Interparty Judicial Appointments

Empirical studies of judges’ ideological voting call for a theory according to which the ideology of the judges can be measured. This article calls into question the assumption that undergirds the measure that currently dominates the legal, economic, and political science literature—the assumption that the ideology of a lower federal court judge is largely predicted by the ideologies of the nominating president and the relevant state's senators who are of the same political party as the president. The article relies on a natural experiment to examine this question empirically. Between 1977 and 1998, New York was represented in the Senate by one Democrat and one Republican who had an agreement to divide appointments to the district courts in the state: the senator who shared party affiliation with the president would be allocated three of every four appointments, while the “out‐of‐party” senator would be allocated the rest. The article employs a novel data set—consisting of all federal district judges appointed to the federal bench in New York during the time period in question, and the senator who recommended each nominee to the nominating president. If the dominant theory—that the party of the recommending senator affects judicial decision making—holds, then one would expect the theory's explanatory power to be at its apex where senators of different parties recommend judges at the same time to the same president. Yet, using median prison sentence length as a proxy for ideology in decision making, the empirical analysis finds no evidence that senatorial ideology has a statistically significant effect on district judge decision making. At the same time, it finds that, indeed, the nominating president's ideology does have a statistically significant effect. The findings are instead consistent with the minority view of lower federal court judges’ ideological leanings—that a lower federal court judge's ideology is in large part a function solely of the nominating president's ideology.