The Courts of Appeals
This is the book's longest chapter (in part because we test the same hypotheses on several ex post databases—that is, databases of judicial votes or decisions), so as with the preceding chapter we begin by sketching its organization.
After discussing the principal data that we'll be using (again supplemented at the end of the chapter, this time by two appendices), we apply the same methodology for determining ideological influence on judicial behavior that we used in the preceding chapter, and we tentatively find significantly less ideological influence in the courts of appeals than in the Supreme Court. We then supplement that methodology with a new one measuring the ideology of judges at the time of their appointment—a methodology that uses information available online rather than just in newspaper editorials—and comparing that to the ideology reflected in their judicial votes. This methodology yields substantial differences between ex ante and ex post ideology: some conservative judges, as one would assess on the basis of their careers and statements up to the time of their appointment, turn out to be liberal judicial voters, and vice versa. We argue that this is more plausibly explained by the force of legalist commitment at the court of appeals level than by ideological drift.
The last section of the chapter deals with group effects. We find a significant conformity effect (a tendency of judges in the minority to go along with judges in the majority) and significant panel composition effects, but no group polarization or political polarization.