This dissertation seeks to determine if, and to what extent, federal judges behave in a partisan manner when deciding politically salient election law cases. Specifically, do judges favor the interests of their political party, after controlling for judicial policy preferences? The main hypothesis that I seek to test is the relationship between case votes and the interests of a judge’s political party in a given election law case. I posit that when the judge’s political party benefits from a ruling for the plaintiff/defendant, a judge will be more likely to rule for the plaintiff/defendant. I also test four additional hypotheses, all of which should moderate the relationship between partisan interests and case votes. I test the effects of political career experience, age, court of appeals membership, and partisan panel composition on the likelihood of a judge ruling in favor of her political party.
To test these hypotheses, I model the case votes of federal district court and court of appeals judges in campaign finance, political party right to association, and redistricting cases from 1962 through 2007. To control for a judge’s policy preferences, I impute first and second dimension common space scores for all federal judges in my dataset. Of the three categories of election law cases examined in this dissertation, only the campaign finance models consistently produce a statistically significant partisanship effect. There is also evidence of a conditional partisanship effect in redistricting cases, which is contingent on the partisan composition of a three-judge district court. Additionally, judicial policy preferences are statistically significant predictors of judicial behavior in political party right to association and redistricting cases, and campaign finance cases that do not involve the interests of the Democratic and Republican Parties. While most political science models of judicial behavior emphasize policy preferences or a combination of law and policy preferences to explain and predict judicial behavior, the results of this dissertation provide evidence that social groups could be another important, systematic influence on judicial behavior. Specifically, these results suggests that judges hold separate sets of preferences: preferences for preferred in-groups (such as a political party) and policy preferences. Therefore, judicial scholars should consider, when appropriate, the influence of personally salient social in-groups when modeling judicial behavior.
While the influence of partisan preferences in any form is troubling, the results demonstrate that the effect of partisanship is not pervasive in all areas of election law. As such, policy makers should decide if any semblance of partisanship is unacceptable in the federal courts before considering any potential reforms or changes to the judiciary.