dissenting:
I agree with the majority that the appellants likely forfeited their vagueness chai-*156lenge, and that the issue is one of plain error. The test for plain error is that there must be (i) error, (ii) that is plain, (iii) that affects substantial rights, and (iv) that seriously affects the fairness, integrity, or public reputation of judicial proceedings. This standard is satisfied here, applying the analysis we employed in United States v. Thomas, 274 F.3d 655, 667 (2d Cir.2001) (in banc). Certainly, conviction under a statute that is unconstitutionally vague on its face is an error of constitutional magnitude. See United States v. Handakas, 286 F.3d 92, 111-12 (2d Cir.), cert. denied, 537 U.S. 894, 123 S.Ct. 168, 154 L.Ed.2d 160 (2002). Reaching the merits, I respectfully dissent because in my view the so-called “honest services” amendment to the wire and mail fraud statute, 18 U.S.C. § 1346, flunks the test for facial vagueness set forth by the Supreme Court in City of Chicago v. Morales, 527 U.S. 41, 119 S.Ct. 1849, 144 L.Ed.2d 67 (1999).
I
The test for facial invalidity of a criminal statute was articulated by the Supreme Court in 1999: “Vagueness may invalidate a criminal law for either of two independent reasons. First, it may fail to provide the kind of notice that will enable ordinary people to understand what conduct it prohibits; second, it may authorize and even encourage arbitrary and discriminatory enforcement.” Id. at 56, 119 S.Ct. 1849 (Stevens, /., writing for the Court, joined by Ginsburg and Souter, JJ.); accord id. at 64-65, 119 S.Ct. 1849 (O’Connor, J., concurring in part and concurring in the judgment, joined by Breyer, J.); see also Kolender v. Lawson, 461 U.S. 352, 357, 103 S.Ct. 1855, 75 L.Ed.2d 903 (1983).
The majority opinion states that the governing standard for facial challenges outside of the First Amendment context is to be drawn from United States v. Salerno, 481 U.S. 739, 107 S.Ct. 2095, 95 L.Ed.2d 697 (1987), which states in dicta that a statute is facially invalid only if there is “no set of circumstances” in which it would be valid. Id. at 745, 107 S.Ct. 2095. The majority’s expressed preference for the 1987 Salerno dicta over the 1999 Morales holding is itself a bit of dicta because the majority holds that the statute in question survives scrutiny under either test. Maj. Op. at 144. Although I believe that section 1346 is so vague that there is “no set of circumstances” in which it is clear enough to be applicable, I think that the test does matter, chiefly to assure a sound analysis of constitutional sufficiency. I therefore undertake to demonstrate that the governing test is the one set forth in Morales.
At most, four Justices in Morales invoked the Salerno test for facial vagueness or words suggestive of that standard. See Morales, 527 U.S. at 77-81 & nn. 1-3, 119 S.Ct. 1849 (Scalia, J., dissenting); id. at 111-12, 114, 119 S.Ct. 1849 (Thomas, J., dissenting, joined by Rehnquist, C.J., and Scalia, /.); id. at 71, 119 S.Ct. 1849 (Breyer, J., concurring in part and concurring in the judgment) (“The ordinance is unconstitutional ... because the policeman enjoys too much discretion in every case. And if every application of the ordinance represents an exercise of unlimited discretion, then the ordinance is invalid in all its applications.”). In any event, Morales did not implement the Salerno dicta. See id. at 81, 119 S.Ct. 1849 (Scalia, J., dissenting) (“Instead of requiring respondents, who are challenging the ordinance, to show that it is invalid in all its applications, [the Justices in the majority] have required [the government] to show that it is valid in all its applications.”).
It is true, of course, that several other propositions discussed in Morales only at*157tracted a plurality. As the majority opinion notes, a three-Justice plurality of the Morales court would ■ apparently allow challenges for facial vagueness outside of the First Amendment context to criminal laws that both lack a mens rea requirement and infringe on constitutional rights. See id. at 55, 119 S.Ct. 1849 (Stevens, J., writing for the Court, joined by Ginsburg and Souter, JJ.); Maj. Op. at 131. However, only those same three Justices believed that the ordinance challenged in Morales implicated such a constitutional right and lacked a specific intent requirement. Compare id. at 55, 119 S.Ct. 1849 (Stevens, J., writing for the Court, joined by Ginsburg and Souter, JJ.) (“[The challenged statute] is a criminal law that contains no mens rea requirement and infringes on constitutionally protected rights.”) (internal citations omitted), with id. at 66, 119 S.Ct. 1849 (O’Connor, J., concurring in part and concurring in the judgment, joined by Breyer, J.) (“To be sure, there is no violation of the ordinance unless a person fails to obey promptly the order to disperse. But, a police officer cannot issue a dispersal order until he decides that a person is remaining in one place ‘with no apparent purpose’ and the ordinance provides no guidance to the officer on how to make this antecedent decision.”) and id. at 69, 119 S.Ct. 1849 (Kennedy, J., concurring in part and concurring in the judgment) (noting that the ordinance “reach[ed] a broad range of innocent conduct” and stating that “[t]he predicate of an order to disperse is not, in my view, sufficient to eliminate doubts regarding the adequacy of notice under this ordinance.”). Thus, three of the six Justices supporting the result in Morales (Justices O’Connor, Kennedy, and Breyer) applied the Morales test outside the First Amendment context without regard to whether the statute had an intent requirement or infringed on a constitutional right. Certainly, none of these propositions nor the invocation of the Salerno standard — each attracting only a plurality in Morales — constitutes Supreme Court precedent.
The only proposition attracting a majority in Morales was that a criminal statute that “reach[es] a substantial amount of innocent conduct” and thereby fails to “establish minimal guidelines to govern law enforcement” is, on its face, unconstitutionally vague. Id. at 60-61, 119 S.Ct. 1849 (Stevens, J., writing for the Court in part V, joined by Ginsburg and Souter, JJ.), id. at 69, 119 S.Ct. 1849 (Kennedy, J., joining in part V, concurring in part and concurring in the judgment); see also id. at 64-65, 119 S.Ct. 1849 (O’Connor, /., concurring in part and concurring in the judgment, joined by Breyer, J.).
We are therefore required to apply Morales here. Although the majority holds that section 1846 withstands either test, it is quite clear that the .statute imposes insufficient constraint on prosecutors, gives insufficient guidance to judges, and affords insufficient notice to defendants. That insufficiency can be illustrated by reference to the cases cited in the majority opinion. As to prosecutors, the majority does not disturb the holding that overturned the conviction in Handakas, 286 F.3d at 96, of a contractor who falsely promised to abide by New York’s wage laws, on the ground that the “honest services” amendment was vague as applied. The majority opinion thus confirms that the prosecution in Han-dakas was misguided, or (in my view) that prosecutorial discretion was unguided altogether because the statute insufficiently describes the offense. The majority opinion similarly demonstrates that judges cannot understand what conduct this statute criminalizes. The majority holds that case law predating McNally v. United States, 483 U.S. 350, 107 S.Ct. 2875, 97 L.Ed.2d 292 (1987), is relevant to deter*158mining the meaning of section 1346 and thus rejects the contrary rule announced in United States v. Sancho, 157 F.3d 918, 922 (2d Cir.1998) (per curiam); the majority holds that reasonable foreseeability of non-cfe minimis economic or pecuniary harm is not a necessary element of the crime and thus rejects the panel’s contrary holding in United States v. Rybicki, 287 F.3d 257, 265-66 (2d Cir.2002); and the majority rejects the view expressed in Handakas that the statute is facially vague (though that panel was constrained by Circuit precedent from so holding). 286 F.3d at 104-06. In short, if the statute means what the majority says it means, eight judges in this Circuit failed to understand it in Sancho, Handakas, and Rybicki. Of course, overturned convictions and in banc rejection of panel rulings do not prove facial vagueness. But this Circuit’s long experience with section 1346 is nevertheless telling evidence that most lawyers and judges, not to speak of ordinary laymen or prospective defendants, cannot be expected to understand the statute.
II
The first question that bears on the vagueness inquiry is whether “a penal statute define[s] the criminal offense with sufficient definiteness that ordinary people can understand what conduct is prohibited.” Kolender, 461 U.S. at 357, 103 S.Ct. 1855. “The plain meaning of ‘honest services’ in the text of § 1346 simply provides no clue to the public or the courts as to what conduct is prohibited under the statute.” Handakas, 286 F.3d at 104. The majority opinion is a prolonged and sustained search for some prior settled meaning for an opaque statutory phrase — “the intangible right of honest services” — so that it can be construed as a term of art. That effort to infuse the putative term of art with meaning is conducted in a painstaking way, and considers an abundant variety of alternative meanings. However, a term of art has one single and apparent meaning, in the same way that a pun has two; it is as odd to conduct a scholarly search for the meaning of a term of art as it would be to hear a pun, conduct research in semantics, etymology and philology for a month, and then laugh.
It may be (as the majority holds) that, in enacting section 1346, Congress intended to reinstate a body of case law that had been overruled by the Supreme Court in McNally. See Handakas, 286 F.3d at 103, 104-05 (citing legislative history). But that insight gets us nowhere in terms of limits on prosecutorial power and notice to the public:
The requirement imposed by the Supreme Court [in McNally] to speak more clearly was not for the benefit of the Circuit Courts which had, in fact, given birth to these concepts in the first place. Rather, the requirement ... was for the benefit of the public, the average citizen, ... who must be forewarned and given notice that certain conduct may subject him to federal prosecution.
United States v. Brumley, 116 F.3d 728, 745-46 (5th Cir.1997) (in banc) (Jolly, /., dissenting) (emphasis added). A statute is unconstitutionally vague unless it provides a “person of ordinary intelligence a reasonable opportunity to know what is prohibited.” United States v. Strauss, 999 F.2d 692, 697 (2d Cir.1993) (internal quotation marks omitted). We have held that notice is insufficient if lay persons are required to “perform! ] the lawyer-like task of statutory interpretation by reconciling the text of [ ] separate documents.” Chatin v. Coombe, 186 F.3d 82, 89 (2d Cir.1999) (invalidating a prison administrative prohibition as unconstitutionally vague as applied). Construing a statute (as the majority does here) to say that scores of *159overruled cases are hereby revived, requires lay persons to do lawyer-like tasks that few lawyers would have the skills to perform:
[N]o one can know what is forbidden by § 1346 without undertaking the “lawyer-like task” of answering the following questions: [1] Can pr e-McNally case law be consulted to illuminate the wording of § 1346? [2] Can any meaning be drawn from the case law, either the uneven pr e-McNally cases or the few cases decided post- § 1346? [3] Is one to be guided only by case law within one’s own circuit, or by the law of the circuits taken together (if that is possible)?
Handakas, 286 F.3d at 105.
It is remarkable how little the majority’s search for meaning has turned up. The term of art for which meaning is sought is essentially the entire mouthful of the statute: “scheme or artifice to deprive another of the intangible right of honest services.” Before McNally, this phrase encompassed four “categories” of honest-services cases:
[1] government officials who defraud the public of their own honest services; [2] elected officials and campaign workers who falsify votes and thereby defraud the electorate of the right to an honest election; [3] private actors who abuse fiduciary duties by, for example, taking bribes; and [4] private actors who defraud others of certain intangible rights, such as privacy.
Id. at 101-02 (citing McNally, 483 U.S. at 362-64 nn. 1-4, 107 S.Ct. 2875). Yet the majority’s search for meaning bears upon no more than one subgroup of the four categories of honest-services cases.
Following an exhaustive, scholarly analysis, my colleagues conclude that one of these four categories of conduct criminalized by pr e-McNally case law — the theft of privacy rights- — cannot be revived as a criminal offense, presumably because the statute would be unconstitutionally vague under Salerno' as well as Morales if so applied. Maj. Op. at 138-139 n. 13. Contra United States v. Condolon, 600 F.2d 7, 8-9 (4th Cir.1979); United States v. Louderman, 576 F.2d 1383, 1387-88 (9th Cir.1978). They also limit their search for meaning to cases of “honest services” fraud in the private sector. No attempt is made (advisedly) to describe the prohibition, if any, in the two remaining categories of public sector “honest services” fraud criminalized in the pr e-McNally case law; any “well-settled meaning” relating to these remaining categories will presumably be supplied — or looked for- — later.
The majority intuits a statutory meaning that is insufficient even to describe the subgroup of private sector cases. Where kickbacks or bribery are involved, the majority holds that a “scheme or artifice to deprive another of the intangible right of honest services” means:
a scheme or artifice to use the mails or wires to enable an officer or employee of a private entity (or a person in a relationship that gives rise to a duty of loyalty comparable to that owed by employees to employers) purporting to act for and in the interests of his or her employer (or of the other person to whom the duty of loyalty is owed) secretly to act in his or her or the defendant’s own interests instead, accompanied by a material misrepresentation made or omission of information disclosed to the employer or other person.
Maj. Op. at 141-142. However, in cases of self-dealing, there “may ” be an additional requirement that the alleged conduct “caused, or at least was capable of causing, some detriment” to the employer. Maj. Op. at 142 (emphasis added). The tentativeness of the, majority’s approach is well justified: a number of pr e-McNally cases *160hold that there is no such requirement of economic detriment. See, e.g., United States v. Bronston, 658 F.2d 920, 927 (2d Cir.1981) (upholding mail fraud conviction against a law firm partner paid for representing a client in contract negotiation despite firm representing competing bidder); see also Rybicki, 287 F.3d at 262 (“[I]t was well-settled law both before and after McNally that the government does not have to establish that a scheme to defraud was successful or resulted in any actual [economic] harm to the victim.”). The majority deems such cases “atypical,” however. Maj. Op. at 141 n. 16. They may be atypical; but even assuming that a term of art can be distilled from the body of case law that was overruled in McNally, surely no unambiguous meaning can be assigned to a phrase that has no meaning except what can be distilled from some pre-McNally cases provided that other pre-McNally cases are ignored, particularly since the designation of overruled cases that are in and those that are out is itself essentially arbitrary. Ordinary people cannot be expected to undertake such an analysis; rare is the lawyer who could do it; and no two lawyers could be expected to agree independently on the elements of an offense that must be defined by such a project.
The majority claims that any ambiguity is of no concern here because defendants’ conduct falls “squarely within the meaning of ‘scheme or artifice to deprive another of the intangible right of honest services’ as distilled from the pr e-McNally private sector cases.” Maj. Op. at 142. But this argument is no answer to a facial challenge for vagueness. The only relevant question is whether “ordinary people can understand what conduct is prohibited.” Kolender, 461 U.S. at 357, 103 S.Ct. 1855.
It is only too obvious that there is no settled meaning to the phrase “the intangible right of honest services” that is capable of providing constitutionally adequate notice. If there were, the judges and prosecutors in this Circuit would certainly know it. Yet, the majority overrules the holding (erroneously characterized as dicta) in Sancho, 157 F.3d at 922, that pr e-McNally case law is irrelevant to determining the meaning of section 1346 — i.e., that “the intangible right of honest services” lacks a well settled meaning. The majority also concludes that the panel opinion in this case was wrongly decided insofar as it held that one element of a section 1346 offense is that a loss of money or property be reasonably foreseeable. See Rybicki, 287 F.3d at 265-66. Finally, the majority preserves the holding of Handakas, 286 F.3d at 96, which means that the prosecutors in the Eastern District of New York did not understand what the statute meant. How can the public be expected to know what the statute means when the judges and prosecutors themselves do not know, or must make it up as they go along?
Ill
The second question that bears on facial vagueness is whether the “legislature [has] established] minimal guidelines to govern law enforcement.” Kolender, 461 U.S. at 358, 103 S.Ct. 1855. This second inquiry is “the more important” of the two, and is alone sufficient to decide constitutional infirmity. Id. at 358, 361 & n. 10, 103 S.Ct. 1855. The governing test is whether the statute “permit[s] ‘a standardless sweep [that] allows policemen, prosecutors, and juries to pursue their personal predilections.’ ” Id. at 358, 103 S.Ct. 1855 (second alteration in original). “An enactment fails to provide sufficiently explicit standards for those who apply it when it im-permissibly delegates basic policy matters to policemen, judges and juries for resolution on an ad hoc and subjective basis.” Handakas, 286 F.3d at 107 (internal quota*161tion marks and citations omitted). Thus, a statute is facially vague if it “necessarily entrusts lawmaking to the moment-to-moment judgment” of law enforcement. Morales, 527 U.S. at 60, 119 S.Ct. 1849 (Stevens, /., writing for the majority) (internal quotation marks omitted).
The majority opinion affirms the result in Handakas, which means (as I pointed out earlier) that the prosecutors in the Eastern District of New York did not understand what the statute meant. Moreover, the meaning supplied by the majority to uphold this use of section 1346 is as elusive as the statute itself. According to the majority, the “honest services” offense is a misrepresentation or omission that (i) is made by a private person who secretly acts in self-interest while purporting to act in the interests of the employer and (ii) is capable of leading a reasonable employer to change its conduct (i.e., it is “material”). Maj. Op. at 141-142. Neither requirement limits prosecutorial discretion.
No limit is placed on the exercise of prosecutorial discretion by requiring a showing that an employee secretly prefers her own interest to the interest of the employer; it is naive to assume that this preference is not the most common premise of private employment. “[Rjelationships in the private sector generally rest upon concerns and expectations less ethereal and more economic than abstract satisfaction of receiving ‘honest services’ for their own sake.” United States v. Frost, 125 F.3d 346, 365 (6th Cir.1997). Every salaried employee can be said to work for her own interest while purporting to act in the interests of the employer. Yet the majority opinion effectively makes “dishonesty by an employee, standing alone, [ ] a crime.” Id. at 368.
Nor is prosecutorial discretion limited by the “materiality” requirement See United States v. Sun-Diamond Growers, 138 F.3d 961, 973 (D.C.Cir.1998) (“Every material act of dishonesty by an employee deprives the employer of that worker’s ‘honest services,’ yet not every such act is converted into a federal crime by the mere use of the mails or interstate phone system.”); Frost, 125 F.3d at 365 (in rejecting the materiality test: “[I]f a ‘change in business conduct’ occurs under the materiality standard when a business alters its behavior merely to avoid the appearance of impropriety ..., the intangible right to honest services doctrine may lack substantive limits in the private sector.”) (emphasis in original).
The majority codifies a doctrine that is as standardless as the statute itself. Nothing in the majority opinion prevents criminalization of any of the following conduct: a regulated company that employs a political spouse; an employee who violates an employee code of conduct; a lawyer who provides sky-box tickets to a client’s general counsel; a trustee who makes a self-dealing investment that pays off; or an officeholder who has made a decision in order to please a constituent or contributor, or to promote re-election, rather than for the public good (as some prosecutor may see the public good).
The majority is unconcerned with the standardless sweep of the statute because supposedly there is “a wide swath of behavior” for which the prohibitions of section 1346 are “clear.” Maj. Op. at 144. This statement turns upside down the Morales test for facial vagueness: whether a statute reaches a wide swath of behavior that no one (yet) deems criminal and fails to provide minimal guidance to law enforcement. See Morales, 527 U.S. at 60, 119 S.Ct. 1849 (Stevens, J., writing for the majority). “ ‘It would certainly be dangerous if the legislature could set a net large enough to catch all possible offenders, and leave it to the courts to step inside and say *162who could be rightfully detained, and who should be set at large.’ ” Kolender, 461 U.S. at 358 n. 7, 103 S.Ct. 1855 (quoting United States v. Reese, 92 U.S. 214, 221, 23 L.Ed. 563 (1875)). I share the confidence implicit in the majority opinion that prosecutors in this Circuit and the Attorney General under whom they serve can be trusted to avoid any systematic abuse of such a statute; but we should construe this statute so that it serves to bind those who nevertheless may need constraint.1
IV
The majority opinion’s search for a meaning of art leans heavily on the overruled pr e-McNally case law of other circuits. But “[e]ven the circuits that have reinstated pr e-McNally law recognize that ad hoc parameters are needed to give the statute shape.” Handakas, 286 F.3d at 109 (collecting cases). Although a number of circuits have upheld section 1346 against a claim of facial vagueness, there is now wide disagreement among the circuits as to the elements of the “honest services” offense. These opinions, taken together, refute rather than support the idea that section 1346 has any settled or ascertainable meaning or that the offense it describes has known contours:
• What mens rea must be proved by the government? The majority follows Second Circuit precedent in holding that an intent to cause economic harm is not required — a defendant need only have intended to deprive another of the “intangible right of honest services.” Maj. Op. at 145. However, in the Seventh Circuit, an intent to achieve personal gain is an element of the offense. See United States v. Bloom, 149 F.3d 649, 656-57 (7th Cir.1998). But see United States v. Welch, 327 F.3d 1081, 1106-07 (10th Cir.2003) (holding that the text and structure of the mail fraud statutes do not support “adding an element” to “honest services” fraud requiring that defendant seek to obtain a personal benefit). The Eight Circuit describes the mens rea element as “caus[ing] or intending] to cause actual harm or injury, and in most business contexts, that means financial or economic harm.” See United States v. Pennington, 168 F.3d 1060, 1065 (8th Cir.1999). One circuit has held that, to secure an honest services conviction, “[t]he prosecution must prove that the employee intended to breach a fiduciary duty.” Frost, 125 F.3d at 368. Other circuits merely require a showing of “fraudulent intent.” See United States v. Cochran, 109 F.3d 660, 667 (10th Cir.1997); United States v. Jain, 93 F.3d 436, 442 (8th Cir.1996).
• Must the defendant have caused actual tangible harm? Compare Jain, 93 F.3d at 442 (“When there is no tangible harm to the victim of a private scheme, it is hard to discern what intangible ‘rights’ have been violated.”), with Frost, 125 F.3d at 369 (“[A] defendant accused of scheming to deprive another of honest *163services does not have to intend to inflict an economic harm upon the victim.”). Some circuits have required that the misrepresentation be material, ie., that the employee have reason to believe that the information would lead a reasonable employer to change its business conduct. See Cochran, 109 F.3d at 667 & n. 3; United States v. Gray, 96 F.3d 769, 775 (5th Cir.1996); Jain, 93 F.3d at 442. Other circuits only require a showing that it was reasonably foreseeable for the victim to suffer economic harm. Frost, 125 F.3d at 368; Sun-Diamond Growers, 138 F.3d at 973-74. We adopted this last requirement in the Ry-bicki panel opinion, and now abandon it. See Rybicki, 287 F.3d at 265.
• What is the duty that must be breached to violate section 1346? The majority holds that it is the duty owed by an employee to an employer, or by “a person in a relationship that gives rise to a duty of loyalty comparable to that owed by employees to employers” (whatever that means). Maj Op. at 141-142. Some circuits only allow prosecutions for breach of an employee’s duty to an employer. See, e.g., Brumley, 116 F.3d at 735. Other circuits require the breach of a fiduciary duty. See Frost, 125 F.3d at 366, 368; Sun-Diamond Growers, 138 F.3d at 974.
• Is the source of that duty state or federal law? The majority does not say, and other circuits are split. Compare Frost, 125 F.3d at 366 (“Federal' law governs the existence of fiduciary duty under the mail fraud statute.” (emphasis added)), with Brumley, 116 F.3d at 735 (“We have held that, services under § 1346 are those an employee must provide the employer under state law.” (emphasis added)).
• Did section 1346 revive pre-McNally case law; if so must each circuit look to its own governing precedent or to some set of rules distilled from the whole body of pr e-McNally cases? See Brumley, 116 F.3d at 733-34 (looking to “plain language” of a statute to discern its “meaning” because “before McNally the doctrine of honest services was not a unified set of rules [a]nd Congress could not have intended to bless each and every pr e-McNally lower court honest services’ opinion”); Frost, 125 F.3d at 364, 365 (holding that “ § 1346 has restored the mail fraud statute to its pre-McNally scope, according to previous opinions interpreting the intangible right to honest services” and looking to “Sixth Circuit precedent issued before McNally in order to discover the precise contours of the right in this circuit”). In Sancho, we held that pre-McNally cases could not be considered in determining the meaning of the statute. See 157 F.3d at 921-22. We now overrule Sancho and adopt an approach divining statutory meaning by analyzing cases overruled by McNally as a whole.
In sum, the circuits are fractured on the basic issues: (1) the requisite mens rea to commit the crime, (2) whether the defendant must cause actual tangible harm, (3) the duty that must be breached, (4) the source of that duty, and (5) which body of law informs us of the statute’s meaning. This lack of coherence has created “a truly extraordinary statute, in which the substantive force of the statute varie[s] in each judicial circuit,” Brumley, 116 F.3d at 743 n. 7 (Jolly, J., dissenting).
Y
As the foregoing section documents, the vagueness of the statute has induced court after court to undertake a rescue operation by fashioning something that (if enacted) would withstand a vagueness challenge. The felt need to do that attests to the *164constitutional weakness of section 1346 as written. And the result of all these efforts — which has been to create different prohibitions and offenses in different circuits — confirms that the weakness is fatal. Judicial invention cannot save a statute from unconstitutional vagueness; courts should not try to fill out a statute that makes it an offense to “intentionally cause harm to another,” or to “stray from the straight and narrow,” or to fail to render “honest services.”
“[Legislatures and not courts should define criminal activity.” United States v. Bass, 404 U.S. 336, 348, 92 S.Ct. 515, 30 L.Ed.2d 488 (1971); see also Handakas, 286 F.3d at 101 (punishment for “constructive offenses” violates the Due Process Clauses of the Fifth and Fourteenth Amendments, which “require the legislature to specify the elements of criminal offenses”); Bloom, 149 F.3d at 654 (describing “a federal common-law crime” as “a beastie that many decisions say cannot exist”). As the splintering among the circuits demonstrates, section 1346 effectively imposes upon courts a role they cannot perform. When courts undertake to engage in legislative drafting, the process takes decades and the work is performed by unelected officials without the requisite skills or expertise; and as the statutory meaning is invented and accreted, prosecutors are unconstrained and people go to jail for inchoate offenses.
The majority complacently cites by analogy similarly vague words in the Sherman Act that make unlawful any “restraint of trade.” Maj. Op. at 137-138. The Sherman Act predates Morales by a century. Moreover, the comparison proves too much. Courts construe the Sherman Act primarily as a civil and regulatory statute, and criminal Sherman Act prosecutions have long been limited essentially to price-fixing. In keeping with its exceptional history, courts have broadly construed the Sherman Act as “a charter of freedom” akin to constitutional provisions. See, e.g., United States v. U.S. Gypsum Co., 438 U.S. 422, 439, 98 S.Ct. 2864, 57 L.Ed.2d 854 (1978) (“Simply put, the Act has not been interpreted as if it were primarily a criminal statute; it has been construed to have a ‘generality and adaptability comparable to that found to be desirable in constitutional provisions.’ ”) (citation omitted)); 2 Phillip E. Areeda, Herbert Hovenkamp & Roger D. Blair, Antitrust Law § 303b4, at 33 (2d ed. 2000) (“Because they were usually dealing with civil proceedings, the courts have implicitly understood the Sherman Act as a mandate to develop a common law of antitrust — as indeed it would have to be in order to fulfill its purpose as a ‘charter of freedom.’ ”). The analogy to antitrust law is valid only insofar as section 1346 is an invitation by Congress for courts to develop a common law of criminal punishment.
Finally, “[t]he absence of discernible standards in the ‘honest services’ doctrine implicates principles of federalism.” Handakas, 286 F.3d at 110. The majority opinion in effect criminalizes all material acts of dishonesty by employees or by persons who owe analogous duties. While the majority opinion concerns itself only with private sector “honest services” fraud, there seems to be no principled basis in the statute’s wording or in (overruled) cases for excluding from section 1346’s reach services rendered in the public sector. Thus, without any obvious limiting principles, the majority opinion invites federal prosecutors to police honesty in the corridors of state government by invoking section 1346 against state employees for their acts of “honest services” fraud. This construction of section 1346 undoubtedly “leaves its outer boundaries ambiguous and involves the Federal Government in setting standards of disclosure and good *165government for local and state officials.” McNally, 483 U.S. at 360, 107 S.Ct. 2875.
I believe that Congress has not heeded the Supreme Court’s admonition to “speak more clearly than it has.” See id. We should not “approve a sweeping expansion of federal criminal jurisdiction in the absence of a clear statement by Congress.” Cleveland v. United States, 531 U.S. 12, 24, 121 S.Ct. 365, 148 L.Ed.2d 221 (2000).
The majority opinion exhibits deference to Congress by conscientiously seeking to understand congressional intent, and the effort and product are scholarly and scrupulous. But the work accomplished by the majority opinion, which is admirable in its way, is properly the work of legislators in statutory drafting and the work of the executive in framing prosecutorial standards. “If the words of a criminal statute insufficiently define the offense, it is no part of deference to Congress for us to intuit or invent the crime.” Handakas, 286 F.3d at 109-10. I respectfully dissent.
. The majority ultimately relies on the reductionist argument that these defendants must have known that it would be illegal “to use the wires and the mails ... to pay off insurance adjustors." Maj. Op. at 142. The natural drift of this observation is that the scheme inflated the settlement of claims to the benefit of the defendants' clients and to the detriment of the insurance companies. However, the government never contended that this occurred. As the majority opinion recites, "the government acknowledged that it would not seek to prove that the amount of any of the settlements connected with a payment to an adjuster had been inflated above what would have been a reasonable range for that settlement." Maj. Op. at 128. There was certainly no reason for these defendants to believe that the gratuities at issue — which the government never contended caused a loss — amounted to a federal mail or wire fraud violation.