Abstract
Although scholars have established that oral arguments play a role in Supreme Court decision making, a fundamental question remains: can oral arguments change justices’ votes? Using data on the positions taken by Justices Blackmun and Powell prior to oral arguments, the authors seek to answer this question while implicitly addressing another: how effectively can attorneys persuade the Court during arguments dominated by justices attempting to persuade each other? The authors find that in a significant minority of cases, justices are persuaded to switch their vote as a result of oral argument and that high-quality attorneys play a central role in that persuasion.
Oral argument is a relatively small and, truth be told, a relatively unimportant part of what we do.
Contrary to Justice Alito’s conjecture, scholars have uncovered many insights about the role and value of oral arguments for U.S. Supreme Court justices’ decision-making process. However, to date Alito’s point has not been fully tested anecdotally or systematically. Indeed, while we understand the many roles these proceedings play for the Court, including providing justices a venue to gather unique information and providing them the first opportunity to persuade their colleagues, scholars have been unable to conclusively answer perhaps the most fundamental question: can the oral arguments presented by the attorneys change justices’ votes?
The reason scholars have been unable to demonstrate whether oral arguments have a direct effect on votes is the necessary condition to test the counterfactual—a justice’s preferred outcome in a case prior to oral arguments—has not been observable. As a result, research that investigates how these proceedings affect justices’ votes (e.g., Johnson, Wahlbeck, and Spriggs 2006; Black et al. 2011) relies on indirect evidence of persuasion, such as justices’ revealed ideological preferences in previous cases, to infer their position going into oral argument. These studies then use this indirect evidence to assess whether justices’ merit votes after oral argument match the inferred preferred outcome. The problem is that when actors behave strategically, sincere preferences may not be equivalent to revealed preferences because strategic actors may vote contrary to their sincere preferences to gain a future advantage. Given evidence of this behavior on the Court generally (see, e.g., Maltzman, Spriggs, and Wahlbeck 2000; Hammond, Bonneau, and Sheehan 2005), and during oral argument specifically (e.g., Johnson 2001, 2004), revealed preferences are likely an inappropriate way to analyze the direct persuasive effects of these proceedings.
Using newly discovered data, we seek to overcome this barrier by comparing the decisions of Justices Harry A. Blackmun and Lewis F. Powell to change positions from the preference they held prior to oral argument with their vote at conference. Specifically, we investigate the influence of these proceedings on Blackmun and Powell’s positions by testing whether two sources of information—attorneys’ arguments and justices’ questions—are persuasive. Our analysis of these data indicates that, in a significant minority of cases, Blackmun and Powell switched positions as a direct result of strong arguments made by the litigants and active participation by their colleagues.
These findings make several unique contributions. As an initial matter, we provide the first evidence that both attorney and justice participation in oral arguments change justices’ minds about how to vote on the merits of cases they decide. Second, we begin to ascertain the conditions under which these proceedings are persuasive. In particular, we examine the persuasive capacity of both litigants and justices. In so doing, we seek answers to several related questions. Are oral arguments more useful when the justices dominate them, as is increasingly the case? Can a skilled attorney use oral arguments to change minds, while simultaneously serving as a mediator for a conversation between the justices? And in what types of cases are oral arguments most likely to have an effect?
In the next section we answer these questions by situating our research in extant literature. We then build a theory of persuasion and demonstrate Blackmun and Powell’s positions were, at times, fluid prior to conference. Next, based on these arguments, we derive hypotheses to explain our phenomenon of interest. The fifth section discusses our data and method. Finally, we turn to results, conclusions, and directions for future research.
Voting Fluidity and Oral Arguments
That justices are open to persuasion is far from a new idea. Murphy (1964) asserted justices’ positions might fluctuate from their vote at conference to the final vote on the merits and scholars demonstrate justices switch votes based on ideology, uncertainty, institutional norms, and strategic motivations (see, e.g., Howard 1968; Brenner 1982; Hagle and Spaeth 1991; Maltzman and Wahlbeck 1996). These anecdotal (see, e.g., Woodward and Armstrong 1979) and scholarly (e.g., Johnson, Spriggs, and Wahlbeck 2005) works suggest strategic considerations play into justices’ first vote at conference but do not hint at whether these considerations (or others) may result in position fluidity prior to conference. As such, this literature cannot speak to exactly when justices are persuaded, if at all, by what transpires early in the process. We thus uniquely contribute to it by examining conditions that lead to early voting fluidity. Specifically, we investigate the degree to which oral arguments influence a justice’s decision to switch positions even before any tentative votes are cast on the merits.
The case that oral arguments may change justices’ votes is not an easy one to make. This is the result of a relatively widespread belief that these proceedings are little more than a dog and pony show, continued for tradition’s sake and meant to entertain, rather than to persuade. 1 For example, in their seminal work, Segal and Spaeth (1993) argue that initial positions on the merits are too strong and oral arguments too short for them to have any measurable impact. Indeed, they conclude, “[T]he justices aver that [oral argument] is a valuable source of information . . . but that does not mean that [it] regularly, or even infrequently, determines who wins or who loses. . . . [W]e know of no systematic evidence indicating the influence of oral argument on the justices’ decisions [on the merits]” (Segal and Spaeth 2002, 280-81).
Recent research on oral arguments has largely sought to combat this claim. The lion’s share of this research can be summarized into a common conclusion: justices use arguments either as an opportunity to gather unique information from the attorneys (Wasby, D’Amato, and Metrailer 1976; Johnson 2004) or to build coalitions and signal their positions to their colleagues (Johnson 2004; Black, Johnson, and Wedeking n.d.). Specifically, this research suggests justices often use their questions to the attorneys as vehicles to make persuasive arguments to their colleagues (see, e.g., Black, Johnson, and Wedeking n.d.). Despite these findings, scholars have virtually ignored the equally important question of what attorneys can do to affect case outcomes. 2 As such, questions remain as to the direct effect attorneys can have on case outcomes and whether they can overcome the predispositions of the justices they seek to persuade. In the remainder of the article we seek to make this case, beginning with a discussion of persuasion more generally.
Measuring Persuasion
O’Keefe (2002, 5) broadly defines persuasion as “a successful intentional effort at influencing another’s mental state through communication in a circumstance in which the persuadee has some measure of freedom.” Successful persuasion must move persuadees from one view to another. 3 Thus, to measure persuasion, we need two pieces of information: an actor’s position on an issue prior to the attempt at persuasion and her or his position afterward. These pieces of information are easy to obtain in work on mass behavior as scholars can invoke surveys and experiments (e.g., Eagly 1974; Cobb and Kuklinski 1997), but gathering them is more complicated when studying elites who are unlikely to participate in experiments or surveys.
A common solution is to examine previous behavior, infer goals from these behaviors, and assume past goals predict future goals (Morrow 1994). 4 The disadvantage of this strategy is that it relies on revealed preferences, which are often influenced by strategic considerations. For example, presidents often strategize about whether to veto a bill and members of Congress make concessions to avoid vetoes (Cameron 2000). Voting in Congress (see Calvert and Fenno 1994) and on the Supreme Court (Johnson, Spriggs, and Wahlbeck 2005) is also often characterized as sophisticated.
Because strategy may affect behavior, it is difficult to utilize past behavior to infer preferences. Thus, we must isolate an actor’s preference prior to the point in a decision making process that may influence a given choice. This was a nearly impossible standard until recently. Ringsmuth’s (2009) analysis of private notes written by Blackmun and Powell prior to the oral arguments in a case provides a mechanism to examine the influence of oral arguments. In these notes, the two justices explicitly stated their initial inclination in a case, and Ringsmuth found they actually switched their votes immediately following oral argument in a significant minority of cases. 5 Because we now have data prior to oral arguments in a case, they provide unique leverage on the degree to which these proceedings are persuasive to (at least two) justices.
Following Ringsmuth’s (2009) lead, we analyze Blackmun and Powell’s decision to alter their position recorded prior to oral arguments, usually within a few days of when the case was heard. We then examine their merit votes at the first conference after oral argument (O’Brien 2005), which are cast within a few days of our “treatment.” As such, we believe Blackmun and Powell’s voting fluidity—when they switched from reverse to affirm or vice versa—can be attributed directly to oral arguments. This is intuitive given that justices do not discuss cases prior to oral arguments, nor do they discuss the case after oral argument and prior to conference. Therefore, these data permit us to focus solely on the intervening effects of oral arguments on a justice’s voting fluidity (Ringsmuth 2009). Extending both the time frame and the sample size of Ringsmuth’s data, the data we present here mimic, as closely as is currently possible, a natural experiment testing the persuasive effects of oral argument.
Initially we demonstrate the justices’ positions can change during oral arguments. While we examine the possibility of a vote switch in every orally argued case from 1971 to 1993, our data are limited in two ways. First, the justices did not take a pre-oral-argument position in every case. Second, there are missing data from Blackmun and Powell’s conference votes on the merits. 6 As such, these data (our dependent variable) contain only cases where Blackmun and Powell’s pre-oral-argument position and conference vote were both nonmissing. 7
Figure 1 depicts a stacked bar plot of how often Blackmun or Powell switched votes per term (light gray) relative to when they did not switch (dark gray). The numbers at the top of each bar are the proportion of vote switches per term (LFP for Powell and HAB for Blackmun). In general, there is a fair amount of term- to-term variation in the justices’ fluidity. Similar to Ringsmuth (2009), we find Blackmun switched his vote in slightly over 10 percent of cases. Powell’s position was less fluid, but he was still persuaded by oral arguments in about 7 percent of cases. Powell exhibited more term-by-term variation than did Blackmun; Powell did not switch once in the 1975 term but did so almost 25 percent of the time in 1976. 8
Stacked bar plot of the number of times Justice Powell or Justice Blackmun’s vote switched per term.
Persuasion during Oral Arguments
Beyond the descriptive evidence of voting fluidity, we analyze the characteristics of oral argument that would make these proceedings most persuasive and disentangle the contributions to this process made by attorneys and justices. It is to the latter task that we first turn.
If information about policy considerations, the preferences of external actors, and institutional constraints is valuable to justices (Johnson 2001, 2004), then the source of the persuasion is the attorneys who provide this information. Alternatively, if justices glean value from oral arguments because it is a mediated conversation between justices and provides information about colleagues’ preferences, the justices themselves are the source of persuasion. Because each side is allocated a finite amount of time (usually thirty minutes) the amount that each “source” may participate is zero sum; the more time justices spend speaking to their colleagues, the less time the attorneys have to persuade the Court. The reverse is, of course, also true.
This argument corresponds to the body of literature that examines the relationship between justice participation in oral arguments and likelihood of success for each side. While many studies examine this relationship (Shullman 2004; Wrightsman 2008; Johnson et al. 2009), their conclusions are consistent: the more justices engage a side, the more likely that side is to lose the case. There are several possible explanations for this phenomenon. First, the more justices engage one side, the more likely Blackmun and Powell are to be exposed to the weaknesses in that side’s position. Second, justices may be signaling their own preferences via statements and questions during arguments, which provides their colleagues with valuable insights into the coalitions that may emerge at conference (Johnson et al. 2009).
More generally, psychologists find that when people actively resist persuasion their attitudes become more ingrained (Tormala and Petty 2002). This effect is especially strong when someone argues against sources with high expertise on a topic (Tormala and Petty 2004)—for our purposes oral advocates. The result, according to this theory, is justices who pose more questions to one side during oral arguments may be persuaded to not cast a vote to support this litigant’s policy position. Combined, this leads us to expect,
Justice Questions Hypothesis: Blackmun and Powell are more likely to switch positions and support the side asked fewer questions by the Court during oral arguments.
Political science and psychology also suggest attorneys should have unique abilities to persuade justices during oral arguments. This ability is born from perceived expertise, and therefore the capacity to provide valuable information to the Court. Hovland, Janis, and Kelly (1953) suggest sources perceived as more credible have a greater ability to produce attitude change. Expertise is especially important when the target and the source disagree. As O’Keefe (2002, 194) writes, “[W]ith a counter-attitudinal message, the high-credibility communicator will tend to have a persuasive advantage over the low-credibility source.” Thus, presenting a credible and high-quality argument is especially important when trying to persuade an adversarial justice, who came into oral arguments disagreeing with the litigant, to change her or his view of the case.
Existing research on oral arguments supports these contentions. McGuire (1995) argues repeat players are more successful because they are perceived as credible sources that provide higher quality arguments. Similarly, Johnson, Wahlbeck, and Spriggs (2006) find attorneys who present higher quality arguments are more successful. We therefore expect,
Argument Quality Hypothesis: Blackmun and Powell are more likely to switch their vote toward the side that presents the higher quality argument.
Beyond the general credibility and expertise of private attorneys, the U.S. Solicitor General (SG) has a unique relationship with the Supreme Court (Black and Owens, forthcoming) that we expect to have the similar effect of either magnifying or blunting the persuasive effect of oral argument. Whether the SG is the ultimate repeat player (Bailey, Kamoie, and Maltzman 2005) or provides unique and unbiased information to the justices (Wohlfarth 2009), decades of research demonstrates the SG is highly successful before the Court. This effect is magnified when the SG participates as amicus curiae because this signals the case is especially important to the SG (and therefore the administration). As such, in addition to the general quality hypothesis we posit,
SG Hypothesis: When the SG appears at oral arguments, Blackmun and Powell are more likely to switch their votes toward the side the SG supports.
Persuasive Effects beyond Oral Arguments
Factors outside of oral arguments may also make a justice more open to persuasion. First, higher levels of involvement with an issue may inhibit acceptance of contrary ideas (Petty and Cacioppo 1986), and those who are more interested in a topic are more likely to resist persuasion (Chaiken 1980). Furthermore, attitude intensity inhibits persuadability (Mehrley and McCroskey 1970). Thus, those who are less active or have not thought about a given policy may be more open to persuasion. In the context of the Supreme Court, “[t]he potential for oral argument to have important effects is clearly the greatest when one or more [j]ustices have not made up their mind prior to argument” (Schubert et al. 1992, 37). Therefore, we hypothesize,
Uncertainty Hypothesis: Blackmun and Powell are more likely to switch votes when their initial position is uncertain.
Characteristics of policy choices may also increase uncertainty. Chief among these factors is previous experience—the more an actor encounters an issue, the more opportunity her or his opinions have to form and harden. Moreover, cognitive consistency theory suggests the more time someone has to position herself or himself on an issue in the past, the less likely she or he is to take contrary positions in the future (Reardon 1981). With respect to Supreme Court decision making, “[u]ncertainties might arise from the . . . novelty of the issues . . . or even the inexperience of the [j]ustices” (Schubert et al. 1992, 37), making familiarity a key component of persuadability. Accordingly, we expect,
Policy Experience Hypothesis: Blackmun and Powell are more likely to switch their vote when they have less experience with an issue area.
Finally, for a message to be persuasive, it must be understood. Indeed, comprehension is a key prerequisite of attitude change (Eagly 1974), and people are less likely to be persuaded by complex arguments or in situations where many options are available to the actor. Justices on the Supreme Court are no different in this respect. Since complex cases are likely to have more complex arguments, and thus more policy options, we expect,
Complexity Hypothesis: Blackmun and Powell are less likely to be persuaded by oral arguments in more complex cases.
Data and Method
To test our hypotheses, we analyze Blackmun and Powell’s decision to switch votes in approximately sixteen hundred cases between 1971 and 1993. 9 Our dependent variable equals one when Blackmun or Powell took a different position at conference than the one he favored prior to oral arguments and zero otherwise. 10
We next turn to our independent variables. Number of questions difference measures the relative number of questions justices ask of each side. Specifically, we generate this variable by subtracting the number of questions posed to the side Blackmun or Powell initially supported from the number of questions asked of the side he initially opposed. 11 We expect a negative relationship since an increase in the relative number of questions asked of one attorney decreases the likelihood this attorney’s position will prevail (Johnson et al. 2009).
Johnson, Wahlbeck, and Spriggs (2006) demonstrate the grades Blackmun gave to each attorney during oral argument serve as a valid proxy for argument quality. 12 Following Johnson et al., we use z scores to standardize these grades across the several different scales. Attorney quality difference is thus the difference between the standardized grade for the side Blackmun or Powell favored in his pre-oral-argument notes minus the grade for the side he opposed. Higher values of this variable indicate a quality advantage for the attorney attempting to persuade Blackmun or Powell away from his initial preference.
Next we capture whether Blackmun or Powell’s position was uncertain. Ordinarily this variable would be difficult to measure because it would require knowing each justice’s state of mind. However, both Blackmun’s notes and Powell’s notes include an indication of their uncertainty and commitment to a given position in a case. In other words, they would explicitly state if they were open to persuasion or if their position going into oral argument was tentative. Blackmun would often note his uncertainty by including a question mark next to his pre-oral-argument position. Other times he would note that he would wait and see whether oral arguments would persuade him. In South Dakota v. Opperman, he wrote, “For the moment, then, I am inclined to affirm, but I am fairly open by way of persuasion from the oral argument.” 13 Powell was more likely to explicitly express a tentative vote. In Navarro Saving Association v. Lee, he noted that “If cases permit I’m inclined to affirm CA5. But I’ll await argument.” 14 Accordingly, we create uncertain position, which equals one when either justice stated his position was provisional.
To test our Policy Experience Hypothesis, we follow Maltzman, Spriggs, and Wahlbeck (2000) and measure prior policy experience as the total number of cases within which Blackmun or Powell wrote an opinion for each narrow issue area in the Supreme Court Judicial Database (Spaeth et al. 2011). 15
Next, we follow Collins (2007) and measure case complexity as the total number of amicus briefs filed in the case. Collins argues that as the number of amicus briefs increases, the number of arguments presented to the justices likewise increases, making the case more complex. While this measure is not ideal because it does not directly measure argument complexity, an increase in amicus briefs increases the number of potential arguments advocates can forward.
Baum (1995, 4) writes that “in major cases oral argument often has little effect on the outcome . . . justices care more about major cases, their predilections are probably stronger.” Therefore, to account for potential relationships between Blackmun’s and Powell’s openness to persuasion and their decision to switch their vote in “important” cases, we include a control for case salience, which we operationalize as Epstein and Segal’s (2000) media-based measure.
Finally, we control for several factors related to majority coalition formation. First, we include the ideological distance between Blackmun or Powell and the Court’s median because of the potential spurious relationship between the justice’s vote switch and his uncertainty about the (arguably) most vital member of any coalition. We measure ideological distance as the absolute value of the difference between Blackmun’s or Powell’s Martin–Quinn (2002) score and the median justice’s score. For the three terms in our data in which Blackmun was the median, we measure the distance between him and the next closest justice on either side. We also control for cases with a minimum-winning certiorari coalition as justices might be uncertain about the policy preferences of their colleagues in these cases. In addition, we account for a switch to majority coalition, as previous evidence suggests a justice is more likely to change her or his pre-oral-argument position at conference if doing so would put her or him in the conference majority (Ringsmuth n.d.; Howard 1968). Finally, we control for the total number of questions asked during the oral argument because there may be substantive differences between cases where the justices asked many questions and cases where the justices largely ignored the attorneys.
Results
Table 1 depicts the results of our analysis, and we begin with the variables that test the potential sources of persuasion during oral arguments: attorneys and justices. First, a clear story emerges of attorney persuasion during these proceedings. Indeed, Blackmun and Powell’s conference votes were more fluid when the attorney who argued for the side they initially supported was less skilled than was the attorney who argued for the side they initially opposed.
Logistic Regression of Blackmun and Powell’s Decisions to Switch Votes after Oral Argument.
p < .05, two-tailed.
To demonstrate this relationship, Figure 2 portrays the predicted probability that Blackmun or Powell switches votes as a function of the quality differential between the two attorneys. The solid black line indicates the predicted probability of a vote switch when all values, except for his expressed tentativeness, are set at their mean or modal values, 16 and the black dashed lines represent the 95 percent confidence intervals. Clearly, when the side opposed by Blackmun or Powell presents significantly lower quality arguments than the side he supported, these two justices were much less likely to alter their positions. As that advantage reverses, and the opposed attorney becomes more skilled, Blackmun and Powell’s probability of switching positions increases nearly 30 percent from the lowest value of quality differential to the highest. This represents a statistically and substantively significant difference, which is strong evidence that oral arguments provide a venue for attorneys to persuade the justices at the same time that their position is the right one.
Predicted probability that the justice switches his vote as a function of quality advantage of attorney that justice entered oral arguments opposing.
Beyond attorneys’ persuasiveness, justices’ comments and questions during oral arguments also affect their colleagues’ votes. Blackmun and Powell are most likely to switch votes when the side they initially oppose is asked fewer questions than the side they supported (negative numbers on the x-axis in Figure 3). As Figure 3 illustrates, if the petitioner and respondent are asked the same number of questions, the predicted probability that the justice switches his position is 0.27. However, if Blackmun or Powell opposes the petitioner’s position going into oral arguments (i.e., he prefers to affirm) and the petitioner’s attorney was asked fifty fewer questions than the respondent (two standard deviations below the mean), the predicted probability of Blackmun or Powell switching votes increases to 0.44. Compared to the opposite situation, when the side Blackmun or Powell initially opposes is asked fifty more questions, their probability of switching positions decreases to 0.15. 17 These results reaffirm that justices’ activity during oral arguments is noticed by their colleagues and suggests this behavior can be persuasive.
Predicted probability that the justice switches his vote as a function of the number of questions asked of each side by justices.
Together these findings provide direct evidence that both attorneys and justices convey persuasive information during oral arguments. They also set up an interesting situation for the justices. Because of the time constraints set for these proceedings, a competition could exist between attorneys and justices for control over this high-level discussion. That said, our results underscore the importance of hearing from both actors because both are directly linked to the stability of Blackmun and Powell’s policy positions before and after the arguments. This, then, supports our contention that oral arguments offer a unique opportunity for justices to gather information from attorneys with an important caveat. Indeed, it is not clear whether justices should entirely cede their time to the attorneys because justices’ contributions can and do influence their colleagues’ policy positions. Rather, the results suggest that if justices dominate these proceedings, doing so could come at the expense of the important information provided by attorneys. Our findings demonstrate justice participation in these proceedings had not yet crossed this line during the period of analysis, but it is an open question whether the increase in the number of questions asked over time (Barnes 2011) could do so.
Beyond our main findings, we find some evidence that factors outside oral arguments are critical for understanding the conditions under which persuasion is likely to occur at this stage. Specifically, we find strong support for our Uncertainty Hypothesis. Perhaps it is not surprising that when Blackmun or Powell takes a tentative, rather than a firm, position prior to oral arguments, the probability of vote switching nearly quadruples, moving from 0.07 to 0.27. In addition, Blackmun and Powell’s experience with an issue area influences their likelihood of changing positions. As both become more experienced with a policy area, and thus more sure of their stances on the legal issues involved, they are less likely to switch their position. The probability of voting fluidity in an issue area with which a justice is totally unfamiliar is approximately 0.30. This decreases to 0.13 when he has written opinions in ten cases (two standard deviations above the mean) in the issue area of the current case.
Finally, two additional control variables affect whether Blackmun’s or Powell’s votes are fluid. First, the justices were more likely to switch positions if doing so would put them in the conference majority coalition. This suggests that, during oral arguments, Blackmun and Powell looked ahead to the coalition formation process and potentially strategic motivations led them to switch positions. In particular, the incentive to change votes may emanate from the fact that being a member of the majority coalition carries with it the important benefit of influencing policy the Court sets. Furthermore, while the presence of the SG favoring the side Blackmun or Powell initially opposes makes them no more likely to switch their positions toward that side, if the SG appears in favor of the side they already support they are less likely to switch their votes. This indicates that while the SG might not be uniquely persuasive, the office does have the ability to reinforce a justice’s preexisting thoughts about a case.
Conclusion
Our results demonstrate attorneys can and do have an impact on how justices view cases they decide. Moreover, their oral arguments play a distinct and vital role in the justices’ decision making process. These proceedings serve the purpose we expect them to: they provide justices a unique venue from which to seek novel information and then for justices to use that information to inform their conclusions. While the justices’ predispositions often win out, oral arguments can serve as a place for gathering information from both attorneys and fellow justices. In short, oral arguments can and do change justices’ minds.
In addition, our findings contribute to a deeper understanding of voting fluidity. While a healthy debate exists over the causes of this phenomenon during the opinion writing process, the literature has not examined the possibility of fluidity earlier in the Court’s decision making process. By doing so here we gain a clearer picture of the conditions under which justices’ positions change as they seek a final decision on the merits. Furthermore, the connection between oral arguments and persuasion sheds light on the hypothetical situation in which the Court does not hear oral arguments. Indeed, our findings suggest the possibility that the policy set in an opinion may be different without the benefit of these proceedings.
At the very least, our analysis makes clear that persuasion during oral arguments not only changes justices’ positions but also, given they are also more likely to switch votes to join the majority coalition, could therefore affect who wins or loses a case. At the same time it is possible that, when justices are initially uncertain in a case, they may shy away from writing the opinion or may write more narrow opinions. While our findings cannot directly speak to these possibilities, the fact that justices can be persuaded to change their conference votes based on what transpires during oral arguments leaves implications, which are ripe for future analysis.
Finally, while we provide initial evidence that attorneys play a vital role in persuasion, we leave largely unanswered the more interesting question of how they do so. The extensive literature on message characteristics is a promising venue for future analysis. What makes a message most persuasive? When are appeals to emotions effective? What components of a message make it most comprehensible? Furthermore, we are interested in shedding more light on the mechanisms through which justices are most likely to have a persuasive impact. For example, are some justices more effective persuaders at oral argument than others? These possibilities also deserve further exploration.
As is often the case, our results are merely a first step and, of course, face limitations. The most notable limitation is that we examine the persuasive effects of oral arguments on only two justices. The lack of comparable data for the entire Court constrains our ability to test our model further. However, past work on voting fluidity demonstrates vote switching from conference to the final votes on the merits occurred for all justices on the Burger Court, which suggests it is unlikely that fluidity during the oral argument stage is unique to Blackmun and Powell (Maltzman and Wahlbeck 1996). Moreover, we believe that because of key differences between Blackmun and Powell, they are the right two justices to examine if we seek to draw generalizable conclusions. While Blackmun was often squarely in the ideological wings of the Court, first as a conservative and later as liberal, Powell was often the median or very near it. As such, we can generalize our findings to ideologues and to seemingly more persuadable “swing” justices. Furthermore, Epstein, Martin, Quinn, and Segal (2007) and others demonstrate that while Powell’s beliefs were relatively moderate, they were also consistent, indicating he may have been less persuadable. Alternatively, while Blackmun’s ideology was more extreme, it also shifted dramatically from the beginning of his tenure to its end. Thus, we can also generalize our findings to justices who are both more and less persuadable.
Blackmun and Powell also had very different views about the importance of oral arguments. Blackmun’s opinion in many ways mirrored Justice Thomas’. Indeed, he once wrote to himself in an oral argument note that Justice Scalia “asks far too many questions and takes over the entire argument of counsel” (quoted in O’Brien 2005, 261). He also often complained in his notes that his colleagues were asking too many questions. Powell took a different stance and believed his colleagues’ contributions during oral arguments “contribute[s] significantly to the development of precedent” (quoted in Johnson 2004, 14-15, internal citation omitted). That two justices who took such divergent views on one of the key questions examined in our piece were persuaded by the same aspects of oral arguments only lends further credence to our ability to generalize our findings more broadly to judicial decision making.
Like most social science research, our results suggest a particular causal mechanism but cannot prove causality. In other words, it is possible an unobserved latent variable may drive both the difference in the number of questions asked to either side and a justice’s vote in the case. Neither previous work on the number of questions asked at oral argument (see, e.g., Roberts 2005; Shullman 2004; Johnson et al. 2009) nor our findings here rule out the possibility that the number of questions may be endogenous to other case characteristics such as strength of legal position or overall complexity of legal argumentation. It is therefore important for future work to examine this possibility with much more nuanced measures of these and other case characteristics.
Furthermore, while we acknowledge our findings shed light on only one aspect of a complex decision-making process (Epstein and Knight 1998; Maltzman, Spriggs, and Wahlbeck 2000), votes are clearly fundamental. Many other aspects of the process, from opinion assignment to opinion content, are grounded in whether a case is reversed or affirmed and who joins the majority coalition. As such, our analysis furthers our understanding of these core attributes that help explain how justices decide.
Despite these limitations and additional questions, our findings provide a first step toward analyzing the extent to which oral arguments play a persuasive role for the justices and the conditions under which they may be open to persuasion. After having demonstrated these proceedings are uniquely valuable, we leave open for future exploration the most intriguing question: what exactly makes a “good” oral argument?
Ultimately, we find direct evidence attorneys can and do use oral arguments to persuade Supreme Court justices in a significant minority of cases. This persuasive capacity is not unique to attorneys. Rather, the results also indicate a direct link between justices’ engagement during these proceedings and the likelihood of voting fluidity. Thus, the evidence that both groups of actors have an effect on conferences votes contrasts with Justice Alito’s contention that these proceedings are relatively unimportant. This analysis merges key findings in the oral argument, voting fluidity, and general Supreme Court decision making literatures, as well as past psychological work on persuasion. It also presents a unified portrait of the persuasive powers of attorneys and justices during oral arguments. These results are particularly interesting in light of the recent trend of justices speaking more at oral argument and thus listening less.
Footnotes
Acknowledgements
For helpful suggestions we thank David Adkins, Justin Wedeking, Ryan Black, Stephen Wasby, the anonymous reviewers, and the editors of Political Research Quarterly. For providing portions of our data we thank Ryan Black. For assistance with data collection we thank Simon Ringsmuth and Kathryn Haglin. We are eternally grateful to John Jacob of the Lewis F. Powell Archives at Washington and Lee University for his help with the Powell Papers and the staff at the Library of Congress for their help with the Blackmun Papers.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research and/or authorship of this article: Ringsmuth thanks University of Minnesota′s Patrick and Kathy Lewis Thesis Research Grant and Doctoral Dissertation Fellowship for facilitating data collection, and Johnson acknowledges assistance from the National Science Foundation: IIS-0324992.