247 F.3d 139 (4th Cir. 2001)
UNITED STATES OF AMERICA, Plaintiff-Appellee,
v.
SEAN LAMAR SANDERS, a/k/a Sean Lamont Sanders, Defendant-Appellant.
No. 00-6281
UNITED STATES COURT OF APPEALS FOR THE FOURTH CIRCUIT
Argued: February 26, 2001
Decided: April 13, 2001
Appeal from the United States District Court for the Eastern District of North Carolina, at Raleigh. W. Earl Britt, Senior District Judge.
(CR-97-149, CA-99-860-5-BR)COUNSEL: ARGUED: Matthew C. Hicks, Student Counsel, Appellate Litigation Program, GEORGETOWN UNIVERSITY LAW CENTER, Washington, D.C., for Appellant. Anne Margaret Hayes, Assistant United States Attorney, Raleigh, North Carolina, for Appellee. ON BRIEF:Steven H. Goldblatt, Director, Adam N. Steinman, Supervising Attorney, Troy B. Klyber, Student Counsel, Appellate Litigation Program, GEORGETOWN UNIVERSITY LAW CENTER, Washington, D.C., for Appellant. Janice McKenzie Cole, United States Attorney, Fenita M. Shepard, Assistant United States Attorney, Raleigh, North Carolina, for Appellee.
Before WILKINSON, Chief Judge, TRAXLER, Circuit Judge, and T. S. ELLIS, III, United States District Judge for the Eastern District of Virginia, sitting by designation.
Affirmed by published opinion. Chief Judge Wilkinson wrote the opinion, in which Judge Traxler and Judge Ellis joined.
OPINION
WILKINSON, Chief Judge:
Sean L. Sanders filed this petition for collateral relief under 28 U.S.C. S 2255 (1994 & Supp. IV 1998). Sanders claimed that his conviction and sentence for conspiracy to distribute cocaine and cocaine base in violation of 21 U.S.C. SS 841 and 846 should be overturned on account of violations of his Fifth and Sixth Amendment rights. The district court dismissed Sanders' motion as untimely under S 2255 because it was filed more than one year after Sanders' conviction became final. Because Sanders' resentencing under Fed. R. Crim. Pro. 35(b) did not renew the clock on S 2255's statute of limitations, because Sanders procedurally defaulted any remaining claims, and because the new rule of criminal procedure announced in Apprendi v. New Jersey, 530 U.S. 466 (2000), does not apply retroactively on collateral review, we affirm the district court's dismissal of Sanders' habeas petition.
I.
On September 23, 1997, Sean L. Sanders was named in a onecount criminal information. The information alleged that Sanders conspired to distribute cocaine and cocaine base in violation of 21 U.S.C. SS 841 and 846. Sanders subsequently waived indictment and pleaded guilty to the criminal information. On January 13, 1998, Sanders was sentenced to 328 months imprisonment and a five-year term of supervised release. The district court also ordered Sanders to pay a $15,792.00 fine and specially assessed him $100. The court entered its judgment on January 15, 1998. Sanders did not appeal.
On December 15, 1998, in light of Sanders' substantial assistance in other prosecutions, the government moved for a reduction in Sanders' sentence pursuant to Fed. R. Crim. Pro. 35(b). On April 16, 1999, the district court granted the government's motion. The court reduced Sanders' term of imprisonment to 188 months and reduced his fine to $9,686.00.
On December 27, 1999, nearly twenty-four months after he had originally been sentenced, Sanders filed a petition for collateral relief under 28 U.S.C. S 2255. Sanders claimed that in light of Jones v. United States, 526 U.S. 227 (1999), his original sentence violated his Fifth Amendment right to due process because the district court did not apply the "beyond a reasonable doubt" standard in determining the type and quantity of the drugs in question. In addition, Sanders claimed that his attorney was ineffective for failing to present this Fifth Amendment argument to the court.
On January 6, 2000, the district court dismissed Sanders' petition for relief. The court noted that the judgment of his conviction became final in January 1998, but that Sanders did not file his petition for collateral relief until December 1999. The court ruled that Sanders' petition was therefore untimely under S 2255's one-year statute of limitations. Sanders now appeals.
II.
Prior to 1996, there was no time limitation on a federal prisoner's ability to collaterally attack his conviction in aS 2255 motion. See United States v. Torres, 211 F.3d 836, 838 (4th Cir. 2000). This changed in 1996 with Congress' enactment of the Antiterrorism and Effective Death Penalty Act, Pub. L. No. 104-132, 110 Stat. 1214 ("AEDPA"). AEDPA amended 28 U.S.C. S 2255 to provide a oneyear limitations period for the filing of S 2255 motions. Section 2255's statute of limitations provides, in relevant part:
A 1-year period of limitation shall apply to a motion under this section. The limitation period shall run from the latest of -
(1) the date on which the judgment of conviction becomes final; . . . [or]
(3) the date on which the right asserted was initially recognized by the Supreme Court, if that right has been newly recognized by the Supreme Court and made retroactively applicable to cases on collateral review.
28 U.S.C. S 2255 (1994 & Supp. IV 1998).1
III.
Sanders contends that his motion was timely under both subsections (1) and (3). Sanders first contends his habeas petition was timely under S 2255 subsection (1). Sanders concedes that if this court construes the date upon which his "judgment of conviction becomes final" to be January 15, 1998, then his motion is untimely. January 15, 1998, was the date on which the district court entered its judgment from which Sanders chose not to appeal. Sanders contends, however, that the one-year limitations period did not begin to run until the completion of his resentencing under Fed. R. Crim. Pro. 35(b). This occurred on April 16, 1999. Since Sanders filed his S 2255 motion on December 27, 1999, roughly eight months after he was resentenced, he claims that his motion therefore is timely.
We disagree. Congress did not explicitly state in the AEDPA when a "judgment of conviction becomes final" for purposes of S 2255 subsection (1). See Torres, 211 F.3d at 838. In Torres, however, this court held that "for purposes of S 2255, the conviction of a federal prisoner whose conviction is affirmed by this Court and who does not file a petition for certiorari becomes final on the date that this Court's mandate issues in his direct appeal." Torres , 211 F.3d at 837. Under the reasoning of Torres, Sanders' conviction became final on the date upon which he declined to pursue further direct appellate review. The district court entered Sanders' judgment of conviction on January 15, 1998. Since Sanders did not file a direct appeal, his conviction became final for purposes of S 2255 subsection (1) on that date.
Contrary to Sanders' assertions, Congress did not intend for Fed. R. Crim. Pro. 35(b) motions to prevent convictions from becoming final for S 2255 purposes. The plain language of 18 U.S.C. S 3582(b) establishes that a modification of a sentence does not affect the finality of a criminal judgment. Section 3582(b) states:
(b) Effect of finality of judgment. -Notwithstanding the fact that a sentence to imprisonment can subsequently be -
(1). modified pursuant to the provisions of subsection (c);
(2). corrected pursuant to the provisions of rule 35 of the Federal Rules of Criminal Procedure and section 3742; or
(3). appealed and modified, if outside the guideline range, pursuant to the provisions of section 3742;
a judgment of conviction that includes such a sentence constitutes a final judgment for all other purposes.
18 U.S.C. S 3582(b) (emphasis added).
As S 3582(b)(1) contemplates, Sanders' sentence was modified pursuant to S 3582(c)(1)(B). That section permits courts to "modify an imposed term of imprisonment to the extent otherwise expressly permitted by statute or by Rule 35 of the Federal Rules of Criminal Procedure." 18 U.S.C. S 3582(c)(1)(B). The plain text of S 3582(b) clearly states that this later modification does not affect the date on which Sanders' judgment of conviction became final"for all other purposes." The conviction which the district court entered on January 15, 1998, included "a sentence to imprisonment." Therefore that judgment of conviction "constitutes a final judgment for all other purposes" under 18 U.S.C. S 3582(b), including the beginning of S 2255 subsection (1)'s limitations period.2
Both Sanders and the government emphasize at length various policy arguments to explain why this court should or should not construe subsection (1)'s limitations period to begin running only after a Rule 35(b) sentence modification is granted. Sanders argues that his interpretation of subsection (1) will streamline court proceedings by allowing defendants to present a single S 2255 motion, rather than one for claims stemming from their original sentences and another for claims stemming from their Rule 35(b) resentencings. Sanders also argues that his construction of subsection (1) will encourage more cooperation between defendants and the government by removing defendants from the uncomfortable position of having to litigate against the government in their S 2255 motions, while simultaneously assisting the government in order to obtain its support for a Rule 35(b) sentence reduction. Finally, Sanders claims that his interpretation will eliminate the burden on defendants of having to litigate their ineffective assistance of counsel claims under S 2255, while simultaneously being represented by that very counsel in their Rule 35(b) resentencings.
For its part, the Government contends that Sanders' construction would result in many stale S 2255 claims. It points out that although Rule 35(b) motions ordinarily must be filed within one year after the sentence is imposed, motions may be made at a later date, sometimes even years afterwards. See Fed. R. Crim. Pro. 35(b). The government also argues that the efficiency gains from Sanders' proposed construction of S 2255 subsection (1) would be de miminis because Rule 35(b) sentence modifications rarely, if ever, lead toS 2255 challenges. The government also argues that Sanders' proposed rule would inadvisably encourage defendants with valid ineffective assistance of counsel claims to persist in being represented by their deficient counsel in their Rule 35(b) proceedings.
It is not our place, however, to weigh the relative merits of these policy arguments. Congress already has considered these very concerns, and it has resolved them with 18 U.S.C. S 3582(b). We are obligated to follow Congress' judgment on this matter and apply the plain text of the statute. Since Sanders' Rule 35(b) resentencing does not affect the date on which the judgment of his conviction became final, Sanders' motion was not timely under S 2255 subsection (1).3
IV.
Sanders next argues that his claims are not time-barred under subsection (3) of S 2255. See In re Vial, 115 F.3d 1192, 1197 n.9 (4th Cir. 1997) (en banc). Even assuming, solely for purposes of this appeal, that Sanders is correct, his habeas petition still faces two insurmountable obstacles.
A.
Sanders did not argue at his original sentencing that a jury must determine beyond a reasonable doubt the types or quantities of drugs involved in his conspiracy. Nor did he raise this claim on direct appeal. As the Supreme Court has admonished, "[h]abeas review is an extraordinary remedy and will not be allowed to do service for an appeal." Bousley v. United States, 523 U.S. 614, 621 (1998) (internal citations omitted). "Where a defendant has procedurally defaulted a claim by failing to raise it on direct review, the claim may be raised in habeas only if the defendant can first demonstrate either `cause' and actual `prejudice,' or that he is `actually innocent.'" Id. at 622 (internal citations omitted).
Sanders first attempts to demonstrate "cause." He argues that the legal basis for his claim was not reasonably available to his counsel at the time his plea was entered. He notes further that the Supreme Court has held "that a claim that `is so novel that its legal basis is not reasonably available to counsel' may constitute cause for a procedural default." Bousley, 523 U.S. at 622 (citing Reed v. Ross, 468 U.S. 1, 16 (1984)).
In Bousley, however, the Supreme Court elaborated on just what constitutes a novel claim. In that case, Bousley claimed on collateral review that his guilty plea for "using" a firearm in violation of 18 U.S.C. S 924(c)(1) was not knowing and intelligent because the district court misinformed him of the nature of the charged crime. Bousley's claim was based on the fact that, five years after his conviction, the Supreme Court held in Bailey v. United States, 516 U.S. 137, 144 (1995), that S 924(c)(1)'s "use" prong requires the Government to show "active employment of the firearm," rather than mere possession. Bousley, 523 U.S. at 616.
Although the Court recognized that its holding in Bailey changed the relevant legal landscape, it held that Bousley's claim was not novel. The Court reached this conclusion because other defendants had previously challenged their S 924(c)(1) convictions on grounds that "use" required more than the mere possession of a firearm. See Bousley, 523 U.S. at 622-23.
The Court further rejected Bousley's argument that his failure to raise the claim on direct review "should be excused because, before Bailey, any attempt to attack his guilty plea would have been futile." Bousley, 523 U.S. at 623 (internal quotation marks omitted). Instead, the Court stated that "futility cannot constitute cause if it means simply that a claim was unacceptable to that particular court at that particular time." Id. (citing Engle v. Isaac, 456 U.S. 107, 130 n.35 (1982)).
Under the Bousley analysis, Sanders simply cannot show cause to explain his failure to raise his Apprendi argument on direct appeal. The Seventh Circuit has recently addressed this precise issue, holding that a petitioner procedurally defaulted his Apprendi claim when he failed to raise it at his trial in 1992. See United States v. Smith, 241 F.3d 546 (7th Cir. 2001). Smith determined that the petitioner's claims were not novel because "the foundation for Apprendi was laid long before 1992. Other defendants had been making Apprendi-like arguments ever since the Sentencing Guidelines came into being, and in McMillan v. Pennsylvania, 477 U.S. 79 (1986), the Court addressed on the merits an argument along similar lines." Smith, 241 F.3d at 548 (internal citations omitted). See also United States v. Nance, 236 F.3d 820, 823 (7th Cir. 2000) ("Indeed, as far back as 1997 defendants were arguing to this court that the type of drug that they were charged with distributing in violation of 21 U.S.C. S 841 was an element of the offense that had to be proven to the jury beyond a reasonable doubt.") (citing United States v. Edwards, 105 F.3d 1179 (7th Cir. 1997)).
Smith further rejected the petitioner's futility argument, which is identical to Sanders' here. The Smith petitioner argued that he showed cause because, prior to Apprendi, the federal circuit courts had held that drug quantity under 21 U.S.C. S 841(b) was a statutory sentencing factor rather than a substantive element of the offense. See, e.g., United States v. Powell, 886 F.2d 81, 85 (4th Cir. 1989); United States v. Edwards, 105 F.3d 1179 (7th Cir. 1997). In light of this contrary precedent, the petitioner claimed that raising his Apprendi claims on direct appeal would have been futile. Applying Bousley, Smith rejected this argument. Instead, the court recognized that "`[c]ause' means some impediment, and [the petitioner] does not contend that any outside force impeded his legal defense." Smith, 241 F.3d at 548.
We agree with the Seventh Circuit's analysis. Adopting petitioner's view of novelty as a cause for procedural default would invite criminal defendants to bypass the preferred procedural avenue of trial and direct appeal in favor of collateral review. Collateral review would come in turn to serve as an all-purposive receptacle for claims which in hindsight appear more promising than they did at the time of trial. Bousley did not intend every change in criminal procedure to become the occasion for reopening a judgment. The germ of Sanders' Apprendi claim had sprouted at the time of his conviction and there is no reason why he could not have raised it then. Although the court may not have been likely to accept Sanders' argument, Sanders plainly had at his disposal the essential legal tools with which to construct his claim. Smith, 241 F.3d at 548. As a result, Sanders has procedurally defaulted his Apprendi claim.
B.
Sanders' argument that his claims are not barred under S 2255 subsection (3) faces an additional impediment -namely, that he seeks to assert a new rule on collateral review in contravention of Teague v. Lane, 489 U.S. 288 (1989). New rules of constitutional criminal procedure "are generally not applied retroactively on collateral review." United States v. Mandanici, 205 F.3d 519, 527 (2d Cir. 2000) (internal citations omitted). Furthermore, neither the Supreme Court nor any of our sister circuits have held that Apprendi is retroactively applicable on collateral attack.4 In fact, the Ninth Circuit has recently concluded just the opposite. See Jones v. Smith, 231 F.3d 1227 (9th Cir. 2000). Several district courts have also held that Apprendi does not apply to initial habeas corpus petitions. See Levan v. United States, 128 F. Supp. 2d 270, 275-76 (E.D. Pa. 2001) (listing cases); but see United States v. Murphy, 109 F. Supp. 2d 1059 (D. Minn. 2000) (holding that Apprendi applies retroactively); Darity v. United States, 124 F. Supp. 2d 355 (W.D.N.C. 2000) (same). We agree with the Ninth Circuit and the majority of district courts that Apprendi does not apply retroactively to cases on collateral review.
In Teague v. Lane, the Supreme Court established a three-step inquiry to determine when new rules of criminal procedure apply retroactively on collateral review. See 489 U.S. at 288. First, the court must determine the date on which the defendant's conviction became final. See O'Dell v. Netherland, 521 U.S. 151, 156-57 (1997) (citing Lambrix v. Singletary, 520 U.S. 518, 527 (1997)). Second, the court must decide whether the Supreme Court's ruling indeed constitutes a "new rule" of constitutional criminal procedure. Third, if the rule is new, then it does not apply retroactively unless it falls within one of the two narrow exceptions to the Teague bar.
Sanders initially argues that Apprendi is not subject to Teague's three-step test because it sets forth a new rule of substantive, rather than procedural, criminal law. To the contrary, Apprendi constitutes a procedural rule because it dictates what fact-finding procedure must be employed to ensure a fair trial. Teague, 489 U.S. at 312; Apprendi, 120 S.Ct. at 2354 ("The substantive basis for New Jersey's enhancement is thus not at issue; the adequacy of New Jersey's procedure is.").
Since Apprendi announced a procedural rule, we now turn to the three-step Teague analysis. First, as noted above, Sanders' conviction became final on January 15, 1998. Second, Apprendi is certainly a new rule of criminal procedure. A new rule is one that was not dictated by precedent at the time the defendant's conviction became final. See O'Dell, 521 U.S. at 156. As of January 15, 1998, the Supreme Court had yet to decide either Jones v. United States, 526 U.S. 227 (1999), or Apprendi v. New Jersey, 530 U.S. 466 (2000), the two cases upon which Sanders bases his S 2255 motion. Rather, the leading Supreme Court case in this area was McMillan v. Pennsylvania, 477 U.S. 79 (1986). In McMillan, the Court ruled that a state could treat the possession of a firearm as a sentencing factor rather than as an element of a particular offense. See McMillan, 477 U.S. at 91. As a result, the Court held that the state did not need to prove the existence of this sentencing factor beyond a reasonable doubt, and the defendant was not entitled to a jury trial on that particular question. Id. at 91-93.
Since only McMillan had been decided by the time Sanders' conviction became final, it is clear that a reasonable jurist in January 1998 would not have felt compelled to adopt the rule later set out in Apprendi; i.e. that any factor which increased the maximum punishment for an offense must be found by a jury beyond a reasonable doubt. This claim is illustrated by the fact that, prior to Apprendi, every federal circuit court considered drug quantity to be a sentencing factor for a judge to determine based on a preponderance of the evidence. See, e.g., United States v. Powell, 886 F.2d 81, 85 (4th Cir. 1989); United States v. Campuzano, 905 F.2d 677, 679 (2d Cir. 1990). As the Ninth Circuit recognized in Jones , "Apprendi certainly established a new rule; that much is clear from the Supreme Court's declaration in Jones v. United States, 526 U.S. 227, 248 (1999), that the issue was, as of 1999, `not yet settled.'" Jones, 231 F.3d at 1236.5
New rules of constitutional criminal procedure are generally not applied retroactively on collateral review. This rule is subject only to the two narrow exceptions discussed in Teague . Teague's first exception addresses new rules which forbid "criminal punishment of certain primary conduct" and new rules which prohibit"a certain category of punishment for a class of defendants because of their status or offense." See O'Dell, 521 U.S. at 157. The first exception clearly does not apply here because Apprendi did not place drug conspiracies beyond the scope of the state's authority to proscribe. Cf. Jones, 231 F.3d at 1237 (finding Teague's first exception inapplicable to petitioner's Apprendi claim because "the state's authority to punish Petitioner for attempted murder is beyond question").
Accordingly, Apprendi does not apply retroactively unless it falls within Teague's second exception. For a rule to qualify, the new rule must be such that, without it, "the likelihood of an accurate conviction is seriously diminished." Teague, 489 U.S. at 313. Improving the accuracy of trial, however, is not sufficient. Rather, a "rule that qualifies under this exception must not only improve accuracy, but also alter our understanding of the bedrock procedural elements essential to the fairness of a proceeding." Sawyer v. Smith, 497 U.S. 227, 242 (1990) (internal quotations omitted). The Supreme Court has pointed to Gideon v. Wainwright, 372 U.S. 335 (1963), as the type of rule that would satisfy Teague's second exception. See Saffle v. Parks, 494 U.S. 484, 495 (1990).
The Supreme Court has stated that it is unlikely that many such rules remain undiscovered. "Whatever the precise scope of this second exception, it is clearly meant to apply only to a small core of rules requiring observance of those procedures that are implicit in the concept of ordered liberty." O'Dell, 521 U.S. at 157. The Court has repeatedly emphasized the rarity of new bedrock rules of procedure. "Because we operate from the premise that such procedures would be so central to an accurate determination of innocence or guilt, we believe it unlikely that many such components of basic due process have yet to emerge." Graham v. Collins, 506 U.S. 461, 478 (1993). Indeed, since Teague, the Court has yet to find a single rule that qualifies under the second exception. See United States v. Mandanici, 205 F.3d 519, 529 (2d Cir. 2000) (discussing eleven new rules or proposed new rules, which the Supreme Court has declined to apply retroactively). The new rule announced in Apprendi is dual-faceted. The Supreme Court held both that a jury, rather than a judge, must determine the facts supporting a statutory sentencing enhancement, and that this finding must be made beyond a reasonable doubt, rather than by a preponderance of the evidence. These rules, however, are not the types of watershed rules implicating fundamental fairness that require retroactive application on collateral attack.
As Sanders recognizes, a rule which merely shifts the fact-finding duties from an impartial judge to a jury clearly does not fall within the scope of the second Teague exception. See Neder v. United States, 527 U.S. 1 (1999). Neder came on the heels of United States v. Gaudin, 515 U.S. 506 (1995), which held that the issue of materiality in false statement prosecutions under 18 U.S.C.S 1001 must be decided by a jury rather than a judge. Neder nonetheless held that a district court's failure to submit a finding of materiality to the jury was harmless error. Neder, 527 U.S. at 15.
Certainly, if having a judge rather than a jury consider an element of the offense amounts only to harmless error, then a new rule mandating a jury to decide an issue rather than a judge cannot fall within the scope of the second Teague exception. The federal circuit courts have recognized this very point, holding that the new rule announced in Gaudin does not apply retroactively on collateral review. See, e.g., Bilzerian v. United States, 127 F. 3d 237, 241 (2d Cir. 1997); United States v. Shunk, 113 F.3d 31, 37 (5th Cir. 1997). As the Second Circuit explained in Bilzerian, the Gaudin rule "merely shift[ed] the determination of materiality from the judge to the jury" and there is "little reason to believe that juries will have substantially different interpretations of materiality than judges." Bilzerian, 127 F.3d at 241. The same reasoning applies to Sanders' case: there is little reason to believe that juries will have substantially different interpretations of drug type and quantity than judges.
Sanders argues, however, that it is not so much the identity of the fact-finder as it is the reasonable doubt standard that profoundly implicates both the fairness and the accuracy of criminal proceedings.6 Sanders likens Apprendi to Cage v. Louisiana, 498 U.S. 39, 40-41 (1990), in which the Supreme Court announced a new rule that a jury instruction equating the reasonable doubt standard with "moral certainty" unconstitutionally diluted the reasonable doubt standard. Sanders correctly points out that this court has held that Cage falls within the second Teague exception and applies retroactively on collateral review. See Adams v. Aiken, 41 F.3d 175, 178 (4th Cir. 1994). Sanders contends that Adams is indistinguishable from this case. He also notes that this term the Supreme Court is reviewing the retroactivity of Cage's new rule. See Tyler v. Cain, No. 00-5961, 69 U.S.L.W. 3393 (U.S. Dec. 12, 2000).
Even assuming the Court applies Cage's rule retroactively in Tyler, the question raised in this case is fundamentally different. Tyler turns on the standard of proof under which the entire case against the defendant is submitted to the jury. Here, in contrast, Sanders cannot claim that his entire conviction is tainted. Rather, his claim concerns the district court's omission of only one "element" of the offense, i.e. drug quantity. Cf. Neder, 527 U.S. at 11. As the Supreme Court explained in Johnson v. United States, 520 U.S. 461, 468-69 (1997), and Neder, 527 U.S. at 9-15, an instruction that gives a proper reasonable doubt instruction but omits an element of the offense does not necessarily render a criminal trial fundamentally unfair, even though it prevents the jury from rendering a "complete verdict" on every element of the offense. Rather, in those cases the Supreme Court held that the district court's failure to submit an element to the jury was subject to harmless-error (Neder) and plain-error (Johnson) analysis.
We may readily assume that the rule that certain sentencing factors must be proven beyond a reasonable doubt will promote marginally more accurate results. Cf. In re Winship, 397 U.S. 358, 363 (1970); Sullivan v. Louisiana, 508 U.S. 275, 281 (1993). In order for a new rule to qualify for the second Teague exception, however, it must not only promote accuracy, but also "alter our understanding of the bedrock elements essential to the fairness of a proceeding." Sawyer, 497 U.S. at 242. A simple comparison between the Apprendi rule and the paradigmatic watershed principle announced in Gideon v. Wainwright indicates that the former lacks the primacy and centrality of the latter. Gideon, which established an affirmative right to counsel in all felony cases, was remarkable in part because of its sweeping breadth. In contrast, Apprendi merely applies to the subsection of criminal defendants whose maximum punishment may be increased by a sentencing factor. If a defendant's sentence is within the statutory limits, i.e. below the maximum sentence, then Apprendi would not apply. See Apprendi, 120 S. Ct. at 2358.
More importantly, we do not read Apprendi to hold that the country's criminal justice system malfunctioned so fundamentally prior to the year 2000, as to merit the retroactive application of the Court's new procedural rule. Apprendi itself recognized that judges in this country, in accordance with the Constitution, had long exercised discretion in imposing sentences within the particular range determined by the legislature. "[B]oth before and since the American colonies became a nation, courts in this country and in England practiced a policy under which a sentencing judge could exercise a wide discretion in the sources and types of evidence used to assist him in determining the kind and extent of punishment to be imposed within limits fixed by law." Id. at 2357-58 (citing Williams v. New York, 337 U.S. 241, 246 (1949)). While the Apprendi principle may spark a closely divided debate among contemporary jurists, it would be unusual for the Supreme Court to identify a bedrock principle of procedure that somehow eluded the nation since its founding. Again, Apprendi holds that so long as the sentence is within the statutory range, a judge can increase a defendant's sentence on the basis of certain facts which need not be proven to a jury beyond a reasonable doubt. This very holding of Apprendi undercuts the argument that it states a bedrock principle as envisioned by Teague. Otherwise it would be unconstitutional to allow vast numbers of criminal defendants to receive sentences based on facts found under a preponderance of the evidence standard.
Further supporting the view that Apprendi does not rise to the level of a watershed change in criminal procedure is the fact that the majority of the federal circuit courts have subjected Apprendi claims to harmless and plain error review. See, e.g., United States v. Terry, 240 F.3d 65, 74-75 (1st Cir. 2001); United States v. White, 240 F.3d 127 (2d Cir. 2001); United States v. Lewis, 235 F.3d 215, 218 (4th Cir. 2000); United States v. Meshack, 225 F.3d 556, 575 (5th Cir. 2000); United States v. Nance, 236 F.3d 820, 823-24 (7th Cir. 2000); United States v. Candelario, 240 F.3d 1300 (11th Cir. 2001). As these courts have recognized, it is possible for a criminal defendant to have a fair and accurate trial without the new procedural protection offered by Apprendi. None of these cases have suggested that failure to submit the question of drug quantities to a jury is structural error. We do not suggest, of course, that all structural errors satisfy Teague's second exception. We merely emphasize that finding something to be a structural error would seem to be a necessary predicate for a new rule to apply retroactively under Teague.
It is important, finally, to keep the ebbs and flows of criminal process in some perspective. The Constitution embodies many important protections for those accused of crime. The rights to counsel, to trial by jury, and to be proven guilty beyond a reasonable doubt are all fundamental rights. These watershed principles in turn spawn numerous subsidiary questions, which are closer to the constitutional margins. These subsidiary questions may qualify as arguable applications of a bedrock principle, but they are not core guarantees themselves.
Taking again the example of Gideon, the Court has had to consider, among other things, whether a district court's order that a defendant not speak with his counsel during a court recess violated his Sixth Amendment right to the assistance of counsel, see Perry v. Leeke, 488 U.S. 272 (1989); whether the Sixth Amendment prohibits a sentencing court from considering a defendant's previous uncounseled misdemeanor conviction in sentencing him for a subsequent offense, see Nichols v. United States, 511 U.S. 738 (1994); and whether a state procedure that permits appointed counsel to withdraw without informing the court that the appeal would be frivolous violates the indigent defendant's right to appellate counsel, see Smith v. Robbins, 528 U.S. 259 (2000). The Supreme Court has never intimated that these offshoot cases would somehow automatically apply retroactively simply because they bear some level of descendancy to the ancestral rule established in the Gideon case. To hold otherwise would nullify Teague's admonition that retroactive application of new constitutional rules on collateral review is reserved only for truly exceptional cases. See Teague, 489 U.S. at 313. In short, the new rule announced in Apprendi does not rise to the level of a watershed rule of criminal procedure which "alter[s] our understanding of the bedrock elements essential to the fairness of a proceeding." Sawyer, 497 U.S. at 242. Because the rule does not apply retroactively on collateral review, Sanders cannot challenge his conviction on the basis of any such error.
V.
All of petitioner's arguments have served a single purpose -to undermine the finality of direct appeal within the criminal justice system. The Supreme Court has resisted having collateral review substitute for the appellate function. Bousley's holding on procedural default, and Teague's pronouncement on the retroactive application of new rules are only the latest of a long line of cases emphasizing the centrality of direct review of criminal convictions. With the AEDPA, Congress has likewise spoken to the question, and established firm time limitations to govern the filing of S 2255 petitions. As a federal court, and an inferior one at that, we are bound to heed the admonitions of the Supreme Court and Congress on this score. For the foregoing reasons, the judgment of the district court is
AFFIRMED.
Notes:
The remaining two parts of S 2255's statute of limitations are not relevant to this appeal. Those provisions state:
(2) the date on which the impediment to making a motion created by governmental action in violation of the Constitution or laws of the United States is removed, if the movant was prevented from making a motion by such governmental action; . . . [or]
(4) the date on which the facts supporting the claim or claims presented could have been discovered through the exercise of due diligence.
Sanders argues that S 3582 does not define when a judgment becomes final for purposes of S 2255. In support, he points out that S 3582 states that a judgment is final notwithstanding the fact that a defendant could still challenge his sentence on appeal. SeeS 3582(b)(3). Sanders contends that this in turn would be inconsistent with our holding in Torres that S 2255's statute of limitations begins to run only after direct appeal has concluded. Torres, however, did not address the effect of S 3582 on S 2255's statute of limitations. And in this case, S 3582 is clear: "a judgment of conviction that includes such a sentence constitutes a final judgment for all other purposes." 18 U.S.C. S 3582(b) (emphasis added). The question of how S 3582 works vis-a-vis an appeal must await another day. It is beyond dispute that under S 3582, Rule 35 sentence modifications do not affect the finality of convictions forS 2255 purposes.
Sanders also contends that under the "Healy doctrine" this court should construe S 2255's limitations period in his favor. See United States v. Healy, 376 U.S. 75 (1964). We disagree. Under Healy, a timely motion for reconsideration in a criminal case resets the deadline for filing the notice of appeal. However, a motion to reconsider and a Rule 35(b) motion are not the same. Unlike a Rule 35(b) motion, the issues asserted in a motion for reconsideration are often the same as those raised on appeal. Allowing a district court to reconsider an alleged legal error can eliminate the need for an appeal. These efficiencies explain the Supreme Court's holding in Healy. However, they do not apply to Rule 35(b) motions, which do not ordinarily assert legal error.
We shall assume, without deciding, that a circuit court can declare a new rule retroactive on collateral review in an initial S 2255 petition. Section 2255 clearly states that in order to bring a second or successive petition, the new rule must be "made retroactive to cases on collateral review by the Supreme Court." 18 U.S.C. S 2255. However, since this is Sanders' initial habeas petition, S 2255 subsection (3) controls. Subsection (3) refers to "the date on which the right asserted was initially recognized by the Supreme Court, if that right has been newly recognized by the Supreme Court and made retroactively applicable to cases on collateral review." Since the language of subsection (3) differs from the language governing second or successive motions, it is possible that lower courts can declare new rules retroactive on initial petitions. However, the language of subsection (3) can also be read to require the Supreme Court to make the decision on retroactivity before a petitioner may file an initial S 2255 motion. In view of our disposition of the petition on other grounds, we need not address this question.
Our holding that Apprendi establishes a "new rule" for purposes of the Teague v. Lane retroactivity analysis, does not also mean that Sanders has shown cause to justify his failure to raise this claim on direct appeal. See section IV.A, supra. The two standards simply do not overlap. See Waldrop v. Jones, 77 F.3d 1308, 1315 (11th Cir. 1996). For purposes of the Teague analysis, a rule is new so long as the precedent existing at the time of the defendant's conviction did not dictate the rule. This is a less rigorous requirement than that for showing cause in order to justify procedural default. As Bousley, 523 U.S. at 623, and Engle v. Isaac, 456 U.S. at 130 n.35, indicate, in order to show cause, the defendant must raise any constitutional claims on direct review even if doing so may seem futile in light of existing precedent. See also Smith, 241 F.3d at 546.
Sanders also finds great significance in the fact that Justice O'Connor, in her dissent in Apprendi, referred to the new rule as a "watershed change in constitutional law." Apprendi, 120 S. Ct. at 2380 (O'Connor, J., dissenting). However, the dissent was addressing the practical effects of the majority's holding. The dissent did not purport to address Teague's second exception. In fact, it does not so much as mention the Teague decision.