What Feeney got right: why Courts of Appeals should review sentencing departures de novo.

• Author: Goldstein, Andrew D.
• Position: United States v. Thurston

Last summer, when a panel of the U.S. Court of Appeals for the First Circuit issued its initial ruling in United States v. Thurston, (1) it plunged into a war between federal judges and Congress, as well as between district and appellate courts, over how much flexibility trial judges have to tailor criminal sentences as they see fit. The war began nearly two decades ago with the passage of the Sentencing Reform Act (SRA) of 1984 (2) and the subsequent enactment in 1987 of the comprehensive Sentencing Guidelines regime. (3) It flared up anew in April 2003, with the passage of the Prosecutorial Remedies and Other Tools To End the Exploitation of Children Today (PROTECT) Act of 2003, (4) Congress's latest attempt to rein in the discretion of sentencing judges.

Before the PROTECT Act, a district court's decision to choose a sentence that departed from the range recommended by the Sentencing Guidelines was to be given considerable deference by courts of appeals. In accordance with the Supreme Court's 1996 decision in Koon v. United States, (5) appellate courts had been directed to review Guideline departures for "abuse of discretion." (6) But the PROTECT Act's "Feeney Amendment," named after its author, Florida Representative Tom Feeney, changed the standard of review to "de novo," and seemingly invited appellate courts to regularly second-guess the sentences imposed by district judges. The Amendment, which was approved after just fifteen minutes of debate on the floor of the House of Representatives, (7) quickly earned the condemnation of nearly the entire legal community. (8) Until Thurston, however, courts of appeals around the country had danced around the new standard of review, asserting in case after case that their decisions would be the same using either de novo or abuse-of-discretion review. (9) The First Circuit, which in the past had often led other courts of appeals when it came to sentencing decisions, (10) tackled the new law head-on in Thurston.

In December 2001, a Massachusetts jury found William Thurston, the vice president of a drug-testing company, guilty of conspiring to defraud the Medicare program of over five million dollars. A direct application of the range prescribed by the Sentencing Guidelines would have given Thurston a prison sentence of between sixty-three and seventy-eight months. But Judge Harrington was moved by Thurston's extensive service to his community and was troubled by the light sentence given to the company's president (who had pied nolo contendere to the same crime). So Judge Harrington departed from the Guideline range and sentenced Thurston to just three months' incarceration, with a recommendation for community confinement. (11) The government appealed, and by the time the First Circuit heard the case, the Feeney Amendment had become law.

In a comprehensive, twenty-five-page opinion issued on August 4, 2003, the Thurston panel exercised its new deference-free powers of review and rejected both of Judge Harrington's reasons for departing from the Guidelines, remanding with instructions to the judge to impose the statutory maximum sentence of sixty months. (12) Two months later, Judge Harrington formally refused to do so. "The Court recuses itself from this case," he wrote. "It is disinclined to mechanically impose a sentence, previously prescribed by the Court of Appeals, which is clearly contrary to the objective of the sentencing guidelines." (13) In a footnote, echoing a widely held feeling among judges across the country, (14) Judge Harrington added:

The newly granted authority given the Court of Appeals under the Protect Act to review de novo a departure decision renders the trial judge superfluous and with only a nominal function at disposition. This is so notwithstanding the fact that the trial judge possesses deep understanding of the case based on knowledge acquired from presiding at two complex criminal jury trials. (15) The purpose of this Note is to suggest that despite Judge Harrington's comments--and the doomsday rhetoric of much of the legal community--departure decisions should be reviewed de novo by the courts of appeals. The "abuse of discretion" standard for reviewing Guideline departures is, I argue, a relic of the pre-Guidelines regime in which judges were authorized--and expected--to consider each offender holistically and base their decisions on any available information, including the individual's character, upbringing, and family life. (16) Only the sentencing judge is situated to conduct this kind of open-ended inquiry, making deferential review on appeal appropriate. For better or worse, however, the enactment of the Sentencing Guidelines in 1987 fundamentally changed the essence of the sentencing decision from an exercise of wide-ranging discretion to an application of carefully delineated rules by which offenders are placed in categories rather than treated as individuals.

Under the Guidelines, once the facts of a given case are established, the decision to depart from a prescribed range requires determining what the Guidelines mean--what types of offenses and offenders the Guidelines are and are not intended to cover. This is, at its essence, a legal assessment that appellate courts are at least as capable of making as district courts. Appellate courts may in fact be better-suited to make these determinations: Not only are they more accustomed to looking at case law across districts and circuits, ensuring more consistent nationwide application of the Guidelines, but they also enjoy the benefit of input from a panel of judges, as well as from the judge below. More importantly, only appellate courts can bring uniformity to normative judgments about the Guidelines' meanings--judgments that form the core of decisions to depart.

The Feeney Amendment's de novo standard of review recognizes these comparative institutional competencies. De novo review also has the benefit of forcing appellate courts to grapple directly with the structure and application of the Guidelines and the purposes of sentencing, as opposed to simply setting boundaries around district court decisionmaking. The courts of appeals, armed with the power of de novo review, have the opportunity to create a principled common law of sentencing departures that can fill the considerable gaps in the existing Guidelines regime.

Part I of this Note provides the background for my analysis, describing the history of appellate review of sentencing decisions from the pre-Guidelines regime through the enactment of the Feeney Amendment. Part II describes why de novo review is the appropriate standard for evaluating Guideline departures, given the purposes and structure of the Sentencing Reform Act, the Guidelines, and the Feeney Amendment. Part III explains why, as a matter of policy, we should embrace greater appellate involvement in sentencing. Part IV addresses the primary justifications that judges and scholars have given in support of the Koon abuse-of-discretion standard. Part V revisits the circumstances of Thurston to show what de novo review looks like in practice (and why it may not be so dangerous after all).

1. A BRIEF HISTORY OF APPELLATE REVIEW OF FEDERAL SENTENCING

   1. Before the Guidelines

      Until 1987, district court judges had enormous discretion to tailor criminal sentences as they saw fit. Judge Marvin Frankel, whose book Criminal Sentences: Law Without Order helped bring about the Guidelines regime, decried the "almost wholly unchecked and sweeping powers" given to sentencing judges--a system that he called "terrifying and intolerable for a society that professes devotion to the rule of law." (17) When determining a sentence, district court judges were instructed to consider not just the crime committed and the offender's relevant criminal history, but also the offender's whole life and character. (18) Professor Kevin Reitz notes that "[w]ith such a free-form thought process in gear, there were effectively no legal principles against which a sentence could be tested on review." (19)

      As a result, so long as sentences fell within the (typically broad) ranges prescribed by statute, just about any sentence issued by a district court judge, for whatever reason, was likely to survive appeal. (20) Judges rarely explained their sentences, and consequently sentencing decisions were made with little or no guidance from case law. (21) The few appellate decisions that did exist mostly dealt not with substantive analysis of the purposes and principles underlying sentencing, but with constitutional issues such as whether the penalty constituted cruel and unusual punishment or whether the sentencing procedures violated due process. (22)

   2. The Early Guidelines Regime

      Then came the revolution. Spurred on by outrage over perceived disparities in sentencing across the country, (23) Congress passed the Sentencing Reform Act of 1984 and enacted the Sentencing Guidelines in 1987. Suddenly the vast majority of sentences were not up to the discretion of individual judges, but instead were dictated by simple application of an exhaustive set of rules.

      Take, for example, the federal bank robbery statute, which prescribes a maximum prison sentence of ten years for nonviolent robberies of more than $100. (24) Before the Guidelines, a judge had discretion to sentence someone who stole, say, $500,000 from a federally insured bank to anywhere from zero to ten years' imprisonment. Under the new regime, however, a sentencing judge must engage in a series of detailed calculations using numeric values assigned by the Guidelines. The judge must determine the "base offense level" (which for nonviolent bank robbery is six), add the number of levels corresponding to the amount stolen (fourteen for a "loss" between $400,001 and $1,000,000), and apply any further adjustments commanded by the Guidelines (if the robbery employed "sophisticated means," for example, two more levels would be added). (25)...