"How Not to Lie with Judicial Votes: Misconceptions, Measurement, and Models"

In Part I, we describe the formal spatial theory often invoked to justify the statistical approach. While spatial theory has the nice feature of synthesizing theory and empirics, legal scholars may remain skeptical of its strong assumptions. Fortunately, measurement models can be illuminating even if the spatial theory is questionable. illustrate this, Part II provides a nontechnical overview of the intuition behind measurement models that take merits votes as an input and return a summary score of Justice-specific behavior as an output. Such scores provide clear and intuitive descriptive summaries of differences in judicial voting. abounds, however, and in Part III we clarify prevailing misconceptions of such scores. We discuss how these scores relate to "ideology, " explain how such models grapple with the complexity and dimensionality of judicial decisionmaking, illustrate the problems of intertemporal extrapolation and cardinal interpretation of the scores, and highlight other common abuses of such measures. Part IV, we demonstrate how modern measurement methods are useful precisely because they empower meaningful examination, data collection, and incorporation of doctrine and jurisprudence. We argue that existing uses are simply a special case of a much more general measurement approach that works synergistically with the qualitative study of case law. We demonstrate in Part V how such measurement approaches-when augmented with jurisprudentially meaningful data-----can advance our understanding of courts, with case studies of the constitutional revolution of 1937, the dimensionality of the Supreme Court, the historical origins of the standing doctrine, statutory interpretation, and backlash against Supreme Court opinions. We conclude with thoughts on the chief virtues of model-based measurement and the study of law.