Argument preview: When is a burglary a “burglary”?

Because its application brings some of the federal system’s harshest mandatory penalties, and requires federal courts to categorize a diverse range of prior state convictions (most of which arose out of guilty pleas on undeveloped records), the Armed Career Criminal Act (“ACCA”), 18 U.S.C. § 924(e), has provided the Court with considerable business (and a fair amount of exasperation).  Descamps v. United States, 11-9540, set for argument on Monday, January 7, presents the Court with yet another categorization exercise that highlights the tension between such exercises and the Court’s developing constitutional doctrine about when judges can find facts in criminal proceedings.

Matthew Descamps was convicted after a jury trial of being a felon in possession of a firearm, in violation of 18 U.S.C. § 922(g).  Ordinarily, such a conviction exposes a defendant to ten years’ imprisonment with no mandatory minimum.  Under the ACCA, however, if the government shows the sentencing judge that a defendant has three prior convictions for a “violent felony,” he faces a mandatory fifteen years, with a maximum of life imprisonment.  (More information about the 592 offenders who qualified as armed career criminals under the ACCA in 2010 is available here.).  Descamps had a number of prior felonies, but the big question at his 2008 sentencing was whether a 1978 California burglary conviction counted as a “burglary” for ACCA purposes.

Although a “burglary” can qualify as a “violent felony” under the ACCA, the Court in Taylor v. United States, faced with a wide diversity of state “burglary” definitions, set out a generic federal one:  A person is convicted of burglary for purposes of the ACCA “if he is convicted of any crime, regardless of its exact definition or label, having the basic elements of unlawful or unprivileged entry into, or remaining in a building or structure, with intent to commit a crime.”  Because the California burglary statute lacked an element requiring that entry be unlawful or unprivileged, Descamps’s 2008 sentencing judge looked to the 1978 guilty plea proceedings.  Back then, the charging document had alleged that Descamps had entered a building “with intent to commit a theft”; the parties had agreed that there was a factual basis for the plea, and when the prosecutor said the crime had “involve[d] the breaking and entering of a grocery store,” Descamps said nothing.  On this record, the district court found adequate evidence that the 1978 offense counted as a “burglary” for ACCA purposes and sentenced Descamps to 262 months’ imprisonment.

In its unpublished memorandum opinion affirming Descamps’s sentence, the Ninth Circuit relied on its recent en banc decision, United States v. Aguila-Montes de Oca, which held that when a prior conviction lacks a required generic element, a federal sentencing court can, under the Supreme Court’s “modified categorical approach,” look to charging documents or a defendant’s plea allocution to see whether the prior conviction counts as a “crime of violence” for purposes of the Sentencing Guidelines or the ACCA, or for purposes of determining whether an alien is removable as an “aggravated felony.” 

Descamps’s certiorari petition raised several questions, but the Court granted on only one: “Whether the Ninth Circuit’s ruling in United States v. Aguila-Montes De Oca [] that a state conviction for burglary where the statute is missing an element of the generic crime, may be subject to the modified categorical approach, even though most other Circuit Courts of Appeal would not allow it.”

While the issue is pitched as one of statutory interpretation, substantial constitutional concerns lurk just beneath the surface.  As the Court explained in Cunningham v. California, under the Apprendi line of cases, the Sixth Amendment right to jury trial prohibits “a sentencing scheme that allows a judge to impose a sentence above the statutory maximum based on a fact, other than a prior conviction, not found by a jury or admitted by the defendant.”  The doctrine’s focus on statutory maximums (which are rarely imposed), but not mandatory minimums (which always are) is an artifact of a line of cases that, on January 14, 2013, will, once again, be up for reconsideration when the Court hears argument in Alleyne v. United States.  But the rule’s carve-out reflects the continued vitality of Almendarez-Torres v. United States, which held that fact-finding as to the existence of prior convictions can be done by judges, not juries, even when such findings can increase a defendant’s statutory maximum. Because this carve-out is a glaring exception from the constitutional rule, the Court has patrolled it carefully, in cases like Shepard v. United States, which limited the universe of materials a sentencing court can consult to determine what the jury in the prior case was actually required to find, or what the defendant necessarily admitted.

Constitutional concerns about judicial fact-finding that increases sentencing maximums might counsel a “categorical” approach to ACCA prior felony calculations (and similar such projects). Under that approach, federal courts would simply classify offenses according to whether, on their face, they fit the requisite federal definition.  But many state statutes are so capacious that they sprawl across the federal categories. Some state offenses are defined to cover multiple subjects, like the Massachusetts statute that covered breaking and entering a “building, ship, vessel or vehicle,” where only certain subjects (in this case, a building) would satisfy the generic federal definition.  In those cases, under its “modified categorical approach,” the Court has allowed the sentencing court to look to such things as the charging document, the plea agreement, and the plea colloquy to determine the prior conviction’s basis and see if it qualifies.

Descamps thus turns on how “modified” the “modified categorical approach” is to be.  Can a judge resort to charging and plea allocution materials only when the state statute has multiple parts, only some of which manifestly satisfy the requisite federal definition?  Or can she also make similar resort simply when the statute doesn’t refer to an element of the generic crime?  The Ninth Circuit saw the absence of a required element in statutory language as an occasion for judicial inquiry that might fill the statutory gap.  The majority en banc opinion in United States v. Aguila-Montes de Oca reasoned:  “The only conceptual difference between a divisible statute and a non-divisible statute is that the former creates an explicitly finite list of possible means of commission, while the latter creates an implied list of every means of commission that otherwise fits the definition of a given crime.” Other circuits have been far more wary about expanding the categorical approach beyond the exceptional modification needed to accommodate statutes with divisible parts.  For a nice example, see United States v. Beardsley.

Petitioner and amicus NACDL get considerable millage out of the “plain language” of the ACCA, which looks to what a defendant was previously “convicted of,” not what he did. Because convictions are legal constructs defined by the elements of the statutory offense, Descamps argues that the statute does not authorize courts to rummage through charging and plea allocution materials to determine the defendant’s actual conduct.  On occasion, an offense with alternative elements may require recourse to such materials, but the focus should remain on the statutory definition.  Moreover, petitioner argues that, because wide-ranging factual inquiries by sentencing judges, of the sort allowed by the Ninth Circuit’s approach, would violate the Sixth Amendment, the ACCA should be interpreted to generally preclude them.  Finally, petitioner argues, going beyond statutory elements would create “a host of practical and fairness problems.”  One cannot, for example, fairly take Descamps’s silence in the face of the prosecutor’s statement that he had committed “breaking and entering” as conclusive proof of that fact, since it was not legally relevant to the offense he was pleading guilty to.  Indeed, the NACDL argues, the designation of statutory elements to which a defendant will plead often lies at the heart of plea bargains.

The government’s brief doesn’t see what the fuss is about.  Far from manufacturing a federally required element from the 1978 plea proceedings, Deschamps’s sentencing judge simply drew on the California Supreme Court’s controlling interpretation of the state’s burglary statute, which, though somewhat broader than the generic offense, was close enough that the gap could be bridged by recourse to the limited materials envisioned by Shepard.  More generally, so long as a sentencing judge’s review stays within those tight constraints whenever confronted with a conviction whose legal basis (whether as a matter of statute or controlling judicial decision) encompasses a generic federal definition, the procedure will be easy, constitutional, and serve both horizontal equity and public safety by ensuring that similar violent criminals across a nation with diverse penal codes are treated the same way.

Although the Court specifically refused to grant that part of Descamps’s petition asking for Almendarez-Torres to be overruled, it will be interesting to see whether hostility to that case drives oral argument.  By Justice Thomas’s tally in Shepard v. United States, “a majority of the Court now recognizes that Almendarez-Torres was wrongly decided.”  So look for the Justices to limit the Almendarez-Torres carve-out by making inquiries into the nature of prior offenses as mechanistic as possible.  Given the warnings by lower courts of the practical and constitutional difficulties raised by the Ninth Circuit’s outlier approach, we should expect the government to face an uphill battle.  And, for all its technical aspects, it is one worth watching, as the heaviness with which the federal hand comes down on a lot of defendants depends on how their prior criminal convictions get categorized.

Daniel Richman is the Paul J. Kellner Professor of Law at Columbia Law School.