COLE, J., delivered the opinion of the court, in which GILMAN, J., joined. SCHWARZER, District Judge, (pp. 263-66), delivered a separate dissenting opinion.
OPINION
COLE, Circuit Judge.
This case reminds us that names can be deceiving. We must determine whether, under the Racketeer Influenced Corrupt Organizations Act (“RICO”), 18 U.S.C. § 1962, the activities of a Detroit-area street gang known as the Cash Flow Posse (“CFP”) had a substantial effect on the nation’s cash flow. Petitioner Robert Waucaush challenges, via 28 U.S.C. § 2255, his conviction and sentence resulting from his pleading guilty to conspiring to violate RICO. He argues that in light of Congress’s limited authority under the Commerce Clause, the conduct underlying his convictions fell short of RICO’s requirement that the regulated enterprise “affect interstate or foreign commerce.” The district court disagreed and denied his petition; we REVERSE.
I. BACKGROUND
On July 16, 1997, federal prosecutors unveiled an indictment against seven members of the CFP, including Waucaush, charging in relevant part that they violated and conspired to violate RICO. Said statute prohibits “any person employed by or associated with any enterprise engaged in, or activities of which affect, interstate or foreign commerce, to conduct or participate, directly or indirectly, in the conduct of such enterprise’s affairs through a pattern of racketeering activity[.]” The indictment alleged that, to protect their turf, Waucaush and his colleagues murdered, conspired to murder, and (during less successful outings) assaulted, with intent to murder, members of two rival gangs that sought to expand their operations in Detroit. On April 16, 1998, Waucaush moved to dismiss the indictment. He argued that, within the meaning of the statute and the Constitution, those acts did not affect interstate commerce. The district court denied the motion five days later, and, on May 7, 1998, Waucaush pled guilty to RICO conspiracy. Waucaush later had second thoughts. On June 9, 1998, acting as his own lawyer, he moved to withdraw his plea, claiming that it was unknowing and involuntary. Disagreeing, the district [254]*254court denied his motion, and on March 9, 1999, sentenced Waucaush to prison for life. With new counsel, Waucaush appealed, and in an unpublished opinion, see United States v. Waucaush, 2000 WL 1478361 (6th Cir. Sept.27, 2000), we affirmed.
In May 2000, the Supreme Court decided two cases (further discussed below) that articulated additional restrictions on the scope of the Commerce Clause. See United States v. Morrison, 529 U.S. 598, 120 S.Ct. 1740, 146 L.Ed.2d 658 (2000); Jones v. United States, 529 U.S. 848, 120 S.Ct. 1904, 146 L.Ed.2d 902 (2000). Following these decisions, charges against one of Waucaush’s co-defendants, whom the district court had yet to sentence, were dismissed. But Waucaush-who on September 27, 2001, pursuant to 28 U.S.C. § 2255, collaterally challenged his conviction, relying on the same commerce arguments as did his co-defendant — did not fare as well. On December 9, 2002, the district court denied his petition. Receiving a certificate of appealability from the district court on all of his claims, Waucaush timely appealed.
II. ANALYSIS
Waucaush argues that his plea was unintelligent, and thus void, due to his and the district court’s misunderstanding of the reach of RICO’s commerce element. At the time of his plea, the scope of Congress’s commerce power was controlled by United States v. Lopez, 514 U.S. 549, 115 S.Ct. 1624, 131 L.Ed.2d 626 (1995). Following his plea, the Supreme Court, in Morrison and Jones, further restricted the reach of the Commerce Clause. And in Bousley v. United States, 523 U.S. 614, 620, 118 S.Ct. 1604, 140 L.Ed.2d 828 (1998), the Court held that a constitutional challenge to a guilty plea may invoke subsequent “decisions of [the Supreme Court] holding that a substantive federal criminal statute does not reach certain conduct.”
A. Procedural Default
Waucaush, however, did not challenge the intelligence of his plea on direct appeal, and the Court held in Bousley that “even the voluntariness and intelligence of a guilty plea can be attacked on collateral review only if first challenged on direct review.” Id. at 621, 118 S.Ct. 1604. Although Waucaush would normally have lost his chance to challenge the plea, Bousley detailed two circumstances that would excuse a defendant’s failure to raise such a claim on direct appeal: (1) there was both “cause” for the default and “prejudice” that would result from failing to consider the challenge; or (2) the defendant was “actually innocent” of the crime to which he pled. Id. at 622, 118 S.Ct. 1604.
We first consider the exception for actual innocence. “To establish actual innocence, [Waucaush] must demonstrate that in light of all the evidence, it is more likely than not that no reasonable juror would have convicted him.” Id. (internal quotations omitted). In other words, we must look at all the evidence in the record, and determine whether — as a matter of law — the Government could establish that Waucaush violated the RICO statute.
Actual innocence does not mean that Waucaush must be innocent of all bad deeds. The question before us is whether Waucaush is actually innocent of violating RICO. Put another way, the inquiry is whether the record contains evidence that the CFP, the enterprise in question, affected commerce within the meaning of RICO. In rebutting the claim of actual innocence, “the Government [is] permitted to present any admissible evidence of [Waucaush’s] guilt even if that evidence was not presented during [Waucaush’s] plea colloquy and [255]*255would not normally have been offered before [the Supreme Court’s decisions in Morrison and Jones.]” Id. at 624, 118 S.Ct. 1604. We look not just at the facts to which Waucaush admitted when he pled guilty, but also at any other evidence of his guilt that the Government has marshaled. Waucaush is actually innocent, and therefore entitled to bring the otherwise defaulted challenge to his plea, only if the entire record before us fails to demonstrate that he violated RICO.
Contrary to the positions of the Government and the district court, Waucaush may be actually innocent even though he admitted as part of his plea that his activities “affected interstate commerce.” To illustrate: imagine that Waucaush had admitted to stealing apples from the Post Office, was advised by his counsel and the court that apples were vegetables, and pled guilty to “stealing vegetables from a federal building.” If the Supreme Court later held that, as a matter of law, apples were not vegetables, Waucaush would be actually innocent of “stealing vegetables.” Just as Waucaush’s misinformed admission of a legal conclusion would not have turned apples into vegetables, his guilty plea in today’s case could not have created an effect on commerce that the law did not otherwise recognize.
At the core, Waucaush argues that he is actually innocent of violating RICO because the statute reaches only enterprises “engaged in, or the activities of which affect, interstate or foreign commerce.” 18 U.S.C. § 1962(c). The Government does not contend that the CFP was itself “engaged in” interstate commerce. See United States v. Robertson, 514 U.S. 669, 115 S.Ct. 1732, 131 L.Ed.2d 714 (1995) (per curiam). It argues only that the CFP’s activities “affeet[ed]” interstate commerce. Because we should avoid interpreting a statute to prohibit conduct which Congress may not constitutionally regulate, RICO’s meaning of “affect[ing] interstate or foreign commerce” cannot exceed the bounds of the Commerce Clause. See Jones 529 U.S. at 852, 120 S.Ct. 1904 (interpreting commerce element of federal arson statute to be coextensive with Commerce Clause). This means that we may draw on cases that interpret (in light of the Commerce Clause) the commerce element of statutes other than RICO, as well as cases that interpret the Commerce Clause directly.
RICO regulates enterprises, not people. Although RICO “does not require the violent acts themselves to have any connection to interstate commerce other than that they were committed for the purpose of establishing or maintaining a position within the enterprise,” United States v. Crenshaw, 359 F.3d 977, 984 (8th Cir.2004), the predicate acts must still further the goals of an enterprise that itself affects commerce. The indictment in our case defines the enterprise as Waucaush and his colleagues, who banded together to form a street gang in Detroit. Accordingly, when we determine whether the enterprise affected interstate commerce, we look to the activities of the CFP.
The parties disagree whether the CFP’s effect on commerce must be substantial, or whether a minimal effect will suffice. The Government contends that it need show only that the CFP’s activities had a minimal effect on commerce. It relies on United States v. Riddle, 249 F.3d 529 (6th Cir.2001), which stated that “a de minimis connection suffices for a RICO enterprise that ‘affects interstate commerce.’ ” Id. at 537. But a minimal connection sufficed in Riddle only because the enterprise itself had engaged in economic activity — it operated an illegal gambling business, extorted money, and fenced stolen merchandise. Id. at 537. See also, e.g., Crenshaw, 359 F.3d at 986 (upholding RICO conviction [256]*256when intrastate acts of violence furthered economic enterprise); United States v. Espinoza, 52 Fed.Appx. 846, 848-49 (7th Cir.2002) (same); United States v. Feliciano, 223 F.3d 102, 118 (2d Cir.2000) (same); United States v. Miller, 116 F.3d 641, 674 (2d Cir.1997) (same). As the Ninth Circuit put it, in upholding a RICO conviction predicated on only a de minimis effect on commerce, “the heart of [the defendant’s] crimes, drug trafficking and extortion, are quintessential illegal economic activities.” United States v. Shryock, 342 F.3d 948, 984 n. 6 (9th Cir.2003).
The problem with the Government’s reliance on Riddle and its cousins is that unlike those cases, there is no evidence in our ease that the CFP was involved in any sort of economic enterprise. The Supreme Court in Morrison “reject[ed] the argument that Congress may regulate noneconomic, violent criminal conduct based solely on that conduct’s aggregate effect on interstate commerce.” Morrison, 529 U.S. at 617-18, 120 S.Ct. 1740. Although one CFP member had previously been arrested for trafficking drugs, an activity that is economic, United States v. Tucker, 90 F.3d 1135, 1140 (6th Cir.1996), the Government admits that these drug charges were unrelated to the activities of the CFP. Nor has the Government produced evidence indicating that CFP did anything like peddle cigarettes, see United States v. Abdullah, 162 F.3d 897, 901 (6th Cir.1998) (upholding federal authority to prohibit intrastate trafficking of untaxed cigarettes); instigate credit fraud, see United States v. Valenzeno, 123 F.3d 365, 367-68 (6th Cir.1997) (same for fraudulently obtaining consumer credit); organize gambling, see United States v. Wall, 92 F.3d 1444, 1447-52 (6th Cir.1996) (same for operating gambling business), or anything else economic. All that is left is violence qua violence — which the Supreme Court in Morrison, 529 U.S. at 613, 120 S.Ct. 1740, plainly classified as conduct of the noneconomic strain.
Consequently, in our case, where the enterprise itself did not engage in economic activity, a minimal effect on commerce will not do. More significant interstate commercial ripples might have arisen, for instance, had the CFP attacked individuals or organizations who were conducting or assisting interstate business. See United States v. Laton, 352 F.3d 286, 301-02 (6th Cir.2003) (upholding federal authority to prohibit arson of fire station that both puts out fires besetting businesses and affects local insurance rates); United States v. Dupree, 323 F.3d 480, 485 (6th Cir.2003) (upholding federal authority to prohibit robbery of an “armored car messenger, engaged in the performance of his duties, inside a department store on his regular route .... [who] had just taken possession of $130,014.03 in cash and checks”); Norton v. Ashcroft, 298 F.3d 547, 557 (6th Cir.2002) (upholding federal authority to prohibit physical obstruction and destruction of a business); United States v. Smith, 182 F.3d 452, 456 (6th Cir.1999) (upholding federal authority to prohibit robbery of stores that sold substantial amounts of products shipped from out-of-state). Along these lines, the Government argues that the CFP’s intrastate acts of violence substantially affected commerce because the murder of rival gang members prevented them from selling drugs. Cf. United States v. Rodriguez, 360 F.3d 949, 953 (9th Cir.2004) (upholding federal authority under the Hobbs Act to punish “defendants who agreed to rob cocaine from the stash house of narcotics traffickers”).
But as evidence that the victims sold narcotics, the Government offers only a decision from an intermediate appellate court in Illinois. See People v. Jamesson, [257]*257329 Ill.App.3d 446, 263 Ill.Dec. 736, 768 N.E.2d 817 (Ill.App.2002). Bousley, however, stressed that the Government’s evidence refuting actual innocence must be admissible. Bousley, 523 U.S. at 624, 118 S.Ct. 1604. An opinion’s “Facts” section plainly is not. In any event, the opinion which the Government classifies as “evidence” tells us only that the Illinois chapter of one of the CFP’s targeted gangs “ha[s] been [involved] in numerous violent incidents involving narcotics, batteries, aggravated batteries, assaults, and numerous other criminal activities.” Jamesson, 263 Ill.Dec. 736, 768 N.E.2d at 821. That the Detroit-area victims belonged to a gang whose affiliates in Illinois sold an unknown quantity of drugs with an unknown frequency at an unknown point in time tells us nothing about whether and to what extent drugs were sold by the Detroit gang members targeted by the CFP.
This is a problem, given the Government’s obligation to show that the CFP’s effect on commerce was substantial. In Jones, the Supreme Court held that, in light of the Commerce Clause, the incineration of a private residence did not affect interstate commerce within the meaning of the federal arson statute. Jones, 529 U.S. at 850, 120 S.Ct. 1904. Presumably, the owners of the torched home held jobs and bought interstate goods as part of their day-to-day lives, but the Court deemed these attenuated effects on commerce insufficient. Likewise, in Wang, we found that no substantial effect on interstate commerce resulted when the defendant “robbed private citizens in a private residence of approximately $4,200, a mere $1,200 of which belonged to a restaurant doing business in interstate commerce.” Wang, 222 F.3d at 240. See also United States v. Corp. 236 F.3d 325, 332 (6th Cir.2001) (federal government may not regulate mere possession of homemade child pornography). Even if we assume that some of the people that the CFP killed were drug-dealers, we have no evidence that they were dealing drugs or carrying drug money when they were killed, or that their deaths significantly disrupted the interstate market for drugs. It is certainly conceivable that the CFP’s rivals sold sizable quantities of drugs, or that their deaths affected the drug trade. But on this question, the record is silent.
The Government’s final argument is that, as the district court found, the CFP “eventually became associated with a national gang.” Neither the Government nor the district court fleshed out the interstate commerce implications of this fact, but the argument presumably would be that by associating with these national gangs, the CFP would have affected interstate commerce through correspondence, travel and the like. As with its bedfellows, this line of reasoning — as applied to this particular case, at least — fails to unearth effects on interstate commerce that are more than minimal. The only evidence of any interstate activity by the CFP is that in 1996, some of its members talked over gang business while in Mexico City. Matched up alongside the ten-year period covered by the indictment, this lone instance of crossing state lines is a needle in a haystack.
Indeed, most individuals and organizations cannot help but buy products that traveled in interstate commerce, or occasionally talk to colleagues in, or travel to, other states for some reason or another. If we were to label these occasional acts of interstate commerce as “substantial,” federal authority under the Commerce Clause would be virtually limitless. “Allowing the government to meet the interstate commerce requirement [in a federal criminal prosecution] through only a nominal showing of a connection to interstate commerce would do as much to ‘completely obliterate’ the distinction between national and local [258]*258authority as if no jurisdictional requirement existed at all.” United States v. Odom, 252 F.3d 1289, 1296 (11th Cir.2001).
At the end of the day, we are left with an enterprise whose activity was intrastate, noneconomic, and without substantial effects on interstate commerce. The CFP’s violent enterprise surely affected interstate commerce in some way — a corpse cannot shop, after all. But we may not “follow the but-for causal chain from the initial occurrence of violent crime (the suppression of which has always been the prime object of the States’ police power) to every attenuated effect upon interstate commerce.” Morrison, 529 U.S. at 615, 120 S.Ct. 1740. Because a reasonable jury could not conclude that Waucaush’s enterprise, the CFP, affected interstate commerce, Waucaush is actually innocent of violating RICO. His actual innocence excuses his failure to challenge his plea on direct appeal, such that we may consider the challenge now.
B. Intelligence of the Plea
We are asked to consider the intelligence of Waucaush’s admission during the plea colloquy that his activities “affected interstate commerce.” According to Bous-ley, a plea is constitutionally unintelligent if “the record reveals that neither [Waucaush], nor his counsel, nor the court correctly understood the essential elements of the crime with which he was charged.” Id. at 618-19, 118 S.Ct. 1604. Waucaush contends that although he admitted that his conduct “affected interstate commerce,” he, along with his lawyer, the prosecutor, and the district court, was mistaken about the legal significance of that element.
When he pled guilty, Waucaush believed, and in fact was told by the district court, that a purely intrastate act of violence that had only minimal, indirect effects on interstate commerce could nonetheless satisfy — as a matter of law — the “affect[ed] interstate commerce” element of RICO. But as we -explained above, Lopez suggested, and Morrison and Jones later confirmed, that this understanding of the statute was legally erroneous: the effect on commerce caused by the CFP’s acts of violence were, as a matter of law, insufficiently “substantial” to establish a violation of RICO. And because he had an incorrect understanding of the reach of RICO’s requirement that the enterprise “affect interstate commerce,” Waucaush pled guilty to conduct which was simply not a federal crime. This type of misunderstanding — a misconception about the statute’s legal scope that results in the defendant pleading guilty to conduct which was not a crime- — typifies an unintelligent guilty plea.
The dissent argues that when Waucaush pled guilty, “he understood that to convict him the government would be required to prove that the Cash Flow Posse’s activities substantially affected interstate commerce.” (Dissent at 23) Yet the requirement that a guilty plea be intelligent would evaporate if intelligence is defined only as the ability to articulate the governing legal rule — however deprived of context that rule may be. For this reason, the Supreme Court has admonished that a guilty plea “cannot be truly voluntary unless the defendant possesses an understanding of the law in relation to the facts." Boykin v. Alabama, 395 U.S. 238, 243 n. 5, 89 S.Ct. 1709, 23 L.Ed.2d 274 (1969) (citations omitted) (emphasis added). In the similar context of qualified immunity, the Supreme Court has explained that when measuring whether an actor understands the legal landscape, the formal ability to recite the governing standard is no substitute for a meaningful understanding of the substance of that standard. For instance, [259]*259“the right to due process of law is quite clearly established by the Due Process Clause, and thus there is a sense in which any action that violates that Clause (no matter how unclear it may be that the particular action is a violation) violates a clearly established right. Much the same could be said of any other constitutional or statutory violation.” Anderson v. Creighton, 483 U.S. 635, 639, 107 S.Ct. 3034, 97 L.Ed.2d 523 (1987). See also Asher v. Baxter Int’l Inc. 377 F.3d 727 (7th Cir.2004) (Easterbrook, J.) (“The fundamental problem is that the statutory requirement of ‘meaningful cautionary statements’ is not itself meaningful.”). That Waucaush knew generally that RICO required a “substantial effect on interstate commerce” sheds little light on whether he meaningfully understood the governing law.
If there is any doubt that Waucaush’s plea was unintelligent, we need look no further than the history of this very case— in which the district court affirmatively ruled that the conduct alleged (and to which Waucaush later admitted) satisfied RICO’s “affecting commerce” element. Although the dissent contends that there is no support in the record that the district court misinformed Waucaush about the scope of RICO’s “affecting commerce” element, the dissent’s conclusion is at odds with much of the case’s procedural history. Waucaush and his co-defendants initially relied on Lopez in moving to dismiss their indictments on the grounds that Congress lacked the authority to regulate their conduct under the Commerce Clause. The district court rejected their motions, holding that Congress could indeed regulate their conduct because [the facts to which Waucaush would later admit] “suffice to demonstrate an interstate commerce nexus sufficient to support the indictment.” In essence, Waucaush asked the district court “Does the conduct to which I am admitting satisfy RICO’s ‘affect[ing] commerce’ element?” and the district court answered “Yes!”
Having already once told Waucaush that the conduct to which he would be admitting would satisfy RICO’s “affecting commerce” element, the district court then accepted Waucaush’s plea of guilty. In so doing, the district court was again required to determine that the conduct to which Waucaush was admitting had satisfied each and every element of RICO, for “[w]hen the district court accepted [the defendant’s] plea, it had a duty ... to ensure that the plea was both voluntary ... and supported by a sufficient factual basis.” In re Hanserd, 123 F.3d 922, 927 (6th Cir.1997) (emphasis added). See also Fed.R.Crim.P. 11(f) (“Notwithstanding the acceptance of a plea of guilty, the court should not enter a judgment upon such plea without making such inquiry as shall satisfy it that there is a factual basis for the plea.”). Of course, there cannot be a sufficient factual basis for the plea unless the facts to which the defendant is admitting satisfy each and every element of the statute — including, in this case, the statute’s requirement that the enterprise “affected] interstate commerce.” We now know — and even the dissent agrees — that the facts did not satisfy this element, and that the district court was simply mistaken. In concluding that Waucaush’s plea of guilty was intelligent, the dissent assumes that Waucaush had insight into the law that exceeded even the very district court that had convicted him.
Nor does the dissent adequately answer the obvious question raised by its position: If it was so clear to Waucaush that the facts to which he was admitting did not constitute a federal crime, why did he plead guilty? Relying on Brady v. United States, 397 U.S. 742, 757, 90 S.Ct. 1463, 25 L.Ed.2d 747 (1970), the dissent suggests [260]*260that Waucaush merely misjudged the strength of the Government’s case, and that “[a] voluntary plea of guilty does not become vulnerable because later developments indicate that the plea rested on an insufficient factual basis.” (Dissent at 266 n. 2) Yet Brady says no such thing. There, the defendant pled guilty to avoid facing the death penalty, which subsequent decisions made clear would have been unavailable even if the defendant had gone to trial. The Court held that his plea was intelligent, however, because although he had misjudged the potential sentence he could have received if he was convicted at trial, he did not misunderstand the substantive law underlying his offense. See Brady, 397 U.S. at 756-58, 90 S.Ct. 1463. Unlike the defendant in Brady, who misjudged only the collateral consequences of his plea, Waucaush mistakenly believed that the conduct to which he was admitting satisfied each element of the statute under which he was charged.
Nor does Brady’s language require a different result. The dissent quotes from a passage in Brady, 397 U.S. at 757, 90 S.Ct. 1463, in which the Supreme Court explained that “[w]e find no requirement in the Constitution that a defendant must be permitted to disown his solemn admissions in open court that he committed the act with which he is charged simply because it later develops that the State would have had a weaker case than the defendant thought.” Thus, if Waucaush had admitted in open court that the CFP ran a commercial gambling enterprise, he would not be able to rescind this admission if it later turned out that the Government did not have any evidence of this. Cf. Parker v. North Carolina, 397 U.S. 790, 796-98, 90 S.Ct. 1458, 25 L.Ed.2d 785 (1970) (rejecting a challenge to a guilty plea based because defendant’s misjudgment about the admissibility of his confession to a burglary went only to the strength of the factual evidence that the government could produce). What would be key in this situation — the type of situation identified by Brady — is that the facts to which Waucaush admitted at the time would have satisfied RICO’s “affecting commerce element.”
The situation in our case is exactly the opposite, and has nothing to do with the rule articulated in Brady. Our determination that Waucuash’s plea was unintelligent turns neither on newly discovered evidence nor from a conclusion that the existing evidence is no longer credible. The facts to which Waucaush admits now are the same facts to which he admitted when he pled guilty. As even the dissent acknowledges, those facts alone — the only ■ facts to which Waucaush has ever admitted — make him actually innocent of violating RICO. The only change in Waucaush’s understanding has come from the legal significance of those facts: his admission that his conduct violated RICO, and the district court’s acceptance of his plea based on that conduct, was simply incorrect as a matter of law.
Thus, our case mirrors Hanserd. There, the defendant pled guilty to using a firearm in furtherance of a drug-trafficking offense. Following the Supreme Court’s decision in Bailey v. United States, 516 U.S. 137, 116 S.Ct. 501, 133 L.Ed.2d 472 (1995), a decision that limited the type of conduct that constitutes “use,” it became clear that the factual conduct to which the defendant admitted did not constitute a crime within the meaning of the statute. We held that the defendant’s plea was involuntary because the subsequent decision in Bailey “ma[de] it clear that the court, counsel, and accused were all operating under what we now know was a too-inclusive view of § 924(c)’s reach.” Hanserd, 123 F.3d at 927. The same mistake infected Waucaush’s plea: judge, counsel, [261]*261and accused were all operating under a too-inclusive view of RICO’s reach. Like in our case, the facts to which the defendant admitted had not changed, but the defendant’s understanding of the legal significance of those facts had.
Attempting to distinguish Bousley and Hanserd, the dissent argues that “in Hanserd, a post-plea decision determined for the first time that the government would have to provide additional proof (that the defendant used the gun ‘during and in relation’ to the drug offense) to obtain a conviction, a fact not known to Hanserd when he pled.” (Dissent at 265-66) The dissent’s distinction is misplaced along two dimensions. First, a plea can be unintelligent even if the law was clear at the time: the question is not whether the law was clear, but whether the defendant was aware of the law (clear or otherwise). If a defendant who stole a pencil from Wal-Mart pled guilty to murder solely on that basis, surely the dissent would not argue that the defendant simply misjudged the strength of the Government’s case. See id. (“The overly broad interpretation of the scope of § 924(c) was as wrong before Bailey as it is now. Bailey did not change the statute’s meaning; it clarified what § 9%U(c) has always meant since its enactment.” (emphasis added)). Indeed, Han-serd itself rested its holding on the fact that “[o]n the record before us, ... the carry prong cannot support either of the [defendant’s] convictions.” Id. According to Hanserd, “[t]hat Bailey had yet to be decided when Hanserd entered his plea serve[d] only to strengthen this conclusion.” Id. (emphasis added) The court in Hanserd thus made explicit that its holding would have been the same even if Bailey had been decided at the time of the defendant’s plea; that Bailey came later was mere gravy.
Second, even if we did not mean what we held in Hanserd, and the dissent’s old/new dichotomy were dispositive, the dissent is incorrect that Morrison and Jones broke no new ground. Lopez certainly sparked a new era in Commerce Clause jurisprudence, explicitly holding for the first time that the Commerce Clause extends to purely intrastate activity only when that activity “substantially affects” interstate commerce. The Government had argued that the statute at issue, the Gun Free School Zone Act, which prohibited the possession of a firearm within 1,000 yards of a school, substantially affected interstate commerce because guns carried the risk of violence, which might interfere with students’ education, which would render them less equipped for the workforce. See Lopez, 514 U.S. at 564, 115 S.Ct. 1624. In rejecting that argument, the Court refused to “pile inference upon inference in a manner that would bid fair to convert congressional authority under the Commerce Clause to a general police power of the sort retained by the States.” Id. at 567, 115 S.Ct. 1624. Yet the holding of Lopez was, at the time, constrained to its facts: it established no categorical prohibition on the federal regulation of the aggregate commercial effects of noncommercial activity, such as violence.
Contemporaneous scholarship confirmed this limited understanding of Lopez. Writing after Lopez but before Morrison, Professor Tribe questioned whether “future applications of Lopez will turn entirely, or even predominately, on deciding whether a regulated activity is sufficiently ‘commercial’ to qualify for the ‘substantial effects’ test and the aggregation principle. The Lopez Court did not expressly hold that only economic or commercial activities could be regulated by Congress whenever they meet these impact tests.” 1 Laurence H. Tribe, American Constitutional Law 823 (3d ed.2000) (emphasis added). [262]*262After concluding that the holding of Lopez did not itself compel what we now know to be the holding of Morrison — -and the basis for our invalidation of Waucaush’s conviction — Professor Tribe noted that “Lopez’s larger significance must therefore remain, at least for now, a matter of speculation.” Id. at 832. See also, e.g., John Copeland Nagle, The Commerce Clause Meets the Delhi Sands Flower-Loving Fly, 97 Mich.L.Rev. 174, 176 (1998) (“Whether Lopez marks a dramatic shift in Commerce Clause jurisprudence or is instead destined to be a ‘but see’ citation remains to be seen.”).
What is important, of course, is not so much that Jones and Morrison broke new ground, but that the new ground they broke is the basis upon which we hold that the CFP’s conduct did not substantially affect interstate commerce. Lopez did not hold that all noneconomic intrastate activity was beyond the pale of the Commerce Clause, nor did it hold categorically that purely intrastate violence was off-limits to regulation by the federal government. Rather, the statute invalidated by Lopez was unsustainable because it regulated the possession of a device that might lead to violence, in a setting where that violence might interfere with learning, with the result that diminished learning might — at some point well into the future — inhibit interstate commerce. Before Jones and Morrison, such a speculative chain of causation was easily distinguishable from the facts in our case, in which the defendants were alleged actually to have murdered several individuals on the city streets (and thereby preventing the victims from working, shopping, or doing anything else commercial).
It was Jones and Morrison that definitively prevented Congress from regulating the CFP’s conduct. In Jones, 529 U.S. at 857, 120 S.Ct. 1904, the Supreme Court held that in light of the Commerce Clause, the federal arson statute could not be construed to encompass the arson of a private, noncommercial dwelling. Because losing one’s house has obvious commercial effects, the refusal to distinguish Lopez on that basis indicated that Lopez would be read broadly. Prior to Jones, this was by no means assured, as evidenced by the contrary holding of the unanimous lower court decision which Jones reversed. See United States v. Jones, 178 F.3d 479, 480 (7th Cir.1999) (Easterbrook, J.) (“Lopez says that the power of Congress is limited to activities that substantially affect commerce, while proof of a small effect will satisfy the statute.” (internal citation omitted)). And in Morrison, the Court held that Congress lacked authority to provide a federal cause of action to victims of gender-motivated violence. The evidence in Morrison indicated that the economic impact of actual gender-motivated violence was far more direct and apparent than that of mere possession of a weapon near a school. That the former did not “substantially affect interstate commerce” revealed again that Lopez could not be constrained to its facts.
Most importantly, Morrison debuted the categorical rule that directly controls the case before us: “We accordingly reject the argument that Congress may regulate noneconomic, violent, criminal conduct based solely on that conduct’s aggregate effect on interstate commerce.” Morrison, 529 U.S. at 617, 120 S.Ct. 1740. See also id. at 656, 120 S.Ct. 1740 (Breyer, J., dissenting) (“The majority holds that the federal commerce power does not extend to such noneconomic activities as noneco-nomic, violent criminal conduct that significantly affects interstate commerce only if we aggregate the interstate effect[s] of individual instances.” (internal quotations omitted)); Julie Goldscheid, United States v. Morrison and the Civil Rights Remedy [263]*263of the Violence Against Women Act: A Civil Rights Law Struck Down in the Name of Federalism, 86 Cornell L.Rev. 109, 111 (2000) (“[Morrison] established that Congress cannot enact laws under the Commerce Clause that regulate noneco-nomic, violent criminal conduct based only-on the conduct’s aggregate effect on interstate commerce.”). Morrison built on Lopez, but also expanded it.
For this reason, the dissent’s observation that Morrison “relied on” Lopez, (dissent at 265) is irrelevant to whether Waucaush meaningfully understood the substance of the charges against him. As commentators recognized following its release, “[Lopez ] did not say absolutely that only commercial activities can be reached by the ‘affects’ branch of Congress’s ‘commerce authority’ .... It only indicated that it would be harder to so reach it. But how much harder? What more would be necessary before the Court will allow Congress to reach an activity, not itself commercial, but which affects interstate commerce?” Lawrence Lessig, Translating Federalism: United States v. Lopez, 1995 Sup.Ct. Rev. 125, 203 (1995). See also, e.g., Larry Kramer, The Supreme Court, 2000 Term-Forward: We the Court, 115 Harv. L.Rev. 4, 141 (2001) (“[In Lopez, t]he Justices had not yet clarified how they planned to determine substantiality, other than to suggest what looked like a principle of proximate causation[.]”); Charles Fried, The Supreme Court, 1995 Term — Forward: Revolutions?, 109 Harv. L.Rev. 13, 41 (1995) (“The projective power of the precedent will depend on how Lopez’s conclusion that the statute’s relation to commerce was not substantial will be understood and applied.”). Morrison addressed these questions — which went to the heart of the charges against Waucaush — for the first time.
Accordingly, and most importantly, the district court’s own conduct in this case confirms that Morrison and Jones added to the understanding of the Commerce Clause that Lopez initially set forth. Prior to Morrison and Jones, the district court refused to dismiss the charges against Waucaush and his co-defendants on the basis of Lopez alone. After Morrison and Jones were decided, the district court granted a rehearing to one of Waucaush’s codefendants who had not yet been sentenced. The district court then ruled that “the recent Supreme Court authority [ (Morrison and Jones) ] relied upon by [Garcia] compels the conclusion that this Court lacks jurisdiction to decide the RICO charges brought by the Government in this case” and further noted that those cases “placefd] limitations on the reach of federal jurisdiction over non-economie crimes.”
Finally, to prohibit Waucaush from taking advantage of Morrison and Jones would ignore the reason that the Supreme Court has applied retroactively decisions that narrow the substantive scope of criminal statutes: to do otherwise would “produce a class of persons convicted of conduct the law does not make criminal.” Schriro v. Summerlin, — U.S. -, 124 S.Ct. 2519, 2523, 159 L.Ed.2d 442 (2004). Because Waucaush’s misunderstanding of the law led him to plead guilty to conduct which the law had not made a crime, his plea was unintelligent and his conviction cannot stand.
III. CONCLUSION
The district court’s judgment is therefore REVERSED, and the case is REMANDED to the district court for entry of an order granting Waucaush’s petition.