The politics of constitutional review: Evidence from the European Court of Justice.

摘要

Judges who perform judicial review have the extraordinary power to strike down laws that do not conform to their own policy preferences. Their political independence is generally regarded as a normative good. In this work, I consider the microfoundations of judicial preferences and how those preferences interact with institutional independence to determine the policy impact of judicial review.;The following argument is developed in the context of the Court of Justice of the European Union (European Court of Justice, or ECJ). Constitutional Courts generally and the ECJ in particular are considered "independent" when they enjoy discretion to act counter to the interests of other policymaking bodies and their political principals. In the European Union, the primary political actors are the member states, which directly appoint the judges and play a significant role in the legislative process. But whether their independence implies policy outcomes that exploit this discretion depends on the preferences of the judges --- which may or may not diverge from those of the principals. Indeed, in standard theories of delegation, broad discretion is likely to be granted when policy preferences of principal and agent align.;A considerably body of scholarly work has asserted that the ECJ's institutional independence has implied behavioral independence: in short, that the ECJ has pursued a prointegration agenda counter to member state governments' preferences. The typical explanation for this apparently independent behavior is that judges share a common preference for expanding the authority of the Court and EU generally. The claim that individual ECJ judges share a uniform preference for integration has never been tested due to the institutional cover of the court's collective decisions, by thwarting scholars' efforts to evaluate individual judicial behavior. But while individual judges' behavior is not directly observable, a feature of the organization of the ECJ, its system of Chambers, provides a potentially valuable window on judges' decisions. Specifically, I content that we can infer individual behavior from the collective judgments made in chambers because most judgments are made by different combinations of judges. I develop a general statistical model for aggregate data produced by subsets of deciders by extending the item-response model to account for selective participation in decisions. In addition, I explicitly model other known features of the parameters of the item-response model to enable inference about both judges and cases.;Results show that judges do not share a common preference for integration -- that institutional independence has provided cover not only for Europhiles but Euroskeptics as well, contrary to the claim that ECJ judges all share a motivation for more integration. In fact, the heterogeneous preferences of judges are predictable based on the preferences of the member states at the time of appointment. Extant results about judges' responsiveness to member state governments is confirmed with microfoundations, and extended to allow for selective responses. I show that institutional independence does not imply behavioral independence -- much less, the behavioral independence that ECJ scholars have assumed the ECJ has engaged in, pushing for greater integration. Instead, because judges are institutionally shielded, member state governments appear to appoint judges with preferences similar to their own.