Vol. 6,No. 3 (March, 1996) pp. 40-44

STARE INDECISIS: THE ALTERATION OF PRECEDENT ON THE SUPREME COURT, 1946-1992 by Saul Brenner and Harold J. Spaeth. Cambridge University Press. 1995.

Reviewed by Donald R. Songer, Department of Government and International Studies, University of South Carolina.

How the Supreme court reaches its decisions has been the focus of scholarly interest for decades. Although political scientists in the empirical tradition have focused on the influence of the justices' attitudes and political preferences on their behavior, a recent spate of articles and books has reflected a resurgence of scholarly interest in the impact of precedent and legal doctrine on Supreme Court justices' decisions.

STARE INDECISIS is the latest attempt to shed new light on this controversy by examining the Court's decisions to overrule its own precedents. As the authors put it, the most important part of their study is an attempt to answer the question, "Why do the justices in the overturning and overturned cases vote the way they do ?" To answer this question, Brenner and Spaeth are particularly concerned with assessing the relative contribution to judicial decisions of the attitudinal model and one important element of the legal model, adherence to stare decisis. The cases examined in this study are an important, although the authors recognize, not a representative sample of the Supreme Court's decisions over a 46 year period. Specifically, the analysis seeks to explain the voting behavior of the justices in all of the cases from the Vinson, Warren, Burger, and the pre-1993 Rehnquist Court in which the Court formally overturned precedent. While these cases constitute a small fraction of the total workload of the Court, they are arguably some of the most important cases from the perspective of the impact of the Court on policy making.

The central thesis of the book is essentially the same as the thesis of THE SUPREME COURT AND THE ATTITUDINAL MODEL: "personal stare decisis" (which is to say the personal policy preferences of the justices) rather than "institutional stare decisis" or precedent is the most important influence on the behavior of a large majority of the justices in the cases examined in the book. Since the authors recognize that the cases they examine are not typical of the agenda of the Court they are careful not to generalize from their findings to all decision making by the justices. However, they suggest that the findings represent an important addition to what they see as a growing body of evidence in support of the argument that precedent exerts little restraint on the decisions of the justices.

Brenner and Spaeth establish the importance of judicial attitudes for their votes on cases that overturn precedent in a straightforward and convincing manner. First, a Guttman scale analysis of the justices votes in each of the four courts examined demonstrates that over 97% of the votes were consistent with the attitudinal model. Additionally, analysis of the ideological direction of the decisions to overturn precedent is consistent with the consensus on the ideological differences among the courts studied. In particular, the precedents overturned by the liberal Warren Court were overwhelmingly conservative while

Page 41 follows:

precedents overturned by the Rehnquist Court were overwhelmingly liberal.

More controversial is the Brenner and Spaeth analysis that leads to their conclusion that the legal model has little significance for the decisions to overturn precedent. To test the legal model, the justices votes are divided into the four possible patterns:
1)A justice may vote with the majority in the overruled case and dissent from the later decision overruling precedent.
2) A justice may dissent in the overruled case and vote with the majority in the later decision overruling precedent.
3) A justice may vote with the majority in both the overruled case and the later decision overruling precedent.
4) A justice may dissent in both the overruled case and the later decision overruling precedent.

Brenner and Spaeth argue that the first scenario is consistent with both the legal model and the attitudinal model, pattern #2 is consistent only with the attitudinal model, pattern #3 is not consistent with either the legal or the attitudinal model, and pattern #4 is consistent only with the legal model. Examining the empirical occurrence of each of these patterns yields the conclusion that "all four Courts manifested more support for personal stare decisis than for the legal model" and "This record hardly betokens attachment to the legal model" (85). Most significant is their finding that only 2.7% of the votes fall into pattern #4 which is consistent with support for precedent but inconsistent with the attitudinal model. Most of the votes in support of precedent fall into pattern #1 which the authors note can be explained equally well by the attitudinal Model.

An attempt to assess the significance of these findings should explore two questions. First, does this study adequately support the conclusion that precedent is of little consequence for the particular set of cases analyzed ? And second, what is the significance of these findings for the larger question about the overall impact of precedent on the Supreme Court ?

The answer to the first question depends on how one conceptualizes the legal model. Most empirical tests of the attitudinal model have implicitly (or sometimes explicitly) contrasted the attitudinal model to a legal model based on mechanical jurisprudence. As a corollary to this view, judges are supposed to face a dichotomous choice between the clear mandates of precedent and their attitudinal preferences. If one understands the legal model in this way, then the argument of STARE INDECISIS is compelling. The evidence presented by Brenner and Spaeth clearly establishes that judicial attitudes have a greater impact on decisions than a legal model so conceived. But Stefanie Lindquist and I have argued elsewhere (e.g., see our article in the forthcoming AJPS symposium) that the effects of precedent are more subtle; that precedent may shape and constrain judicial choice without necessarily providing a single clear answer to every legal question. Given this conception of the legal model it is reasonable to expect that in many cases the

Page 42 follows:

position adopted by a justice will reflect the influence of both precedent and their personal attitudes. From this perspective, the data presented by Brenner and Spaeth suggest that precedent may have influenced (without being the sole determinant of) more than 40% of the votes even in these cases to overrule precedent. But the impact of precedent may be even greater. The least satisfying aspect of the Brenner and Spaeth analysis is their treatment of the votes of justices who voted with the majority in the original precedent and then later voted with the majority that overruled the precedent. The authors note briefly that such cases appear to be inconsistent with both the legal model and the attitudinal model. The fact that 28% of all the votes fall into this category is used as a key part of the evidence against the legal model. However, the implications of this finding for the attitudinal model are not adequately explored and there is little examination of the nature of the cases that fall into this puzzling category. If the vote of a justice for the original precedent reflected his or her policy preference, it is hard to rationalize a later vote to overrule the precedent as consistent with the attitudinal model (unless the ideology of the justice changed in the intervening years). From a POLICY point of view, the two positions are inconsistent even if the VOTES of the justice scale in both cases. On the other hand (as both Lindquist and Haire pointed out in the APSA round table on STARE INDECISIS), it is possible that a justice in the original majority would vote to overturn the precedent later for jurisprudential reasons that are consistent with the legal model. For example, if anomalies have arisen under the existing precedent, such that contradictions have been created as either the Supreme Court or the lower courts have attempted to apply the law to new fact situations, the justices may feel it necessary to eliminate the offending precedent in order to correct these anomalies. Thus, some portion of the 28% of cases falling in pattern #3 may also reflect some influence of the legal model. Without a detailed analysis of the published opinions and the comments of the judges during conference and opinion writing in these majority-majority cases it is impossible to know how frequently such jurisprudential concerns motivated these seemingly paradoxical voting patterns. If a substantial proportion of them do reflect such jurisprudential concerns, then the legal model may actually have influenced a majority of the votes in these cases in which precedent was overturned.

Second, what is the significance of these findings for the larger question about the overall impact of precedent on the Supreme Court ? Do these findings seriously undermine the validity of the legal model as a general explanation of Supreme Court decision making ? I think the answer is clearly no. First, one would expect that decisions overruling precedent should provide the clearest support of all cases for the attitudinal model. Given this expectation, the finding that over 30% of even these cases fall into Patterns # 3 and #4 that appear to be inconsistent with the attitudinal model is not strong evidence that the attitudinal model provides a sufficient general explanation of decision making. Second, as Brenner and Spaeth admit, it is inappropriate to generalize from such a small, unrepresentative proportion of the Court's docket. More importantly, one of the most interesting conclusions of STARE INDECISIS is that the

Page 43 follows:

justices are able to reach "striking" agreement about what the precedent means and that the continued existence of the precedent presents a roadblock to the resolution of the pending case (and presumably other cases in the future) in the manner they would prefer. Why should precedent present a roadblock if precedent is completely malleable as the attitudinalists proclaim ? In contrast, a critical assumption of those that argue that precedent constrains the justices is that it is possible in some cases for the justices to reach inter-subjective agreement about what precedent means within the context of a specific set of case facts. In the cases studied in STARE INDECISIS the justices vote to overturn the precedent after they agree on its meaning. But the discovery of this agreement makes it plausible that the justices are able to agree on the meaning of precedent for other cases as well. If so, a critical question would be how often the justices follow precedent once they have agreed on what it means. This question is not addressed by Brenner and Spaeth except to note that it is difficult to determine how often the justices may have agreed on the meaning of precedent and decided not to overturn it. This is a question that is worthy of future attempts at systematic investigation so that the findings of STARE INDECISIS can be seen in better perspective. On the surface, however, two categories of decisions would appear to be examples of cases in which the Court agreed on the meaning of precedent and decided to support the precedent rather than overturn it: 1) formally decided unanimous decisions in which the Court reversed or vacated the decision below, and 2) summary decisions in which the decision below is reversed or vacated "in light of" an explicitly identified precedent. These cases would appear to occupy the opposite extreme from the cases analyzed in STARE INDECISIS in terms of support for the legal model versus the attitudinal model. Significantly, however, there appear to be many more cases in these latter categories than in the category of cases formally overturning precedent. The remainder of the docket of the Court may be composed of cases in which it is precisely the meaning of precedent that is fundamentally at issue. For these cases, neither the findings of STARE INDECISIS nor predictions based on a model of mechanical jurisprudence are particularly relevant . For these "in-between" cases that probably compose a significant majority of all formally decided cases, further research is needed to explore the possibility that outcomes are shaped by both precedent and attitudes while being determined by neither.

While the central theoretical argument of STARE INDECISIS centers on the superiority of the attitudinal model to the legal model as an explanation of Supreme Court decision making, even students of the Court who are unpersuaded by the authors' thesis will find many reasons to add this book to their personal libraries. First, it is a delight to read. The writing is lucid, lively, and few words are wasted conveying a lot of interesting information. Even the 30 pages of appendices are worth examining. In addition, this short work will be an indispensable aid to any other scholars who do subsequent work on Supreme Court decisions that overturn precedent. In fact, it should be the first work read. The chapter on previous empirical research is so well done that future scholars may be tempted to skip going back to the original sources after reading the eight pages

Page 44 follows:

in which Brenner and Spaeth tie it all together. Moreover, the identification and explication of the cases in which precedent was overruled will prove to be a valuable resource for innumerable projects. As the authors demonstrate through a thorough and careful analysis, the questions of how and what to count as an instance of precedent overruled is not as simple as it may seem on the surface, and several earlier studies suffer from reliance on an inadequate identification of the relevant cases. The method employed by Brenner and Spaeth convincingly overcomes the shortcomings of prior studies and results in the production of a list of cases for analysis that will be accepted as the standard for future analyses.

STARE INDECISIS also follows in the tradition of earlier works by Segal and Spaeth and Rohde and Spaeth by including a set of introductory chapters that provide a wealth of useful descriptive material. Anyone who wants to know when, how often, and how the Court overruled precedent or what types of precedent were reversed after how many years and by which margins will find a wealth of material that is not readily available from other convenient sources. Of particular interest is the chapter on the conference votes of the justices. While a number of analyses of conference voting have been produced from recent examinations of the papers of the justices (including a number of notable earlier works by Brenner himself), I believe that this is the first time that a thorough analysis of conference votes has been so well integrated into an empirical study of the final votes on the merits. Adequate empirical description is an essential precondition for scientific analysis and advances in theory, and those on all sides of the theoretical controversies addressed by this book will benefit from the thorough and systematic description of Court decision making presented.


Songer, Donald R. And Stefanie A. Lindquist. Forthcoming. "Not the Whole Story: The Impact of Justices' Values on Supreme Court Decision Making." AMERICAN JOURNAL OF POLITICAL SCIENCE.

Copyright 1996