with whom CLARK, Chief Judge, and REAVLEY, KING, JOHNSON, WILLIAMS, and DUHE, Circuit Judges, join, dissenting:
By adopting the defendant’s position in this case, the majority has misread United States v. Freed, 401 U.S. 601, 91 S.Ct. 1112, 28 L.Ed.2d 356 (1971), by relying upon a concurring opinion therein. I conclude, to the contrary, that the long-accepted interpretation of Freed made by this court in United States v. Vasquez, 476 F.2d 730 (5th Cir.), cert. denied, 414 U.S. 836, 94 S.Ct. 181, 38 L.Ed.2d 72 (1973), more faithfully adheres to Freed’s majority view. Even if this were not the case, the doctrine of stare decisis requires us, in this case of statutory construction, to adhere to our earlier holding absent compelling circumstances not demonstrated by this defendant or today’s majority. Thus, I dissent from what I perceive to be a serious failure by this court to apply the Supreme Court’s pronouncements regarding stare decisis and the legal process.
I.
A close reading of Freed convinces me that the Court purposefully eschewed interpreting the act to require knowledge of a weapon’s characteristics. In comparing the Act with other public-welfare statutes importing criminal penalties, the Court quoted from Justice Frankfurter’s opinion in United States v. Dotterweich, 320 U.S. 277, 284, 64 S.Ct. 134, 138, 88 L.Ed. 48 (1943), upholding a conviction under the Food and Drug Act “ ‘though consciousness of wrong-doing be totally wanting.’ ” 401 U.S. at 609, 91 S.Ct. at 1118. The Court also noted that in United States v. Balint, 258 U.S. 250, 254, 42 S.Ct. 301, 303, 66 L.Ed. 604 (1922), it had reasoned, similarly, that the Narcotic Act’s “ ‘manifest purpose is to require every person dealing in drugs to ascertain at his peril whether that which he sells comes within the inhibition of the statute, and if he sells the inhibited drug in ignorance of its character, to penalize him.’ ” 401 U.S. at 609, 91 S.Ct. at 1118 (emphasis added). The Bal-int Court had surmised that “[C]ongress weighed the possible injustice of subjecting an innocent seller to a penalty against the evil of exposing innocent purchasers to danger from the drug, and concluded that the latter was the result preferably to be avoided.” Freed, id. (quoting Balint, 258 U.S. at 254, 42 S.Ct. at 303). The Freed Court adopted this reasoning verbatim with respect to the Act. Id. 401 U.S. at 609-10, 91 S.Ct. at 1118-19.
Given this language, I must conclude today, as this court did sixteen years ago in Vasquez,1 that there is no requirement of proving knowledge of a weapon’s specific characteristics. Justice Brennan’s concurrence in Freed, relied upon by the majority, is just that, a concurrence: It sets out the law as he believes it should be but neither represents nor clarifies the majority view.2 It is a “strange practice to seize upon passages from such a writing — a mere expression of additional views ... — and found a decision upon them.” United States v. Edwards, 554 F.2d 1331, 1340 (5th Cir.1977) (Gee, J., dissenting), vacated, 577 F.2d 883 (5th Cir.), cert. denied, 439 U.S. 968, 99 S.Ct. 458, 58 L.Ed.2d 427 (1978).
*1257The majority attaches special significance to the fact that Vasquez requires knowledge that a weapon is a firearm in a general sense instead of imposing strict liability. The majority asserts that it is anomalous to use a general notion of “firearm” as the standard where the Act explicitly defines that term in a manner that would exclude many weapons (such as revolvers and hunting rifles) and include many others (such as silencers and landmines) that are not commonly thought of as firearms.
While this argument has some merit, it should not ultimately be persuasive, as it does not properly take account of the implications of Freed: There, the Court concluded that the possession of a hand grenade, given its obvious dangers, was sufficient to dispense with any knowledge requirement that such weapon was unlawfully possessed. In Vasquez, we used the same reasoning to accommodate the case of machine guns: The knowing possession of a gun, given its inherent dangers and the frequency of regulation under state and federal law, was considered sufficient to impose upon defendant the burden of ascertaining both the true character of his weapon and whether it is of the type that must be registered.
The only material difference between Freed and Vasquez is that in the former, no argument was made that that defendant legitimately could have mistaken the hand grenades for something that could be lawfully owned without registration. The Court reasoned, in part, that hand grenades could not reasonably be taken for something else and that awareness of their inherent dangers sufficed to charge the possessor with knowledge of the Act’s provisions. Here and in Vasquez, however, the subject automatic pistols, at least superficially, could have been confused, by a reasonable person, with the lawful, semiautomatic variety. Accordingly, defendant argues that the possession that might be sufficient to charge knowledge in the case of grenades is not satisfied in the case of machine guns by knowledge that a pistol is a firearm in the general sense. I disagree.
I begin by noting that it is possible, although admittedly less likely, to possess a grenade innocently, thinking it is a toy or a replica, in the same sense that is possible to possess a machine gun thinking that it is a semi-automatic pistol. In the former situation, the Court in Freed impliedly determined that the indications, to a possessor, that a grenade is a grenade and is a highly dangerous weapon, the possession of which might be unlawful, overcame any concern for convicting the innocent possessor. In the instant case, the rule in Vasquez comprehends that ownership of semi-automatic weapons, although lawful, carries with it the responsibility to be aware of their characteristics.
Handguns that are not subject to the Act nevertheless are subject to other forms of federal, state, and local regulation.3 Owners of such weapons should be aware generally that there may be laws affecting their rights of ownership. Given this presumed awareness, Vasquez properly follows the logic of Freed in requiring gun owners to know what their weapons are about or face possible penalties for their ignorance. The likelihood of such ignorance in the case of machine gun owners is not enough greater than in the case of grenade owners to call for different rules. And in both circumstances, confusion is improbable.
In an effort to put a wedge between Freed and Vasquez, the majority lamely asserts that the two opinions address different issues. But nothing supports that contention except the majority’s ipse dixit: In each case, the question is whether a defendant who knew that he actually possessed a particular thing must also know that that thing was an item required by federal law to be registered. The difference between Freed and Vasquez is one of degree, not kind. I acknowledge that it is more likely that a person possessing a grenade will understand that he owns something that is subject to regulation; but such likelihood as applied to grenades *1258does not exclude the likelihood that owners of certain other types of dangerous weapons will have the same understanding, or that charging them with that knowledge is somehow unfair.
In any event, if this court in Vasquez misread Freed in the manner suggested by the majority, so have virtually all of our sister circuit courts that have addressed this issue. Those circuits have interpreted Freed similarly and have adopted rules almost identical to that of Vasquez. In United States v. DeBartolo, 482 F.2d 312, 316 (1st Cir.1973), for example, the court held that the defendant’s knowledge that he possessed a gun in the general sense sufficed to meet any mens rea requirement under the Act, even without knowledge that his shotgun had been modified so as to qualify as a firearm under the Act.4 As here, the defendant lawfully could have possessed the weapon but for the modifications bringing it under the Act.
In only requiring proof that the defendant knew the shotgun to be a generic firearm, DeBartolo referred to Freed’s discussion of the special status of regulatory offenses and concluded, as we did in Vasquez, that “[t]he government need not prove that a defendant knows he is dealing with a ... weapon possessing every last characteristic which subjects it to regulation.” DeBartolo, id. at 315.
It is enough to prove he knows that he is dealing with a dangerous device of such type as would alert one to the likelihood of regulation. If he has such knowledge, and if the particular item is in fact regulated, he acts at his peril. A shotgun in today’s society plainly falls into this category. One knowingly participating in the sale of such a lethal instrumentality cannot escape liability by failing to inspect the length of its barrel any more than by failing to inquire whether it is registered. In Freed, it is true, the court assumed that the Government would have to prove that the defendant knew the grenade was a grenade, hence a ‘firearm’. Because all grenades are classified as ‘firearms’, the issue never arose whether knowing possession of a grenade would give rise to a duty to inspect to see if it was the kind of grenade regulated. But we see no difference in rationale between the duty of one possessing a grenade to ascertain if it is registered, and of one, knowingly transferring a shotgun, to ascertain if by reason of its barrel length it must be registered.
Id. at 316-17 (citations omitted).5
As of the time this matter was submitted en banc, only in United States v. Herbert, *1259698 F.2d 981 (9th Cir.), cert. denied, 464 U.S. 821, 104 S.Ct. 87, 78 L.Ed.2d 95 (1983), had a court of appeals fashioned a rule contrary to Vasquez. There, the court created a limited exception to the general rule of not having to prove knowledge of a weapon’s characteristics. The exception extends only to guns that otherwise may be lawfully possessed but have been internally modified so as to become “firearms” under the Act. The Herbert court perceived a meaningful difference between readily observable and purely internal modifications because, in the latter situation, the defendant does not know “ ‘that he is dealing with a dangerous device of such type as would alert one to the likelihood of regulation.’ ” Id. at 986 (quoting DeBartolo, 482 F.2d at 316).
As the majority notes, the Sixth Circuit, departing from its opinions in Decker and Cowper, recently has sided with Herbert in United States v. Williams, 872 F.2d 773 (6th Cir.1989). As for prior Sixth Circuit cases, the opinion mentions but does not attempt to distinguish Decker, see id. at 775 n. 2, and dismisses Cowper as dictum, see id. at 777.
The reasoning in Herbert and Williams is unpersuasive. By the logic of Freed and DeBartolo, possession of a gun constitutes possession of a dangerous weapon that, even if completely lawful, alerts one to the likelihood of regulation. Herbert and Williams, on the other hand, appear completely to excuse the possessor where the modifications to the weapon are purely internal. According to that reasoning, a shotgun that has been cut off so as to be a fraction of an inch too short and hence just barely a “firearm” under the Act would not be subject to proof of knowledge of its characteristics, even though a person possessing it might easily be unaware of the unlawful condition; but a semi-automatic pistol that had been internally modified so as to become a machine gun would require such proof simply because the modification is not superficially discernable. Moreover, to fashion a distinction based upon the external visibility of the modification is to engage in legislative line-drawing more befitting Congress than the courts.
I would decline the argument that separate rules are justified on this basis and conclude that the holdings in Herbert and Williams rely upon a distinction without a difference. The better view, instead, is to adhere to Vasquez and to most other circuits’ reading of the Act rather than to fashion an exception to cover the instant facts.6
II.
The majority suggests that Vasquez improperly dispenses with the venerable common-law tradition of requiring mens rea for felony offenses. I readily acknowledge the well-established principle of mens rea; I note, as well, that the Supreme Court, as a matter of statutory interpretation, often reads a mens rea requirement into a statutory offense where none is provided. E.g., Liparota v. United States, 471 U.S. 419, 425-26, 105 S.Ct. 2084, 2087-88, 85 L.Ed.2d 434 (1985); United States v. United States Gypsum Co., 438 U.S. 422, 438-43, 98 S.Ct. 2864, 2874-77, 57 L.Ed.2d 854 (1978).
However, these considerations are inap-posite to this court’s reconsideration of Vasquez. First of all, as noted above, Vasquez does require mens rea to the extent that a defendant must realize that his or her weapon is a firearm in the general sense. Moreover, the Court in Freed directly addressed whether to read a stricter culpability requirement into the Act when it analogized that case to its earlier decisions in Dotterweich and Balint involving other public-welfare statutes. The certain thrust of the Court’s reasoning is that irrespective of common-law tradition, regulatory offenses where the act in question is *1260unequivocally prohibited do not require the same magnitude of fault in order to justify a felony conviction. Thus, it is the extent of mens rea, and not its existence vel non, which we should determine here, as we did in Vasquez.
Although the majority apparently does not dare to make a due process challenge to the instant conviction, the defendant suggests that convicting an individual for possession of a weapon without having to prove that such person knew what the weapon was may violate due process. I cannot agree. Shortly after deciding Freed, the Court considered the proper mens rea for the offense of unlawfully shipping sulfuric acid. In United States v. International Minerals & Chem. Corp., 402 U.S. 558, 564-65, 91 S.Ct. 1697, 1701-02, 29 L.Ed.2d 178 (1971), the Court referred to Freed and to Balint, involving the illegal sale of narcotics, in rejecting a due process challenge based upon the lack of sufficient mens rea requirements. The Court reasoned as follows:
In Balint the Court was dealing with drugs, in Freed with hand grenades, in this case with sulfuric acid and other dangerous acids. Pencils, dental floss, paper clips may also be regulated. But they may be the type of product which might raise substantial due process questions if Congress did not require ... ‘mens rea ’ as to each ingredient of the offense. But where, as here and as in Balint and Freed, dangerous or deleterious devices or products or obnoxious waste materials are involved, the probability of regulation is so great that anyone who is aware that he is in possession of them or dealing with them must be presumed to be aware of the regulation.
As discussed supra, I conclude, as the court should, that guns, even if they may in some instances be owned lawfully without registration, qualify as sufficiently “dangerous or deleterious devices” so as to require individuals to be aware of their characteristics and the possibility of their inclusion under the Act. Possessing a semi-automatic pistol is surely not the equivalent of possessing pencils, paperclips, or dental floss.
Moreover, aside from the easy conclusion that the rule in Vasquez satisfies due process, this court’s only proper charge here is to interpret Congressional intent in the light of the underlying goals of the Act. I conclude, consistently with Vasquez, that the Act was intended to control traffic in firearms without placing an undue burden upon those wishing to use firearms for lawful purposes.7 As it is the role of Con*1261gress, not the courts, to make that policy determination, it is useful to examine whether the burden which we imposed upon the gun owner in Vasquez is greater than that which Congress reasonably may have intended in effecting its intent.
At the outset, it appears extremely unlikely that anyone could unwittingly possess a machine gun believing it to be a semi-automatic pistol. If a gun is purchased through proper lawful channels, such as from a reputable sporting goods store, the possibility of its being automatic is remote. Such guns presumably become automatic in the vast majority of cases because, as apparently with defendant Anderson, they are modified sometime after the initial lawful purchase; hence, it is perfectly reasonable to assume that the owners are well aware of their characteristics without the need for specific proof. One may reliably infer that in most such instances the owners either modified their weapons themselves or deliberately purchased them that way. And in the instance of a street purchaser of a gun, the suspicious circumstances should more than suffice to put him or her on notice that there may be legal problems with the weapon.
There is always, of course, the scant possibility that a truly innocent purchaser will not realize that a weapon is automatic. This arguably was the case in Williams, where the defendant, licensed to sell semiautomatic rifles, sold automatic weapons which he had never unpacked or examined and claimed that he was unaware of their automatic characteristics. But again, the pertinent question is whether Congress intended to place on the owner a reasonable duty of inquiry; by far the better view is that it did.
An additional concern that Vasquez may subject innocent persons to serious criminal penalties stems from the possibility, noted by the majority, that certain semi-automatic pistols apparently may become automatic by accident as a result of wear and tear on their mechanisms. I do not suggest that Congress intended the draconian result of convicting such law-abiding citizens without some showing of fault beyond the knowledge that they owned a perfectly lawful firearm. However, the answer again is that a minimal burden of inquiry imposed upon the gun owner is a reasonable component of Congress’s policy of limiting firearms traffic.
The results of Congress’s policy choice in fashioning the Act appear to have been salutary and (more importantly for present purposes) devoid of unfair consequences to innocent persons. Even were this not so, it would not be this court’s proper role to alter that legislative policy. I note, nonetheless, that with the possible exception of Williams, I am unaware of any case in the long history of the Act involving the prosecution of a truly “innocent” gun owner. Thus, any possibility that the statute will be misused is indeed remote; at oral argument, defendant’s counsel admitted that he could point to no historical instance of such a prosecution.
Instead, the Act comes into play primarily in combating the illegal arms trade where the players are typically gangsters, soldiers of fortune, or terrorists, and increasingly, in prosecuting inner-city drug dealers whose prestige, and ability to compete with rivals, demand that they carry automatic weapons.8 Thus, the majority’s *1262suggestion “that law-abiding, well-intentioned citizens” could be unfairly prosecuted — despite the dearth of such prosecutions over the long history of the act — is at best hyperbole and at worst unwarranted hysteria.
To the contrary, I fail to see how the innocent gun owner gets swept up in this enforcement scheme. In any event, it stands to reason that only frequently-used guns could become modified through wear on the mechanism; such weapons presumably would be fired by experienced shooters who would know immediately upon next discharging them that they had become fully automatic. Thus, the ultimate burden imposed upon gun owners by Vasquez’s reading of Congressional intent is the duty to test-fire their weapons or have them examined by an expert. It may be true that the typical non-expert cannot easily tell by looking at a pistol whether it has been internally modified, but surely anyone who actually shoots a gun will know, immediately upon firing, that it is automatic.
III.
Having concluded that Vasquez neither raises colorable due process concerns nor poses a realistic chance of imposing criminal liability upon innocent individuals, I now turn briefly to a discussion of the benefits flowing from the no-specific-knowledge rule which the court today rejects. , Its primary virtue is that it appreciably facilitates enforcement of the Act by eliminating the reasonable-doubt-as-to-knowledge loophole that otherwise could be exploited. In the case here of defendant Anderson, the government probably could have obtained a conviction for possession of at least one of the pistols even under a specific-knowledge rule, as the gun was found accompanied by a weapon-modification manual. However, in evaluating the effect of the rule, it is best to look at the typical case rather than at the instant facts.
Where law enforcement officials find only weapons, unaccompanied by corroborating evidence such as the modification manual found here, it may be difficult to prove the possessor’s knowledge of the weapon’s characteristics. The defendant may not testify, or may testify that he or she was unaware that the guns were automatic. The prosecution then would depend upon circumstantial inferences that may be difficult to establish beyond a reasonable doubt. In such cases, the will of Congress to control the possession of crime-facilitating dangerous firearms will be unduly thwarted by virtue of today’s ruling.
In assessing the implications of the majority opinion, one is left with a balancing of two factors: the illusory negative impact of Vasquez upon law-abiding gun owners versus the decreased law-enforcement effectiveness of abandoning this rule. Because I believe that the first factor is negligible, while the second is a real and increasingly penetrating concern, I would hold that the rule which we established sixteen years ago in Vasquez is not only supportable as a matter of law, but as a practical matter has yielded a salutary result consistent with the design of Congress.
IV.
Vasquez has been on the books for sixteen years. The rule established by that case is both useful and fair. At the very least, it represents a constitutional and plausible interpretation of the Act. Accordingly, out of respect for the judicial process that produced Vasquez and its progeny, and as a matter of adherence to the well-established doctrine of stare deci-sis, I would be reluctant to overturn such a rule even if defendant’s arguments against it were stronger.
In my view, the most distressing aspect of today’s decision is its failure to recognize the importance of stare decisis in cases such as this. Yet remarkably, the majority devotes only a few cursory sentences to this maxim. It is clear from that truncated treatment that the majority views stare decisis not as a basic tenet of the legal method but simply as a matter of *1263convenience — or like a flag to be raised high in the sunshine and taken down in the rain.
Only when presented with a compelling constitutional attack or dire practical ramifications do we readily overturn well-established precedent. As Judge Higginbotham of this court recently observed, “Courts must be particularly circumspect in reconsidering decisions interpreting statutes.” Bhandari v. First National Bank of Commerce, 829 F.2d 1343, 1353 (5th Cir.1987) (en banc) (Higginbotham, J., concurring), vacated, — U.S.-, 109 S.Ct. 3207, 106 L.Ed.2d 558 (1989).9 In this case, defendant’s constitutional case based upon due process is dubious. His best arguments that Vasquez was wrongly decided are, firstly, that it is anomalous to impose only a requirement of knowledge that a weapon is a generic firearm under a statute that explicitly defines this term in an unusual manner and, secondly, that the case arguably misinterprets Freed. However, these assertions relate to statutory interpretation and sound in an arena where we should be much less likely to abandon our prior interpretation.
In Vasquez v. Hillery, 474 U.S. 254, 265-66, 106 S.Ct. 617, 624, 88 L.Ed.2d 598 (1986), one of its more recent expressions on stare decisis, the Court set forth some general principles that should have guided this court’s disposition of the instant case. The Court described stare decisis as “the means by which we ensure the law will not merely change erratically, but will develop in a principled and intelligent fashion.” The Court continued at length,
That doctrine permits society to presume that bedrock principles are founded in the law rather than in the proclivities of individuals, and thereby contributes to the integrity of our constitutional system of government, both in appearance and in fact. While stare decisis is not an inexorable command, the careful observer will discern that any detours from the straight path of stare decisis in our past have occurred for articulable reasons, and only when the Court has felt obliged ‘to bring its opinions into agreement with experience and with facts newly ascertained’ [citing Burnet, 285 U.S. at 412, 52 S.Ct. at 449 (Brandéis, J., dissenting) ].
Our history does not impose any rigid formula to constrain the Court in the disposition of cases. Rather, its lesson is that every successful proponent of overruling precedent has borne the heavy burden of persuading the Court that changes in society or in the law dictate *1264that the values served by stare decisis yield in favor of a greater objective.
Id. (citation omitted). The Court concluded that it would not break with a well-established rule of law unless it is “outdated, ill-founded, unworkable, or otherwise vulnerable to serious reconsideration.” Id. 474 U.S. at 266, 106 S.Ct. at 625.
More significantly, the Supreme Court— in an opinion issued subsequently to oral argument herein — has reiterated the compelling role of stare decisis in cases of statutory construction. In Patterson v. McLean Credit Union, — U.S. -, 109 S.Ct. 2363, 105 L.Ed.2d 132 (1989), the Court had directed reargument on the question of whether to overrule Runyon v. McCrary, 427 U.S. 160, 96 S.Ct. 2586, 49 L.Ed.2d 415 (1976), involving the construction of 42 U.S.C. § 1981. Declining to overrule Runyon, the Court noted that “ ‘the doctrine of stare decisis is of fundamental importance to the rule of law.’ ” 109 S.Ct. at 2370 (quoting Welch v. Texas Dep’t of Highways & Pub. Transp., 483 U.S. 468, 494, 107 S.Ct. 2941, 2956-57, 97 L.Ed.2d 389 (1987)). “[SJtare decisis is a basic self-governing principle within the Judicial Branch, which is entrusted with the sensitive and difficult task of fashioning and preserving a jurisprudential system that is not based upon ‘an arbitrary discretion.’ ” Id. (quoting The Federalist No. 78, at 490 (A. Hamilton) (H. Lodge ed. 1988)).
Importantly for the instant case, the Court observed, consistently with Justice Brandeis’s view in Burnet, that
the burden borne by the party advocating the abandonment of an established precedent is greater where the Court is asked to overrule a point of statutory construction. Considerations of stare decisis have special force in the area of statutory interpretation, for here, unlike in the context of constitutional interpretation, the legislative power is implicated, and Congress remains free to alter what we have done.
Id. (citations omitted).
The Court found no “special justification” for overruling Runyon, noting that “[i]n cases where statutory precedents have been overruled, the primary reason for the Court’s shift in position has been the intervening development of the law, through either the growth of judicial doctrine or further action taken by Congress.” Id. The Court reasoned that stare decisis can be overridden (i) where “later law has rendered the decision irreconcilable with competing legal doctrines or policies”; (ii) where there is “inherent confusion created by an unworkable decision”; (iii) where “the decision poses a direct obstacle to the realization of important objectives embodied in other laws”; or (iv) where the precedent has been “ ‘tested by experience [and] has been found to be inconsistent with the sense of justice or with the social welfare’ ” (quoting Runyon, 427 U.S. at 191, 96 S.Ct. at 2604 (Stevens, J., concurring) (quoting B. Cardozo, The Nature of the Judicial Process 149 (1921))). Id. 109 S.Ct. at 2371. Thus, the Court concluded that “whether or not Runyon was correct as an initial matter, there is no special justification for departing here from the rule of stare decisis.” Id. at 2372 n. 1.
We are, of course, less free to break with the past than is the Supreme Court, as we are bound by that Court’s opinions. Subject thereto, it seems that we should adhere generally to the same constraints of stare decisis which the Court acknowledges.10 If *1265we were to apply these principles to the instant case, stare decisis would require that we leave United States v. Vasquez intact. As a preliminary matter, I note that the rule of Vasquez is about as well established as a precedent concerning an infrequently-arising point of statutory interpretation is likely to become. That case was decided in 1973 shortly after, and in reliance upon, the Court’s decision in Freed. Since that time, circuit after circuit has adopted the same or a similar analysis.
Furthermore, despite this development of authority primarily along the lines of Vasquez, the Court has given no indication that any of these cases has either misconstrued Freed or presented other problems of statutory interpretation. The issue certainly has had time to ferment among the courts of appeals. I therefore consider the nearly unanimous movement in the direction of Vasquez, and the Supreme Court’s attendant silence, as an indication that the case was rightly decided and that the law on this point has been settled and should not be disturbed absent the sort of “special justification” described in Patterson. Here defendant does not, and cannot plausibly, contend that either society or the law has in fact changed so as to justify the abandonment of stare decisis in favor of adopting a new rule.11
This is not to say that stare decisis should be viewed as talismanic in every case, and there are circumstances in which we do, and should, reject its application as inappropriate — as, for example, when the established rule has proven to be “outdated, ill-founded, unworkable, or otherwise vulnerable to serious reconsideration.” Vasquez v. Hillery, 474 U.S. at 266, 106 S.Ct. at 625.12 Thus, as “[o]ur law is neither moribund nor muscle-bound[,] [tjhere are justifiable escapes and liberations from the rigidities and inflexibilities of stare de-cisis,” as when rules “ ‘are found after fair trial to be inconsistent ... with an attainment of the ends which law is meant to *1266serve’ ” (quoting B. Cardozo, The Growth of the Law 120 (1924)), or where “ ‘the rule of our circuit has commanded no following in other federal courts of appeals and has been much criticized by the commentators’ ” (quoting Chappell & Co. v. Frankel, 367 F.2d 197, 200-01 (2d Cir.1966)). United States v. Cocke, 399 F.2d 433, 448-49 (5th Cir.1968) (en banc) (Goldberg, J.), cert. denied, 394 U.S. 922, 89 S.Ct. 1187, 22 L.Ed.2d 455 (1969).13
But none of these conditions obtains here. As I have noted, United States v. Vasquez was well-founded upon the principle established by Freed. The rule which it created is by no means outdated. To the contrary, as automatic weapons become more and more prevalent among criminal elements,14 the virtues of the Vasquez rule in potentially helping to fight this problem have become more, not less, apparent over time.
By all indications, the principle of United States v. Vasquez has proven quite workable, as well. To date, it has permitted effective enforcement of the Act without subjecting “innocent” persons to prosecution. Thus, under the Patterson analysis, I certainly cannot surmise that our prior decision has been “tested by experience [and] ... found to be inconsistent with the sense of justice or with the social welfare.” 109 S.Ct. at 2371.
The majority’s evident frustration is that it is wholly unable to identify any adverse experience in the years since United States v. Vasquez was decided. To the contrary, I can only conclude, as the Court did in Vasquez v. Hillery, that “the need for such a rule is as compelling today as it was at its inception.” 474 U.S. at 266, 106 S.Ct. at 625. And we are reminded that “it is better the law should be certain, than that every Judge should speculate upon im-provements....” Sheddon v. Goodrich, 32 Eng.Rep. 441, 447 (Ch. 1803). If the rule is to be changed at this point, it should be for the Supreme Court or Congress to do so.
The court today too lightly dispenses with the principle of stare decisis for the sake of a danger that is wholly speculative. I respectfully dissent.
. References hereinafter to “Vasquez” are to United States v. Vasquez, not Vasquez v. Hillery, 474 U.S. 254, 106 S.Ct. 617, 88 L.Ed.2d 598 (1986), discussed infra as part of the stare deci-sis analysis.
. Accord, United States v. Cowper, 503 F.2d 130, 132 n. 2 (6th Cir.1974) ("The majority opinion [in Freed], however, which had eight votes, is controlling.”), cert. denied, 420 U.S. 930, 95 S.Ct. 1133, 43 L.Ed.2d 403 (1975). In Freed, Justice Brennan specifically stated, "I do not join part II of the opinion; although I reach the same result as the Court on the intent the Government must prove to convict, I do so by another route" 401 U.S. at 610, 91 S.Ct. at 1118-19 (emphasis added) (Brennan, J., concurring).
. See, e.g., 18 U.S.C. §§ 921-928; Tex. Penal Code Ann. §§ 46.01-46.04 (Vernon 1989).
. Shotguns with barrels less than 18 inches in length are covered under the Act. § 5845(a)(1).
. Accord, United States v. Mittleider, 835 F.2d 769, 774 (10th Cir.1987) ("The offenses with which defendant was charged were not specific intent crimes, and the government was not required to prove either that he knew that possession of the weapon was against the law or that registration of the weapon was required.”), cert. denied, — U.S. -, 108 S.Ct. 1279, 99 L.Ed.2d 490 (1988); United States v. Shilling, 826 F.2d 1365, 1368 (4th Cir.1987) (rejecting "appellant’s contention that the Government must prove actual knowledge of a weapon’s characteristics”), cert. denied, — U.S. -, 108 S.Ct. 777, 98 L.Ed.2d 863 (1988); United States v. Taylor, 728 F.2d 864, 867-70 (7th Cir.1984) ("Accordingly, proof that a person actually or constructively possessed the unregistered firearm, in violation of 26 U.S.C. section 5861(d), is sufficient proof that the person ‘knowingly possessed’ the unregistered firearm in question.”); United States v. Gonzalez, 719 F.2d 1516, 1522 (11th Cir.1983) (holding that "the government does not have to prove that the defendant knew that the weapon in his possession was a ‘firearm’ within the meaning of the statute, or that he knew registration was required") (emphasis in original), cert. denied, 465 U.S. 1037, 104 S.Ct. 1312, 79 L.Ed.2d 710 (1984); Morgan v. United States, 564 F.2d 803, 806 (8th Cir.1977) ("In concluding that it need not be proved that Morgan had actual knowledge that the M16 was designed to fire automatically or semiautomatically, we approve a rationale now accepted by several courts of appeal.”); United States v. Ranney, 524 F.2d 830, 832 (7th Cir.1975) (proof that defendant “knew that he possessed a firearm in the general meaning of the term was sufficient"), cert. denied, 424 U.S. 922, 96 S.Ct. 1130, 47 L.Ed.2d 330 (1976); United States v. Cowper, 503 F.2d at 132 (Freed does not require knowledge that firearm is automatic rifle); United States v. Decker, 292 F.2d 89, 90 (6th Cir.) (“It is not necessary for the government to prove that defendant knew that the weapon in his possession was a firearm within the meaning of the statute. Scienter is not involved.”), cert. denied, 368 U.S. 834, 82 S.Ct. 58, 7 L.Ed.2d 36 (1961).
. Except for Williams and today’s majority, all of the courts that subsequently have addressed this issue have either rejected or criticized Herbert. E.g., Mittleider, 835 F.2d at 774 (concluding that "the rationale underlying the court's holding in Herbert is of questionable validity”); Shilling, 826 F.2d at 1368 ("We decline to follow the Herbert decision in this Circuit.”). While this court is not bound to follow the law of other circuits, I think it significant that the weight of such authority so overwhelmingly supports upholding the rule in Vasquez,
. Similarly, in Balint, as I have noted, the Supreme Court concluded that "Congress weighed the possible injustice of subjecting an innocent seller to a penalty against the evil of exposing innocent purchasers to danger from the drug, and concluded that the latter was the result preferably to be avoided.” 258 U.S. at 254, 42 S.Ct. at 303.
The majority concludes that Congress could not possibly have intended the result reached in Vasquez. While there is no legislative history directly on point, this bald assertion is undermined by a review of associated legislative actions indicating that Vasquez is entirely consistent with Congress's express and implicit purposes in its regulation of automatic weapons.
The 1986 revisions to the Gun Control Act were entitled the Firearms Owners’ Protection Act, P.L. 99-308. These amendments, inter alia, added scienter requirements to parts of 18 U.S.C. §§ 922 and 924. Despite its clear general intention to protect gun owners’ interests, Congress did not see fit to add any mens rea component to § 5861(d).
To the contrary, § 104 of the Firearms Owners' Protection Act reveals exactly the opposite intention with respect to machine guns. Subsection (a) created additional penalties for possession of machine guns while committing violent or drug-related crimes. Even more to the point, § 101, by incorporation of the definition appearing in § 5845(b) of title 26 U.S.C., broadened the definition of machine gun in the Gun Control Act to include machine gun parts and conversion kits. Citing the Report of the Attorney General’s Task Force on Violent Crime, the legislative history to this subsection reveals a clear recognition of, and desire to curb, "the easy conversion of semi-automatic weapons into fully automatic weapons.” H.R. No. 495, 99th Cong., 2d Sess., reprinted in 1986 U.S.Code Cong. & Ad.News 1327, 1354.
The Attorney General’s Report in turn refers to 1981 statistics indicating that approximately 20% of illegal machine guns were converted semi-automatic weapons. Recommendation 20, Final Report of the Attorney General’s Task Force on Violent Crime 29, 32 (United States Dep’t of Justice Aug. 17, 1981) (citing Dep’t of the Treasury, Bureau of Alcohol, Tobacco & Firearms, Firearms Case Summary (Washington: U.S. Gov’t Printing Off. 1981)). I cannot help but *1261note that this percentage has likely increased, perhaps markedly, as a result of the tragic and well-documented proliferation, among criminals over the past several years, of easily-convertible "assault rifles.” See infra n. 14.
The above materials are not dispositive of whether Vasquez was correctly decided. However, they more than establish that the rule of that case comports with the attitude expressed by Congress three years ago when it globally reviewed the question of federal firearms regulation with the intention of protecting lawful gun owners, and in so doing specifically considered the issue of the conversion of semi-automatic weapons, with the intention of eradicating this problem. The majority’s suggestion that Congress could not have intended the result reached by Vasquez hence is without support.
. The Congressional purpose in the initial enactment of the firearm registration requirement was "to make it more difficult for the gangster element to obtain certain types of firearms.” Sipes v. United States, 321 F.2d 174, 176 (8th Cir.) (quoting the legislative history) (cited in Freed, 401 U.S. at 607, 91 S.Ct. at 1117), cert. *1262denied, 375 U.S. 913, 84 S.Ct. 208, 11 L.Ed.2d 150 (1963).
. The Supreme Court has long acknowledged a difference in its attitude toward precedent where it is presented with a constitutional question rather than one of statutory interpretation. Justice Brandéis, in his dissent in Burnet v. Coronado Oil & Gas Co., 285 U.S. 393, 405-06, 52 S.Ct. 443, 446-47, 76 L.Ed. 815 (1932) (Brandéis, J., dissenting), surmised that "in most matters it is more important that the applicable rule of law be settled than that it be settled right” (quoted approvingly in, e.g., United States v. Dayton, 592 F.2d 253, 256 n. 1 (5th Cir.1979) (per curiam)) but that “in cases involving the Federal Constitution, where correction through legislative action is practically impossible, this Court has often overruled its prior decisions.” Later, in his celebrated opinion in Erie R.R. v. Tompkins, 304 U.S. 64, 77-78, 58 S.Ct. 817, 822, 82 L.Ed. 1188 (1938), Justice Brandéis remarked, "[i]f only a question of statutory construction were involved, we would not be prepared to abandon a doctrine so widely applied throughout nearly a century. But the unconstitutionality of the course pursued has now been made clear and compels us to do so.”
Justice Brandeis’s dissent in Burnet was the seminal expression of what is now the generally-accepted notion that stare decisis weighs more heavily in statutory than in constitutional adjudication; we have accepted his view repeatedly, e.g., Rios v. Dillman, 499 F.2d 329, 334 n. 8 (5th Cir.1974), although the distinction has been subject to some criticism. See Easter-brook, Stability & Reliability in Judicial Decisions, 73 Cornell L.Rev. 422 (1988) (but acknowledging the prevalence of the constitutional/statutory distinction). But see Levi, An Introduction to Legal Reasoning, 15 U.Chi.L.Rev. 501 (1948) (supporting the predominant rule).
Thus, in general the Supreme Court has long afforded what might be described as an extremely strong presumption of correctness to its decisions concerning statutory interpretation. See Toolson v. New York Yankees, Inc., 346 U.S. 356, 356-57, 74 S.Ct. 78, 79, 98 L.Ed. 64 (1953) (per curiam); United States v. South Buffalo Ry., 333 U.S. 771, 68 S.Ct. 868, 92 L.Ed. 1077 (1948); Helvering v. Griffiths, 318 U.S. 371, 63 S.Ct. 636, 87 L.Ed. 843 (1943); see generally Eskridge, Overruling Statutory Precedents, 76 Geo.L.J. 1361 (1988).
. Another circuit has spoken convincingly of the proper constraints upon an en banc court of appeals:
... The Court en banc is not, of course, bound by prior opinions of panels. We have both the power and the right, in appropriate cases, to overrule panel opinions. This power, however, should be exercised sparingly and with great caution, especially ... when the issue is one of statutory construction, the kind of question on which Congress can easily correct us if it wishes. See, e.g., Cottrell v. Commissioner, 628 F.2d 1127, 1131 (8th Cir.1980) (en banc): ‘The doctrine of stare decisis, weighty in any context, is especially so in matters of statutory construction. For in such cases Congress may cure any error made by the courts. Until it does, the bar and the public are justified in expecting the courts, except in the most egregious cases, neither to depart, nor to give them a grudging application’ (footnote omitted).
Fast v. School Dist. of the City of Ladue, 728 F.2d 1030, 1034 (8th Cir.1984) (en banc).
. Moreover, there is a colorable argument that reversing a well-established point of statutory interpretation that builds upon the bare bones of a statute presents constitutional separation-of-powers problems, as any subsequent change of position has the practical effect of amending the statute, an act that is legislative rather than judicial in nature. See Francis v. Southern Pac. Co., 333 U.S. 445, 450, 68 S.Ct. 611, 613, 92 L.Ed. 798 (1948); Boys Markets, Inc. v. Retail Clerks Union, 398 U.S. 235, 257-58, 90 S.Ct. 1583, 1595-96, 26 L.Ed.2d 199 (1970) (Black, J., dissenting); United Gas Improvement Co. v. Continental Oil Co., 381 U.S. 392, 406, 85 S.Ct. 1517, 1525, 14 L.Ed.2d 466 (1965) (Douglas, J., dissenting); see generally Horack, Congressional Silence: A Tool of Judicial Supremacy, 25 Tex.L.Rev. 247 (1947).
. The majority correctly notes that only a month before deciding Patterson, the Supreme Court overruled one of its prior statutory-construction opinions in Rodriguez de Quijas v. Shearson/American Express, Inc., — U.S. -, 109 S.Ct. 1917, 104 L.Ed.2d 526 (1989). But that case, unlike the case sub judice, amply satisfies Patterson’s "special justification” requirement: The court concluded that the earlier precedent should be overruled "to achieve a uniform interpretation of similar statutory language ... and to correct a seriously erroneous interpretation of statutory language that would undermine congressional policy as expressed in other legis-lation....” Id. 109 S.Ct. at 1922. I note, as well, that Justice Kennedy authored the majority opinions in both Rodriguez de Quijos and Patterson.
I find it nothing short of bizarre that the majority has cited Bhandari v. First Nati Bank of Commerce (Gee, J.) as an example of this court’s overruling of statutory precedent. That case indeed is an example of a recent denigration of stare decisis, but apparently the Supreme Court found it in error: The judgment was vacated in light of Patterson. — U.S. -, 109 S.Ct. 3207, 106 L.Ed.2d 558 (1989).
Hence, it accomplishes little to cite, in fini-tum, cases in which the courts have, or have not, adhered to stare decisis. I readily acknowledge that there are times when the doctrine should not be applied, even in instances of statutory construction — but only when the Patterson "special justification” test is met. In the Supreme Court cases noted by the majority, "special justification” was shown. In Monell v. Department of Social Servs., 436 U.S. 658, 98 S.Ct. 2018, 56 L.Ed.2d 611 (1978), Continental T.V., Inc. v. GTE Sylvania, Inc., 433 U.S. 36, 97 S.Ct. 2549, 53 L.Ed.2d 568 (1977), and Nye v. United States, 313 U.S. 33, 61 S.Ct. 810, 85 L.Ed. 1172 (1941), the decision being overruled was recognized as a departure from past Supreme Court precedent. And in Peyton v. Rowe, 391 U.S. 54, 55, 88 S.Ct. 1549, 1550, 20 L.Ed.2d 426 (1968), the prior decision was found to represent "an indefensible barrier to prompt adjudication of constitutional claims in the federal courts.” But here, neither the defendant nor the majority has pointed to a single "special justification” for overruling the longstanding precedent of Vasquez,
. See Helvering v. Hallock, 309 U.S. 106, 119, 60 S.Ct. 444, 451, 84 L.Ed. 604 (1940) (Frankfurter, J.). The test for overruling precedent has been stated variously. E.g., " ‘[A] justice should consider overturning a prior decision only when the decision is clearly wrong, has significant effects, and would otherwise be difficult to remedy.’” Monaghan, Stare Decisis and Constitutional Adjudication, 88 Colum.L.Rev. 723, 762 (1988) (quoting Wallace, Whose Constitution?, in Still the Law of the Land? 10 (C. Roche ed. 1987)). “If the original reasons for the rule have disappeared or weakened, the rule has been persuasively criticized by judges and commentators, and practical experience suggests that the statutory goals are being undermined by the existing rule and can be better served by a new rule, the precedent should be overruled unless there has been substantial legislative or private reliance on the rule.” Eskridge, supra note 9, at 1364. But see IB J. Moore, J. Lucas & T. Currier, Moore’s Federal Practice ¶ 0.0402[5] at 101-02 (2d ed. 1988): “[Aj rule provided by statute ... is not to be rejected because the passage of time and affairs lead the court to believe that the rule was improvident and should be changed_ [I]n the case of statutory interpretation Blackstone will simply not completely disappear.”
. See, e.g. “Epidemic in Urban Hospitals: Wounds from Assault Rifles,” New York Times, Feb. 21, 1989, at 1.