Why Federal Prosecutors Charge: A Comparison of Federal and New York State Arson and Robbery Filings, 2006-2010

Academic, judges, lobbyists, special interest groups, and the defense bar all love to complain about the undue discretion held by federal prosecutors. Criticism has intensified over the last few decades, as the federal criminal code has grown to more than 4,500 prohibitions, a fair number of which replicate nearly identical state offenses. Little empirical evidence, however, attempts to discern what, if anything, is distinctive about the cases charged in federal rather than state court, and what might be motivating federal prosecutors to make their charging decisions. Our study aims to shed some light on this subject. In Part II, we describe our efforts to collect data on the characteristics of cases prosecuted under arson and robbery statutes from three sources: (1) the United States Sentencing Commission (“USSC”); (2) the New York State Division of Criminal Justice Services (“DCJS”); and (3) Federal Bureau of Investigation Uniform Crime Reports.In Part III, we explain how we combined the USSC and New York State DCJS data before proceeding to our empirical analysis. First, we conduct a simple, bivariate analysis comparing the frequency with which our independent variables are observed in federal versus state arson and robbery cases. We note where we believe the observed, bivariate relationship is likely explained by confounding variables. Second, we proceed to utilize a more sophisticated logistic regression model to simultaneously examine the effect of our independent variables on the choice between federal versus state prosecution for arson and robbery. We find statistically significant evidence that cases prosecuted under federal arson and robbery statutes are more likely to include circumstances such as a conspiracy, a minor victim, use of a weapon, and serious recidivism.In Part IV, we conclude by discussing the higher plea rates and longer sentences imposed under federal as opposed to state criminal justice systems. We argue that where crimes involve the above-noted more egregious circumstances, federal prosecutors are more likely motived to prosecute the crime in expectation of a likely guilty plea and longer sentence. Our study provides much needed empirical evidence to support this rational view of federal prosecutorial discretion.