PER CURIAM.

Appellee Joshua Acoff pled guilty to possessing five or more grams of cocaine base with intent to distribute, in violation of 21 U.S.C. § 841. Although the district court accepted Acoff's plea of guilty to that offense, it declined to sentence him pursuant to Section 841(b)(1)(B), the penalty provision that covers the conduct charged in the indictment and admitted to by Acoff. In lieu of the sixty-month sentence mandated by the statute, the district court sentenced Acoff to fifteen months in prison, over the government's objection. The district court justified its decision by observing that the 100-to-1 ratio between crack and powder cocaine sentences established by the statute then in force "does not make sense at all." The government appealed.

The district court manifestly erred in sentencing Acoff to a term below the statutory minimum. As the Supreme Court has explained, "the scope of judicial discretion with respect to a sentence is subject to congressional control." Mistretta v. United States, 488 U.S. 361, 364 (1989). Accordingly, except in circumstances not applicable here,1 district courts lack the authority to impose a sentence below the statutory minimum. See Kimbrough v. United States, 552 U.S. 85, 108 (2007).

Acoff contends that the mandatory minimum sentence no longer applies to him in light of intervening congressional legislation that reduced sentences for certain crack cocaine offenses. See Fair Sentencing Act of 2010, Pub. L. No. 111-220, § 2, 124 Stat. 2372 (amending 21 U.S.C. § 841) ("FSA"). This argument is unavailing. Under the general savings statute, 1 U.S.C. § 109, "[t]he repeal of any statute shall not have the effect to release or extinguish any penalty... incurred under such statute, unless the repealing Act shall so expressly provide, and such statute shall be treated as still remaining in force for the purpose of sustaining any proper action or prosecution for the enforcement of such penalty." Although Acoff argues that the savings statute does not foreclose retroactive application of the FSA, we have recently held otherwise. See United States v. Diaz, 627 F.3d 930 (2d Cir. 2010).

The fact that Acoff, unlike the defendant in Diaz, had not yet exhausted his appeals when the FSA came into force does not change our analysis. Relying on Griffith v. Kentucky, 479 U.S. 314 (1987), Acoff argues that principles of equal protection require us to read the FSA as applying not only to future offenders, but also to those who violated the statute before it was amended but whose sentences were not yet final when the FSA was enacted. That is not correct. The constitutional concern that occupied the court in Griffith was "the actual inequity that results when the Court chooses which of many similarly situated defendants should be the chance beneficiary of a new rule." Id. at 323 (internal quotation marks and emphasis omitted). The Court's holding, which required lower courts to apply new constitutional rules of criminal procedure to all cases not yet final, was intended to account for the injustice that would result if the Court were to grant certiorari and reverse one defendant's conviction, while otherwise applying the new rule only prospectively. There is no suggestion in Griffith that similar constitutional concerns would apply to a new rule announced by Congress. To the contrary, the Court found it necessary to adopt the rule that it did precisely because "[u]nlike a legislature, we do not promulgate new rules of constitutional criminal procedure on a broad basis." Id. at 322. It is not irrational for Congress to impose a penalty on those who committed their offenses at a time when they knew or should have known the severity of the applicable penalty, even while reducing the penalty as to future offenders. Accordingly, "because the FSA took effect... after [the defendant] committed his crimes 1 U.S.C. § 109 bars the Act from affecting his punishment." Diaz, 627 F.3d at 391, quoting United States v. Gomes, 621 F.3d 1343, 1346 (11th Cir. 2010) (omission in the original).

Acoff next contends that the mandatory sentencing scheme in former 21 U.S.C. § 841(b) violates the Equal Protection Clause of the Fourteenth Amendment, because there is no rational basis for the disparity between sentences for crack and powder cocaine. We have repeatedly rejected this argument. See United States v. Regalado, 518 F.3d 143, 149 n. 3 (2d Cir. 2008) (per curiam); United States v. Then, 56 F.3d 464, 466 (2d Cir. 1995); United States v. Moore, 54 F.3d 92, 97-99 (2d Cir. 1995); United States v. Stevens, 19 F.3d 93, 96-97 (2d Cir. 1994). Nothing in the text or legislative history of the Fair Sentencing Act undermines the validity of these prior decisions. As we have noted in another context, "[a] congressional decision that a statute is unfair, outdated, and in need of improvement does not mean that the statute when enacted was wholly irrational or, for purposes of rational basis review, unconstitutional." Smart v. Ashcroft, 401 F.3d 119, 123 (2d Cir. 2005).

We have considered Acoff's remaining arguments and find them to be without merit. Because we reject Acoff's arguments on the merits, we need not address the government's argument that Acoff waived his right to defend the sentence on appeal in his plea agreement.

CONCLUSION

For the foregoing reasons, the judgment of the district court is VACATED and the case is REMANDED for resentencing consistent with this decision.

GUIDO CALABRESI, Circuit Judge, concurring:

I believe the per curiam opinion accurately describes U.S. law and that of our circuit, so I join it fully. That is, its treatment of the effect of a change in law on previously imposed sentences in the absence of a clear statement of retroactivity is the normal—and normally appropriate—procedure and leads to the result in this case. And, yet, there is something troubling about this result with regard to a statute whose grossly different treatment of chemically identical drugs—the rock and powder forms of cocaine—has been criticized and questioned, particularly on grounds of racial injustice. E.g., U.S. Sentencing Comm'n, Special Report to Congress: Cocaine and Federal Sentencing Policy 192 (1995) (concluding that "the vast majority of those persons most affected by such an exaggerated ratio are racial minorities," which creates a perception—if not a reality—of injustice). As we have learned more about the drugs' similarities in terms of effect and addictiveness and about the racially disparate impact the statute's mandatory minimum provisions have had, these criticisms have intensified. See, e.g., ACLU, With the Stroke of a Pen, a Fairer Criminal Justice System, Aug. 3, 2010 (calling the crack-powder disparity "one of the most dysfunctional and needlessly cruel aspects of the federal criminal justice system").

European courts have developed a way of dealing with statutes that though valid when enacted, come, over time, to raise significant constitutional questions. These courts have engaged in a dialogue with their legislatures, explaining that though the courts were not prepared—or possibly even able—to say that the statute was unconstitutional, nevertheless, it was the court's role to inform the legislature that the statute was "heading towards unconstitutionality." Not surprisingly, in such situations, it sometimes happens that the legislature responds to cure the rising defects in the statute. See United States v. Then, 56 F.3d 464, 468-69 (2d Cir. 1995) (Calabresi, J., concurring) (explaining this process).

One could describe the crack-powder disparity in the same way. The dialogue here has involved not just courts but the Sentencing Commission, the academy, and the press. See, e.g., U.S. Sentencing Comm'n, Special Report to Congress: Cocaine and Federal Sentencing Policy (1995, 1997, 2007) (recommending reducing or eliminating the crack-powder disparity); Steven L. Chanenson, Booker on Crack: Sentencing's Latest Gordian Knot, 15 Cornell J. L. & Pub. Pol'y 551, 583-86 (2006) (describing the difficulties and disparities in the crack sentencing Guidelines and urging Congress to act); Editorial, Bad Science and Bad Policy, N.Y. Times, at A30 (Mar. 2, 2010).

To the extent that one could have viewed what occurred in Congress as a response to a suggestion by courts that the sentencing statutes were heading towards unconstitutionality, one might question whether the traditional presumption against retroactivity should apply. In circumstances where the legislature has responded to a judicial suggestion of unconstitutionality, the appropriate starting point might well be the opposite: to assume that the change reaches back—at the very least to cover cases pending on appeal at the time of enactment (and perhaps further)—in the absence of a specific statement that some other metric should be used. The import of this shift in presumption would be to force Congress to focus specifically on the impact of a legislative change resolving a potential constitutional problem, a focus that is not necessary in the run-of-the-mill situation where no countervailing constitutional-level values suggest that a statute's official "effective date" and its practical application date should be different. If the statute's validity was becoming dubious, why should we assume that the legislature wished the statute's constitutional dubiousness to apply in any case?

Nonetheless, U.S. courts have failed to adopt this European dialogic approach. Indeed, with regard to the very question of the crack-powder disparity, it has been rejected by our circuit. Then, 56 F.3d at 466 (majority opinion). In Then, the majority expressly declined to adopt this technique.1 I believe they were wrong then and that following this approach would be desirable now. But, I am bound by that decision and cannot argue that this case should be treated differently from the standard presumption.

GERARD E. LYNCH, Circuit Judge, concurring:

As our per curiam opinion demonstrates, through the combination of the savings statute, 1 U.S.C. § 109, and the provisions of the Fair Sentencing Act of 2010, Pub. L. No. 111-220, § 2, 124 Stat. 2372 (amending 21 U.S.C. § 841) ("FSA"), which fails to provide for any retroactive application of its reform of the sentences for crack cocaine offenses, Congress has commanded that offenders who committed such offenses before the effective date of the FSA but are to be sentenced after that date must be sentenced under the harsh terms of the prior law, now recognized by virtually everyone, including Congress, to have imposed unnecessarily and unfairly severe mandatory sentences. As Judge Calabresi acknowledges, this result is consistent with the Constitution, and — short of adopting a novel theory of constitutional adjudication — courts have no power to alter it.

There is a reasonable argument that Congress's recognition that the prior law was unfair should have led to complete retroactivity, that is, to revisiting the sentences of the thousands of prisoners serving long terms of imprisonment pursuant to the now-abandoned policy. At the same time, it is understandable why Congress would have been reluctant to make its salutary reform fully retroactive. Doing so would not only have imposed a huge burden on the criminal justice system; prosecutorial and judicial decisions in many of those cases — including decisions to drop other, valid charges in return for a plea to a crack offense in the belief that the severe narcotics sentences would adequately protect society — may have been made in reliance on the existence of the prior law. Congress may well have decided that it is simply too difficult to rewind these cases to the beginning, unscramble all of the decisions that had been made, and reprosecute the cases as they might have played out had the provisions of the FSA been in effect all along.

It is more difficult, however, to understand why Congress would want to continue to require that courts impose unfair and unreasonable sentences on those offenders whose cases are still pending. Such defendants still need to be sentenced, and there are few persuasive reasons why they should be sentenced pursuant to an unjust law when Congress has already replaced it with a more just one. It seems likely that simple congressional inattention produced this result: understandably focused on the much larger question of full retroactivity, when Congress decided against making the provisions of the FSA fully retroactive, it may simply have overlooked the distinguishable, and much smaller, category of past offenders who are still being sentenced for pre-FSA crimes.

This is simply a transitional problem. The class of affected past offenders who are still subject to mandatory sentences calculated pursuant to the old and unjust 100-to-1 ratio is presumably small. But it is no comfort to those, like the defendant in this case, who are sentenced unduly harshly under a now-discredited and repealed law, to know that a relatively small number of offenders share their predicament.

It may now be too late for Congress to affect many of these cases. The number of pre-FSA crack offenders whose sentences are not yet final shrinks with every day that Congress fails to act. But at least some such offenders will remain in the system for years to come. With the enactment of the FSA, Congress has finally remedied a glaring injustice in our narcotics laws. Perhaps some day it will revisit the fates of all those who are serving excessive sentences under those now-repealed laws. But even if the arguments against full retroactivity continue to prevail, Congress would do well to quickly and seriously consider whether it has inadvertently required courts to continue an unjust practice in a small but significant number of on-going cases.