Abstract
This analysis seeks to understand the decline of Supreme Court consensual norms often attributed to the failed leadership of Chief Justice Stone. A new unit of analysis—justice-level dissent and concurrence rates—supports an alternative view of observed increases in dissensual decision making. When these measures are estimated with time-series techniques, results offer evidence of multiple changepoints in this norm of the Court that both lead and lag Stone’s elevation. Broader contextual explanations related to the alteration of the Court’s discretionary issue agenda and its ideological and demographic composition also contribute to fractures in the once-unanimous voting coalitions.
Stone’s leadership appears to be a prime cause of the Court changing from an institution emphasizing consensus to one characterized by a high level of individual expression.
Institutional changes in decision-making norms are questions of interest to political scientists within the numerous subfields, but perhaps none is as peculiar as the one that challenges those scholars interested in the Supreme Court. Current accounts of the demise of the Court’s norm of consensus all begin with a common conceptualization. Namely, this norm emerged with Chief Justice Marshall’s early abandonment of seriatim opinions, which was a strategic attempt to establish legitimacy for a judicial branch lacking implementation powers. The practice of speaking with one voice became institutionalized and then endured in the form of unanimous opinions that obscured private disagreement taking place during conference deliberations. In the first half of the twentieth century, the norm of consensus was shattered when justices publicly began disclosing internal conflicts in the form of concurring and dissenting opinions. And this newly-visible conflict was the impetus for legal realist critiques (e.g., Pritchett 1948), thus making it a topic of considerable importance for our knowledge of judicial decision making.
Clearly, the seminal approach to this topic can be found in Walker, Epstein, and Dixon’s (1988) analysis. Utilizing an opinion-based measure of the proportion of cases with at least one concurring or dissenting opinion (Blaustein and Mersky 1978), the authors separately evaluate a number of competing hypotheses, including (1) the Judiciary Act of 1925 (hereafter, the Judges’ Bill), which established the discretionary docket (Halpern and Vines 1977), (2) changes in Court caseload, (3) promotion of a sitting associate to chief justice, (4) alteration of the composition of the bench, and (5) the varying leadership (Danelski 1986) capabilities of the chief justice. In this instance, the failed social and task leadership of Chief Justice Stone—who as an associate justice bridled under the no-dissent norm and refused to enforce it—ascends to prominence as the explanation for the growing level of dissensus on the Court (see Walker, Epstein, and Dixon quotation above).
Although Walker, Epstein, and Dixon caution against singular focus on the Stone-did-it narrative, 1 it has become a cornerstone of the conventional wisdom for understanding this historical transformation, despite the fact that other studies challenge the inference with time-series intervention techniques that examine these same opinion-based measures. For example, Haynie (1992) moves the transition point earlier, suggesting that the leadership of Chief Justice Hughes also played a significant role. Caldeira and Zorn (1998) add considerable nuance by modeling the cointegration that exists between concurrence and dissent relationships, with results showing that transformation exhibits a long-run dynamic along with chief-justice-specific variance, including significant effects for Chief Justice Hughes. Most recently, Spirling (2007) has applied a new Bayesian changepoint technique to these series and identified the point of transition as the 1941 term, which once again places the onus on Stone’s failed leadership.
These conflicting conclusions regarding trends in the Court’s dissensual decision making are in one sense the product of competing research designs. Walker, Epstein, and Dixon (1988), for instance, address a wide array of potential hypotheses but lack a multivariate estimation strategy to test the strength of each against the other. Haynie (1992), Caldeira and Zorn (1998), and Spirling (2007) exhibit methodological rigor with respect to the timing of changes but do not delve into the range of possible explanations that Walker, Epstein, and Dixon propose. Whether the failed leadership of Stone is a simplistic explanation or represents a straw man, past studies of the decline of consensual norms have been wanting in some regard. Either such studies did not account for all of the variables or they failed to assess them simultaneously. Thus, we believe that an opportunity exists to further our understanding of the consensual norms of the Court with a new research design that addresses differing aspects of current strategies.
We further suspect that these alternative accounts are not solely a function of varying research designs but may be the result of a less-than-ideal measurement of the breadth of dissensus within the Court. Opinion-based measurement offers a valid means of capturing one component of consensus (i.e., a lack of unanimity). Within annual macro studies, however, it can act to overstate levels of disagreement since the existence of a concurring or dissenting opinion essentially models a single justice-level vote. Because other justices’ support for concurring and dissenting positions is not captured, opinion-based measures contain an upward bias with respect to the levels of dissensus within the Court. For an opinion-based measure, a largely unified Court that includes a consistently obstinate dissenter operationally is equivalent to a sharply divided Court that decides all cases by a 5–4 minimum winning coalition.
A second issue with opinion-based measurement is the reliability of the measurement strategy across sample periods. The aggregation of the Blaustein and Mersky (1978) and Spaeth (2009) data on decisional outcomes tends to incorporate underlying disparities in case types that adversely affects the variables modeled by current studies. 2 The historical nature of this question requires a sufficiently long period of analysis, but it also encompasses several changes in working practices of the Court. The most obvious institutional adjustment is the transition to a discretionary docket, which represents a primary hypothesis; but, other less salient norms (such as opinion format and the issuance of signed opinions) may equally be important. Clear indicators of oral argument dates do not arise until the 1899 term, and, in general, signed opinions were much more prevalent in the period before the Judges’ Bill. In conjunction, these two norms suggest that existing opinion-based measures may subsume early cases more akin to modern-day per curiam opinions, and, thus, the current measurement strategy likely is temporally biased.
To address these issues, we have conducted an intensive data-gathering effort to construct annual justice-level concurrence and dissent rates for orally argued cases between 1899 and 2004. A plot of the existing opinion-based, and new justice-level, rates of dissent for the Court can be found in Figure 1, which depicts a temporal bias within current measures that may hold considerable implications for our knowledge of consensual norms.
Court-level dissent—Opinion measures versus justice-level votes.
In practice, an opinion-based measure should always be greater than a corresponding justice-level rate (see Supplemental Appendix A, Supplemental Figure 1, and Supplemental Tables 1 and 2, at http://prq.sagepub.com/supplemental/, for additional discussion and analysis of opinion-based measures). While this is the case for decision making both during and after the Stone Court, the preceding period shows that a justice-level rate actually is greater than the opinion-based measure. This methodological disparity, which coincides with the Stone Court, is evidence that the term-by-term samples of decisions differ between the Blaustein and Mersky (1978) data and more recent Spaeth (2009) data. When viewing dissenting behavior on the basis of opinion production, a sharp break at the beginning of the Stone Court is observed in the opinion measure. With an alternative unit of analysis and more consistent sample construction, however, the plot reveals a more incremental rate of change across the Hughes, Stone, and Vinson Courts. Therefore, a broader account of change in the norm of consensus may be warranted.
In the following analysis, we sketch out a new narrative of this seminal event. Utilizing the methodology that Spirling (2007) employed, we find evidence that changes in dissenting behavior take place both at the end of the Hughes Court and during the Stone Court. Changes in the rate of concurring votes are found under Chief Justice Stone and later under Chief Justice Warren. The application of traditional Box–Jenkins (1976) techniques to these new series suggests that this transformation in decision making is associated with the discretionary docket and agenda change, the demographic and ideological constructs of the Court, and differences attributable to the leadership styles of Chief Justices Stone and Burger.
Hypotheses for the Demise of Consensual Norms
Although prior studies conflict over the sources and timing of the Court’s movement toward dissensual decision making, they have adequately identified a number of suspects that are implicated in this mystery. Our focus in this instance is to adapt and modify these working hypotheses, associating them with valid explanatory variables and simultaneously evaluating them to leverage their independent effects on the decline of the norm of consensus.
The Judges’ Bill and the Discretionary Docket
Of the existing explanations, the influence of the Judges’ Bill has created the greatest disagreement among scholars. Halpern and Vines (1977) have primarily advocated this perspective, which asserts that it was the Court’s newfound authority to control its docket that acted as the mechanism behind the subsequent growth of dissenting and concurring opinions. In many ways, other scholars’ differing views turn on the issue of timing. Walker, Epstein, and Dixon (1988) note that substantial increases in dissent take place as many as fifteen years after the adoption of the Judges’ Bill, suggesting that the legislation did not have an immediate effect. However, we also know that adjustment to the Court’s docket was incremental. Shifts in exigent and volitional agendas (Pacelle 1991; Lanier 2003), and the Court’s role-change from economic policy maker to arbiter of civil liberties and rights disputes, took place over an extended period (Hurwitz 2006). Concurrent with the showdown over New Deal economic regulation (e.g., the Court-packing episode and “switch in time that saved nine”; McKenna 2002), the Court signaled that such a change was forthcoming.
In the now-famous footnote 4 of United States v. Carolene Products (1938), then Associate Justice Stone argued that the Court should adopt a dual standard of review. In economic regulation cases, the Court should follow judicial restraint. But in civil rights and individual liberties cases, the Court should exercise judicial activism such that the interests of political minorities could be protected. This so-called “preferred position” doctrine moved from a footnote, to a dissent, and eventually to a majority opinion (Pacelle 1991). From this perspective, the heightened variance in the justice-level dissent rate taking place at this time provides visual clues that agenda formation may prove to be a more robust explanation of the demise of consensual norms than currently thought.
Demand for Litigation and “Hard Cases”
Passage of the Judges’ Bill may be crucial from a slightly different stance since it acted to erode existing constraints on the justices’ freedom to pen dissenting and concurring opinions. When the final determination of more routine matters was relegated to the circuit courts of appeals, the time resources necessary to voice public disagreement became more available. Subsequent increases in the demand of litigation through certiorari then presented it with a menu of more challenging (and, thus, contentious) disputes to resolve. With the Court’s docket responding to the percolation taking place in lower courts, it increasingly comprised so-called “hard cases” (Perry 1991), and the emergence of distinct voting blocs may be associated with changes in the relative difficulty of questions to consider. Therefore, in addition to changes in the issue types under consideration (i.e., a transition from economic regulation to civil liberties and rights), a growing docket of potentially more vexing questions offers an alternative explanation for the demise of consensual norms.
Associate Justice Demographic and Ideological Composition
Distinct from those rationales involving the Court’s docket are hypotheses associated with the composition of the Court itself. The appointment of youthful, inexperienced, and therefore dissent-prone associate justices (Walker, Epstein, and Dixon 1988) during the Stone Court is thought to have created an atmosphere in which disagreement could no longer be contained within conference. The concepts of youth and inexperience obviously exhibit considerable levels of correlation with each other (i.e., those lacking experience tend to be youthful). Given its stronger theoretical foundation, we emphasize the levels of judicial experience found throughout the sample period since it is not only youthfulness but also a lack of prior judicial background that may make a difference here (e.g., Justices Black, Douglas, Frankfurter, and others were drawn from executive and legislative branches).
This particular mix of new justices lacked the socialization that they might have received as lower court judges, and such inexperience can serve cross purposes with respect to the working norms of the Court. Being less indoctrinated with institutional expectations of behavior, inexperienced justices could be more susceptible to the goal of speaking and being heard (Maveety, Turner, and Way 2010). From the alternative perspective, the literatures on freshman, or acclimation, effects (e.g., Snyder 1958) indicate that the opinion writing of newly-appointed justices is constrained as they adjust to their new positions and that they may defer to more experienced members of the Court. If the latter is in fact the case, then we should expect greater amounts of individual expression as the level of experience among the associate justices increases.
The exogeneity of the dissent-prone justice hypothesis may be somewhat uncertain since these same appointments take place within a period of historic interbranch conflict. Franklin Roosevelt’s Court-packing plan and observed changes in the vetting of nominees (Abraham 2008) point to strategic efforts to reshape the judicial branch and make its rulings more consistent with the tenets of the new majoritarian alliance (Dahl 1957). The addition of new and inexperienced justices to the bench offers a demographic explanation of this transformation, but the ideological preferences of these appointees are significant, too. The common law system and institutional adherence to precedent presaged greater levels of disagreement as these newly appointed liberal justices sought to overturn existing conservative precedents.
It is interesting that justices allude to such effects in the period just prior to the Stone Court. In Graves v. N.Y. ex rel. O’Keefe (1939, 487), Justice Frankfurter writes separately and speaks to the Court’s past use of seriatim opinions within the context of reversals of precedent: I join in the Court’s opinion but deem it appropriate to add a few remarks. The volume of the Court’s business has long since made impossible the early healthy practice whereby the Justices gave expression to individual opinions. But the old tradition still has relevance when an important shift in constitutional doctrine is announced after a reconstruction in the membership of the Court. Such shifts of opinion should not derive from mere private judgment. They must be duly mindful of the necessary demands of continuity in civilized society. A reversal of a long current of decisions can be justified only if rooted in the Constitution itself as an historic document designed for a developing nation.
Thus, efforts to confer legitimacy on emerging New Deal economic precedents, and to legitimize subsequent forays into liberties and rights issue areas, may provide another compositional explanation.
Given focus on the role of dissent in the rise of legal realist views of decision making, it is somewhat unusual that this ideological hypothesis has not become a more central focus of the literature in this area. However, support for ideological influences within the demise of the consensual norm can be found (Hurwitz and Lanier 2004), as well as in contemporary studies of collegial decision making (Maltzman, Spriggs, and Wahlbeck 2000). Accordingly, we hypothesize not only that the relative experience of the associate justices can explain this transition in part, but also that disparity in their ideological preferences may explain increases in the rates of dissenting and concurring votes.
The Idiosyncrasies of the Chief Justice
The conventional wisdom on this topic emphasizes the characteristics of the presiding chief justice. Danelski (1986) maintained that Stone’s lack of leadership skills was largely responsible for the norm’s decline, contending that the social and task leaderships of Stone’s predecessor, Hughes, were outstanding. From this standpoint, it is chief justices’ abilities to effectively frame cases within conference and clear the docket (i.e., task leadership) and, at the same time, to alleviate the inevitable tension that arises from the decision-making process (i.e., social leadership) that contribute to unanimous outcomes.
The chief justice almost certainly exerts considerable influence in the establishment of a congenial work environment and in the transmission of Court norms during periods of membership change. Yet this position holds institutional advantages, such as priority in vote order and opinion assignment, that can either contribute to or detract from a unified Court. As Johnson, Spriggs, and Wahlbeck (2005) show, the strategic use of passing votes to maintain control over opinion assignment can engender bitterness (e.g., the Burger Court) and fair-play approaches contribute to greater levels of amiability (e.g., the Rehnquist Court).
Past attention on Chief Justice Stone’s role in the demise of consensual norms makes the leadership hypothesis crucial to this study, and we approach it in a twofold manner. First, we take advantage of the justice-level unit of analysis to separately evaluate the idiosyncrasies of the chief justice position. By segregating separate series attributable to the chief and associate justices, we explore any institutional disparities in chief justice dissensus. Second, we introduce interventions for changes in chief justice leadership that will capture idiosyncratic relationships attributable to specific chief justices after controlling for competing hypotheses.
Research Design
Since the effect of the Judges’ Bill is a primary point of contention, we focus on those cases in which the Court exhibited its greatest discretion. While the Court was burdened with a mandatory docket prior to the advent of the contemporary certiorari process, it still maintained some discretion over the way it could approach its workload. Although the control was limited, the Court exhibited control over those cases it deemed worthy of oral argument, and, as today, we expect that these orally argued cases were the most salient and pressing disputes on the docket. Together, then, we argue that the most valid unit of analysis for the study of the demise of consensual norms is a justice-level measurement of dissent and concurrence voting behavior within those cases granted oral argument.
To generate such measures, we have undertaken an intensive data-gathering effort to generate the population of orally argued Supreme Court decisions between 1899 and 2004. 3 We began by conducting Westlaw searches to identify cases receiving an oral argument before 1953. 4 The results of these searches were then merged with two databases: the Wood et al. (1998a, 1998b) data for terms between 1899 and 1940 and the Wood (1994) data for terms between 1941 and 1952. The merged data set was then reconciled, and all unmatched cases were cross-checked against Westlaw to determine whether an oral argument occurred. Any missing case was coded and added to the database, and a search for duplicate entries was conducted with additional entries eliminated. The Spaeth (2009) database was then employed for the term years 1953 to 2004. Cases that received an oral argument and contained justice-level voting breakdowns were included in the analysis. 5 Following this search, the outcome of each case was reconciled with the justice-level votes, and any discrepancy was corrected by referencing the opinion through Westlaw. 6 These data were then used to create the Court-level rates of concurrence and dissent for each term, as well as concurrence and dissent rates for the associate justices (see Figure 2) and the chief justice (see Figure 3). Both of these figures indicate that substantial increases in concurring and dissenting votes take place prior to the Stone Court.
Associate justice dissent and concurrence rates in orally argued cases.
Chief justice dissent and concurrence rates in orally-argued cases.
We analyze our newly generated data with two different time-series techniques. First, we replicate the changepoint analysis of Spirling (2007) with Martin, Quinn, and Park’s (2011) Bayesian application. We then evaluate hypotheses related to agenda transformation, workload, demographic and ideological composition of the bench, and chief justice leadership with traditional Box–Jenkins (1976) Autoregressive Integrated Moving Average (ARIMA) specifications for the Court-level, associate justice, and chief justice rates. This Box–Jenkins methodology is dependent on the estimation of stationary series, and both the dependent and independent variables exhibit unit-root, or trend, relationships. To account for this lack of stationarity, we take advantage of fractional differencing techniques (DeBoef and Granato 1997) that are more suitable to long memory processes, such as the behavior that underlie these forms of dissensual decision making. A complete discussion of the fractional differencing related to this analysis can be found in Supplemental Appendix B and Supplemental Tables 3 and 4 (at http://prq.sagepub.com/supplemental/).
Independent Variable Construction
The data-gathering effort used to generate the above dependent variables also provides a more valid means to evaluate the effect of agenda transformation following the Judges’ Bill, which eventually brought about changes in the Court’s long-run docket. Issue coding for these orally argued cases is used to control for the Court’s varying attention to economic issues and civil liberties and rights disputes (Pacelle 1991; Lanier 2003). We divide the orally argued docket into five categories that identify the number of criminal, civil liberties and rights, economic, institutional power, and original or miscellaneous cases heard in each term. 7 These series are plotted in Figure 4, where the long-run transformation of the Court’s agenda is evident. Economic issues account for the vast majority of orally argued cases in the period both before and after the 1925 Judges’ Bill, but the Stone and Vinson Courts jointly represent a period of transition, reflecting the decline in attention to economic issues that precedes the modern concentration on criminal and civil liberties disputes.
Supreme Court agenda transformation.
To capture effects related to workload and demand for litigation, we control for the number of docketed cases for each term of the Court (Epstein et al. 2007, 62). Demand for litigation incrementally increases from approximately 700 cases to 8,500 by the end of the sample.
To assess the experience hypothesis, we used the Federal Judicial Center’s biographical database to quantify the levels of associate justice experience on the Court at the beginning of each term. 8 We present the mean experience measure in Figure 5. The Stone Court clearly represents an outlier in terms of experience; the mean values are the lowest of any other set of natural courts. In addition to this variable, we include an exponentially declining intervention series that identifies changes in chief justice leadership. 9 This measure controls for freshman effects related to a general change in leadership of the Court. Supplementing this acclimation series are separate intervention controls, 10 for each chief justice after Fuller, that capture residual variance that coincides with the leadership of subsequent chief justices (i.e., from Chief Justice White through Chief Justice Rehnquist). 11
Associate justice experience and ideological dispersion.
The final independent variable captures changes in the ideological context of the Court. The sample period in our analysis limits the use of conventional measures of justices’ preferences (e.g., Segal and Cover 1989). Accordingly, we incorporate an alternative measure of ideological disparity through an adaptation of the Poole and Rosenthal (1997) DW-Nominate values. We use the mean first dimension position of the appointing president’s party in the Senate as a proxy for each justice’s ideological position. These positions were then used to calculate the standard deviation of the collective of associate justices. This series also is presented in Figure 5, which shows the Hughes, Stone, and Vinson Courts embodying a period of considerable ideological flux. The Taft and Hughes Courts exhibit considerable amounts of dispersion among the associates, but FDR’s subsequent series of appointments acted to reconstitute the Court, and the standard deviation value collapses until an alternating series of Democratic and Republican appointments reestablishes degrees of dispersion within the modern era.
Changes in the Court’s Norm of Consensus
Estimation results of the changepoint analysis can be found in Supplemental Table 5 (at http://prq.sagepub.com/supplemental/) and are graphically presented in Figure 6. Overall, neither of the newly generated series exhibits a singular break in equilibrium. Changes in the propensity to dissent take place both in the 1938 and 1943 terms, indicating that the Court’s norm of consensus was first challenged by growing levels of dissent in the later years of the Hughes Court. Dissent then reaches a more contemporary level three years after the inception of the Stone Court. Unlike the rapid-fire changes exhibited in the dissent rate, the Court’s concurrence activity reflects an extended period of transformation. A new equilibrium is established in 1942, or the second year of the Stone Court, and then reaches more contemporary levels twenty-two years later, during the last half of the Warren Court.
Changepoint analysis of Supreme Court dissent and concurrence.
Our initial evidence suggests that the Court’s decision making under Stone’s leadership was different than that under his predecessors. Levels of dissent and concurrence both reach new equilibria, but the first assault on the long-standing norm of consensus actually took place in 1938. This was a period marked by interbranch conflict and realization of the Court’s new discretion in constructing its docket. Modern levels of support for concurring opinions transpire during the Warren Court’s attention to a civil liberties agenda made possible, in part, by this new discretion. The short and sharp interval of change in dissent that traverses the Hughes and Stone Courts and the extended period of transformation in concurrence activity both point to broader explanations for the demise of the norm of consensus.
Sources of Dissent
To evaluate these broader influences, we now turn to a Box–Jenkins (1976) analysis of annual changes in Court-level, associate-justice-level, and chief-justice-level dissent rates from 1899 to 2004. These results are presented in Table 1.
Box–Jenkins Estimates of Dissent Rates in Orally Argued Cases 1899–2004.
Estimations conducted after fractionally differencing of the series. Newey-West HAC (heteroscedasticty and autocorrelation consistent)/prewhitened standard errors ( Akaike Information Criterion [AIC]; four lags) are presented along with two-tailed probability tests.
Series divided by 1,000 for parameter scale.
Series multiplied by 100 for parameter scale. Breusch–Godfrey serial correlation test with eleven lags (H0: residuals do not exhibit serial correlation). White heteroscedasticity test (H0: residuals are homoscedastic).
Starting with the dissent rate for the Court as a whole, we find significant parameters for the effects of agenda transformation (Pacelle 1991; Lanier 2003). Increases in the numbers of civil liberties cases are positively related to growth in the rate of dissent, and greater attention to the Court’s criminal agenda shows a similar but slightly less robust relationship. Economic cases, and to a lesser extent issues of institutional power, exhibit negative relationships, meaning that greater attention to these issues tended to suppress dissent on the Court. Thus, the Court’s transfer of economic cases to an exigent agenda and a growing focus on civil liberties issues concomitantly contribute to the growth of dissent on the Court.
Strong support for the workload hypotheses is not found in these estimates, but the composition of the Court is a factor. Growth in ideological dispersion among the associate justices is significantly related to increases in the dissent rate, meaning that ideological polarization contributes to the production of dissent. Evidence of a freshman effect and chief-justice-specific variance is found. Transition in leadership of the Court induces a short, and then exponentially declining, decrease in the level of dissent that suggests a period of acclimation.
Intervention controls for the chief justices generally reflect the increased variance of the dissent rate over time, but they do offer some leverage on the relative differences between chief justices’ leadership of the Court with respect to Chief Justice Fuller, who represents the null category of the model specification. The intervention parameters double in magnitude between Chief Justice Hughes and Chief Justice Stone. After controlling for agenda change, workload, and compositional explanations, Stone is in fact unique with respect to his predecessor. Relating this to leadership capability, it would seem that Stone either lacked the social and task skills necessary to reinforce the norm of consensus or, as suggested, simply refused to enforce it. 12
As one might expect, models attributable to associate justices emulate the Court-level specification, but the segregation of the associate justices improves model performance, meaning that there are institutional differences in dissent with respect to the chief justice. Agenda change parameters are consistent in direction but reveal more robust significance relationships that show the increasing attention to criminal and civil liberties issues, and corresponding decline in economic and institutional power cases, contribute to greater dissent from the associates specifically.
Ideological polarization continues to be a factor in the production of dissent, but the model now yields some evidence for an effect related to justice experience. Greater levels of associate justice experience are positively related to increases in the dissent rate. 13 This is opposite to the dissent-prone justice expectation and indicates that associate justices are less willing to abide by the no dissent norm as they gain experience on the Court and its membership remains stable. Similarly, some marginal evidence for a positive effect of demand for litigation and the hearing of “hard cases” on the dissent rate now exists in this specification.
The isolation of the chief justice dissent rate presents an alternative view with respect to model performance. Annual time-series techniques do not perform as well for this singular justice rate, 14 but they do provide a handful of useful insights. Effects related to the alteration of the Court’s agenda are absent with the exception of economic cases where the parameter shows a positive relationship. Although economic cases tend to suppress the level of dissent at the Court and associate justice levels, they exacerbate dissent within this chief justice series. At first blush, this could appear to be an aberration, but it more likely is specific to the New Deal era, when interbranch conflict over economic policy was at a peak. The initial spike in the chief justice dissent rate takes place at this time, suggesting that Chief Justice Hughes may have been caught between a bloc of holdover justices (e.g., the Four Horsemen) and FDR’s New Deal appointees. Thus, the result alludes to an initial fracturing of the norm of consensus by the chief justice as he led the Court through a set of volatile economic disputes.
The remaining findings show a strong freshman effect for the chief justice, where newly appointed leaders of the Court hold their dissent as they acclimate (Lanier 2011). Individual intervention parameters again show a sizeable increase in the magnitude between the Hughes and Stone Courts but also show substantial growth in scale between the Warren and Burger Courts. Like Stone, Chief Justice Burger is often cited for his lack of leadership on the Court (Kobylka 1989), and our results provide evidence of unique approaches to the leadership role for both.
Sources of Concurrence
Concurrences may represent a form of judicial egocentrism (Caldeira and Zorn 1998), and our results manifest their idiosyncratic nature in contrast to concurrent patterns of dissent. Systematic relationships within the concurrence series generally are weaker and confined to the Court-level and associate justice estimates. The associate justice results (see Table 2) again mimic and offer some improvement over the Court-level estimates, so they are the most relevant.
Box–Jenkins Estimates of Concurrence Rates in Orally Argued Cases 1899–2004.
Estimations conducted after fractionally differencing of the series. Newey-West HAC (heteroscedasticty and autocorrelation consistent)/prewhitened standard errors ( Akaike Information Criterion [AIC]; four lags) are presented along with two-tailed probability tests.
Series divided by 1,000 for parameter scale.
Series multiplied by 100 for parameter scale. Breusch–Godfrey serial correlation test with eleven lags (H0: residuals do not exhibit serial correlation). White heteroscedasticity test (H0: residuals are homoscedastic).
Associate justice estimates provide some support for the effects of agenda transformation. Greater attention to institutional power issues significantly decreases the rate of concurrence. Increases in the number of economic cases similarly yield a negative and borderline significant finding. Parameters controlling for attention to civil liberties and criminal issues both exhibit null relationships. It appears that a general restructuring of the Court’s agenda affected the rate of concurrence, but it is not necessarily closely tied to the core components of the Court’s modern agenda. Ideological dispersion among the associate justices was a significant source of growing dissent, but similar effects are not found within concurrence activity. Thus, it seems that production and support for concurring opinions may be related to nonattitudinal aspects, such as the maintenance of legal doctrine and circuit management (Corley 2010).
However, like the above findings for dissent, growing levels of experience on the Court do contribute to increases in the associate justice concurrence rate. Together these results indicate that a general fracturing of majority coalitions takes place as members of the Court gain experience and its membership remains stable. Support for an acclimation effect under new chief justices actually is more robust in terms of concurrence activity. While some evidence of a tendency for associates to hold their dissents during a change in leadership of the Court could be found, it seems that the justices are more willing to forgo opportunities, albeit temporarily, to write and support concurrences in such instances.
Finally, the chief-justice-specific parameters provide some insight on long-run changes in concurrence and leadership capabilities. Parameters for the chief justice interventions again reflect a substantial increase in concurrence activity under Chief Justice Stone’s leadership. This is to be expected given the preceding changepoint analysis, but a similar increase is not found for Chief Justice Warren. In fact, his parameter exhibits a reduction in scale versus his predecessor, Vinson. After controlling for agenda, workload, and Court composition, it seems that the establishment of modern concurrence levels did not take place during Warren’s leadership. Instead, the subsequent increase associated with Chief Justice Burger indicates that his leadership may have been responsible for cutting the remaining threads of the norm of consensus.
Discussion and Conclusions on the Stone Hypothesis
Our results suggest a more complete account of those factors behind the demise of consensual norms on the Court. Although the Judges’ Bill, with a movement to a discretionary docket, is substantively important, it does not represent an immediate cataclysmic event. Instead, it is an influence that can be found in both the Hughes and Stone Courts, where the roots of this decline tend to be distinct, and one that continues into the subsequent Vinson and Warren Courts.
The altered political context stemming from the Great Depression and the subsequent New Deal realignment established a new American majoritarian alliance. From Dahl’s (1957) perspective, it would have been the Court’s task to use its discretionary docket to confer legitimacy on the new forms of economic regulation that were then emerging from the elected branches. However, a dated and highly experienced collection of associate justices composing Republican laissez-faire appointees held fast to established precedents that would nullify federal intervention within the economy. The Court-packing episode and the “switch in time that saved nine” (McKenna 2002) alleviated this salient interbranch conflict, but ideological polarization and distinct voting blocs (i.e., the Four Horseman versus FDR’s new appointees) more frequently placed Chief Justice Hughes in a minority position on these economic cases. As a leader facing threats to the Court’s legitimacy and norm of consensus, Hughes appears to have placed greater emphasis on the former, which conflicted with his duty of encouraging unanimity.
The fracturing of the majority coalition under Hughes previously was obscured by the use of opinion-based measures of dissent that placed the onus for declines in unanimity squarely on Chief Justice Stone’s lack of leadership skills. These results do confirm that Stone was unique, and potentially lacking, with respect to his predecessors. In fact, these results suggest that several factors could have engendered a return to equilibrium with respect to consensual decision making. A new series of FDR appointees acted to reduce levels of ideological polarization, and in time his justices constituted the Court. In contrast to the dissent-prone justice hypothesis, the addition of these inexperienced justices potentially would have curtailed the production of separate opinions, but it seems reasonable to expect that such an effect would have been predicated on stronger leadership from the center chair.
Nonetheless, other factors taking place in this period would contribute to dissensual opinions both under Stone and into the foreseeable future. The seeds that Stone himself planted in footnote 4 would come to fruition in the Court’s charting of a new course wherein it realized the potential of the discretionary docket by creating agenda space for civil liberties and criminal cases. The rise of novel, contentious civil rights and individual liberties cases against a backdrop of world war certainly contributed to the rise of dissenting behavior. But the call to arms that was footnote 4 meant that the Court initially would be systematically and incrementally ridding itself of economic disputes by settling the law and delegating them to others. As this process was being consummated, publicly salient individual rights cases drove new wedges into a collective of associate justices that was appointed primarily on criteria related to economic policy (Abraham 2008). Throughout the Stone, Vinson, and Warren Courts, then, an increasingly experienced Court established new equilibria for rates of dissent and concurrence.
The eventual death knell of the norm of consensus, however, seems to have taken place under the leadership of Chief Justice Burger, who too is seen as a lacking social and task leader. After controlling for agenda transformation, workload, and compositional explanations, the associate justice concurrence rate reveals yet another significant increase that represents the rise of the modern choral Court (Maveety, Turner, and Way 2010). Still, other factors continue to support these historically high levels of dissensus. The alternating series of Democratic and Republican Party appointees help maintain levels of ideological polarization that feeds dissent. In addition, a modern and more systematic vetting process for nominees, which places emphasis on candidates with prior judicial experience and clear decision-making track records, contributes to the individual expression found within modern dissenting and concurring coalitions.
As a consequence of being in the center chair, Stone principally has been held responsible for all of these institutional changes. We conclude that Chief Justice Stone was not a solitary source of dramatic changes in the dynamics of Supreme Court decision making. We do find evidence consistent with a lack of social leadership and task leadership, but these are perhaps no better or worse than those of Chief Justice Burger. He did, however, play a central part in this change. The number of unanimous decisions declined in part because of the increased discretion that the Court had over its docket, whereby consensual, easy cases could be relegated to the lower courts. The Court was increasingly consistent in its regulation and federalism decisions, paving the way for more attention to individual rights and liberties once the New Deal transition took hold. Ironically, Associate Justice Stone was the driving force behind the inception of the preferred position doctrine, which changed the dominant philosophy of the Court and helped transform its agenda. These new issues arising in a time of a hot war and a cold war created conditions that led to sustained growth in concurring and dissenting votes. We believe it is less Stone’s failure as a chief justice than his initiative in changing the paradigm of the Court that launched a thousand separate opinions.
Footnotes
Acknowledgements
We would like to express our gratitude to Sandra Wood, Linda Camp Keith, and Ayo Ogundele, who gathered original data utilized in this analysis, and to the anonymous reviewers who contributed to the project. Arthur Spirling and Jong Hee Park were instrumental in conducting the Bayesian changepoint analysis. Patricia Morgan of the Levin Legal Information Center and Chris Schmidt facilitated access to the Westlaw database.
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) received no financial support for the research, authorship, and/or publication of this article.