THOMPSON
Vacated and remanded.
Judith H. Mizner, Assistant Public Defender, for appellant. Dina Michael Chaitowitz, Assistant United States Attorney, with whom Carmen M. Ortiz, United States Attorney, was on brief, for appellee.
Before TORRUELLA, HOWARD, and THOMPSON, Circuit Judges. THOMPSON, Circuit Judge.
Overview
Jeremy Barnes is here again, this time because the Supreme Court granted his certiorari petition, vacated our judgment, and remanded his case for reconsideration in light of Alleyne v. United States, ––– U.S. ––––, 133 S.Ct. 2151, 186 L.Ed.2d 314 (2013). See Barnes v. United States, –––U.S. ––––, 133 S.Ct. 2851, 186 L.Ed.2d 903 (2013). We asked for and received supplemental briefing from the parties regarding Alleyne and heard oral argument too. So the matter is teed up for decision. And after summarizing the case's background and discussing the parties' arguments, we vacate Barnes's sentence and remand for resentencing.
Barnes appealed, arguing (among other things) that drug quantity must be treated as an element of the offense and thus proved beyond a reasonable doubt. We affirmed because then-controlling caselaw let a judge use a preponderance standard in finding facts that increase a defendant's mandatory-minimum sentence. See Goodine, 326 F.3d at 32. Barnes sought certiorari. A little later the Supreme Court held in Alleyne that most (but not all) facts that increase statutory minimum penalties must (if the defendant does not admit them) be proved beyond a reasonable doubt. See 133 S.Ct. at 2161–63. Not surprisingly, the Court “GVR'd”—granted certiorari, vacated the judgment, and remanded the case—for further consideration in view of Alleyne.
Alleyne did not change the preexisting rule that a judge may find the fact of a prior conviction. See 133 S.Ct. at 2160 n. 1; see also United States v. Rodriguez, 759 F.3d 113, 122 (1st Cir.2014).
Analysis
So here we are. Helpfully, the parties do not dispute that the judge's drug-quantity calculation reflects Alleyne error. Rightly so, because the judge reached the 1,000–plus kilogram figure—the amount needed to trigger mandatory minimums—using a preponderance standard. See United States v. Harakaly, 734 F.3d 88, 93–94 (1st Cir.2013) (discussing Alleyne error). They also do not dispute that an Alleyne error is of constitutional magnitude. About this, they are right again. See United States v. Pena, 742 F.3d 508, 514 (1st Cir.2014). And they do not dispute that Barnes preserved his claim of Alleyne error regarding the 210–month jail term—which they agree means that we must vacate his sentence unless the government proves the error's harmlessness beyond a reasonable doubt. Right they are. See United States v. PérezRuiz, 353 F.3d 1, 17 (1st Cir.2003) (explaining that the government must prove “beyond any reasonable doubt that the assigned error did not contribute to the result of which the appellant complains”); see also United States v. Melvin, 730 F.3d 29, 32 (1st Cir.2013) (calling the government's burden here a “heavy” one).
Recently we held that no Alleyne error arises if “a defendant's sentence is based entirely on [g]uidelines considerations without changing the applicable mandatory minimum.” United States v. Ramírez–Negrón, 751 F.3d 42, 49 (1st Cir.2014). Neither the judge nor the parties there ever mentioned at sentencing that a mandatory minimum was in play. Id. at 50. The government here does not make any Ramírez–Negrón–type claim, telling us at oral argument that, unlike that case, Barnes's case does involve Alleyne error.
Two issues divide the parties, however. The first is whether the Alleyne error involving the 210–month prison term is harmless—the government says it is; Barnes says it is not. The second is whether Barnes preserved the Alleyne error regarding the 5–year supervised-release term—the government says he did not, and so argues plain-error review applies; Barnes says he did, and so argues harmless-error review controls. We think Barnes has the better of the arguments.
Barnes also argues that the Alleyne error is structural, requiring reversal even if everyone thinks it is harmless. He says that his argument appears foreclosed by Harakaly, 734 F.3d at 94–95, and raises it—without developing it in any meaningful way—simply to preserve it for possible en banc or certiorari review. So we need say no more about that issue.
The Prison Term
For those keeping track, the prison term imposed—210 months—is 90 months above the 10–year mandatory minimum. The government pounces on this fact, insisting it shows the judge based the jail sentence on guideline considerations, like the applicable guideline-sentencing range and the relevant section 3553(a) factors. The applicable mandatory minimum, in other words, “had no practical effect” on the prison sentence that the judge chose, at least in the government's mind. Ergo, the government contends, the Alleyne error was harmless beyond a reasonable doubt. We view matters quite differently, however. Here's why:
The key players at sentencing, recall, talked about mandatory minimums. True, the probation officer did not use the words “mandatory minimum,” but she did advise the judge that for 1,000 kilograms or more of marijuana, “the supervised release range would be 5 years”—which is the mandatory minimum under section 841(b)(1)(A). And if, as probation suggested, that amount activated a mandatory minimum supervised-release term, then everyone at sentencing could assume probation thought that amount activated a mandatory minimum jail term too. The prosecutor did use the words “mandatory minimum” hard on the heels of probation's comment, telling the judge that the government's recommended prison sentence of 210 months would “capture[ ] the applicable 10–year mandatory minimum.” The judge later used “mandatory minimum” lingo too, asking Barnes's lawyer whether “a 10–year mandatory minimum” applied given the drug-quantity findings, to which counsel responded “yes.” All of this throws cold water on the government's attempt to downplay the mandatory minimum's role at sentencing.
But wait, protests the government. The judge also said there that he had actually considered giving Barnes a “higher sentence.” Surely that comment shows the mandatory minimum did not affect the sentence selected, the government argues. Not quite. Remember, the judge followed his “higher sentence” remark by saying, in almost the same breath, that he decided not to jack up the prison stay because he “usually” does not “exceed the government's recommendation” in plea-agreement cases—a recommendation, we repeat, that the prosecutor made while explaining how a 210–month sentence “captures the applicable 10–year mandatory minimum.” Remember too that in his formal statement of reasons (which judges must complete when sentencing defendants), the judge indicated that he had handed out a “[m]andatory minimum sentence.” So the government's protest fails.
An Alleyne error can be harmless if “no reasonable jury”—hearing the same evidence that the sentencing judge heard—could hold the defendant responsible for a drug amount “below the ... threshold [ ] triggering the mandatory minimum[ ].” Ramírez–Negrón, 751 F.3d at 51 n. 8 (citing Harakaly, 734 F.3d at 95–96). But the government makes no argument like that here. And so we move on. See, e.g., United States v. Politano, 522 F.3d 69, 75 n. 5 (1st Cir.2008) (holding that an argument never made on appeal is waived).
The bottom line here is this. The government offers a variety of theories why, in its view, the Alleyne error had no effect on the 210–month prison term. But none proves the error's harmlessness under our stiff test— i.e., none proves beyond a reasonable doubt that the error did not “contribute” to the complained-about sentence. See Pérez–Ruiz, 353 F.3d at 17. Consequently we must vacate the prison component of Barnes's sentence.
The Supervised–Release Term
And we must do the same with the supervised-release part of Barnes's sentence too. That 5–year term, we remind the reader, corresponds exactly to the statutory mandatory minimum. Faced with this circumstance, the government pins its principal hope on persuading us that the defense acquiesced to the 5 years. The record, however, undercuts that theory.
Yes, as the government notes, defense counsel did not object when the probation officer told the judge a 5–year supervised-release term applied, assuming the judge found a drug weight of at least 1,000 kilograms. And yes, as the government also notes, defense counsel later told the judge that he had to “impose at least five years of supervised release.” But those things happened after the defense lost the drug-quantity battle, with the judge's findings activating mandatory minimums. Well, the government says, when the judge handed down the sentence, the defense never asked him to explain whether (to quote its latest brief) “the term was statutorily required or simply desirable.” Essentially, the government wants us to ignore everything that preceded the sentence's imposition— e.g., the defense's fighting and losing on drug quantity and the probation officer's then telling the judge that 1,000 kilograms or more of marijuana (the threshold for triggering a 5–year supervised-release term) made Barnes eligible for at least 5 years of supervised release. And that we will not do. Ultimately, then, this concatenation of events kiboshes the government's forfeiture argument.
And with the forfeiture issue out of the way, the rest of the analysis is easy. On the harmlessness question, considering everything in context—with the talk of mandatory minimums in the air at sentencing, with a check next to “[m]andatory minimum sentence” on the judge's statement-of-reasons form, etc.—we cannot say beyond a reasonable doubt that the Alleyne error did not contribute to the complained-of supervised-release result. So we must vacate that part of the sentence as well.
Final Words
Our work over, we vacate Barnes's sentence in its entirety and remand the case for new sentencing proceedings consistent with this opinion. -------- Notes: