Empirical Criminal Law Scholarship and the Shift to Institutions

Article excerpt

INTRODUCTION
I.   DESIGNING PUNISHMENT
II.  DESIGNS OF DECISIONMAKING
     A. Choosing Decisionmakers
     B. Structures of Cost Bearing in Decisionmaking
     C. Empirical Opportunism and Institutional Experimentation
III. STAGE PROCESSING AND THE LEGACY OF MCCLESKEY
CONCLUSION

INTRODUCTION

The broadly defined field of "criminal law" or "criminal justice" has always provided ripe material for empirical studies, but in assessing the most interesting new research from the Seventh Annual Conference on Empirical Legal Studies (CELS), it is useful to underscore the word "legal" and thereby demarcate a more specific category of research. Legal academics have long benefited from the venerable line of social science research on the causes of crime or changes in crime rates. This work, as recently exemplified by inquiries into the causes of America's dramatic post-1990 decline in crime, (1) usually involves social or other factors often exogenous to the legal system. (2) Another tradition of research involves the study of broad penal legislation--such as the death penalty (3) or gun laws (4)--where empiricists test the efficacy of laws in accomplishing goals that are uncontested as a matter of social policy or morality (e.g., protecting victims from violent crime). This research has tended to implicate empirical studies in volatile public and cultural debates at too high a level of generality (often with highly equivocal results) to bear on the operations of criminal justice systems. (5)

While acknowledging the roughness of some of these distinctions, (6) I will use some of the exemplary papers presented at CELS to describe and assess a trend toward more empirical "legal" research--papers that help demonstrate how statistical findings can inform government in the context of institutional decisions within the formal legal system. This research, focusing on the political economy of criminal justice agencies, examines legal decisionmaking under the constraints of constitutional demands, managerial and fiscal challenges, and, sometimes, the peculiar deontological forces that distinguish criminal law from other forms of regulation. Most notably, these studies examine the unavoidable discretion that the legal system invests in officials at key decision points, with special attention to prosecutors--the most important and unexamined of all categories of officialdom. They identify patterns and outcomes at early and low-level stages of adjudication, especially where choices of both means and ends implicate nonutilitarian values. In so doing, this research frames and informs questions for the officials who delegate, supervise, or exercise discretion, but it ultimately cannot answer the questions it poses. (7) Empirical research can provide what can be measured, but criminal law consists of more than just the measurable. At some point, research can do no more than arm decisionmakers with a sharpened sense of the scope and nature of the residual factors that empirics cannot measure with the reassurance of precision.

In this Essay, I will review, describe, and assess this trend through some CELS examples sorted into some of the key varieties of institutional empirical research about criminal justice. Part I considers how empiricists can inform the design of penal and sentencing systems through microanalysis. Such work takes advantage of quasi-experiments in the administration of criminal justice that exploit very specific legal changes under the methods of regression discontinuity, testing out marginal instances of crime reduction and efficiency, and cautiously generalizing upward. Part II considers research that evaluates (and can recommend) structural or bureaucratic changes in criminal justice administration, including selection of officials and allocation of cost bearing among agencies or levels of authority. Part III turns to the crucial phenomenon of official discretion as it is exercised at distinct and incremental stages of the criminal justice process. …