Reset A A Font size: Print

United States Court of Appeals,First Circuit.

UNITED STATES of America, Appellee, v. Jorge REYES–SANTIAGO, Defendant, Appellant.

Nos. 12–2372, 12–2381.

Decided: September 23, 2015

Before Torruella, Lipez, and Barron, Circuit Judqes. Vivianne M. Marrero–Torres, Assistant Federal Public Defender, with whom Hector E. Guzman–Silva, Federal Public Defender, and Hector L. Ramos–Vega, Assistant Federal Public Defender, Supervisor, Appeals Section, were on brief, for appellant. Luke Cass, Assistant United States Attorney, with whom Rosa Emilia Rodriquez–Velez, United States Attorney, and Nelson Perez–Sosa, Assistant United States Attorney, Chief, Appellate Division, were on brief, for appellee.

Appellant Jorge Reyes–Santiago (“Reyes”) was among 110 defendants charged in a two-count indictment with drug and firearms offenses arising from a massive drug ring operating in public housing projects in Bayamón, Puerto Rico. Most of the high-level members of the conspiracy, Reyes among them, pled guilty pursuant to plea agreements. Other than for Reyes, the sentences imposed on Count One, the drug count, ranged from 78 months to 324 months,1 the latter imposed on the chieftain of the enterprise. Reyes received the stiffest Count One sentence: 360 months. In this appeal, he seeks resentencing on Count One on three grounds: the government's alleged breach of his plea agreement, the sentencing court's alleged inappropriate conduct in demanding witness testimony, and the disparity between his sentence and those of similarly situated co-defendants. Reyes also claims the district court erred in ordering a 24–month consecutive sentence for his violation of supervised release conditions imposed in an earlier case.

We find merit in the disparity argument. Ultimately,2 in sentencing the lead conspirators, the district court refused to accept stipulated drug amounts only for Reyes, listed as Defendant # 9 in the indictment, and for the conspiracy's kingpin, Defendant # 1. Although sentencing courts have the discretion to reject recommendations made in plea agreements, and need not uniformly accept or reject such stipulations for co-defendants, they nonetheless must impose sentences along a spectrum that makes sense, given the co-defendants' criminal conduct and other individual circumstances. In this case, after reviewing Presentence Investigation Reports (“PSRs”) and sentencing transcripts for the leaders in the conspiracy, we conclude that the rationale offered by the district court for the substantial disparity between Reyes's sentence and the sentences of others above him in the conspiracy's hierarchy is unsupported by the record. We therefore must remand this case to the district court for reconsideration of Reyes's sentence.

Given that resentencing must occur, we need not decide whether the government breached Reyes's plea agreement in the prior sentencing proceedings. However, we discuss certain aspects of the government's performance to provide guidance for the proceedings on remand. Finally, as explained below, resentencing also is necessary for Reyes's violation of his. conditions of supervised release.


Before delving into the substance of this case, we address the government's motion to dismiss the appeal on the ground that it was not timely filed.3 Under Federal Rule of Appellate Procedure 4(b), with exceptions inapplicable here, a notice of appeal in a criminal case must be filed within fourteen days of the entry of judgment. In United States v. Gonzalez–Rodriguez, 777 F.3d 37 (1st Cir.2015), we explained that the filing of a motion seeking reconsideration of a criminal sentence does not extend Rule 4(b)'s filing period. Id. at 41. Reyes properly recognizes that his appeal is untimely under Gonzalez–Rodriguez, as he filed his notice of appeal more than five months after judgment entered.4 He argues, however, that the government both waived and forfeited its untimeliness challenge. As explained below, we agree that the time limit set by Rule 4(b) may be waived and that the government did so in this case.

A. The Consequence of Untimely Filing

Although we previously have described the time limits in Rule 4(b) as “mandatory and jurisdictional,” United States v. Rapoport, 159 F.3d 1, 3 (1st Cir.1998) (internal quotation marks omitted); see also Gonzalez–Rodriguez, 777 F.3d at 40 n.4, more recent Supreme Court cases have pointed out the difference between “ ‘a rule governing subject-matter jurisdiction and an inflexible claim-processing rule,” Eberhart v. United States, 546 U.S. 12, 13 (2005) (per curiam) (quoting Kontrick v. Ryan, 540 U.S. 443, 456 (2004)). In the latter instance, a failure to adhere to prescribed time limits does not foreclose jurisdiction, but may bar the tardy litigant from securing the relief sought if the opposing party properly objects. See Kontrick, 540 U.S. at 456 (noting that “a claim-processing rule, ․ even if unalterable on a party's application, can nonetheless be forfeited if the party asserting the rule waits too long to raise the point”); see also Eberhart, 546 U.S. at 19.5

We have not had occasion to revisit our earlier precedent describing Rule 4(b) as jurisdictional, but every circuit to decide the issue since Kontrick and Eberhart has concluded that Rule 4(b) is a claims-processing rule and, hence, may be waived or forfeited. See United States v. Gaytan–Garza, 652 F.3d 680, 681 (6th Cir.2011) (per curiam) (citing cases). As the other circuits have interpreted the Supreme Court's cases, the distinction rests on whether the time-bar at issue originated in a statute. Id. We see merit in that view, which draws support from the Court's comparison in Bowles v. Russell, 551 U.S. 205 (2007), between the “longstanding treatment of statutory time limits for taking an appeal as jurisdictional,” and the conclusion in Kontrick that the limitations period at issue was a claims-processing rule. Bowles, 551 U.S. at 210–11. In Bowles, the Court observed that critical to the Kontrick holding was “the fact that ‘[n]o statute ․ specifies a time limit for filing a complaint objecting to the debtor's discharge.’ “ (quoting Kontrick, 540 U.S. at 448) (alteration and omission in original)). Id. at 211; see also id. at 211–12 (referring to “the jurisdictional distinction between court-promulgated rules and limits enacted by Congress”).

Rule 4(b) does not arise from a statutorily imposed time constraint,6 and we see no rationale for crossing the line the Supreme Court seemingly has drawn between statute-based time limits and those without legislative origin.7 Hence, we hold that Rule 4(b)'s time limits are not “mandatory and jurisdictional” in the absence of a timely objection from the government. We thus must consider whether the government waived or forfeited its challenge to appellant's tardiness.

The government's motion to dismiss reflected a belief that it may object to a late filing under Rule 4(b) at any time, and, once it does, the court is obliged to enforce the rule. It therefore asserted that its motion—filed three months after this case was submitted for decision following oral argument—required us to dismiss Reyes's appeal. In so arguing, the government relied on the statement in Gonzalez–Rodriguez that the time limits in Rule 4(b), “even if not jurisdictional, are mandatory when raised by the government.” 777 F.3d at 40 n.4. In Gonzalez–Rodriguez, however, the government had contended in its original response brief that the court had no jurisdiction to consider the defendant's sentencing appeal. Here, by contrast, the government suggested in a footnote in its original brief that Reyes's appeal was timely because the fourteen-day clock for filing a notice of appeal did not begin to run until after the district court denied Reyes's motion for reconsideration. In pertinent part, the footnote states:

“[T]he Supreme Court has held that the timely filing of such a motion [for reconsideration] in a criminal action tolls the time for filing a notice of appeal and the time begins to run anew following disposition of the motion.” United States v. Vicaria, 963 F.2d 1412, 1413–14 (11th Cir.1992 [) ] (citing United States v. Dieter, 4 29 U.S. 6, 8–9 (1976[) ]; see also United States v. Ortiz, No. 12–2190, 2014 WL 278510, at *5, n.2 (1st Cir. Jan. 27, 2014) (“motions for reconsideration in criminal cases are not specifically authorized either by statute or by rule”); United States v. Healy, 376 U.S. 75, 84 (1964[) ]. “A motion for reconsideration in a criminal case must be filed within the period of time allotted for filing a notice of appeal in order to extend the time for filing the notice of appeal.” See United States v. Russo, 760 F.2d 1229, 1230 (11th Cir.1985).

Gov't Br. at 3 n.2.

The government attempted to characterize the footnote as something other than a concession of timely filing and appellate jurisdiction. Although the passage may not be an explicit concession, it is nearly so. The government sought to justify its acquiescence with the fact that Gonzalez–Rodriguez was decided after its brief was filed, insinuating that the prosecutors did not have reason before then to raise a timeliness objection. But the brief in Gonzalez–Rodriguez raising the jurisdictional argument was filed by the same United States Attorney's Office responsible for this case more than two months before it submitted the response brief that contains the footnote. The government did not explain why it could not have made the same time-bar argument here. Moreover, the panel's opinion in Gonzalez–Rodriguez simply pulled together pre-existing authorities. See, e.g., United States v. Dotz, 455 F.3d 644, 648 (6th Cir.2006) (“In the sentencing context, there is simply no such thing as a ‘motion to reconsider’ an otherwise final sentence ․“ (quoted in United States v. Ortiz, 741 F.3d 288, 292 n.2 (1st Cir.2014)). Plainly, as its argument in Gonzalez–Rodriguez demonstrates, the government had ample basis to challenge the jurisdiction for this appeal in its original response brief. Thus, we have here a situation that is fairly characterized as a waiver.

Moreover, even absent the jurisdictional footnote, the government's request for dismissal would have confronted the maxim that any issue not raised in a party's opening brief is forfeited. See, e.g., United States v. Tum, 707 F.3d 68, 72 n.2 (1st Cir.2013). The government cited no case in which a court has found adequate the government's request for dismissal of an appeal, based on Rule 4(b), when the objection was raised for the first time after the government had filed its response brief. The forfeiture discussion in other cases has focused on whether the government needed to raise the issue before filing its brief. See, e.g., Sadler, 480 F.3d at 941 (agreeing with other circuits that “raising the untimeliness argument in briefing, as opposed to in a motion to dismiss, was sufficient to invoke Rule 4's protections”); see also United States v. Garduño, 506 F.3d 1287, 1292 n.7 (10th Cir.2007) (advising the government to consider filing a motion for dismissal under Fed. R.App. P. 27 and the applicable Tenth Circuit rule “when the government recognizes a violation of Rule 4(b)(1)(A),” but stating that failure to do so “does not constitute a forfeiture where, as here, the appellee seeks dismissal for failure to timely appeal in its response brief”); cf. Huerta v. Gonzales, 443 F.3d 753, 757 (10th Cir.2006) (finding that the government had forfeited objection to a claim-processing rule because it responded on the merits to petitioner's late-filed appeal).8

Applying ordinary forfeiture principles in this context is consistent with the Supreme Court's assertion in Eberhart that “[t]hese claim-processing rules ․ assure relief to a party properly raising them.” 546 U.S. at 19 (emphasis added). Further, in Kontrick, the Court endorsed the notion that “[t]ime bars ․ generally must be raised in an answer or responsive pleading.” 540 U.S. at 458. Although the Court's reference was to trial court proceedings, the obligation to “timely assert[ ] the untimeliness” of a litigation adversary's conduct is equally applicable on appeal. Id. Hence, even if the government had not waived the timeliness issue, we would hold that the claim was forfeited because it was not raised in its opening brief.


In assessing Reyes's challenge to his sentence in the conspiracy case, we draw the facts from his and his co-defendants' change-of-plea and sentencing hearings, PSRs, and plea agreements. See United States v. Rivera–González, 776 F.3d 45, 47 (1st Cir.2015).9

A. Sentencing Overview

Count One of the indictment against Reyes and his numerous co-defendants alleged possession with the intent to distribute substantial quantities of various drugs—including heroin, crack cocaine, and cocaine—in a conspiracy that operated at four public housing projects in Bayamón at least from 2005 to 2010. See 21 U.S.C. §§ 841(a), 846, 860. Count Two charged the defendants with aiding and abetting their co-conspirators in using and carrying firearms in connection with the drug trafficking enterprise identified in Count One, in violation of 18 U.S.C. §§ 924(c)(1)(A) and 2.

The indictment divided the defendants into multiple groups: the Leaders (defendants1–5), Managers/Drug Owners (defendants6–14), Supplier (defendant # 15), Enforcers (defendants16–17), Runners (defendants18–34), Sellers (defendants35–84), Lookouts (defendants85–106), and Facilitators (defendants 107–110). Reyes, # 9, was in the second tier and designated a Manager/Drug Owner.10 The PSRs for the top defendants stated that, during the enterprise's five-year span, it was responsible for distributing approximately the following amounts of drugs: 210 kilograms of heroin, 485 kilograms of crack cocaine, 505 kilograms of cocaine, 2,000 kilograms of marijuana, and undetermined amounts of Percocet and Xanax pills.

Most of the high-level defendants pled guilty pursuant to non-binding plea agreements and were sentenced by the district court judge here. The plea agreements contained recommended sentences or sentencing ranges based in part on stipulated amounts of cocaine,11 with some recommendations contingent on the defendant's not-yet-determined Criminal History Category (“CHC”). Most of the agreements also included stipulations on whether a role-in-the-offense enhancement should be added to the Base Offense Level (“BOL”), and the extent of any such increase. The stipulated drug amounts, the sentences or sentencing ranges recommended by the government, and the sentences actually imposed for nine of the top ten defendants, including Reyes, are as follows:12

Thus, as the chart reveals, Reyes's 360–month sentence is the longest of any of the defendants, including both the conspiracy's leader and a career offender, and his term exceeds most of the other sentences by wide margins. Also pertinent to this appeal are two facts applicable only to Reyes, Sevilla–Oyola and José Colón de Jesús—i.e., the three co-defendants with the longest sentences. First, the district court heard witness testimony at sentencing only for those three, relying in the remaining cases on the PSRs, the plea agreements, and the parties' arguments. Second, those three defendants were implicated in notorious killings: the so-called “Pitufo Murder” and the “Pájaros Massacre.” Reyes and Colón de Jesús were acquitted in state court on charges stemming from the latter, in which a drug dealer, his two-year-old daughter, and another individual were killed in a barrage of shots in Toa Alta, Puerto Rico, in 2009. Sevilla–Oyola was identified by a cooperating witness as the shooter in the Pitufo incident, in which another co-conspirator under federal supervision (José Manuel Torres–Morales, known as “Pitufo”) was murdered in front of the federal courthouse in Hato Rey, Puerto Rico, in 2007. See United States v. Sevilla–Oyola, 770 F.3d 1, 5 (1st Cir.2014).

B. Reyes's Sentencing

1. Plea Agreement and PSR

As reflected in the chart above, Reyes's plea agreement stipulated that he was accountable for at least 2 but less than 3.5 kilograms of cocaine, which produced a BOL of 28. With a two-level enhancement because the drug trafficking occurred in a protected location, see U.S.S.G. § 2D1.2(a)(1), and a three-level reduction for acceptance of responsibility, see U.S.S.G. § 3E1.1, his proposed total offense level was 27. The government agreed to recommend a sentence of 87 months if Reyes's CHC was III or below, and if the CHC was higher, the government would recommend a sentence at the low end of the applicable guidelines range. No role enhancement was proposed. Because Reyes's CHC turned out to be IV, the Guidelines range consistent with the plea agreement was 100 to 125 months.

Reyes's PSR, however, calculated a substantially longer Guidelines sentence. Although the PSRs for most of the defendants listed in the chart above used the cocaine amounts stipulated in the plea agreements for their BOL calculations, the PSRs for Reyes and the conspiracy's head, José Colón de Jesús, used the higher quantities alleged by law enforcement authorities. Based on roughly 241,000 kilograms of marijuana,15 Reyes's BOL was 38. The PSR added both the two-level enhancement for protected location proposed in the plea agreement and three levels for Reyes's role in the offense. With the three-level reduction for acceptance of responsibility, the total offense level was 40. The PSR calculated Reyes's CHC to be IV, producing a Guidelines range of 360 months to life. The PSR also included information about the Pájaros Massacre as criminal conduct that might warrant a departure or variance from the Guidelines range.

In a February 2012 order applicable to both Reyes and Colón de Jesús, prior to their sentencing hearings, the district court directed the government to produce evidence on the amount of drugs trafficked during the conspiracy and to secure testimony from three sources: (1) a “cooperating witness” to the Pitufo Murder, (2) “the principal witness” to the Pájaros Massacre, and (3) witnesses familiar with “the facts behind the extensive criminal activity that ended up dismissed in state court throughout the years .” The court advised the two defendants that it had obtained the full transcript of the Pájaros trial and that it was available for examination by the parties. The court also made available to those two defendants, along with Sevilla–Oyola, police and FBI reports about the Pitufo Murder that it had obtained from the United States Marshal. Subsequently, however, after an informal status conference with counsel for Colón de Jesús, the district court notified both defendants that it would not take the Pitufo Murder or Pájaros Massacre into account for their sentencings. The court stated that it would hear testimony only on “drug amount, dismissed criminal prosecutions and nature of the drug trafficking charge in referen[ce] to firearms and violence.”

2. The Hearing and Aftermath

At the outset of Reyes's sentencing hearing, held in May 2012, the district court noted that it was not bound by the plea agreement and observed that “the amount of drugs in this case is so huge, so huge that it makes it up to a certain point unacceptable to adopt the stipulation.” In response, defense counsel pointed out that the same judge had accepted the drug stipulations during prior sentencings of Defendants # 2 (Angel Colón de Jesús), # 3 (Jimenez–Echevarria), # 6 (Rivera Bermudez), and # 7 (Nazario–Pedroza). The court responded by noting that the conspiracy's leader had “relied a lot” on Reyes and that “he was also like a Lieutenant to number one.” The court further explained that he was “not happy ․ with the stipulation in this case regarding narcotics, because there is an issue as to whether he was something larger, bigger in the organization than what he pled to,” and noted its understanding that Reyes “was the owner of th[e] Boston Red Sox marijuana [marijuana sold for $6/bag], and he was like an assistant to number one.” The court directed the government to present its witness to clarify “the issue of role and the issue of drugs.”

Before the witness took the stand, however, the parties and the court discussed defense counsel's contention that the government's examination of the witness would be a breach of the plea agreement. After the objection was addressed,16 and before the witness testified, counsel again raised the differential treatment among defendants, prompting the following exchange:

COUNSEL: But then my question, Your Honor, that we have with these people that are numbers higher in the ranking of this Indictment—

COURT: It doesn't matter. The cases are different.

COUNSEL: The cases are the same, Your Honor.

COURT: Presentence Reports are different. Criminal records are different.

COUNSEL: Exactly, but then I go back, why is it, why is it that the Presentence Report for my client is different than probably the other ones? I wonder if the other ones received the same calculations as to the whole amount of the drugs, and not—than Mr. Reyes did? And I can only say, I can only understand or figure out that it might be because he was acquitted of the P [á]jaros massacre.

We can always say that we are not considering this, we are not considering that; but it's still there, because no one wants to be—

COURT: Absolutely, it's still there. Yes.

COUNSEL: It's still there. So that's why I wanted to place the Court in a position to understand that my client was the only one that presented an alibi defense [in the Pájaros case], and that alibi defense was not shady. It was good.

And then because that probably we—the Court and the Probation Office and the Government might have a different eye to my client, instead, and probably he would get there like the other ones that were already sentenced.

Defense counsel sought to provide the court with evidence substantiating Reyes's Pájaros alibi, but the court said it was “not going to go anymore into the P[á]jaros massacre,” noting that the crime “is a dead case for me.” The court then directed the government to call its witness, Carlos Burgos Rodriguez (“Burgos”), to elicit testimony on Reyes's “involvement with this particular brand of marijuana that we've been talking about” and “his involvement with number one in the context of role.”

Burgos, a cooperating co-defendant who was a runner at the Virgilio Dávila Public Housing Project,17 testified that Reyes's “position there was to find the workers, to assign the shifts, to organize the runners, and to look out for the stash that was kept there, the drugs.” He reported that Reyes at one point received a commission from the sale of four drugs—including “the Boston”—that “belonged” to Colón de Jesús, but that the profits from that brand of marijuana (i.e., the Boston Red Sox variety) were divided among five individuals after the death of Pitufo: Reyes, Sevilla–Oyola, Jimenez–Echevarria, Rodriguez–Rodriguez, and an individual whom Burgos identified as “Menovito” and added, “may he rest in peace.”

Based on this testimony, the district court found “by more than a preponderance of the evidence” that Reyes was a supervisor, and it concluded that the three-level role enhancement included in the PSR calculation was appropriate. The court also stated that “a person with this degree of supervision and involvement in the conspiracy knew or should have known the extent of the drug amounts that were handled by this conspiracy, and it wasn't what is stipulated.” In response to these findings, defense counsel returned to the issue of disparity:

COUNSEL: Then, Your Honor, I would just renew my objection as to the other people, the other defendants that were sentenced prior to Mr. Reyes Santiago, who have administrator or supervision roles, and they were not attributed this amount of drugs as—and they were probably foreseeable to them, too. And that treatment—

COURT: I don't remember that being an issue in those cases. I don't remember—

COUNSEL: Because it was—

COURT:—ever having a situation like that in the other cases, how much were they responsible—

COUNSEL: It was not an issue, because the Court followed the Plea Agreement, did not bring any witness to testify as to the amount of drugs. So that's why. And probably the presentence report did not show that information. So that is basing my objection that there is a disparity [in] treatment as to Mr. Reyes Santiago and the other co-defendants in this case as—

Counsel again speculated that the different treatment occurred because the Pájaros Massacre “was still in the back of the mind of everyone.” The probation officer, who attended the hearing, disclaimed reliance on the Pájaros Massacre “to achieve the calculation of the drugs, and his participation in the conspiracy.”

As the discussion continued, defense counsel noted that Sevilla–Oyola was similarly situated to Reyes. The court responded that Sevilla–Oyola was serving a “hefty sentence” and, in addition, did not have the same role as Reyes. Counsel replied that Sevilla–Oyola's long sentence was attributable to his career offender status and observed that, in addition, Burgos had testified that Sevilla–Oyola “was in charge when the other ones were in jail.” The court again attempted to draw a distinction between the two defendants:

COURT: And I did not enhance his [Sevilla–Oyola's] drugs, because the record did not allow that—I don't think that I had to do it. There was plenty there to deal with objectively without getting into this kind of—but here you have a situation whereby there is big time drug dealing for a number of years here, and the stipulation—if the stipulation were 15 kilos, 30 kilos, or something of the sort, but two but not more than 3.5 kilos? That's the stipulation? Imagine.

Additional colloquy followed, in which the court noted its responsibility to evaluate plea agreements and its prerogative to reject drug amount stipulations. The court assured counsel it would not depart or vary from the Guidelines, but said it could not accept the stipulated drug amount and instead would sentence Reyes based on the PSR calculation. Accordingly, using a sentencing range of 360 months to life imprisonment, the court imposed the low-end 360–month sentence.

Reyes filed a motion seeking reconsideration of the sentence, arguing that it was procedurally defective and substantively unreasonable. He again raised an objection to the disparity among co-defendants, which he attributed to his prosecution for the Pájaros Massacre. Reyes also asserted, inter alia, that the government had breached the plea agreement by eliciting Burgos's testimony, which was directed toward increasing his total offense level. In its response, the government emphasized that it stood behind the plea agreement and the parties' stipulations, and it noted the court's explicit disclaimer of reliance on the Pájaros Massacre. The government did not address the disparity argument. The district court denied the motion, and this appeal followed.18

C. Discussion

Reyes asserts three errors that he claims require resentencing on the drug count: (1) the government breached the parties' plea agreement by actively eliciting prejudicial testimony from Burgos while remaining silent on the disparity issue; (2) the district court abused its discretion by “act[ing] as a prosecutor” and demanding witness testimony to support a sentence higher than the term of imprisonment proposed in the plea agreement, Br. at 43; and (3) the court abused its discretion by treating Reyes uniquely among the top defendants in the charged conspiracy and rejecting the drug quantity stipulation in his plea agreement, a disparity Reyes' attributes to “reprisal for his acquittal in the Pajáros massacre trial.” Br. at 36 (internal quotation marks omitted).

We address the government's compliance with the plea agreement in Section II. D below. Reyes's second contention—that the district court overstepped its role by requiring the government to present witness testimony—is easily dispatched. Reyes entered into a non-binding plea agreement under Federal Rule of Criminal Procedure 11(c)(1)(A) and (B), which preserved the court's discretion over the sentence to be imposed. See United States v. Díaz–Bermúdez, 778 F.3d 309, 311 & n.2 (1st Cir.2015) (describing the different types of plea agreements available under Fed.R.Crim.P. 11(c)(1)); United States v. Torres–Oliveras, 583 F.3d 37, 41 (1st Cir.2009). We find no reversible error in the court's decision to hear testimony before deciding whether the agreement, which stipulated to a drug amount substantially lower than reported in the defendant's PSR, was appropriate. See, e.g., United States v. Eirby, 262 F.3d 31, 38 n.3 (1st Cir.2001) (“[T]he essence of a non-binding plea agreement is that the judge may override the parties' agreements.”); U.S.S.G. § 6B1.4(d) (“The court is not bound by [a stipulation of facts in a plea agreement], but may with the aid of the presentence report, determine the facts relevant to sentencing.”); see also 18 U.S.C. § 3661 (“No limitation shall be placed on the information concerning the background, character, and conduct of a person convicted of an offense which a court of the United States may receive and consider for the purpose of imposing an appropriate sentence.”). Apart from the separate issue of disparity, the court's inquiry was within the bounds of its discretion.

We thus turn to Reyes's contention that his sentence was unreasonable, and must be vacated, because of the disparity between his term of imprisonment and those of similarly situated codefendants.

1. Sentencing Disparity: Legal Principles

In fashioning an appropriate sentence, judges are directed by statute to consider “the need to avoid unwarranted sentencing disparities among defendants with similar records who have been found guilty of similar conduct.” 18 U.S.C. § 3553(a)(6); see also United States v. Reverol–Rivera, 778 F.3d 363, 366 (1st Cir.2015). We have observed that this provision is “primarily aimed at national disparities, rather than those between co-defendants,” United States v. Rivera–Gonzalez, 626 F.3d 639, 648 (1st Cir.2010) (internal quotation marks omitted), but we have recognized that a sentence may be “substantively unreasonable because of the disparity with the sentence given to a codefendant,” Reverol–Rivera, 778 F.3d at 366. See also United States v. Correa–Osorio, 784 F.3d 11, 28 n.25 (1st Cir.2015) (“[S]entencers can consider disparities between codefendants, we have noted—even though § 3553(a)(6) chiefly addresses disparities among defendants nationwide.”); United States v. Martin, 520 F.3d 87, 94 (1st Cir.2008) (“[D]istrict courts have discretion, in appropriate cases, to align codefendants' sentences somewhat in order to reflect comparable degrees of culpability—at least in those cases where disparities are conspicuous and threaten to undermine confidence in the criminal justice system.”).

We have routinely rejected disparity claims, however, because complaining defendants typically fail to acknowledge material differences between their own circumstances and those of their more leniently punished confederates. To present “[a] well-founded claim of disparity,” a defendant must compare apples to apples, United States v. Mateo–Espejo, 426 F.3d 508, 514 (1st Cir.2005), and the myriad factors that come into play at sentencing make it difficult to isolate “identically situated” co-defendants, Rivera–Gonzalez, 626 F.3d at 648 (internal quotation marks omitted). We have noted, for example, the permissible distinction between co-defendants who go to trial and those who plead guilty, see, e.g., United States v. Rodríguez–Lozada, 558 F.3d 29, 45 (1st Cir.2009), between those who cooperate and those who do not, see United States v. Rossignol, 780 F.3d 475, 478 (1st Cir.2015), and between those whose cooperation is “prompt and full” and those whose cooperation is “belated and grudging,” Mateo–Espejo, 426 F.3d at 514. We have pointed to various other facts that can undermine an assertion of unjustified disparity. See, e.g., Rivera–Gonzalez, 626 F.3d at 648 (noting that some of appellant's “co-defendants were first-time offenders, ․ others were convicted of possessing a smaller quantity of drugs [,][s]till others received point reductions for the minor role that they played in the offense”). Nonetheless, although “our general rule of thumb is that a defendant is not entitled to a lighter sentence merely because his co-defendants received lighter sentences,” there may be “reason for concern” if “two identically situated defendants receive different sentences from the same judge.” Id. (internal quotation marks omitted).

2. Standard of Review

Sentencing challenges often include both a complaint of procedural error and an assertion that the sentence imposed is substantively unreasonable. Rivera–Gonzalez, 626 F.3d at 646. Where both types of objections are made, we ordinarily will consider first “whether the district court committed any procedural missteps and, if the sentence is procedurally sound, we then ask whether the sentence is substantively reasonable.” Rossignol, 780 F.3d at 477. In both contexts, we review the district court's discretionary judgments for abuse of discretion, its findings of fact for clear error, and its conclusions of law de novo. Reverol–Rivera, 778 F.3d at 366. “Ultimately, the linchpin of a reasonable sentence is a plausible sentencing rationale and a defensible result.” Rivera–Gonzalez, 626 F.3d at 647 (internal quotation marks omitted).

Reyes asserts that the disparity between his sentence and the sentences of his co-defendants reflects both procedural and substantive error. Most of his contentions target procedural flaws. He argues that the district court relied on improper facts (i.e., the Pájaros Massacre) when it rejected the stipulations in his plea agreement, see id. at 646 n.9; failed to adequately explain why he should be held responsible for the distribution of substantially more drugs than higher-level members of the conspiracy, see id.; and neglected its statutory obligation to avoid unwarranted disparity among similarly situated defendants, see, e.g., United States v. Flores–Machicote, 706 F.3d 16, 24 (1st Cir.2013) (citing 18 U.S.C. § 3553(a)(6)). At bottom, however, his objection is that his sentence is substantively unreasonable because it far exceeds the sentences imposed on equivalently culpable co-defendants without any justifiable rationale for the greater severity. Thus, because Reyes's procedural concerns can be captured in our discussion of substantive reasonableness, we move directly to consider whether “the court has provided a plausible explanation, and the overall result is defensible.” United States v. Torres–Landrúa, 783 F.3d 58, 68 (1st Cir.2015) (internal quotation marks omitted).19

3. The Reasonableness of Reyes's Sentence

To briefly recap the disparity of which Reyes complains: he was sentenced to a term of 360 months on the drug count based on a substantially greater drug quantity than other similarly situated members of the conspiracy,20 producing a longer sentence than even the conspiracy's chieftain (324 months) and the one high-level co-defendant who was classified as a career offender (240 months). Reyes does not, and could not, argue that his sentence would be unlawful in all circumstances. Even the limited quantity of cocaine for which he accepted responsibility (2 to 3.5 kilograms) exposed him by statute to a term of imprisonment up to 40 years. See 21 U.S.C. § 841(b)(1)(B)(ii).

Reyes argues, however, that his personal background, criminal history, and involvement in the charged conspiracy do not stand out among the circumstances of his co-defendants as warranting a uniquely harsh sentence. He points, in particular, to Burgos's testimony revealing—as the district court acknowledged—that he was not the “owner” of the Boston Red Sox marijuana, as his PSR had suggested. Rather, he was one of five co-defendants who shared in the profits of that particular drug, which was owned by Colón de Jesús. He asserts that other defendants were identified as drug point owners and that Burgos's testimony did not substantiate the district court's assumption that Reyes was the overall first lieutenant for Colón de Jesús.

Without question, Burgos's testimony confirms that Reyes was a high-level member of the conspiracy. Burgos identified Reyes as a manager at the Virgilio Dávila Public Housing Project (finding workers, assigning shifts, organizing runners), and, as the district court noted at the hearing, the evidence showed that Reyes “acted as one of the right hands” of Colón de Jesús. Indeed, the district court could reasonably infer that Reyes's position in the conspiracy's hierarchy was higher than suggested by his # 9 ranking in the indictment. Of the five individuals whom Burgos said received profits from the Boston Red Sox marijuana, Reyes was the only one listed in the indictment's second tier—i.e., an alleged Manager/Drug Owner—rather than in the Leader group.

Nonetheless, the record does not show that Reyes played a uniquely prominent role in the conspiracy such that, in calculating offense levels, there was a factual basis for holding only him accountable for the full quantity of drugs allegedly distributed by the conspiracy. As defense counsel emphasized, the district court heard details about Reyes's conduct because it insisted that the government produce a witness at his sentencing hearing—a procedure the court followed only with Reyes and the two other defendants who also had been implicated in the Pitufo murder or the Pájaros Massacre. The court did not solicit comparable information about the responsibilities of the five other top defendants listed in the chart above. Although the conspiracy operated at four housing projects, Burgos's testimony placed Reyes at the helm of only one, making it likely that others played similar roles at other locations. Indeed, Burgos testified that Wilfredo Rodriguez–Rodriguez (# 5, who has yet to be sentenced), was “in charge of the crack” and the leader at the Brisas Housing Project. Reyes's PSR identifies a dozen other defendants who were named as owners of particular types of cocaine and marijuana, including two who were referred to as leaders at the Falín Torrech and Las Gardenias Public Housing Projects.

Notably, the court's concern about the stipulated drug quantities was not limited to Reyes's plea agreement. At Rivera–Bermudez's sentencing, for example, the court observed that it had “a number of issues with the gradation ․ of several defendants in this case.” Elaborating, the court stated:

I do think that the plea negotiations resulted in a very beneficial relevant conduct stipulation, which in some of these cases I may not accept, and in some I will, and I have․

․ [I]n this case, I am accepting at least 3.5 kilos but less than five kilos of cocaine, with the understanding that I do think that it would have been very, very, very, very, easy to—for this to be a lot, lot higher had this case been tried. But that's fine. Let's leave it at that.

He definitely was a leader. What I'm going to do is, he had a heroin and cocaine drug point. He accepted being a leader, but even if he didn't accept being a leader, he was a leader. What I am going to do ․, rather than use four levels, I'm going to use three levels for that. More than anything else, not because four is not the number, but as an act of good faith trying to figure out a way to make some sense to the plea bargaining process.21

The court similarly accepted the amount stipulated for Sevilla–Oyola, despite misgivings. The court explained:

The conspiracy to possess, even though he was a drug point owner, gave him in the context of the plea negotiations a very, very, very beneficial stipulation regarding relevant conduct. He pled with a recommendation that I accept at least two kilos but less than 3.5 kilos of cocaine․ And when I took his plea, a reader of the plea colloquy must clearly see that I was kind of puzzled, a little bit concerned with the stipulation, but that's the stipulation. And the Government is bound by it.

And other than the testimony of Carlos Burgos in one of the sentencing hearings that he mentioned a little bit more drug, perhaps nine kilos or something of the sort, the truth of the matter is I don't have anything that I can put my finger on other than logic and inference as to how much drug there is. So I have to go by the stipulation, at least two kilos, but less than 3.5 kilos of cocaine; ․ but with the clear understanding, and I will say it for the record, that I have no doubt in my mind that there was a lot more drug attributable to him than what he was able to stipulate, because that's something that is so obvious from the facts of this case. But, anyway, let's go with at leas[t] two kilos, less than 3.5 kilos.

Meanwhile, for two of the highest ranked defendants, the court accepted the stipulations with little comment. In a short sentencing proceeding for Angel Colón de Jesús (# 2), the court made the following remarks about drug quantity:

When you go to the drug counts, and due to the stipulation that appears in the record regarding relevant conduct which we're willing to accept for purposes of the discussion, I know there's a lot more drug than this involved, but in this case, I am willing to say I'll take 3.5 but less than five kilos of cocaine.

In an even shorter proceeding held to sentence Jimenez–Echevarria (# 3), the court addressed the issue as follows:

This case involves a huge conspiracy in a number of housing projects in Bayam [ó]n involving at least four of these projects that have been no man's land in reference to drugs for more than at least since in the ′90s. Early ′90s, we already were having cases from Virgilio D[á]vila, major cases involving Virgilio D[á]vila.

Anyway, he got a stipulation, which I will honor in this case, not necessarily in every case, for at least two but not more than 3.5 kilograms of cocaine; distribution within one thousand feet of a protected location; use of and carrying firearms in furtherance of the drug crime.

At the sentencing hearing for the conspiracy's leader, Colón de Jesús, the government acknowledged that the plea deal took into consideration, inter alia, “the fact that the case was a big case, 110 defendants.” In explaining its decision to reject the recommended sentence, the court recognized the government's need for plea agreements:

I think that the problem is that when you are dealing with cases of this magnitude, and you have an Indictment that calls in 110 people, you are dealing with an unmanageable case, an unmanageable situation.

And of course when you are dealing with that, and you have also in the docket of this Court many, many, many cases of that same nature going on at the same time, not counting the death penalty cases and all, it's impossible, impossible to try them all. Therefore, what happens is that in order to get rid of the cases, people involved in the process, in the best of faith, in the best of faith, try to find solutions. They stipulate amounts that are palatable, that are acceptable to the people who are about to take the responsibility.

The court concluded, however, that it could not accept the stipulated amount—5 to 15 kilos of cocaine—in light of Colón de Jesús's position as head of the entire operation, though it refrained from the most severe punishment:

I'm going to be mindful of the fact that I never counted the drugs myself. I'm mindful of the fact that I sentenced other people in a particular way. I am mindful of the fact that he was the leader, organizer, and in charge of this, the chairman of the board as I called it before. And I could use easily the top end of the guidelines in this case, the lower end of the guidelines in this case, but I cannot honestly use the stipulation.

Using the highest BOL (38) associated with drug quantities, see supra n.11, the court calculated a sentencing range of 324 to 405 months and initially announced a 360–month term for Colón de Jesús on Count One. His attorney pointed out that even the low end of the range, when combined with the mandatory five years on Count Two, produced a sentence 15 years or more longer than anticipated by the plea agreement (i.e., a sentence totaling 32 years, compared to “an expectation between 15 and 20”)—a difference she said was far greater than typical when a sentencing judge rejects a drug trafficking plea deal. The court then agreed to reduce the term on Count One to 324 months.

We think it evident from these excerpts that the court's attitude toward Reyes and his plea agreement—and particularly the stipulated drug amount—was decidedly different from its approach toward most, if not all, of the other high-level defendants. In multiple instances, the court accepted the stipulation without dwelling on it. On two occasions, it accepted the stipulation after observing that it had no independent knowledge of the actual drug amount. Even in selecting a sentence for Colón de Jesús, the court was “mindful of the fact that [it] never counted the drugs myself” and “mindful of the fact that [it] sentenced other people in a particular way.”22

In Reyes's case, by contrast, the court explained its rejection of the stipulated drug quantity and its use of the highest drug-related offense level (i.e., BOL 38) by pointing to Reyes's high position in the conspiracy: “All I am relying on is a person who is a supervisor at that level has the foreseeability of how much drug is involved when he has been in a conspiracy since at least 2005.” The court further noted that the 30,000 kilograms of marijuana that trigger BOL 38 would be reached even without counting all the drugs identified in the PSR: “[I]t's easy, easy, easy to reach the 30,000 [kilo] figure when you convert to marijuana.” It also observed that “no matter how you calculate it, he's a level 38. There is no way to do it differently.” This depiction of the drug quantity attributable to Reyes could be applied as well to the other defendants in the chart, yet the court's approach was markedly more lenient in other sentencings.

To be sure, there are significant differences in criminal histories among the co-defendants. Reyes's PSR calculated that he was in CHC IV based on two prior convictions and his violation of supervised release. Most of the others had less substantial criminal backgrounds. The Probation Office placed Nazario–Pedroza (# 7), Medina–Rivera (# 8), Ramos–Oyola (# 10), and Jose Colón de Jesús in CHC I, and calculated a CHC of II for Jimenez–Echevarria (# 3) and Angel Colón de Jesús (# 2). On the other hand, the PSR for Rivera–Bermudez (# 6) noted three prior convictions, two of which involved possession of heroin and cocaine, yielding the same CHC as Reyes, IV.23

These differences in criminal history, however, are reflected in the Guidelines ranges apart from the BOL determined by drug quantity. The Guidelines Sentencing Table sets out six possible sentencing ranges for each offense level, depending upon the defendant's CHC. See U.S.S.G. Ch. 5 Pt. A. Offense level 38, for example, produces a sentencing range of 235–293 months in CHC I and a range of 360 months to life in both CHC V and CHC VI. Indeed, Rivera–Bermudez's sentence—151 months—was the longest imposed on the nine co-defendants in the chart except for Reyes and the two defendants with unique circumstances (i.e., the top position in the conspiracy or career offender status), reflecting his higher CHC. If Reyes had been held responsible for 3.5 to 5 kilograms of cocaine—the largest amount used for most of the codefendants—his BOL would have been 30 instead of 38 and, with the adjustments applied by the court, his total offense level would have been 32 instead of 40. Using his CHC of IV, his Guidelines sentencing range would have been 168 to 210 months instead of 360 months to life. If the court had used 2 to 3.5 kilograms as the applicable drug amount—the quantity the government stipulated for Reyes—his sentencing range would have been 135 to 168 months.24 Even if Reyes had been held responsible for 5 to 15 kilograms—the quantity the government stipulated for Jose Colón de Jesús—his sentencing range would have been only 210 to 262 months.

In sum, we find no rationale articulated by the court, and supported by the record, that justifies the uniquely harsh approach to drug quantity taken by the district court in sentencing Reyes. As discussed above, we cannot conclude, based on Burgos's account, that Reyes played a significantly larger role in the drug trafficking than other high-level defendants. To the contrary, Burgos's testimony suggests that others were equally culpable. Moreover, it is striking that, like his two co-defendants who also received relatively long terms of imprisonment, Reyes was accused of participating in a notorious violent crime. Although the court eventually decided not to factor the murders into the defendants' sentences for the drug conspiracy, Reyes's prosecution for the Pajáros Massacre appears to have been a significant factor in the court's decision to hear testimony about his role and the quantity of drugs foreseeable to him. Indeed, when defense counsel noted that the incident is “still there” despite assertions that it would not be considered, the district court acknowledged, “Absolutely, it's still there.” See supra Section II.B.2. It thus appears that Reyes's disparate sentence may be indirectly traceable to events the court said it was not taking into account, resulting in a conspicuous and facially unjustified disparity between his sentence and those of co-defendants with “comparable degrees of culpability.” Martin, 520 F.3d at 94.

This case thus presents the unusual circumstance of a sentence that is substantively unreasonable and, hence, an abuse of discretion, because of its substantial disparity with the sentences given to co-defendants and the absence of any identified, supportable basis for the inconsistency. See Reverol–Rivera, 778 F.3d at 366 (noting that a sentence may be “substantively unreasonable because of the disparity with the sentence given to a co-defendant”). This is not a matter of an incomplete explanation by the court. With respect to drug quantity, the record does not support the comparatively harsh approach taken toward Reyes by the same judge who sentenced the other high-ranking members of the conspiracy. The disparity is particularly pronounced and detectable here because of the numerous co-defendants who entered into similar plea agreements with the government.

Our colleague agrees that a resentencing is necessary, but he believes the district court should be given another opportunity to provide a supportable rationale for the disparate treatment of Reyes. His view is that the court's explanation for the disparity is inadequate. But the problem in this case is not the lack of explanation for the court's sentencing decision or the lack of clarity of its reasoning. In such cases, we do remand to allow the district court to provide the missing explanation. See United States v. Gilman, 478 F.3d 440, 446–47 (1st Cir.2007) (noting that the appeals court may remand “to provide the district court an opportunity to explain its reasoning at resentencing” if the expressed rationale was “less-than-explicit” or “a court has provided no explanation at the sentencing hearing” (citing cases)).

Here, however, the court's rationale is clear: because Reyes was closely tied to Colón de Jesus—i.e., “like a Lieutenant to number one”—he should be treated like him with respect to drug quantity. The problem, instead, is the lack of record support for the court's finding that Reyes played a larger role in the conspiracy as a manager at the Virgilio Dávila Public Housing Project than the other high-level defendants played as leaders, owners, or managers at other locations. Although Reyes happened to work at the same housing project where Colón de Jesus owned the drugs, that proximity alone cannot justify attributing to Reyes, uniquely among defendants with similar roles in the conspiracy, the quantity of drugs calculated in each of their PSRs. There are no other facts in the record that could support the differential treatment.25

We therefore conclude that it is necessary to vacate Reyes's 360–month sentence on Count One and remand for resentencing so that Reyes's sentence can be appropriately aligned with those of his co-defendants.26

D. Breach of the Plea Agreement

Reyes argues that the government breached his plea agreement when it: (1) went beyond brief and neutral questioning of Burgos to elicit testimony that would support a much higher BOL, (2) informed the court it had evidence to support a higher sentence, (3) objected to defense counsel's questions during cross-examination, (4) failed to address the disparity issue, and (5) did not oppose sentencing enhancements that were not contemplated by the parties. Where, as here, the defendant properly objects to the government's conduct in the district court, “breaches of plea agreements present questions of law for plenary review.” United States v. Gonczy, 357 F.3d 50, 52 (1st Cir.2004).

We have long recognized the competing responsibilities that confront the government at sentencing when it has entered into a plea deal with the defendant. “[T]he prosecution's solemn duty to uphold forthrightly its end of any bargain that it makes in a plea agreement” must be balanced against “its equally solemn duty to disclose information material to the court's sentencing determinations.” United States v. Saxena, 229 F.3d 1, 5 (1st Cir.2000) (citation omitted); see also United States v. Miranda–Martinez, 790 F.3d 270, 275 (1st Cir.2015) (referencing the “equilibrium [that must be] struck by these competing tugs”). In this case, however, we need not decide whether the government properly calibrated its “obligation to furnish relevant information” with its “corollary obligation to honor commitments made under a plea agreement” because we have determined that Reyes's sentence must be vacated for another reason. Saxena, 229 F.3d at 6. We therefore do not review the government's conduct in full, and consequently offer no judgment on the ultimate question of breach. See United States v. Almonte–Nuñez, 771 F.3d 84, 91 (1st Cir.2015) (“We consider the totality of the circumstances in determining whether a prosecutor engaged in impermissible tactics .”). Instead, we limit our comments to two objectionable aspects of the government's behavior to provide guidance for its approach to the proceedings on remand.

First, the government had an obligation to address the disparity issue, which it did not do either at the sentencing hearing or in its response to the defendant's motion for reconsideration. As described above, defense counsel repeatedly raised the discrepancy between the court's handling of drug quantity for Reyes and its treatment of drug quantity for other high-level defendants. Although the prosecutor reaffirmed the government's commitment to the plea deal multiple times during the colloquy, the government neither explained why it believed a low quantity stipulation was appropriate for Reyes despite his high rank in the indictment nor acknowledged that Reyes's sentence significantly diverged from those of other top-ranked defendants.27 By its silence, the government almost certainly reinforced the court's view that the record provided a basis for attributing more drugs to Reyes than to other defendants, and it impliedly endorsed the court's belief that the low stipulation was irrational and unjustifiable. Cf. Reverol–Rivera, 778 F.3d at 365 (noting the government's argument at the sentencing hearing that “any sentence higher than 108 months would create an unjustified disparity with his accomplice's already-imposed sentence”). It is not enough for the government to voice adherence to a plea bargain if its other actions—or inaction—belie its support for the deal. See Gonczy, 357 F.3d at 53 (“A plea agreement is a binding promise by the government and is an inducement for the guilty plea; a failure to support that promise is a breach of the plea agreement, whether done deliberately or not.”); United States v. Canada, 960 F.2d 263, 271 (1st Cir.1992) (“In the circumstances, ․ the Assistant United States Attorney's silence about [the defendant's] cooperation ․ becomes another factor which, when added with [others], suggests an implied repudiation of the government's bargain.”).28

Second, while the government cannot be faulted for complying with the district court's demand that it produce a witness at the sentencing hearing, it went beyond its obligation to the court by objecting when defense counsel, seeking to impeach Burgos's credibility, questioned him about inconsistencies between his testimony at the hearing and his testimony before the grand jury. The government's “duty to bring all facts relevant to sentencing to the judge's attention,” Gonczy, 357 F.3d at 53, does not come with license to impede defense counsel's effort to persuade the court to adopt the parties' agreed—upon sentencing recommendation. Indeed, the government in effect acknowledged at argument that silence in response to the cross-examination would have been more consistent with the government's duty to support the plea agreement,29 and we agree.30

We thus expect that, on remand, the government will be more circumspect and will adhere as strictly as possible to its obligation to support the terms of the plea agreement. As we have emphasized in the past, “a defend ant entering into a plea agreement with the government undertakes to waive certain fundamental constitutional rights; because of that waiver, the government is required to meet the most meticulous standards of both promise and performance.” Id. (internal quotation marks omitted).


A. Revocation Sentence: Background

In September 2000, based on a guilty plea he entered in an earlier drug conspiracy case, Reyes was sentenced to 60 months imprisonment followed by four years of supervised release. He was released from custody in March 2004, and twice thereafter violated the conditions of his release. Each violation resulted in revocation of supervised release and a term of imprisonment: six months the first time and nine months the second time. After completing the required periods of incarceration, Reyes again was released from federal custody and placed on supervised release in February 2009.

In May 2009, Reyes's probation officer alerted the district court to new violations of his supervised release requirements and also reported Reyes's arrest by local authorities in connection with the Pájaros Massacre. In May 2010, following Reyes's acquittal in the Massacre trial, he was immediately taken into federal custody for revocation proceedings for the latest supervised release violations. A few weeks later—in July 2010—Reyes was indicted in the 110–member drug conspiracy case that is the primary subject of this appeal, and the pending revocation hearing was first postponed and then combined with the sentencing proceedings in the later case. See No. 3: 99–cr–00264–JAF–CVR, Dkt. 339 (Nov. 12, 2010) (stating that “[t]he pending revocation request will be handled ․ as part of the proceedings in Criminal No. 10–251”).

The combined sentencing hearing took place in May 2012. In addition to the 420–month term of imprisonment imposed in the more recent drug conspiracy case (i.e., 360 months on Count One and 60 months on Count Two), the district court imposed a 24–month, consecutive sentence for Reyes's 2009 violation of conditions. The court concluded that the Guidelines range for the type of violations Reyes committed was 24 to 30 months, and it chose the lower end of that range. In imposing the 24–month term, the court explained that it was crediting Reyes—as his attorney had requested—for the year he spent in jail leading up to the Pájaros Massacre trial. Reyes subsequently argued to the district court, in an unsuccessful motion for reconsideration, that the 24–month sentence is unlawful because it exceeds the applicable statutory cap governing penalties for violations of supervised release conditions.

On appeal, Reyes reiterates his contention that, under the applicable version of 18 U.S.C. § 3583(e)(3), 36 months is the maximum aggregate term of imprisonment that may be imposed for violations of conditions related to his 2000 conviction. He thus claims that, because he previously served terms of six months and nine months for violating the same conditions, the maximum sentence available for his third transgression is 21 months (i.e., the 36–month maximum less the combined 15 months already served). In its brief, the government inaptly argues that the 24–month sentence is lawful because the court properly used the Guidelines manual in effect at the time of Reyes's original sentencing, and sentenced him below the 36–month statutory cap. In a memo responding to questions posed by the panel at oral argument, the government asserts that the court was not obligated to credit Reyes for time served for his prior supervised release violations.

B. Discussion

Notwithstanding the government's contrary view, the legal parameters of Reyes's revocation sentence are clear. Reyes is correct that, under well-established law, he must be sentenced based on the statutory limits in effect at the time of his original sentencing proceeding in September 2000. See United States v. Tapia–Escalera, 356 F.3d 181, 188 (1st Cir.2004) (citing Johnson v. United States, 529 U.S. 694, 701–02 (2000), where the Supreme Court noted that, “[s]ince postrevocation penalties relate to the original offense,” imposing sanctions retroactively would raise an ex post facto issue); see also, e.g., United States v. Thomas, 600 F.3d 387, 389 (5th Cir.2010) (per curiam). In addition, we have concluded, in accord with other circuits, that § 3583(e)(3) as it existed in 2000 set cumulative maximum sentences for all revocations of supervised release relating to the same original offense. See Tapia–Escalera, 356 F.3d at 187–88; see also United States v. Perry, 743 F.3d 238, 241 (7th Cir.2014) (noting the uniform interpretation of the circuits).31 Here, both parties agree that the applicable statutory maximum is three years.32 Hence, Reyes's sentence for the 2009 violation of release conditions should have been capped at 21 months, and the 24–month sentence imposed by the district court must therefore be vacated.

We offer an observation concerning the resentencing that will need to occur for the supervised release violation. We are uncertain how the district court determined that the applicable Guidelines range was 24 to 30 months. The court labeled Reyes's conduct as a Grade A violation based on criteria set forth in U.S.S.G. § 7B1.1 (1999), which sets out three levels of supervised release violations (Grades A, B, and C). Grade A violations are the most serious, and they include conduct constituting a federal controlled substance offense or firearm possession, if punishable by a term of imprisonment exceeding one year. Id. The Guidelines specify different sentencing ranges for each grade of violation, based on the defendant's CHC “at the time the defendant originally was sentenced to a term of supervision,” id. § 7B1.4 (Policy Statement) (Revocation Table), and on the class of felony committed, see supra n.32.

Reyes was on supervised release for a Class B felony, and it is undisputed that his CHC at the applicable time was I. Under Guidelines § 7B1.4, the recommended sentencing range is 12 to 18 months for a defendant in CHC I who commits a Grade A violation of supervised release conditions related to a Class B felony conviction. Hence, in choosing the appropriate sentence for Reyes's violation of conditions on remand, the district court should look to the correct Guidelines range, see, e.g., United States v. Figueroa–Figueroa, 791 F.3d 187,191 (1st Cir.2015) (noting that “district courts must still give respectful consideration to the now-advisory Guidelines” (internal quotation mark omitted)), and may not impose a term that exceeds 21 months of imprisonment.


To recap our holdings:

1. Federal Rule of Appellate Procedure 4(b) is a claims-processing rule whose time limitations may be waived or forfeited if the opposing party fails to challenge appellate jurisdiction in a timely manner. Here, the government both waived and forfeited the timeliness issue.

2. Reyes's 360–month sentence on Count One of the 110–defendant drug conspiracy indictment must be vacated because the record shows no justification for the wide disparity between Reyes's sentence and those of similarly situated co-defendants insofar as the gap is attributable to the substantially higher drug quantity for which Reyes was held responsible.

3. Reyes's 24–month sentence for violating conditions of supervised release must be vacated because, when aggregated with his prior sentences for violating the same conditions, it exceeds the applicable statutory maximum of 36 months.

Although both of Reyes's sentences must be corrected, we note that the district court attempted to take a moderate approach, in certain respects, in imposing those punishments. As recounted above, the court chose a sentence at the low end of the Guidelines range it deemed appropriate on the drug conspiracy charge, and it gave Reyes credit for his time in state custody when sentencing him for violating his conditions of release. In addition, the court decided against imposing another period of supervised release to follow the latter, consecutive term of imprisonment. Hence, despite flaws in its assumptions about the facts and the law, the court manifested an intention to “impose ․ sentence[s] sufficient, but not greater than necessary, to comply with the purposes” of sentencing identified by Congress. 18 U.S.C. § 3553(a). We expect that, on remand, the court will again endeavor to achieve that objective, aided by the government's more complete adherence to its obligations under the plea agreement.

Accordingly, we vacate both of the sentences challenged on appeal and remand for further proceedings consistent with this opinion.

I fully join the majority's persuasive analysis save for one part, regarding sentencing, where I see things somewhat differently.


The defendant, Reyes, contends that he has been treated disparately in sentencing from his co-conspirators for no good reason. And the majority is understandably troubled—as am I—that he was.

Reyes points to the fact that a number of his co-conspirators stipulated in their plea agreements—as he did in his—to a relatively small amount of drugs. But while those co-conspirators were sentenced on the basis of those stipulated quantities, the District Court refused to accept Reyes's stipulation.

As a result, the District Court used a dramatically higher drug quantity in calculating Reyes's base offense level under the Sentencing Guidelines. The District Court set Reyes's base offense level at 38 due to the higher drug quantity it found should be attributed to him. That level of 38 was much higher than that assigned to some number of his co-conspirators, who received base offense levels of 28 or 30 in consequence of the low drug quantity amounts to which they had stipulated in their plea agreements and which the District Court had accepted in calculating their guidelines ranges and thus in determining their sentences.

Moreover, by assigning this higher base offense level to Reyes (due to the higher drug quantity attributed to him), the sentencing judge assigned a much higher guidelines calculation to Reyes. And, in consequence, Reyes received a sentence on the drug conspiracy count of 360 months, which was far greater than his coconspirators' sentences (ranging from 108 months to 151 months).33 Even the co-conspirator that the District Court acknowledged was the conspiracy's true leader only received a prison sentence of 324 months on the drug conspiracy count. Yet Reyes was not identified in either the indictment or the plea agreement as being as high up in the conspiracy's hierarchy as others who received lesser sentences based on guidelines calculations that rested on the acceptance of low drug-quantity stipulations. Hence, Reyes contends that his sentence reflects an impermissible disparity in the way he was treated relative to other co-conspirators.

But while I agree with the majority that we should vacate and remand the sentence, the complexities underlying the kind of follow a somewhat different analysis in getting to that result. And the consequence of this different analysis is that I would allow the District Court more discretion on remand to explain the basis for this sentence than the majority seems willing to allow.


As the majority recognizes, the mere fact that a defendant is sentenced differently from his co-conspirators is not a reason to conclude that he has been sentenced impermissibly. As we have explained many times before, the sentencing statute aims primarily at addressing nationwide disparities, not ensuring equal treatment among co-conspirators. See, e.g., United States v. Reverol–Rivera, 778 F.3d 363, 366 (1st Cir.2015). And here there is no question that the sentence imposed on this defendant is in line with the sentences received by defendants who engage in the same conduct in which he was found to have engaged. After all, he was sentenced within a guidelines range that was itself predicated on a base offense level pegged to the quantity of drugs involved in the conspiracy for which he was responsible. And the interest in uniformity supports keying the base offense level in a drug conspiracy to drug quantity. Such keying avoids the disparity in treatment that could arise if the base offense levels—and thus the resulting recommended sentencing ranges—were not so keyed.

But the fact that the sentencing statute aims primarily at addressing national disparities does not mean that the statute aims exclusively at addressing that concern—a point we have also made numerous times. See, e.g., id. And so a sentencing disparity can arise from the disparate sentences handed out to co-conspirators—even if all received sentences that were in line with the guidelines.

The question, then, is whether there was such a disparity here, given the harshness of Reyes's sentence relative to those that some co-conspirators received. And that in turn raises the question whether there is something wrong with treating Reyes as having known about such a large drug quantity when others, also high up in the conspiracy, were sentenced as if they did not have such knowledge.

The simple answer is that there is certainly nothing wrong with this outcome if Reyes is differently situated in some relevant way from those co-conspirators whose stipulations as to drug quantity were accepted. See id. (“[D]ifferences in culpability can justify disparate sentences among co-defendants․”). But, unfortunately, the answer is not in fact so simple in a case like this one. And that is because of the difficulty in determining what should suffice as a showing that a defendant is differently situated—and thus not identical—to his co-conspirators in such circumstances.

The sentencing statute itself offers little guidance. See 18 U.S.C. § 3553. It instructs us to avoid “unwarranted” disparities in sentences. See id. § 3553(a)(6) (“The court, in determining the particular sentence to be imposed, shall consider ․ the need to avoid unwarranted sentence disparities among defendants with similar records who have been found guilty of similar conduct.”). But it leaves unclear exactly what avoiding unwarranted sentencing disparities means with respect to calculations made under non-binding sentencing guidelines that in turn depend on calculations drawn from judgments about whether to accept generous stipulations or not.

To be sure, guidelines calculations that are wrong can lead to sentencing errors, precisely because the sentence may be imposed only after a correct guidelines calculation is made. See Gall v. United States, 552 U.S. 38, 51 (2007). But the sentencing range recommended by the guidelines is not binding, and § 3553(a)(6) states that it is the “sentence” that must not be disparate. So the statute says little that bears directly on our issue, beyond making clear that disparities in sentences (of some kind) should be avoided.

Similarly, it is clear that factual findings on which guidelines calculations rest may be reviewed for clear error. See United States v. Gracie, 731 F.3d 1, 2 (1st Cir.2013). And so factual findings that are clearly wrong—when challenged by the defendant—cannot support a sentence. But the defendant here does not claim that the District Court clearly erred in finding him responsible for the entire quantity of drugs in the conspiracy. (Perhaps he could have made such a claim, though I doubt it would have been persuasive, based on the evidence on the record of his involvement in this conspiracy and the quantity of drugs sold by the conspiracy.) Instead of challenging the factual basis for finding that he, in fact, was responsible for such a large amount of drugs, the defendant contends only that even though such a finding of fact would not be clearly erroneous, it was error for the District Court to attribute that quantity to him when the District Court could have accurately attributed that same quantity to the other conspirators but did not do so.

Precedent also offers little guidance. The truth is that—for as many times as we have said that identically situated co-conspirators may not be sentenced disparately—I have yet to find a case before this one in which we have actually concluded that co-conspirators were identical yet were disparately sentenced.34 Nor have I found a case suggesting with any clarity how to go about determining what makes co-conspirators identical or distinguishable in a case with facts meaningfully analogous to ours. And a search beyond our own Circuit turns up nothing useful either. So while there are cases (scores really) noting how defendants may be different—for example, because one pled and another did not, see United States v.. Ayala–Vazquez, 751 F.3d 1, 31 (1st Cir.2014)—I have not come across any precedent (and the parties cite to none) that seems to give much guidance here.35

To be sure, I do not doubt that a clear disparity would exist if two people with identical roles in the same conspiracy and with identical offense levels and identical criminal histories—and thus identical guideline sentencing ranges—were sentenced by the same judge to different terms of imprisonment without an adequate justification. See United States v. Saez, 444 F.3d 15, 19 (1st Cir.2006). But that is not our case.

Reyes received the longest sentence of any of the conspirators. But he had a higher criminal history than most other conspirators. We thus deal here with the very different question of how to administer the calculation of the guideline sentencing range itself on the basis of what seems to be a claim of disparate fact-finding leading to disparate guidelines' calculations of base offense levels. And that is the question for which guidance is wanting.

Reyes does argue that, on disparity grounds and disparity grounds alone, the District Court should have held him to the stipulated drug quantity. And there is an intuitive appeal to that contention. Why should co-conspirators with roughly equivalent knowledge of the amount of drugs involved in the conspiracy be sentenced as if they in fact had different levels of awareness?

But I am nonetheless wary of mandating a rigid rule of equal treatment in determining whether stipulated drug quantities should be accepted. For although I recognize drug quantity determinations do influence sentences indirectly, insofar as guideline sentencing ranges influence final sentences, I am concerned about the risks of injecting a too-rigid disparity analysis into our sentencing review in cases of this sort, for fear of how it might affect the process of making and accepting stipulated quantity determinations.

As we have recognized, and as the record in this case reveals, guilty pleas to stipulated drug quantities serve a number of practical purposes even if they do not reflect the truth. So, too, do decisions to allow such stipulations to stand even when they are accurate only in the sense that the defendant certainly was responsible for at least that quantity (though the stipulated amounts are otherwise, sometimes wildly, understated). See United States v. Yeje–Cabrera, 430 F.3d 1, 23 (1st Cir.2005) (“Fact bargaining may arise when there are different views of the facts, counsels' ability to prove them, and their consequences.”).

That is in part why we have said there is no constitutional problem with the prosecutor and the defendant bargaining over factual stipulations as to drug quantity. See United States v. Hall, 557 F.3d 15, 21 (1st Cir.2009). And those practical considerations also help to explain why there is also not a problem with a judge choosing to accept the bargain—at least when it does not result in a defendant taking responsibility for more than he did—even if a more-searching inquiry would show that the stipulation was generous. Simply put, fact bargaining is central to the plea bargaining system, as it provides an incentive to plea rather than go to trial, and a means of potentially reducing the harshness of the punishment that otherwise might be imposed. See Yeje–Cabrera, 430 F.3d at 25.

In shaping the scope of our review of disparity-based challenges to a sentencing judge's decision to accept or reject factual stipulations in plea agreements that are favorable to defendants, it seems to me we should be equally sensitive to these practical considerations. Importing a rigid, substantive disparity analysis into our review of a district court's decision to reject a bargained-for fact, I worry, may undermine the foundations on which the fact-bargaining system operates. For a District Court with a strong suspicion that the real weight of drugs in a conspiracy is far higher than the factual stipulations reveal might then be in a curious position. The District Court might feel the need to reject stipulations it would otherwise accept in order to maintain its discretion to impute the full weight to particular co-conspirators to be sentenced in the future. And, in turn, those involved in the bargaining might have their incentives altered, too.

Were that the consequence, I am dubious that there would be much virtue in our having altered the calculus for sentencing judges in that way. It may be that a district court, considering the quality of cooperation of the various defendants, their particular roles in the conspiracy, and their past criminal histories, is trying to settle on individualized quantity determinations for each defendant that reflects a reasonable assessments of the individualized sentence that each warrants, given the characteristics of the defendant that the sentencing statute instructs the district court to consider. And, in the process of trying to arrive at those individualized sentences, the district court may choose, therefore, to accept certain stipulations and reject others not for no reason, but instead with an eye on the ultimate sentencing range it considers to be appropriate for each defendant in light of the considerations that the sentencing statute directs the sentencing judge to consider and that the guidelines do not otherwise seem to prohibit.

So, for example, a district court might chose to accept a low stipulated drug quantity for a conspirator with a high criminal history category. The judge might figure that the criminal history category, paired with the offense level dictated by the lower quantity, will effectively bring the defendant's guidelines sentencing range in line with what—all things considered, including levels of cooperation—it should be under the statute. And thus, the District Court might choose not to probe further into what the true amount was and accept the lower amount contained in the plea agreement. By contrast, the district court might choose to look past the stipulated amount for another conspirator whose role seemed greater but whose criminal history category was less, figuring that such an approach will, again, best capture the sentence that is appropriate for that defendant, given that defendant's individual role and characteristics. As to that defendant, therefore, a deeper dive into the facts might seem appropriate to ferret out the higher quantity that is true.

It is far from clear to me that a District Court may not take account of such holistic sentencing judgments—when rooted in the statutory considerations set forth in 18 U.S.C. § 3553—when considering whether to accept or reject stipulated drug quantities that are defendant-friendly. I thus would be wary of setting a rule regarding disparities among co-conspirators in calculating sentencing guidelines that could discourage the process through which low stipulations are forged and allowed to stand, especially given the uncertain (and unbriefed) impact such a rule might have on the practical administration of the criminal justice system.

I doubt we improve things by reducing the discretion of already-restricted sentencing judges (given mandatory minimum sentences) through an approach to disparity analysis that is not sensitive to these realities. Such an approach could have the perverse effect of making a plenty-punitive criminal justice system yet more punitive still.

Having said all that, I do not read the majority to be imposing the kind of rigid approach that most worries me. This case is fact-bound, and the majority's analysis is, appropriately, dependent on these particular facts as well. And the facts here are troubling. Nonetheless, rather than characterizing the District Court's sentence as substantively unreasonable—that is, inherently unjustifiable—as the majority does, I prefer to analyze this case through the lens of procedural reasonableness and inadequate explanations. Doing so, I believe, poses the least risk of unintended consequences while still fulfilling our obligation to ensure that a sentence meets the “the basic requirement of rationality[.]” Saez, 444 F.3d at 19.

Of course, as the majority rightly notes, the distinction between matters of substance and procedure is less than clear in this context—and, at the margins, may not matter. But it also might, particularly with respect to the scope of remand. And here I believe that is the case. Thus, in my view, the reason we must vacate and remand is only that the District Court failed to provide an adequate explanation for its conclusion that Reyes was not identically situated to those for whom drug-quantity stipulations were accepted. I would not go so far as to say (as I read my colleagues to conclude) that no such explanation is possible on this record, and thus that on remand the District Court has no choice but to sentence this defendant on the basis of a quantity that is closely aligned with the quantities on which his co-conspirators were sentenced. In my view, my approach—by treading more lightly—best accounts for the lurking complexities that inhere in this type of disparity-based challenge.


Still, I do agree with the majority that the sentence must be vacated and remanded. And so I need to say more about the nature of the disparity concern that I believe must be redressed on remand, which, in turn, requires that I describe in more detail both what I understand to be the District Court's explanation for the sentence that it chose and why I do not believe that explanation is adequate.

The chain of logic undergirding the District Court's sentencing in this case—as best I can tell—seems to go like this. There was a large conspiracy with many members, and they played different roles. There was an overall leader. There were leaders of many drug points in various housing projects. And there were other significant participants—like Reyes—who performed neither the role of overall leader nor the leader of a particular drug point. The key for the District Court, therefore, seems to be what it suggested was the close relationship between Reyes and the overall leader of the conspiracy. And, as best I can tell, the District Court thought that relationship matters for determining drug quantity in the following way.

The overall leader of the conspiracy, José Colon de Jesus, like Reyes, stipulated to a very low drug quantity in his plea agreement. But the District Court thought that number—which was dramatically at odds with the figure contained in the presentence report—was risible. As I read the record, the District Court quite plausibly explained that Colón de Jesus was, in effect, the “Chairman of the Board” and, like the head of a large franchising corporation, “relicensed ․ housing projects to people who were going to be the drug point owners of those ․ different housing projects, [and] would in turn be part of his organization.” For that reason, the District Court thought there was no way Colón de Jesus was not aware of the drug quantities involved in the conspiracy as a whole, which were plainly much larger than the stipulation reflected.

To be sure, other participants in the conspiracy plainly knew more than the stipulated quantities in their plea agreements reflected, as the District Court indicated while sentencing those co-conspirators. And some of those co-conspirators also actually ran particular housing project franchises. But the District Court concluded that Colón de Jesus was different from the other drug point leaders in his responsibility for the whole conspiracy from its founding.

I think that distinction suffices to make Colón de Jesus—or No. 1 as he is known—distinct from the other members of the conspiracy, such that he could fairly be charged with a quantity much higher than his stipulation reflected and much higher than the stipulation reflected in his confederates' plea agreements.36 I reach that conclusion notwithstanding that these co-conspirator stipulations were by no means accurate accounts of what those individuals actually knew. And I reach this conclusion notwithstanding that those stipulations likely also could have been rejected for far higher quantities, perhaps including quantities yielding the highest base offense level for a drug conspiracy charge. Given the realities of fact-bargaining over drug quantities, I would not say that a sentencing judge lacks the discretion to ignore a stipulation for someone like Colón de Jesus, given his different role in the conspiracy, even though it might be hard to say that, in fact, his confederates did not also know the conspiracy involved drug amounts sufficient to raise their base offense level to be as high as his.

Since in my view the District Court had a basis for holding Colón de Jesus responsible, in calculating his base offense level under the guidelines, for a higher drug quantity than the other leaders in the conspiracy—notwithstanding the absurdly low quantities assigned to those other leaders—the question that remains is whether there is any similar rationale for treating Reyes like Col6n de Jesus with respect to his quantity stipulation. The District Court seemed to think there was, and it is in the lack of clarity on that point where, in my view, the error lies.

In sentencing Reyes, the District Court was plainly concerned initially about Reyes's potential involvement in other murders unrelated to this conspiracy, as he was concerned about the involvement of Colón de Jesus in the murders as well. But the District Court eventually made it plain in the face of objections to considering those murders that it was not basing Reyes's sentence on the alleged murders in any respect. Instead, the sentence would rest in significant part on Reyes's role in this conspiracy and his knowledge of the drugs involved.

In that regard, as with Colón de Jesus, the District Court was troubled, according to the transcript, by the gap between the stipulated drug quantity in Reyes's plea agreement and the quantity attributed to Reyes in the presentence report. And so the District Court dug into that issue, rejecting several times the contention from Reyes's counsel that the District Court was effectively bound by the stipulation in Reyes's plea agreement due to the District Court's prior acceptance at the sentencing of other co-conspirators of the low stipulations set forth in their plea agreements.

The District Court unfortunately was less than direct in explaining why it was rejecting this disparity-based argument for accepting the drug-quantity stipulation that Reyes and the government had made. The District Court seemed much more focused on justifying its authority to reject a stipulation, contending that it was not required to mechanically adopt all stipulations and that it could base its sentence on an understanding of the actual quantity for which Reyes was responsible and not simply the quantity set forth in the plea agreement.

But the sentencing judge eventually did get around to offering a reason for treating Reyes differently from others with respect to the quantity and for finding him to be, like Colón de Jesus, fully responsible for the full amount of drugs in the conspiracy. The District Court seemed to be suggesting that Reyes was different from these other leaders in the conspiracy because, even though he was in some respects lower down in the conspiracy's chain of command, he was in fact so closely tied to Colón de Jesus that he should be treated differently from all the others save for No. 1 himself.

In this regard, the District Court identified Reyes as “like a Lieutenant” and “one of the right hands” of Colón de Jesus. And the District Court pointed out that Reyes worked directly with Colón de Jesus at the drug point that Colón de Jesus owned while Colón de Jesus carried out his duties as chairman of the board for the conspiracy as a whole. The District Court also noted that testimony even showed that Reyes “look[ed] out for the stash that was kept” in Colón de Jesus's housing project, though it is not clear to me whether the testimony referred to the entire conspiracy's central stash house or to the stash house for just the local housing project. Finally, the District Court remarked that Reyes was apparently involved in the conspiracy for a large number of years, dating back all the way to 2005, and cited Reyes's particular role in the conspiracy as the reason for holding him responsible for the higher quantity of drugs.

But if the District Court meant to make a finding that Reyes was like a chief of staff or aide-de-camp to No. 1—rather than like the rest of the department heads—he did not do enough to make that clear. I say that while recognizing the difficult position district judges are in, busy as they are in general and particularly when it comes to criminal sentencing. Certainly they are more burdened in issuing sentences than we are in reviewing them.

In the face of the seeming disparity in how the District Court treated the quantity stipulated by Reyes relative to other drug point heads, it is not enough to say that Reyes should be treated differently because Reyes is like “a” lieutenant or “one of the” right-hand men to the overall leader. That characterization does not-in and of itself-distinguish Reyes from other lieutenants and right-hand men whose stipulations were accepted. Nor did the District Court explain why he thought the conspiracy was organized in such a manner that a close aide to Colón de Jesus was plausibly more responsible for what went on than a drug point leader for an entire housing project.

So while the District Court's observations do point the way to a possible rationale for distinguishing Reyes from the others and for treating him, like Colón de Jesus, as responsible for a higher quantity, the District Court's explanation is too unclear and imprecise to suffice. And if the District Court meant to make some other argument about why Reyes should be treated so differently, it did not offer any such explanation.

Accordingly, I would vacate and remand for the District Court to explain why Reyes—and Reyes alone—should be treated like the conspiracy's overall leader as to a matter so critical to determining his sentence. For while apples only need be compared to apples, the District Court has some obligation to explain why it treated Reyes, uniquely, like an orange, given that no distinction may be inferred.

Perhaps the District Court on remand would determine that no answer is possible, in which case there is no question but that the District Court would have the discretion to impose a sentence more in line with Reyes's co-conspirators. United States v. Martin, 520 F.3d 87, 94 (1st Cir.2008) ( “[D]istrict courts have discretion, in appropriate cases, to align codefendants' sentences somewhat in order to reflect comparable degrees of culpability․”). But were the District to do so at that point, we would then have no need to probe further the limits of what, substantively, constitutes an unjustified disparity and what does not.

Perhaps, instead, the District Court would conclude that the defendant is an aide-de-camp to Colón de Jesus, and would further find that that fact makes him more responsible than his co-conspirators. Reyes could then appeal those findings to us, which we would review for clear error. Or the District Court might have some other explanation to offer for why this defendant's stipulation—like Colón de Jesus's but unlike the others—should be disregarded. If so, hearing it might help illuminate just what the effect on sentencing generally would be from denying the District Court the discretion to rely on such a reason for treating Reyes differently.

Whichever course the District Court might pursue on remand, I do not believe we should cut short the process by which we often proceed when we find a rationale for a sentence wanting—a process by which we point out why the explanation does not suffice and then remand for the District Court to reconsider the sentence. See United States v. Medina, 779 F.3d 55, 64 (1st Cir.2015); United States v. Mendez–Colon, 15 F.3d 188, 191 (1st Cir.1994); see also United States v. Cirilo–Muñoz, 504 F.3d 106, 127 (1st Cir.2007) (Lipez, J., concurring in the judgment).


By statute, we have discretion to limit (or not) the issues to be addressed on remand by fashioning orders “as may be just under the circumstances,” 28 U.S.C. § 2106, a power we have interpreted broadly “to allow appellate courts the flexibility to adapt their mandates to the particular problem discerned on appeal and to provide an efficient and sensible solution.” United States v. Garafano, 61 F.3d 113, 116 (1st Cir.1995). But, because the problem I discern here is the lack of an adequate explanation and not the impossibility of one, and because of my concerns about adopting an overly-rigid disparity analysis with regard to drug quantity calculations for guidelines purposes, I would vacate Reyes's sentence and remand for reconsideration and a more-fulsome explanation, based on adequate findings, rather than for imposition of a sentence based on a drug quantity like those stipulated to by some of Reyes's co-conspirators. I thus respectfully dissent as to the majority's disparity analysis, though I join in full the majority's resolution of the remaining issues presented in this case.


1.  Nineteen individuals in addition to Reyes also were charged in Count Two, including the first seventeen listed in the indictment. Each of those discussed in this opinion also received a 60–month consecutive term for aiding and abetting the use and carrying of firearms in furtherance of the drug conspiracy. See 18 U.S.C. § 924(c)(1)(A)(i).

2.  We use the term “ultimately” because at least one defendant was resentenced (to a lower term) as a result of challenges to his original sentence. See United States v. Sevilla–Oyola, 770 F.3d 1, 9, 16 (1st Cir.2014) (appealing a sentence of 345 months on Count One and 60 months on Count Two); United States v. Sevilla–Oyola, No. 3:10–cr–00251–JAF–4, Dkt. 3278 (Dec. 16, 2014) (imposing 300–month term of imprisonment on both counts).

3.  We denied this motion in a summary order issued on March 6, 2015. We now explain the basis for that decision.

4.  Reyes filed a motion for reconsideration within the initial fourteen-day period after judgment entered, and he filed a notice of appeal one day after the district court denied his motion for reconsideration. He thus adhered to the time limits he incorrectly thought applicable.

5.  As we have observed, nonjurisdictional claims-processing rules “do not define what cases courts can and cannot hear, and a modicum of flexibility attaches to their administration.” Alphas Co., Inc. v. William H. Kopke, Jr., Inc., 708 F.3d 33, 37 (1st Cir.2013).

6.  In United States v. Sadler, 480 F.3d 932 (9th Cir.2007), the court noted that the only statutory reference to the timing of criminal appeals it could find was at 18 U.S.C. § 3732. Id. at 938 n.6. That provision has no substantive content. The section is entitled “Taking of appeal; notice; time—(Rule)” and it has a subheading in all capital letters: “SEE FEDERAL RULES OF CRIMINAL PROCEDURE,” followed by the following text: “Taking appeal; notice, contents, signing; time, Rule 37(a).” The substance of Rule 37(a) now appears in the Federal Rules of Appellate Procedure. See Fed. R.App. P. 4 advisory committee's note (1967 adoption) (stating that Rule 4(b) is “derived” from Federal Rule of Criminal Procedure 37(a)(2) “without change of substance”).

7.  Arguably, the Supreme Court has itself indicated that Rule 4(b) is a non-jurisdictional claims-processing rule by suggesting that its predecessor, Federal Rule of Criminal Procedure 37(a)(2), was such a rule. In United States v. Robinson, 361 U.S. 220 (1960), the Court characterized the filing of a notice of appeal within the period prescribed by Federal Rule of Criminal Procedure 37(a)(2) as “mandatory and jurisdictional.” Id. at 224. However, in Eberhart, 546 U.S. at 18, the Court, in describing “the central point of the Robinson case,” linked “mandatory and jurisdictional” to the government's objection: “[W]hen the Government objected to a filing untimely under Rule 37, the court's duty to dismiss the appeal was mandatory.” The Court in Eberhart then continued its thought with a reference to its observation in Kontrick that it had been “less than meticulous” in using the label “jurisdictional,” Id. (internal quotation marks omitted):The net effect of Robinson, viewed through the clarifying lens of Kontrick, is to admonish the Government that failure to object to untimely submissions entails forfeiture of the objection, and to admonish defendants that timeliness is of the essence, since the Government is unlikely to miss timeliness defects very often.Id. We agree with the Ninth Circuit in Sadler that “[t]he clear implication of this statement is that the timeliness dictates of Rule 4(b) are forfeitable, because Rule 4(b) is a nonjurisdictional claim-processing rule.” 480 F.3d at 940.

8.  In United States v. Watson, 623 F.3d 542 (8th Cir.2010), the court apparently noticed the violation of Rule 4(b) on its own when the defendant's pro se appeal was filed. The court appointed counsel to represent the defendant and asked the parties to brief whether Rule 4(b) is jurisdictional and, if not, whether the government forfeited any objection to the timeliness of the notice of appeal. Id. at 544. The government moved to dismiss the appeal, and the parties then filed their briefs. The court concluded that Rule 4(b) is not jurisdictional and that the government “properly objected” in its motion to dismiss and in its responding brief. Id. at 544, 546.

9.  Although Reyes was sentenced for violating his conditions of supervised release in the same proceeding as for the conspiracy, we address separately his challenge to that additional term of imprisonment. See infra Section III.

10.  thus, varying levels of responsibility for the criminal activity. See, e.g., United States v. Fornia–Castillo, 408 F.3d 52, 57–58 & n.4 (1st Cir.2005).

11.  Under the Sentencing Guidelines, a defendant's Base Offense Level for a controlled substance crime is linked to the drug quantity for which he is deemed responsible. See, e.g., United States v. Rodriguez, 731 F.3d 20, 26 (1st Cir.2013) (noting that drug quantities have “corresponding guideline range[s]”); U.S.S.G. § 2D1.1(c) (Drug Quantity Table). The plea agreements presumably relied solely on cocaine amounts because that drug alone sufficed to achieve the punishments sought.

12.  We present the details for these defendants because their sentences are the ones relevant to Reyes's argument that he was improperly sentenced more harshly than other highly ranked members of the conspiracy. For convenience, we have limited the chart to those within the top ten positions in the indictment.DefendantStipulated Drug Amt.Recommended TermCoun t One Sentence# 1 José Colón de Jesús5–15 kilos180 months324 months# 2 Angel Colón de Jesús3.5–5 kilos168 months135 months# 3 Jimenez–Echevarria2–3.5 kilos108 months108 months# 4 Sevilla–Oyola2–3.5 kilosunspecified240 months 13

FN13. Each defendant listed in the chart, with the exception of Sevilla–Oyola, also received a consecutive 60–month sentence on the firearms count. See 18 U.S.C. § 924(c)(1)(A)(i) (imposing minimum consecutive sentence of five years for use and carriage of a firearm in connection with a drug trafficking offense). Sevilla–Oyola received an aggregated sentence of 300 months on both counts when he was re-sentenced following an appeal. See supra n.2. Hence, given the statutory mandatory minimum of five years that the others received for the firearms count, Sevilla–Oyola effectively received a 240–month sentence on the drug count and a 60–month sentence on the firearms count. Sevilla–Oyola also is distinctive because he qualified as a career offender, which substantially elevated his Guidelines range despite the low stipulated drug quantity..  FN13. Each defendant listed in the chart, with the exception of Sevilla–Oyola, also received a consecutive 60–month sentence on the firearms count. See 18 U.S.C. § 924(c)(1)(A)(i) (imposing minimum consecutive sentence of five years for use and carriage of a firearm in connection with a drug trafficking offense). Sevilla–Oyola received an aggregated sentence of 300 months on both counts when he was re-sentenced following an appeal. See supra n.2. Hence, given the statutory mandatory minimum of five years that the others received for the firearms count, Sevilla–Oyola effectively received a 240–month sentence on the drug count and a 60–month sentence on the firearms count. Sevilla–Oyola also is distinctive because he qualified as a career offender, which substantially elevated his Guidelines range despite the low stipulated drug quantity.

FN14. The higher stipulated drug quantity for Medina–Rivera (# 8)—the same as for Defendant # 1—may reflect that Medina–Rivera was a fugitive from the time the indictment was issued in 2010 until his arrest on October 28, 2014. See No. 3:10–cr–00251–JAF–8, Dkt. 3282 (Dec. 12, 2014). He was sentenced in April 2015.Defendant # 5, Wilfredo Rodriguez–Rodriguez, was arrested in March 2015 and is scheduled to be sentenced on September 24, 2015. His plea agreement stipulated a drug quantity of 15 to 50 kilograms of cocaine, perhaps reflecting his lengthy fugitive status and the fact that, at the time of his arrest, he was found in possession of firearms and drugs. See No. 3:10–cr–00251–JAF–5, Dkt. 3404 (April 21, 2015). The agreement calculates a total offense level of 37, but does not stipulate a CHC..  FN14. The higher stipulated drug quantity for Medina–Rivera (# 8)—the same as for Defendant # 1—may reflect that Medina–Rivera was a fugitive from the time the indictment was issued in 2010 until his arrest on October 28, 2014. See No. 3:10–cr–00251–JAF–8, Dkt. 3282 (Dec. 12, 2014). He was sentenced in April 2015.Defendant # 5, Wilfredo Rodriguez–Rodriguez, was arrested in March 2015 and is scheduled to be sentenced on September 24, 2015. His plea agreement stipulated a drug quantity of 15 to 50 kilograms of cocaine, perhaps reflecting his lengthy fugitive status and the fact that, at the time of his arrest, he was found in possession of firearms and drugs. See No. 3:10–cr–00251–JAF–5, Dkt. 3404 (April 21, 2015). The agreement calculates a total offense level of 37, but does not stipulate a CHC.

15.  Under the Sentencing Guidelines, when various controlled substances are involved in a crime, the quantity of each type of drug ordinarily is converted into its marijuana equivalent to calculate the BOL. See U.S.S.G. § 2D1.1 cmt. n.8(B), (D) (2012). Reyes's PSR thus used a total of 241,021.3265 kilograms of marijuana for his BOL calculation. Colón de Jesús's PSR instead relied solely on a quantity of cocaine—150 kilograms—to set his BOL. The BOL for both defendants was the same because the Guidelines provide that a BOL of 38 is triggered by 30,000 kilograms or more of marijuana and 150 kilograms or more of cocaine. See id . § 2D1.1(c). Although the particular drug amount attributed to each defendant is thus irrelevant—because both quantities meet the threshold for BOL 38—we note that the quantity used in Reyes's PSR was roughly eight times the quantity used in Colón de Jesús's based on the Guidelines' Drug Equivalency Table. The Table equates one gram of cocaine with 200 grams of marijuana (i.e., 241,000 kilograms of marijuana divided by 200 = 1,205 kilograms of cocaine; 1,205 kilograms is roughly eight times 150 kilograms).

16.  We provide additional detail on this colloquy in Section II. D below.

17.  Among other tasks, runners provided drugs to the sellers and collected sales proceeds.

18.  Reyes's disparity claim covers only the sentence imposed on the drug charge (Count One) because, as noted above, all defendants received the same 60–month consecutive sentence on the firearms charge (Count Two).

19.  The line between procedural and substantive sentencing issues is often blurred. We have placed the disparity argument in both categories. Compare, e.g., Rossignol, 780 F.3d at 477 (substantive, based on court's inconsistent application of an enhancement), with, e.g., United States v. Dávila–González, 595 F.3d 42, 47, 49–50 (1st Cir.2010) (procedural, based on codefendant's lighter sentence). Similarly, we have noted that “[t]he lack of an adequate explanation can be characterized as either a procedural error or a challenge to the substantive reasonableness of the sentence.” United States v. Crespo–Ríos, 787 F.3d 34, 37 n.3 (1st Cir.2015) (citing cases).

20.  As discussed supra, the drug amounts used to sentence Reyes and Colón de Jesús were effectively the same, even though Reyes's PSR used an amount eight times larger than the amount used in Colón de Jesús's PSR. Indeed, the district court at Reyes's sentencing noted that 30,000 kilograms of marijuana was enough to reach BOL 38 and, hence, it was unnecessary to rely on the PSR's total of 241,021.

21.  Earlier in the hearing, the court noted that it had been puzzled by the government's plea offers and sentencing recommendations for the conspirators in this case, describing them as “extremely lenient pleas for something that I think is not that lenient and that simple.” In discussing a role enhancement for Rivera–Bermudez, the court observed that he already was benefitting from the stipulated drug amount:He's being responsible for only 3.5 kilos less than five kilos of cocaine, when it is obvious from the knowledge that we have of this file that any of these leaders would be responsible for a lot more drug than that, a lot more drug.

22.  All of the sentencing proceedings described above occurred before Colón de Jesús's sentencing in July 2012, with Reyes's in May 2012 the latest of those. Although Medina–Rivera's sentencing occurred much later, in April 2015, the record does not currently contain a transcript of that proceeding. A docket entry reports that the hearing began at 10:38 AM and ended at 10:45 AM. See No. 3:10–cr–00251–JAF–8, Dkt. 3392 (April 14, 2015).

23.  The addendum to Rivera–Bermudez's PSR noted that the report erroneously stated that his criminal history points added up to CHC III. The addendum observed that, “in the abundance of caution,” the court calculated his Guidelines range using CHC III.

24.  The sentence recommended in the plea agreement was lower—100 months—because the government agreed not to seek a role-in-the-offense enhancement, which both the PSR and district court applied.

25.  Our colleague notes that we should be wary of adopting a “rigid approach” to disparities in the treatment of drug-quantity stipulations among co-conspirators. We agree. Sentencing judges have discretion to accept or reject stipulations in plea agreements, and a judge may treat co-defendants differently-if they are not similarly situated. Here, the district court identified a factual basis for treating Reyes uniquely. If the court's rationale were supported by the record, Reyes would not have a viable disparity claim. Our decision thus establishes no general rule about the nature of remediable sentencing disparity.Indeed, even on remand, the district court need not accept the drug quantity stipulated in Reyes's plea agreement (2 to 3.5 kilograms of cocaine). However, the record lacks support for exceeding the quantity (3.5 to 5 kilograms) that was the highest amount attributed to the others (besides Colon de Jesus) listed in the chart in Section II.A.

26.  Reyes also asserts unjustified disparity between him and his co-defendants in the district court's application of the Guidelines' role-in-the-offense adjustment. See U.S.S.G. § 3B1.1. The court, however, did not take a consistent approach for all defendants. For example, like Reyes, Rivera–Bermudez (# 6) received a three-level adjustment, though his plea agreement did not recommend any leadership adjustment. Angel Colon de Jesús (# 2) received a three-level adjustment, though his plea agreement contemplated a two-level increase. Jimenez–Echevarria (# 3) was given a two-level adjustment, as recommended by his plea agreement. In other instances (e.g., Nazario–Pedroza (# 7) and Ramos–Oyola (# 10)), the court followed the plea agreements and did not include role adjustments in its Guidelines calculations. Although we would not vacate Reyes's sentence solely on the basis of these inconsistencies, it would be appropriate on remand for the district court to revisit Reyes's role adjustment.

27.  The prosecutor offered an explanation—of sorts—for the terms of the plea agreement only in the context of discussing the appropriateness of a role enhancement. Although affirming the government's commitment not to recommend a role enhancement, the prosecutor reported that authorities had sufficient evidence to justify an enhancement but “took into consideration other things in the case.”On appeal, the government does not attempt to factually justify the drug quantity disparity, instead defending the sentence by noting that the district court “disagreed that the defendants were similarly situated.” The government also cites the court's reference to Reyes's criminal history and relies on the court's statement that it found the stipulation “in two or three of these defendants absolutely unacceptable.” As explained above, neither of those points offsets the disparity problem.

28.  Of course, if speaking up would be detrimental to the defendant's position, restraint would be an appropriate tactic. See, e.g., Miranda–Martinez, 790 F.3d at 274 (“[W]e have acknowledged that certain factual omission[s], helpful to the defendant, may be an implicit part of the bargain.” (second alteration in original) (internal quotation marks omitted)). The record reveals no reason for such restraint in this instance.

29.  The pertinent exchange between the appellate panel and the Assistant United States Attorney (“FAUSA”) was as follows:COURT: There are various points at which the defense counsel on cross-examination is asking certain questions, and the prosecution objects—AUSA: Yes, Your Honor, I noticed that in the transcript.COURT: How is that consistent with just providing the evidence to the judge?AUSA: I don't remember the specific basis for the objections, Your Honor, but I would agree it probably would have been better not to object, to affirmatively contest—COURT: And why do you think that would have been better?AUSA: It would be more consistent and akin with our duties to follow and abide by the terms of the plea agreement. There's a certain point where we have to fulfill our duty to the court. But I'm not sure the objection was on a substantive issue, but ․COURT: By that logic, when defense counsel sought reconsideration of the sentence you probably shouldn't have opposed the motion for reconsideration, which you did, isn't that correct?AUSA: We didn't, Your Honor, actually until the court ordered us to oppose it. He ordered the government to respond to it. Initially we did not file an opposition to it, Your Honor.

30.  Of course, we do not mean to suggest that the government must remain silent if defense counsel seriously misrepresents the facts. That was not the situation here.

31.  Before 2003, § 3583(e)(3) stated that “a defendant whose term is revoked under this paragraph may not be required to serve more than [the applicable maximum].” 18 U.S.C. § 3583(e)(3) (2000). In 2003, Congress added the words “on any such revocation,” so that the provision now states that “a defendant whose term is revoked under this paragraph may not be required to serve on anv such revocation more than [the applicable maximum].” 18 U.S.C. § 3583(e)(3) (2015) (emphasis added); see also PROTECT Act, Pub.L. 108–21, § 101(1), 117 Stat. 650, 651 (2003). Every circuit to consider the issue has held that the amendment “eliminate[s] the credit for terms of imprisonment resulting from prior revocations.” Perrv, 743 F.3d at 241–42 (citing cases) (alteration in original) (internal quotation marks omitted).

32.  The statute prescribes a three-year maximum for a “class B felony.” See 18 U.S.C. § 3583(e)(3). Pursuant to 18 U.S.C. 5 3559(a), “[a]n offense that is not specifically classified by a letter grade in the section defining it” is classified on the basis of its maximum term of imprisonment. A Class B felony is one for which the maximum term is 25 years or more, but not including life imprisonment (which signifies a Class A felony).Under this classification scheme, the drug conspiracy crime underlying Reyes's 2000 conviction is a Class B felony. He was sentenced to the statutory minimum of 60 months, with the related statutory maximum set at 40 years. See 21 U.S.C. §§ 841(a)(1) (B), 846.

33.  One co-conspirator, Carlos Sevilla Oyola—a career offender—was initially sentenced to life in prison but the District Court then changed that sentence to 327 months, 960 months, and then 345 months. United States v. Sevilla–Oyola, 770 F.3d 1, 6–9 (1st Cir.2014). We vacated and remanded, id. at 16, and the District Court ultimately sentenced Sevilla to 300 months on the drug conspiracy count.

34.  In United States v. Cirilo–Mufioz, 504 F.3d 106 (1st Cir.2007) (per curiam), two judges voted to vacate and remand the sentence while a third voted to affirm. Id. at 107. One of the two in the majority concluded both that the District Court's explanation was inadequate and that the defendant's sentence was substantively unreasonable for being unjustifiably disparate from a co-conspirator. Id. at 126 (Torruella, J., concurring in the judgment). The other judge concluded the explanation was merely inadequate and declined to address the substantive disparity issue. Id. at 127 (Lipez, J., concurring in the judgment).

35.  Reyes does cite United States v. Barnes, 602 F.3d 790 (7th Cir.2010), but that case does not help our consideration of this question in this Circuit. In Barnes, most participants in a cocaine conspiracy pled guilty with low-level quantity stipulations that the District Court accepted at sentencing. Id. at 794. Barnes was convicted after a trial and obtained the same low-level quantity stipulation from the government going in to sentencing. Id. But the District Court rejected the stipulation and imputed a much higher drug quantity to Barnes, which affected his guidelines calculation, without offering any reason for having done so. Id. The Seventh Circuit concluded that this disparate treatment as to quantity stipulations was error, vacated the sentence, and remanded for resentencing. Id. at 797. Barnes is of no help in our case for two reasons. First, the Seventh Circuit framed its analysis around review of the District Court's factfinding, concluding the drug quantity it attributed to Barnes was clearly erroneous, rather than around sentencing disparity per se. And second, under First Circuit precedent, Barnes, who went to trial, would have been differently situated from the co-conspirators who pled guilty. See Ayala–Vazquez, 751 F.3d at 31.

36.  Colón de Jesus's sentence is not before us in this appeal, and I take no position on any claim he may bring in the future. I do note, however, that Colón de Jesus appealed his sentence to this court and we affirmed, concluding in an unpublished judgment that there was no abuse of discretion. See United States v. Colón de Jesus, No. 12–1936 (1st Cir. June 23, 2014). In particular, we rejected Col6n de Jesus's argument that the District Court created an unwarranted disparity when it looked behind his stipulated drug amount and sentenced him using a higher drug quantity than was used for his co-conspirators. Id. And we did so because we concluded that Col6n de Jesus, as the head of the conspiracy, was differently situated than his co-conspirators.

LIPEZ, Circuit Judqe.

Copied to clipboard