The arc of the pendulum: judges, prosecutors, and the exercise of discretion.

• Author: Stith, Kate

FEATURE CONTENTS INTRODUCTION I. TRY AND CATCH THE WIND: EFFORTS TO LIMIT DISCRETION IN FEDERAL CRIMINAL SENTENCING A. The Sentencing Reform Act of 1984: Sentencing Rules To Control Judges B. The Sentencing Commission: "Real Offense" Sentencing To Control Prosecutorial Undermining of Sentencing Rules C. The Inquisitorial Implications of "Real Offense" Sentencing D. Main Justice: Constraints on Prosecutors and the Pursuit of Centralized Control II. TAKING MEASURE OF THE GUIDELINES REGIME A. Sentencing Cases in the Courts of Appeals B. Sentencing Data from the District Courts C. Measuring Other Consequences of the Guidelines III. THE CHALLENGE OF THE FEENEY AMENDMENT IV. THE SUPREME COURT'S POWERS TO LIMIT PROSECUTORIAL DISCRETION A. Booker: Recharging Judges and Prosecutors B. Booker Is for Real CONCLUSION INTRODUCTION

In the federal criminal justice system, both prosecutors and judges have historically exercised broad discretion--prosecutors in charging (or not charging), and judges in sentencing. Both prosecutorial and judicial discretion in the criminal process date back to the very beginnings of the Republic. (1) For most of our history, the exercise of discretion has simply been taken for granted by judges, by prosecutors, and most importantly, by Congress, which has created a system of criminal laws that requires--and has always required--the exercise of discretion. Unlike the civil system in continental Europe, the common law has never featured or claimed to feature mandatory exercise of prosecutorial power.

In the modern era, we have grown suspicious of discretion. To a formalist, discretion seems the very antithesis of law. To a realist who views law as simply power, discretion is, at best (in Judge Marvin Frankel's memorable book title), "law without order." (2) A central campaign of the modern age--extending far beyond sentencing and the criminal justice system--has been to reduce the discretion of government officials. (3)

I use the term "power" to refer to lawful authority to take action against an individual. "Discretion," on the other hand, is the authority not to exercise power. In the context of the criminal law, to exercise discretion means, most simply, to decide not to investigate, prosecute, or punish to the full extent available under law. Discretion in federal criminal law enforcement is so great and so difficult to constrain because it is a necessary concomitant of the substantive federal criminal law. (4) That is, federal statutory criminal law has great breadth and has always included both lesser-included offenses and overlapping offenses. Moreover, the federal criminal law has always been an adjunct to state criminal law; most conduct that violates federal law also violates state law. Thus, in many instances, federal prosecutors must decide both whether to intervene in potential state prosecutions and, if they do choose to intervene, which crimes to charge. Federal prosecutorial decision makers (whoever they may be--line prosecutors, U.S. Attorneys, or officials and bureaucrats in the Department of Justice) necessarily have broad charging discretion. Concomitantly, sentencing authorities (whoever they may be--judges, administrative agencies, or prosecutors) necessarily have broad discretion over punishment. As Congress well understands when it enacts federal criminal proscriptions, both prosecutorial and sentencing discretion are inevitable because of the broad reach of these proscriptions and the severity of authorized punishments. (5) Resource constraints as well as prudence dictate the conclusion that the federal criminal law cannot be applied in its full rigor. (6) Someone has to exercise authority to decide what to investigate, what to prosecute, what to charge, and how great punishment will be.

The inevitable exercise of charging and sentencing discretion in the federal criminal justice system has been a recurring theme in the saga of the Federal Sentencing Guidelines, whose recent transformation by the Supreme Court from a "mandatory" to an "advisory" regime (7) I consider in this essay. I do not view the Court's recent Guidelines decisions only from an internal perspective--that is, in terms of the competing constitutional doctrines expounded in these cases. Rather, I consider the recent decisions against the backdrop of inevitable, ongoing institutional rivalries. The institutions in play include not only the inferior federal courts (both trial and appellate), Congress, and the U.S. Sentencing Commission, but also the Supreme Court, federal prosecutors in the ninety-four U.S. Attorneys' offices, and, importantly, the U.S. Department of Justice in Washington, D.C. ("Main Justice"). Early scholarship on the Federal Sentencing Guidelines focused on the transfer of sentencing authority from judges to the Sentencing Commission; later studies have examined the transfer of discretion from judges to prosecutors. Of equal significance are two other ongoing competitions for power: one between local federal prosecutors and officials in Main Justice, the other between Congress and the Supreme Court. In its 2005 decision in United States v. Booker (8) and its recent decisions elaborating the new sentencing regime constructed in Booker, the Supreme Court asserted the significant responsibility and authority of sentencing judges, local prosecutors, and the Supreme Court itself.

In Part I, I seek to identify the critical decisions made in constructing and implementing the Guidelines, decisions that ultimately resulted in increased prosecutorial power and discretion. This discretion could, and would, be used to influence defendants to plead guilty or face remarkably severe Guidelines sentences. Although it was not the goal either of sentencing reformers or of Congress, the actual result of the Guidelines regime that took effect in late 1987 was to transfer sentencing authority not to the U.S. Sentencing Commission, but to federal prosecutors in general and--particularly in recent years--to Main Justice.

Because I have elsewhere expressed skepticism about the project of uniform application of sentencing rules, (9) I do not dwell here on the issue that motivated the Sentencing Reform Act (10)--the existence of "disparity" among judges in sentencing. Disparity unquestionably exists. But requiring judges to apply national sentencing rules risks masking both the continued significance of the individual judge in sentencing and the increased leverage over defendants afforded to prosecutors in plea bargaining. The federal effort to stamp out judicial disparity through the Guidelines was probably not successful. (11) In any event, the decades-long enterprise provided prosecutors with indecent power relative to both defendants and judges, in large part because of prosecutors' ability to threaten full application of the severe Sentencing Guidelines.

Part II explains why neither appellate decisions nor raw sentencing data are an accurate tool to measure the Guidelines' success in achieving greater national uniformity in sentencing--nor even for measuring the extent to which the Guidelines are actually implemented. Each criminal sentencing is ultimately highly "local," a result of the strategic decisions of the prosecutor, the defense attorney, and the judge--all acting within the factual confines of the case at hand as well as the larger norms and practices of the judge's courtroom, of the federal district, and of the relevant circuit. Further, the sentencing decisions of the courts of appeals--including the "win/loss" ratio for defendants and the government--tell us very little about law on the ground. Few sentencing decisions are appealed by defendants, and even fewer are appealed by the government. (12) While courts of appeals may use these cases to signal to district courts how rigorously they should apply the Guidelines, this signal is imperfect at best and may be ignored altogether in cases that are not likely to be appealed. Indeed, even ground-level sentencing data--the sort of data assiduously compiled by the Sentencing Commission for every sentence in the federal courts--is a poor measure of the extent of Guidelines implementation and compliance. Although we can count the case reports submitted by judges, and thereby determine the ratio of reported Guidelines sentences to reported non-Guidelines sentences, there is no way to judge how accurate these reports are--or even what "complying" with the Guidelines would mean.

The unreliability of appellate decisions and raw sentencing data as portrayals of actual practice has not always been appreciated. Interested political observers, in particular, have looked to appellate case law and to the frequency of reported non-Guidelines sentences as a measure of the extent to which judges have "complied" with the Guidelines and thus implemented Congress's design to reduce sentencing disparity. Part III recounts Congress's 2003 decision--in reaction to sentencing decisions in particular white-collar cases and to nationwide data that appeared to reveal that sentencing judges were willfully ignoring the Guidelines in a growing proportion of cases--to enact legislation that represented a direct challenge to every level of the federal judiciary, the Sentencing Commission, and local prosecutors. By design, this legislation, known as the Feeney Amendment ("Feeney"), (13) simultaneously empowered Main Justice, which was Congress's partner in the endeavor to limit if not eliminate the exercise of discretion by decision makers in the field. Feeney added language to the Sentencing Reform Act to overturn a unanimous Supreme Court case, Koon v. United States, (14) that appeared to encourage judicial disregard of the Guidelines. Feeney also directed the Sentencing Commission to amend the Guidelines to reduce judicial discretion to impose below-Guidelines sentences, and directed the Department of Justice to monitor the sentencing advocacy of prosecutors...