At the heart of attitudinal and strategic explanations of judicial behavior is the assumption that justices have well-defined policy preferences. To test implications of either explanation, researchers need measures of the justices' underlying policy preferences. In the literature these preferences have been measured in a handful of ways, including looking at past votes in a single policy area (Epstein, Walker, and Dixon 1989), content-analyzing newspaper editorials at the time of appointment to the Court (Segal and Cover 1989), and recording the background characteristics of the justices (Tate and Handberg 1991). Scholars have used these measures succesfully to explain behavior in some policy areas, including civil rights, but have been less successful in others, such as economics cases (Epstein and Mershon 1996).
In this paper we posit a simple formal-theoretic model of voting on the Court that we use to specify a Bayesian measurement model of judicial policy preferences. In so doing, we simultaneously estimate a bliss point for each justice in a multidimesional space and the cutting hyperplane for each non-unanimous case. The Bayesian approach also allows us to include background information about the justices' ideal points in a straightforward manner, and allows us to gauge the uncertainty of each estimate by summarizing the posterior density. We employ Bayes factors (Kass and Raftery 1995) to gauge the dimensionality of the issue space, and to test various model specifications. With our measures in hand, we further explore: (1) the reliability and validity of the measures; (2) how these estimates compare to extant preference measures; (3) the dimensionality of the Supreme Court issue space; and (4) how substantive issues, coded by Spaeth (1999), map into this multidimensional space.
