"The wisdom we have lost": sentencing information and its uses.

• Author: Miller, Marc L.

Where is the wisdom we have lost in knowledge? Where is the knowledge we have lost in information?--T.S. Eliot (1)

INTRODUCTION I. 1984: A KNOWLEDGE-DRIVEN RULEMAKER II. USERS, USES, AND USABILITY OF DATA AND ANALYSIS A. The Commission and Congress B. Judges and Attorneys C. The Scientific Sentencing Community III. THE MISSING NATIONAL SENTENCING REFORM CONTEXT A. The National Goal of Standardized Data B. A National Sentencing Institute CONCLUSION INTRODUCTION

In the Sentencing Reform Act of 1984 (SRA) Congress envisioned federal sentencing with a technocratic cast, with policies designed and revised based on "advancement in knowledge of human behavior as it relates to the criminal justice process." (2) The value of data and expertise in the sentencing enterprise jumps off the pages of the statute. (3) Congress directed the U.S. Sentencing Commission to "establish a research and development [R&D] program" and to serve as a clearinghouse and information center for the collection, preparation, and dissemination of information on Federal sentencing practices." (4) One key activity of this sentencing R&D program was to "collect systematically" various forms of sentencing data and to publish those data. (5)

Who would use all of this sentencing information ? Most of the statutory provisions and the relevant legislative history point towards one primary user: the Sentencing Commission. Congress did not envision the Commission as the only user of the information, for the statute labeled the Commission as a "clearinghouse" and an "information center" and gave it data publication duties. Nevertheless, the statute includes specific directives to the Commission--and only to the Commission--about how it should use sentencing data. The external uses of the data stored in the "clearinghouse" remain unspecified, and the potential users of the data remain unnamed in the statute.

The last thirty-five years of sentencing reform have generated a lot of data. They have also generated a good bit of knowledge and, perhaps, a measure of wisdom. Both federal and state experiences in sentencing over the last three decades suggest that sentencing data and knowledge have the most impact--and most often lead to wisdom--when they are collected and analyzed with particular uses and users in mind.

Ironically, greater reliance on data and expertise can democratize the making and testing of sentencing policy. When data are collected and published with many different users in mind, they all can contribute as sentencing practices take shape. A variety of participants in the sentencing process can join the Commission as creators of sentencing wisdom, including Congress, state legislatures, state sentencing commissions, sentencing judges, appellate judges, prosecutors, defense attorneys, probation officers, and scholars. One of the central lessons learned from thirty-five years of structured-sentencing law is that input from a broader range of participants makes a system more durable and balanced. (6)

We believe that Congress can improve the federal sentencing system by directing the U.S. Sentencing Commission to provide better and more timely information and to link that information explicitly to a broader range of specified users and uses. Because rulemaking and research may have become incompatible tasks in the federal sentencing context, perhaps Congress should separate these functions and transfer the responsibilities for national data collection, dissemination, and research to a separate National Sentencing Institute, ideally to be located in the judicial branch. Such a separation of functions would parallel the separation of federal prosecution policy and federal crime data collection and distribution into different units within the U.S. Department of Justice.

An expanding range of uses for sentencing data also has implications for the sources of those data and analyses. The Commission has been too parochial about sentencing data and research. Over the past thirty-five years dozens of structured-sentencing experiments have emerged at the state level throughout the United States. Yet these experiments have been isolated from each other, and from the federal system, so that lessons drawn from the states' experiences have spread too slowly. As in many areas of criminal justice (as with key aspects of public health, the environment, and the economy), there is a critical federal role to play in creating national knowledge. That federal role often takes the form of encouraging the collection of comparable data from state actors. Although no single state has the incentive to pay for standardized collection, the widespread benefits to all jurisdictions from such efforts make this an ideal target for modest federal funding.

1. 1984: A KNOWLEDGE-DRIVEN RULEMAKER

   The marching orders that Congress gave to the Commission in 1984 were quite detailed and hopeful about the role of knowledge in sentencing. The SRA instructed the Commission to base its new sentencing guidelines and policy statements on specialized knowledge. The statute directed the Commission to ascertain average sentences imposed and served under old federal law and to use these data as a "starting point" in the creation of the Federal Sentencing Guidelines. (7) Once the Guidelines were drafted, the SRA called for the Commission to "develop means of measuring the degree to which the sentencing, penal, and correctional practices are effective in meeting the purposes of sentencing." (8)

   The Commission never made much headway on this statutory duty to measure the effectiveness of sentences, but it did faithfully carry out its ongoing data reporting chores. The Commission and its staff now create valuable annual statistical reports about federal sentencing practices, (9) and they regularly deliver more detailed case-level sentencing data sets to an interuniversity data consortium that is nominally accessible to the public, though in practice prohibitively expensive and difficult to use. (10) The Commission staff also evaluates high-profile sentencing practices from time to time. (11)

   The statute goes beyond the collection and publication of data, giving a prominent role to experts and to system participants during the Commission's periodic review and revision of the existing Guidelines. The Commission is to formulate any needed revisions in light of "comments and data coming to its attention," after consulting "authorities on, and individual and institutional representatives of, various aspects of the Federal criminal justice system." (12)

   This vision of a Sentencing Commission that "consults" with authorities to create sentencing rules that reflect knowledge and experience has several distinct components. Some aspects of a genuine consultation have developed well over the last fifteen years, while others have not.

   One aspect of a successful consultation is the willingness of an agency to receive comments and other input from interested parties, particularly those with specialized knowledge. (13) On this score, the Commission has performed reasonably well. It regularly receives commentary from advisory groups already in existence, including the Judicial Conference of the United States, various lawyer associations, defense organizations, business groups, and other organizations. The Commission also created its own "Practitioners' Advisory Group" to comment on proposed changes to the Guidelines. (14)

   It is not enough, however, for an agency merely to receive comments when they are offered. A healthy consultation between an agency and those on the outside also requires the agency to respond to the input. It must demonstrate to those who provide comments that their input routinely makes a difference. (15)

   The Sentencing Commission can claim less success as a listener and respondent, and a leading example is its grudging response to feedback from judges. The Commission surveyed the views of judges about the operation of the Guidelines but then failed to respond concretely to criticisms that appeared in the survey responses. When judges make suggestions to the Commission in dicta of written opinions deciding a single case, they have no assurance that they are heard, for the Commission does not systematically collect or analyze these missives.

   Finally, a fruitful consultation between an agency and outside experts requires the agency to explain the basis for its decision in enough detail so that an observer can determine whether the agency accounted for the available data and listened to the available expert advice. (16) This is the most disappointing aspect of the Commission's consulting function over the years. While the Commission has been a busy agency, producing 674 amendments and many reports over the years, it has developed weak habits of explanation and justification. (17) In particular, the Commission has failed to explain rules or amendments in light of research findings. (18)

   The Commission's peremptory style made its decisions difficult to accept in a judicial world where legitimacy of a decision depends on the quality of explanation for the decision. When the Commission made it a priority early in its existence to settle "circuit splits" during its amendment process, (19) it acted like a Supreme Court for Sentencing, but without issuing opinions or reasons. (20)

   The Sentencing Commission should go beyond accepting comments on its own proposals. Under a reasonable and good faith reading of the SRA, the Commission should have provided the public with relevant data in a format that makes it possible for outsiders to ask their own questions and to make their own proposals. (21) The obligation to collect sentencing data and make them available might have encouraged a knowledge-driven Commission through tests of Commission hypotheses, the promulgation of competing theories, and policy and scholarly debate. In this conception of the role of data, the scientific method would be...