Evaluating Data on Corporate Sentencing: How Reliable are the U.S Sentencing Commission's Data?

Abstract:     
During a recent study of how 1991 federal sentencing guidelines have affected the penalties that federal courts impose on public corporations, we performed an independent evaluation of the quality of the data on corporate sanctions that the U.S. Sentencing Commission releases to the public. Our initial findings led us to use other, independently-compiled data for our own research. This paper presents the main findings of our evaluation, which focused on the quality of the Commission's 1988-1996 (ICPSR) data on public corporations. First, the Commission's post-Guidelines data on penalties for public corporations appear to be incomplete and non-representative of the underlying case population. For example, the ICPSR post-1991 data appear to exclude a disproportionate number of large fines imposed on public corporations. No similar difficulties in the ICPSR pre-Guidelines (1988-1989) data were found. Shortfalls in the post-Guidelines data on other kinds of defendants, such as individuals, appear to be less marked. Second, the Commission's data are missing variables that may explain a substantial part of the case-by-case variation that occurs in sentencing. The data reveal little about the harm caused by the offense, which is often estimated in court papers. Also missing is information about the identity of the sentencing judge and about the identity of the corporation being sentenced. We review the history of the Commission's efforts to collect data on federal sentencing, highlighting institutional constraints and other factors that appear relevant to the difficulties we have found in the data that the Commission releases to the public.

Number of Pages in PDF File: 23

Accepted Paper Series