Dec. 20, 2004, at 1.
225. Jonathan Ringel, Republicans Recruiting for Sears Opponent, FULTON COUNTY DAILY REP. (Atlanta), Apr. 21, 2004, at 1.
226. Camille M. Tribble, Comment, Awakening a Slumbering Giant: Georgia's Judicial Selection System After White and Weaver, 56 MERCER L. REV. 1035, 1068-69 (2005).
227. Greg Bluestein, State Bar May Back Campaign Reform Bill, FULTON COUNTY DAILY REP. (Atlanta), Dec. 13, 2004, at 1 (quoting Judge Benjamin W. Studdard III).
228. See Schotland, supra note 8, at 1082 (noting Georgia's proposed change).
229. BONNEAU & HALL, supra note 13, at 132.
230. See HARRY P. STUMPF & JOHN H. CULVER, THE POLITICS OF STATE COURTS 46 (1992) (In partisan races, the political party label may give most voters all the information they seek.).
231. BONNEAU & HALL, supra note 13, at 132.
232. See id. (Through the lens of campaign spending, nonpartisan elections are not the best choice among elections for minimizing the politics of big money in state supreme court contests.).
233. See National Overview, NAT'L INST. ON MONEY ST. POL., http://www. followthemoney.org/database/nationalview.phtml?l=0&f=J&y=2010&abbr=0 (last visited Feb. 17, 2010) (showing total campaign expenditures for high court candidates across states); supra Table 2 (identifying states with partisan elections as highest spending). For example, in 2010, spending was greatest in Texas, Ohio, Alabama, and Arkansas, all four of which have either partisan elections or elections in which parties choose the nominees.
234. Bonneau, supra note 5, at 63-64 (In every election from 1990 to 2004, partisan elections were more expensive, on average, than nonpartisan elections.).
235. BONNEAU & HALL, supra note 13, at 131-32.
236. See supra Table 7.
237. See Streb, supra note 179, at 102.
238. Id. at 109 tbl.6.5.
239. Id. at 109 tbl.6.6.
240. Id. at 105 tbl.6.2.
241. These figures represent the average contribution to candidates across all judicial races, including many noncompetitive races. When we exclude candidates who received no party contributions at all, the average combined contribution of the major parties per candidate was $120,386 in partisan states, compared to $46,337 in nonpartisan states.
242. See CHARLES H. SHELDON & LINDA S. MAULE, CHOOSING JUSTICE: THE RECRUITMENT OF STATE AND FEDERAL JUDGES 69-85 (1997) (describing party recruitment of judicial candidates for partisan elections).
243. We do not take a position here, based on our findings, about the advisability of public financing of judicial elections, whether they are partisan or nonpartisan, but we also do not think our findings cut against public financing at all. Of course, public financing of campaigns limits the need for private financing and therefore has the potential, depending on the design, to limit the influence of campaign contributions as well. However, there are many reasons to choose between nonpartisan and partisan elections on other grounds. Our study concludes only that the uncomfortably tight relationship between campaign money and judicial decisions applies mainly to partisan elections and that nonpartisan elections present fewer concerns on this score.
244. Charles Gardner Geyh, The Endless Judicial Selection Debate and Why It Matters for Judicial Independence, 21 GEO. J. LEGAL ETHICS 1259, 1272 (2008).
245. Caufield, supra note 37, at 627-28.
246. See generally Laura Denvir Stith & Jeremy Root, The Missouri Nonpartisan Court Plan: The Least Political Method of Selecting High Quality Judges, 74 MO. L. REV. 711 (2009) (describing and advocating for merit selection).
247. Schotland, supra note 8, at 1082.
248. See Hall & Aspin, supra note 124, at 343 (noting that, between 1964 and 1984, judges facing retention elections were retained over 98% of the time).
249. See Nevada Voters Reject Judicial Ballot Measures, LAS VEGAS SUN, Nov. 3, 2010, available at http://www.lasvegassun.com/news/2010/nov/03/nv-nevada-measures-2nd-ldwritethru/(noting that supporters promoted proposal as way to reduce role of campaign contributions in judicial elections).
250. See Henry R. Glick & Craig F. Emmert, Selection Systems and Judicial Characteristics: The Recruitment of State Supreme Court Judges, 70 JUDICATURE 229, 235 (1987) (finding no evidence that objective credentials of judges selected by merit plan differ significantly from those of elected judges).
251. The table reports the marginal effects of all variables on the probability of a judge voting for the business litigant, based on probit estimates. Standard errors clustered by state are reported underneath each marginal effect. The issue indicators and industry indicators are not reported in the interest of brevity. The symbols "*," "+," and "a" represent significance at the 1%, 5%, and 10% levels, respectively.