Ronnie Maurice Howard v. United States

[PUBLISH] IN THE UNITED STATES COURT OF APPEALS FOR THE ELEVENTH CIRCUIT FILED ________________________ U.S. COURT OF APPEALS ELEVENTH CIRCUIT No. 03-11919 June 25, 2004 THOMAS K. KAHN ________________________ CLERK D. C. Docket Nos. 02-00856-CV-CB-M and 97-00178-CR-CB RONNIE MAURICE HOWARD, Petitioner-Appellant, versus UNITED STATES OF AMERICA, Respondent-Appellee. ________________________ Appeal from the United States District Court for the Southern District of Alabama _________________________ (June 25, 2004) Before CARNES and WILSON, Circuit Judges, and HANCOCK *, District Judge. CARNES, Circuit Judge: Ronnie Maurice Howard appeals from the denial of his 28 U.S.C. § 2255 * Honorable James H. Hancock, United States District Judge for the Northern District of Alabama, sitting by designation. motion as time-barred. The district court concluded that the Supreme Court’s ruling in Alabama v. Shelton, 535 U.S. 654, 122 S. Ct. 1764 (2002), did not recognize a “new” right and therefore did not restart the running of the one-year period of limitation under § 2255 ¶ 6(3). The district court’s conclusion is not an unreasonable one, but we are obligated to exercise de novo review, Castro v. United States, 290 F.3d 1270, 1272 (11th Cir. 2002), and doing so we reach the opposite conclusion. We believe that Shelton did recognize a new right which is retroactively applicable on collateral review, thereby bringing his case within the rewind provision of § 2255 ¶ 6(3). As a result, Howard’s motion was timely filed. We also believe that he procedurally defaulted his claim by not raising it at his sentencing proceeding where the error occurred, but the government procedurally defaulted Howard’s procedural default by failing to raise this affirmative defense in the district court. The bottom line is that we will reverse and remand. I. On March 25, 1997, Howard was convicted of Assault 3rd degree in the Municipal Court of Selma, Alabama. On May 6, 1997, in an unrelated proceeding, he was convicted of the unauthorized use of a motor vehicle in the state district court of Autauga County, Alabama. In neither proceeding was Howard 2 represented by counsel. Each conviction resulted in a suspended sentence and probation. In 1998, Howard pleaded guilty in federal court to bank robbery in violation of 18 U.S.C. § 2113(a) and (d), and to use of a firearm during a crime of violence in violation of 18 U.S.C. § 924(c). At sentencing, the court calculated Howard’s base offense level to be 20. The court then assessed one criminal history point for the assault conviction and one for the unauthorized use of a vehicle conviction. It added two more because Howard was on probation at the time of the bank robbery. Those four points raised Howard’s criminal history category from I to III, which increased the sentencing range for his bank robbery conviction from 33-41 months to 41-51 months. U.S.S.G. Ch. 5 Pt. A (Nov. 1997) (sentencing table). The court sentenced Howard to 41 months for that conviction. The four points had no effect on the sentence for his firearm conviction, which was statutorily set at 60 months. The sentences were made to run consecutively. The sentence proceeding was on May 19, 1998, and the judgment became final June 15, 1999. On November 6, 2002, more than a year after final judgment but within a year of the Supreme Court’s May 20, 2002 Shelton decision, Howard filed in the district court what he styled as a 28 U.S.C. § 2241 federal habeas petition. The district court correctly construed the petition as a motion to vacate pursuant to § 3 2255. See Medberry v. Crosby, 351 F.3d 1049, 1056-59 (11th Cir. 2003) (explaining the relationship between § 2241 and § 2255). Relying on Shelton, Howard claimed that the sentencing court had violated his Sixth Amendment right to counsel by considering the two uncounseled state court convictions in calculating his criminal history category. Believing that Shelton did not recognize a new right, the district court denied Howard’s motion as time-barred. It did grant him a certificate of appealability which, coupled with a notice of appeal, brought the case to us. II. When the district court sentenced Howard for his robbery and firearm convictions, Howard failed to object to the court’s counting his uncounseled state court convictions as part of his criminal history. He acknowledges that but says it should not matter because the issue is one that cannot be procedurally defaulted. The use of an uncounseled conviction is, Howard contends, a jurisdictional defect. We have noted that a jurisdictional defect cannot be waived or procedurally defaulted and that a defendant need not show cause and prejudice to justify his failure to raise one. McCoy v. United States, 266 F.3d 1245, 1249 (11th Cir. 2001). Relying on language from Gideon v. Wainwright, 372 U.S. 335, 83 S. Ct. 792 (1963), and its progeny, Howard contends that, in the words of the Supreme 4 Court, the ‘“failure to appoint counsel for an indigent [is] a unique constitutional defect . . . ris[ing] to the level of a jurisdictional defect.”’ Lackawanna County Dist. Attorney v. Coss, 532 U.S. 394, 404, 121 S. Ct. 1567, 1574 (2001) (quoting Custis v. United States, 511 U.S. 485, 496, 114 S. Ct. 1732, 1738 (1994)). He maintains that because his sentencing was tainted by a defect that rose to the level of a jurisdictional defect, he can bring up that defect for the first time in this collateral proceeding. The legal premise for Howard’s position is based upon language wrenched from its context in the Lackawanna and Custis opinions. Lackawanna held that because of its special status a Gideon-type defect in a prior conviction may be raised collaterally in a sentence proceeding where that prior conviction is offered as a basis for enhancement. 532 U.S. at 404-05, 121 S. Ct. at 1574. But the Supreme Court did not decide in Lackawanna or any other case that procedural defenses do not apply to claims of Gideon-type errors. Just the opposite. The Court said in Lackawanna that: “As with any § 2254 petition, the petitioner must satisfy the procedural prerequisites for relief including . . . exhaustion of remedies.” Id. at 404, 121 S. Ct. at 1574. Compliance with contemporaneous objection rules is a procedural prerequisite for relief on Gideon-related grounds in a § 2255 proceeding, just as it is in a § 2254 proceeding. That much is clear from the Supreme Court’s opinion in 5 Daniels v. United States, 532 U.S. 374, 121 S. Ct. 1578 (2001). There the Court said: A defendant may challenge a prior conviction as the product of a Gideon violation in a § 2255 motion, but generally only if he raised that claim at his federal sentencing proceeding. See United States v. Frady, 456 U.S. 152, 167-68, 102 S. Ct. 1584 (1982) (holding that procedural default rules developed in the habeas corpus context apply in § 2255 cases); see also Reed v. Farley, 512 U.S. 339, 354-55, 114 S. Ct. 2291 (1994). Id. at 382-83, 121 S. Ct. at 1583-84. The use of the qualifier “generally” in the quoted passage recognizes the possibility of an exception where “a habeas petition directed at the enhanced sentence may effectively be the first and only forum available for review of the prior conviction.” Lackawanna, 532 U.S. at 406, 121 S. Ct. at 1575. That is not the situation in the vast majority of cases, like this one, where the defendant could have raised the claim about the prior conviction in the sentencing proceeding in which that conviction was used against him. It is true that Supreme Court decisions from the time of Gideon to the present day have reflected “a theme that failure to appoint counsel for an indigent defendant was a unique constitutional defect.” Custis, 511 U.S. at 496, 114 S. Ct. at 1738. That theme has been manifested in a willingness to allow the defect to be raised collaterally in a sentence proceeding in which the conviction in question is being offered for use. It has not, however, been manifested in a willingness to 6 disregard applicable procedural defenses, one of which arises from the failure to raise the claim in the sentence proceeding. It is true that the language of jurisdiction was used in some early opinions to describe this type of error and claim. See id. (citing Johnson v. Zerbst, 304 U.S. 458, 468, 58 S. Ct. 1019, 1024-25 (1938)). But the Supreme Court dealt with and minimized the significance of that language in Custis. There the Court explained that at the time that jurisdiction language was first used, “the underlying habeas statute was construed to allow collateral attacks on final judgments of conviction only where the rendering court lacked ‘jurisdiction’ – albeit a somewhat expansive notion of ‘jurisdiction.’” Id. at 494, 114 S. Ct. at 1737 (citing Moore v. Dempsey, 261 U.S. 86, 43 S. Ct. 265 (1923)). Saying that a Gideon error was “jurisdictional” meant no more than that habeas relief could be granted based upon it. See id. As the construction of the habeas statute changed, the need to phrase entitlement to relief in terms of jurisdiction ended. A modern day petitioner cannot rely on outdated language about Gideon errors rising to the level of jurisdictional defects in order to get past procedural defenses, because the Supreme Court has stated in two modern decisions – Daniels and Lackawanna, both decided in 2001 – that procedural defenses do apply to Gideon-based claims. 7 This does not mean that there are no exceptions to procedural bar defenses. There are. “A habeas petitioner can escape the procedural default doctrine either through showing cause for the default and prejudice, or establishing a fundamental miscarriage of justice.” Bailey v. Nagle, 172 F.3d 1299, 1306 (11th Cir. 1999) (internal quotation marks and citations omitted). In order to show the type of “miscarriage of justice” that will excuse a procedural bar, a petitioner must make “a colorable showing of actual innocence.” Id. Howard has not attempted to do that. We don’t need to address prejudice, because he has not established cause. The only arguable cause Howard hints at is novelty. We have explained what is required to successfully establish the novelty of a claim as cause: In Reed v. Ross, 468 U.S. 1, 16, 104 S. Ct. 2901, 2910 (1984), the Supreme Court held “that where a constitutional claim is so novel that its legal basis is not reasonably available to counsel, a defendant has cause for his failure to raise the claim in accordance with acceptable state procedures.” In order to establish the novelty of a constitutional claim sufficient to provide cause, a defendant must initially demonstrate that his situation is one where a court has “articulated a constitutional principle that has not been previously recognized but which has been held to have retroactive application.” Id. at 17, 104 S. Ct. at 2911. . . . A new retroactive decision must be a sufficiently “clear break with the past,” so that an attorney representing the defendant would not reasonably have had the tools for presenting the claim in the state courts. [Id. at 16-17, 104 S. Ct. at 2910-11.] Hargrave v. Dugger, 832 F.2d 1528, 1530-31 (11th Cir. 1987). Where a number of others had raised the claim before the petitioner failed to do so, the claim is not sufficiently novel to meet the cause requirement. See Turner v. Crosby, 339 F.3d 8 1247, 1282 (11th Cir. 2003) (citing Bousley v. United States, 523 U.S. 614, 622- 23, 118 S. Ct. 1604, 1611 (1998)). Here, the legal basis for the right later recognized in Shelton was readily available at the time of Howard’s federal sentencing hearing. That is why the Shelton Court was able to observe that “[c]ourts have divided on the Sixth Amendment question presented in this case.” 535 U.S. at 660, 122 S. Ct. at 1768. The opinion in Shelton cites three federal courts of appeals decisions – all on the books at the time of Howard’s sentencing proceeding – recognizing that appointment of counsel is a constitutional prerequisite to imposition of a conditional or suspended prison sentence. Id. (citing United States v. Reilley, 948 F.2d 648, 654 (10th Cir. 1991); United States v. Foster, 904 F.2d 20, 21 (9th Cir. 1990); and United States v. White, 529 F.2d 1390, 1394 (8th Cir. 1976)). The existence of these decisions at the time of Howard’s sentencing establishes that other defendants had long been raising the issue. Howard could have as well. In fact, it would have been easier for him to raise the issue, because those three earlier decisions themselves provided Howard with additional building blocks with which to construct a Shelton claim at his sentencing proceeding. The claim was not sufficiently unheard to be novel for cause purposes. 9 This means that Howard’s claim is procedurally barred, unless the government is itself barred from raising that affirmative defense because of its own default. And it is. The government failed to raise the defense of procedural default in the district court, and the court did not bring it up either. In these circumstances Gray v. Netherland, 518 U.S. 152, 165-66, 116 S. Ct. 2074, 2082 (1996), prevents the government from benefitting now from a defense it did not raise then. For that reason, we turn to the defense that the government did raise in the district court and upon which the court did base its dismissal. III. Section 2255 imposes a one year limitation period on the filing of motions for relief under that section. As it relates to this case, the limitation period runs from the later 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. § 2255 ¶ 6(1),(3). This case turns on whether the Shelton decision fits within that third provision, because Howard filed his motion for relief more than a year after his conviction became final but less than a year after the Shelton decision 10 was released. The district court ruled that the right recognized in Shelton was not a “newly recognized” one. We disagree. A. In deciding “newly recognized” right issues arising under § 2255 ¶ 6(3), we have applied decisions involving the Teague retroactivity doctrine. See Garcia v. United States, 278 F.3d 1210, 1212-15 (11th Cir. 2002) (accepting parties’ concession that right was newly recognized to satisfy Teague’s new rule requirement, and applying Teague’s retroactivity analysis); Dodd v. United States, 365 F.3d 1273, 1278 (11th Cir. 2004) (concluding that a right was newly recognized based on precedent establishing that a new rule had been announced for Teague purposes). Under that doctrine, “a case announces a new rule when it breaks new ground or imposes a new obligation on the States or the Federal Government. To put it differently, a case announces a new rule if the result was not dictated by precedent existing at the time the defendant’s conviction became final.” Teague v. Lane, 489 U.S. 288, 301, 109 S. Ct. 1060, 1070 (1989) (internal citations omitted). A result is not dictated by precedent just because “the result the habeas petitioner seeks is within the logical compass of a prior Supreme Court decision,” or because “prior Supreme Court decisions inform, or even control or govern, the 11 analysis of the claim.” Spaziano v. Singletary, 36 F.3d 1028, 1042 (11th Cir. 1994) (internal quotation marks omitted). For these purposes, a result is dictated by precedent only if the court considering the claim at the time the conviction became final “would have felt compelled by existing precedent to conclude that the rule [the defendant] seeks was required by the Constitution.” Glock v. Singletary, 65 F.3d 878, 884 (11th Cir. 1995) (internal quotation marks omitted). It is not a dictated result if the case’s outcome was “susceptible to debate among reasonable minds.” Id. The Supreme Court held in Shelton that “a suspended sentence that may ‘end up in the actual deprivation of a person’s liberty’ may not be imposed unless the defendant was accorded ‘the guiding hand of counsel’ in the prosecution for the crime charged.” Shelton, 535 U.S. at 658, 122 S. Ct. at 1767 (quoting Argersinger v. Hamlin, 407 U.S. 25, 40, 92 S. Ct. 2006, 2014 (1972)). It said that two prior decisions “controlled” its judgment in the Shelton case: Argersinger v. Hamlin, 407 U.S. 25, 92 S. Ct. 2006 (1972), and Scott v. Illinois, 440 U.S. 367, 99 S. Ct. 1158 (1979). Shelton, 535 U.S. at 657, 122 S. Ct. at 1767. In Argersinger, the Supreme Court held that defense counsel must be appointed in any criminal prosecution “that actually leads to imprisonment even for a brief period.” Argersinger, 407 U.S. at 33, 37, 92 S. Ct. at 2006, 2012. In Scott, the Supreme 12 Court “drew the line at ‘actual imprisonment,’ holding that counsel need not be appointed when the defendant is fined for the charged crime, but is not sentenced to a term of imprisonment.” Shelton, 535 U.S. at 657, 122 S. Ct. at 1767 (quoting Scott, 440 U.S. at 373-74, 99 S. Ct. at 1162). The district court in this case concluded that Shelton was dictated by Argersinger and Scott. It viewed Shelton as a routine application of the “actual imprisonment” rule of those two earlier decisions, even though Shelton applied the requirement of counsel to a suspended sentence where an actual deprivation of liberty is entirely contingent. Shelton had been convicted in an uncounseled proceeding of third-degree assault and sentenced to a jail term of 30 days. Id. at 658, 122 S. Ct. at 1767-68. The trial court had suspended that sentence and placed Shelton on probation for two years. Id. Shelton appealed his suspended sentence, a sentence which had not resulted in even a minute’s incarceration. In no sense had Shelton been subjected to “actual imprisonment.” Id. In that way, the Shelton case was different from the Argersinger case. And, of course, it was different from the Scott case because Shelton did receive a suspended sentence and not merely a fine as Scott had. In deciding whether the Shelton decision was dictated by prior decisions, it is helpful to break its holdings down into two parts. The Shelton Court first held 13 that where the state had not provided counsel to an indigent defendant in a proceeding resulting in a suspended sentence, it violates the Sixth Amendment to jail the defendant thereafter for some or all of the term of that sentence because of a subsequent probation violation. Id. at 662, 122 S. Ct. at 1770. The Court reasoned that when a suspended sentence is unsuspended following a probation violation, the resulting incarceration is not for the probation transgression but for the original offense. Id. It is the uncounseled conviction that results in actual imprisonment, the Court said. Id. We doubt that first holding of Shelton was dictated by Argersinger, the rule of which is that “absent a knowing and intelligent waiver, no person may be imprisoned for any offense . . . unless he was represented by counsel at his trial.” Id. at 662, 122 S. Ct. at 1770 (quoting Argersinger, 407 U.S. at 37, 92 S. Ct. at 2012). Before the Court spoke in Shelton, it was not clear that Argersinger’s actual imprisonment rule applied when there would have been no imprisonment but for a subsequent probation violation. Argersinger itself was a non-contingent actual imprisonment case. It was not a case of contingent imprisonment that became actual only after another event occurred. The defendant in Argersinger, unlike the one in Shelton, was going to jail even if his post-conviction conduct was purely angelic. 14 Even if we could say that the first holding of Shelton was dictated by Argersinger, the Supreme Court went further to reach the result it did. The appeal in Shelton did not involve a defendant who had actually been sent to jail. Shelton was unhappy because he had the threat of imprisonment hanging over him; he wanted to avoid the prospect of jail time if he did violate probation. To reach its ultimate holding – that “a suspended sentence that may ‘end up in the actual deprivation of a person’s liberty’ may not be imposed unless the defendant was accorded ‘the guiding hand of counsel’ in the prosecution for the crime charged,” Shelton, 535 U.S. at 658, 122 S. Ct. at 1767 (quoting Argersinger, 407 U.S. at 40, 92 S. Ct. at 2014) – the Supreme Court had to go beyond its first holding, and way beyond the holding in Argersinger. The Court went from a rule requiring counsel in proceedings which directly result in a sentence of actual imprisonment to one requiring counsel in proceedings which result in a sentence that contains only the possibility of imprisonment. The journey from Argersinger to Shelton may be good constitutional law, but it is still a journey of some distance. The destinations are different. The Court did say that Argersinger and Scott “control[led]” its judgment in Shelton, id. at 657, 122 S. Ct. at 1767, but we know that statement does not mean that either of those two earlier decisions dictated the result the later one reached. 15 We know that because the Supreme Court itself has told us that for these purposes “controlled” does not equate with “dictated.” Saffle v. Parks, 494 U.S. 484, 491, 110 S. Ct. 1257, 1261 (1990); see also Spaziano v. Singletary, 36 F.3d 1028, 1042 (11th Cir.1994) (citing Saffle for the same proposition). As the Shelton dissent pointed out, Scott identified as the central premise of Argersinger “that actual imprisonment is a penalty different in kind from fines or the mere threat of imprisonment,” Scott, 440 U.S. at 373, 99 S. Ct. at 1162, and Scott drew the line defining the right to counsel at actual imprisonment. Shelton, 535 U.S. at 675, 122 S. Ct. at 1776 (Scalia, J., dissenting). The Court in Shelton erased that line and drew another one far enough out to encompass a mere threat of imprisonment. Of course, none of this lessens the force of the rule crafted by the Shelton majority which is, by definition, the law of the land. And we take at full value the Shelton majority’s statement that Argersinger and Scott controlled the result in that case. But neither Shelton nor any other Supreme Court decision has ever said that the rule in that case was dictated by Argersinger and Scott or any other decision. Nor has the Supreme Court ever suggested that the outcome in Shelton was not “susceptible to debate among reasonable minds,” which is another measure of whether a decision is dictated by prior precedent, see Glock v. Singletary, 65 F.3d at 884. The susceptibility of the Shelton issue to debate among reasonable minds 16 is shown by the status of that issue among the lower courts before the Supreme Court resolved the matter. Compare United States v. Reilley, 948 F.2d 648, 654 (10th Cir. 1991) (appointment of counsel is a constitutional prerequisite to imposition of conditional or suspended prison sentence), United States v. Foster, 904 F.2d 20, 21 (9th Cir. 1990) (same), and United States v. White, 529 F.2d 1390, 1394 (8th Cir. 1976) (same), with Cottle v. Wainright, 477 F.2d 269, 274-75 (5th Cir. 1973) (rejecting counsel prerequisite to imposition of suspended sentence), vacated on other grounds, 414 U.S. 895, 94 S. Ct. 221 (1973), Griswold v. Commonwealth, 472 S.E.2d 789, 791 (Va. 1996) (same), and State v. Hansen, 903 P.2d 194, 197 (Mont. 1995) (same). The pre-Shelton split on the question shows that it was “susceptible to debate among reasonable minds,” which means that the answer had not been dictated previously. See Glock, 65 F.3d at 884. The situation here is similar to the one we faced in Turner v. Crosby, 339 F.3d 1247 (11th Cir. 2003). There we held that the Supreme Court’s decision in Ring v. Arizona, 536 U.S. 584, 122 S. Ct. 2428 (2002), was not dictated by its earlier decision in Apprendi v. New Jersey, 530 U.S. 466, 120 S. Ct. 2348 (2000). Turner, 339 F.3d at 1284. In Apprendi, the Court had held that “[o]ther than the fact of a prior conviction, any fact that increases the penalty for a crime beyond the prescribed statutory maximum must be submitted to a jury, and proved beyond a 17 reasonable doubt.” Apprendi, 530 U.S. at 490, 120 S. Ct. at 2362-63. Ring applied Apprendi to the capital sentencing context, holding that aggravating factors at sentencing, because they act as the “functional equivalent of an element of a greater offense,” must be found by a jury beyond a reasonable doubt. Ring, 536 U.S. at 609, 122 S. Ct. at 2443. Ring’s treatment of aggravating factors as elements of a greater offense extended Apprendi, much as Shelton’s treatment of suspended sentences as actual sentences extended Argersinger. Both Ring and Shelton extended an existing rule into a new and different context. For all of these reasons, we conclude that the rule of the Shelton case does involve a “newly recognized” right within the meaning of § 2255 ¶ 6(3). B. Our conclusion that Shelton announced a “newly recognized” right is not the end of the analysis, however, because the rewind provision of § 2255 ¶ 6(3) does not apply unless the newly recognized right has been “made retroactively applicable to cases on collateral review.” Made retroactively applicable by whom? In Tyler v. Cain, 533 U.S. 656, 121 S. Ct. 2478 (2001), the Supreme Court interpreted another AEDPA provision with somewhat similar wording. It held that the “made retroactive to cases on collateral review by the Supreme Court” language in 28 U.S.C. § 2244(b)(2)(A) plainly means that the predicate 18 retroactivity holding can come only from the Supreme Court, not from any other court. Id. at 662, 121 S. Ct. at 2482; id. at 668, 121 S. Ct. at 2485 (O’Connor, J., concurring). Neither the Supreme Court nor this Court has decided if the same is true of § 2255 ¶ 6(3)’s less specific language that the right must have been “newly recognized by the Supreme Court and made retroactively applicable to cases on collateral review.” We have no need to decide that issue here, and we imply no view about it. The government has conceded for purposes of this case that if we ourselves determine that the right announced in Shelton should be applied retroactively, Howard’s motion is not time-barred under § 2255 ¶ 6. We realize that in the future the government may change its mind and press that issue, see Hunter v. United States, 101 F.3d 1565, 1574 (11th Cir. 1996) (“past experience has taught us that on occasion the government's position on criminal law issues is fluid”), but it has not done so in this case. We accept for here and now the government’s waiver of any argument that § 2255 ¶ 6(3) applies only when the Supreme Court itself has made a newly recognized right retroactively applicable to cases on collateral review. Because of that, we proceed to undertake the task of deciding the retroactivity issue ourselves. C. 19 Under Teague, new rules of constitutional law are not to be applied retroactively to cases on collateral review unless they fall into one of two exceptions. Teague, 489 U.S. at 311-13, 109 S. Ct. at 1075-76; see also Saffle, 494 U.S. at 494-95, 110 S. Ct. at 1263-64 (same); Garcia v. United States, 278 F.3d 1210, 1214 (11th Cir. 2002) (same). Only the second exception is in play here. Under it, a new rule should be applied retroactively if it “requires the observance of those procedures that . . . are implicit in the concept of ordered liberty.” Teague, 489 U.S. at 311, 109 S. Ct. at 1076 (internal quotation marks and citations omitted). This exception is “reserved for watershed rules of criminal procedure . . . [that] properly alter our understanding of the bedrock procedural elements.” Id. We have stated that ‘“[t]o fall within the [second] exception, the new rule must satisfy a two-pronged test: (1) it must relate to the accuracy of the conviction; and (2) it must alter our understanding of the bedrock procedural elements essential to the [fundamental] fairness of a proceeding.”’ Garcia, 278 F.3d at 1215 (quoting Nutter v. White, 39 F.3d 1154, 1157 (11th Cir.1994)). Overshadowing our consideration of whether Shelton’s extension of the right to counsel should be made retroactively applicable is one momentous fact: Every extension of the right to counsel from Gideon through Argersinger has been applied retroactively to collateral proceedings by the Supreme Court. The holding 20 of Gideon itself, which established the right to counsel in all felony convictions, 372 U.S. at 344-45, 83 S. Ct. at 796-97, was judged to be retroactively applicable in Kitchens v. Smith, 401 U.S. 847, 847, 91 S. Ct. 1089, 1090 (1971). The right to counsel at plea hearings, recognized in White v. Maryland, 373 U.S. 59, 83 S. Ct. 1050 (1963), was held to be retroactively applicable in Arsenault v. Massachusetts, 393 U.S. 5, 6, 89 S. Ct. 35, 36 (1968). The right to counsel at probation revocation hearings, announced in Mempa v. Rhay, 389 U.S. 128, 88 S. Ct. 254 (1967), was held to be retroactively applicable in McConnell v. Rhay, 393 U.S. 2, 3-4, 89 S. Ct. 32, 33-34 (1968). The right to counsel on appeal, recognized in Douglas v. California, 372 U.S. 353, 83 S. Ct. 814 (1963), has also been retroactively applied. See McConnell, 393 U.S. at 3, 89 S. Ct. at 33. Finally, Argersinger’s extension of the right to counsel to any prosecution leading to actual imprisonment was deemed retroactively applicable in Berry v. City of Cincinnati, 414 U.S. 29, 29-30, 94 S. Ct. 193, 194 (1973). A score that is perfect packs punch in any analysis. The implication of all those retroactivity decisions dealing with Gideon- related rights is arguably lessened because they were made in the pre-Teague era. The Supreme Court has not decided the retroactivity of any rule expanding Gideon since the Teague regime began in 1989 – there have been no expansions of Gideon since then except for Shelton. Before Teague retroactivity issues in criminal cases 21 were governed by the guidelines set out in Linkletter v. Walker, 381 U.S. 618, 85 S. Ct. 1731 (1965). Under the Linkletter guidelines the Court considered the purposes of the new rule, any reliance on the old rule, and the effect retroactive application of the new rule would have on the administration of justice. Id. at 636, 85 S. Ct. at 1741; Johnson v. State of New Jersey, 384 U.S. 719, 727, 86 S. Ct. 1772, 1777 (1966). Because of the substantial difference in analysis, the pre- Teague decisions applying Gideon-related rights retroactively do not control whether a post-Teague decision announcing a new one is retroactively applicable. But those pre-Teague decisions are hard to ignore. There are statements in them, and in later decisions characterizing them, that stress the importance of the right to counsel in the retroactivity context. Examples of various paeans to the right to counsel abound. “The Supreme Court typically offers the right to counsel . . . as the paradigm of a ‘bedrock procedural element’ falling within the second exception.” Nutter v. White, 39 F.3d 1154, 1157-58 (11th Cir. 1994) (quoting Mackey v. United States, 401 U.S. 667, 693-94, 91 S. Ct. 1160, 1180-81 (1971) (Harlan, J., concurring)). “The right to counsel at the trial . . . on appeal, and at the other ‘critical’ stages of the criminal proceedings have all been made retroactive, since the ‘denial of the right must almost invariably deny a fair trial.’” Arsenault, 393 U.S. at 6, 89 S. Ct. at 36 22 (citations omitted). The right to counsel relates to “the very integrity of the fact- finding process.” McConnell, 393 U.S. at 3, 89 S. Ct. at 33. The Supreme Court has “underscored the narrowness of [Teague’s] second exception by using as a prototype the rule of Gideon.” Spaziano, 36 F.3d at 1043 (citing Teague, 489 U.S. at 313, 109 S. Ct. at 1077); see Beard v. Banks, 542 U.S. ___, No. 02-1603, at 11 (June 24, 2004) (“In providing guidance as to what might fall within this [second Teague] exception, we have repeatedly referred to the rule of Gideon v. Wainwright . . . and only to this rule.”); Saffle, 494 U.S. at 495, 110 S. Ct. at 1264 (“Although the precise contours of this [second Teague] exception may be difficult to discern, we have usually cited Gideon . . . to illustrate the type of rule coming within the exception.”). Significantly, the Supreme Court has never distinguished between different contexts in judging whether an extension of the right to counsel should be made retroactive. It appears that for these purposes at least one right to counsel case is indistinguishable from another. See Arsenault, 393 U.S. at 6, 89 S. Ct. at 36. The Supreme Court has instructed us that the right to representation by counsel is inevitably tied to the accuracy of a conviction. McConnell, 393 U.S. at 3-4, 89 S. Ct. at 33-34. We have said outright that the right to counsel is a bedrock procedural element for Teague purposes. See Nutter, 39 F.3d at 1157. 23 The government does not dispute much, if any, of this, but instead pegs its position to the proposition that Shelton did not really alter our understanding of the right to counsel. Having already held in this opinion that Shelton’s application of the right to counsel in a new context constitutes a “newly recognized” right, it would be odd to hold now that our understanding was not altered by the Shelton decision. Before Shelton this circuit had no rule on whether it violated the Sixth Amendment right to counsel to use an uncounseled conviction that had not resulted in jail time to enhance the sentence imposed for a counseled conviction. Other courts had answered that question in different ways. Now, the split of authority has been healed. Along with every other court in the country, we must follow the Shelton rule. Our own understanding has been altered because it went from a blank slate to one on which is written the Shelton rule. Another consideration in deciding this retroactivity issue is the realization that Teague is a remarkably restrictive doctrine, and its second exception exceedingly narrow. As we have explained: This exception is a narrow one, and its narrowness is consistent with the recognition underlying Teague that retroactivity “seriously undermines the principle of finality which is essential to the operation of our criminal justice system.” Teague, 489 U.S. at 309, 109 S. Ct. at 1074. To fit within the second exception, it is not enough that the rule “preserve the accuracy and fairness of capital sentencing judgments,” Sawyer, 497 U.S. at 24[2], 110 S. Ct. at 2831, or that it “is aimed at improving the accuracy of trial.” Id. The new rule also must be so fundamentally important that its announcement is a 24 “groundbreaking occurrence.” Caspari, 510 U.S. at 396, 114 S. Ct. at 956. It must be a “watershed rule” that “alter[s] our understanding of the bedrock procedural elements essential to the fairness of a proceeding.” Sawyer, 497 U.S. at 241, 110 S. Ct. at 2831 (internal quotation marks omitted). Thus, there is a requirement of “the primacy and centrality of the rule,” Saffle, 494 U.S. at 495, 110 S. Ct. at 1264. Spaziano, 36 F.3d at 1042-43. The Supreme Court has often examined, announced, or proposed new rules of law to see if they fit within the strictures of Teague’s second exception, but it has never found one that does. Beard, No. 02-1603, at 10; see id. at 13-14 (holding that the new rule announced in Mills v. Maryland, 486 U.S. 367, 108 S. Ct. 1860 (1988), and McKoy v. North Carolina, 494 U.S. 433, 110 S. Ct. 1227 (1990), does not fit within the second Teague exception); Schriro v. Summerlin, 542 U.S. ___, No. 03-526, at 10 (June 24, 2004) (same holding regarding the new rule announced in Ring v. Arizona, 536 U.S. 584, 122 S. Ct. 2428 (2002)); O’Dell v. Netherland, 521 U.S. 151, 167, 117 S. Ct. 1969, 1978 (1997) (same holding regarding the rule of Simmons v. South Carolina, 512 U.S. 154, 114 S. Ct. 2187 (1994)); Gray v. Netherland, 518 U.S. 152, 170, 116 S. Ct. 2074, 2085 (1996) (same holding regarding a proposed new rule concerning notice to a defendant of evidence to be used against him in a capital sentencing proceeding); Goeke v. Branch, 514 U.S. 115, 120, 115 S. Ct. 1275, 1278 (1995) (per curiam) (same holding regarding a proposed new rule relating to the fugitive disentitlement doctrine); Caspari v. 25 Bohlen, 510 U.S. 383, 396, 114 S. Ct. 948, 956 (1994) (same holding regarding a proposed new rule that Double Jeopardy Clause applies to noncapital sentencing proceedings); Gilmore v. Taylor, 508 U.S. 333, 345, 113 S. Ct. 2112, 2119 (1993) (same holding regarding a new rule about jury instructions on mitigating mental state in a murder case); Graham v. Collins, 506 U.S. 461, 478, 113 S. Ct. 892, 903 (1993) (same holding regarding a proposed new rule concerning jury questions in Texas’ capital sentencing scheme); Sawyer, 497 U.S. at 244, 110 S. Ct. at 2832 (same holding regarding the new rule of Caldwell v. Mississippi, 472 U.S. 320, 105 S. Ct. 2633 (1985)); Saffle, 494 U.S. at 495, 110 S. Ct. at 1264 (same holding regarding a proposed new rule regarding jury consideration of sympathy in a capital sentencing proceeding); Butler v. McKellar, 494 U.S. 407, 416, 110 S. Ct. 1212, 1218 (1990) (same holding regarding the new rule of Arizona v. Roberson, 486 U.S. 675, 108 S. Ct. 2093 (1988)). We have been slightly more liberal in our application of Teague’s second exception and have on two occasions found a rule to fit within it, but in one of those the Supreme Court later disagreed. In Nutter, 39 F.3d at 1157-58, we held that the new rule of Cage v. Louisiana, 498 U.S. 39, 111 S. Ct. 328 (1990), which prohibited certain jury instruction language that undermined the beyond a reasonable doubt standard, did fall into the second Teague exception because the 26 rule is central to an accurate determination of innocence or guilt, and, like Gideon, implicated a fundamental guarantee of a fair trial. In Clark v. Dugger, 901 F.2d 908, 912-13 (11th Cir. 1990), we held that the rule announced by the Supreme Court in its Caldwell decision fit within the second Teague exception, but the Supreme Court itself later disagreed. See Sawyer, 497 U.S. at 244, 110 S. Ct. at 2832. More often we, like the Supreme Court, have found that new rules cannot squeeze within the narrow confines of the second Teague exception. See Turner v. Crosby, 339 F.3d 1247, 1285 (11th Cir. 2003) (holding that the new rule of Ring v. Arizona does not fall into the second Teague exception); Housel v. Head, 238 F.3d 1289, 1298 (11th Cir. 2001) (same holding regarding proposed new rule that Eighth Amendment forbids jury from weighing unadjudicated crimes in capital sentencing proceeding); Glock, 65 F.3d at 890 (same holding regarding the new rule of Espinosa v. Florida, 505 U.S. 1079, 112 S. Ct. 2926 (1992)); Spaziano, 36 F.3d at 1043 (same holding regarding proposed new rule to bar or curtail the use of a witness’ hypnotically refreshed testimony against a defendant); Collins v. Zant, 892 F.2d 1502, 1512 (11th Cir. 1990) (per curiam) (same holding regarding the new rule of Michigan v. Jackson, 475 U.S. 625, 106 S. Ct. 1404 (1986)). 27 The lesson of all these decisions, we believe, is that the second Teague exception is so tight that very few new rules will ever squeeze through it. The exception that proves the exception, however, is a new Gideon-related rule. Over and over again, the Supreme Court and this Court have held up Gideon as the paradigm case for the second Teague exception. See, e.g., Saffle, 494 U.S. at 495, 110 S. Ct. at 1264 (“[W]e have usually cited Gideon . . . to illustrate the type of rule coming within the exception.”); Nutter, 39 F.3d at 1157-58 (“The Supreme Court typically offers the right to counsel . . . as the paradigm of a ‘bedrock procedural element’ falling within the second exception [of the Teague rule]”). The pre-Teague retroactivity decisions dealing with right to counsel indicate that each extension of that groundbreaking decision has itself been treated with the worshipful respect accorded Gideon itself. The inference we draw is that it is the sheer importance of the right to counsel that is primary in the analysis, not the incremental extension of that right in the case at hand. At the risk of oversimplification, for purposes of the second Teague exception there are new rules, and then there are new Gideon-extension rules. The Shelton decision fits within the second category. IV. 28 Because the government procedurally defaulted its procedural bar defense against the Shelton claim in this case, the result of this appeal turns on whether the Shelton decision is a “right [that] has been newly recognized by the Supreme Court and made retroactively applicable to cases on collateral review” within the meaning of § 2255 ¶ 6(3). The government having waived in this case any argument that the rewind provision applies only when the Supreme Court itself has held the new right to be retroactively applicable, we have made that determination ourselves. In our judgment, the new rule of the Shelton decision does apply retroactively to cases on collateral review. As a result, it is a decision which restarted the one-year clock of § 2255 ¶ 6’s statute of limitations. The district court’s dismissal of Howard’s petition as untimely is REVERSED, and the case REMANDED for further proceedings consistent with this opinion. 29