Variable Voting Behavior on the Supreme Court: A Preliminary Analysis and Research Framework*
Collins, Paul M., Jr., Justice System Journal
On June 26, 2003, the Supreme Court ruled in Lawrence v. Texas that a Texas statute, making it a crime for homosexuals to engage in "deviate sexual intercourse," was an unconstitutional violation of the Fourteenth Amendment's Due Process Clause.1 In so doing, the Court struck down the 1986 precedent of Bowers v. Hardwick.2 Writing for the Court, justice Kennedy explained, "The doctrine of stare decisis is essential to the respect accorded to the judgments of the Court and to the stability of the law. It is not, however, an inexorable command."3 Five justices supported overruling Bowers (Breyer, Ginsburg, Kennedy, Souter, and Stevens). Three justices, in dissent, argued that Bowers should remain good law (Rehnquist, Scalia, and Thomas). The lone concurrer, justice O'Connor, agreed with the judgment of the Court (that the Texas statute was unconstitutional), but disagreed with the overruling of Bowers. all of the justices' opinions spoke to the need for stability in the law, a concept not lost on academics and practitioners.
Both attitudinal (e.g., Segal and Spaeth, 1993, 2002) and legal (e.g., Dworkin, 1978) approaches to Supreme Court decision making assume that the justices' voting behavior is rational and stable.4 This is perhaps no more evident than in studies of the justices' voting behavior in precedent-setting and precedent-overruling cases (e.g., Brenner and Spaeth, 1995; Segal and Spaeth, 1996; Spaeth and Segal, 1999).5 From the attitudinal perspective, justices are assumed to reveal their preferences in precedent-setting cases and simply apply these preferences to subsequent cases. From the legal perspective, justices are expected to announce the rule of law that governs the precedent-setting case and apply it to the progeny (thus upholding the precedent).6 The purpose of this note is to examine how often the justices' voting behavior fits these expectations in precedent-setting and precedent-overruling cases and create a research framework for better understanding variable voting behavior in such cases (e.g., that behavior explained neither by the attitudinal or legal models).
To be sure, the Supreme Court rarely overrules itself. During the tenure of the Warren, Burger, and Rehnquist Courts, the Court overruled past decisions in less than 2 percent of cases heard (Reddick and Benesh, 2000). But such cases are of import precisely because of the rarity with which they occur. By overruling itself, the Supreme Court violates the doctrine of stare decisis, the canon that the Court has a duty to follow its own precedents. While scholars and students of the Court are well aware of the infrequency with which cases are overruled, little attention has been paid to examining the behavior of individual justices in these cases. Further, those studies that have examined the behavior of the justices have done so by focusing primarily on the frequency with which justices exhibit behavior that is explainable by either the legal or the attitudinal models (e.g., Brenner and Spaeth, 1995; Segal and Spaeth, 1996; Spaeth and Segal, 1999). Thus, little attention has been paid to situations in which individual justices change their views toward cases they initially supported: those occasions where justices vote to set a precedent, but then vote to overrule the very precedent they helped establish. This study attempts to address these occurrences by creating a framework for better understanding the choices justices make in precedent-setting and precedent-overruling cases.
VOTING BEHAVIOR IN PARENT AND PROGENY caseS
Justices who participate in precedent-setting cases, parents, and also in cases that reconsider the viability of precedents, progenies, have two choices available to them in both the parent and the progeny (see Figure 1). In the parent case, justices may choose to vote either in support of or against the establishment of precedent. In the progeny, justices may choose to vote to uphold or overturn the precedent-setting case. …