with whom ROVNER and WILLIAMS, Circuit Judges, join, dissenting from denial of rehearing en banc.
The panel’s opinion in this case addresses and resolves a question of general importance for a recurring issue in collateral *1074attacks on criminal convictions: how do the rules of procedural default operate when a later Supreme Court decision changes the legal landscape faced by a defendant at the time of trial? According to the panel, even if there is no way that a defendant, or more to the point, defense counsel, could have anticipated the later legal ruling, the defendant cannot raise the point in a later collateral attack. I find this conclusion troubling from many perspectives, which I explain briefly below: first, contrary to the panel’s suggestion, it is neither compelled by nor is it even consistent with the Supreme Court’s pair of decisions in Bailey v. United States, 516 U.S. 137, 116 S.Ct. 501, 133 L.Ed.2d 472 (1995), and Bousley v. United States, 523 U.S. 614, 118 S.Ct. 1604, 140 L.Ed.2d 828 (1998); second, as the annual filing statistics from this court and our sister circuits attest, habeas corpus petitions and the rules of procedural default consume enormous resources for the court and are of vital importance to the parties; and third, from a pragmatic standpoint, the panel has imposed an impossible burden on defense counsel.
First and most importantly, the panel has not paid sufficiently close attention to the rules the Supreme Court was following in Bailey and Bousley. Bailey, everyone will recall, resolved a conflict in the circuits over the interpretation of the “use” part of one of the federal firearms statutes, 18 U.S.C. § 924(c). The Court’s decision adopted a narrower interpretation than many courts, including this one, had used, by holding that “use” connoted active employment. The Bailey ruling was followed by a flood of petitions for collateral relief filed by prisoners who had been convicted under the disapproved standard. These were straightforward enough when the petitioner had been convicted after a full trial, and when there was no alternate ground upon which to uphold the conviction, but it was unclear whether Bailey should be applied in cases where the defendant had entered a guilty plea to the offense.
In Bousley, the Court confronted that question. It first held that the retroactivity doctrine of Teague v. Lane, 489 U.S. 288, 109 S.Ct. 1060, 103 L.Ed.2d 334 (1989), was not applicable and did not bar Bousley’s claim, because Teague applies only to procedural rules. 523 U.S. at 620, 118 S.Ct. 1604. Furthermore, the essence of Bousley’s argument was that his guilty plea was not knowing and intelligent (because he pleaded thinking that “use” meant passive association with the gun, and that was not the correct interpretation of the statute). There was nothing “new” about the rule that guilty pleas must be knowing and intelligent, which was another reason why Teague was not helpful. Id. Next, the Court turned to the point that is important for the case now before us: procedural default. Even if Bousley’s claim was not Teague-barred, it was procedurally defaulted because Bousley had not challenged the validity of his plea on direct appeal.
This led the Court to consider whether any exception to the rules of procedural default applied that would permit Bous-ley’s claim to go forward. One possibility is a showing of cause and prejudice, as required by Wainwright v. Sykes, 433 U.S. 72, 87, 97 S.Ct. 2497, 53 L.Ed.2d 594 (1977); the other is a showing that the constitutional error resulted in the conviction of one who is actually innocent. With respect to the “cause and prejudice” avenue, the Court looked at two potential arguments: first, that an argument based on the “true” meaning of § 924(c) was unavailable to Bousley at the time of his plea; and second, that an effort to show that “use” was being treated too broadly would have been futile at the time of the plea. The Court concluded that Bousley *1075could not show “cause” in either of those ways, but it found that he was essentially making a claim of actual innocence, which sufficed to overcome his procedural default. Contrary to the implication in the panel’s opinion, however, the Court did not rule as a matter of law that the first two theories were meritless in all cases. Instead, as the following (somewhat lengthy) passage from the Bousley opinion illustrates, it took a more nuanced view of those arguments:
Petitioner [ie. Bousley] offers two explanations for his default in an attempt to demonstrate cause. First, he argues that “the legal basis for his claim was not reasonably available to counsel” at the time his plea was entered.... This argument is without merit. While we have 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, Reed v. Ross, 468 U.S. 1, 16, 104 S.Ct. 2901, 82 L.Ed.2d 1 (1984), petitioner’s claim does not qualify as such. The argument that it was error for the District Court to misinform petitioner as to the statutory elements of § 924(c)(1) was most surely not a novel one. See Henderson [v. Morgan,] 426 U.S. [637], at 645-646, 96 S.Ct. 2253, 49 L.Ed.2d 108 [(1976)]. Indeed, at the time of petitioner’s plea, the Federal Reporters were replete with cases involving challenges to the notion that “use” is synonymous with mere “possession.” ... Petitioner also contends that his default should be excused because, “before Bailey, any attempt to attack [his] guilty plea would have been futile.” ... This argument, too, is unavailing. As we clearly stated in Engle v. Isaac, 456 U.S. 107, 102 S.Ct. 1558, 71 L.Ed.2d 783 (1982), “futility cannot constitute cause if it means simply that a claim was ‘unacceptable to that particular court at that particular time.’ ” Id. at 130, n. 35, 102 S.Ct. 1558. Therefore, petitioner is unable to establish cause for his default.
Bousley, 523 U.S. at 622-23, 118 S.Ct. 1604 (some citations omitted).
This discussion would have been considerably shorter if the Court had really meant to say that legal unavailability and futility are simply never enough to show “cause” excusing a procedural default. It would have needed only to say so, without all the qualifications that appear in the paragraph quoted above. In the case before it, it was rather easy to find that neither legal unavailability nor futility applied. After all, the Court had taken Bailey to resolve a conflict in the circuits, and as the Court pointed out, the Federal Reporters were “replete” with decisions on which Bousley’s lawyer could have relied.
In order to apply Bousley to our case, we must take a careful look at the criteria the Court established and the reasons why Bousley’s own effort to demonstrate cause failed. Taking the Court at its word and considering the “legal unavailability” argument first, there are several points that it flagged for consideration before a court may conclude that an argument was legally “available.” First, was the proposed rule fairly suggested by precedent? Second, had any court accepted the proposed rule? Third, if the answer to the first two is yes, how widely accepted had the rule become at the time of the guilty plea (or other pertinent time)? With respect to the futility point, the Court drew a similar distinction between arguments that are unacceptable to particular courts, at particular times, and arguments that are more generally unacceptable. That this is what the Court meant becomes even more apparent when one looks at footnote 2 to the Bous-ley opinion, on the same page as the quoted passage, where the Court quotes the passage from Engle that stresses the “availability” and hence non-futility of a *1076claim that “other defense counsel have perceived and litigated.... ” Bousley, 523 U.S. at 623 n. 2, 118 S.Ct. 1604 quoting Engle, 456 U.S. at 134, 102 S.Ct. 1558.
Applying these criteria to the argument based on the Supreme Court’s decision in Apprendi v. New Jersey, 530 U.S. 466, 120 S.Ct. 2348, 147 L.Ed.2d 435 (2000), and its applicability to the drug type and quantity issue in prosecutions under 21 U.S.C. § 841 and related statutes, I believe that Smith makes a compelling argument for legal unavailability. (I do not discuss the futility ground separately, both because it is unnecessary to do so and because as I show below there was nothing “particular” about any given court’s attitude toward Apprendi-like arguments for drug cases.)
Prior to Apprendi, the Federal RepoR-ers were anything but “replete” with cases holding or even suggesting that drug type and quantity were elements of the offense under 21 U.S.C. § 841, and thus that these details had to be charged in the indictment and submitted to the jury under the reasonable doubt standard for its determination. On the contrary, prior to Apprendi, no court of appeals in the nation had held that drug type and quantity were elements of the offense. The first court of appeals to so hold was the Eighth Circuit, in a decision issued July 18, 2000. United States v. Aguayo-Delgado, 220 F.3d 926 (8th Cir.2000). Even after the Supreme Court issued Jones v. United States, 526 U.S. 227, 119 S.Ct. 1215, 143 L.Ed.2d 311 (1999), in March 1999, every circuit that decided this question held, in accordance with long-standing precedent, that drug type and quantity were sentencing factors. See United States v. Jackson, 207 F.3d 910, 920 (7th Cir.2000); United States v. Thomas, 204 F.3d 381, 384 (2d Cir.2000); United States v. Swiney, 203 F.3d 397, 404 n. 5 (6th Cir.2000); United States v. Smith, No. 99-4454, 2000 WL 139250 (4th Cir. Feb. 8, 2000) (unpublished); United States v. Hester, 199 F.3d 1287, 1291-93 (11th Cir.2000); United States v. Jones, 194 F.3d 1178, 1183-86 (10th Cir.1999); United States v. Williams, 194 F.3d 100, 104-07 (D.C.Cir.1999). In the interim between Jones and Apprendi, only the Fifth and the Eighth Circuits suggested that the question was at least open, and both of these circuits concluded that if Jones affected the treatment of drug type and quantity, the effect was not obvious, so a district court that treated them as sentencing factors did not commit plain error. See United States v. Grimaldo, 214 F.3d 967, 974-75 (8th Cir.2000); United States v. Rios-Quintero, 204 F.3d 214, 215 (5th Cir.2000). Even assuming that Chñmaldo and Rios-Quintero adequately placed the treatment of drug type and quantity up for debate, however, those opinions were not issued until the spring and summer of 2000, eight years after the panel’s opinion would have had Smith’s lawyer raise the issue. In my view, the panel has substantially tightened the test the Supreme Court used in Bousley and Ross, and in so doing it has placed an impossible burden of prescience on defense lawyers.
The issue before us can be phrased succinctly as follows: is the test for “cause” the lack of any reasonable basis for making a legal argument that would scrape by Rule 11 scrutiny, or is the test for “cause” the lack of a suggestion in precedent that would support the particular argument? Bousley, as I see it, takes the latter approach. It compels us to read the term “legal precedent” in a way that fairly leaves some room for the legal unavailability argument recognized in Reed v. Ross and the futility .argument discussed in En-gle v. Isaac. The panel’s opinion has ignored these distinctions. It holds instead that counsel should have anticipated Ap-prendi and its further application to drug cases, based solely on the fact that McMillan v. Pennsylvania, 477 U.S. 79, 106 S.Ct. *10772411, 91 L.Ed.2d 67 (1986), had been decided at the time of Smith’s trial. Respectfully, I think McMillan cannot bear this weight. A brief look at McMillan shows that this 1986 decision went no further than to describe the distinction between “sentencing factors” and “elements of an offense.” The Court did not even say that due process would ever require something to be treated as an element of the offense if Congress had labeled it as a sentencing factor. Instead, its actual holding was to reject the conclusion that Pennsylvania had created an element of the offense. In that sense, arguably all the discussion about elements was therefore dicta.
Even if I am wrong about Bousley, we have before us a choice between two possible rules for deciding whether a person has shown “cause” for purposes of a collateral attack. One is the rule the panel has apparently adopted, and the other the one I believe Bousley requires that I have sketched out here. The difference for criminal defendants between- these two rules is of the utmost importance — surely something I would have thought was worth the en banc court’s time and plenary consideration. Whichever rule is adopted will apply to every collateral attack that comes before us under 28 U.S.C. § 2255, not just drug cases where the Apprendi rule is at issue. Thus, even if on remand Smith were to fail because the court decided that Apprendi should not be applied retroactively to drug cases on collateral review, the importance of the issue now before us would remain.
Last, I fear that the panel’s rule will create an administrative nightmare not only for defense counsel trying to represent them clients responsibly, but also for the district courts and this court. After this, defense counsel will have no choice but to file one “kitchen sink” brief after another, raising even the most fanciful defenses that could be imagined based on long-term logical implications from existing precedents. The Supreme Court may never go down most of those paths, but -that will not matter, because otherwise the defendant will find him or herself staring at a procedural default that cannot be overcome for good cause. I have the deepest concern about the consequences of this approach for the courts. Because I believe that this case raises an important question for a major part of our docket, and I disagree with the panel’s interpretation of Bousley, I respectfully dissent from the court’s decision not to rehear this case en banc.