Has Congress ever made the federal courts do more work to little so good effect than it did when it passed the Armed Career Criminal Act in 1984? The ACCA imposes a fifteen-year mandatory minimum on certain federal defendants who have three prior convictions for a violent felony or serious drug crime, which are defined terms in the statute. The basic application problem is that we have fifty different state criminal codes, and state legislatures never saw fit to amend their laws so as to fit their crime definitions to the ACCA terminology. As a result, figuring out which state convictions count as ACCA predicates has consumed — and continutes to consume — an enormous amount of judicial time and effort. A few lines of statutory text have generated a marvelously intricate, uncertain, and ever-changing body of jurisprudence.
The Supreme Court offered its latest foray into the ACCA quagmire yesterday in Descamps v. United States (No. 11-9540). At issue was whether Descamps’s prior burglary conviction in California could be used as a predicate for the fifteen-year ACCA mandatory minimum. The statutory definition of “violent felony” does include “burglary,” but the Court has previously held that not all burglary convictions count; rather, the crime of conviction must have the elements of “generic burglary” — if a state has chosen to define the crime of burglary in an unusually broad manner, then convictions of burlgary in that state may not be treated as burglary convictions for ACCA purposes.
And it turns out that California does have an idiosyncratic burglary definition.
While the crime of burglary traditionally requires unlawful entry, California’s version does not. Thus, the Court observed, the California law “covers, for example, a shoplifter who enters a store, like any customer, during normal business hours” — hardly the image one normally has of a burglar and surely not what Congress had in mind when it classified burglary as a “violent felony.”
Using Descamps’s burglary conviction against him would thus run afoul of what the Court has called the “categorical approach” to ACCA analysis. Under this approach, all convictions under a given state law are either in or out; the courts are not to engage in a defendant-by-defendant inquiry into the details of who did what. Because Descamps might have received his burglary conviction for mere shoplifting, the conviction cannot be used against him, even though it is possible that what he actually did involved unlawful entry.
Yet, Descamps nonetheless lost in the lower courts. Why? Because the Supreme Court had previously recognized an exception to the categorical approach for “divisible” crimes. A divisible crime is one that has elements in the alternative. For instance, imagine a statute that defines burglary as “breaking and entering either a home or a vehicle for the purpose of committing a felony.” Breaking and entering a home would fit the generic definition of burglary, but breaking and entering a vehicle would not. In such circumstances, the Court has approved a limited examination of charging documents and other records from the earlier case in order to determine whether the defendant was convicted of doing something that amounts to generic burglary. This is the “modified categorical approach,” and it was used by the Ninth Circuit in examining a transcript from Descamps’s burglary case to determine that he was no mere shoplifter.
The Supreme Court reversed in an 8-1 decision, holding that the modified categorical approach really is limited to formally divisible statutes. A statute that is simply overbroad, like California’s, does not qualify.
Was all of this effort really worth it? The Ninth Circuit’s decision was based on an earlier en banc opinion dealing with the same California burglary statute. The Supreme Court’s effort to sort the mess out resulted in a set of opinions totaling 47 pages. All of this to resolve the status of one crime in one state. This is to say nothing of the pages and pages of the federal reporters devoted to questions like whether fleeing a law enforcement officer or escaping from prison count as violent felonies.
All of this might be okay if the ACCA actually served some principled end, but it does not. Like California’s notorious three-strikes law, the ACCA imposes an exceedingly harsh penalty on the basis of an extraordinarily crude use of criminal history. Criminal history has long been viewed as an appropriate consideration at sentencing, but there is widespread expert consensus on two limiting principles: (1) the seriousness of the current offense, and not criminal history, should be the primary determinant of sentence severity — criminal history should not become the tail wagging the dog — and (2) the weight given to prior convictions should vary depending on several different factors, including recency of the conviction, age at the time of conviction, seriousness of the prior offense, and similarity of the prior offense to the current offense. The basic criminal history provisions of the federal sentencing guidelines provide a good, empirically validated model for the nuanced assessment of prior convictions. Unfortunately, the ACCA overrides the normal guidelines treatment of criminal history whenever it comes into play. No recognized purpose of sentencing is advanced by substituting the ACCA’s crude treatment of criminal history for the guidelines’ more sophisticated approach. All of the judicial effort in cases like Descamps goes in service of a statute that makes federal sentencing less just, less cost-effective, and, by diverting scarce criminal-justice resources in unproductive ways, less protective of public safety.