The decision rests on a line of Seventh Circuit cases going back to United States v. Cunningham, 429 F.3d 673 (7th Cir. 2005).  These cases, which have not been followed in some other circuits, require district judges to address nonfrivolous arguments for a sentence below what is recommend by the sentencing guidelines.  As I discussed in this article, I think the Cunningham rule should be adopted more widely and enforced more rigorously.  For that reason, I’m glad to see the Seventh Circuit reaffirm the rule in Robertson.

Although it does not purport to break any new legal ground, the decision nonetheless has some noteworthy aspects.  

First, the court strongly embraced self-motivated rehabilitation as a sentencing factor.  Here’s some of the language:

The Supreme Court recently reiterated “the principle that ‘the punishment should fit the offender and not merely the crime.’” Pepper v. United States, 131 S. Ct. 1229, 1240 (2011), quoting Williams v. New York, 337 U.S. 241, 247 (1949). “Highly relevant — if not essential — to [the] selection of an appropriate sentence is the possession of the fullest information possible concerning the defendant’s life and characteristics.” Pepper, 131 S. Ct. at 1235, quoting Williams, 337 U.S. at 247. This aim is codified in 18 U.S.C. § 3553(a), which requires that any sentence imposed be “sufficient, but not greater than necessary” to serve the sentencing goals of punishment, deterrence, protection of the public, and rehabilitation, and which requires the court to consider “the history and characteristics of the defendant.” Adequate consideration of a defendant’s evidence of rehabilitation fits squarely within these parameters. Demonstrated self-motivated rehabilitation is direct and relevant evidence of “the need for the sentence imposed . . . to afford adequate deterrence to criminal conduct; to protect the public from further crimes of the defendant; [and to] provide the defendant with needed educational or vocational training . . . or other correctional treatment in the most effective manner.” 18 U.S.C. § 3553(a)(2)(B)-(D).

The power of evidence of self-rehabilitation was evident inGall, where the Supreme Court noted that it was reasonable for the district court to attach “great weight” to a defendant’s decision to change his life and withdraw from a drug distribution conspiracy: “Compared to a case where the offender’s rehabilitation occurred after he was charged with a crime, the District Court here had greater justification for believing [the defendant’s] turnaround was genuine, as distinct from a transparent attempt to build a mitigation case.” 552 U.S. at 57. Such self-motivated rehabilitation “lends strong support to the conclusion that imprisonment [is] not necessary to deter [a defendant] from engaging in future criminal conduct or to protect the public from his future criminal acts.” Id. at 59.  (12-13)

It’s interesting to see such reliance on the Supreme Court’s recent decision in Pepper.  As I discussed here, Pepper strikes me as a potentially significant break from the Court’s recent federal sentencing jurisprudence.  If lower courts were to read Pepper for all it’s worth, we might start to see a real shift in federal sentencing practices.

Also notable in Robertson is the fact that the defendants’ arguments were not entirely passed over in silence, as they were in Cunningham.  Whereas some Cunningham-type cases involve district judges who literally say nothing to indicate that they have even heard a defendant’s argument, the judge in Robertson expressly indicated some awareness of the facts emphasized by Mr. and Mrs. Robertson:

Concerning the Robertsons’ criminal histories, the court acknowledged that Henry had not committed any crimes since 2002 and that Elizabeth lacked any criminal history. But other than noting, without further detail or explanation, that Elizabeth had provided “excellent service . . . as a professional in the medical field,” it is not apparent that the sentencing court considered the Robertson’s unusually strong evidence of self-motivated rehabilitation over the past ten years. Because the court’s silence makes it impossible to discern that it appropriately balanced the Robertsons’ rehabilitated lives and characters against the seriousness of their offense for purposes of 18 U.S.C. § 3553(a), we find this minimal treatment to be insufficient.

. . . The probation office and the government agreed that it would be appropriate to treat Henry’s criminal history as overstated due to the passage of time, yet the district court’s only acknowledgement of this argument was its comment that Henry was not “youthful or immature” when he committed the reckless driving offense in 2002.  (15-16)

I think it unfortunate that the Cunningham rule is sometimes treated as satisfied when there is mere acknowledgement of a defendant’s argument, as opposed to substantive responsiveness.  I’m glad to see Robertson suggesting a more rigorous approach to Cunningham.

In this regard, I thought it interesting that the Seventh Circuit made nothing of the fact that “the district judge agreed to modify Elizabeth’s date to report to prison to allow her to continue to work as a nurse long enough to become eligible for retirement benefits.”  (15 n.3)  This contrasts with the Sixth Circuit’s opinion in United States v. Liou, 491 F.3d 334 (6th Cir. 2007), in which the district judge’s decision to give the defendant a more favorable report date was held to constitute an adequate response to the defendant’s argument for a below-guidelines sentence.

An interesting question about Robertson is why the Seventh Circuit formally treated the district judge’s error as merely procedural, remanding for a resentencing at which the original sentence could be reimposed.  It is seems clear enough that the panel felt the district judge erred substantively, not just procedurally, in imposing multiyear prison sentences on two defendants who led exemplary lives for a decade after their crimes were committed.  For instance, it’s hard not to read this view between the lines of the final words of the Seventh Circuit’s opinion: “the [district] court should carefully weigh and explain its consideration of the Robertsons’ evidence of self-motivated rehabilitation.”  (17)  Why not end the possibility of misunderstanding or mischief at the district court level and forthrightly hold that a guidelines sentence would be substantively unreasonable in this case?

Appellate courts have been loathe to hold guidelines sentences substantively unreasonable.  (As a side note, it is actually a debatable question whether the Robertsons’ sentences can be fairly characterized as guidelines sentences, since their guidelines ranges were calculated using a newer and harsher version of the guidelines than existed at the time they committed their crimes.)  This reluctance doubtlessly owes much to the presumption of reasonableness that may be accorded guidelines sentences under Rita v. United States, 127 S. Ct. 2456 (2007), and to the underlying premise of Rita that the guidelines embody the research and expertise of the Sentencing Commission.  Appellate courts should recognize, however, that is does no violence to the logic of Rita to hold guidelines sentences substantively unreasonable in unusual circumstances that the Commission did not contemplate in crafting the guidelines.  To start holding more guidelines sentences substantively unreasonable — where there are sound, principled grounds for doing so — would in effect start to build the common law of sentencing for which many scholars have been advocating for a very long time.  Such a common law holds out the hope for greater transparency, consistency, and proportionality in federal sentencing.

But are there persuasive grounds for holding the Robertsons’ sentences substantively unreasonable?  But my lights, this is a complex and uncertain question.  The underlying principle would be that defendants who have led productive, crime-free lives for many years before being charged present almost no recidivism risk, rendering multiyear prison terms a needless burden on the both the defendants and our overcrowded federal prison system.

However, as someone who believes that punishment should be based more on the severity of the crime than the risk of the criminal, this principle leaves me a little cold.

On the other hand, as I’ve been exploring in some of my recent writing, I also think that retributive approaches to punishment may be compatible with crediting defendants for acts that have a penitential character.  I don’t know, though, whether anything the Robinsons did could fairly be characterized as penitential.

Aside from the questions relating to self-motivated rehabilitation, Robinson also raised a couple of other interesting questions.  First, the Seventh Circuit adhered to its precedent in holding that the Ex Post Facto Clause is not violated when defendants are sentenced under a harsher version of the guidelines adopted after their crimes were completed.  Several other circuits take a contrary view, which may make the question ripe for Supreme Court consideration.

Second, the Seventh Circuit noted, but felt it did not have to resolve, continuing uncertainty over whether the aggravating role enhancement of U.S.S.G. § 3B1.1 requires that the defendant have exerted control over other participants in a criminal activity.

Cross posted at Life Sentences.

The question has particular importance in fentanyl cases, as illustrated by the Seventh Circuit’s recent decision in United States v. Alvarado-Tizoc (No. 10-1613). In sentencing the wholesaler-defendants, the district court chose to attribute to them the full retail quantities, which were 11 to 16 times greater than the wholesale quantities.

This was improper, the Seventh Circuit held.  

The district court had relied on the relevant conduction provision of the guidelines, which makes defendants responsible for the reasonably foreseeable acts of others in furtherance of a “jointly undertaken criminal activity.” However, drawing on established principles of conspiracy law, the Seventh Circuit held that wholesalers and retailers do not necessarily engage in a “jointly undertaken criminal activity.” In general, a buyer-seller relationship does not establish a conspiracy – a point my Criminal Law students will immediately recognize (I hope) – and the same principle applies to the “jointly undertaken criminal activity” guideline. Moreover, the record in Alvarado-Tizoc did not indicate anything special in the relationship between wholesaler-defendants and their buyers so as to warrant a finding that they were jointly engaged in the retail sale of fentanyl.

The courts have not always been clear or consistent in their description of the relationship between the vicarious liability doctrine for coconspirators and the relevant conduct provision of the sentencing guidelines. For that reason, the unusually lucid treatment of the topic in Alvarado-Tizoc (penned by Judge Posner) warrants quoting at length:

The only vulnerable point in the sentencing of two of these defendants . . . was the judge’s finding that the jointly undertaken criminal activity included the retail sale of the fentanyl. There was insufficient evidence that the retailers to whom the defendants sold heroin and fentanyl were, so far as their relation to the defendants was concerned, anything more than buyers. The government points out that the defendants “specifically sought out, and received, information about [the retailers’] heroin business . . . and thus purposefully kept apprised of their operation. For instance, they asked specific questions about how much money [one of the retailers] made from the heroin he obtained from them” and informed them that “the ‘new heroin’ [which was actually fentanyl] could be diluted even further.” All this just shows a wholesaler’s natural motivation to gauge demand for his product and if possible increase that demand and so be able to raise his price. The government also notes that the defendants were “exclusive supplier[s]” of the retailers and therefore had a “vested interest in the success and profitability” of the operation. But exclusive dealing is common and every wholesaler has a vested interest in the success of his retailers. And finally the fact that the buyers diluted the fentanyl they received (and that this was foreseeable to the defendants) no more proved a conspiracy than the fact that a seller of chocolate syrup to a soda fountain knows that the syrup will be mixed with milk or soda to make chocolate milk shakes or chocolate sodas rather than being sold in its original, undiluted form makes the seller a conspirator in the retail sale of adulterated chocolate drinks.

A seller is not a party to a conspiracy with a mere buyer from him. E.g., United States v. Vallar, 635 F.3d 271, 286-87 (7th Cir. 2011); United States v. Colon, 549 F.3d 565, 567-71 (7th Cir. 2008). And while the applicable Sentencing Guideline as we know uses the term “jointly undertaken activity” rather than “conspiracy,” and indeed provides that the jointly undertaken activity need not be “charged as a conspiracy,” the case law generally treats the terms “jointly undertaken activity” and “conspiracy” as interchangeable.  See, e.g., Gray-Bey v. United States, 156 F.3d 733, 740-42 (7th Cir. 1998); United States v. McDuffy, 90 F.3d 233, 235-36 (7th Cir. 1996). The concept of conspiracy is frequently employed in criminal cases without a conspiracy actually being charged, as when proof of a conspiracy is used to render a statement by a coconspirator admissible against the defendant; and so it is with the Guidelines’ equivalent, a “jointly undertaken activity.”

Some cases point out that “jointly undertaken activity” should not be equated to “conspiracy” because a defendant could have joined a conspiracy without having joined in or agreed to all the activities undertaken by it. United States v. Soto-Piedra, supra, 525 F.3d at 531-32, and cases cited there; U.S.S.G. § 1B1.3 Application Note 2. But there is no actual conflict; the cases we cited earlier (Gray-Bey andMcDuffy) impose a sentencing enhancement on a conspirator for a jointly undertaken activity only if the activity was reasonably foreseeable to him, for foreseeing or being charged with foreseeing an activity makes him a joint participant with the other conspirators. United States v. Hernandez-Santiago, 92 F.3d 97, 100 (2d Cir. 1996). But this qualification on equating conspiracy to jointly undertaken activity cannot help the government; if there was no conspiracy between the defendants and the retailers, a fortiori the former were not engaged in a jointly undertaken activity with the latter.  (5-8)

But Posner also had a bigger and more challenging point to make about sentencing in cases involving fentanyl and other highly potent drugs.  As Posner sees it, since the guidelines’ drug equivalency table already takes into account differences in drug potency, there is potentially a form of double counting whenever a defendant is held responsible for the full weight of the diluted form of a drug:

But a point of more general significance for cases involving fentanyl and other superstrong narcotics is that attributing the amount of the diluted retail product to the seller (whether the seller is a retailer, or a wholesaler conspiring with a retailer) in computing the Guidelines sentence involves double counting. The quantity of the diluted retail product, if attributed to a wholesaler defendant, will already account, in part anyway, for the fact that fentanyl is more potent than heroin; if the same weight of fentanyl and heroin bought by a retailer makes 50 retail doses of fentanyl versus 5 of heroin, the seller of fentanyl will be “credited” with 10 times the quantity as the seller of heroin. To multiply 10 by 2.5 is to double count— more precisely to 2.5-count. And double or other multiple counting—at least when the judge is unconscious of doing it—is improper . . . .

Thus, irrespective of the fact that the quantities sold by the retailers could not be attributed to the defendants on the theory that they were conspiring with the retailers (rather than merely supplying them), the fact that by selling fentanyl instead of heroin the defendants were responsible for a larger number of doses sold to the ultimate consumers was a basis for a higher sentence—but a basis already partly reflected in the drug equivalency tables in the Guidelines.  (8-9)

In light of dilution-based difficulties with the weight-based sentencing guidelines, Posner invited district judges to use their post-Booker discretion to use a more sensible dose-based approach: sentences should not vary based on weight alone, but also take into account potency and purity.  The real question to answer in determining the relative gravity of a drug trafficking offense is not the weight sold, but the number of doses sold.  (10)  Posner observed:

Adjusting for potency makes more sense than adjusting for weight. Emphasis on the weight of a defendant’s drugs (in this case the weight of the dilute drugs sold by customers of defendants), whether or not they are diluted, has the perverse effect of giving drug dealers an incentive to possess and sell drugs of high purity or potency and makes the length of sentences depend perversely on the weight of the inactive ingredients in the drugs. Jonathan P. Caulkins et al., “Mandatory Minimum Drug Sentencee: Throwing Away the Key or the Taxpayers’ Money?” 22 (RAND Corp. Drug Policy Research Center 1997).  (11)

Cross posted at Life Sentences.

The Seventh Circuit’s position followed that of the Department of Justice.  However, since the initial retroactivity ruling, DOJ has changed its position and now supports partial retroactivity.  Additionally, three other circuits have since rejected the Seventh Circuit’s position.  In light of these developments, one of the Seventh Circuit judges proposed that the initial ruling be reconsidered en banc.  Last week, however, the court announced that the initial ruling would stand.

Remarkably, the court was split 5-5, just barely short of the majority required for rehearing.  One wonders how the late Judge Evans would have voted if he had lived a little while longer.  To be sure, he was part of the panel that initially rejected retroactivity, but both of the other members of the panel switched sides and supported rehearing.  (I can’t recall ever seeing such a switch by multiple panel members in connection with a rehearing decision — it is perhaps an indication of how important DOJ’s switch was in the minds of some of the judges.)

Judge Easterbrook wrote an opinion opposing rehearing, while Judge Williams wrote the principal opinion for rehearing.  United States v. Holcomb (No. 11-1558).  These are both remarkably lucid, cogent opinions.  They effectively present quite different approaches to statutory interpretation and touch on some deep jurisprudential questions.  This would make a very nice teaching case in a statutory interpretation class.

The general rule is against retroactivity for a new statute repealing penalties, although Congress is free to override the general rule.  As to the FSA, the Williams position is basically this: in the statute, Congress made clear that the Sentencing Commission should move as quickly as possible to reduce crack penalties in the sentencing guidelines, with the understanding that the reduced guidelines penalties would apply to everyone sentenced after the new guidelines took effect (Nov. 1, 2010).  Why, Williams asked, would Congress want the new guidelines sentences to be applied as quickly as possible to all new sentencings, but not want the same prompt implementation of the new statutory sentences?  Retroactivity for both types of sentence reductions is the only way to make sense of Congress’s intentions.

Ours is not to question why, responded Easterbrook.  For all we know, what seems an arbitrary distinction between retroactivity for guidelines changes and retroactivity for statutory changes was part of a backroom deal that was necessary to secure the FSA’s passage.  Bringing principled coherence to the law lies beyond the judicial role in interpreting statutes.  Arbitrary though the result may be, nothing in the FSA provides a sufficiently clear signal that Congress wanted retroactivity for the changes in the statutory penalties.

Cross posted at Life Sentences Blog.

That the explanation requirement would apply here is perhaps not a given, since, as the court observed, the district judge has even more discretion in this setting than in an original sentencing.  (4)  The court ruled, however, that the district judge must indeed “say something that enables the appellate court to infer that he considered both [the recommendations of the sentencing guidelines and the statutory sentencing factors].”  (4)

In Robertson, the guidelines recommended a term of 12-18 months following the defendant’s revocation for growing marijuana, but the district judge instead imposed a sentence of 34 months.  Here is the “explanation” for the sentence that the Seventh Circuit found inadequate:

The judge didn’t give a reason for the 34-month term, and made just four, brief remarks that might supply clues to his thinking. The first is that he was “baffled” that the defendant would continue growing marijuana after spending eight years in prison for that crime. The second remark, which followed immediately and merely repeated the first in different words was: “why did you even consider doing this again?” The defendant replied that he had grown marijuana because “he just liked the way the plant looked” and he “liked to smoke it,” whereupon the judge asked him sarcastically whether he had “ever thought about growing gardenias or something legal versus growing marijuana” and added (this was his fourth remark) “you could have grown roses.” Without further explanation the judge stated: “with respect to the revocation, the Court is going to sentence you to 34 months.”  (2-3)

Sending a man to prison for nearly three years surely requires more of an explanation than a couple of sarcastic observations about his horticultural preferences.  The Seventh Circuit properly vacated the sentence and remanded for resentencing.

Cross posted at Life Sentences Blog.

Here are the highlights of the Commission’s work.

First, the big, good news for crack defendants: The Commission chose to make retroactive the changes to the drug quantity table that were promulgated in April.  The Commission also made retroactive another guidelines amendment that reduces sentences for crack defendants convicted of simple possession.  (To be technically precise, these are Parts A and C of Amendment 750.)  These were the two decisions that I (and many other witnesses) advocated most forcefully for at the June hearing on retroactivity (see my post here), and they will make a big difference for a large number of people.  According to Commission analysis, “approximately 12,000 offenders would be eligible to seek a reduced sentence and the average sentence reduction would be approximately 23 percent.”  To be sure, district judges will have discretion to turn down any sentence-modification requests they receive, but the experience with retroactivity for the 2007 crack amendment indicates that the great majority of eligible defendants will indeed be granted sentence reductions.

Second, the Commission wisely rejected the Administration’s misguided request to disqualify defendants above Criminal History III or with firearms involvement.  (See my post here.)

Third, the Commission chose not to make any part of Part B retroactive.  Part B adds a variety of new aggravating and mitigating specific offense characteristics to the drug trafficking guideline.  Although I think the new “minimal role cap” of Part B could have been implemented retroactively with fairness and relative ease, the remainder of Part B would have presented more significant administrative burdens and fairness objections.

Fourth, the bad news for defendants: while the previous version of the retroactivity guideline (1B1.10) authorized sentence reductions below the new guidelines range if the original sentence had been a departure, the Commission has now indicated that the bottom of the amended guidelines range is a firm floor in all cases except cases involving a downward departure for substantial assistance on the government’s motion.  Apart from a desire to minimize the administrative burdens of retroactivity, I’m not sure the restriction makes much sense.  Imagine a defendant who received a downward departure from the previous crack guidelines, with the sentencing judge concluding that a variety of unusual mitigating circumstances rendered the defendant substantially less culpable than other defendants responsible for a comparable quantity of crack.  Now, with the new drug quantity table made retroactive, typical crack defendants have their sentences reduced to the same level as our super-mitigated defendant.  If the super-mitigated defendant cannot also get a sentence reduction, then the sentences no longer reflect important distinctions in culpability.

Even less do I see merit in making an exception in cases of substantial assistance.  I’ve always had mixed feelings about the highly preferential treatment snitches get under the guidelines, and it’s hard to see why they get yet another benefit here relative to other defendants with more genuinely mitigating circumstances.  Certainly, when we are talking about sentence modifications, there is no added incentive for snitching — it seems highly unlikely that a defendant deciding whether or not to cooperate would take into account the remote possibility that the applicable guidelines range might someday be reduced and that the Commission would then make the change retroactive.

In any event, judges and practitioners should note that the new version applies to all requests for sentence modification decided on or after the effective date of the amendment (Nov. 1, 2011) — including any requests based on the 2007 amendment that are still pending then.  Defendants who are eligible for a reduction based on the 2007 amendment and who received a non-substantial-assistance downward departure at their original sentencing should do everything they can to push their sentence modification requests forward in the next five months.

Finally, in amended application note 1(a) to the retroactivity guideline, the Commission resolved a circuit split on whether there are any types of departures that should be taken into account in calculating a defendant’s new guidelines range (e.g., under 4A1.3 for inadequacy of criminal history category).  The answer is “no.”  This changes the law in the First, Second, and Fourth Circuits.

Cross posted at Life Sentences Blog.

The Court’s decision seems pretty clearly correct as a matter of law (the government actually declined to defend the ruling below), but what about policy — should Congress repeal that language from § 3582(a)?  Added to the law as part of the Sentencing Reform Act of 1984, this language reflected two then-common criticisms of the rehabilitative paradigm in criminal law: (1) the pragmatic objection that prison-based rehabilitative programming did not work, and (2) the ethical objection that defendants should not be sentenced in excess of their just deserts in order to force social services on them.  The first objection has less force today than it did in the 1980’s, as good studies now document at least modest levels of success by some prison-based programs in reducing recidivism.  However, the second objection remains no less important today than it was three decades ago.  On the other hand, desert is hardly a precise concept in practice; there may be enough play in the joints to permit some consideration of rehabilitative programming at sentencing without exceeding desert-based constraints.  If so, § 3582(a) as written (and as interpreted in Tapia) may be more limiting than is necessary.

Cross posted at Life Sentences Blog.

The Supreme Court’s recent decisions in Begay and Chambers substantially narrowed the reach of the Armed Career Criminal Act’s fifteen-year mandatory minimum.  (For background, see this post.  Ironically, shortly after Narvaez was decided, the Court issued its opinion in Sykes v. United States, which seemed to back away from Begay.)  Five years before Begay, Luis Narvaez pled guilty to bank robbery and was sentenced as a career offender under the sentencing guidelines based on his prior convictions for “violent felonies,” including two convictions for failure to return to confinement in violation of Wis. Stat. § 946.42 (3)(a).  Later, in Chambers, the Supreme Court ruled that the Illinois crime of failing to report for confinement did not count as a “violent felony.”  Narvaez then filed a motion under 28 U.S.C. § 2255 to vacate his sentence in light of Chambers.  The district judge held that Chambers did not apply retroactively, but granted Narvaez a certificate of appealability.

On appeal, the government conceded that Begay and Chambers, as decisions of substantive criminal law, applied retroactively, and that, under Chambers, Narvaez’s failure-to-return convictions were not “violent felonies” for purposes of the career offender guideline.  However, the government argued that Narvaez could not take advantage of Chambers because he raised only a statutory issue in his petition — in order to get a certificate of appealability, a § 2255 petitioner must make “a substantial showing of the denial of a constitutional right.”

The Seventh Circuit rejected the government’s theory, finding that, as a result of Chambers, Narvaez’s sentence violated the Due Process Clause:

We have recognized that sentencing errors are generally not cognizable on collateral review. See Scott v. United States, 997 F.2d 340, 342-43 (7th Cir. 1993). Mr. Narvaez’s case, however, does not come within this general rule.  It presents a special and very narrow exception: A post-conviction change in the law has rendered the sentencing court’s decision unlawful. See Welch, 604 F.3d at 412–13 (recognizing that relief is available “where a change in law reduces the defendant’s statutory maximum sentence below the imposed sentence”). In Begay and Chambers, the Supreme Court resolved an open question regarding the kinds of crimes that fall within the meaning of “violent felony” under the ACCA, and, by implication, the career offender guideline. See Templeton, 543 F.3d at 380. The Court’s pronouncements did not simply constitute an “intervening change in the law” in the traditional sense. Rather, the decisions make clear that, at the time of his sentencing, Mr. Narvaez did not qualify as a career offender under the guidelines. An additional five years of incarceration was imposed upon him without any legal justification. Such gratuitous infliction of punishment is a fundamental defect in the court’s judgment that clearly constitutes a complete miscarriage of justice and a violation of due process.  (9-11)

The Seventh Circuit’s reasoning here seem quite broad and might open up collateral relief any time any time the Supreme Court adopts a narrowing construction of a guidelines provision.  There is perhaps a limitation, though, to sentences imposed when the guidelines were mandatory.  Consider this argument by the government and Seventh Circuit’s response:

The Government submits, however, that the sentencing court’s error in this case does not warrant § 2255 relief. Unlike the situation under the ACCA, Mr. Narvaez’s 170- month sentence was actually within the authorized 20-year statutory maximum for his crime. Therefore, the Government reasons that, because Mr. Narvaez would be exposed to the full range of punishment authorized by Congress for his crime at resentencing, and would remain eligible for the identical 170-month sentence under the advisory guidelines, his claim does not present a fundamental defect.

We cannot accept this argument. The fact that Mr. Narvaez’s sentence falls below the applicable statutory-maximum sentence is not alone determinative of whether a miscarriage of justice has occurred. The sentencing court’s misapplication of the then-mandatory § 4B1.1 enhancement in Mr. Narvaez’s case was central to its guidelines calculation. Nothing in the record indicates that the court would have sentenced Mr. Narvaez to five additional years of incarceration had the judge not been under the legal misapprehension, shared by the rest of the circuit, that Mr. Narvaez was a career offender and that the corresponding guidelines required such an enhancement. The application of the career offender provision increased the sentencing range for Mr. Narvaez. Speculation that the district court today might impose the same sentence is not enough to overcome the fact that, at the time of his initial sentencing, Mr. Narvaez was sentenced based upon the equivalent of a nonexistent offense. This error clearly constitutes a miscarriage of justice and a due process violation.  (12-13)

Whether and under what circumstances the retroactivity ruling of Narvaez would apply to defendants sentenced post-Booker thus remains an open question.

Cross posted at Life Sentences Blog.

What many observers took from Begay and Chambers is that a prior conviction does not count under the ACCA unless it satisfies both a subjective test (purposeful, violent, and aggressive) and an objective test (statistically demonstrated likelihood of injury).

But, today, in Sykes v. United States (No. 09-11311), the Court threw this understanding into doubt, suggesting a considerably more expansive interpretation of the residual clause.

In Sykes, the Court held that the Indiana crime of vehicular flight from a law enforcement officer counts as a violent felony.  Here’s the way the Court framed the issue:

The question, then, is whether Indiana’s prohibition on flight from an officer by driving a vehicle—the violation of Indiana law for which Sykes sustained his earlier conviction—falls within the residual clause because, as a categorical matter, it presents a serious potential risk of physical injury to another. The offenses enumerated in §924(e)(2)(B)(ii)—burglary, extortion, arson, and crimes involving use of explosives—provide guidance in making this determination.  (6)

Even in this preliminary statement of the case, the Court notably omitted any reference to the “purposeful, violent, and aggressive” language of Begay, suggesting that the objective test of Chambers did not so much supplement the subjective test as replace it.

In deciding that Sykes’s crime was sufficiently dangerous to count, the Court emphasized what it felt was the inherent character of the crime, rather than relying primarily on data:

Risk of violence is inherent to vehicle flight. Between the confrontations that initiate and terminate the incident, the intervening pursuit creates high risks of crashes. It presents more certain risk as a categorical matter than burglary. It is well known that when offenders use motor vehicles as their means of escape they create serious potential risks of physical injury to others. Flight from a law enforcement officer invites, even demands, pursuit. As that pursuit continues, the risk of an accident accumulates. And having chosen to flee, and thereby commit a crime, the perpetrator has all the more reason to seek to avoid capture.

Unlike burglaries, vehicle flights from an officer by definitional necessity occur when police are present, are flights in defiance of their instructions, and are effected with a vehicle that can be used in a way to cause serious potential risk of physical injury to another.  (7-8)

Absent here is any sense that the government must prove dangerousness in a rigorous, empirical fashion, as was suggested byChambers and clearly endorsed by Justice Scalia in his concurring opinion in Begay.

Although the government did, in fact, provide some decent statistical evidence in Sykes, the Court seemed to treat this evidence as secondary to “common sense”:

Although statistics are not dispositive, here they confirm the commonsense conclusion that Indiana’s vehicular flight crime is a violent felony. See Chambers, 555 U. S., at 129 (explaining that statistical evidence sometimes “helps provide a conclusive . . . answer” concerning the risks that crimes present).  (8)

So what of the purposeful, violent, and aggressive test?  As I discussed in this post, the “aggressive” component seemed to be an especially big sticking point for some of the justices at oral argument.  But those concerns are simply nowhere to be seen in the majority opinion.  Rather, the Court — in a not very clear fashion — seemed to want to limit or modify the Begay test:

Sykes argues that, regardless of risk level, typical vehicle flights do not involve the kinds of dangers that the Armed Career Criminal Act’s (ACCA) residual clause demands. In his view this Court’s decisions in Begay and Chambers require ACCA predicates to be purposeful, violent, and aggressive in ways that vehicle flight is not. Sykes, in taking this position, overreads the opinions of this Court.

ACCA limits the residual clause to crimes “typically committed by those whom one normally labels ‘armed career criminals,’” that is, crimes that “show an increased likelihood that the offender is the kind of person who might deliberately point the gun and pull the trigger.” Begay, 553 U. S., at 146. In general, levels of risk divide crimes that qualify from those that do not. See, e.g., James, 550 U. S. 192 (finding attempted burglary risky enough to qualify). Chambers is no exception. 555 U. S., at ___–___ (slip op., at 5–6) (explaining that failure to report does not qualify because the typical offender is not “significantly more likely than others to attack, or physically to resist, an apprehender”).

The sole decision of this Court concerning the reach of ACCA’s residual clause in which risk was not the dispositive factor is Begay, which held that driving under the influence (DUI) is not an ACCA predicate. There, the Court stated that DUI is not purposeful, violent, and aggressive. 553 U. S., at 145–148. But the Court also gave a more specific reason for its holding. “[T]he conduct for which the drunk driver is convicted (driving under the influence) need not be purposeful or deliberate,” id., at 145 (analogizing DUI to strict-liability, negligence, and recklessness crimes). By contrast, the Indiana statute at issue here has a stringent mens rea requirement. Violators must act “knowingly or intentionally.” Ind. Code §35–44– 3–3(a); see Woodward, 770 N. E. 2d, at 901 (construing the statute to require “a knowing attempt to escape law enforcement” (internal quotation marks omitted)).

The phrase “purposeful, violent, and aggressive” has no precise textual link to the residual clause, which requires that an ACCA predicate “otherwise involv[e] conduct that presents a serious potential risk of physical injury to another.” §924(e)(2)(B)(ii). The Begay phrase is an addition to the statutory text. In many cases the purposeful, violent, and aggressive inquiry will be redundant with the inquiry into risk, for crimes that fall within the former formulation and those that present serious potential risks of physical injury to others tend to be one and the same. As between the two inquiries, risk levels provide a categorical and manageable standard that suffices to resolve the case before us.

Begay involved a crime akin to strict liability, negligence, and recklessness crimes; and the purposeful, violent, and aggressive formulation was used in that case to explain the result. The felony at issue here is not a strict liability, negligence, or recklessness crime and because it is, for the reasons stated and as a categorical matter, similar in risk to the listed crimes, it is a crime that “otherwise involves conduct that presents a serious potential risk of physical injury to another.” §924(e)(2)(B)(ii).  (10-11)

So where does this leave us?  ”In general, levels of risk divide crimes that qualify from those that do not,” the Court said.  Objective dangerousness thus seems to be the “general” test to use; other considerations will come into play, at most, in unusual circumstances.  Such circumstances may be present when there is “a crime akin to strict liability, negligence, and recklessness crimes.”  But a felony with a “knowing” mens rea, like Indiana’s vehicular fleeing offense, will not require anything more than objective dangerousness in order to count as an ACCA predicate.

Even as to crimes in the strict liability/negligence/recklessness category, it is not clear that the Court is still committed to the “purposeful, violent, and aggressive” test.  The majority went out of its way to highlight the test’s lack of a “textual link” and to observe that Begay itself was the Court’s “sole decision” to rely on the test.  Moreover, in noting that “[i]n many cases the purposeful, violent, and aggressive inquiry will be redundant with the inquiry into risk,” the Court may be signaling an interest in collapsing the one test into the other.

Bottom line: it’s hard to say whether we will be seeing the Court use the “purposeful, violent, and aggressive” formulation ever again.  This seems a remarkable place to end up in Sykes, given how much effort was expended in the briefs and oral argument on the question of whether vehicular flight satisfied the test — which everyone, including the government, seemed to assume would apply.

In her dissenting opinion, Justice Kagan noted this issue, but seemed more sanguine about the future of the “purposeful, violent, and aggressive” test:

I understand the majority to retain the “purposeful, violent, and aggressive” test, but to conclude that it is “redundant” in this case. See ante, at 11. Like JUSTICE SCALIA, see ante, at 3 (dissenting opinion), I find this conclusion puzzling. I do not think the majority could mean to limit the test to “strict liability, negligence, and recklessness crimes.” Ante, at 11 (majority opinion). As JUSTICE SCALIA notes, see ante, at 3, that would be to eliminate the test’s focus on “violence” and “aggression.” And it would collide with Chambers v. United States, 555 U. S. 122 (2009)—a decision the majority cites approvingly, see ante, at 8—which applied the test to an intentional crime. See 555 U. S., at 128 (opinion of the Court), 130 (Appendix A to opinion of the Court) (holding that “knowin[g] fail[ure] to report to a penal institution” does not involve “purposeful, violent, or aggressive conduct” (internal quotation marks omitted)). So I assume this test will make a resurgence—that it will be declared non-redundant—the next time the Court considers a crime, whether intentional or not, that involves risk of injury but not aggression or violence.  (Kagan dissent, 2 n.1)

Justice Scalia’s dissent is quite remarkable and deserves much attention for its aggressive use of the “void for vagueness” doctrine.  Here is what he says:

Today’s opinion, which adds to the “closest analog” test (James) the “purposeful, violent, and aggressive” test (Begay), and even the risky-as-the-least-risky test that I had proposed as the exclusive criterion, has not made the statute’s application clear and predictable. And all of them together—or even the risky-as-the-least-risky test alone, I am now convinced—never will. The residual-clause series will be endless, and we will be doing ad hoc application of ACCA to the vast variety of state criminal offenses until the cows come home.

That does not violate the Constitution. What does violate the Constitution is approving the enforcement of a sentencing statute that does not “give a person of ordinary intelligence fair notice” of its reach, United States v. Batchelder, 442 U. S. 114, 123 (1979) (internal quotation marks omitted), and that permits, indeed invites, arbitrary enforcement, see Kolender, 461 U. S., at 357. The Court’s ever-evolving interpretation of the residual clause will keep defendants and judges guessing for years to come. The reality is that the phrase “otherwise involves conduct that presents a serious potential risk of physical injury to another” does not clearly define the crimes that will subject defendants to the greatly increased ACCA penalties. It is not the job of this Court to impose a clarity which the text itself does not honestly contain. And even if that were our job, the further reality is that we have by now demonstrated our inability to accomplish the task.  (Scalia dissent, 7)

His dissent closes with a well-deserved shot across Congress’s bow:

We face a Congress that puts forth an ever-increasing volume of laws in general, and of criminal laws in particular. It should be no surprise that as the volume increases, so do the number of imprecise laws. And no surprise that our indulgence of imprecisions that violate the Constitution encourages imprecisions that violate the Constitution. Fuzzy, leave-the-details-to-be-sorted-out-by-the-courts legislation is attractive to the Congressman who wants credit for addressing a national problem but does not have the time (or perhaps the votes) to grapple with the nitty-gritty. In the field of criminal law, at least, it is time to call a halt. I do not think it would be a radical step—indeed, I think it would be highly responsible—to limit ACCA to the named violent crimes. Congress can quickly add what it wishes. Because the majority prefers to let vagueness reign, I respectfully dissent.  (9)

Cross posted at Life Sentences Blog.

Count the Court’s decision today in McNeill v. United States (No. 10-5258) in the latter category.

Here’s the background on McNeill from an earlier post:

McNeill was convicted of being a felon in possession of a firearm. He had two prior convictions for violent felonies. In order to get a third ACCA predicate, the government pointed to McNeill’s drug trafficking convictions in North Carolina in 1992 and 1995. At the time he committed those offenses, North Carolina law specified a maximum sentence of ten years for each. Thus, at first blush, the convictions seem to fall pretty clearly within the ACCA’s definition of “serious drug offense”: “an offense under State law . . . for which a maximum term of imprisonment of ten years or more is prescribed by law.”

But here’s the catch: North Carolina changed its sentencing law, and the crimes for which McNeill was convicted now carry a maximum of only 25 months in prison. If McNeill did today exactly what he did before, the resulting convictions would plainly not count as ACCA predicates.

The change-in-law problem has produced a circuit split, which the Supreme Court will now presumably resolve.

McNeill relies on the use of the present tense in the statutory definition of “serious drug offense”: “ten years or more is prescribed.” He also argues that the ACCA was intended to defer to state legislative judgments regarding offense severity — the North Carolina legislature now apparently believes that McNeill’s crimes were not all that serious, and federal courts applying the ACCA should respect that judgment.

On the other hand — and the Fourth Circuit seemed to think this was crucial in rejecting McNeill’s arguments — the legislature did not make the reduced penalties retroactively applicable to conduct committed before the effective date of the sentencing reform law. Because McNeill’s convictions were based on things he did before the effective date, he would apparently be subject to the same ten-year maximum even if he was just being prosecuted now for what he did in the 1990’s.

The Supreme Court affirmed McNeill’s ACCA sentence, but adopted a slightly different approach than the Fourth Circuit, holding that the sentencing scheme on the date of the earlier conviction governs without regard even to later retroactive changes in the law. 

Justice Thomas, writing for a unanimous Court, treated the question as a simple matter of “plain text”:

The statute requires the court to determine whether a “previous conviction” was for a serious drug offense.  The only way to answer this backward-looking question is to consult the law that applied at the time of the conviction.

This is the “only way” to interpret the statute?  How can it be that easy when both the Second and the Fourth Circuits interpreted it differently?

The Court further reasoned that its “natural reading of ACCA also avoids the absurd results that would follow from consulting current state law to define a previous offense.”  The Court seemed particularly concerned that, under McNeill’s interpretation, a prior conviction might “disappear” for ACCA purposes if a state ”reformulates its criminal statutes in a way that prevents precise translation of the old conviction into the new statutes.” 

But “translation” difficulties — deciding which state-law offenses fit into the federal-law definitions of “violent felony” and “serious drug offense” — are endemic to the ACCA regime, and the case law draws arbitrary distinctions all the time between what counts and what doesn’t count.  Against the backdrop of a poorly conceived and drafted statute and an incoherent body of precedent, “absurdity” hardly seems an appropriate interpretive criterion.

When we decide that a prior conviction counts as an ACCA predicate, we are functionally punishing the defendant a second time in federal court for the earlier state conviction.  In this context, it may be especially appealing to look to the rule of lenity — the principle that statutory ambiguities are resolved in favor of the defendant — as a way to sort out the many uncertainties that arise in applying the ACCA.

Cross posted at Life Sentences Blog.

This was my first time appearing before the Commission, and I was quite impressed by how engaged and well-prepared the Commissioners were.  Through a long morning of testimony by a dozen witnesses, the Commissioners asked many questions, and not one of the questions seemed ill-conceived or poorly articulated.  They had obviously read with real care the written submissions by the witnesses, and they went right at the key problems with each witness’s position.  Testifying was like oral argument before an exceptionally good appellate panel.

The Commissioners seemed pretty clearly inclined to make Parts A and C of the amendment package retroactive.  (Part A reduces sentences for crack offenders based on drug quantity; it’s another two points from the offense level for many current prisoners, as also happened with the  2007 amendment.  Part C ends the treatment of simple possession of crack as a trafficking offense.)  On several occasions, one commissioner or another referred to a “consensus” in favor of retroactivity for A and C.  That view, in some form or another, was favored even by most of the witnesses who were there to represent law-enforcement perspectives.

But it’s not clear how far the retroactivity decision will reach.  Attorney General Holder led off the hearing with something of a surprise from the Department of Justice.  The Department is taking a position in favor of retroactivity, but with two major exclusions: offenders in criminal history categories IV-VI or with a weapons enhancer (guidelines or 924(c)).  This would exclude about half of the offenders who are otherwise eligible for a sentence reduction.

Although the Department obviously carries a lot of weight with the Commission, I don’t think the Commission is going to go for the exclusions.  I was surprised by how uniformly and sharply critical the Commissioners were in their questioning regarding the exclusions.  (There was a nice lesson in the politics of these things.  The AG led off with a prepared statement announcing the Department’s position, then excused himself, leaving an unlucky AUSA behind to field the Commission’s tough questions.)  One major objection is that the excluded offenders already had their criminal history and weapons involvement taken into account in setting their original sentences, and those enhancements will remain in place even if their sentences are now modified.  But if they are denied the opportunity to benefit from the new drug quantity table, their criminal history or weapons involvement will effectively be double-counted against them.

I spoke extemporaneously against the Department’s proposed exclusions in my testimony.  The main reason offered by the Department in favor of the exclusions is that prosecutors do not have the time to assess and litigate dangerousness on a case-by-case basis; therefore, public safety can only be protected if crude proxies for dangerousness are used.  I pointed out, though, that most eligible inmates will not have a release date until at least year three after the amendment, so there’s no real hurry in most cases.  Prosecutors should have plenty of time to triage and then take a closer look at the cases with high criminal history or other indicators of dangerousness.  If judges are pressing too hard, I suggested the Commission could, in consultation with the Department, issue recommendations for court scheduling and case prioritization so as to permit adequate case-by-case assessment.  If this is unsatisfactory, I suggested that the Commission might consider a tiered approach to setting the effective date for retroactivity; for instance, the effective date might be set an extra six months out for offenders in criminal history categories V and VI, giving an opportunity for the system first to clear out a lot of the easy cases.

I also pointed out that brand-new recidivism data (release by the Commission just yesterday afternoon) shows that offenders in CH category IV who were released under the 2007 amendment actually have a lower recidivism rate than offenders in CH category III.  It’s a bit odd, then, to invoke public safety as a reason for making IV the cutoff.

Finally, I argued that vicarious liability makes mere weapons possession an exceptionally poor proxy for dangerousness — you can get a weapons enhancement based merely on the reasonable foreseeability that one of your criminal confederates might possess a gun.  If weapons involvement is used as a basis for exclusion, I suggested that the exclusion focus on brandishing or more aggravated use under 924(c) rather than mere possession.

Even if the Department’s exclusions are rejected, some of the Commissioners seemed surprisingly interested in making the new aggravating specific offense characteristics of Part B retroactive.  This would have the effect of precluding or limiting sentence reductions for crack offenders who are subject to one of the new aggravators, e.g., use of violence, bribery of a law-enforcement officer, use of a minor.  This could make the sentence-modification proceedings considerably more cumbersome, as new fact-finding would be required.  (Parts A and C could be implemented retroactively based on existing paper records from the initial sentnencing.)

I opposed retroactivity for the new SOCs in my written testimony and reiterated the main points in my oral testimony today.  Even if not made formally retroactive, they can still be considered by district judges as a basis for denying or reducing the size of a sentence modification.  The real question for retroactivity is whether district judges will be required to attach a particular weight to the new SOCs, or whether they can be assessed in a more flexible, discretionary way.  Consistent with the general premises of post-Booker federal sentencing, I favor the more discretionary approach.  I think this is especially important in light of the fact that some of the “new” aggravators may have already been used to set a high sentence within a guidelines range; retroactivity thus creates a potential double-counting problem.

Cross posted at Life Sentences.

In holding that post-sentencing rehabilitation is a permissible consideration at resentencing, the Court addressed a couple of notable legal questions.  What is perhaps most remarkable about Pepper, however, is not the legal analysis, but the prefatory rhetoric with which it was framed.

In brief, the Court offered what Justice Alito, dissenting in part, characterized as a “paen to that old regime” of highly discretionary, individualized sentencing.  For instance, here is the very first sentence of Justice Sotomayor’s opinion for the majority:

The Court has long recognized that sentencing judges “exercise a wide discretion” in the types of evidence they may consider when imposing sentence and that “[h]ighly relevant — if not essential — to [the] selection of an appropriate sentence is the possession of the fullest information possible concerning the defendant’s life and characteristics.”  Williams v. New York, 337 U.S. 241, 246-47 (1949).

The citation to Williams is telling.  The 1949 case is widely seen as emblematic of the Court’s hands-off approach to sentencing issues in the era when criminal law was dominated by the rehabilitative ideal — sentencing judges had to be given almost unlimited discretion at sentencing in order to weigh each defendant’s treatment needs and prospects.

The Pepper Court went on at some length regarding this tradition of discretionary, individualized sentencing before getting to the real issues in the case.  That tradition, of course, predates the Sentencing Reform Act of 1984, which was intended to overturn the Williamsregime.  The Court then overturned the SRA in 2005 in Booker v. United States.  But Booker didn’t offer a clear alternative vision to the SRA, and subsequent cases have hardly embraced the pre-SRA,Williams-era model.

Instead, what seemed to be emerging from the Court’s 2007 trilogy of Rita, Kimbrough, and Gall was a pragmatic vision of sentencing as a shared responsibility of district judges and the Sentencing Commission, with appellate courts recognizing and respecting the particular institutional strengths and weaknesses of both — sometimes emphasizing deference to the district court and sometimes emphasizing deference to the Commission, as appropriate in the circumstances.  This vision is quite close in spirit to the basic sentencing framework embraced by the Court in 1996 in Koon v. United States, which relied on an earlier First Circuit decision (United States v. Rivera) authored by then-Judge Breyer.  Breyer, I suspect, continues to be the Court’s chief proponent for this vision, which also comes through in his opinion for the majority in Rita.  Breyer’s concurring opinion in Pepper once again pushes this vision.

But no one joined Breyer’s opinion in Pepper — that is surprising to me.

Although there are now quite a few Booker progeny cases, I can’t recall any so whole-heartedly embracing district-judge discretion as an affirmative value and treating Sentencing Commission expertise so dismissively.  I wonder if Justice Sotomayor is driving this.  Not only is she the author of Pepper, but she is the only justice who participated in Pepper who was not around for the 2007 trilogy that gave greater weight to Commission expertise.

Is there a fundamental shift taking place in the Court’s thinking about federal sentencing?  I suppose that may depend in part on where Justice Kagan comes out, which remains a mystery — she sat out Pepper.

Onto the legal analysis.  The main question in Pepper was whether 18 U.S.C. § 3742(g)(2) survived Booker.  The Court held that it did not.  The statutory provision prohibits a sentence outside the applicable Guidelines range on resentencing unless it is based on a ground that was “specifically and affirmatively included in the written statement of reasons” provided for the original sentence.  Although the provision says nothing on its face about post-sentencing rehabilitation, the provision effectively precludes a below-Guidelines sentence based on that factor since post-sentencing rehabilitation, by definition, cannot be mentioned by the judge at sentencing.  In Pepper, the Court held that § 3742(g)(2) violated Booker since it has the effect of making the Guidelines mandatory in certain cases.  For instance, if a judge declined to sentence below the Guidelines because the judge erroneously believed that she could not consider a particular mitigating factor, and the judge was then reversed on appeal, the judge would still be required by  § 3742(g)(2) to impose the Guidelines sentence on remand — the advisory Guidelines would then become effectively mandatory by virtue of  § 3742(g)(2).  Although the Pepper Court might have made some distinctions and held the statute constitutional in some circumstances but not others, the Court chose as a matter of administrative convenience to throw out  § 3742(g)(2) entirely.

And good riddance to § 3742(g)(2) — a sorry relic from Congress’s misguided assault on judicial discretion in the 2003 PROTECT Act.

Note, however, that jettisoning § 3742(g)(2) does not necessarily work to the advantage of defendants.  Now, not only may the judge take post-sentencing rehabilitation into account, but also post-sentencing misconduct that potentially warrants a sentence above the Guidelines.  Defendants will be well advised to be on their best behavior while their appeals are pending!

The other big question the Pepper Court had to deal with was how much weight, if any, to give to the Sentencing Commission’s statement in § 5K2.19 of the Guidelines that post-sentencing rehabilitation is not an appropriate ground for a below-Guidelines sentence.  To my mind, the Court was surprisingly dismissive of the Commission’s views:

[A] district court may in appropriate cases impose a non-Guidelines sentence based on a disagreement with the Commission’s views.  That is particularly true where, as here, the Commission’s views rest on wholly unconvincing policy rationales not reflected in the sentencing statutes Congress enacted.

“Wholly unconvincing?”  That’s strong language, but probably deserved here.  I imagine the phrase will now be used extensively in briefs by defendants in other cases arguing that other aspects of the Guidelines, such as the child pornography provision, should not be followed by sentencing judges.

In the end, the Court did not provide much clear guidance on when sentencing judges may reject the Commission’s policy choices, but the general tone of the opinion seems to suggest that judges have more freedom in this regard than some appellate courts have been assuming.

So, Pepper will get yet another resentencing, at which his post-sentencing rehabilitation may be considered — the Court gave its permission, but did not require the resentencing judge to attach any particular significance to his impressive record in turning his life around.

Cross posted at Life Sentences Blog.

Their thesis is simply stated.  A vastly disproportionate number of federal death sentences come from counties with high minority populations that are located in districts that are heavily white overall.  Think diverse urban cores surrounded by lily-white suburbs.  Given that federal juries are typically drawn from the entire district, this means that capital trials in these districts are apt to involve minority defendants being judged by white-dominated juries.  Having minimal racial diversity on the jury means that black defendants have little protection from the unconscious racial biases that most of us carry around.  This, in turn, drives both the racial and geographic disparities in federal death sentences.

The patterns are striking. 

For instance, both federal districts in Missouri display the racial demographics that are of interest to Cohen and Smith (racially diverse urban county surrounded by heavily white suburban counties), and Missouri has returned more federal death sentences than New York, California, and Florida combined (p. 436).  In fact, Cohen and Smith contend that all eight of the districts that have returned more than two federal death sentences exhibit pronounced county-district racial disparities.

By contrast, the three districts in which it has been hardest for the feds to get a death sentence are all majority-minority: District of Columbia, Puerto Rico, and the Southern District of New York.  ”These three federal districts account for 55 of the 460 death-authorized cases but are not responsible for a single death sentence” (465).  Expanding the view to the ten districts in which it has been hardest to get a death sentence, eight have “similar demographic profiles between the federal district and the most populous county.”

Cohen and Smith have uncovered a fascinating pattern, although it surely does not tell the whole story.  What about the Eastern District of Wisconsin, for instance?  While the largest city in the District, Milwaukee, is 37 percent black, the District as a whole is only nine percent black.  This disparity would seem to put the District at considerable risk for the racial dynamics that are of concern to Cohen and Smith, but we have no death sentences.  In fact, the numbers for Milwaukee and the Eastern District of Wisconsin are almost identical to the numbers for Kansas City and the Western District of Missouri, which leads the nation in federal death sentences.

Indeed, while I haven’t cranked the numbers, I strongly suspect there are a great many other death-free districts with similar profiles to these two.  High county-district racial disparities may be necessary for federal death sentences, but I doubt they are sufficient.

Even at that, what Cohen and Smith have uncovered should heighten concerns about the role of racial bias in the administration of the federal death penalty.  For that reason, their reform proposals (especially drawing the venire for federal capital trials from the county of the offense, as federal law mandated prior to the Civil War) deserve attention.

I wonder, too, if the race-geography dynamics they have uncovered are apparent more broadly in federal criminal trials.  If racial bias is a problem in capital trials in some districts, why would it not also be a problem in noncapital trials?  As federal law enforcement has become more oriented to responding to street crime, which is really a local problem, it makes sense for federal juries to be drawn on a more local basis, too.

Cross posted at Life Sentences Blog.

The Court created the state-of-mind problem in Begay v. United States, 553 U.S. 137 (2008), which held that a prior conviction does not count as a “violent felony” under the ACCA unless the crime was “purposeful, violent, and aggressive.”  This is a rather mysterious phrase.  Although the word “purposeful” is a familiar culpability term, it is not clear what “violent” and “aggressive” are meant to connote in this context.  And even “purposeful” has some ambiguity, as any law student who has ever wrestled with the elusive distinction between “general intent” and “specific intent” will tell you.

Begay itself indicated that DUI does not satisfy the PVA test because DUI is a strict liability offense.  This teaches that some culpability is indeed required for an offense to count as a “violent felony,” but Begay provided little guidance beyond that.

Then came Chambers v. United States, 129 S. Ct. 687 (2009).  

Chambers held that the crime of failing to report to serve a sentence did not count as a violent felony.  The Court’s relatively brief analysis paid very little attention to state-of-mind considerations, but focused more on objective dangerousness, including data showing low levels of violence associated with the offense.  Indeed, the Court approvingly cited Justice Scalia’s concurring opinion in Begay, which rejected a culpability requirement in favor of a purely objective approach.  Remarkably, the Court made no mention of the subjective-objective debate in Begay. Based on the citation to Scalia’s concurrence and the use of an objective approach in Chambers, I’ve wondered whether the Court is quietly moving towards Scalia’s view.  Scalia himself did not write in Chambers, suggesting that he is satisfied with the Court’s trajectory from Begay to Chambers.

But there is another possibility.  Rather than selecting an objective approach over a subjective approach, Chambers may instead tacitly endorse a two-prong test: to count as an ACCA predicate, an offense must satisfy both a subjective culpability requirement and an objective dangerousness requirement.  Because the offense in Chambers failed the objective test, there was no need to get into the subjective test – on this view, the Court’s silence on culpability did not indicate that culpabilty has fallen out of the equation generally, but only on the specific facts of Chambers.

Although I’ve not seen much express consideration of the question, my impression is that the lower federal courts have generally understood Chambers in the latter, both/and sense.

In the immediate aftermath of Begay, I wrote here in favor of Scalia’s purely objective approach.  Now I’m not so sure.  For one thing, Scalia’s opinion was framed as an either/or choice between the objective and subjective approaches.  If the question is presented that way, I suppose I would still favor the objective approach as more consistent with the statutory language.  But, as a third option, the two-pronged approach does have some appeal.  In the past, I (like Scalia) may have been a little too dismissive of the culpability requirement as lacking a clear textual basis.

Here’s the relevant statutory language:

[T]he term “violent felony” means any crime punishable by imprisonment for a term exceeding one year . . . that . . . is burglary, arson, or extortion, involves use of explosives, or otherwise involves conduct that presents a serious potential risk of physical injury to another . . . .

The italicized portion constitutes the so-called residual clause, the interpretation of which has been the task of Begay, Chambers, and now Sykes. The way that this clause is paired up with three familiar crimes that do have substantial culpability requirements (burglary, arson, and extortion) provides support for interpreting the residual clause also to have a culpability requirement.  Doing so is also consistent with the evident purpose of the ACCA, which is to deliver extended incapacitation to the most dangerous offenders; recidivists whose record indicates a pattern of intentionally dangerous or harmful conduct would seem more suitable for incapacitation than those who have only acted, say, negligently.  Yet, a purely objective approach might draw serial carelessness into the ACCA’s incapacitation scheme.

Indeed, if we are to ascribe any significance to the title of the statute, the phrase “career criminal” — consistent with the references to burglary, arson, and extortion — connotes a repeat offender who is committing crime for a living, i.e., is motivated to achieve pecuniary gain by wrongfully depriving others of their property.  Again, this points to an interpretation of the residual clause that would require a truly blameworthy state of mind (although, to be fair, I’m not so sure Begay’s “purposeful, violent, and aggressive” standard precisely captures the nature of the culpability suggested by the phrase “career criminal”).

There may also be some appeal to a subjective prong insofar as it permits some priors to be knocked out as ACCA predicates quickly and easily without requiring the extensive litigation that might be required to answer the complex empirical question of how dangerous a given category of crime is.  This was the dynamic in Begay itself, in which the Court used the culpability requirement as a way to avoid answering the uncertain question of how dangerous is the typical instance of DUI.  Although I’ve not yet noticed any indication of this in the case law, such questions of objective dangerousness seem to invite the use of expert witnesses on statistics and criminology.  Again, judges and lawyers alike might appreciate the ability to resolve the status of some priors without going down the path of such complex litigation.

Whatever the merits of a subjective prong, Sykes may provide a good opportunity for the Court to clarify whether such a prong exists and, if so, what exactly the PVA standard means.

Sykes was convicted of the Indiana crime of using a vehicle while knowingly or intentionally fleeing from a law enforcement officer after being ordered to stop.  Was this a “violent felony”?  The Seventh Circuit held yes, although at least one other circuit has concluded that a similar crime was not.

In Chambers, the Court was able to avoid answering any hard questions about the culpability required for an offense to count as a “violent felony” because the offense at issue in that case did not satisfy the requirement of objective dangerousness.  The Court is less likely to reach a similar conclusion in Sykes.

Having read the government’s brief, I’m impressed that there is a much more solid case for objective dangerousness in Sykes than inChambers.  Not airtight, mind you, but still pretty strong.

The key to the government’s argument is the assumption that police typically give chase when a motorist refuses an order to pull over, because there is a fair bit of data showing that police pursuit is a rather dangerous proposition.  And it does seem at least plausible that pursuit typically ensues, although I suppose it is also possible that (a) many motorists who initially refuse to pull over think better of it pretty quickly, obviating the need for pursuit; and (b) that many orders to pull over are for such minor infractions that officers (either as a matter of their own discretion or departmental policy) decline to initiate pursuit.  I don’t think the government has nailed down how common (a) and (b) are.  Still, the dangerousness argument here seems considerably stronger than in Chambers, which means that the Court is likely to have to confront the state-of-mind questions.

The threshold question is whether the Begay culpability requirement survives Chambers.  Assuming it does, the Court may then have an opportunity to explain more clearly what state of mind is implied by Begay’s ”purposeful, violent, and aggressive” test.

The government’s position on culpability seems to be that it is enough that Sykes was convicted of a crime having as an element that he “knowingly or intentionally” fled from a law enforcement officer.

Sykes, however, cites Judge Posner’s cogent dissent in another ACCA case involving a similar Illinois fleeing statute, Welch v. United States, 604 F.3d 408, 434 (2010).  Here is Posner’s position:

Thus I don’t agree that just because the defendant intended to flee from the police his action was “purposeful” within the meaning of the Supreme Court’s formula. Given that the purpose of the catch-all provision in the Armed Career Criminal Act is to enable courts to identify crimes that are similar to the enumerated ones, “purposeful” should be interpreted to mean trying to harm a person’s person or property, which is characteristic of the enumerated crimes. Burglary requires proof of intent to commit a crime following unlawful entry, arson proof of intent to destroy property without legal authority, extortion proof of intent to obtain another person’s property by a threat. These crimes do not merely create a risk of harm, as aggravated fleeing does.

It’s not that the enumerated crimes necessarily are motivated by a desire to hurt anyone. The criminal may simply want the victim’s property. (In an arson case, the victim is often an insurance company-and arsonists don’t dislike insurance companies-rather the contrary!) But to get what he wants he has to harm the victim; purpose to harm is intrinsic to the crime although it often is not the motive. That is not true in a flight case any more than it is true in a DUI case. In both the perpetrator is behaving in a dangerous manner but in neither is he trying to take anything from anyone or otherwise harm anyone. And “although the [fleeing] statute does require intent, the required mental state is only intent to be free of custody, not intent to injure or threaten anyone. It is easy to violate [such a statute] without intending or accomplishing the destruction of property or acting in an aggressive, violence-provoking manner that could jeopardize guards or bystanders.” United States v. Templeton, supra, 543 F.3d at 383.

Dismuke calls aggravated fleeing “aggressive” because it involves defiance of authority. 593 F.3d at 595. But that is true of all escapes-the point of Chambers was that we can’t treat all escapes alike. Dismuke calls fleeing “active” rather than “passive,” id., but all actions are active, yet most are not aggressive. To fail to report to prison when ordered is to defy the authority of the sentencing judge and the Bureau of Prisons. Many acts of civil disobedience are emphatically active and defiant of authority at their core, but are miles away from being aggressive. Dismuke notes the possibility of a future confrontation with authority but that is also a likely consequence of a walkaway escape or a failure to report and is an especially likely consequence of many acts of civil disobedience. Adopting a rule that would exclude from the category of “violent felony” a crime that while it may be dangerous does not involve any intention of harming anyone would go some distance toward clarifying the meaning of “violent felony” and by doing so perhaps check the avalanche of litigation over that meaning.

Assuming the Court plans to stick with a culpability requirement, Posner’s formulation strikes me as quite sensible and consistent with the best understanding of the statute’s purpose.

After being convicted of alien smuggling and bail jumping, Tapia was sentenced to 51 months in prison.  The judge made clear that the sentence was based, at least in part, on what the judge anticipated would be necessary for Tapia to complete a drug treatment program:

I am going to impose a 51-month sentence[:] 46 months [for smuggling] plus five months for the bail jump[.]  [O]ne of the factors that affects this is the need to provide treatment.  In other words, so she is in long enough to get the 500 Hour Drug Program, number one.

The dispute over the permissibility of the judge’s reasoning has its roots in the origins of the Sentencing Reform Act of 1984.  

As I discuss in detail in this article, the idea of the federal sentencing guidelines system emerged from intense dissatisfaction with the rehabilitative paradigm that dominated American penology in the middle decades of the twentieth century.  Careful reviews of the social scientific literature in the 1970’s purportedly demonstrated that “nothing works” — no available treatment programs for criminal offenders could be proven to reduce recidivism by any significant degree.

Sen. Ted Kennedy, the leading legislative proponent of sentencing reform in the 1970’s, plainly agreed with the “nothing works” thesis.  For instance, an early version of his reform bill (S. 181 in 1977) pointedly identified only three permissible purposes of punishment: deterrence, incapacitation, and desert.  Rehabilitation is off the table.

But the legislative process is messy, and Kennedy’s skepticism of rehabilitation was not expressed so clearly in the final version of the bill enacted in 1984.  Thus, 18 U.S.C. § 3553(a)(2)(D) declares that one permissible purpose of a sentence is “to provide the defendant with needed educational or vocational training, medical care, or other correctional treatment in the most effective manner.”

On the other hand, 18 U.S.C. § 3582(a) imposes what might at first seem a contrary mandate:

The court, in determining whether to impose a term of imprisonment, and, if a term of imprisonment is to be imposed, in determining the length of the term, shall consider the factors set forth in section 3553(a) to the extent that they are applicable, recognizing that imprisonment is not an appropriate means of promoting correction and rehabilitation.  (emphasis added)

One plausible way of reconciling these provisions — adopted, for instance, by the Third Circuit — is this: rehabilitation is not a proper purpose of imprisonment, but it is a permissible purpose for other sorts of sentences, such as probation.  Under this reading, Tapia’s prison sentence would be improper.

Another reading, though, would cast the rehabilitative purpose ban of § 3582(a) in much narrower terms.  Although imprisonment per semay not be viewed permissibly as rehabilitative — it’s hopeless to think that merely removing an offender from the community will do anything to turn the offender’s life around — imprisonment might be a venue for particular rehabilitative interventions whose success a judge may permissibly seek to facilitate in determining sentence length.  Under this reading, Tapia’s sentence would be okay since the judge was not thinking of imprisonment per se as a means of promoting rehabilitation, but rather as a setting for a specific drug treatment program that was believed to promote rehabilitation.

I’m not sure what the right answer is from the standpoint of statutory interpretation, but I am skeptical of the wisdom of a sentence like Tapia’s.  My concerns are both pragmatic and principled.

On the pragmatic side, I am doubtful of the ability of judges to make accurate decisions about which defendants are able to benefit from what sorts of prison-based programming.  Although we’ve progressed beyond the “nothing works” mentality of the 1970’s, rehabilitation remains a highly uncertain science.  Many programs claim impressive results in reducing recidivism rates, but few have been rigorously tested.  Those that have been carefully assessed tend to show modest benefits, if any at all.  Moreover, even with respect to good programs, judges have little control over the assignment of particular defendants to those programs.  Given all of the uncertainties, I have a hard time seeing the justice (or cost-effectiveness, for that matter) in depriving a person of his liberty for an extra few months in the hope that the person will be rehabilitated while in prison.

On the principled side, a sentence like Tapia’s seems to muddy the message of moral condemnation that criminal punishment ought to send.  Is Tapia being punished or treated?  Does the sentence primarily communicate a message that Tapia deserves blame for her conduct, or that she is in need of social services to address a mental illness?  Does the 51 months reflect what she deserves?  If so, why did the judge seek to justify it by reference to treatment needs?  And if the 51 months really does exceed what is deserved based on the severity of the offense, then the sentence is (at least arguably) morally untenable.

Cross posted at Life Sentences.

Here’s what happened.  Jennifer Krieger was prescribed fentanyl, a powerful opioid, to help her with severe back pain.  She gave some of the drug to her friend Jennifer Curry for recreational use.  Curry misused the fentanyl, as well as a variety of other substances, and died the next day.  Krieger was then indicted for distributing fentanyl with death resulting.  That’s when things got really weird. 

It turns out that the government’s main witness, the medical examiner who concluded that Curry died of fentanyl toxicity, had some serious legal problems of his own and fled the country. 

Eventually, the government returned a superseding indictment removing the “death resulting” language and charging only distribution.  Krieger pled guilty.  The pre-sentencing report recommended a sentencing range of ten to sixteen months.

At sentencing, however, the government once again pushed the “death resulting” provision (21 U.S.C. § 841(b)(1)(C)) that it had removed from the indictment.  The provision requires a minimum sentence of twenty years.  The government managed to produce the errant medical examiner for the sentencing hearing, and the judge found by a preponderance of the evidence that death had resulted from Krieger’s distribution of fentanyl.  With the 841(b)(1)(C) mandatory minimum, the judge felt compelled to sentence Krieger to twenty years, even though he believed the sentence was “too harsh.”  Indeed, the judge went so far as to indicate that he would not have found that death had resulted from fentanyl beyond a reasonable doubt; it was only because death was a “sentencing factor” (hence found using the preponderance standard) and not an element (which would be subject to the higher BRD standard) that Krieger received the concededly too harsh sentence.  Rarely is the significance of burden of persuasion made so clear.

The sentencing judge also nicely captured the “this just can’t be right” flavor of the case: “Krieger, while convicted of distribution of divers amounts of narcotics, is being sentenced for homicide.”  (7)

The Seventh Circuit nonetheless affirmed.  Although Apprendi indicates that a defendant has a right to jury fact-finding using the BRD standard when facts increase the range of punishment, Harris later specified that Apprendi only applies when a fact increases the maximum sentence, not the minimum.  As a result, the BRD standard governs when a fact increases the maximum by only a small amount (as might happen in a system of mandatory sentencing guidelines), but not necessarily when a fact increases the minimum by an enormous amount — as in Krieger, where causing death increased the minimum from zero to twenty years.  In practice, Harris means that Apprendi does not apply precisely where it could be of greatest benefit to defendants.

Interestingly, the Krieger court went out of its way to criticize the current state of the jurisprudence, echoing Justice Stevens’ concurring opinion last May in United States v. O’Brien, 130 S. Ct. 2169 (2010).  Although O’Brien was decided on statutory grounds, Stevens took advantage of the opportunity to reiterate his opposition to Harris and to highlight a provocative comment by Justice Breyer at oral argument (“Well, at some point I guess I have to accept Apprendi, because it’s the law and has been for some time.”).  The comment is important because Breyer was the crucial fifth vote in the Harris majority, and his vote was premised on his continued disagreement with Apprendi.  If he now accepts Apprendi, then it would seem likely that he would reverse his position on mandatory minimums.

In any event, by pointing the finger of blame at the Supreme Court for the current state of the constitutional law, the Seventh Circuit distracts attention from the statutory analysis that was equally important to the outcome in Krieger – analysis that may actually be more troubling to me than the constitutional analysis.

Given the absence of constitutional constraint, the key question in the case was whether § 841(b)(1)(C) ought to be interpreted such that it created an element of a new offense or merely a sentencing factor.  Harris itself involved a parallel question of statutory interpretation, and the Seventh Circuit relied on the analysis in Harris, as well as in the Supreme Court’s more recent O’Brien decision, in determining that “death resulting” was merely a sentencing factor.  The court treated Harris and O’Brien as interchangeable in this regard, characterizing O’Brien as simply a “reiterat[ion]” of Harris (13).

From Harris and O’Brien, the court derived a five-factor test.  Although the court found that two factors (fairness and severity of the sentence) cut “sharply” in Krieger’s favor, the court ultimately concluded that it was

hard pressed to ignore the most important considerations: first, the clear command of the language listing “death resulting” in the “penalties” section of the statute, and second, our precedent (at least when considering drug quantity) of defining the considerations in § 841(b) as sentencing factors . . . . (18-19)

I find the analysis here troubling in at least three respects.

First, it is not clear to me why we should have a special five-factor test for deciding whether a statute creates an element or a sentencing factor.  This is a question of statutory interpretation that ought to be decided using all of the normal tools of statutory interpretation, including traditional canons of construction.  Here, in particular, the “rule of lenity” might appropriately come into play to resolve a question that Congress did not speak to in a clear manner in the statute.  (The avoidance canon might also come into play, given that the increase in Krieger’s minimum sentence was so much greater than the increases found constitutional in Harris and McMillan.)  Although O’Brien did enumerate and consider the five factors, it did not hold that these were intended to supplant conventional interpretive approaches or that they should be regarded as an exhaustive list of considerations in all “sentencing factor” determinations.  Indeed, Harris treated the factors in an even less formal, nonenumerated fashion, and expressly considered the relevance of at least one canon (the avoidance canon).  In Krieger, it seems that a loose listing of considerations from prior cases is hardening into a formal test.   

Second, even if we regard the five O’Brien factors as constituting a controlling test, I see no basis in O’Brien for treating the first two factors as “the most important considerations.”  Indeed, if anything, I think O’Brien meant to lay special emphasis on the fourth factor, the severity of the sentencing increase.  O’Brien dealt with an increase in a mininum sentence from five years to thirty years, which seems hard to distinguish from the zero to twenty in Krieger.  Here’s what the Court said about the five-to-thirty increase:

But the severity of the increase in this case counsels in favor of finding that the prohibition is an element, at least absent some clear congressional indication to the contrary.  (2178)

This important language went undiscussed in Krieger, but I would read it as creating a presumption that “death resulting” is an element, at least in cases with such enormous increases in the minimum. 

Finally, I find it merely question-begging that the “death resulting” language appears under the heading “penalties.”  (I should also note that O’Brien downplayed the significance of formal aspects of statutory structure in deciding whether something is a sentencing factor.  In this regard, O’Brien arguably marks an important departure from Harris, which ultimately laid controlling significance on formal structure.)  No one seems to think that Congress actually contemplates these procedural issues when enacting mandatory minimum statutes.  So why should “penalty” be equated with “sentencing factor” — a technical term of art invented by the Supreme Court, whose procedural significance would not be guessed at even by most lawyers, let alone laypeople and politicians?  There is certainly no “plain meaning” warrant for reading into the term “penalties” an intent for judicial fact-finding using the civil evidence standard.  Indeed, if anything, the term “penalties” might conjure for most people the image of the full panoply of procedural protections that have traditionally attended the imposition of punishment.

I think David is correct to trace the jurisprudential difficulties to the tension between two lines of Supreme Court decisions.  

The first line, exemplified by Taylor v. United States, 495 U.S. 575 (1990), requires the “categorical” approach to determining whether a prior conviction qualifies: based on its elements, an offense is or is not an ACCA predicate without regard to the specifics of how a defendant perpetrated the offense.  The second line, exemplified by Chambers v. United States, 129 S. Ct. 687 (2009), indicates that courts should determine whether a prior felony was violent based on whether that type of felony ordinarily involves violence.  This inquiry contemplates that judges will go beyond the bare elements of the offense and draw on statistics, personal experience, or common-sense intuitions.  Adding to the difficulty is Begay, which adopted a vague mens rea requirement that was not very well anchored in the statutory language.

Here are David’s thoughtful recommendations for lower courts struggling with these difficulties, particularly as they relate to the ACCA’s “residual clause,” which extends the definition of “violent felony” to any crime that “otherwise involves conduct that presents a serious potential risk of physical injury to another”:

First, courts should narrowly construe Begay’s mens rea holding and read it as excluding only strict liability crimes from the residual clause while including crimes of negligence and recklessness. Second, courts should readBegay’s “aggressive” requirement as a rhetorical flourish without any meaningful distinction from “violent.” Third, despite Begay’s apparent invitation to do otherwise, courts should strictly follow the categorical approach and apply the residual clause to only those crimes with elements that require the underlying conduct be violent while excluding those crimes with elements that do not require violence or any mens rea.

David’s proposal is framed as advice to lower courts that are stuck with the Supreme Court precedent.  But I wonder how the Supreme Court itself might best clean up the mess it is has created.

I think a good first step would be for the Court to abandon the categorical approach, which seems inevitably to lead to many arbitrary distinctions based on the fine-grained parsing of criminal statutes that were not written with the ACCA in mind and that do not readily lend themselves to categorization under the Begay test.  Courts ought to be able to look through the formal crime of conviction to determine what defendants’ actually did — this is the only sensible way to accomplish what Congress intended through the ACCA, that is, the incapacitation of armed felons who have demonstrated their dangerousness through a prior course of violent crime.

The Court has resisted an actual-conduct test because of Sixth Amendment concerns — a sentencing judge should not be making findings of fact about a defendant’s prior conduct when those findings increase a defendant’s sentencing exposure.  There would indeed be a conflict between an actual-conduct test and the Sixth-Amendment purism of cases like Apprendi v. New Jersey.  On the other hand, the Court’s most recent pronouncement in this area, Oregon v. Ice, indicates that the Court may be shifting to a more flexible, pragmatic approach.  In any event, there is a simple solution to the Sixth Amendment dilemma: give defendants the right to a jury trial on their alleged ACCA predicates under the residual clause.  There would be some additional transaction costs associated with this, but probably not much — the ACCA issues would surely be resolved by plea agreement in the vast majority of cases.

The defendants in Abbott illustrate how the same conduct that triggers 924(c) can also trigger other mandatory minimums.  

Abbott himself was a felon with an extensive criminal history.  As a result, his possession of a firearm in connection with a drug trafficking crime also subjected him to the fifteen-year mandatory minimum of the Armed Career Criminal Act.  Adding this to the five years under 924(c), Abbott was sentenced to twenty years in prison.

Meanwhile, Gould, the other defendant in Abbott, possessed his firearm in connection with a crack offense.  Under the stiff mandatory minimum statute for dealing crack, Gould faced an additional ten-year minimum on top of the 924(c) five.

In the Supreme Court, Abbott and Gould argued that, as a matter of statutory interpretation, their 924(c) minimums need not have been imposed consecutively to their other, longer minimums.  The Court, however, affirmed both sentences.

The defendants relied on the prefatory language to 924(c): “Except to the extent that a greater minimum sentence is otherwise provided by this subsection or by any other provision of law . . . .”  In Abbott’s view, the “except clause” renders 924(c) inoperative when another, longer mandatory minimum applies to the same conduct that would otherwise function as a predicate for the 924(c) minimum.  (Gould advanced a slightly different interpretation of the clause.)

The Court, however, adopted the government’s interpretation, under which the “except clause” only kicks in when the other, longer minimum is for using, carrying, or possessing a firearm in connection with a crime of violence or drug trafficking crime.  In other words, 924(c) applies unless there is a longer minimum that includes exactly the same elements as 924(c).

Although the statutory language is messy, it strikes me that the Court’s interpretation better fits the language than the defendants’ interpretations.  In particular, 924(c)(1)(D)(ii) seems to speak pretty clearly in favor of the Court’s view: “[N]o term of imprisonment imposed on a person under this subsection shall run concurrently with any other term of imprisonment imposed on the person, including any term of imprisonment imposed for the crime of violence or drug trafficking crime during which the firearm was used, carried, or possessed.”

I am nonetheless troubled by an aspect of the Court’s reasoning: that is, the Court’s rejection of the defendants’ interpretations because they would result in “anomalies,” “oddities,” and outcomes that are “bizarre” and not “equitable.”  Although the Court did not use the term, what it really had in mind seems to be disproportionality — treating less culpable offenders more harshly than more culpable offenders.

To be sure, the defendants’ interpretation would provide a windfall of sorts for those 924(c) offenders who were “fortunate” enough to be subject to even longer minimums.  But 924(c) — like all mandatory minimums — has nothing to do with proportionality.

In the federal system, we have a generally serviceable sentencing scheme intended to achieve proportionality — the guidelines.  Section 924(c) is not intended to accomplish proportionality, but to overrideproportionality.  Like other mandatory minimums, 924(c) takes a single sentencing variable — here, possession of a firearm — that may mean dramatically different things in different cases, and gives that single variable a dominant, uniform weight.  The guidelines provide for far more nuanced and sensible distinctions.

It is hard to say exactly what the point of 924(c) and other mandatory minimums is — these election-year enactments are probably more about anti-crime political posturing than anything else — but it seems more than a bit perverse to invoke proportionality ideals when rejecting a defendant’s proposed interpretation of the statute.  When the application of 924(c) systematically produces “anomalies,” “oddities,” and outcomes that are “bizarre” and not “equitable,” a limiting interpretation may bring more, not less, proportionality to our overall system for punishing drug and gun offenders.

At one point, the Court itself came close to acknowledging as much: “We do not gainsay that Abbott and Gould project a rational, less harsh, mode of sentencing.  But we do not think it was the mode Congress ordered.”  This points to a more honest interpretive approach.  When the statutory language does not bear a reading that makes for sensible sentencing policy, why pretend that it does?

Cross posted at Life Sentences Blog.

Complementing what is happening in the collateral-consequences cases, the Tenth Circuit recently ruled that a defendant’s right to effective assistance was violated when his lawyer did not warn him of the dangers of confessing to uncharged criminal conduct during a presentence investigation meeting with a probation officer. 

In United States v. Washington (No. 08-3313), the defendant was convicted of distributing about 62 grams of cocaine base.  Between conviction and sentencing, Washington submitted to a routine interview with a probation officer as part of the presentence investigation.  Washington’s lawyer neither attended the interview in person nor advised Washington that admission of uncharged criminal conduct might aggravate his sentence as “relevant conduct” under the then-mandatory federal sentencing guidelines.  Washington ultimately admitted to distributing an additional 2.5 kilograms of cocaine base, which helped to boost his sentence to the statutory maximum of forty years per count of conviction.

In arguing that his right to effective assistance was violated, Washington had a serious problem: Tenth Circuit precedent established that the presentence investigation is not a “critical stage” in the criminal process that triggers a right to counsel.  The precedent seems of questionable wisdom to me, but, in any event, the majority concluded it was distinguishable.  Although the presentence investigation may not be a critical stage, sentencing itself plainly is.  Effective representation at a critical stage implies that appropriate preparations are made to maximize the chances of a good outcome at the critical stage.  Thus, while an attorney need not be physically present during the presentence investigation interview, the attorney at least needs to provide sound advice in advance of the interview regarding the potential effect of admissions on the sentence.

Washington seems clearly correct to me in its reasoning, but two considerations may limit its applicability to other cases.  First, the court emphasized the lawyer’s extraordinary ignorance regarding the whole guidelines system (“counsel’s failure to understand the basic mechanics of the sentencing guidelines,” as the court put it).  So what would happen if counsel understood and related to the defendant the concept of relevant conduct in a general way, but did not provide specific counsel regarding admissions or provided misinformation regarding the weight given to relevant conduct?  Second, Washington was sentenced under the old mandatory guidelines system.  Now that the guidelines are advisory, it is arguably less important for counsel to provide accurate advice about relevant conduct.  Given the continuing requirement that sentencing judges calculate and “consider” a guidelines sentence, as well as the data indicating that the guidelines sentence is still routinely imposed in many districts, I would think effective representation still requires good advice about how the presentence investigation affects the guidelines calculus.  It is not clear, however, whether the Washington court would agree.

Cross posted at Life Sentence.

At least sixteen districts, including the Mexican border districts, have developed fast-track programs that offer extraordinary sentencing benefits for illegal reentrants who plead guilty in an especially expedited fashion.  (For background, see my article at 27 Hamline L. Rev. 357.)  However, many other districts, including all of the Seventh Circuit districts, do not offer defendants the fast-track option, which creates wide sentencing disparities in illegal reentry cases.  When the federal sentencing guidelines were converted from mandatory to advisory in 2005, many defendants in non-fast-track districts argued that judges ought to give them the fast-track benefit in order to mitigate the disparities.  Appellate courts, however, uniformly rejected these arguments prior to 2007, when the Supreme Court reemphasized the discretionary nature of federal sentencing in Kimbrough v. United States, 552 U.S. 85.  Post-Kimbrough, three circuits, now joined by the Seventh, have ruled that sentencing judges may consider the fast-track disparities.   

For the reasons set forth in Reyes-Hernandez, I think this position is clearly the correct position if Kimbrough is taken seriously.  The Kimbrough Court held that sentencing judges are permitted to take into account the disparities created by the guidelines’ harsher treatment of crack than powder cocaine, notwithstanding various intimations from Congress that it approved of the disparities.  (For more recent developments in the area of crack sentencing, see my posts here and here.)  Likewise, the subtle ways in which Congress has arguably approved of fast-track disparities should not be enough to overcome the broad discretion sentencing judges have in the post-Booker advisory system.  If attempting to mitigate crack-powder disparities is permissible as a matter of law, then so should attempting to mitigate fast-track disparities.

As the Seventh Circuit made clear, Reyes-Hernandez only permits, and does not require, judges to take fast-track disparities into account.  Moreover, the court signaled some doubt about the appropriateness of giving a below-guidelines sentence based solely on fast-track disparities:

[W]e provide a word of caution that a departure from the guidelines premised solely on a fast-track disparity may still be unreasonable.  To withstand scrutiny, a departure should result from a holistic and meaningful review of all relevant § 3553(a) factors.  (33)

Although now permitted to do, it is a separate (and more difficult) question whether judges should take the fast-track disparity into account.  The problem with lowering one reentrant’s sentence in a non-fast-track district is that it exacerbates disparities relative to other defendants in non-fast-track districts at the same time that it mitigates disparities relative to fast-track districts.  One solution that I proposed a few years ago (see the article here) would be to select a sentence based on the national average sentence imposed in all illegal reentry cases, which would presumably be somewhere between the fast-track and non-fast-track norms.

The Reyes-Hernandez court suggested another way of thinking about the problem, in light of the parsimony principle embodied in § 3553(a).  As the court put it, “It has been observed even by strong defenders of the guidelines that the sentencing ranges called for under the guidelines for unlawful reentry cases are often unreasonably harsh and disproportionate to the seriousness of the offense” (32).  The steep discount in fast-track programs helps to confirm that the full guidelines sentence is not really necessary to achieve proportionality objectives in illegal reentry cases.  This gives sentencing judges a good reason to err on the side of lenience in deciding which set of disparities are most worthwhile to mitigate.

Cross posted at Life Sentences.

Although “cocaine base” cases normally involve what is undisputably crack, defendants have from time to time litigated whether a particular susbtance really triggers the ten-year minimum.  These cases have produced a longstanding circuit split, with six circuits (the First, Second, Third, Fourth, Fifth, and Tenth) reading “cocaine base” to encompass all forms cocaine that are chemically classified as a base, and five circuits holding that “cocaine base” means more narrowly what Congress was really concerned about, i.e., crack and other types of smokable cocaine base.  With today’s cert. grant in DePierre v. United States, 599 U.S. 25 (1st Cir. 2010), the Supreme Court appears poised finally to resolve the issue. 

The facts of DePierre nicely illustrate the pernicious effects of the artificial distinction created by the 1986 law.  A government informant purchased powder cocaine from DePierre in February 2005.  Federal agents could have arrested DePierre at that time, but decided instead to invite a crack deal with the deliberate purpose of increasing the sentence.  Cocaine is easily cooked into crack, so the informant’s request to receive his next delivery of cocaine in crack form presented no large difficulty for DePierre, who eventually sold the informant 55.1 grams of what the government alleged was crack — just a little more than necessary to trigger the ten-year minimum.  Whether or not these facts satisfy the demanding legal requirements for the entrapment defense — DePierre’s jury decided not, and the question is not now before the Supreme Court – the government’s conduct was at the least unseemly.  Yet, given the enormous consequences the law attaches to an arbitrary distinction, it is easy to see the temptation for the government to induce crack deals.

Although we ought to welcome any narrowing of the ”cocaine base” law, even if the Court does rule on the side of the defendant in DePierre, we will still have a basically arbitrary distinction in place.  (For an illustration of the difficulties created by using “crack” as the key term, see my earlier post here.)  More promising is the recently enacted Fair Sentencing Act of 2010, which raised the triggering quantity for the ten-year minimum to 280 grams.  It will be interesting to see if and how the new law affects law-enforcement practices.  For instance, would the agents pursuing DePierre have engineered several additional crack deals in order to reach the 280-gram threshhold?

Cross posted at Life Sentences.

In United States v. Glosser (No. 08-4015), the judge made a promise to the defendant at his change-of-plea hearing that he would impose the statutory minimum 120-month sentence in the case, notwithstanding the prosecutor’s suggestion that the government might seek more.  And, indeed, it turned out that the government sought a 210-month sentence in light of firearms found at Glosser’s resident.  The judge, however, mostly stuck to his promise and imposed a 121-month sentence.  The government appealed. 

Although predetermining the sentence before hearing a party’s arguments is not exactly the same thing as passing over a party’s arguments in silence, they seem functionally much the same.  And, in Glosser, the Seventh Circuit indeed indicated that the former is a reversible procedural error.  The government’s success in Glosser thus suggests the government has much the same opportunity to challenge sentence explanations as defendants.

Another aspect of Glosser also strikes me as noteworthy, that is, the court’s consideration of whether the error was harmless.  The sentence explanation cases do not often include express consideration of harmlessness, which is not surprising because a failure to address a party’s sentencing arguments does not lend itself to a harmlessness analysis — post-Booker federal sentencing is so discretionary that it is difficult to see how harmlessness could be either proven or disproven.

The Glosser court seemed ambivalent about whether harmless error analysis was appropriate.  On the one hand, the court did conclude (in a rather conclusory way) that the error was not harmless (“Although the district court have several reasons at the sentencing hearing for imposing a below-guidelines sentence of 121 months, we are not confident that it would have done so if it had not decided it would impose the mandatory minimum sentence when Glosser changed his plea to guilty.” (14)).  On the other hand, the court also suggested that the error might be more properly analyzed as “fundamental procedural error” that would not require harmless error analysis (15).  Because the outcome would be the same either way, the court did not need to definitively resolve whether the error was fundamental or not — an interesting open question.

Cross posted at Life Sentences.

• A household member is present in about one-quarter of residential burlgaries.
• A household member is violently vicitmized in about seven percent of residential burglaries (or about one-quarter of the burglaries in which a household member is present).
• In residential burglaries, simple asault is the most common violent crime (3.7 percent of all burlgaries), while more serious violent crimes like rape (0.6 percent) and aggravated assault (1.3 percent) are far less frequent.
• In a majority of even the “violent” burglaries, the victim indicates there is no injury; a “serious injury” is sustained in only 8.5 percent of the violent burglaries.
• In the violent burglaries, fewer than forty percent of the offenders are armed.

Now, for the ACCA link. 

As I describe in this post, Justice Scalia’s preferred method for deciding whether a prior conviction counts as a “violent felony” (three of which trigger the ACCA fifteen-year mandatory minimum prison term) involves statistical comparison between the dangerousness of the prior offense and the dangerousness of burglary.  Moreover, as I describe in this post, there is reason to think that Scalia’s statistical approach is gaining traction on the Court.  (With cert. recently granted in a new ACCA case, we may soon see more evidence of the shift.)  If the statistical approach becomes more important in applying ACCA, then the BJS’s new burlgary data may prove a critically important touchstone in the analysis.

Here are a few additional interesting bits of data from the BJS report:

• The U.S. averages about 3.7 million household burglaries per year.
• Renters are almost twice as likely to experience burglary as home-owners.
• Poor people are more likely to experience burglary than the well-off, and young people more than old.
• In nearly half of burglaries, the property loss is less than $250.
• Burglaries frequently go unreported to the police (more than forty-seven percent of cases when no household member was present and more than forty-one percent of cases when a household member was present).
• Belying the common association between burglary and shadowy strangers, offenders are known to their victims in nearly two-thirds of violent burglaries.

Cross posted at Life Sentences.

This issue also includes Rachel Delaney’s comment analyzing the use of deferred prosecution agreements (DPAs) in the corporate crime context, ultimately calling for congressional regulation of prosecutorial discretion.  

Tracing the brief history of DPAs from the 1992 Salomon Brothers case to the 2008 AGA Medical Corporation agreement, Delaney demonstrates how the DPA process has been largely governed by, initially, ad hoc decisions of prosecutors and, currently, internal guidelines that grant too much prosecutorial discretion on a variety of levels, are unenforceable, and subject to revision at any time.  She suggests that there is broad legal support for reform of the process and advocates for the passage of recently proposed legislation by Congress to curb prosecutorial abuses and effectively punish and deter white collar crime.

Both of these pieces can be accessed at the Law Review’s website.

Filip, who also is a former federal judge and now practices with a Chicago law firm, said that United States v. Booker in 2005 reduced the import of sentencing guidelines that dated to the late 1980s, “returning us to an era of indeterminate sentencing.” While he said that commentary on Booker from both judges and defense lawyers has been generally favorable, data on sentencing patterns since the decision show that in different parts of the country, significantly different sentences are being given for comparable convictions. 

Overall, Filip said, in fiscal year 2008, only 59 percent of sentences fell into the non-binding range given in the guidelines, with the figure ranging from about 40 to 70 percent in different federal circuits.

Filip criticized “the corrosive effect of this disparity,” adding, “It hardly promotes respect for the judicial system” when, as has been reported, defendants and inmates talk in jails and prisons about which judge  gives more lenient sentences.

Filip urged that federal judges spend time collaborating on how to set sentences and discussing what is appropriate, and that the federal Sentencing Commission take strong positions when it observes broad patterns of sentencing below the recommended ranges. “The systematic benefits of having relatively uniform sentences” should be kept in mind by judges, he said.  

Filip also expressed concern about two other trends, the use of business concepts in measuring whether the legal system is doing a good job and growing reliance on corporate compliance monitors as part of the settlement or sentence in disputes involving businesses.

“Law enforcement is not like a business,” he said. Applying concepts such as return on investment or cost-benefit analysis is often not appropriate. Measuring the performance of a judge, prosecutor, or law enforcement agency by data such as how many cases were brought, the length of sentences, or arrest or conviction totals may distort wiser and more appropriate judgment of whether people are doing their jobs in the best manner.

Corporate compliance monitors are not always inappropriate, Filip said, but the use of them has been expensive, burdensome, and not really productive in many instances. “At minimum, we have come to a point where caution is appropriate,” he said.

Filip was Deputy Attorney General under Attorney General Michael Mukasey in the last two years of the administration of President George W. Bush.

No, said the Seventh Circuit in United States v. Hines (No. 08-3255).  Writing for the court, Judge Posner offered a little chemistry lesson, explaining that the process of cooking powder into crack removes hydrochloride from the drug.  Under ideal conditions, cooking results in an eleven-percent weight loss.  But, given the potential for careless waste during cooking, it is not clear that even the eleven-percent loss should be assumed.  The court concluded, “[If] the government wants the sentencing judge to infer the weight of the crack from the weight of the powder from which the crack was manufactured, it has to present evidence, concerning the cooking process, that would enable a conversion ratio to be estimated” (3).  (Judging by the mess on my kitchen counter most mornings, I am guessing that the “conversion ratio” when my six-year-old cooks oatmeal into hot cereal is about 2:1.  Fortunately, and notwithstanding its cholesterol-lowering benefits, the street value of oatmeal remains a lot less than cocaine.)