Persuasion in the Decision Making of U.S. Supreme Court Justices
Abstract and Keywords
This chapter proposes that persuasion is more effective in influencing votes in certain types of cases than others. A distinction is made between ideological (i.e., hot-button issue) cases and non-ideological cases. Using data from several recent Supreme Court terms, the chapter shows that latency of the decision, frequency of unanimous votes, and behavior in oral arguments are affected by type of case. In oral arguments, justices' eventual votes can be predicted from the nature and frequency of their questions, to a greater degree in ideological cases than in non-ideological cases. Thus it is concluded that both the legal model and the attitudinal model of judicial decision making are applicable, depending on the type of case.
Keywords: persuasion, ideological versus non-ideological cases, criterial referents, automatic responses, legal model of judicial decision making, attitudinal model of judicial decision making, oral arguments, office of Solicitor General, "More-questions-to-loser" rule, pressures toward uniformity, dissenting decisions
The purpose of this paper is to provide some thoughts about the role of persuasion in judicial decision making. Herein, persuasion is treated broadly as the process of influence, which covers not only overt attempts at influence (by the media, by the advocates, and by other judges) but also the judges' responses to attempts at influence. Emphasis is on the question: What influences how a judge forms an initial opinion about a case? The coverage and examples refer to the United States Supreme Court, but the ideas are applicable to other appellate panels and in some respects to trial judges.
This paper assumes that justices are more susceptible to persuasion depending on the nature of the case; some cases deal with matters to which the justices have given much prior thought. Certain cases may reflect issues on which they have developed strong views long in their past. Research on persuasion certainly indicates that the nature of the material affects the willingness to be responsive to an attempt to persuade. But individual differences may exist in justices' readiness to be persuaded on all issues; some may possess personality characteristics that inhibit change. Thus, this paper reflects two assumptions:
1. Persuasion operates differently depending on the type of case.
2. Persuasion operates differently depending on the particular justice.
Testing the First Assumption: Ideological Versus Nonideological Cases
Cases that are granted certiorari by the Supreme Court can be classified in many different ways: by their topic, by whether they reflect ordinary litigation or political litigation (Baum, 2007), by the presence or absence of the government as a party, by whether they come from the liberal 9th Circuit Court or the conservative 4th and 5th Circuits, and so on. This paper hypothesizes that persuasion, and thus the decision-making process, operates differently in ideological cases than in nonideological cases. By “ideological cases,” I mean cases whose content is related to an attitude or value held by a justice. Hot-button issues lead to ideological cases. In general, agreement exists on just where positions on salient ideological issues can be placed along a liberal-versus-conservative political dimension. For example, “in criminal cases, liberals are relatively sympathetic toward criminal defendants and their procedural rights, while conservatives give more emphasis to the effectiveness of the criminal justice system in fighting crime” (Baum, 1990, Table 1–3, p. 13). This paper employs Lawrence Baum's distinctions between liberal and conservative positions on judicial issues as a convenient summary.
Nonideological cases are those in which the central issues do not tap into deeply felt values of the justices. In the studies to be described subsequently I asked two raters to independently classify cases as nonideological if their content did not deal with the various topics that distinguished between liberals and conservatives, using Baum's detailed distinctions. Thus, nonideological cases are less easily classified by content, but disputes between two states, patent disputes, and conflicts between two businesses usually are nonideological cases. The case of Wachovia Bank v. Schmidt (2006) dealt with the question of whether a bank with many branches was a “person” in every state where it had a branch or only in its headquarters state. The case was decided unanimously and did not seem to provoke any ideological biases. Another nonideological example is the case of Kansas v. Colorado, to which the Supreme Court granted cert in 2001. The Arkansas River begins in the Rockies, flows across Kansas, part of Oklahoma, and across Arkansas before emptying into the Mississippi. A dispute between Colorado and Kansas over diversion of river water had seemingly been settled by the passage of the Arkansas River Compact by Congress in 1949. But for a number of years Kansas had complained that Colorado had violated the compact. After the Court had so ruled in 1995, it remanded the case to a Special Master. But Kansas claimed that Colorado continued to violate the (p.59) rules. In 2001 the Court again sided with Kansas, including a judgment that Colorado should pay interest on the damages accrued. While the question of an interpretation of the Eleventh Amendment was part of the issue at hand, the Supreme Court ruled unanimously in favor of Kansas in 2001, and the basis appears to be simply what the law says. Thus the case appears to be free of ideological triggers.
How does persuasion operate differently in ideological cases than in nonideological cases? First, it is proposed that justices are aware of ideological cases earlier than they are about nonideological cases. Well before their conference to decide which cases to grant cert, even well before a petition is submitted, the typical ideological case has generated publicity. A state passes a law that critically restricts abortions, a university's admissions program to increase diversity is challenged in a lower court, the Bush administration places prisoners of war at Guantanamo Bay and denies them rights to a trial—such decisions draw attention from the media, and the justices read the newspapers and watch television news as many members of the citizenry do. In contrast, for many nonideological cases, the first awareness may come when the justice reads the recommendations from a law clerk who has processed the case as part of the cert pool.
Second, at the point of granting cert, justices know more about the issues in ideological cases than nonideological ones. Issues of search and seizure, for example, are frequently before the Court; in the October 2003 Term the Court dealt with ten cases involving claims of a violation of Fourth Amendment rights. More often, nonideological cases deal with a relatively obscure federal law, so that during the process leading up to the decision to grant cert, most justices have to do more review. For example, a case during the October 2003 Term (BedRoc Limited v. United States) dealt with the terminology in the Pittman Underground Water Act of 1919, a legislative decision probably not on the forefront of each justice's awareness prior to preparing for the cert conference.
Justices' Values and Ideological Cases
But the most important difference is what makes the case ideological—whether its issues generate a value-based predisposition. Many years ago psychologist Fred Kerlinger (1967) proposed that in conceptualizing attitudes and values, it was useful not to think of each as a bipolar continuum, but rather to focus on what he called “criterial referents.” Certain objects, topics, or issues may serve as anchors, or criterial referents, which define one's values. For each justice, different issues may serve as criterial referents; for some, abortion; for some, the death penalty; and for some, racial or gender discrimination. These serve as triggers to at least a preliminary leaning in one direction. Sometimes it is more than a leaning; it is an irrevocable response. For example, toward the end of Justice Harry Blackmun's service on the (p.60) Court, his position on any case coming before the Court that dealt with the death penalty was clear. In 1994, his dissent in Callins v. Collins states:
From this day forward, I no longer shall tinker with the machinery of death. … It is virtually self evident to me now that no combination of procedural rules or substantive regulations can ever save the death penalty from its inherent constitutional deficiencies. … The problem is that the inevitability of factual, legal, and moral error gives us a system that we know must wrongly kill some defendants, a system that fails to deliver the fair… and reliable sentences of death required by the Constitution. (p. 1145)
More recently, Justice Antonin Scalia has made it clear that on certain matters before the Court, he had made his mind up before the process of considering the issues had been completed. On one occasion—the issue was the constitutionality of the words “under God” included in the Pledge of Allegiance—he recused himself because of public statements he had made.
But in another notable case, that of Hamdan v. Rumsfeld, he did not. This case required the Court to decide whether an enemy combatant who was detained at Guantanamo Bay was protected by the articles of the 1949 Geneva Conventions as a prisoner of war. In a speech a few weeks before this case's oral arguments, Scalia ridiculed the suggestion that detainees captured “on the battlefield” should receive a trial in civil courts; that proposition, he said, was a “crazy idea.” He interrupted a subsequent question by claiming: “If he was captured by my army on a battlefield, that is where he belongs. I have a son on that battlefield and they were shooting at my son. And I am not about to give this man who was captured in a war a full jury trial. I mean it's crazy” (Isikoff, 2006, p. 6).
Justice Scalia's comments drew strong criticism and calls for recusal from several sources (Isikoff, 2006). But Justice Scalia was not deterred; he participated actively at the oral arguments and voted in the case. In fairness to Justice Scalia, he is certainly not the only justice and this was not the only case in which a justice's eventual vote was fixed in concrete before the oral arguments; the matter came to light because of his provocative style and willingness to express his opinion in public. For him, if not every justice, this was an ideological case, and his actions support the argument that in such cases, opinions are formed early and not susceptible to persuasion. Thus, even though the focus of this paper is on persuasion, it questions how much opportunity exists to persuade justices on certain cases; their minds may be made up very quickly.
Ideological Cases and Automatic Responses
Thus, for some justices on some cases, it is proposed that their reaction is emotional and their response is instantaneous. (For a sharp contrast in very well formed values between two justices, read Justice Scalia's opinion for the Court and Justice Brennan's dissent in the case of Michael H. v. Gerald (p.61) D., 1989.) Social psychologists have concluded that the evaluations of stimuli—events, persons, issues—are often automatic; that is, they are so primed by the stimulus that they are given without further processing (Bargh & Ferguson, 2000; Duckworth, Bargh, Garcia, & Chaiken, 2002). There is even some evidence that we may have two systems
for evaluating various aspects of the social world; one that operates in an automatic manner and the other that operates in a systematic and controlled manner. … In fact, studies conducted from the perspective of social neuroscience indicate that these differences exist. Certain parts of the brain, especially the amygdala, may be involved in automatic evaluative reactions, simple good-bad judgments that occur in a rapid and nonconscious manner. In contrast, portions of the prefrontal cortex (especially the medial prefrontal cortex and ventrolateral prefrontal cortex) may play a key role in more controlled executive reactions—the kinds about which we think carefully and consciously. (Baron, Byrne, & Branscomb, 2006, p. 54)
Are some values so entrenched and so relevant to case decisions that they produce automatic reactions that do not receive review? If Justice Scalia (rather than the ubiquitous freshman psychology student) were a respondent in the kind of experiment used by psychologists to demonstrate automatic responses, how would he respond if “abortion” were flashed on the screen? We cannot say, but the emphasis on automaticity of responses is provocative.
Attitudinal and Legal Models of Judicial Decision Making
The distinction between ideological and nonideological cases may help to understand the long-standing conflict between two models of judicial decision making. Both the legal model and the attitudinal model have been extensively considered and debated in the political-science literature. (Other chapters in this book describe shifts in the attitudinal model toward a more strategic orientation.) Indirect support for the attitudinal model has also been demonstrated in several extensive analyses of the voting records of federal district court and circuit court judges; for example, Rowland and Carp (1996) showed how district judges appointed by Republican presidents decided cases on civil liberties and civil rights, for example, in a different direction from similar cases decided by judges appointed by Democrats. More recently, Sunstein, Schkade, Ellman, and Sawicki (2006) presented data on how the composition of judges on federal circuit-court panels (specifically the political party of the president who appointed them) affected their votes on what we would call ideological issues.
Each of these models (especially the attitudinal model) has its adherents (see, for example, Segal & Spaeth, 1993, 2002), but the possibility that each may apply in certain types of cases is less often emphasized. Another (p.62) possibility is that in a particular case, the attitudinal model may describe the reaction of some justices, while others may struggle to maintain a basis in the legal model.
The wealth of empirical data used by adherents of the attitudinal model (consistency in votes, correlation of votes of individual justices with their ideological pronouncements in speeches, books, or articles) to support their claim may apply in ideological cases, but in those cases where no personal values are salient, the justices are more likely to examine the statutes, the precedents, and, if applicable, the Constitution in a dispassionate manner. (See, for example, the recent article by Lindquist and Klein, 2006, that revives support for the role played by legal reasoning in Supreme Court decisions.)
In summary, it is proposed that, compared to their response to nonideological cases, individual justices in ideological cases are less responsive to the persuasion expressed in merit briefs and oral arguments. Specifically, it is hypothesized that they:
1. Form initial opinions more quickly.
2. Are less likely to change their opinions.
3. Have formed opinions by the time of oral arguments, so that their questioning reflects their already-formed opinions.
4. Are less responsive to outside influences.
5. Are more predictable with regard to their final votes.
Furthermore, this paper suggests that the degree to which a case is ideological can be reflected in several qualities of the dispensation of the case. Specifically, in ideological cases, compared to nonideological cases:
1. The final vote will less often be a unanimous vote, because the current composition of the Court is divided ideologically.
2. It will take longer for the Court to announce the decision in the case, because the minority will more likely write dissents and the opinion of the Court will go through more drafts.
As part of an ongoing project (see Wrightsman, 2008), data have been collected from several recent terms of the Supreme Court that sometimes directly and sometimes indirectly respond to these hypotheses.
For four recent terms each case has been classified as ideological, borderline, or nonideological. The following variables have been determined:
1. In the oral arguments, the length of time (measured in words) that the petitioner spoke and the respondent spoke before they were interrupted by a justice.
2. The number of questions asked each advocate by the justices during the oral arguments.
3. Whether or not a representative of the Office of the Solicitor General participated in the oral arguments, and if so, on which side.
4. The length of time (measured in days) from the oral argument to the announcement of the decision.
5. The announced vote.
6. For the October 2002 Term, the predicted votes and case outcomes as determined by the Supreme Court Forecasting Project at the School of Law at Washington University in St. Louis. (Although this project calculated statistical predictions of the outcomes in some cases in the next term, the October 2003 Term, the data analysis was discontinued in the middle of this second term.)
Empirical Tests of the Predictions
Data from the Supreme Court website were used to test the above predictions. For example, it was hypothesized that in ideological cases, justices form opinions quickly, they are less likely to change their opinions, and their questioning during oral arguments reflects their already-formed opinions. How can these hypotheses be tested empirically?
Over the last three years, two findings have been published that have concluded that during oral arguments, justices as a group ask more questions to the advocates representing the eventual losing side than they do to the advocates whose side eventually wins. The first was a study done by Sarah Shullman (2004), who observed ten oral arguments during the October 2002 Term. She reported, “All nine justices seemed to ask fewer questions of the party in whose favor they would ultimately decide” (2004, p. 278). But there are numerous methodological problems in her study. With only one observer, there is no test of inter-rater reliability. The task of doing these ratings on the spur of the moment seems very challenging; a typical 30-minute presentation by one side generates anywhere from 30 to 100 questions/comments by the justices. And when my students and I set forth to replicate and extend her findings, we found it is not easy to say just what is a question. Justices interrupt each other, their questions are sometimes answered before they are finished, there are numerous brief comments (“Okay.” “I see.” “Right.”) which may or may not be counted as questions. But despite these, Shullman's study drew attention. Tony Mauro devoted an article in the American Lawyer to it, and it certainly provoked me to determine if the effect was genuine and widespread.
Just about the same time as the publication of Shullman's article, John Roberts (before he was named Chief Justice) addressed the Supreme Court Historical Society at its annual meeting. Although most of his talk dealt with the reemergence of a Supreme Court bar, he did report on an analysis he had conducted. He took the first and last cases in each of the two-week argument (p.64) sessions for the October 1980 Term and for the October 2003 Term, and determined, for each case, the number of questions directed at the petitioner and the number directed at the respondent. Then he examined which side won. He reported: “In the 28 cases I looked at, 14 from the 1980 Term and 14 from 2003, the most-questions-asked ‘rule’ predicted the winner—or, more accurately, the loser—in 24 of those 28 cases, an 86 percent prediction rate” (2005, p. 75). Then he drily added, “The secret to successful advocacy is simply to get the Court to ask your opponent more questions” (2005, p. 75). Like Shullman, Roberts did not define what a “question” was, and neither study distinguished between ideological and nonideological cases. But the methodology seemed to provide an entry into determining if any difference related to persuadability existed. Does the more-questions-to-the loser rule hold when all cases in a term are examined? Does it hold more strongly for ideological cases?
Testing the Accuracy of the “More Questions to the Loser” Rule
In our research (Wrightsman, 2008), we quantified “questions” by referring to the transcript of the case and counting any comment or question by a justice, even if it was interrupted or fragmentary. The virtue of this operational definition is that it provides reliability. We investigated other ways of counting questions, such as combining interrupted statements or questions; such procedures lacked reliability; it is not as easy to say what is and what is not one “question” as one might think. The analysis my students and I did revealed the following: For the October 2001 Term, the eventual losing side was asked more questions in 29 of 41 ideological cases, or 70.7%; this success rate contrasts strongly with those in the borderline ideological cases—6 of 14, or 42.8%—and the nonideological cases—8 of 16, or 50%. For the October 2002 Term, the same differences were observed; the eventual losing side was asked more questions in 21 of the 33 ideological cases, or 63.6%; in 9 of 16 borderline ideological cases, or 56.2%, and in only 11 of 23 nonideological cases, or 47.8% (Overall, in each of these terms, while the eventual loser did get more questions, the percentage of outcomes predicted accurately, 60% in each term, was much lower than what was found in the previous studies.)
For the October 2005 Term, the results were consistent with the October 2001 and 2002 terms, although the differences between types of cases were smaller. More questions were asked of the losing side in 28 of 43 ideological cases (65.1%), in 9 of 14 borderline ideological cases (64.3%), and 9 of 15 nonideological cases (60%). Overall, the rate was 46 out of 72 cases, or 64%.
Thus, the rule does seem to have some validity, although not as great a predictability as the early studies promised. The fact that it holds more strongly in ideological cases (albeit the differences are not huge) indicates that to some extent, the nature of the case has led to an earlier formation of opinions by justices.
(p.65) Agreement with the Office of the Solicitor General
Another measure that supports the distinction between ideological and non-ideological cases is the degree to which the eventual decision was consistent with the position advocated by the Office of the Solicitor General. As has been well documented (Baum, 1997; McGuire, 1996), the side supported by the solicitor general, as advocate or amicus curiae, more often wins. In those terms studied in this project, this was true: in the October 2002 Term, the solicitor general's side won in 43 of 61 cases, or 65%, and in the October 2005 Term, 38 of 56 cases, or 68%. But in each of these terms, in ideological cases the solicitor general's side won less often than it did overall—only 60.7% in 2002 and 61% in 2005. In the October 2001 Term, the Court's decision was consistent with the solicitor general's position in a whopping 89% of the ideological cases and 93% of the nonideological cases. Again, the differences are small, and subject to several interpretations, but one possible interpretation is that in ideological cases the justices are less susceptible to persuasion from the Office of the Solicitor General, despite the high regard with which these attorneys are held, because their minds have been made up.
It was also predicted that differences in the dispensation of cases would be based on the degree to which their content was ideological. Two measures were employed to test these predictions: percentage of cases in which the final decision was unanimous, and length of time to reach a decision.
Do Ideological Cases More Often Lead to Nonunanimous Decisions by the Court?
During the October 2001 Term, of 72 decisions, only 26, or 36%, were unanimous. As predicted, unanimous decisions were reached less often in ideological cases (32.5%) than in nonideological cases (43.7%). Borderline-ideological cases produced just about the same level of unanimity (33.3%) as ideological cases.
While the overall degree of unanimity increased slightly in the October 2002 Term, the pattern was consistent with the earlier term. Only 12 of 33 ideological cases resulted in unanimous decisions (36.4%), compared to 7 of 16 borderline-ideological cases (43.7%), and 13 of 23 nonideological cases (56.5%). Overall, 32 of 72 decisions, or 44.4%, were unanimous. For the October 2005 Term, the effect of the type of case is quite similar: unanimity in 20 of 43 ideological cases (45.6%), in 8 of 14 borderline-ideological cases (57.1%), and 10 of 15 nonideological cases, or 66.7%). Overall, in 2005–2006, 38 of 72 decisions were unanimous, or 52.8%. (The latter term, Chief Justice Roberts' first, led to a higher rate of unanimous decisions than in any recent term, perhaps reflecting his aspirations for greater consensus in decisions.)
(p.66) How Long Does the Court Take to Reach a Decision?
It was also predicted that the justices would take longer to announce decisions in ideological cases. Decision latency was measured by the number of days between the oral argument and the announcement of the decision. For the October 2002 Term, the average latency for ideological cases was 8l days, for borderline ideological 75 days, and for nonideological 67 days. For the October 2005 Term, these average latencies were 84, 78, and 66 days. In both terms, ideological cases took longer to reach final resolution. But in the October 2001 Term, the justices took less time to decide ideological cases—82 days versus 92 days for nonideological cases.
Reanalyzing Data from the Supreme Court Forecasting Project
Finally, with regard to predictions generated from the initial hypothesis that persuasion operates differently based on the nature of the case, a further analysis was done of some data generated by researchers at the School of Law at Washington University in St. Louis (Ruger, Kim, Martin, & Quinn, 2004). The Supreme Court Forecasting Project has been an exceedingly useful vehicle for anyone interested in Supreme Court decision making. For all cases in the October 2002 Term, the staff developed a prediction of votes and decisions based on a statistical formula that employed only six generally straightforward variables. These were the following:
1. The circuit court of origin.
2. The issue area of the case (using the 15 topic areas developed by Harold J. Spaeth in his U.S. Supreme Court Judicial Database).
3. The type of petitioner (e.g., the federal government, an employer, a defendant, etc.)
4. The type of respondent.
5. The ideological direction (liberal or conservative) of the ruling by the lower court.
6. Whether the petitioner argued that the law or practice was unconstitutional (Ruger et al., 2004, p. 1163).
This information was fed into classification trees and generated predictions for the votes of each justice and hence the outcome of the case. The classification trees differed from justice to justice; a variable that was prominent in the decision tree of one justice might be “relatively unimportant or altogether absent in another” (Ruger et al., 2004, p. 1165). But also, the decision trees of the different justices are not independent of one another; for example, a branch point in Justice Thomas's decision tree is based on Justice Scalia's anticipated vote (Ruger et al., 2004, Figure 9, p. 1198).
The staff also identified a pool of 83 experts, 71 law professors, and 12 appellate attorneys (including 38 former Supreme Court law clerks), and, for each case, asked as many as three who had specialized knowledge (p.67) in the type of case to predict the votes and outcome. Experts were asked to predict the outcomes of cases within only their areas of expertise. Like the predictions from the statistical formula, all predictions by the experts were made prior to oral arguments. Experts were provided a copy of the lower court opinion and citations to the parties' Supreme Court briefs, but they were free to consider any sources of information they considered relevant.
Overall, the statistical model correctly predicted 75% of the decisions in the October 2002 Term, while the experts, as a group, were correct only 59.1% of the time (Ruger et al., 2004, Table 1, p. 1171). It was hypothesized that decisions in the ideological cases would be more predictable than those in the nonideological cases, and so a further analysis was done of data that were available on the project's website (www.wusct.wustl.edu). Considering the statistical model first, it was found that it was correct in 26 of 33 ideological cases or 78.8%, correct in 11 of 15 borderline-ideological cases, or 73.3%, and correct in 16 of 23 nonideological cases, or 69.6%. Thus the predicted difference was obtained. For the experts, in ideological cases, 53 of 88 were correct or 60.2%, while 18 of 35 were correct in borderline-ideological cases, or 51.4%, and 32 of 55 were correct in nonideological cases, or 58.2%. For the experts, the differences are not linear, as they are with the statistical model, and the experts did not do appreciably better in the ideological cases than in the nonideological ones.
In summary, persuasion does appear to operate somewhat differently based on whether the case is an ideological one or not. Data from several terms indicate that:
1. In ideological cases, the final vote less often is unanimous.
2. In ideological cases, the time it takes to reach a decision is longer, at least in the majority of terms.
3. In oral arguments in ideological cases, justices direct more questions to the advocate or advocates who represent what later becomes the losing side, implying that to some extent justices have already formed an opinion in ideological cases before the oral argument.
4. In ideological cases, the decision of the Court is less often consistent with the position of the Office of the Solicitor General, again reflecting the power of the justices' own ideologies in deciding such cases.
Most of the empirical tests of the hypothesis produce small differences, although the pattern is strongly in the expected direction. The smallness of the differences is perhaps not surprising, given the rather broad means of distinguishing between ideological and nonideological cases. For example, abortion cases were classified as ideological; in actuality, the topic of abortion triggers (p.68) an instantaneous, value-drive reaction in some justices more than it does in others. More work is needed in specifying the interaction between type of case and the individual justice.
To conclude that justices react in different ways to ideological and nonideological cases may, to many readers, seem to be less than a surprising conclusion. But this variable may be useful in helping us understand the role of persuasion in judicial decision making. A next step is to examine the content (rather than the frequency) of justices' questions during oral arguments, to determine if evaluative comments are made more often in ideological cases. In an analysis of 24 oral arguments during the October 2004 Term, Jacqueline Austin and I recorded 109 instances of a justice's question or comment that was unsympathetic to the advocate's position. In 87 of the 109, the justice later voted against that side. Do these reflect values that play a role in ideological cases?
Testing the Second Assumption
The second assumption offered at the beginning of this paper proposed that persuasion operates differently depending on the particular justice. Here an individual-differences approach may be fruitful. What makes certain justices more persuasive and what make certain justices more resistant to persuasion?
Justices Who Were Effective Persuaders
If we consider justices over the last 50 years, certain ones stand out for their ability to persuade their colleagues. Earl Warren is recognized for taking a conflict-riddled Court that was divided on Brown v. Board of Education and persuading its holdouts so that the Court was able to announce a unanimous decision on May 17, 1954. Chief Justice Warren was not a legal scholar, but the other justices were influenced by his charisma and his political skills. During that period, and even after Warren had left the bench, William Brennan was very influential, even bringing conservatives to his side in some cases, because of his genuine interest in people and his willingness and ability to craft majority opinions that reflected the wishes of justices who did not completely agree with everything that Brennan would have wished to achieve. Consideration of the Court during those years leads to a conclusion that sheer brilliance is not, in and of itself, enough to make justices effective in their attempts to persuade their colleagues (Rosen, 2007). Felix Frankfurter came to the Court with everything going for him: a professorship in constitutional law at Harvard Law School, a number of articles and books on the Supreme Court, and a network of friends and former students in high places in the government. Yet Frankfurter's attempts to ingratiate and manipulate other justices were largely unsuccessful, and, in fact, some justices (Douglas and Black and even, eventually, Whittaker) came to ignore and even ridicule his efforts to persuade.
(p.69) Justices as Recipients of Persuasion
Judges, as recipients of persuasion, can be studied through several approaches. For example, an appellate court is like any other small problem-solving group in that pressures toward uniformity exist and that some group participants succumb to them. Solomon Asch (1955, 1956), a social psychologist, demonstrated in what has become a classic study, that it is very hard for a sole participant in a group project to maintain his or her response when all the other participants differ in response, even if he or she is the only participant who is correct. That such pressures to conformity often cause the outlier to succumb has been demonstrated in everything from jury deliberations to decisions by the advisers to the United States president. Justice O'Connor is quoted in a recent book as saying that justices would never change their vote simply to be a part of the majority (Greenburg, 2007) but it does happen. Certainly on occasion justices join the majority opinion even when they have reservations. A memo from Chief Justice Burger to Justice Black in a 1971 case said: “I do not really agree but the case is narrow and unimportant except to one man . … I will join you in spite of my reservations” (Maltzman, Spriggs, & Wahlbeck, 2000, p. 22). The justices even have a name for this reaction, as illustrated in a communication from Justice White to Justice Marshall: “I was the other way, but I acquiesce, i.e., a graveyard dissent” (Maltzman et al., 2000, p. 7).
Responses to Pressures Toward Uniformity
The claim that even justices frequently respond to pressures to uniformity is illustrated in an analysis of the data generated by the Supreme Court Forecasting Project (Wrightsman, 2006, Chapter 10). Recall that this project generated predictions of each justice's vote in each case for the October 2002 Term. The project then took these anticipated votes to make predictions of the outcomes. But the procedure did not have any means to add in what might be called a “conformity correction.” Thus the statistical model predicted that in this term there would be only 7 unanimous decisions out of 72 cases. In actuality there were 31. The model predicted there would be 20 8-to-1 decisions, but in actuality there were only 4 (Wrightsman, 2006). Clearly, when most justices vote one way, pressures exist on the holdout justice to go along, and often they do. An analysis by Granberg and Bartels (2005) of voting patterns in the Court extending back to the 1950s found that 8-to-1 votes were the least frequent type, accounting for only 10% of the decisions. In contrast, over this period of 48 terms, unanimous votes occurred in 35% of the cases and 5-to-4 votes in 21%. William O. Douglas, who served on the Court for 36 years, the longest of any justice, also has the record for the most sole dissents, 106, or about 3 a year. But William Brennan, on the Court for almost as long as Justice Douglas, had only 11 in 34 years, reflecting his desire to be conciliatory. And Chief Justice Burger, who did not like other justices to write dissents, or even concurrences, was a sole dissenter only 4 times in his 17 years on the Court.
(p.70) Such individual differences are also apparent among current justices; in the last five terms, Justice Stevens has been a sole dissenter 10 times, Justice Thomas 6 times, and Justice Scalia 4 times. In contrast, Justice Ginsburg never was a sole dissenter in that time period, and neither Chief Justice Roberts nor Justice Alito have been sole dissenters in their briefer times on the Court. (As noted, Chief Justice Roberts, during this term, has been advocating for more narrow, unanimous decisions.)
What contributes to a justice's being a sole dissenter? Strongly held opinions and a relative lack of concern about the opinions of others would seem to be determinants. In his 30 years on the Court, Justice Stevens has always gone his own way. Kenneth Starr has written that “he has taken on the role of a naysayer” (2002, p. 43). Jan Crawford Greenburg's recent book calls him “an iconoclast” and has this description of him:
Stevens was a maverick who didn't ascribe to a particular theory. He was fiercely independent in his writings and actions. When the justices donned their robes before taking the bench, Stevens was the only one who refused assistance from the aides in the robing room. He always insisted on putting on his own robe. He took his own path in his opinions, too. (2007, p. 180)
Justice Thomas has, of course, been subjected to intense scrutiny from the time of his nomination to the Court; within the last year two books have been published that direct attention to his style of decision making. Supreme Discomfort (2007) by Kevin Merida and Michael Fletcher provides an insight into Justice Thomas's background and possible reasons for his adamant position on ideological issues. Jan Crawford Greenburg's (2007) book illustrates that even from his first months on the Court, Justice Thomas had an independent streak. At his very first conference after the oral arguments (Foucha v. Louisiana, 1992) he chose to dissent from the majority, and it initially appeared that he would be the sole dissenter. (Several months later, after he had circulated his written dissent—which was sharply critical of the majority opinion by Justice White—three other justices shifted from the majority to his minority side.) In his second week of oral arguments, he again chose to be a sole dissenter, in the Eighth Amendment case Hudson v. McMillian (2002). His dissent drew wide condemnation in the media (the New York Times editorialized that he was “the youngest cruelest justice”) but again it drew the support of one other justice (Scalia).
Greenburg's view is the following:
Though quiet on the bench during public sessions, Thomas wasted no time sharing his views in conference. Pundits and analysts would disparage Thomas as Scalia's intellectual understudy, but from the beginning that portrayal was grossly inaccurate. … Thomas made clear that he was willing to be the solo dissenter, sending other justices the strong signal that he would not moderate his opinions for the sake of comity. (Greenburg, 2007, p. 115)
(p.71) According to Mark Tushnet, Justice Thomas's strong will and uncompromising positions created problems when he was assigned majority opinions by Chief Justice Rehnquist. Tushnet writes:
Thomas was rarely unsure about the positions he took, so strategy never counseled in favor of giving him an important opinion. Indeed, his very certainty sometimes recommended against doing so. Thomas tended to write strong opinions, drawing sharp lines and rarely acknowledging that different circumstances might produce different outcomes. O'Connor and Kennedy sometimes preferred a more nuanced doctrine than Thomas's opinions articulated. Also, Thomas was more reluctant that other justices to accommodate such concerns. As a result, Thomas “lost a Court” more often than other justices given opinion assignments. That is, instead of gaining the five votes that seemed to be available when the opinions were assigned, his opinions might get only four votes. (Tushnet, 2005, p. 86)
Tushnet makes an interesting observation. It seems to predict that, at least when Chief Justice Rehnquist was assigning opinions, Justice Thomas was more likely to be assigned “safe” ones. An examination of the opinion assignments for the last two terms when Rehnquist was Chief Justice (October 2003 and October 2004 terms) finds that Justice Thomas wrote 16 opinions for the Court, and 9 of the 16, or 56%, were unanimous. Of the remaining 7, 3 were 7-to-2 or 6-to-2, 2 were 6-to-3, and 2 were 5-to-4. The percentage of 9-to-0 decisions in cases in which Justice Thomas authored the Court's opinion was much higher than the 35% to 40% unanimity in those terms.
Justice Thomas's response to his critics is to say, “I don't care what they think. I am free to live up to my oath” (Greenburg, 2007, p. 121). And it is certainly true that his independence is manifested in a number of ways, some related to his work on the Court and some not. His failure to participate in oral arguments has been noted here and elsewhere. During a visit to the University of Kansas Law School, he was interviewed by the local newspaper; when asked about oral arguments, he told the reporter, “I don't see the need for all those questions. I think justices, 99 percent of the time, have their minds made up when they go to the bench” (Rombeck, 2002, p. 5B). Justice Thomas also demonstrates his nonconformity in his off-the-Court preferences, including—in what seems to be a deliberate act of perversity in Washington, D.C., where everyone lives and dies with the results of the Redskins' football fortunes—his identification as a Dallas Cowboys fan.
Individual Differences Variables and Personality Variables
Psychologists interested in personality and individual differences have generated concepts applicable to resistance to persuasion, going all the way back to work on dogmatism in the 1950s. Tetlock's (1983a; Tetlock, Bernzweig, & Gallant, 1985; Gruenfeld, 1995) work on integrative complexity or cognitive (p.72) complexity led him to analyze majority and dissenting opinions by the Supreme Court, and such concepts could be applied to resistance to persuasion. As discussed in the Mitchell and Tetlock chapter in this book, Isaiah Berlin's prototypes of the hedgehog and the fox can be used to identify formulaic approaches to decision making, including that by judges and justices. These and other related traits, such as need for cognition, dogmatism, and the need for cognitive closure are worthy of the attention of scholars studying persuasion in judging.
In the preface to The Puzzle of Judicial Behavior, Lawrence Baum expresses his belief “that we are a long way from achieving explanations of judicial behavior that are fully satisfactory” (1997, p. xi). That was more than10 years ago; this chapter has presented a modest effort toward an explanation, by identifying ways that persuasion operates differently on judges as they form their opinions, depending on their biases and the nature of the case. Beyond this, scholars are beginning the hard work of examining the files of several recent justices, especially the detailed records of Justice Blackmun, to illustrate the role of persuasion during the process of moving from the initial decision draft to the final opinion of the Court. The recent article by Johnson, Spriggs, and Wahlbeck (2007) is an excellent example. We can look forward to a greater understanding of the role of persuasion at all stages of judicial decision making.