How to Fix USSC’s Sentencing Guidelines for Career Offenders

Overview

Under 28 U.S.C. 994(h), Congress has directed the United States Sentencing Commission (USSC) to create federal sentencing guidelines that punish certain repeat offenders “at or near the maximum term authorized.” These are known as “career offenders.” Career offenders are those who have at least two previous convictions (state or federal) for certain violent and controlled substance offenses, and their current conviction is either a violent felony or controlled substance offense. If an individual is found to be a career offender, then the sentencing guidelines call for a significantly longer sentence. This is called a “sentencing enhancement.” When determining who is a career offender for sentencing enhancements, whether found in statute or in the Sentencing Guidelines, courts use the “categorical approach.” This forces judges to ignore the actual way the individual defendant committed the crime and to focus only on the elements of the statute for which the defendant was convicted. This approach can result in “arbitrary and inequitable” prison terms that can either over-punish or under-punish a defendant.

Although not technically required, judges still apply this categorical approach when determining whether someone is a career offender under the Sentencing Guidelines. While there are reasonable justifications for the categorical approach for sentencing enhancements found in statute, there is little logic in using this method to determine who is a career offender under the Sentencing Guidelines, as the same Constitutional concerns and rules of evidence do not apply. Accordingly, the USSC should move away from a categorical approach and adopt guidelines that allow judges to consider a defendant's actual conduct when determining whether a defendant is a career offender. This will ensure that those who commit truly violent offenses do not get to avoid an appropriately lengthy term of incarceration.  

Background

Many federal criminal statutes include punishment enhancements based on a person’s instant offense (i.e., the offense for which they are being sentenced) and prior convictions. For example, 18 U.S.C. §924(e), known as the “Armed Career Criminal Act” (ACCA), requires a 15-year mandatory minimum prison sentence if 1) the person’s instant offense is a violation of 18 U.S.C. 922(g) (oftentimes a felon in possession of a firearm); and 2) the person has three previous convictions for either (or both) a “violent felony” or a “serious drug offense.” Issues can arise when determining whether previous crimes should be classified as violent or as a serious drug offense, particularly when they were state convictions with varying definitions of crimes.

To resolve these issues, the Supreme Court in the 1990 case Taylor v. United States established the categorical approach (and later the modified categorical approach in Shepard v. United States) to determine whether a state statute falls into one of these categories for purposes of the enhancement. Generally speaking, the categorical approach requires courts to look only at the statutory elements of the crime for which the defendant was previously convicted, rather than the underlying facts of the case that led to the conviction. Then, the court must consider whether it matches the elements of a generic definition of that crime. If the elements of the prior offense match or are narrower than the generic definition, then the conviction qualifies. However, if the statute is broader than the generic definition, it does not. These generic definitions are often court-constructed and rely on reviewing state criminal codes, the intent of Congress, case law, and even dictionaries.

To illustrate this point, consider the facts in Taylor. In Taylor, the defendant pled guilty to possession of a firearm by a convicted felon in federal court. He had four prior convictions, two of which were for second-degree burglary under Missouri law. Federal prosecutors wanted to apply the ACCA enhancement. Under the ACCA, a violent felony was defined in part as “burglary, arson, or extortion, involves use of explosives, or otherwise involves conduct that presents a serious potential risk of physical injury to another.” The question before the court was whether Missouri’s statute for second-degree burglary was considered a violent felony under the ACCA. The Supreme Court held that courts must consider only the elements of the prior conviction and disregard the underlying facts of the case or the conduct involved in the defendant’s conviction of that crime, and whether it matches a generic definition of burglary. Here, Missouri had seven statutes laying out how a person could be guilty of second-degree burglary, and not all of them contained the required elements of the generic federal definition of burglary. It was unclear from the record which statute the defendant was found guilty of second-degree burglary, so the court reasoned that it could not categorize these as a violent felony, and, therefore, the enhancement could not apply based on the current record.

Courts have articulated several reasonable justifications for applying a categorical approach rather than a predominantly conduct-based one. First, the text of the Armed Career Criminal Act and similar enhancement statutes mentions the defendant's "previous convictions" for certain categories of crime rather than the underlying conduct, strongly suggesting that Congress intended courts to look only to the elements of the statute of conviction. Second, a conduct-based inquiry would raise potential Sixth Amendment concerns because permitting a sentencing judge to make factual findings about prior conduct never found by a jury beyond a reasonable doubt risks being unconstitutional. Third, a conduct-based approach presents practical difficulties, as judges may be required to reconstruct what a defendant actually did during a prior offense based on records that may be incomplete or difficult to obtain.

While there may be valid reasons for the categorical approach, it can yield outcomes that seem to undermine the Congressional intent behind these criminal enhancements. For example, in Descamps v. United States, a defendant's prior California burglary conviction did not qualify as an ACCA offense because California's burglary statute criminalized entry into certain locations without requiring the unlawful entry element that a generic burglary definition contains, meaning that even a defendant who indisputably committed conduct matching generic burglary did not receive the enhancement solely because of how California’s statute was constructed. Similarly, in Mathis v. United States, a defendant with five prior Iowa burglary convictions avoided an ACCA enhancement because Iowa's burglary statute listed alternative locations broader than those required by the generic definition, regardless of what the record showed that the defendant had actually done.

The Courts' Use of the Categorical Approach in Sentencing Guidelines

Courts use the categorical approach when determining if a defendant is eligible for enhancements under the Sentencing Guidelines as well. For example, to determine who is a career offender under §4B1.1 of the USSC Guidelines, the judge must determine that 1) the defendant was at least 18 years old when they committed the instant offense that the person is being sentenced for; 2) the instant offense is either a crime of violence or a controlled substance offense; and 3) the person has at least two previous felony convictions (state or federal crimes) for a crime of violence or controlled substance offense.

The same issues that plague the application of the categorical approach to statutory enhancements are present when it is applied to the Sentencing Guidelines. As the Department of Justice noted in its comments to the Sentencing Commission in 2018, the categorical approach has resulted in offenses that, on the surface, seem to easily qualify for the career offender enhancement under the Guidelines being routinely denied inclusion in federal courts. United States v. Edling illustrates this problem. The Ninth Circuit held that a prior Nevada robbery conviction could not qualify as a "crime of violence" under the Guidelines, not because the defendant's conduct was non-violent, but solely because Nevada's robbery statute encompassed force directed at property in addition to force directed at a person.

Although courts are not actually required to use the categorical approach in the Sentencing Guidelines, judges have continued to do so, justifying the practice on grounds of Sixth Amendment constraints, administrative issues, and uniformity. However, these justifications are less compelling in a discretionary sentencing regime where judges exercise broad factfinding authority. In 2000, the Supreme Court in Apprendi v. New Jersey required that any fact that increased a penalty beyond the maximum sentence had to be proven by a jury beyond a reasonable doubt. That included the Sentencing Guidelines, meaning a judge who determined facts not submitted by a jury to increase a sentence beyond the Guideline maximum would be unconstitutional. However, in 2005, in United States v. Booker, the Supreme Court held that the Sentencing Guidelines are now discretionary, meaning that judges may depart from the recommendations based on the individual facts of the case. Therefore, the previous justifications for the categorical approach under the Sentencing Guidelines are considerably weaker.

The Sentencing Commission has itself recognized that because no barrier prevents a sentencing judge from considering what a defendant actually did during a prior offense, there is little justification for a rule that artificially limits what the court may consider when determining whether that prior conduct warrants a career offender designation.

Policy Recommendations

Congress can amend the Sentencing Reform Act to explicitly direct the USSC to abandon the categorical approach in favor of a conduct-based classification system for prior convictions. This would ensure that sentencing enhancements more accurately reflect the underlying conduct of a defendant's convictions rather than the formal elements of an overbroad statute.

Fortunately, the USSC is actively considering changes to the Sentencing Guidelines that would end the categorical approach (or at least significantly amend it). However, without Congressional action, there is no obligation to do so, and even if changes are made, they could be amended later on.  

The USSC should modify the Sentencing Guidelines to adopt a conduct-based approach that considers an individual's actual actions during the commission of the crime when determining sentencing enhancements. The Guidelines are advisory, and sentencing courts already exercise broad discretion to consider a defendant's background and conduct. A rigid elements-based rule that excludes clearly violent prior offenses from a court's analysis serves no legitimate purpose in this context and undermines Congress's explicit intent.

Currently, U.S.S.G. §4B1.2(a) defines a crime of violence in two ways: through a force clause, which covers offenses involving the use or threatened use of physical force against another person, and an enumerated offenses clause, which lists specific crimes such as murder, robbery, and kidnapping. If the Sentencing Commission amends the Guidelines, it should retain the force clause rather than replacing it with a static list of offenses alone. Criminal law evolves constantly, and a fixed list risks becoming incomplete or obsolete as Congress amends existing statutes and as new forms of violent conduct emerge that do not fit neatly into any predetermined category. More fundamentally, a list-only approach gives judges no ability to account for violent conduct that falls outside its boundaries, potentially allowing dangerous, repeat offenders to escape the career offender designation, not because their conduct was non-violent, but because it was not specifically named. This would undermine Congress's intent to ensure that repeat violent offenders receive sentences that reflect the seriousness of their records.

Retaining the force clause is especially important under a conduct-based framework. Today, the force clause serves largely as a fallback when a prior offense does not match an enumerated crime. Under a conduct-based approach, however, courts would be asked to evaluate what a defendant actually did, making a clear definition of what constitutes sufficient violence essential. Without the force clause to provide that standard, courts would have no consistent benchmark for deciding when prior conduct is serious enough to warrant an enhancement.

Guidelines that permit judges to consider the actual nature of a defendant's prior conduct rather than the technical elements of the statute of conviction would better reflect the realities of how sentencing already works in practice. Continued reliance on the categorical approach in the Guidelines context produces outcomes that appear arbitrary to defendants, victims, and the public, and it results in courts imposing sentences that bear little relationship to the risk and culpability that a defendant's prior record actually represents. Reform is needed and long overdue.