United States v. Ricardo Epps

United States Court of Appeals FOR THE DISTRICT OF COLUMBIA CIRCUIT Argued October 11, 2012 Decided February 12, 2013 No. 11-3002 UNITED STATES OF AMERICA, APPELLEE v. RICARDO EUGENE EPPS, ALSO KNOWN AS MAN, ALSO KNOWN AS FAT MAN, APPELLANT Appeal from the United States District Court for the District of Columbia (No. 1:99-cr-00175-1) Mary Manning Petras, Assistant Federal Public Defender, argued the cause for appellant. With her on the briefs was A.J. Kramer, Federal Public Defender. Bernard J. Apperson III, Assistant U.S. Attorney, argued the cause for appellee. On the brief were Ronald C. Machen Jr., U.S. Attorney, and Roy W. McLeese III, James S. Sweeney, and Elizabeth H. Danello, Assistant U.S. Attorneys. Sarah Chasson and Elizabeth Trosman, Assistant U.S. Attorneys, entered appearances. Before: ROGERS and BROWN, Circuit Judges, and WILLIAMS, Senior Circuit Judge. — 2 — Opinion for the Court by Circuit Judge ROGERS and Senior Circuit Judge WILLIAMS. Dissenting opinion by Circuit Judge BROWN. ROGERS, Circuit Judge and WILLIAMS, Senior Circuit Judge: In Freeman v. United States, 131 S. Ct. 2685 (2011), the Supreme Court held that the district court is not categorically barred from reducing a defendant’s sentence under 18 U.S.C. § 3582 (c)(2) where the defendant entered into a plea agreement pursuant to Federal Rule of Criminal Procedure 11(c)(1)(C). The decision was splintered, however, with the plurality and concurring opinions adopting different reasoning. Prior to Freeman, the district court denied Ricardo Epps’ § 3582(c)(2) motion for a reduction of his Rule 11(c)(1)(C) sentence. United States v. Epps, 756 F. Supp. 2d 88 (D.D.C. 2010). Epps appeals, contending that there is no controlling opinion in Freeman and that because the district court (as well as the Rule 11(c)(1)(C) agreement) relied upon the crack-cocaine Guidelines range when determining whether to accept the stipulated sentence, his sentence was imposed “based on” the Guidelines range and the district court was authorized under § 3582(c)(2) to reconsider and reduce his sentence in light of the Sentencing Commission’s reduction of the sentencing range applicable to him. For the following reasons, we reverse and remand the case to the district court. I. On October 29, 1999, Epps was sentenced to 188 months’ imprisonment for violation of 21 U.S.C. §§ 841 (a)(1), 841(b)(1)(A)(iii), and 846 and, in view of the quantity of illegal drugs for which he was responsible, to five years’ supervised release, see id. § 841(b)(1)(A)(viii). Epps had entered a Rule — 3 — 11(c)(1)(C) plea agreement.1 The district court accepted Epps’ plea, ordered a presentence investigation, and stated that it would determine whether to accept the stipulated 188 month sentence upon reviewing the presentence report. Tr. Aug. 6, 1999 at 11–12. At the sentencing hearing, upon reviewing the presentence report, the district court recalculated the Guidelines sentencing range applicable to Epps. Tr. Oct. 29, 1999 at 12–13. Rejecting a two-level addition to the base level offense for possession of a firearm, the district court calculated the offense level at 35. Id. With a criminal history category of III, Epps’ Guidelines sentencing range was 210 to 260 months. The prosecutor joined defense counsel in seeking a downward departure from the Guidelines range to 188 months, explaining their agreement to that term was a way to avoid the need to litigate disputes regarding the Guidelines calculations. Id. at 6–8. The district court agreed to depart from the Guidelines range and sentenced Epps to 188 months’ imprisonment. Id. at 14. At the time, the district court expressed concern about the disparity between the Guidelines sentencing range for crack and powder cocaine offenses, noting that Epps’ sentence would have — 1 — Rule 11(c)(1)(C) provides that, if the parties reach agreement on a plea, the plea agreement may specify that an attorney for the government will: agree that a specific sentence or sentencing range is the appropriate disposition of the case, or that a particular provision of the Sentencing Guidelines, or policy statement, or sentencing factor does or does not apply (such a recommendation or request binds the court once the court accepts the plea agreement). FED. R. CRIM. P. 11(c)(1)(C) (emphasis added). — 4 — been “substantially less” if his offense had involved powder rather than crack cocaine. See id. at 11, 14. On October 16, 2008, Epps filed a motion to reduce his sentence pursuant to 18 U.S.C. § 3582 (c)(2)2 in light of the Sentencing Commission’s amendments to the crack cocaine Guidelines in November 2007 and March 2008, see U.S.S.G. (2011), Supp. to App. C, Amend. 706, 711, which the Commission in 2008 made retroactive, id. at Amend. 713, 716. Applying the amendments would reduce Epps’ offense level to 33 and the applicable Guidelines sentencing range to 168 to 210 months. The government opposed the motion on the ground that Epps’ sentence was based on the 188 months stipulated in his Rule 11(c)(1)(C) plea agreement, not on the Guidelines range that was applicable to him. Epps responded that because his sentence and the stipulated range were calculated in relation to, and therefore “based on,” a Guidelines range that was subsequently reduced, § 3582(c)(2) authorized the district court to reduce his sentence. The district court denied Epps’ motion. See Epps, 756 F. Supp. 2d at 89 . Epps appealed on January 4, 2011; on January 11, 2011, this court held his appeal in — 2 — Section 3582(c)(2) provides: [I]n the case of a defendant who has been sentenced to a term of imprisonment based on a sentencing range that has subsequently been lowered by the Sentencing Commission pursuant to 28 U.S.C. 994(o), upon motion of the defendant or the Director of the Bureau of Prisons, or on its own motion, the court may reduce the term of imprisonment, after considering the factors set forth in section 3553(a) to the extent that they are applicable, if such a reduction is consistent with applicable policy statements issued by the Sentencing Commission. 18 U.S.C. § 3582 (c)(2) (emphasis added). — 5 — abeyance pending the Supreme Court’s decision in Freeman, where the stated question was “whether a defendant is ineligible for a sentence reduction under 18 U.S.C. § 3582 (c)(2) solely because the district court accepted a Rule 11(c)(1)(C) plea agreement.” Pet. for Writ of Cert., Freeman, 131 S. Ct. 2685 (No. 09-10245). II. As a threshold matter, the government maintains, on three grounds, that this court lacks jurisdiction now that Epps has completed his period of imprisonment and commenced his five-year term of supervisory release that is mandatory.3 Epps responds that his appeal is not moot because its resolution could affect his term of supervised release in view of 18 U.S.C. § 3583 (e)(1), which provides that a district court “may . . . terminate a term of supervised release and discharge the defendant released at any time after the expiration of one year of supervised release . . . if it is satisfied that such action is warranted by the conduct of the defendant released and the interest of justice.” We conclude, notwithstanding Epps’ release from incarceration while his appeal was pending and the commencement of his term of supervised release, that we have jurisdiction. 1. Because 21 U.S.C. § 841 (b)(1)(A) mandates five years’ supervised release, the government maintains there is no relief this court could order, even if it agrees with the merits of Epps’ motion, because the district court “has no authority to reduce [Epps’] term of supervised release.” Appellee’s Supp. Br. at 7. — 3 — On October 11, 2012, upon determining that Epps had been released by the Bureau of Prisons on May 11, 2012, the court ordered the parties to file supplemental briefs on whether this appeal is moot. — 6 — Before § 841 was amended in 2002, the circuits were split on the relationship between two provisions prescribing the terms of supervised release for overlapping categories of felonies. In 18 U.S.C. § 3583 (b)(2), Congress stated that “[e]xcept as otherwise provided, the authorized terms of supervised release are . . . for a Class C or Class D felony, not more than three years.” But in 21 U.S.C. § 841 (b)(1)(A)(1994), which applied to Epps at the time of his Rule 11(c)(1)(C) plea agreement, Congress described the appropriate term of supervised release as follows: “Any sentence under this subparagraph shall . . . impose a term of supervised release of at least 5 years in addition to such term of imprisonment.” This led courts to take conflicting views on which statutory provision prevailed when both were applicable to a defendant being sentenced. Compare, e.g., United States v. Kelly, 974 F.2d 22, 24 (5th Cir. 1992) (three years is the maximum term), with United States v. Garcia, 112 F.3d 395 , 397–98 (9th Cir. 1997) (term of supervised release can be greater than three years). In 2002, Congress resolved the uncertainty between § 3583's ceiling and § 841's floor in favor of § 841, rendering § 841(b)(1)(A) to read: “Notwithstanding section 3583 of Title 18, any sentence under this subparagraph shall . . . impose a term of supervised release of at least 5 years in addition to such term of imprisonment.” Id. (emphasis added). In United States v. Johnson, 331 F.3d 962 , 967 n.4 (D.C. Cir. 2003), this court noted both the circuit split and that after the 2002 amendment the term of supervised release for a § 841 conviction can exceed three years, see also H. R. CONF. REP. NO. 107-685, at 188–89 (2002). The government’s position that Congress’s clarification of the § 3583/§ 841 conflict removed the district court’s discretion in determining the length of Epps’ term of supervised release assumes (without discussion) that the 2002 amendment to § 841 was retroactive. Yet “the presumption against — 7 — retroactive legislation is deeply rooted in our jurisprudence, and embodies a legal doctrine centuries older than our Republic.” Landgraf v. USI Film Products, 511 U.S. 244, 265 (1994). Indeed, the Court recently reiterated its commitment to the “deeply rooted presumption against retroactive legislation” in Vartelas v. Holder, 132 S. Ct. 1479, 1484 (2012). In Justice Story’s classic formulation, an act is impermissibly retroactive “when such application would . . . ‘attac[h] a new disability, in respect to transactions or considerations already past.’” Id. at 1486–87 (quoting Soc’y for Propagation of Gospel v. Wheeler, 22 F. Cas. 756, 767 (1814)). Although deriving from constitutional principles including the Ex Post Facto Clause, the Contract Clause, and the Due Process Clause of the Fifth Amendment, see id. at 1486, the operative principle is one of interpretation: where the statutory text is ambiguous courts presume that Congress did not intend retroactive application in the sense defined by Justice Story, see Landgraf, 511 U.S. at 280 . The § 3583/§ 841 clash was unresolved in this circuit at the time of the 2002 amendment to § 841. The Supreme Court, however, as it noted in Vartelas, had applied the principle against retroactivity in INS v. St. Cyr, 533 U.S. 289 (2001), to justify a prospective-only interpretation of a statute that replaced a discretionary decision with an automatic negative.4 — 4 — In St. Cyr, a lawful permanent resident had pleaded guilty to a criminal charge that made him deportable. Under the immigration law then in effect, he would have been eligible to apply for a waiver of deportation, but his removal proceeding commenced after Congress withdrew such dispensation. 533 U.S. at 292 . The Supreme Court held that disallowance of discretionary waivers “attaches a new disability, in respect to transactions or considerations already past,” id. at 321 (internal quotation marks omitted), observing that aliens “almost certainly relied upon th[e] likelihood [of receiving discretionary relief] in deciding [to] forgo their right to a trial,” id. at — 8 — 132 S. Ct. at 1491–92. Here too, by resolving the conflict between discretion under § 3583 and its absence under § 841, and uncertainty in this circuit as to which controlled, Congress in 2002 “attac[hed] a new disability in respect to transactions or considerations already past.” The 2002 amendment should accordingly be interpreted as prospective only. In Johnson, 331 F.3d at 967 n.4, this court did not wade into the circuit split to harmonize the apparent conflict between the statutes before 2002, nor has it subsequently done so; we need not do so now. The government’s reliance on United States v. Lafayette, 585 F.3d 435 (D.C. Cir. 2009), for the proposition that this court has already decided that the five-year term of supervised release is mandatory is misplaced. Although Lafayette was sentenced and resentenced under § 841 prior to the 2002 amendment, the court neither addressed the problematic retroactivity questions with respect to the appropriate term of supervised release nor acknowledged the circuit split; the parties’ briefs did not refer to the split. See Brief of Appellant, United States v. Lafayette, 2008 WL 6742228 (focusing on “the inaccurate and unlawful calculation” of defendant’s prison sentence as leading to erroneous term of supervised release); Brief of Appellee, United States v. Lafayette, 2009 WL 2633676 (focusing on distinction between supervised release term and imprisonment term). Here, as in Lafayette, the issue of resolving the pre-2002 amendment circuit split regarding the proper interpretation of § 841 and its relationship to § 3583 has not been briefed, and the question can be determined by the district court when it addresses Epps’ pending § 3583 motion, see Appellant’s Supp. Br. at 2–3, or some successor motion. The point is simply that the 2002 325. Because applying the withdrawal of waiver would have “obvious and severe retroactive effect” and Congress had not made its intention clear, the Court held the amendment applied prospectively only. Id. at 325–26 & n.55. — 9 — amendment’s prospective application means that reduction of Epps’ term of imprisonment would, as he maintains, enhance his prospect for securing a similar reduction in his term of supervised release because the district court’s discretion in determining the length of his supervised release is unaffected by the 2002 amendment. See U.S.S.G. § 1B1.10. 2. Under 18 U.S.C. § 3583 (e)(1), “[t]he district court may . . . terminate a term of supervised release and discharge the defendant released at any time after the expiration one year of supervised release.” The government notes that Epps has yet to complete one year of his supervised release term. Nothing in the text of § 3583(e), however, prevents a defendant from moving for a § 3583(e)(1) termination prior to the completion of his first year. Indeed, the provision is ambiguous as to whether it limits when the district court might act on that motion; it requires only that the effect of any termination begin “after the expiration [of] one year of supervised release.” 18 U.S.C. § 3583 (e)(1). Additionally, 18 U.S.C. § 3583 (e)(2) provides an avenue by which the district court “may modify, reduce, or enlarge the conditions of supervised release, at any time prior to the expiration or termination of the term of supervised release.” Id. (emphasis added). While Epps has moved for a § 3583(e)(1) termination, see Appellant’s Supp. Br. at 2, nothing prevents him from moving, in the alternative, for a reduction pursuant to § 3583(e)(2). The only temporal restriction associated there is that the judicial action must occur, logically, prior to the term’s termination or expiration. Epps is obviously within that term. 3. In our unpublished opinion in United States v. Bundy, 391 F. App’x 886 (D.C. Cir. 2010), the court stated, as the government notes, that the prospect of reduction of the defendant’s term of supervised release “is so speculative that any decision on the merits [of his claim to a reduced prison — 10 — term] would be merely advisory and not in keeping with Article III’s restriction of power to live cases or controversies.” Id. at 887 (citing Burkey v. Marberry, 556 F.3d 142, 149 (3d Cir. 2009)) (internal quotation marks omitted). Bundy is not binding here. “Although our circuit does not have a local rule directly on point, . . . unpublished dispositions should not strictly bind panels of the court . . . [and] do not constrain a panel of the court from reaching a contrary conclusion in published opinion after full consideration of the issue.” In re Grant, 635 F.3d 1227, 1232 (D.C. Cir. 2011). More substantively, that Epps over-served his sentence — as we must assume for jurisdictional arguments, see Am. Nat’l Ins. Co. v. FDIC, 642 F.3d 1137, 1139 (D.C. Cir. 2011) — is of paramount importance to whether he should continue under supervised release for five years. In Bundy the court relied on Burkey for the proposition that challenges such as Epps’ are necessarily moot upon a defendant’s release from prison. Burkey, in turn, purports to apply Supreme Court precedents on the extent to which courts will presume “collateral consequences” in comparable cases. The cases considered in Burkey are not comparable. In Sibron v. New York, 392 U.S. 40 (1968), the (alternative) holding was that a defendant’s challenge to his conviction is not mooted by his release; the prospect that the conviction would have “collateral consequences,” such as impeachment of his character in a later criminal trial, was sufficient. Id. at 54–56. En route to this conclusion the Court summarized Pollard v. United States, 352 U.S. 354 (1957), as a case where “the Court abandoned all inquiry into the actual existence of specific collateral consequences and in effect presumed that they existed.” Sibron, 392 U.S. at 55 . Then, in Lane v. Williams, 455 U.S. 624 (1982), the Court declined to apply Sibron in a case where the defendant challenged imposition of a mandatory parole term (on the ground that he had not been warned of this consequence of pleading guilty) that had expired by the time of the district — 11 — court’s final ruling. Finding no generally applicable legal consequences from the expired parole term, the Court held the case was moot, although it observed that if the defendant had challenged the guilty plea itself, release from the parole would not have mooted the case. Id. at 633 . Similarly, in Spencer v. Kemna, 523 U.S. 1 (1998), where the defendant challenged parole revocation and was released from the renewed incarceration (his term of imprisonment having expired) before the district court ruled, the Court refused to presume collateral consequences from the parole revocation (or, necessarily, from its invalidation), and found the defendant’s attempted showing of an actual likelihood of consequences too weak to support justiciability. See id. at 985–86. Epps’ case does not fit precisely into either the Sibron or the Lane-Spencer paradigm. Unlike in Sibron, Epps is not challenging his conviction; he is merely claiming the benefit of an opportunity to have his sentence retroactively reduced. But, because his five-year term of supervised release is still unserved, and because of the relationship between a prison sentence and supervised release (notwithstanding United States v. Johnson, 529 U.S. 53 (2000)), there seems to be a very substantial likelihood that a ruling that Epps’ incarceration should have been shorter would influence the district court’s readiness to reduce his term of supervised release. Epps’ circumstances thus differ from those examined in Lane and in Spencer. At least two courts of appeals clearly regard the enhanced prospects for a reduced term of supervised release under § 3583 as adequate to hold non-moot a released prisoner’s claim to a lesser period of incarceration: the Second Circuit in Levine v. Apker, 455 F.3d 71 , 76–77 (2d Cir. 2006), held that the prospect renders non-moot a released prisoner’s challenge to a Bureau of Prisons regulation cutting off his change of placement in a half-way house, while the Ninth — 12 — Circuit in Mujahid v. Daniels, 413 F.3d 991 , 993–95 (9th Cir. 2005), viewed the prospect to render non-moot a prisoner’s attack on the Bureau’s “good time” regulations. Similarly, the Fourth Circuit held in Townes v. Jarvis, 577 F.3d 543 , 546-49 & n.3 (4th Cir. 2009), relying on Levine and Mujahid, that a prisoner’s challenge to a finding of parole ineligibility was not moot due to the likelihood that a favorable ruling would yield a reduction in ongoing parole. Against these cases, Burkey’s assertion that “[w]here . . . the appellant is attacking a sentence that has already been served, collateral consequences will not be presumed, but must be proven,” 556 F.3d at 148 , is overbroad and suggests a failure to focus on the distinctive features of the cases on which it relied. This court has not yet weighed in on the subject of whether a defendant’s motion for a sentence reduction under § 3582(c)(2) is rendered moot upon completion of his term of imprisonment (beyond our unpublished treatment in Bundy), but the logic of Levine and Mujahid seems far more persuasive than that of Burkey. Our conclusion that Epps is eligible for a reduced sentence under § 3582(c)(2), if it led to an actual sentence reduction, would necessarily inform the district court’s evaluation of a motion for termination or reduction of his term of supervised release under § 3583(e)(1) or (e)(2). We note the Supreme Court in Johnson identified relief under § 3583(e)(1) or (e)(2) as potential means for addressing the injustice of a prisoner’s being incarcerated beyond the proper expiration of his prison term. 529 U.S. at 60 . Our dissenting colleague argues that we “conflate[] two separate issues: whether Epps should have been allowed to pursue § 3582(c)(2) relief, and whether he actually spent too much time in prison,” and that it is only “the latter that would be relevant to a district court deciding whether to modify the term of Epps’ supervised release.” Dissent Op. at 2. This is a false distinction. Epps’ claim is not the almost metaphysical — 13 — issue of whether he has “spent too much time in prison.” He claims instead that the sentence is excessive in the sense that, but for the matter of timing, he would be legally entitled under § 3582(c)(2) to have the district court consider the application to him of the Sentencing Commission’s decision to retroactively reduce the crack-powder distinction. He is entitled to have that question answered. III. In Freeman v. United States, 131 S. Ct. 2685 (2011), the Supreme Court addressed whether defendants sentenced in accordance with a Rule 11(c)(1)(C) plea agreement may be eligible for a reduction of sentence under § 3582(c)(2). A four Justice plurality held that “the text and purpose of the three relevant sources — the statute [i.e., the Sentencing Reform Act], the Rule, and the governing policy statements — require the conclusion that the district court has authority to entertain § 3582(c)(2) motions when sentences are imposed in light of the Guidelines, even if the defendant enters into an 11(c)(1)(C) agreement.” Id. at 2693 (Kennedy, J., joined by Ginsburg, Breyer, and Kagan, JJ.). “In every case the judge must exercise discretion to impose an appropriate sentence. This discretion, in turn, is framed by the Guidelines.” Id. at 2960. Rejecting the interpretation of the four Justices in dissent that a Rule 11(c)(1)(C) sentence is based only on the plea agreement and not the Guidelines, see id. at 2701 (Roberts, C.J., joined by Scalia, Thomas, and Alito, J.J., dissenting), the plurality explained that concern about upsetting the bargain struck between the defendant and the prosecutor “has nothing to do with whether a sentence is ‘based on’ the Guidelines under § 3582(c)(2).” Id. at 2963. Further, such concern was “overstated” because “[r]etroactive reductions to sentencing ranges are infrequent, so the problem will not arise often,” and “[m]ore important, the district court’s authority under — 14 — § 3582(c)(2) is subject to significant constraints, constraints that can be enforced by appellate review.” Id.5 Notably, the plurality veered from the approach adopted by the concurring opinion of Justice Sotomayor — requiring the plea agreement itself to contemplate sentence reduction — on the ground that “[t]he statute . . . calls for an inquiry into the reasons for a judge’s sentence, not the reasons that motivated or informed the parties,” noting that “[t]he parties cannot by contract upset an otherwise-final sentence.” Id. at — 5 — The plurality explained that “the governing policy statement confirms that the [district] court’s acceptance [of the Rule 11(c)(1)(C) plea agreement] is itself based on the Guidelines.” Freeman, 131 S. Ct. at 2692 . The commentary to § 6B1.2 instructs that the district court may accept the plea “only if the court is satisfied either that such sentence is an appropriate sentence within the applicable guideline range or, if not, that the sentence departs from the applicable guideline range for justifiable reasons.” Id. (quoting U.S.S.G. § 6B1.2 commentary). Allowing the district court under § 3582(c)(2) “to revisit a prior sentence to whatever extent the sentencing range in question was a relevant part of the analytic framework the judge used to determine the sentence or to approve the agreement,” the plurality concluded, “is the only rule consistent with the government policy statement . . . that rests on the premise that a Guideline range may be one of the many factors that determine the sentence imposed.” Id. at 2692-93. The plurality also found support for its approach in the policy statement that applies to § 3582(c)(2) motions, which placed “considerable limits on district court discretion,” id. at 2693: U.S.S.G. § 1B1.10 instructs that the district court may “modify[] a sentence to substitute only the retroactive amendment,” Freeman, 131 S. Ct. at 2692 (quoting policy statement), thus “isolating whatever marginal effect the since-rejected Guideline had on the defendant’s sentence,” id. Also, under § 1B1.10(b)(2), below-Guidelines modifications are forbidden in § 3582(c)(2) proceedings, except where the original sentence was a downward departure. Id. at 2693; see also id. (citing § 1B1.10, comment n). — 15 — 2694. Discussing the concurrence’s approach, the plurality emphasized that “the consequences of this erroneous rule would be significant. By allowing modification only when the terms of the agreement contemplate it, the proposed rule would permit the very disparities the Sentencing Reform Act seeks to eliminate.”6 Id. The plurality thus construed § 3582(c)(2) to contribute to reducing sentencing disparities “by ensuring that district courts may adjust sentences imposed pursuant to a range that the Commission concludes are too severe, out of step with the seriousness of the crime and the sentencing ranges of analogous offenses, and inconsistent with the Act’s purposes.” Id. It cited the crack-cocaine range as “a prime example of an unwarranted disparity that § 3582(c)(2) is designed to cure.” Id. In sum, the plurality held that “when a defendant enters into an 11(c)(1)(C) agreement, the judge’s decision to accept the plea and impose the recommended sentence is likely to be based on the Guidelines; and when it is, the defendant should be eligible to seek § 3582(c)(2) relief.” Id. at 2695. “Even where the judge varies from the recommended range . . . if the judge uses the sentencing range as the beginning point to explain the decision to deviate from it, then the Guidelines are in a real sense a basis for the sentence.”7 Id. at 2692. — 6 — The dissenting justices, like the plurality, concluded that examination of the parties’ intent is largely irrelevant to the inquiry of whether a Rule 11(c)(1)(C) agreement was “based on” the Guidelines and “agree[d] with the plurality that the approach of the concurrence to determining when a Rule 11(c)(1)(C) sentence may be reduced is arbitrary and unworkable.” Freeman, 131 S. Ct. at 2700–01 (Roberts, C. J., joined by Scalia, Thomas, Alito, JJ., dissenting). — 7 — Some circuit courts of appeals have concluded, because the district court is required to use the Sentencing Guidelines as a baseline for evaluating plea agreements, see Gall v. United States, 552 U.S. 38 , 50–51 (2007); Rita v. United States, 551 U.S. 338, 348 (2007); — 16 — In contrast, the concurring opinion adopted a different approach: [I]f a [Rule 11(c)(1)(C)] agreement expressly uses a Guidelines sentencing range applicable to the charged offense to establish the term of imprisonment, and that range is subsequently lowered . . . the term of imprisonment is “based on” the range employed and the defendant is eligible for sentence reduction under § 3582(c)(2). Freeman, 131 S. Ct. at 2695 (Sotomayor, J., concurring). This is so because a term of imprisonment imposed pursuant to a Rule 11(c)(1)(C) agreement is “‘based on’ the agreement itself, not on the judge’s calculations of the Sentencing Guidelines.” Id. Thus, the district court has jurisdiction to consider a sentence reduction pursuant to § 3582(c)(2) only if the Rule 11(c)(1)(C) plea agreement either (1) explicitly “call[s] for the defendant to be sentenced within a particular Guidelines range” or (2) “make[s] clear that the basis for the specified term is a Guidelines sentencing range applicable to the offense to which the defendant pleaded guilty” and “that sentencing range is evident from the agreement itself.” Id. at 2697. A. Due to the fragmented nature of the Supreme Court’s holding in Freeman, it is not immediately obvious whether the Court set a standard for evaluating whether sentences pursuant U.S.S.G. § 6B1.2, that the logic of the plurality’s approach is “that a district court can always grant § 3582(c)(2) relief to a defendant who enters a Rule 11(c)(1)(C) plea agreement.” United States v. Brown, 653 F.3d 337, 339 (4th Cir. 2011); see also United States v. RiveraMartinez, 665 F.3d 344, 347 (1st Cir. 2011). — 17 — to Rule 11(c)(1)(C) agreements are eligible for § 3582(c)(2) reductions that is controlling on lower courts, or, if it did, what the precise contours of that standard are. In Marks v. United States, 430 U.S. 188 (1977), the Supreme Court instructed: “When a fragmented Court decides a case and no single rationale explaining the result enjoys the assent of five Justices, the holding of the Court may be viewed as that position taken by those Members who concurred in the judgments on the narrowest grounds.” Id. at 193 (internal quotation omitted). The Court has subsequently recognized that this seeminglysimple rule is “more easily stated than applied,” noting that “the Marks inquiry . . . has so obviously baffled and divided the lower courts that have considered it” that it has created a “degree of confusion” such that it is not always “useful to pursue . . . to the utmost logical possibility.” Nichols v. United States, 511 U.S. 738 , 745–46 (1994); see also Grutter v. Bollinger, 539 U.S. 306, 325 (2003). Parsing Freeman in light of Marks is required to distill what impact that precedent has in deciding Epps’ appeal. This court has interpreted Marks to mean that the narrowest opinion “must represent a common denominator of the Court’s reasoning; it must embody a position implicitly approved by at least five Justices who support the judgment.” King v. Palmer, 950 F.2d 771, 781 (D.C. Cir. 1991) (en banc) (emphasis added). Stated differently, Marks applies when, for example, “the concurrence posits a narrow test to which the plurality must necessarily agree as a logical consequence of its own, broader position.” Id. at 782 (emphasis added). In King, the en banc court addressed the question of when counsel should receive a contingency enhancement under a fee-shifting statute in addition to the “lodestar” award computed by multiplying hours worked by a reasonable hourly fee. Id. at 775 . The court thus had to assess whether there was sufficient common ground between the plurality opinion and the — 18 — concurring opinion in Pennsylvania v. Delaware Valley Citizen’s Council for Clean Air, 483 U.S. 711 (1987) (“Delaware Valley II”), to control its decision. See King, 950 F.2d at 775 . In the splintered decision in Delaware Valley II, the plurality concluded that contingency enhancements were appropriate only in “exceptional cases.” 483 U.S. at 730 (White, J., joined by Rehnquist, C.J., Powell and Scalia, J.J.). The concurring opinion, in contrast, concluded that “Congress did not intend to foreclose consideration of contingency in setting a reasonable fee,” agreeing with the dissent’s analysis on this point and with respect to the view that “compensation for contingency must be based on the difference in market treatment of contingent fee cases as a class, rather than on an assessment of the ‘riskiness’ of any particular case.” Id. at 731 (O’Connor, J., concurring in part and concurring in the judgment)(emphasis in original). Nevertheless, the concurring opinion agreed with the plurality that “no enhancement for risk is appropriate unless the applicant can establish that without an adjustment for risk the prevailing party ‘would have faced substantial difficulties in finding counsel in the local or other relevant market,’” id. at 733 (quoting plurality opinion at 731), and concluded there was insufficient record evidence to justify an enhancement, id. at 734 . Analyzing the precedential force of Delaware Valley II, the en banc court in King overruled a prior holding (and portions of prior cases) that had treated the concurring opinion as controlling under Marks. See King, 950 F.2d at 785 . The en banc court noted that this circuit had not previously “focused on the fact that there are two analytically distinct questions involved in awarding a contingency enhancement” if its prior view of Delaware Valley II were perpetuated. Id. at 777 . Reexamining Delaware Valley II, the en banc court concluded: When . . . one opinion supporting the judgment does not fit entirely within a broader circle — 19 — drawn by the others, Marks is problematic. If applied in situations where the various opinions supporting the judgment are mutually exclusive, Marks will turn a single opinion that lacks majority support into national law. When eight of nine Justices do not subscribe to a given approach to a legal question, it surely cannot be proper to endow that approach with controlling force, no matter how persuasive it may be. The Court itself does not appear to apply Marks in cases of this type. Id. at 782 (referencing Coolidge v. New Hampshire, 403 U.S. 443 (1971), as an example). In sum, “all the analytically necessary portions of a Supreme Court opinion” must overlap in rationale in order for a controlling opinion to be discerned pursuant to Marks; if no such common rationale exists the Supreme Court precedent is to be read only for its persuasive force. Id. at 784. On the merits of whether an enhancement was available, the en banc court concluded there was “no practical middle ground between providing enhancements routinely and not providing them at all,” and adopted a view similar to that of the plurality in Delaware Valley II that contingency enhancements were unavailable in this circuit, noting that “a majority of the Supreme Court clearly agrees that the question of attorney’s fees must not turn into major litigation in itself.” Id. This court has not heretofore applied the Marks standard articulated by the en banc court in King to the splintered decision in Freeman. Most recently, the court applied the concurring opinion of Justice Sotomayor in a similar case, where a defendant, who had entered a Rule 11(c)(1)(C) plea agreement, had filed a motion to reduce his sentence under § 3582(c)(2), and where the parties agreed that the concurring opinion in Freeman controlled. In United States v. Duvall, — 20 — 2013 WL 276016 (D.C. Cir., Jan. 25, 2013), the court affirmed the district court’s decision that the defendant was ineligible for § 3582(c)(2) relief, stating: “For purposes of this appeal, both parties agree that Justice Sotomayer’s opinion controls our analysis in light of the Supreme Court’s decision in Marks v. United States, 430 U.S. 188 , 193 (1977). Accordingly, we do not further address that question.” Id. at *4. The concurring opinion discussed the en banc opinion in King, but concluded relief was unavailable in light of United States v. Berry, 618 F.3d 13 (D.C. Cir. 2010). See Duvall, 2013 WL 276016 at *611 (Williams, J., concurring in the judgment) (suggesting Berry warrants en banc reconsideration). In other words, in Duvall, because the parties agreed that Justice Sotomayor’s concurring opinion controlled, the court had no occasion to reject alternative arguments or to hold that the concurring opinion in Freeman was controlling precedent. Absent such a holding there is no law of the circuit to apply on that question. See LaShawn A. v. Barry, 87 F.3d 1389 (D.C. Cir. 1996) (en banc). Under Marks then, we conclude that there is no controlling opinion in Freeman because the plurality and concurring opinions do not share common reasoning whereby one analysis is a “logical subset,” King, 950 F.2d at 781 , of the other. The plurality opinion rejects the concurring opinion’s approach, stating its rationale is fundamentally incorrect because § 3582(c)(2) “calls for an inquiry into the reasons for a judge’s sentence, not the reasons that motivated or informed the parties.” Freeman, 131 S. Ct. at 2694 (plurality opinion). Although suggesting the approach in the concurring opinion is an “intermediate position,” id., the plurality understands the reasoning of the concurrence — that courts should examine the intent of the parties to a Rule 11(c)(1)(C) agreement to determine whether a sentence pursuant to such a plea is “based on” the Guidelines — to be incompatible with its own because the concurring opinion offers an analytically distinct rationale to justify its approach. Indeed, eight of the nine Justices — 21 — rejected the framework of the concurring opinion, with the dissent correctly predicting that it would “foster[] confusion in a area in need of clarity,” 131 S. Ct. at 2704 (Roberts, C.J., joined by Scalia, Thomas, and Alito, J.J., dissenting), because it initially examined what the district court judge did but then “suddenly” shifted its focus to the parties’ intent, id. at 2702 . Thus the dissent “agree[d] with the plurality that the approach of the concurrence . . . is arbitrary and unworkable.” Id. at 2700–01. We have previously noted that other courts of appeals have held that the concurring opinion in Freeman, as the narrower interpretation of “based on,” is the holding of the Court. See Duvall, 2013 WL 276016 ; supra n.1. But these courts, like the government here, see Appellee’s Br. at 21, appear not to have considered circumstances where § 3582(c)(2) relief would be available under the concurring opinion but not the plurality opinion. Epps, referencing acknowledgment of this possibility by the plurality in Freeman, 131 S. Ct. at 2694 , offers examples where the concurring opinion is not a subset of the plurality opinion. See Reply Br. at 8-108; see also United States v. Duvall, 2013 WL 276016 at — 8 — As examples Epps suggests: For example, the parties may state in the plea agreement that a particular range applies and agree to a sentence at the bottom of that range, but the district court may not agree that the range determined by the parties applies, finding for example that the career offender range is applicable instead, but notwithstanding this finding accept the plea because it is to a term that is acceptable to the court for reasons unrelated to the guideline range determined by the parties. Using Justice Sotomayor’s standard, if the sentencing range used by the parties is subsequently reduced, the defendant would be eligible for a sentence reduction because the plea agreement — 22 — *8-9 (Williams, J., concurring in the judgment) (offering illustrative example where the concurring opinion is not a subset of the plurality opinion). In other words, the set of cases where the defendant prevails under the concurrence is not always nestled within the set of cases where the defendant prevails under the plurality as the Marks framework requires, whether articulated as in King or in terms of “one opinion . . . always lead[ing] to the same result that a broader opinion would reach,” Jackson v. Danberg, 594 F.3d 210, 222 (3d Cir. 2010). was accepted and provided for a stipulated sentence based on a subsequently reduced range – according to Justice Sotomayor, eligibility is determined based on the agreement. The plurality, however, would find this defendant ineligible because the range that the parties agreed to played no role in the court’s determination that this was an appropriate sentence, despite the fact that the court imposed the agreedupon term of imprisonment. Reply Br. at 8-9. Another example Epps suggests is where: The sentencing court . . . might consider and reject the guideline range used by the parties, not because the court finds that a different guidelines range (such as the career offender range) applies, but because, having considered the applicable guidelines range, the court rejects it as a matter of policy and selects its sentence without regard to it. If under these circumstances the court decides for reasons unrelated to the guidelines range to impose the setence the parties agreed upon, under the plurality’s analysis, the defendant would not be eligible even if the guideline range is later reduced. Under Justice Sotomayor’s analysis, however, the defendant would be eligible. Id. at 9-10. — 23 — In sum, while five Justices in Freeman agreed that the district court is not categorically barred from reducing a sentence under § 3582(c)(2) in Rule 11(c)(1)(C) cases, the concurring opinion is not controlling in this circuit and the question for the court, upon independent analysis of the statute, is when a sentence is “based on a sentencing range that has subsequently been lowered by the Sentencing Commission.” We consider, then, which, if any, of the rationales in Freeman is persuasive. Cf. King, 950 F.2d at 784 . In so doing, we are bound only by the result in Freeman, namely that § 3582(c)(2) relief is not invariably barred when a sentence was imposed pursuant to a Rule 11(c)(1)(C) plea agreement. See id. (referencing Nat’l Mut. Ins. Co. v. Tidewater Transfer Co., 337 U.S. 582, 655 (1949) (Frankfurter, J., dissenting)). B. Whether Epps is eligible for a § 3582(c)(2) reduction of his sentence is a question of law, which this court reviews de novo. See United States v. Cook, 594 F.3d 883, 886 (D.C. Cir. 2010). We conclude that the Freeman plurality’s interpretation, looking to the Sentencing Reform Act, Rule 11(c)(1)(C), and applicable Guidelines policies, is more persuasive than that of the concurring opinion. The Sentencing Reform Act, which seeks through Guidelines to frame the district court sentencing discretion, requires, even under the advisory Guidelines sentencing scheme, that the district court begin by calculating the defendant’s sentence under the Guidelines. See Gall, 552 U.S. at 50–51. In light of the Act, Rule 11(c)(1)(C)’s requirement for court approval, and the instructions in applicable Guidelines policy statements, the plurality in Freeman interpreted § 3582(c)(2) to mean that the focus, even when there is a Rule 11(c)(1)(C) plea agreement, ought to be on the reasons given by the district court for accepting the sentence that it ultimately imposed, not on the parties’ agreement. Freeman, 131 S. Ct at 2694 (plurality opinion). A contrary focus on the parties’ intentions would contribute to the — 24 — unwarranted disparity that the Act was designed to reduce. Id. The plurality thus reasonably viewed § 3582(c)(2) as a mechanism for helping to reduce unwarranted sentencing disparities, such as the crack-cocaine range. Id. at 2965; see also Duvall, supra, at *9 (Williams, J., concurring) (suggesting Congress’s concern was with the phrase — “a sentencing range that has subsequently been lowered by the Sentencing Commission” — in order “to make clear that the district court’s sentence reduction authority was not to be all-purpose, but linked to a range that an amendment had ‘lowered’”). Epps’ Rule 11(c)(1)(C) plea agreement repeatedly refers to the Sentencing Guidelines as the basis for determining Epps’ sentence. Paragraph 8, for example, states that “the sentence in this case will be imposed in accordance with” the Guidelines. See Plea Agreement ¶ 8 (emphasis added). Although prior to Freeman, this court held that the phrase “accordance with” is ambiguous and insufficient alone to anchor the inference that a sentence was determined by an otherwise-unspecified Guidelines range, see Cook, 594 F.3d at 888 , there is further evidence in Epps’ case that the Guidelines formed the basis of his sentence. For example, Paragraph 12 recites the parties’ agreement, applying the Guidelines, to depart downward from the base level of the offense pursuant to Epps’ acceptance of responsibility, id. ¶ 12; see U.S.S.G. § 3E1.1, and the district court, prior to imposing Epps’ sentence, recalculated his Guidelines range before granting a downward departure to the 188 months stipulated in the plea agreement. See Tr. Oct. 29, 1999, at 12–14. Additionally, in evaluating Epps’ plea agreement, the district court stated that it considered the sentence imposed “sufficient” “in view of the fact that the crack cocaine guidelines are what they are.” Id. at 14 (emphasis added). Thus, in departing downward from the Guidelines sentencing range and explaining the basis for its departure, the district court anchored the inference that Epps’ sentence, unlike — 25 — the sentence in Cook, was determined “based on” a specific Guidelines range. The district court’s view that its findings were not determinative of whether Epps’ sentence was “based on” the Guidelines, see Epps, 756 F. Supp. 2d at 92–93, is persuasively refuted by the Freeman plurality. To the extent Sentencing Guidelines § 1B1.10 Application Note 5 advises that overserving a sentence is alone insufficient to warrant early termination of supervised release, we recognize that the district court is to consider many factors in ruling on a § 3582(c)(2) motion and note only that over-serving a sentence is a strong factor that, for the reasons explained above, is not necessarily “too speculative” to have a substantial influence on remand. Under the circumstances, upon “full consideration of the issue,” Grant, 635 F.3d at 1232 , nothing in this court’s precedent requires adoption of a different approach. But cf. Berry, 618 F.3d at 46–47 (avoiding question of “when, if ever, a defendant who enters a Rule 11(c)(1)(C) plea agreement is sentenced ‘based on a sentencing range’”); United States v. Heard, 359 F.3d 544, 548 (D.C. Cir. 2004) (dictum). Nor is there evidence that statutory considerations trumped the applicable Guidelines range in forming the basis of Epps’ sentence. Compare Cook, 594 F.3d at 883 (defendant ineligible for § 3582(c)(2) reduction because sentence based on statutory mandatory minimum); United States v. Tepper, 616 F.3d 583 (D.C. Cir. 2010) (same, because sentence based on career offender status); Berry, 618 F.3d at 13 (same). To recap: The court has jurisdiction of Epps’ appeal notwithstanding his release from incarceration and the commencement of his term of supervised release. His appeal is not moot because applying the amended version of the supervisory release provision would be impermissibly retroactive and, in not applying this amended provision, it becomes likely that Epps’ term of supervisory release may be — 26 — impacted by the outcome of this appeal; it remains for the district court to address the pre-amendment inter-circuit conflict as to which of two provisions on supervisory release applies to Epps in considering his pending motion to reduce his supervisory term. In the absence of necessary overlap between the reasoning of the plurality and concurring opinions in the Supreme Court’s decision in Freeman to discern a narrower opinion that constitutes binding precedent, see King, 950 F.2d at 771 (interpreting Marks), Epps qualifies for relief under § 3582(c)(2). Accordingly, we reverse and remand the case to the district court. BROWN, Circuit Judge, dissenting: Having been released from prison, Epps can neither make good on the court’s conclusion that he was eligible for a reduction in his sentence under 18 U.S.C. § 3582 (c)(2) nor credit any excess prison time served against his term of supervised release, see United States v. Johnson, 529 U.S. 53 , 56–58 (2000). Realizing this, the court attempts to evade the mootness doctrine by invoking the collateral effects its holding might have in future discretionary proceedings to reduce Epps’s terms of supervised release. Because today’s decision fails to offer a “more-than-speculative chance” of affecting Epps’s rights in the future, I would dismiss his appeal as moot. Transwestern Pipeline Co. v. FERC, 897 F.2d 570 , 575 (D.C. Cir. 1990). A district court’s discretion to decide whether to terminate a defendant’s supervised release period is broad: if, having considered the sentencing factors listed in 18 U.S.C. § 3553 (a), the court “is satisfied that” termination of supervised release “is warranted by the conduct of the defendant released and the interest of justice,” then the district court may elect to do so. 18 U.S.C. § 3583 (e)(1). But as we previously concluded, “the inability to obtain a reduced sentence on account of the completion of a prison term, though potentially relevant, is only one of many factors guiding the district court’s exercise of its discretion under § 3583(e)(1),” rendering a decision here merely advisory. United States v. Bundy, 391 F. App’x 886, 887 (D.C. Cir. 2010) (per curiam) (internal citation omitted). Bundy may not bind us, but that does not mean its reasoning is faulty. The collateral consequences of the court’s decision inhabit the realm of the hypothetical: just as a decision in Epps’s favor does not guarantee him relief under § 3583(e), neither would a decision reaching the opposite conclusion have foreclosed it. Nor, for that matter, does the court raise any argument Epps cannot make for himself. What we have, then, is an opinion that neither forces a district — 2 — court to grant Epps a reduction in his supervised release period nor offers him arguments that would otherwise be unavailable to him. At most, the opinion lends Epps’s case the clout of a judicial imprimatur—something a law review article or op-ed by a well-respected jurist might similarly accomplish. The court’s attempts to sidestep mootness rely not on any precedential effect the opinion might have, but on the mere fact that it endorses Epps’s analytical position. Cf. Telecomms. Research & Action Ctr. v. FCC, 917 F.2d 585 , 588 (D.C. Cir. 1990) (holding, in the administrative law context, that a party may not predicate its Article III standing on the content of an agency’s legal reasoning). By assuming its decision “would necessarily inform the district court’s evaluation of a motion for termination or reduction of his term of supervised release,” Maj. Op. 12, the court conflates two separate issues: whether Epps should have been allowed to pursue § 3582(c)(2) relief, and whether he actually spent too much time in prison. The court decides the former, but it is the latter that would be relevant to a district court deciding whether to modify the term of Epps’s supervised release. See Johnson, 529 U.S. at 60 (“There can be no doubt that equitable considerations of great weight exist when an individual is incarcerated beyond the proper expiration of his prison term.”). And if Epps served an excessive prison term, that is because his original sentence was imposed when the sentencing guidelines for crack cocaine were harsher than the Sentencing Commission now deems appropriate; it does not turn on the availability of § 3582(c)(2) relief. True, had a district court reduced Epps’s sentence, it would have remedied—or at least mitigated—his injury, thereby obviating the basis for seeking subsequent modification of his supervised release term. But now that Epps has completed his prison term, whether he should have had access to this procedural avenue is irrelevant to the — 3 — § 3583(e) inquiry. That Epps had no previous opportunity to reduce his sentence may have deprived him of a chance to correct a punishment that was potentially excessive in light of the amended sentencing guidelines. It is not, however, the reason the punishment was or was not excessive. What is instead important is the actual term of imprisonment served. For the purpose of § 3583(e), it matters not that Epps was erroneously denied recourse to a sentence reduction. Even if Congress had never enacted § 3582(c)(2), Epps could still invoke the sentencing guidelines’ later amendment as a reason to reduce his term of supervised release. If the sentence he served was too long, it was too long. Today’s decision in no way changes that fact; any collateral effect its holding may have is illusory. Because Epps’s release from prison renders ineffectual any relief this Court might provide, his case is moot. See Church of Scientology of Cal. v. United States, 506 U.S. 9, 12 (1992). I would therefore dismiss the appeal.