The Supreme Court
In chapter 1, we offer a generalized judicial utility function. For Supreme Court Justices, the function is simpler because Justices cannot be promoted to a higher court, few Justices have any realistic expectation of being promoted to Chief Justice, and the large size and quality of sta. relative to the number of Justices and the Court's light caseload in recent years imply that a Justice's leisure activities and nonjudicial work activities are not significantly constrained by his or her judicial duties.
Assuming that the time devoted to judicial work) is a more or less fixed number based on the Justice's energy, work habits, and devotion to the job, the principal tradeoff a Justice makes is between time devoted to leisure activities and time devoted to nonjudicial activities that enhance the external satisfactions of his judicial job—prestige, power, influence, and celebrity.
We can expect both the internal and external satisfactions of the judgeship to be strongly influenced by a Justice's ideological leanings. The Supreme Court's principal role is to resolve large political controversies, and a Justice is likely to derive personal satisfaction, as well as reap prestige, exert power and influence, and achieve celebrity, from attempting to align the law with his ideological commitments; in contrast, caseload pressures, the threat of reversal or eventual overruling (and so of not having the last word), desire for promotion, a different case mix, and lower visibility combine to dampen the ideological ambitions of lower court judges.
The influence of a Justice's ideology on his voting behavior on the Supreme Court is therefore the principal focus of this chapter. After a brief description of our data and of problems with them (also discussed in the appendix to the chapter), we use a variety of measures of ideology to determine the correlation between a Justice's judicial voting and his or her ideological leanings. We find strong evidence that ideology does influence the Justices' judicial votes, and thus the Court's outcomes, in a variety of cases, and that this ideological influence has been growing.
Turning next to changes in Justices' ideology over the course of their service on the Court, we distinguish between ideological drift—changes relative to the Justice's ideology when appointed—and ideological divergence— changes relative to the ideology of the appointing President. Such a study may seem more relevant to the behavior of the appointing authorities (for example, does the threat of drift or divergence influence the weight that a President will give to age in deciding whom to nominate for a position on the Court?) than of the Justices. But drift and divergence are behaviors of Supreme Court Justices, and we relate these behaviors to our model of the judicial worker.
The chapter then forks into separate analyses of unanimous and non-unanimous decisions in the Supreme Court. The former are thought to challenge the realist theory of judicial behavior. But the Supreme Court has been ideologically divided throughout the period of our study—how then can it be unanimous in more than 30 percent of its cases? We show that unanimity is indeed inconsistent with an extreme realist model of the Justices' behavior. We trace the inconsistency to a combination of legalist commitment, absence of high ideological stakes in many Supreme Court cases, and modest effort aversion, which leads to reluctance to dissent at the drop of a hat. We find stronger ideological influence in non-unanimous decisions, but also moderating influences, such as a Justice's prior socialization as a judge of a lower appellate court (though the evidence of this is weak). We end with a discussion of group influences. In a non-unanimous decision, there are two groups of Justices: the majority and the minority, though the minority group may have only a single member. We inquire whether changes in the relative size of groups have an independent influence on Justices' votes and find they do not. We do find such influences in the courts of appeals, as discussed in Chapter 4.