Case Visibility and the Electoral Connection in State Supreme Courts

Transcription

1 Case Visibility and the Electoral Connection in State Supreme Courts Damon M. Cann Department of Political Science Utah State University Teena Wilhelm Department of Political Science University of Georgia August 7, 2009 Electronic copy available at:

2 Abstract: Critics traditionally portray state supreme court elections as low-information events that fail to accomplish the stated goal of engendering accountability to the public. Recent changes in the intensity of contestable judicial elections have led scholars to consider the effect of public opinion on state court decision making. We delineate necessary conditions for judicial responsiveness to public opinion, integrating research on state court decision making with the broader literature on representation. We then empirically test our framework for judicial responsiveness. Our findings suggest that the strength of the electoral connection between state supreme court justices and their constituents is quite dependent on method of judicial retention and the visibility of the case. Electronic copy available at:

3 1 Introduction The notion that voters have the ability to hold elected officials accountable for their actions is central tenet of democratic theory. This electoral connection between citizens and elected officials is intended to generate more responsive policy and subsequently offer greater legitimacy for political institutions (Mayhew 1974). It has been the subject of intense study in the legislative context (see, e.g., Erikson 1978; Erikson and Wright 2000; Miller and Stokes 1963; Stimson, MacKuen, and Erikson 1995). A nearly exclusive focus on the electoral connection in Congress seems natural given the founders design of the legislative branch as the most democratic branch of government. However, with pronounced increases in the visibility of judicial elections, scholars have begun to consider the judicial branch as a unique and important venue in which to explore the electoral connection. While the federal judicial system is staffed exclusively by appointment, efforts to balance accountability and judicial independence at the state level continue to fuel debates about whether state judges should be elected (to maximize accountability), appointed (to maximize independence), or selected in a hybrid process where judges are selected by appointment and later required to face a retention election. Traditionally, judicial elections have lacked competition, substantive discourse, and high levels of voter participation. However, in the last 40 years, this has changed. While the politicization of judicial elections opened a new front of attacks on judicial elections for their political nature, proponents of judicial elections argue that new-style judicial elections have increased voter participation (Hall 2007; Hall and Bonneau 2008), and that they may better inform voters. As a result, proponents of judicial elections argue that elected judges are more responsive to the policy opinions of their constituents than those who are appointed. While it may not settle the debate regarding judicial selection, these claims do beg the empirical question of whether judicial elections 1

4 generate a stronger electoral connection between elected judges and their constituents. Recent scholarship provides some support for this notion. Hall (2001) finds a strong association between the murder rate in a state and incumbents electoral fortunes, showing that votes in judicial elections can turn on key criminal issues. Brace and Boyea (2008, 2007) show that there is a strong association between votes to overturn death sentences and public support for the death penalty in states that elect judges. They find no such association in states that appoint judges. As the evidence provided by these scholars is convincing, we seek to situate these results in the broader literature on representation. Particularly, scholars of representation consistently find that the strength of the electoral connection varies greatly across context, particularly the salience of the issues involved (consider Iyengar 1990; Krosnick 1990; Burstein 2002, 2003). We suspect that there is similar variation in the degree of judicial responsiveness depending on the nature of the case being decided. In recent years, state supreme courts have become arbiters of some of the most important (and controversial) policy issues of the day, including same-sex marriage, education policy, torts, and the extension of civil rights beyond those recognized by the federal government. While many of these significant decisions receive a great deal of public attention, a surprising amount of the workload of state courts including cases with significant policy implications operates out of the public s view. While the current literature on judicial accountability provides solid evidence of an electoral connection on key criminal issues, the degree of accountability may vary across cases depending on the salience of the case, or the visibility of the issue being decided. With this in mind, we propose to re-examine the question of citizen influence on state supreme court decision making. We first revisit the general literature on representation and develop a theoretical framework for the conditions under which citizen preferences may influence judicial action. We then test this framework on data drawn from the State Supreme Court Data Archives (SSCDA) and discuss the find- 2

5 ings. Finally, we conclude with a discussion of our results and their possible implications for the debate over whether judges should be selected in competitive elections. 2 The Electoral Connection and State Supreme Courts The basic logic of the electoral connection is simple: politicians will be motivated by the revealed preferences of their constituencies toward some form of policy responsiveness. The prevailing, although not isolated, theory to explain this motivation comes from a rational choice perspective: single-minded seekers of reelection will pursue electoral goals and make policy choices accordingly (Fenno 1973; Mayhew 1974). The linkage of constituency opinion and representative behavior has been primarily scrutinized in the Congressional literature, as Miller and Stokes (1963) classic study became the first of many attempts to determine the extent to which representative behavior measurably reflects the opinions of the electorate (see also Kuklinski and Elling 1977; Erikson 1978; Page and Shapiro 1983; Stimson, MacKuen and Erikson 1995; Hurley and Hill 2003, among many others). Of course, given that any electorally-minded politician should theoretically share such motivation, scholars have extended the discussion to include assessments of all political leaders, including the executive branch (e.g. Kingdon 1984; Erikson, MacKuen and Stimson 2002; Druckman and Jacobs 2006). More recently, scholars have considered the possibility of an electoral connection between the public and the judicial branch (Brace and Boyea 2008; Hall 1987; Hall and Brace 1992; Traut and Emmert 1998). While significant research has investigated the link between the U.S. Supreme Court and public opinion (Mishler and Sheehan 1993; Norpoth and Segal 1994; and McGuire and Stimson 2004), the electoral insulation and appointment process specific to the federal courts makes this line of research a somewhat altered question. Recently however, 3

6 the ongoing practice of electing justices in some American states, coupled with increasingly vigorous elections, has presented a unique opportunity to test whether an electoral connection exists in the judicial branch. Since state supreme court justices, in particular, may be subjected to a similar electoral reality as state legislators and state executives, it stands to reason that these justices may also have motivation to behave with single-minded, reelectionseeking strategies. As there is considerable variation in the states with regard to how justices are selected, the comparative advantage of examining behaviors of justices who are elected versus those who are not provides additional motivation to study these courts. Of course, notwithstanding many similarities in the conditions under which state supreme court justices and other elected representatives function, there are obvious differences. While the average person may have some fair amount of knowledge regarding the legislative or executive representation in his/her state, knowledge about courts and judges (both specifically and in general) is far less. In a 2001 national survey of registered voters, only 22% indicated that they knew a great deal about the work of courts and judges in their state. Moreover, almost half (49%) of those surveyed indicated they had very little or no information at all about the candidates in the last judicial election within their state. 1 Statistics such as this suggest that people do not know a great deal about the personnel, or the operation of the judicial branch, within their state. What people do know tends to be associated with the criminal justice aspect of the courts (Sheldon and Lovrich 1983). For these reasons, when scholars have chosen to examine electoral pressures in state supreme courts, they have often chosen highly visible issues, such as death penalty decisions, for analysis (see Hall 1987 and Hall and Brace 1992 for competitive electoral systems; Traut and Emmert 1998 for retention electoral systems; and Brace and Boyea 2008 for an integrated approach). 1 The survey of 1000 registered voters was conducted by Greenberg Quinlan Rosner Research for the Justice at Stake Campaign. Summary statistics for the survey are available online at 4

7 In addition to the differences in terms of citizen knowledge, there is also a marked difference between citizen participation in judicial elections versus other elections. In recent elections, ballot roll-off averaged about 20 percent for judicial candidates (though there is considerable variation, see Hall and Bonneau 2008). This could reflect the differences in how the public perceives the role of the judicial candidate versus the legislative or executive candidate. Citizens tend to voice their concerns to other elected officials, not their elected judges. Whereas an active citizen may feel inclined to contact his/her elected representative and voice an opinion about how the official should vote or behave, citizens do not correspond with elected judges to ask for support for certain outcomes. Instead, U.S. citizens are socialized to direct policy considerations toward the legislative or executive branch. In this admittedly low-knowledge environment where citizen understanding is limited to criminal issues and where limited citizen engagement exists, we have reason to wonder whether the conditions that make for accountability exist generally within state judiciaries. Still, the theoretical motivation for strategic, electorally-minded behavior in justices does exist as it does with elected politicians in the other branches of government. We hypothesize that on salient issues, circumstances are such that judges respond to public sentiment, which accounts for the compelling evidence that public opinion constrains judicial decision making on high-profile murder cases. However, we suspect that the influence of public opinion may not be uniform across these types of high-information cases and less prominent ones. In the section that follows, we explicate a framework for the conditions under which strong correspondence may exist between public opinion and judicial decision making. 5

8 3 A Conditional Theory of Judicial Responsiveness Based on a synthesis of the literature on representation generally and state judicial elections specifically, we propose four necessary conditions which must be met to observe a strong correspondence between judicial action and citizen opinion. While some of the conditions will either be universally met for all issues or not, other conditions may be better met in some instances than in others, leaving open the possibility that judicial responsiveness will vary across cases. The first condition is that judges must be interested in re-election. The simple fact that elected officials generally seek to retain their office implies that this condition will be met. Additionally, the claim is buttressed by the literature on judicial elections, which suggests that incumbents often seek reelection and that incumbents enjoy a reasonable degree of success (Hall 2001; Hall and Bonneau 2006). The second condition is that citizens must have stable, defined policy preferences for the choices that elected judges should make. Scholars have demonstrated that this is almost certainly the case for some issues, but that there is equal certainty this condition is not met on other issues. Zaller (1992) shows that on salient issues, citizens are more likely to have well-informed, stable preferences (see also Iyengar 1990 and Krosnick 1990). In the context of state supreme court elections, these issues are often (but not always) prominent criminal cases, due in part to the visibility of these cases. For example, Hall (1987) found that justices worry about citizen opinion on death penalty cases in particular. However, on a wide range of important issues decided by courts, citizens have little understanding of the actions of their courts. Kritzer (2001) shows that the public has a limited understanding of damages awarded in torts cases. Previous work on presidents responsiveness to public opinion suggests that politicians are most responsive on issues where 6

9 they expect presidential action (Burstein 2002, 2003; Canes-Wrone and Schotts 2004). We expect the same to be true in the judicial context. On a handful of highly visible issues before a state high court, citizens will likely have clear and stable preferences, but for the vast majority of a state high court s docket, most citizens will be unaware of the proceedings and the outcomes. Thus, it seems wise to expect this condition to be best met on visible cases, but not on low visibility cases. For strong electoral linkages to exist, an assitional third condition must be met. Not only should citizens have well-defined policy preferences, but judges must also be aware of those preferences, and perceive that abiding by those preferences is critical to obtaining reelection. In the absence of opinion data, judges would be unaware of the policy decisions the public wants them to pursue. While polling offers some opportunity for judicial candidates to learn about their constituency, there are limits on the extent to which we should believe this condition is met. First, the scope of polling in judicial elections is limited. One study of judicial candidate spending shows that only one-tenth of a percent of judicial campaign spending is devoted to candidate polling efforts (Goldberg, Holman, and Sanchez 2002). Abbe and Herrnson (2002) show that less than 15% of judicial campaigns are issue-oriented, so of the polling efforts that do take place, a good portion is likely directed at gauging the closeness of the race rather than assessing the public s opinions on issues. While judges may not have a great deal of specific public opinion data available to them, the public opinion literature suggests a possible alternative that allows for some measure of responsiveness. Druckman and Jacobs (2006) argue while politicians sometimes behave as splitters, examining specific policy opinions and acting in accordance with specific revealed preferences (see also Eisinger 2003; Page and Shapiro 1983; Geer 1991, 1996), they often behave instead as lumpers, using a general policy mood to assess public opinion rather than some specific measure of opinion (consider Kingdon 1984; Erikson MacKuen 7

10 and Stimson 2002; Stimson, MacKuen, and Erikson 1995). Druckman and Jabobs contend that politicians act as splitters on a handful of critical issues (particularly those issues on which reliable opinion data are available to them), but find that on most issues politicians instead rely on the more readily available sense of general policy opinion or mood. Because elected judges have relatively little specific indications of public opinion available to them, if judges are responsive to public opinion, we expect them to primarily rely on a general sense of public opinion rather than issue-specific polling data. Our third condition maintains that judges must not only have some sense of general opinion, but they must also perceive that following that opinion is critical to obtaining reelection. Since citizens typically only have well-formed opinions on the most salient issues, judges will sense a longer electoral leash on cases that are not highly salient. This notion is supported by a broader literature showing that elected officials are more responsive on highly salient issues (e.g., Burstein 2002, 2003; Kuklinski and Elling 1977; Kuklinski and McCrone 1980; Manza and Cook 2002; Page and Shapiro 1983; Wlezien 2004). This condition thus reinforces our expectation that judicial responsiveness will be most pronounced on highsalience issues. As a final condition for a strong electoral connection, citizens must be aware of what incumbent justices have done while in office (or what non-incumbents indicate they intend to do if elected) to enable them to hold judges accountable at the ballot box. For judges who never face election, but are appointed by their governor or legislature, there is little chance of ties being forged to the electorate. Even judges who face retention elections have little reason to believe that citizens will retaliate at the ballot box; only a handful of judges have lost their retention elections. It is in contestable elections that this final condition may most fully be met. For some time, contestable judicial races were characterized as personal, quiet, and 8

11 inexpensive, but they have now become noisier, nastier, and costlier (Schotland 1985). While some have charged that the combative tone of these elections is bad for the legitimacy of state judiciaries (Benesh 2006; Cann and Yates 2008; Gibson 2008), others have argued that vigorous competition has positive democratic benefits like increased turnout (Hall and Bonneau 2006) and the promotion of electoral accountability (Hall 2001). Notwithstanding the politicization of judicial elections, several institutional constraints make this aspect of the electoral connection somewhat weaker in the judicial realm. First, judges face restraints on the types of campaigning that they may engage. While the Supreme Court s 2002 ruling in Republican Party of Minnesota v. White (536 U.S. 765) struck down a campaign speech restriction that had barred judicial candidates from announcing their views on disputed legal issues, a number of other restrictions remain in force. In example, nearly every state with judicial elections has adopted a provision from the 1972 ABA Judicial Code that prohibits candidates from making pledges or promises of conduct other than faithful and impartial performance of the duties of the office. Additionally, most states prohibit judges from intentionally misrepresenting the qualifications or positions of their opponents in judicial election campaigns (Hasen 2007). Whether because of the existence of these restrictions or out of other considerations, the average judicial election tends to have less issue content than the average congressional election. In a survey of candidates for state judicial office, Abbe and Herrnson (2002) find that nearly 90% of judicial candidates focus their campaigns on their personal qualifications, with only a handful of candidates campaigning on their own issue positions or their opponents positions. In a review of judicial television ads run in the 2006 election cycle, Sample, Jones, and Weiss (2007) found that the modal state supreme court campaign television ad had a traditional theme that emphasized qualifications and experience rather than issues. They also found that issue-oriented advertising itself largely centered on criminal or moral issues. 9

12 This is not to say judicial elections have not become more rancorous, but simply that many judicial elections have a more mild tenor than the average congressional or presidential election. Among existing studies of the electoral connection in judicial elections, Brace and Boyea (2008) and Caldarone, Canes-Wrone, and Clark (2009) stand out as being among the most systematic. Their evidence is strong and their results convincing that public opinion influences judicial decisionmaking on death penalty cases (in the case of Brace and Boyea) and abortion cases (in Caldarone, Canes-Wrone, and Clark) in elective state supreme courts (but is irrelevant in states using appointment plans). Notwithstanding the many virtues of both studies, they each consider only a single visible issue under what Brace and Boyea describe as very special conditions. The specific settings involved in these studies meets our theoretical conditions for generating judicial responsiveness. In the instance of death penalty cases, electorally motivated judges face citizens who have well-defined policy preferences on a visible issue. Because judges know death penalty cases matter to their constituents (Hall 1987), judges are almost certainly aware of specific policy considerations of their constituents and make use of that information in making their decisions both in office (for incumbents) and in the campaign. Finally, because criminal issues are of tremendous importance to voters, citizens pay attention to murder-related issues and use this information in the voting decision for state high court elections (Hall 2001). While the extant literature on judicial responsiveness on visible issues is laudable, the conditions of influence outlined above suggest that the same level of responsiveness will not be observed across all cases. Specifically, for less visible (less salient) policy areas, citizens may lack stable, defined policy preferences for what elected judges should do. This is partially because citizens are not necessarily well-informed about the specifics of court activity (Kritzer 2001), but also because citizens are not socialized to direct policy expectations to- 10

13 ward the courts. While judges may attempt to use general (as opposed to specific) opinion in these cases, if citizens are unaware of court actions, judges have less incentive to behave accountably. In the absence of an informed electorate, judges sense less need for responsiveness. Finally, because citizens are prone to direct general policy considerations to legislators instead of judges, we have less reason to believe that citizens will use non-criminal issue information (especially on non-salient cases) in formulating their voting decisions. We propose to re-evaluate the judicial electoral connection in a manner that accounts for the conditional nature of judicial responsiveness outlined above. Drawing from a broad range of criminal and civil cases, we expect to confirm that in some circumstances elected judges are rather responsive to the preferences of their constituents, while in other instances they are not. 4 Testing the Conditional Theory of Judicial Responsiveness 4.1 Data and Hypotheses To examine the relationship between citizen preferences and judicial decisionmaking, we use data from Brace and Hall s Justice-Level State Supreme Court Data Archives (Brace and Hall, n.d.). To this extensive database of cases decided in the courts of last resort in all 50 states, we added data including public opinion, judge ideology, method of judicial selection/retention, and case visibility. The dependent variable in the analysis is whether a judge reached a liberal decision (coded as 1) or a conservative decision (coded as 0). 2 Because the dependent variable is a 2 We follow convention in coding criminal cases a conservative if the judge votes in favor of the state and 11

14 binary choice, we analyze the data with a probit model, with standard errors clustered by case. We subdivide cases on the basis of whether the case was decided by judges who face contestable elections for retention and those who do not. 3 We run our probit model on each set of cases to evaluate the conditional effect of public opinion and judge ideology on judicial decision making across these two contexts. Our primary hypothesis deals with the conditional influence of public opinion. We expect that public opinion should only affect the decisions of judges who face contestable elections. Because citizens are most concerned about criminal issues, we expect this effect to be most pronounced in criminal cases. Our theoretical framework gives little reason to expect strong linkages between public opinion and judicial decision making among judges whose retention is not dependent on winning a contestable election (but see Baum 2006). Because we expect judges to generally act as public opinion lumpers, we measure general liberal-conservative citizen preferences using the Berry et al (1998) Citizen Ideology scores. The scale hypothetically ranges from 0 to 100 with higher scores indicating more liberal citizen preferences. Our framework also suggests that elected judges will be most sensitive to public opinion on cases that are highly visible. To assess this, we include Vining and Wilhelm s (2009) dichotomous Case Visibility measure (coded 1 = visible and 0 otherwise). This measure follows the standard set by Epstein and Segal (2000) for United States Supreme Court (USSC) cases. In their analysis, Epstein and Segal examined the front page of the New York Times (over nearly 50 years) to determine which cases were reported on the day immediately following the decision. From this, they created a dichotomous measure of case visibility, often liberal otherwise. For civil cases, we follow convention in coding decisions for the original plaintiff as liberal and decisions for the original defendant as conservative. 3 This practice is consistent with the literature that suggests merit selection states (those that use noncontestable retention elections) generally have more in common with appointed states than states that hold competitive elections (e.g. Bonneau 2005, Cann 2007, Hall and Bonneau 2008). 12

15 equated with case salience. Vining and Wilhelm (2009) replicated this measure for the state supreme courts by examining the front pages of the largest circulated newspapers in all 50 states over a four-year period (corresponding with the years of the SSCDA, ). The measure of case visibility is based on whether court decisions were given front-page coverage in these newspapers on the day following the decision. Thus, the Vining and Wilhelm measure indicates whether cases received intense scrutiny in the form of immediate, prioritized media attention. Ultimately, these cases (of all cases on the docket) should be the ones that receive the most public scrutiny, and are thus arguably the most visible for the electorate. 4 We interact this measure with citizen ideology scores. This allows us to assess the effect of ideology conditional on case visibility. We expect the strongest correspondence between public opinion and judicial decision making among elected judges deciding highly visible cases. We expect little or no correlation for elected judges deciding less visible cases and no correlation for judges who do not face contested retention campaigns (whether the case is highly visible or not). Mounting evidence shows that state high court judges also consider their own policy preferences when making decisions (see for example Brace and Hall 1995; Brace and Boyea 2008, Langer 2002). Because our framework predicts elected judges will emphasize the preferences of their constituency in high-visibility cases, we expect judge ideology to be less influential in these cases. However, on low-visibility cases we expect elected judges ideological values to matter. For judges who do not face contestable elections, we expect to see consistent effects of judge ideology. We measure judges preferences using party-adjusted ideology (PAJID) scores (Brace, Langer and Hall 2000). To capture the effect of ideology conditional on case visibility, we also include an interaction between PAJID scores and case 4 For a complete discussion of this measure, see Vining and Wilhelm 2009). 13

16 visibility. 4.2 Results Results of model estimation appear in Table 1. The results are strongly supportive of our hypotheses regarding the conditional nature of judicial responsiveness. The coefficients in the model for judges who face contestable retention elections imply that citizen ideology does not effect decision making in low visibility cases, but that on highly visible cases there is a strong relationship. In contrast, the model for judges who do not face a contestable election to retain their seat appear to pay little attention to public opinion. The coefficient on citizen ideology (indicating the effect of citizen ideology in less visible cases) is actually negative, though small enough to be substantively inconsequential. 5 Even in highly visible cases heard in states using merit selection or appointment methods of judicial retention we do not see a significant effect. [Table 1 about here] Figures 1 and 2 offer a more easily interpretable presentation of the conditional effects of public opinion and judge ideology in terms of changes in predicted probabilities. We compute the discrete change in the predicted probability of a liberal decision caused by changing citizen ideology from 5 points below the mean citizen ideology (45.37) to 5 points above the mean, holding judge ideology at its mean. The effect of this mean-centered 10- point change in citizen ideology is conditional upon case visibility. In Figure 1, we see the conditional effects of a mean-centered 10-point change in citizen ideology on judges who must face contestable elections to retain position. On less visible cases, the 10-point change in 5 One interpretation of the negative coefficient could be that citizen interests in these states are addressed in the legislatures, leaving courts to fill a counter-majoritarian role where they actually make decisions contrary to public opinion. 14

17 citizen ideology has a slightly negative but statistically insignificant effect on the probability of making a liberal decision. In contrast, on visible cases, the 10-point change in citizen ideology raises the predicted probability of a liberal decision by.073. Figure 2 shows the effect of the same mean-centered 10-point change in citizen ideology on judges who do not face a contestable election to retain their seat (e.g. states that use either merit selection plans or appointment to retain judges). In less visible cases, we see a slight negative effect of citizen ideology on the probability of a liberal decision (though it is so small as to be substantively unimportant). Even in visible cases, though, we find no statistically significant effect of public opinion as the 95% confidence interval around the discrete change in the predicted probability contains zero. [Figures 1 and 2 about here] 5 Discussion Our results generally support the theoretical framework for the conditions of judicial responsiveness. Where circumstances are such that state supreme court judges seek reelection, citizens have enough awareness of a case that they have stable preferences, and judges know that citizens have such preferences and could use them at the ballot box, we tend to see significant judicial responsiveness to public opinion. These conditions are well-met for highly visible cases heard by judges who face contestable elections. However, for a much larger proportion of cases that are either less visible or are heard by judges who do not face contestable elections, there little reason to expect judicial responsiveness. The development of this framework, and the empirical results supporting it, help to explain the disparate literatures that suggest that voters have relatively little knowledge of what state courts do (Kritzer 2001, Sheldon and Lovrich 1983) but that elected judges are quite accountable on 15

18 at least some key issues (Brace and Boyea 2008; Caldarone, Canes-Wrone, and Clark 2009; Hall 1987, 2001). These results are consistent with the broader literature on representation that suggests the elected officials responsiveness varies on the basis of citizens awareness and expectations (Burstein 2002, 2003; Canes-Wrone and Shotts; Kuklinski and Elling 1977; Kuklinski and McCrone 1980; Manza and Cook 2002; Page and Shapiro 1983; Wlezien 2004). Even in light of these results, evaluating the utility of judicial elections remains a difficult task. Reformers have historically criticized judicial elections because they fail to promote accountability; others have criticized them for their intense political nature. Judicial election proponents have responded to these charges by showing that vigorous elections do promote accountability, at least on some key issues. Our results suggest that the degree to which judicial elections promote accountability is not only conditional on the method of selection and retention, but also on the nature of the issues involved in the case. Reformers may view these results as vindication for their perspective because our evidence shows that judicial elections fail to promote accountability across the full spectrum of issues. This characterization is not entirely accurate. While judicial elections may not promote accountability on all issues, the promotion of accountability on the most visible issues (where citizens may have clear, well-defined expectations for state supreme court judges via the saliency of the issue) may be all one needs to justify the use of electoral mechanisms. Still, from a democratic perspective, there is something unsatisfying about a situation where judges set citizen preferences aside simply because an issue will likely receive less media attention. Moreover, it is a mistake to equate visibility with importance. While many important cases are certainly visible, only a fraction of even important state supreme court cases breaks into the public s consciousness. While we cannot end the dispute over the ideal method for selecting and retaining judges, this paper makes a significant contribution to the literature by outlining and empirically 16

19 testing necessary conditions for judicial responsiveness to public opinion. Just as certainly as elective judiciaries are not void of citizen influence, we must acknowledge that these institutions are certainly not dominated by the prevailing views of the electorate on all issues. 17

20 References [1] Abbe, Owen G. and Paul S. Herrnson How Judicial Election Campaigns Have Changed. Judicature 85: [2] Baum, Lawrence Judges and Their Audiences: A Perspective on Judicial Behavior. Princeton, NJ: Princeton University Press. [3] Benesh, Sarah Understanding Public Confidence in American courts. Journal of Politics 68: [4] Berry, William, Evan Ringquist, Richard Fording and Russell Hanson Measuring Citizen and Government Ideology in the American States, American Journal of Political Science 42: [5] Brace, Paul and Brent D. Boyea State Public Opinion, the Death Penalty and the Practice of Electing Judges. American Journal of Political Science 52: [6] Brace, Paul and Brent D. Boyea Judicial Selection Methods and Capital Punishment in the American States. In Running for Judges: The Rising Political, Financial, and Legal Stakes of Judicial Elections, M. Streb, ed. New York: New York University Press. [7] Brace, Paul and Melinda Gann Hall. n.d. State Supreme Court Data Archive. Available online at pbrace/statecourt/, accessed March [8] Brace, Paul and Melinda Gann Hall Studying Courts Comparatively: The View from the American States. Political Research Quarterly 48: [9] Brace, Paul, Laura Langer, and Melinda Gann Hall Measuring the Preferences of State Supreme Court Judges. Journal of Politics 62(2):

21 [10] Burstein, Paul Public Opinion and Congressional Action on Labor Market Opportunities, In Navigating Public Opinion: Polls, Policy, and the Future of American Democracy, ed. Jeff Manza, Fay Lomax Cook, and Benjamin I. Page. New York: Oxford University Press. [11] Burstein, Paul The Impact of Public Opinion on Public Policy: A Review and an Agenda. Political Research Quarterly 56:2940. [12] Caldarone, Richard P., Brandice Canes-Wrone, and Tom S. Clark Partisan Labels and Democratic Accountability: An Analysis of State Supreme Court Abortion Decisions. Journal of Politics 71: [13] Canes-Wrone, Brandice and Kenneth W. Shotts The Conditional Nature of Presidential Responsiveness to Public Opinion. American Journal of Political Science 48(4): [14] Cann, Damon M Beyond Accountability and Independence: Judicial Selection and State Court Performance. Judicature 90: [15] Cann, Damon M. and Jeff Yates Homegrown Institutional Legitimacy: Assessing Citizens Diffuse Support for State Courts. American Politics Research 36(2): [16] Druckman, James N. and Lawrence R. Jacobs Lumpers and Splitters: The Public Opinion Information that Politicians Collect and Use. Public Opinion Quarterly 70(4): [17] Eisinger, Robert M The Evolution of Presidential Polling. New York: Cambridge University Press. 19

22 [18] Robert S. Erikson Constituency Opinion and Congressional Behavior: A Reexamination of the Miller-Stokes Representation Data. American Journal of Political Science 22(Aug.): [19] Robert S. Erikson and Gerald C. Wright Representation of Constituency Ideology in Congress. In Continuity and Change in Congressional Elections. David Brady and John Cogan (eds.). Palo Alto, CA: Stanford University Press. [20] Erikson, Robert S., Michael B. MacKuen, and James A. Stimson The Macro Polity. New York: Cambridge University Press. [21] Fenno, Richard F Congressmen in Committees Boston: Little, Brown. [22] Geer, John From Tea Leaves to Opinion Polls: Politicians, Information, and Leadership. New York: Columbia University Press. [23] Geer, John Do Open-Ended Questions Measure Salient Issues? Public Opinion Quarterly 55(3): [24] Gibson, James L Challenges to the Impartiality of State Supreme Courts: Legitimacy Theory and New-Style Judicial Campaigns. American Political Science Review 102: [25] Goldberg, Deborah, Craig Holman, and Samantha Sanchez The New Politics of Judicial Elections Justice at Stake: Washington, DC. [26] Hall, Melinda Gann Constituency Influence In State Supreme Courts: Conceptual Notes and a Case Study. Journal of Politics 49 (November): [27] Hall, Melinda Gann State Supreme Courts in American Democracy: Probing the Myths of Judicial Reform. American Political Science Review 92 (June):

23 [28] Hall, Melinda Gann Competition as Accountability in State Supreme Court Elections. Running for Judges: The Rising Political, Financial, and Legal Stakes of Judicial Elections, M. Streb, ed. New York: New York University Press. [29] Hall, Melinda Gann and Chris W. Bonneau Does Quality Matter? Challengers in State Supreme Court Elections. American Journal of Political Science 50 (January): [30] Hall, Melinda Gann and Chris W. Bonneau Mobilizing Interest: The Effects of Money on Citizen Participation in State Supreme Court Elections. American Journal of Political Science 53(3): [31] Hall, Melinda Gann and Paul Brace Toward an Integrated Model of Judicial Voting Behavior. American Politics Research 20(2): [32] Hasen, Richard L First Amendment Limits on Regulating Judicial Campaigns. In Running for Judge: The Rising Political, Financial, and Legal Stakes of Judicial Elections, M. Streb, ed. New York University Press: New York. [33] Huber, Gregory A. and Sanford C. Gordon Accountability and Coercion: Is Justice Blind when It Runs for Office? American Journal of Political Science 48 (Apr.): [34] Hill, Kim Quaile, and Patricia A. Hurley Dyadic Representation Reappraised. American Journal of Political Science 43: [35] Iyengar, Shanto Shortcuts to Political Knowledge: The Role of Selective Attention and Accessibility. In Information and Democratic Processes, ed. John A. Ferejohn, and James H. Kuklinski. Urbana: University of Illinois Press. 21

24 [36] Kingdon, John W Agendas, Alternatives, and Public Policies. Boston: Little, Brown. [37] Kritzer, Herbert M Public Perceptions of Civil Jury Verdicts. Judicature 85(2): [38] Krosnick, Jon A Government Policy and Citizen Passion: A Study of Issue Publics in Contempo- rary America. Political Behavior 12:5992. [39] Kuklinski, James H., and Richard Elling Representational Role, Constituency Opinion, and Legislative Roll Call Behavior. American Journal of Political Science 21: [40] Kuklinski, James H., and Donald J. McCrone Policy Salience and the Causal Structure of Representation. American Politics Quarterly 8: [41] Langer, Laura Strategic Considerations and Judicial Review: The Case of Workers Compensation Laws in the American States. Public Choice 116 (August): [42] Manza, Jeff, and Fay Lomax Cook A Democratic Polity? Three Views of Policy Responsiveness to Public Opinion in the United States. American Politics Research 30: [43] Mayhew, David Congress: The Electoral Connection New Haven, CT: Yale University Press. [44] McGuire, Kevin T., and James A. Stimson The Least Dangerous Branch Revisited: New Evidence on Supreme Court Responsiveness to Public Preferences. Journal of Politics 66(4):

25 [45] Miller, Warren E., and Donald E. Stokes Constituency Influence in Congress. American Political Science Review 57:4556. [46] Mishler, William, and Reginald S. Sheehan The Supreme Court as a Countermajoritarian Institution? The Impact of Public Opinion on Supreme Court Decisions. American Political Science Review 87 (1): [47] Norpoth, Helmut, and Jeffrey A. Segal Comment: Popular Inuence on Supreme Court Decisions. American Political Science Review 88 (3): [48] Page, Benjamin I., and Robert Y. Shapiro Effects of Public Opinion on Policy. American Political Science Review 77: [49] Sample, James, Lauren Jones, and Rachel Weiss The New Politics of Judicial Elections Washington, DC: Justice at Stake. [50] Schotland, Roy A Elective Judges Campaign Financing: Are State Judges Robes the Emperor s Clothes of American Democracy? Journal of Law and Politics 2 (Spring): [51] Sheldon, Charles H. and Nicholas P. Lovrich, Jr Knowledge and Judicial Voting: The Oregon and Washington Experience. Judicature 67: [52] Stimson, James A., Michael B. MacKuen, and Robert S. Erikson Dynamic Representation. American Political Science Review 89(Sept.): [53] Carol Ann Traut and Craig F. Emmert Expanding the Integrated Model of Judicial Decision Making: The California Justices and Capital Punishment. Journal of Politics 60(4):

26 [54] Vining, Richard and Teena Wilhelm Issue Salience in State Courts of Last Resort. Paper presented at the 2009 Annual Meetings of the Southern Political Science Association. [55] Wlezien, Christopher Patterns of Representation: Dynamics of Public Preferences and Policy. Journal of Politics 66:124. [56] Zaller, John The Nature and Origins of Mass Opinion. New York: Cambridge University Press. 24

27 Table 1: Probit Models of the Effects of Citizen and Justice Ideology on Judges Decisions Judges Face Contestable Judges Face No Contestable Elections for Retention Elections for Retention Variable Coef. Coef. (Std. Err.) (Std. Err.) Citizen Ideology * (.0013) (.0007) Case Visibility.019*.008 Citizen Ideology (.009) (.005) Judge Ideology * (.001) (.0004) Case Visibility Judge Ideology (.007) (.003) Case Visibility -.777* (.356) (.232) Intercept -.212* -.088* (.049) (.031) n Notes: Cell entries are probit coefficients with robust standard errors, clustered by case, in parantheses. The dependent variable is coded 1 = liberal decision, 0 = conservative decision. * denotes p <.05, two-tailed. 25

28 Discrete Change in Pr(Liberal Decision) Low High Visibility Level of Case Visibility Discrete Change in Pr(Liberal Decision) Low Visibility Level of Case Visibility High Visibility Figure 1: The Conditional Effect of Public Opinion on Judges Facing Contestable Retention Elections. Points represent the discrete change in the probability of a liberal decision given a mean-centered 10 point increase in the the liberalness of citizen ideology. The capped range bars depict 95% confidence intervals. 26

29 Discrete Change in Pr(Liberal Decision) Low High Visibility Level of Case Visibility Discrete Change in Pr(Liberal Decision) Low Visibility Level of Case Visibility High Visibility Figure 2: The Conditional Effect of Public Opinion on Judges Not Facing Contestable Retention Elections. Points represent the discrete change in the probability of a liberal decision given a mean-centered 10 point increase in the the liberalness of citizen ideology. The capped range bars depict 95% confidence intervals. 27