Full opinion text
WIENER, Circuit Judge: Albert Lipscomb, a former member of the Dallas City Council, appeals his convictions for conspiracy and program bribery, in violation of 18 U.S.C. § 666 (“§ 666”). Whether he raises a constitutional challenge to his convictions, and, if so, how we should rule on that challenge, are questions that have divided our panel three ways, as will become clear from our separate writings. Despite this tripartite fractionation, however, different majorities of the panel conclude that (1) the question of § 666’s as-applied constitutionality is properly before the panel and should be addressed; (2) the district court had subject-matter jurisdiction of this criminal bribery case against Lipscomb; and (3) the court abused its discretion in transferring the trial sua sponte over Lipscomb’s objections. We therefore reverse his conviction, vacate his sentence, and remand for a new trial. I. FACTS Both Lipscomb’s conduct and particular jurisdictional facts are important to the varying views of the members of this panel. We therefore recount them in considerable detail. A. Lipscomb’s Offense Conduct Lipscomb served on the Dallas City Council (the “Council”) from 1984 to 1993 and again from 1995 until 2000. During his first period of service, Lipscomb vigorously opposed any measure favorable to taxicab companies, including Yellow Cab and Checker Cab (together, “Yellow Cab”), both owned by his co-conspirator, Floyd Richards. Lipscomb’s animus against cab companies apparently was grounded in a belief that cab companies perennially failed to serve the minority community adequately- During his second period of service on the Council, however, Lipscomb demonstrated a considerably kinder disposition toward cab companies, especially Yellow Cab. In 1994, during Lipscomb’s hiatus from the Council, Richards asked Lipscomb to help improve Yellow Cab’s reputation in the minority community and offered to pay Lipscomb $1,000 a month in cash for that help. Lipscomb assented to this proposal. Richards and Lipscomb agreed to continue this arrangement as long as it was mutually agreeable. All this transpired orally. Richards continued to make the monthly payments to Lipscomb after he was reelected to the Council. At times, Richards would receive phone calls from Lipscomb indicating that he needed a payment, after which Lipscomb would visit Yellow Cab’s office and receive cash that Richards took from the company safe. Sometimes during these meetings, Richards and Lipscomb would discuss taxicab issues then pending before the Council. The government alleged that in addition to making these monthly payments to Lipscomb, Richards gave Lipscomb free use of cars, free cellular telephone service, and free cab rides worth more than $3,300. When Lipscomb ran again, his advisers heard Richards declare that he was willing to spend up to $30,000 to get Lipscomb elected. When Richards learned that corporations could not contribute to campaigns and that individuals could contribute no more than $1,000, however, he decided to “lend” $20,500 to a business owned by Lipscomb’s daughter and son-in-law. That money was intended by all concerned to help fund Lipscomb’s campaign, and it did so; but Lipscomb did not report the campaign “loan” or any of the payments in his campaign finance reports or his personal financial statements. Richards testified that although he never made the quid pro quo explicit, he expected that, in return for the monthly payments and the campaign funding, Lipscomb would cast votes favorable to Yellow Cab. Richards testified further that he and Lipscomb had an understanding, and that Richards was satisfied that Lipscomb knew that the payments would stop if he voted the wrong way. Lipscomb’s support of Yellow Cab went far beyond the casting of favorable votes at meetings of the Council. Over time, he and Richards discussed each of the taxicab issues on which Lipscomb allegedly was influenced by this bribery: (1) operating authority and fleet increases, (2) location of dispatch offices, (3) age limits and inspections, and (4) insurance ratings. Lipscomb had opposed Yellow Cab on these issues before 1994, but when he returned to the Council, he supported that company vigorously and often. For example, in 1994 Lipscomb, as a private citizen, had spoken out against authority for Yellow Cab and two other cab companies to operate in Dallas. Once he returned to the council, though, he supported Yellow Cab’s requests for increases in the size of its cab fleets. Yet when cab companies unaffiliated with Richards sought authority to operate in Dallas, Lipscomb urged that their applications be removed from the council’s agenda. When another cab company’s request for operating authority was taken up by the council, Lipscomb tried to require a voice vote on the matter. Yellow Cab also needed relief from a city ordinance requiring cab companies to maintain their dispatch offices inside the Dallas city limits. After a city staffer learned that Yellow Cab was violating this policy, she sought to enforce it, but the Council referred the matter to its Transportation Committee. Even though Lipscomb did not serve on that committee, he attended its meeting and browbeat the staffer, going so far as to ask her when she would retire. Eventually, with Lipscomb’s encouragement, the Council permitted cab companies to operate dispatch offices in the Dallas suburbs, thus legitimating Yellow Cab’s office, the only one in violation, in which Yellow Cab had invested $15,000. Because Yellow Cab had the newest fleet among the cab companies serving Dallas, the City was encouraged by Yellow Cab energetically to enforce against its competitors the City’s age limit on vehicles for hire and its requirement that they be inspected. In 1992, Lipscomb had favored relaxing both rules, but in 1996, after he was told by Richards that he wanted stricter enforcement, Lipscomb began to support age limits on sedan-style limousines similar to the limits that applied to taxicabs. He also sought to remove older shuttles and limousines from service more quickly, and he opposed the Council’s effort to revisit its earlier vote — favorable to Yellow Cab — to approve stricter age limits. Lipscomb also acted on Yellow Cab’s behalf with respect to insurance issues. Yellow Cab lobbied the Council to require that the insurance coverage mandated for taxis be written by insurers with favorable financial ratings. This proposal proved to be controversial: The City’s Director of Human Services, whose department handled insurance matters, was concerned that a rating requirement might favor large firms and exclude small businesses owned by minorities or women. Lipscomb nevertheless sought to put the rating requirement on the Council’s agenda, and both seconded and voted for a motion to increase the minimum rating. In sum, Lipscomb energetically used many of the tools at the disposal of a Council member — his vote, his oversight authority, his agenda-setting power, and his other parliamentary privileges — to support policies favorable to Yellow Cab, even though these policies conflicted with his previous positions. B. Jurisdictional Facts During Lipscomb’s second period of council service, the City, through many of its agencies and departments, received substantial federal funds. In the year ending in September 1996, Dallas received $44.3 million and spent $48.1 million in federal financial assistance which funded a wide range of joint priorities: community development, farmer’s market infrastructure, emergency shelter, housing, community policing, airport and freeway improvements, arts development, pollution control, emergency management, interlibrary cooperation, child immunization, homeless health care, and substance abuse control, among others. Federal support in 1996 dwarfed state support, which totaled only $3.7 million received and $3.1 million spent. Other years were similar: in 1997, the city received $54.3 million and spent $53.3 million in federal funds, but received only $3.0 million and spent $3.8 million in state funds. Testimony of the city’s chief financial officer showed that in Dallas’s efforts to obtain and then allocate federal funds, the Council played an integral role: Q. And once the City gets the Department of Housing money or grant funds, does the City then disburse those funds? A. Yes, we do. Q. And is the disbursement by approval of a City Councilmember or the City Council at large? A. If the individual expenditure is greater than $50,000, or $15,000 in the case of professional services, it would come back to the Council for approval of that specific contract. Q. And does that frequently happen? A. Yes, uh-huh. Q. All right. And, in fact, does the Council have to approve, vote for and approve the application to HUD and the other agencies of the federal government to get federal money? A. Yes. They vote for the application and the acceptance of the money. The Council as a whole thus controlled— and individual' council members influenced — the City’s applications for, and receipt and expenditure of, at least forty million federal dollars each year. II. PROCEEDINGS The government secured a lengthy indictment against Lipscomb. Counts 2 through 33 charged him with specific substantive bribery violations of § 666(a)(1)(B) and charged Richards with aiding and abetting those offenses. Conversely, counts 34 through 65 charged Richards with bribery violations of § 666(a)(2) and charged Lipscomb with aiding and abetting. Count 1 charged Lipscomb with conspiring to violate § 666. Notably, the government did not charge Lipscomb with the misuse of state or federal funds. Three weeks before the long-scheduled trial date, the district court, acting sua sponte, without giving notice to the parties or holding a hearing, and over Lipscomb’s strenuous objections, transferred the trial from the Dallas Division of the Northern District of Texas to the Amarillo Division. Thereafter, Richards entered into a plea agreement which, among other things, required him to testify at trial. The jury convicted Lipscomb on all counts. The district court sentenced him to 41 months’ imprisonment, imposed a $7,500 fine, and ordered him to pay a $6,500 special assessment. The court also directed that the sentence be served under home confinement because of Lipscomb’s failing health and advanced age. Lipscomb appeals his conviction on several grounds. The government cross-appeals the home-confinement aspect of his sentence. III. STATUTORY INTERPRETATION A. Lipscomb’s Challenge to the Jurisdictional Reach of § 666 As it stood at the time and now stands, § 666 contains two monetary thresholds. Section 666 reads, in principal part: § 666. Theft or bribery concerning programs receiving Federal funds (a) Whoever, if the circumstance described in subsection (b) of this section exists— (1) being an agent of an organization, or of a State, local, or Indian tribal government, or any agency thereof— (B) corruptly solicits or demands for the benefit of any person, or accepts or agrees to accept, anything of value from any person, intending to be influenced or rewarded in connection with any business, transaction, or series of transactions of such organization, government, or agency involving anything of value of $5,000 or more; shall be fined under this title, imprisoned not more than 10 years, or both. (b) The circumstance referred to in subsection (a) of this section is that the organization, government, or agency receives, in any one year period, benefits in excess of $10,000 under a Federal program involving a grant, contract, subsidy, loan, guarantee, insurance, or other form of Federal assistance. Lipscomb insists that we should narrowly construe § 666 to avoid the constitutional question that arises if we interpret the statute to prohibit activity not directly related to federal spending or federally funded programs. He proposes that we construe the statute to require a nexus between his offense conduct and federal funds — or, put differently, that his conduct implicate a tangible federal interest. He also contends that, when so construed, the statute does not reach his conduct. Neither contention succeeds. The phylogeny of § 666 jurisprudence does reflect a growing tension between two possible focuses of the statute. One, which another court has dubbed the “funds focus,” would concentrate on deterring direct depletion of federal funds; the other, the so-called “corruption focus,” would combat “the corrupting, public-trust eroding effects of bribery” and would not require that federal funds be depleted or misallocated as a direct result of the bribe. Lipscomb’s proposal that we adopt the narrower, funds focus, however, would require us to ignore our consistently broad interpretation of § 666 as targeting corruption qua corruption. Furthermore, even if we were to read § 666 and our cases to construe it narrowly, to superimpose a nexus element, we would still conclude that there is a sufficient linkage between Lipscomb’s conduct and federal funds to support jurisdiction of Lipscomb’s case. B. Westmoreland and Its Progeny: The Corruption Focus — No Further Nexus Required We first interpreted § 666 in United States v. Westmoreland. The defendant, Westmoreland, was a county supervisor who was convicted of accepting bribes and kickbacks in connection with the purchasing of supplies for the county’s highway construction projects. The county received slightly more than $200,000 in total federal revenue-sharing funds, of which roughly 15% was allocated to Westmore-land’s district. Westmoreland contended that “the federal revenue sharing funds received [by her district] ... were segregated and not expended for the types of purchases she made.” She therefore argued that the bribery “concerned only state monies and did not fall within the purview of the statute.” We rejected such a construction as contrary to the statute’s text: Despite Westmoreland’s protestations, we find the relevant statutory language plain and unambiguous. By the terms of section 666, when a local government agency receives an annual benefit of more than $10,000 under a federal assistance program, its agents are governed by the statute, and an agent violates subsection (b) when he engages in the prohibited conduct “in any transaction or matter or series of transactions or matters involving $5,000 or more concerning the affairs of’ the local government agency. 18 U.S.C. § 666(b) (Supp. 1984) (emphasis added). Subsection (b) contains nothing to indicate that “any transaction involving $5,000” means “any federally funded transaction involving $5,000” or “any transaction involving $5,000 of federal funds[.]” Westmoreland also made the argument that Lipscomb makes here: “[A]n expansive interpretation [of § 666] ... extends federal power in a manner that, in many instances, the federal interest at stake does not warrant.” The Westmoreland panel responded: Once Congress has spoken, however, we do not sit to judge the wisdom of its action. It is sufficient that Congress seeks to preserve the integrity of federal funds by assuring the integrity of the organizations or agencies that receive them.... [T]he direct involvement of federal funds in a transaction is not an essential element of bribery under section 666(b); the government need not prove that federal monies funded a corrupt transaction. Westmoreland thus held that no connection was required between the federal funds allocated to the county and the supervisor’s illegal conduct. Instead, the only requisite involvement of federal funds was the county’s receipt of more than $10,000 per year. Since Westmoreland, we have sometimes applied its broad reading of § 666 unconditionally. For example, in United States v. Moeller, the government appealed the dismissal of § 666 counts against employees of the Texas Federal Inspection Service (“TFIS”), a cooperative venture of the agriculture departments of Texas and the United States, in which state workers were empowered to conduct federal inspections. Although we said that “there must be some nexus between the criminal conduct and the agency receiving federal assistance,” that nexus was purely textual: It was present when the Texas Department of Agriculture, a government agency for purposes of § 666, received more than $10,000 a year in federal funds, and the defendants, TFIS employees, were agents of that federally-funded agency. Thus Moeller cannot be read to have imposed the extratextual nexus that Lipscomb urges us to engraft on § 666. Some uncertainty seeped into our § 666 jurisprudence as a result of United States v. Marmolejo. There, we upheld the conviction of a county sheriff in Texas who had accepted bribes in return for permitting conjugal visits to a federal prisoner whom the State of Texas, in return for a federal per diem fee, housed in a state prison renovated with federal funds. In addressing whether § 666 gave jurisdiction to prosecute, we noted that “[w]e have previously held that § 666(a)(1)(B) does not require the government to prove that federal funds were directly involved in a bribery transaction, or that the federal monies funded the corrupt transaction.” Nevertheless, when discussing whether conjugal visits were “anything of value” under § 666, we stated that [b]ecause the conduct in this case involves serious acts of bribery by agents of a local government who were carrying out their duties under a Federal program, we conclude that this case is within the scope of conduct Congress intended to encompass with 18 U.S.C. § 666. We did not identify whence we derived any limits on the “scope of conduct Congress intended to encompass.” The dissent argued that Westmoreland interpreted § 666 to reach “only those acts of bribery that could somehow be traced, directly or indirectly, to the integrity of federal program funds.” The Supreme Court granted certiorari to address this argument and affirmed the panel majority’s holding, but beclouded our § 666 jurisprudence in the process. C. The Salinas Speculation and Its Se-quellae: The Funds Focus, Requiring a Further Nexus In reviewing Mannolejo, under the caption Salinas v. United States, the Supreme Court asked whether § 666 is “limited to eases in which the bribe has a demonstrated effect upon federal funds.” The Court stated that “[t]he statute’s plain language fails to provide any basis” for such a limitation and that the legislative history forecloses it. The Court thus agreed with our Mannolejo holding that federal funds need not be directly involved in a violation of § 666. The Court nonetheless obliquely suggested that there might be obstacles to applying § 666 to different facts: We need not consider whether the statute requires some other kind of connection between a bribe and the expenditure of federal funds, for in this case the bribe was related to the housing of a prisoner in facilities paid for in significant part by federal funds themselves. And that relationship is close enough to satisfy whatever connection the statute might required Even so, the Court disposed of any constitutional question: [T]here is no serious doubt about the constitutionality of § 666(a)(1)(B) as applied to the facts of this case. [The briber] was without question a prisoner held in a jail managed pursuant to a series of agreements with the Federal Government. The preferential treatment accorded to him was a threat to the integrity and proper operation of the federal program. Whatever might be said about § 666(a)(l)(B)’s application in other cases, the application of § 666(a)(1)(B) to Salinas did not extend federal power beyond its proper bounds. Since Salinas, the Supreme Court has decided only one more § 666 case: Fischer v. United States, which also sent mixed messages. Echoing Salinas, the Fischer Court described § 666 as “expansive, both as to the conduct forbidden and the entities covered” and read the statute to reveal Congress’s “expansive, unambiguous intent to ensure the integrity of organizations participating in federal assistance programs”—clearly embracing the “corruption focus.” The Court therefore affirmed the conviction of a defendant who had defrauded a municipal hospital authority that participated in the federal Medicare program. In so doing, however, the Court once again mentioned in passing a conceivable constitutional problem: To read the statutory term “benefits” too broadly, the Court cautioned, so as to mean “[a]ny receipt of federal funds,” could “turn almost every act of fraud or bribery into a federal offense, upsetting the proper federal balance.” Justice Thomas, joined in dissent by Justice Sca-lia, likewise warned that “[w]ithout a jurisdictional provision that would ensure that in each case the exercise of federal power is related to the federal interest in a federal program, § 666 would criminalize routine acts of fraud or bribery” and threaten principles of federalism. Although Salinas and Fischer did not unconditionally validate our view that once a local government accepts more than $10,000 per year from the federal government, no further federal interest is needed to justify prosecution under § 666, neither did either of those cases condemn our broad approach. The Salinas Court merely observed in passing that, even if a federal interest were required, such an interest clearly existed in preventing federal prisoners from bribing local jail officials participating in a federal incarceration program. Similarly, the Fischer Court construed a term in § 666 broadly, simply musing that federalism principles might somehow, limit the statute’s sweep. As either a statutory or constitutional matter, then, the Court might be seen as harboring inchoate qualms about whether, for § 666 to apply, there might be some need for a direct interest in the funds involved in the prohibited conduct (or, alternatively, a need for either a nexus between the federal dollars and the offense conduct or an extra-textual jurisdictional element to § 666). Lipscomb argues this inference forcefully, noting that Salinas left open the question whether § 666 “requires some other kind of connection between a bribe and the expenditure of federal funds.” He urges us to overlook West-moreland and answer this question in the affirmative. This, of course, we could not do even if we were so inclined. Mere ruminations in Supreme Court opinions do not empower a subsequent panel of our court to disregard, much less overrule, the holding of a prior panel. And, as we noted just last year, “[w]e are not convinced that Salinas wrought a change upon our earlier precedents.” Because Salinas and Fischer went no further than to advert in dicta to the mere- possibility that the argument now advanced by Lipscomb might someday be favored, we are bound to adhere to Westmoreland’s statutory holding. Likewise, our post-Salinas decisions interpreting § 666 must be read as adhering to this rule. Nevertheless, the cautionary words in Salinas and Fischer, combined with our prior opinions’ silence on the constitutional question, divided the next panel of this court to interpret § 666. The panel majority in United States v. Phillips reiterated Moeller’s requirement of a nexus between the misconduct and the agency (as distinct from a nexus between the misconduct and the federal funds themselves), but added some extra-textual teeth in holding that defendant Phillips, a tax assessor, was not an “agent” of Louisiana’s St. Helena Parish, which received over $10,000 in federal funds, so as to be liable himself under § 666 for putting on his payroll a political ally who then did no work. The panel instead viewed Phillips as an agent of the Louisiana Tax Commission, which received no federal funds, and concluded that the statute did not reach Ms activity. Underlying this definitional question about “agent,” however, ^lurked the majority’s concern that the defendant was too functionally distant from the flow of federal funds to the parish: We know from the Supreme Court’s decision in Salinas that the funds in question need not be purely federal, nor must the conduct in question have a direct effect on federal funds. The statute possibly can reach misuse of virtually all funds of an agency that administers the federal program in question. It is a different matter altogether, however, to suggest that the statute can reach any government employee who misappropriates purely local funds, without regard to how organizationally removed the employee is from the particular agency that administers the federal program., We acknowledge that it is at least arguable, albeit tenuously, that this “organizationally removed” language conflicts with Westmoreland and Moeller, even though the Phillips majority purported to distinguish those two cases factually, and the Phillips panel may be perceived as having favored the “funds focus” for § 666. To the extent that there is a conflict, however, the older case controls, as the Phillips dissenter correctly noted. Only by interpreting “agent” narrowly was the Phillips majority able to avoid the constitutional question. The Phillips dissent read our own precedents as rejecting any nexus requirement whatsoever and took issue with the panel majority’s narrow definition of “agent.” The dissent asserted that a “specific nexus — between Phillips and the federal funds inside Parish coffers — is not required” and furthermore that “it is sufficient that the criminal conduct affect the agency receiving federal assistance: in essence, we have determined that there is an inherent federal interest in insuring that agencies receiving significant amounts of federal funding are not corrupt.” In a nutshell, this is precisely the “corruption focus” that we had firmly adopted in Westmoreland, a focus that has never been overruled either by this court en banc or by the Supreme Court. D. Reyes and Williams: Either Way, § 666 Covers Lipscomb Two cases decided last year demonstrate our continued commitment to applying § 666 to members of municipal and parochial governing bodies. These cases provide additional support for the proposition that § 666 easily reaches Lipscomb’s conduct. In United States v. Reyes, we affirmed the § 666 conviction of city council members who had been bribed to vote in favor of awarding a municipal construction project to a particular contractor. The government noted that a federal loan would have supported the project had it gone forward, but we explicitly declined to rely on that fact, stating instead that federal support of the three city departments involved in the project — the finance, housing, and legal departments — justified prosecuting under § 666: Applying Westmoreland and Moeller ..., we conclude that the connection between federal benefits and the charged conduct is sufficient to uphold Reyes’s convictions under § 666.... Like the county supervisor in Westmore-land and the senior agency officials in Moeller, here the charged criminal conduct related to city council members, who, by voting up or doim on bids, ultimately decide hoiv federal money will be spent, Such an analysis firmly supports Lipscomb’s susceptibility to conviction under § 666: In Dallas, federal money supports the City’s transportation and human services departments — the very agencies of city government that Lipscomb sought corruptly to influence. Reyes reaffirms, as a statutory matter, that whatever nexus § 666 requires — if any — is present in this case. More recently, we decided United States v. Williams' without discussing any jurisdictional element or nexus requirement at all — despite the fact that if there had been a jurisdictional flaw, it would have been incumbent on the Williams panel to address that problem, even sua sponte. Williams involved facts virtually identical to those present in Reyes, in Westmore-land, and here. Williams, a former member of the Jackson, Mississippi City Council, was convicted under § 666 of aiding and abetting the solicitation and acceptance of bribes — specifically, $150,000 in exchange for a re-vote on a cable television license contract. The Williams court did not describe any direct federal fiscal interest at stake in that re-vote. Such questions, however, may not have been briefed or argued in Williams, as the opinion in that case expressly rejected only other challenges-^-those grounded in equal protection, due process, and sufficiency and admissibility of the evidence — to Williams’s conviction. Taken in isolation, Williams has little if any precedential value on the nexus issue, one way or the other. But of course Williams does not stand alone. It is merely the most recent in a series of our opinions — Westmoreland, Moeller, Marmolejo, Reyes, and Williams — that have consistently applied the broad “corruption focus” of § 666. The Phillips panel did construe the term “agent” to avoid the constitutional question, but we cannot do that here: As a textual matter, the term “agent” plainly includes city council members. Westmore-land applied § 666 to a county supervisor; Reyes and Williams both applied it to city council members. Hence Westmoreland’s view of § 666 continues to be the law in this circuit and to preclude a more narrow construction of the statute. Even though we get to the question from different jurisdictional perspectives, Judge Duhé and I are in complete accord on the result for Lipscomb of the foregoing analysis of § 666: He was subject to being tried in federal district court for violating that statute, and he was subject to being convicted by a jury. IV. DID LIPSCOMB RAISE THE CONSTITUTIONAL ISSUE? Before addressing the constitutional problem that Lipscomb’s statutory-construction argument presages (and writing for myself alone, although supplementing Judge Smith’s analysis), I must make three observations on our intrapanel disagreement over whether the constitutional issue is properly before us. To me, this debate: (1) is more semantical than substantive, (2) is in tension with controlling Supreme Court precedent, and (3) overlooks the real nature of the constitutional question at issue. Semantics first: As the Supreme Court recently said, “jurisdiction ... is a word of many, too many, meanings.” This imprecision is one source of our panel’s split here. Judge Duhé reads “jurisdiction” in the pleadings, briefs, and record to mean adjudicative jurisdiction only — the authority of federal courts to hear only those categories of cases (subject-matter jurisdiction) authorized by Congress, between those categories of persons (personal jurisdiction) permitted by the Constitution. But in the context of the expressly constitutional arguments that Lipscomb sometimes makes, Judge Smith and I read his use of “jurisdiction” — at least on those occasions — to mean legislative jurisdiction, the “authority” of Congress “to make its law applicable to particular persons or activities.” Lipscomb also uses the ambiguous phrase “federal jurisdiction,” which could be either adjudicative or legislative. That ambiguity is not only terminological, but also conceptual. To state the obvious, legislative jurisdiction flows from the Constitution to the Congress and limits, in today’s context, the subject matter and the classes of persons that Congress may regulate by statute. In contrast, adjudicative jurisdiction generally flows from Congress to the courts as grants of subject-matter jurisdiction, grants made by Congress in enacting laws pursuant to its power to constitute inferior federal courts. In the instant context, the judicial power extends constitutionally to cases arising under federal criminal laws. Consequently, a court’s adjudicative jurisdiction to convict a defendant of a federal crime cannot exist in the absence of Congress’s legislative jurisdiction to criminalize the particular conduct of which the particular defendant is accused. The reach of Congress’s legislative jurisdiction, of course, is sometimes bounded by structural constitutional provisions. For example, grants of jurisdiction are limited by the Necessary and Proper Clause, which covers laws that “carry into Execution ... all other powers vested by this Constitution in the Government of the United States or in any Department or Officer thereof.” I cannot even imagine how it could be “necessary and proper” to the exercise of either the judicial power or the power to constitute inferior courts for us to have adjudicative jurisdiction over a case implicating a statute that Congress lacked the legislative jurisdiction to enact. It should go without saying, therefore, that our subject-matter jurisdiction has constitutional as well as statutory limits: It involves “the courts’ statutory or constitutional power to adjudicate the case.” To repeat, then: A federal forum simply must lack adjudicative jurisdiction to hear a case based on a federal statute that Congress lacked the legislative jurisdiction (translation: constitutional power or authority) to apply to the situation in question. If I am correct in my position that this case implicates our constitutional duty, at every level and at every stage of the proceedings, to ensure the existence of our adjudicative jurisdiction, then that duty trumps the canon of constitutional avoidance that Justice Brandéis discussed in Ashwander v. TVA, a canon that on other occasions I have dutifully obeyed. At the risk of exposing my own intellectual shortcomings, then, I confess that neither semantically nor substantively can I understand the distinction, which Judge Duhé detects in Lipscomb’s pleadings and briefs, between adjudicative, subject-matter jurisdiction and legislative jurisdiction (structural constitutionality) — a distinction that is clear to Judge Duhé but in this case remains blurred to me. If a successful as-applied challenge to the constitutionality of § 666 would limit Congress’s legislative jurisdiction, i.e., would identify someone or some act beyond Congress’s authority, it cannot help but limit our adjudicative jurisdiction to the same degree. In this sense, because Lipscomb appears to raise a structural jurisdictional challenge, I question whether his is the kind of constitutional argument that may be waived through delay or disuse; if it “involves a court’s power to hear a case, [it] can never be forfeited or waived.” As I read them, Lipscomb’s pleadings and briefs do raise — and thus do not waive — the constitutional issue. Rather, they question both Congress’s legislative jurisdiction (constitutional authority) to enact § 666 and our adjudicative power to apply § 666 here. Lipscomb has raised a classic challenge to subject-matter jurisdiction: He “argues that the extension of federal jurisdiction over acts such as [his] would exceed the power of Congress.” My belief that we should consider this argument finds support in Salinas itself. There the Supreme Court easily undertook to determine whether § 666 was constitutional, and squarely held that it was, as applied, despite the fact that neither we nor the district court had addressed the statute’s constitutionality. In support of its “holding,” the Court explained: [sjtatutes should be construed to avoid constitutional questions, but this interpretative canon is not a license for the judiciary to rewrite language enacted by the legislature. Any other conclusion, while purporting to be an exercise in judicial restraint, would trench upon the legislative powers vested in Congress .... These principles apply to the rules of statutory construction we have followed to give proper respect to the federal-state balánce.... [W]e cannot press statutory construction to the point of disingenuous evasion even to avoid a constitutional question. This is why, with all due respect, I find it odd, as we labor to interpret 18 U.S.C. § 666, for Judge Duhé to urge obeisance to the Ashwander canon, which the Supreme Court itself in Salinas first acknowledged and then declined to observe or apply. To the extent that the real question is whether Lipscomb adequately raised constitutionality, I trust Judge Duhé would concede two premises: first, that Lipscomb urged the district court (and this one) so to construe § 666 as to avoid a serious and identified constitutional flaw; and second, that this panel has unanimously concluded that we cannot so construe the statute. Starting with these two premises, I cannot avoid the conclusion that Lipscomb did raise the constitutional flaw. By construing the statute as all three of us do, we are sailing into the very waters that Lipscomb warned us were constitutionally uncharted. I knowingly and willfully proceed to endeavor to chart them. V. AS-APPLIED CONSTITUTIONALITY UNDER DOLE We review the constitutionality of a federal statute de novo. . My solo review here will focus on whether § 666 is necessary and proper to the spending power, but the proper foundation for that analysis is a review of the Supreme Court’s spending-clause jurisprudence. That jurisprudence has focused on whether Congress may condition grants of federal funds. Even if § 666 is not a conditional-grant statute — a conclusion of which I am less certain than is Judge Smith — the conditional-grant cases establish both that “internal limits on congressional spending power are difficult to discern” and that, to whatever extent the Tenth Amendment is an external limit on the spending power, that Amendment does not function as “a prohibition on the indirect achievement of objectives which Congress is7 not empowered to achieve directly.” The Tenth Amendment thus is not as great an obstacle to the necessity and propriety of § 666 as Judge Smith believes it to be. A. Conditional-Grant Precedents Congress likely enacted § 666 pursuant to the Spending Clause of the Constitution. Under that clause, it is settled, Congress may regulate the states by conditioning grants. Cases on such conditions have established that the structural limits on federal power that often arise in the commerce-clause context do not operate with the same force against conditional-grant provisions. United States v. Butler, for example, is still good law for its announcement that Congress’s spending power, like its power to tax, is “to provide for the general welfare,” and is therefore untrammeled by the specific grants of legislative power found elsewhere in Article I, Section 8: While, therefore, the power to tax is not unlimited, its confines are set in the clause which confers it, and not in those of section 8 which bestow and define the legislative powers of the Congress. It results that the power of Congress to authorize expenditure of public moneys for public purposes is not limited by the direct grants of legislative power found in the Constitution. Although the Butler Court did hold that the Tenth Amendment cabined Congress’s spending power, the Court quickly abandoned this view, in Oklahoma v. United States Civil Service Commission, which rejected a constitutional challenge to the Hatch Act. That Act then forbade political activities by any “officer or employee of any State or local agency whose principal employment is in connection with any activity which is financed in whole or in part by loans or grants made by the United States.” Oklahoma and its State Highway Commissioner challenged the Civil Service Commission’s attempt to force on the State the choice between dismissing the Commissioner, who had engaged in political activities, or forgoing highway funds in the amount of twice the commissioner’s salary. The Court responded: "While the United States is not concerned with and has no power to regulate local political activities as such of state officials, it does have power to fix the terms upon which its money allotments to states shall be disbursed. The Tenth Amendment does not for-hid the exercise of this power in the tvay that Congress has proceeded in this case.... [T]he Tenth Amendment has been consistently construed “as not depriving the national government of authority to resort to all means for the exercise of a granted power which are appropriate and plainly adapted to the permitted end.”.... The offer of benefits to a state ... dependent upon cooperation by the state with federal plans, assumedly for the general welfare, is not unusual. Oklahoma, the Court said, could evade the condition by the “simple expedient” of not yielding to the enticement of federal funds. The apex of the Court’s conditional-grant jurisprudence is South Dakota v. Dole, which involved a statute conditioning a small portion of each state’s federal highway aid on the state’s establishing a minimum drinking age. The Court upheld the drinking-age requirement as an exercise of Congress’s Spending-Clause authority to condition federal grants. The Court also announced that when Congress chooses to go beyond its enumerated powers, and to use its spending power “to further broad policy objectives by conditioning receipt of federal monies upon compliance with federal statutory ... directives,” the statutory condition must itself meet four conditions, the failure to meet any one of which might render a statute unconstitutionally broad. B. The Dole Test Is Instructive Here Given Dole’s, context, applying its test to § 666 could be trebly problematic. First, like Judge Smith, two district courts have concluded that § 666 is not a conditional-grant statute at all, because it does not require the state (here, Texas) or its political subdivision (here, Dallas) to do anything. As the court in United States v. Cantor noted, § 666 “does not impose a condition on the receipt of federal funds. The statute neither requires a state’s compliance with federal ... directives nor prevents state action.” Like the Cantor court, however, I believe that this lack of direct effect on states and localities actually supports the statute’s constitutionality. Furthermore, the Supreme Court has not held that, for a statute to be a conditional-grant provision and stand or fall under a Dole analysis, the statute must require states or localities either to take or to refrain from taking any action. Dole may describe Congress’s spending power generally, not just its power to condition grants. Second, § 666 is a freestanding ban: It neither grants any funds nor takes part in a broader funding statute. This fact has prompted the objection that its criminal sanction cannot be a condition. Although superficially appealing, this argument elevates form too highly over substance. The anticorruption principle in § 666 applies equally to every federal dollar granted, and § 666 logically cuts across all federal grants to states and localities. To require Congress to insert a mini-§ 666 into every chapter of the United States Code that authorizes intergovernmental financial assistance would constitute excessive scrupulosity. Third, several judges have objected that Congress’s spending power cannot include the power to criminalize conduct by third parties, and that Dole therefore cannot apply. (This argument begs the broader question, which I address below, whether § 666 is necessary and proper to the spending power.) Many courts, including this one in Phillips, have nevertheless interpreted § 666 using Dole’s factors. Therefore, although we may debate whether the § 666 peg fits the conditional-grant hole, I shall test it under the four prongs of Dole. C. The Dole Analysis Dole first requires that “exercise of the spending power must be in pursuit of the general welfare.” In assessing whether this is so, Dole cautions, “courts should defer substantially to the judgment of Congress.” Congress has stated that the purpose of § 666 is to “protect the integrity of the vast sums of money distributed through Federal programs from theft, fraud, and undue influence by bribery.” Mindful of the deference due this judgment, I accept that Congress easily could have thought that § 666 advanced the general welfare by protecting the federal fisc and by ensuring that state and local decisions regarding federal programs are not made by corrupt officials. I do not doubt, then, that Congress enacted § 666 “in pursuit of the general welfare.” Second, Dole warns that “if Congress desires to condition the States’ receipt of federal funds, it must do so unambiguously ..., enabling] the States to exercise their choice knowingly, cognizant of the consequences of their participation.” Even though § 666 does not require the states to act, it does make state and local government officers criminally liable for specific misdeeds. Thus the states arguably have a dignity interest at stake, and if so, they have a right to know the threat to that interest that § 666 would pose — and the language of § 666, which is anything but ambiguous, surely lets them know. To the extent that § 666 is a conditional-grant statute, both the grant (of $10,000 or more in federal funds) and the condition (criminalizing official bribery and theft) are pellucid. I see little danger that a state or locality that receives federal funds could mistake the potential of § 666 to criminalize conduct by its officials. Third, Dole mandates that conditions on federal spending be related “to the federal interest in particular national- projects or programs,” or that conditions “bear some relationship to the federal spending.” “The required degree of this relationship is one of reasonableness or minimum rationality.” It suffices here to observe that many courts have held that § 666 is reasonably related to the federal interest in safeguarding federal dollars from control of dishonest administrators, and that § 666 therefore passes spending-power muster. At least one court has so concluded when the offense conduct did not involve federal funds. Some of the other decisions arriving at this conclusion, however, may have dismissed facial, rather than as-applied, challenges to the statute; and other cases-have affirmed convictions for conduct that implicated federal funds more directly than did Lipscomb’s actions here. Because Dole’s relatedness inquiry merges with my analysis of whether applying § 666 is necessary and proper to the spending power, I discuss both questions together in detail below. Fourth, in Dole’s final prong, the Court cautioned that “other constitutional provisions may provide an independent bar to the conditional grant of federal funds.” Yet the Court then reiterated its Oklahoma holding that “a perceived Tenth Amendment limitation on congressional regulation of state affairs did not concomitantly limit the range of conditions legitimately placed on federal grants,” Rather, the “independent bar” simply means that Congress may not use its spending power “to induce the States to engage in activities that would themselves be unconstitutional.” In this case, no action by Texas or Dallas is alleged to be unconstitutional, so the fourth Dole prong is plainly not at issue. In sum, to the extent that Dole controls whether § 666 can apply here, the only problem lies in the third part of the Dole test: reasonable relationship to a federal interest. Because this reasonably-related prong of Dole is a specific application of the more general test for whether an act of Congress is necessary and proper to an enumerated power, I treat these questions together. VI. AS-APPLIED CONSTITUTIONALITY UNDER McCULLOCH In addition to assigning Congress the spending power, which brings with it the power to condition grants, the Constitution also gives Congress the power “[t]o make all Laws which shall be necessary and ;proper for carrying into Execution” the powers expressly delegated to the federal government. Prosecuting Lipscomb under § 666 is therefore constitutional if § 666 is “necessary and proper” to Congress’s spending power. A. McCulloch and the Necessary and Proper Clause In testing for necessity and propriety, courts should remain mindful of Justice John Marshall’s prescient explanation, in McCulloch v. Maryland, of what “necessary and proper” means: Let the end be legitimate, let it be within the scope of the constitution, and all means which are appropriate, which are plainly adapted to that end, which are not prohibited, but consist with the letter and spirit of the constitution, are constitutional. Importantly for the instant case, Marshall derived an expansive meaning of “necessary” from the principle that Congress can derive from its enumerated powers the power to impose criminal sanctions. From the enumerated power to “establish Post Offices and post Roads,” Congress had “inferred the right to punish those who steal letters from the post-office, or rob the mail.” In other words, Congress’s postal power carried with it the ability to impose criminal penalties to protect federal interests advanced' by that power. To the McCulloch Court, this example demonstrated that “necessary” has a range of meanings, including “needful, requisite, essential, or conducive to.” It was through the lens of this broad construction of the Necessary and Proper Clause that Marshall saw justification for Congress’s creation of the national bank, the power to create which is nowhere enumerated in Article I. Whether that broad construction justifies applying § 666 here depends on Congress’s intent in enacting the statute, as well as on the nature of the federal interest embodied in this case and the relationship between that interest and Lipscomb’s conduct. B. Legislative History History often tells us why Congress deemed a statute necessary and proper. Not so for § 666, however, because it was enacted as part of an omnibus spending bill of the type that makes the search for legislative history Sisyphean. What history exists is multilayered, sparse, equivocal, and even mysterious. By no means, I respectfully submit, is it capable of supporting Judge Smith’s contention that “Congress did not find it necessary that § 666 be applied in cases not involving federal funds or programs.” 1. The 1986 Technical Amendment We owe the current language of § 666 to the Criminal Law and Procedure Technical Amendments Act of 1986. As that Act’s title suggests, and as we recognized in Westmoreland, Congress did not intend the Act to change § 666 substantively in ways that would affect our reading of it here. This is important, because the 1986 amendment rewrote language that reveals how Congress would have answered our constitutional question in 1984. 2. The 198k Enactment As first enacted, § 666(b) read: Whoever, being an agent of an organization, or of a State or local government agency ... [that receives more than $10,000 a year in federal funds], solicits, demands, accepts, or agrees to accept anything of value from a person or organization other than his employer or principal for or because of the recipient’s conduct in any transaction or matter or a series of transactions or matters involving $5,000 or more concerning the affairs of such organization or State or local government agency, shall be imprisoned. .. , The emphasized phrase strongly suggests that in 1984 Congress believed it necessary and proper for § 666 to reach bribery that had no relation to federal funds. 3. The 1988 Report To counter the broad original and current language of § 666, Judge Smith relies heavily on a Senate Judiciary Committee report, but this report described a different bill that never became law. As eventually enacted, § 666 was a small part of a large crime bill which was engrafted on a huge omnibus spending bill that funded many departments and agencies. None of this bill’s reports, written as they were by the Appropriations Committees, gives context for § 666, a criminal statute which, of course, appropriated no funds. Section 666 as enacted was identical to a provision in the Comprehensive Crime Control Act of 1984, which passed the Senate but never made it out of the House Judiciary Committee on its own and evidently had to piggyback on the omnibus spending bill to gain legislative momentum. The Senate report on the crime bill, printed in 1983, can be taken as an authoritative statement of the Senate Judiciary Committee’s intent for what became § 666. It is tenuous at best, however, to rely, as does Judge Smith, solely on one committee report — on a wholly separate bill — as stating the views of the entire Congress. The questionable probative weight of the Senate report aside, that report is still not determinative here, for the evidence goes both ways. The relevant passage is titled “Part C — Program Fraud and Bribery,” and states that § 666 is designed to create new offenses to augment the ability of the United States to vindicate significant acts of theft, fraud, and bribery involving Federal monies that are disbursed to private organizations or State and local governments pursuant to a Federal program. The report notes, however, that under the prior law banning theft of federal property, prosecuting was often impossible because title has passed to the recipient [government] before the property is stolen, or the funds are so commingled that the Federal character of the funds cannot be shown. This gives rise to a serious gap in the law, since even though title to the monies may have passed, the Federal Government clearly retains a strong interest in assuring the integrity of such program funds. Even though the report’s emphasis on program funds would support a narrow reading of the necessity and propriety of § 666, its emphasis on commingling supports a broad one. In fact, the Senate Judiciary Committee’s most explicit direction actually suggests that the Committee intended to limit the scope of § 666, but in a way that still would cover Lipscomb-like conduct: The Committee intends that the term “Federal program involving a grant, a contract, a subsidy, a loan, a guarantee, insurance, or another form of Federal assistance” be construed broadly, consistent with the purpose of this section to protect the integrity of the vast sums of money distributed through Federal programs from theft, fraud, and undue influence by bribery. However, the concept is not unlimited. The term “Federal program” means that there must exist a specific statutory scheme authorizing the Federal assistance in order to promote and achieve certain policy objectives. Thus, not every Federal contract or disbursement of funds would be covered. For example, if a government agency lawfully purchases more than $10,000 in equipment from a supplier, it is not the intent of this section to make a theft of $5,000 or more from the supplier a Federal crime. Thus one of the two lines that the Senate Judiciary Committee expressly drew — to exclude theft from a supplier from the coverage of § 666 — would not exclude Lipscomb’s conduct, quintessentially “undue influence by bribery.” 4. The Specified Cases The immediate next sentence in the report, subject to much exegesis by Judge Smith, states: “It is, however, the intent [‘of this section’] to reach thefts and bribery in situations of the types involved in the Del Toro, Hinton, and Mosley cases.” With continued due respect to Judge Smith, I do not discern in this sentence any clear direction to us. Both Hinton and Mosley sustained convictions of bribed local officials whom courts considered to be federal officials under the prior bribery statute because they exerted federal authority and controlled the disbursement of federal funds. Simple logic dictates that just because the Committee intended § 666 in part to codify these cases does not meant that it sought to limit § 666 to these cases exclusively. In Del Toro, the federal interest was more attenuated. The defendants were convicted of bribing a New York City official to ensure that they would supply office space to a city program that was eligible for federal funds. The Del Toro court reversed these bribery convictions, noting that even if the official had succeeded in provisionally securing the lease as desired, three local agencies would have had to approve the lease before the city could apply to the federal government for funds, so that “[t]here were no existing committed federal funds for the purpose.” The Senate Committee’s intent to overrule Del Toro thus reflects that the Committee thought it necessary and proper for § 666 to reach bribery even before the federal government had committed funds. The most that can be concluded from the report, then, is that the Senate Judiciary Committee delimited the scope of § 666 in part by seeking to exclude theft from suppliers but to include bribery of officials running programs that might receive federal funds. I do not see the report as shedding much light on our question. Lipscomb’s conduct falls into a middle ground that the report simply does not address. 5. The 1981 Bill The plot thickens still further when an effort is made to verify the assertion in the 1988 report that the language of § 666 was derived from a 1981 bill that never became law. The 1981 bill and its report emerged from Senate Judiciary — the same committee that later wrote the 1983 bill and report. The Committee omitted from both the 1983 bill and the 1984 act, however, the very language in the 1981 bill that would have answered our question: (c) Jurisdiction. — There is federal jurisdiction over an offense described in this section if— (6) the public servant is an agent of a State or local government charged by a federal statute, or by a regulation issued pursuant thereto, with administering monies or property derived from a federal program, and the official action or legal duty [with respect to which the bribe is taken] is related to the administration of such program. Thus, in 1983, the Senate Judiciary Committee had in hand — and even mentioned — a two-year-old bill that would have required a federal interest or nexus as a jurisdictional predicate. Yet the 1983 bill and 1984 enactment contained none of that language or anything similar. The reason for that absence is unclear. In 1981, when the Committee clearly sought to require a federal nexus, it had sufficient command of the English language to do so. To suppose that the Committee lost that faculty over either two or four years is ludicrous. Section 666 as enacted and amended, therefore, might have reflected a change in the Senate Judiciary Committee’s view on whether to require a federal nexus, but we cannot say this with certainty: For all we know, the Committee might well have sought to exercise federal criminal jurisdiction up to its constitutional limits, leaving the issue to the courts to decide. C. The Views of Other Courts Whatever the reason for § 666’s silence on this question, the courts have struggled to produce the answer. Some district courts have tested § 666 against the Tenth Amendment, treating the statute as an emanation of the spending power, and have come to varying conclusions. Additionally, four of our fellow appellate courts have examined the sweep of § 666, either as a statutory matter or a constitutional one, and are also divided. Lipscomb relies on United States v. Zwick, in which the Third Circuit declined to apply § 666 to conduct such as his: Interpreting § 666 to have no federal interest requirement produces serious concerns as to whether Congress exceeded its power under the Spending Clause in enacting this statute. See McCormack, 31 F.Supp.2d at 187-89. To pass muster under the Spending Clause, legislation regulating behavior of entities receiving federal funds must, among other things, be based upon a federal interest in the particular conduct. See South Dakota v. Dole, 483 U.S. 203, 207, 107 S.Ct. 2793, 97 L.Ed.2d 171 (1987). Applying § 666 to offense conduct, absent evidence of any federal interest, would appear to be an unconstitutional exercise of power under the Spending Clause. To avoid this supposed constitutional problem, the Zimck court believed that to read § 666 literally is to err, and