Collaborative Research: Individual-Level Analysis of Supreme Court Justices: A Modification to the United States Supreme Court Judicial Databases

点击翻译按钮获取中文摘要

Much is to be said for judicial databases that treat a court's decision as the unit of analysis-as is evidenced by the widespread use over the past twelve years of the two NSF-Supported U.S. Supreme Court Judicial Databases. Indeed, almost no empirical studies fail to avail themselves of the data contained in these two databases. However, neither allows for research pertaining to the behavior of the individual justices, as distinct from the aggregate analysis of the Court's decisions/outcomes. This shortcoming is a major limitation to understanding the behavior of the justices. Empirical studies of the behavior of the individual justices arguably provide the most fruitful route to understanding judicial behavior. With the databases so transformed, scholars will be able to efficiently investigate such theoretically important matters as individual-level analysis of voting behavior, models predicting and explaining dissent or concurrence, opinion assignment, as well as any other question that requires choice among several options using independent variables that vary both by justice and by case -- i.e., those amenable to the use of conditional logit analysis or hierarchical models.Recently formulated procedures allow the two databases to be "flipped" so that their focus becomes the justice rather than the case. Such transformations require systematic coding of the individual justice's voting and opinion behavior that does not pertain to the Court's decisions; i.e., the 4500 special opinions (dissents and concurrences) and the 11,000 votes represented by the opinions that the justices wrote and cast between 1946 and 1999. The result will be two complementary user-friendly databases, paralleling the existing ones, but increased in size by a factor of approximately ten, extending across a period from 1946 to the end of the currently existing Rehnquist Court.