ABSTRACT-The fact that a substantial percentage of Supreme Court decisions are unanimous is often used to undermine the theory that the Court's decisions are ideologically driven. We argue that if the ideological stakes in a case are small, even slight dissent aversion is likely to produce a unanimous decision. The data support this interpretation but also establish the existence of an ideological effect in unanimous decisions. These findings are consistent with a realistic conception of the Court as a mixed ideological-legalistic judicial institution.
The Supreme Court is widely regarded, and not only by political scientists, as a highly political Court, an impression often reinforced by the ferocious dissents of one or more Justices in many cases. This makes the Court's unanimous decisions an important subject for study-for how is unanimity achieved in a political court?-and it is a subject that has been neglected.1 This Essay seeks to advance understanding of the phenomenon of Supreme Court unanimity. Despite the title of the Symposium, this is not an essay about Justice Stevens as such, but it is not irrelevant to his service on the Court. He was noted for frequent dissents.2 Of the Justices who served in the 2005 through 2009 Terms3 (the years of the Roberts Court before Justice Stevens retired), he had the highest dissent rate-30.3% of the 320 cases in which he participated (Justice Souter was second, at 27.2%-69/254).4 Yet he voted more often to join a unanimous decision than he voted to dissent-39.1% of the votes he cast were in unanimous decisions (125/320). Whether this should be thought paradoxical is one of the questions we try to answer in this Essay.
The Supreme Court Database (also known as the Spaeth Database) contains data on cases decided since the 1946 Term (the last Term we study is the 2009 Term). We define unanimous decisions as ones in which no Justice dissented, even if there were also one or more concurring opinions. Concurring opinions are actually more common in unanimous decisions than in non-unanimous ones-41% of the unanimous decisions in The Supreme Court Database include concurring opinions, compared to 38% for non-unanimous decisions-although the reason may be arithmetical: in a 5- 4 decision, the maximum number of concurring opinions is four; in a 9-0 decision it is eight. An alternative definition of a unanimous decision- which would be more realistic in recognizing that concurring opinions often indicate disagreement with, rather than merely supplementation or extension of, the majority opinion-would be a decision in which all the Justices joined the majority opinion, whether or not any of them also wrote a concurring opinion. We have not broken down the data sufficiently to enable us to analyze unanimous decisions so defined.
Except in the 1969 Term, in which there were only eight Justices, we exclude unanimous cases in which one or more Justice was absent; the reason is that we cannot be certain that if a ninth Justice had sat, the case would still have been decided unanimously.
Under our definition, about 30% of the Court's orally argued decisions from 1946-2009 were decided unanimously. Figure 1 shows that the percentage has been trending upward: from 21% in 1946-1952 (the Vinson Court) to 34% since 2005 (the Roberts Court). The graph shows that part of the upward trend is the result of an increasing fraction of unanimous decisions reversing the Ninth Circuit.5 Over the entire period, about 73% of Ninth Circuit decisions that the Supreme Court reviewed were reversed unanimously, compared to 63% for the other circuits. If we exclude them, the upward trend falls from 1.35% to 1.07% per year.6 Even if the decisions reversing the Ninth Circuit are included, there is no significant positive trend over the last twenty-five years (since Rehnquist's first Term as Chief Justice). The percentage peaked in 1997 at 51.1% (40.2% if excluding Ninth Circuit reversals) and has since declined to 37%. …