U.S. v. PHILIP MORRIS USA, INC

SENTELLE, Circuit Judge.

A group of cigarette manufacturers and related entities ("Appellants") appeal from a decision of the District Court denying summary judgment as to the Government's claim for disgorgement under the Racketeer Influenced and Corrupt Organizations Act ("RICO" or "the Act"), 18 U.S.C. §§ 1961-68. The relevant section of RICO, 18 U.S.C. § 1964(a), provides the District Courts jurisdiction only for forward-looking remedies that prevent and restrain violations of the Act. Because disgorgement, a remedy aimed at past violations, does not so prevent or restrain, we reverse the decision below and grant partial summary judgment for the Appellants.

I. Background

In 1999 the United States brought this claim against appellant cigarette manufacturers and research organizations, claiming that they engaged in a fraudulent pattern of covering up the dangers of tobacco use and marketing to minors. The Government sought damages under the Medical Care Recovery Act ("MCRA"), 42 U.S.C. §§ 2651-53, and the Medicare Secondary Payer ("MSP") provisions of the Social Security Act, 42 U.S.C. § 1395y to recover health-care related costs Appellants allegedly caused. The United States also claimed that Appellants engaged in a criminal enterprise to effect this cover-up, and sought equitable relief under RICO, including injunctive relief and disgorgement of proceeds from Appellants' allegedly unlawful activities. The Government sought this relief under 18 U.S.C. § 1964(a), which gives the District Court jurisdiction

to prevent and restrain violations of [RICO] by issuing appropriate orders, including, but not limited to: ordering any person to divest himself of any interest, direct or indirect, in any enterprise; imposing reasonable restrictions on the future activities or investments of any person, including, but not limited to, prohibiting any person from engaging in the same type of endeavor as the enterprise engaged in, the activities of which affect interstate or foreign commerce; or ordering dissolution or reorganization of any enterprise. . . .

18 U.S.C. § 1964(a).

Appellants moved to dismiss the complaint in 2000. The District Court did dismiss the MCRA and MSP claims, but allowed the RICO claim to stand. United States v. Philip Morris, Inc., 116 F.Supp.2d 131, 134 (D.D.C. 2000).

Section 1964(a) conferred jurisdiction on the District Court only to enter orders "to prevent and restrain violations of the statute." In considering wither disgorgement came within this jurisdictional grant, the court relied on a decision of the Second Circuit, the only circuit then to have considered "whether . . . disgorgements . . . are designed to `prevent and restrain' future conduct rather than to punish past conduct." United States v. Carson, 52 F.3d 1173, 1182 (2d Cir. 1995) (emphasis in original). After noting that "RICO has a broad purpose [and] the legislative history of § 1964 indicates that the equitable relief available under RICO is intended to be `broad enough to do all that is necessary,'" id. at 1181, the Carson court went on to observe that it did not see how it could "serve any civil RICO purpose to order disgorgement of gains ill-gotten long ago. . . ." Id. at 1882. The portion of Carson relied upon by the District Court in the present controversy suggested that disgorgement might "serve the goal of `preventing and restraining' future violations," but flatly held that the remedy would not do so "unless there is a finding that the gains are being used to fund or promote the illegal conduct, or constitute capital available for that purpose." Id. at 1182. The Second Circuit went on to caution that disgorgement would be better justified under this analysis where the "gains [were] ill-gotten relatively recently." Id. The District Court accepted the Second Circuit's suggested holding that the appropriateness of disgorgement depends on whether the proceeds are available for the continuing of the criminal enterprise, but ruled that the question was premature, and denied the motion for dismissal on the RICO-disgorgement claim. Philip Morris, 116 F.Supp.2d at 151-52. Neither party sought leave to file an interlocutory appeal of that ruling.

While the Carson language may appear to be dicta, the Second Circuit remanded for determination of which disgorgement amounts were sufficiently directed to prevention and restraint to qualify under § 1964(a), thus treating the language on availability of disgorgement as essential to the outcome of the case, and therefore a holding. Some other courts have followed Carson. See, e.g., Richard v. Hoechst Celanese Chem. Group, Inc., 355 F.3d 345, 354 (5th Cir. 2003) (observing that "the Second Circuit noted that disgorgement is generally available under § 1964"); United States v. Private Sanitation Indus. Ass'n, 914 F.Supp. 895, 901 (E.D.N.Y. 1996) ("[T]he disgorgement in this case is clearly directed towards the prevention of future illegal conduct, and is therefore a permissible remedy for civil RICO violations under the limitations imposed by Carson.").

The case proceeded, and the Government sought disgorgement of $280 billion that it traced to proceeds from Appellants' cigarette sales to the "youth addicted population" between 1971 and 2001. This population includes all smokers who became addicted before the age of 21, as measured by those who were smoking at least 5 cigarettes a day at that age.

After discovery, Appellants moved for summary judgment on the disgorgement claim arguing that (1) disgorgement is not an available remedy under § 1964(a), (2) even if disgorgement were available, the Government's model fails the Carson test for permissible disgorgement that will "prevent and restrain" future violations, and (3) even if disgorgement were available, the Government's proposed model is impermissible because it includes both legally and illegally obtained profits in violation of SEC v. First City Financial Corp., 890 F.2d 1215 (D.C. Cir. 1989). The District Court denied this motion in a memorandum order designated "# 550." United States v. Philip Morris USA, Inc., 321 F.Supp.2d 72 (D.D.C. 2004). On motion of the defendants, the District Court certified Order # 550 for interlocutory appeal pursuant to 28 U.S.C. § 1292(b). That section provides for interlocutory appeal where a district judge has certified that "an order not otherwise appealable . . . involves a controlling question of law as to which there is substantial ground for difference of opinion and that an immediate appeal from the order may materially advance the ultimate termination of litigation. . . ." Under § 1292(b), the Court of Appeals may then decide whether to permit the appeal to be taken from such order. In the present case, we allowed the appeal.

II. Analysis

A. Scope of Review

At the outset, the Government urges that our review should be limited to the narrow question of whether the disgorgement it seeks is consistent with the standards of Carson, not whether disgorgement vel non is an available remedy under civil RICO. The Government bases this argument on the theory that the order on appeal — that is the memorandum order denying "defendants' motion for partial summary judgment dismissing the Government's disgorgement claim" — was reiterating a prior order on the general question of availability of disgorgement Further, the Government argues, the order spoke anew only to the measure of disgorgement, assuming such disgorgement to be otherwise available. In support of its proposed limitation of our review, the Government relies upon Yamaha Motor Corp., USA v. Calhoun, 516 U.S. 199, 116 S.Ct. 619, 133 L.Ed.2d 578 (1996). In Yamaha, the Supreme Court dealt with the breadth of review properly conducted by a court of appeals under 28 U.S.C. § 1292(b). Id. at 204, 116 S.Ct. 619. The Government selectively quotes from Yamaha the sentence that, "The court of appeals may not reach beyond the certified order to address others made in the case." Id. at 205, 116 S.Ct. 619. Based on this sentence, the Government then argues that because the first order denying a motion to dismiss had dealt with the question of the availability of disgorgement, this certified interlocutory review of the subsequent summary judgment order is restricted to the new theory considered by the court on that occasion — that the disgorgement the Government pursued exceeded the standard available for such disgorgement as set by the Second Circuit in Carson.

Unfortunately for the Government's position, the Yamaha opinion did not end with the sentence upon which the Government relies. The Supreme Court went on to say in the same paragraph: "But the appellate court may address any issue fairly included within the certified order because `it is the order that is appealable, and not the controlling question identified by the district court.'" Id. (emphasis in original) (quoting 9 J. MOORE B. WARD, MOORE'S FEDERAL PRACTICE § 110.25[1] at 300 (2d ed. 1995) and citing 16 C. Wright, A. Miller, E. Cooper, E. Greshman, Federal Practice Procedure § 3929 at 144-45 (1977)). Appellants' motion below was for "Summary Judgment Dismissing the Government's Disgorgement Claim," and granting this motion would have resulted in complete dismissal of the Government's claim for disgorgement with prejudice. See Appellee's App. at 19, 79. Thus the District Court's denial was on the question of whether disgorgement would be allowed at all, and we may review it as such regardless of the grounds the District Court gave for its decision. In the memorandum accompanying its denial of this motion, evidencing an accurate understanding of the summary judgment standard provided by Rule 56 of the Federal Rules of Civil Procedure, the District Court noted that "summary judgment is appropriate if the pleadings, depositions, answers to interrogatories, and admissions on file, together with the affidavits, if any, show that there is no genuine issue as to any material fact and that the moving party is entitled to judgment as a matter of law." Philip Morris, 321 F.Supp.2d at 74 (citing Fed.R.Civ.P. 56(c)). Significantly, the court further noted that "Defendants argue that any disgorgement which might be ordered upon a finding of liability must be limited by both the text of Section 1964(a) itself and the holding in United States v. Carson . . . interpreting that section." Philip Morris, 321 F.Supp.2d at 74 (emphasis added). Thus the court clearly implied the possibility that none might be ordered, and that statutory issues outside Carson were before the court.

Our dissenting colleague argues that the availability of the disgorgement claim vel non is not before us because Appellants did not fully restate their earlier arguments in their motion, but only expressed their reservation in a footnote referencing the District Court's prior rejection of their position. While it is true, as our colleague reminds us, that we have held that a "litigant does not properly raise an issue by addressing it in a `cursory fashion,' with only `bare-bones arguments,'" Cement Kiln Recycling Coalition v. EPA, 255 F.3d 855, 869 (D.C. Cir. 2001) (per curiam), our prior holdings on that subject have been in very different contexts. In Cement Kiln, for example, and in Wash. Legal Clinic for the Homeless v. Barry, 107 F.3d 32, 39 (D.C. Cir. 1997), relied upon by the dissent, we were determining whether an issue was properly before us that had been raised in no other fashion. In the present case, we are reviewing a summary judgment decision, presumably according to the standards set forth by the Supreme Court in such decisions as Yamaha, and the issue in question was clearly decided by the District Court in the first rejection of the motion to dismiss. The issue was called to the attention of the court as a necessary antecedent in the second summary judgment order, now under direct review, and expressly pointed out in the footnote which our colleague disdains. Furthermore, the motion leading to the order presently before us sought summary judgment of dismissal of the disgorgement claim, not simply a limitation to such disgorgement as might have been supported by the Carson test or other factors. Given the Supreme Court's plain teaching in Yamaha, particularly its adoption from a learned treatise of the language "it is the order that is appealable, and not the controlling question identified by the district court," Yamaha, 516 U.S. at 205, 116 S.Ct. 619 ( see other authorities, supra), Cement Kiln and Barry have no applicability. Yamaha controls. We therefore proceed to review the denial of summary judgment, under the usually applicable standards, not simply the sole question to which the Appellees and the dissent would restrict us.

Our dissenting colleague suggests that we are limited by "our general policy of declining to consider arguments not made to the district court in the motion leading to the order under appeal." Dissent at 1211. We know of no such "general policy" that the particular issue addressed has to have been raised in the particular motion. Rather, we understand our general policy to be following the instructions of the Supreme Court that we are to "address any issue fairly included within the certified order." Yamaha, 516 U.S. at 205, 116 S.Ct. 619. Insofar as our colleague's differing understanding rests on United States v. British Am. Tobacco (Invs.) Ltd., 387 F.3d 884, 892 (D.C. Cir. 2004) (citing United States v. Hylton, 294 F.3d 130, 135-36 (D.C. Cir. 2002)), cited by Dissent at 1213, we do not read that case as supporting a general policy that limits consideration to those arguments raised in the particular motion leading to the certified order, as opposed to being "fairly included" within that order, or even to address the point. The court in British American Tobacco held only that an intervenor that had raised a privilege issue with respect to an entire collection of documents at one stage of the litigation, but that failed to participate at all in later proceedings focused on one of the documents, despite having notice, had not adequately preserved its objection as to that single document. 387 F.3d at 887-88. It had nothing to do with the scope of review on an interlocutory appeal under § 1292(b). Neither it nor Hylton dealt in any fashion with the breadth of interlocutory review, nor was establishing any standard for the papers in which an argument must have been raised. Each rejected an attempt by an appellant to raise a new ground for the first time on appeal. Appellants before us raised and preserved their argument as set forth in the text above. We read nothing in British American Tobacco or Hylton to suggest a general policy barring our review under the Yamaha standard.

We find no history of such a general policy that would bar us from considering questions logically antecedent and essential to the order under review. Especially is this so given the Supreme Court's instructions in Yamaha that we are to "address any issue fairly included within the certified order." That must include at least issues that are logically interwoven with the explicitly identified issue and which were properly presented by the appellant. Even ignoring the apparent allusion to the broader issue of summary judgment preserved in the caption of the motion, the relief sought, and the footnote provided above, it is difficult to see how we could establish such a policy that would cause us to affirm a decision denying summary judgment when a ground compelling its grant is fairly encompassed within the order. Our colleague's interpretation of general policy would seem to compel us to return for trial a case before us for review of a denial of summary judgment, no matter how plain the absence of substantial question of material fact, on the grounds that the denial of summary judgment had been based on rejection of some other reasoning in a previous motion, even though the trial court had earlier erred in denying the first motion to dismiss — even when the appellant had called that denial to the court's attention in the caption of its motion, and a proposed order accompanying the second motion.

Our dissenting colleague finds in Yamaha support for the proposition that "the only issues `fairly included' within a certified order are those decided in the district court's accompanying memorandum. . . ." Dissent at 1213. We understand the law to be, as suggested in Yamaha, that issues are not decided in memoranda at all, but rather in orders. Therefore, consistent with Yamaha, we review orders, not memoranda. Our colleague asserts that in Yamaha the Court "found `fairly included' an issue that the district court had resolved in the same opinion in which it decided the issue identified as the controlling question of law." Dissent at 1213. While this may well be the case, the Supreme Court not only did not stress that circumstance, it did not even mention it. Indeed, we note that our colleague had to repair to the unpublished opinion of the District Court to discover the truth of his proposition. We seriously doubt that the Supreme Court intended to establish a precedent that difficult to discover, let alone apply.

Nothing in United States v. Stanley, 483 U.S. 669, 107 S.Ct. 3054, 97 L.Ed.2d 550 (1987), is to the contrary. The passage relied upon by our dissenting colleague to the effect that courts considering interlocutory appeals under § 1292(b) should "not consider matters that were ruled upon in other orders," id. at 677, 107 S.Ct. 3054, did not address a situation like the one before us. Here the order appealed from reiterated, and totally depended upon, an issue fairly encompassed within the motion before that court and the order now before us. In Stanley, the court of appeals undertook interlocutory review of an order dealing with one claim of a multi-claim complaint. In that order, the district court had refused to dismiss a claim asserted under the authority of Bivens v. Six Unknown Fed. Narcotics Agents, 403 U.S. 388, 91 S.Ct. 1999, 29 L.Ed.2d 619 (1971). On appeal, the court of appeals not only affirmed the district court's conclusion as to the Bivens claim, but reached back in the record to order the district court to reinstate another claim for relief asserted under the Federal Tort Claims Act, 28 U.S.C. § 2671 et seq. In the present case, the disputed "prior order" had denied judgment of dismissal on the disgorgement claim. The order concededly before us denied judgment of dismissal on the same disgorgement claim. We see nothing in Stanley inconsistent with the later instruction in Yamaha recognizing our jurisdiction to "address any issue fairly included within the certified order." Yamaha, 516 U.S. at 205, 116 S.Ct. 619. We therefore proceed, obedient to our understanding of Yamaha, to review the order before us denying summary judgment.

We review an order denying summary judgment de novo. Cicippio-Puleo v. Islamic Republic of Iran, 353 F.3d 1024, 1031 (D.C. Cir. 2004). Obedient to Yamaha, we will review Order # 550 denying summary judgment applying anew the standards of Rule 56, and will not simply review that part of the District Court's thinking directed to the applicability of the Carson standard or the consistency of the Government's proffers with that standard. Therefore, we must address the issue, logically prior to the Carson question, of whether disgorgement is available at all. We hold that the language of § 1964(a) and the comprehensive remedial scheme of RICO preclude disgorgement as a possible remedy in this case.

B. The Availability of Disgorgement

The Government argues that § 1964 contains a grant of equitable jurisdiction that must be read broadly to permit disgorgement in light of Porter v. Warner Holding Co., 328 U.S. 395, 66 S.Ct. 1086, 90 L.Ed. 1332 (1946), and its progeny. The Porter Court considered reimbursement awards under the Emergency Price Control Act of 1942 ("EPCA") and concluded that where a statute grants general equitable jurisdiction to a court, "all the inherent equitable powers . . . are available for the proper and complete exercise of that jurisdiction." Porter, 328 U.S. at 398, 66 S.Ct. 1086. This grant is only to be limited when "a statute in so many words, or by a necessary and inescapable inference, restricts the court's jurisdiction." Id. In this case the text and structure of the statute provide just such a restriction.

As the Supreme Court has repeatedly observed: "Federal courts are courts of limited jurisdiction. They possess only that power authorized by Constitution and statute, which is not to be expanded by judicial decree." Kokkonen v. Guardian Life Ins. Co. of America, 511 U.S. 375, 377, 114 S.Ct. 1673, 128 L.Ed.2d 391 (1994) (citations omitted). Reading Porter in light of this limited jurisdiction we must not take it as a license to arrogate to ourselves unlimited equitable power. We will not expand upon our equitable jurisdiction if, as here, we are restricted by the statutory language, but may only assume broad equitable powers when the statutory or Constitutional grant of power is equally broad.

As our dissenting colleague correctly notes, the Court in Porter was considering whether a district court acting under the authority granted in the EPCA had the authority to order restitution for overcharges. The implication of broad equitable authority in Porter came from a statute which empowered the district court to grant "a permanent or temporary injunction, restraining order, or other order." EPCA § 205(a), 56 Stat. 23, 33 (1942). The action before the Court in Porter was brought under a section providing that "the Administrator" could bring action against persons engaged in overcharges for "an order enjoining such acts or practices, or for an order enforcing compliance with such provision, and upon a showing by the Administrator that such person has engaged or is about to engage in any such acts or practices a permanent or temporary injunction, restraining order, or other order shall be granted without bond." Id.

The Supreme Court did not have to make much of a stretch to determine that the phrase "enforcing compliance with such provision," and expressly referring to "a permanent or temporary injunction, restraining order, or other order," would include restitution for amounts collected exceeding the ceilings determined under the statute. The Government in the present case asks us to work a far greater expansion of the statutory grant enabling the District Court in a civil RICO action brought by the Government under § 1964(a). We further note that the Court in Porter was ordering restitution, under a statute designed to combat inflation. Restitution of overcharge works a direct remedy of past inflation, directly effecting the goal of the statute. The Court in Porter set forth two theories under which "[a]n order for the recovery and restitution of illegal rents may be considered a proper `other order'" under the applicable statute. 328 U.S. at 399, 66 S.Ct. 1086. First, the recovery of the illegal payment by the victim tenant "may be considered as an equitable adjunct to the injunction decree," as it effects "the recovery of that which has been illegally acquired and which has given rise to the necessity for injunctive relief." Id. (noting that "such a recovery could not be obtained through an independent suit in equity if an adequate legal remedy were available."). The equitable jurisdiction of the Court having been properly invoked, the Court then had the power "to decide all relevant matters in dispute and to award complete relief. . . ." Id. Also, and more to the point, the Court was authorized "in its discretion, to decree restitution of excessive charges in order to give effect of the policy of Congress." Id. at 400, 66 S.Ct. 1086. The policy of Congress under the EPCA was to prevent overcharges with inflationary effect. The goal of the RICO section under which the government seeks disgorgement here is to prevent or restrain future violations. We therefore must consider the forward-looking nature of the remedy in a way not applicable to a different remedy in Porter for the accomplishment of a different goal under a different statute.

Section 1964(a) provides jurisdiction to issue a variety of orders "to prevent and restrain" RICO violations. This language indicates that the jurisdiction is limited to forward-looking remedies that are aimed at future violations. The examples given in the text bear this out. Divestment, injunctions against persons' future involvement in the activities in which the RICO enterprise had been engaged, and dissolution of the enterprise are all aimed at separating the RICO criminal from the enterprise so that he cannot commit violations in the future. Disgorgement, on the other hand, is a quintessentially backward-looking remedy focused on remedying the effects of past conduct to restore the status quo. See, e.g., Tull v. United States, 481 U.S. 412, 424, 107 S.Ct. 1831, 95 L.Ed.2d 365 (1987). It is measured by the amount of prior unlawful gains and is awarded without respect to whether the defendant will act unlawfully in the future. Thus it is both aimed at and measured by past conduct.

The Government would have us interpret § 1964(a) instead to be a plenary grant of equitable jurisdiction, effectively ignoring the words "to prevent and restrain" altogether. This not only nullifies the plain meaning of the terms and violates our canon of statutory construction that we should strive to give meaning to every word, see, e.g., Murphy Explor. Production Co. v. United States Dept. of the Interior, 252 F.3d 473, 481 (D.C. Cir. 2001), but also neglects Supreme Court precedent. In Meghrig v. KFC Western, Inc., 516 U.S. 479, 488, 116 S.Ct. 1251, 134 L.Ed.2d 121 (1996), the Court held that compensation for past environmental cleanup was ruled out by the plain language of the Resource Conservation and Recovery Act which authorized actions "to restrain" persons who were improperly disposing of hazardous waste. If "restrain" is only aimed at future actions, "prevent" is even more so.

Mitchell v. DeMario Jewelry, 361 U.S. 288, 80 S.Ct. 332, 4 L.Ed.2d 323 (1960), relied on by the Government, is not to the contrary. The Mitchell case was brought under the Fair Labor Standards Act of 1938, 29 U.S.C. § 215, 52 Stat. 1060 (1938) ("FLSA"). In that action, the Government was invoking the court's jurisdiction to restrain violations of a section making it unlawful for a covered employer to discharge or discriminate against employees who had filed complaints or instituted actions under the FLSA. The Court reviewed the whole breadth of that broad Act to conclude that the available remedies included not only injunction against further discrimination and mandatory injunctions of reinstatement, but also a "make whole" reimbursement for lost wages because of the discriminatory discharge. As in Porter, the Court reiterated that in equitable jurisdiction "[u]nless otherwise provided by statute, all the inherent equitable powers of the District Court are available for the proper and complete exercise of that jurisdiction." Mitchell, 361 U.S. at 291, 80 S.Ct. 332 (quoting Porter, 328 U.S. at 398, 66 S.Ct. 1086). In the RICO Act, Congress provided a statute granting jurisdiction defined with the sort of limitations not present in the FLSA or the EPCA. The statute under which the Government sued Appellants, 18 U.S.C. § 1964(a), granted only the jurisdiction which we set forth above. The District Court, so far as is relevant to actions under that section, has jurisdiction only

to prevent and restrain violations of [RICO] by issuing appropriate orders, including, but not limited to: ordering any person to divest himself of any interest, direct or indirect, in any enterprise; imposing reasonable restrictions on the future activities or investments of any person, including but not limited to, prohibiting any person from engaging in the same type of endeavor as the enterprise engaged in, the activities of which affect interstate commerce; or ordering dissolution or reorganization of any enterprise. . . .

18 U.S.C. § 1964(a) (emphasis added). The order of disgorgement is not within the terms of that statutory grant, nor any necessary implication of the language of the statute.

In considering the broad language from Porter upon which our dissenting colleague relies for the proposition that we should find disgorgement available because Congress has not taken it away, we note that the Supreme Court considered a similar argument in Meghrig. The High Court nonetheless limited the available remedies under CERCLA to those provided in the statute, declaring that

where Congress has provided "elaborate enforcement provisions" for remedying the violation of a federal statute, as Congress has done with RCRA and CERCLA, "it cannot be assumed that Congress intended to authorize by implication additional judicial remedies. . . ."

516 U.S. at 487-88, 116 S.Ct. 1251 (quoting Middlesex County Sewerage Auth. v. Nat'l Sea Clammers Ass'n, 453 U.S. 1, 14, 101 S.Ct. 2615, 69 L.Ed.2d 435 (1981)).

In RICO, as in RCRA and in CERCLA, Congress has laid out elaborate enforcement proceedings. One of those proceedings is a government action brought under § 1964(a). That one does not provide for disgorgement. That one provides only for orders which "prevent or restrain" future violations. Disgorgement does not do that.

It is true, as the Government points out, that disgorgement may act to "prevent and restrain" future violations by general deterrence insofar as it makes RICO violations unprofitable. However, as the Second Circuit also observed, this argument goes too far. "If this were adequate justification, the phrase `prevent and restrain' would read `prevent, restrain, and discourage,' and would allow any remedy that inflicts pain." Carson, 52 F.3d at 1182.

The remedies available under § 1964(a) are also limited by those explicitly included in the statute. The words "including, but not limited to" introduce a non-exhaustive list that sets out specific examples of a general principle. See Dong v. Smithsonian Inst., 125 F.3d 877, 880 (D.C. Cir. 1997). Applying the canons of noscitur a sociis and ejusdem generis, we will expand on the remedies explicitly included in the statute only with remedies similar in nature to those enumerated. See Wash. State Dep't of Soc. Health Servs. v. Guardianship Estate of Keffeler, 537 U.S. 371, 384, 123 S.Ct. 1017, 154 L.Ed.2d 972 (2003). The remedies explicitly granted in § 1964(a) are all directed toward future conduct and separating the criminal from the RICO enterprise to prevent future violations. Disgorgement is a very different type of remedy aimed at separating the criminal from his prior ill-gotten gains and thus may not be properly inferred from § 1964(a).

The structure of RICO similarly limits courts' ability to fashion equitable remedies. Where a statute has a "comprehensive and reticulated" remedial scheme, we are reluctant to authorize additional remedies; Congress' care in formulating such a "carefully crafted and detailed enforcement scheme provides strong evidence that Congress did not intend to authorize other remedies that it simply forgot to incorporate expressly." Great-West Life Annuity Ins. Co. v. Knudson, 534 U.S. 204, 209, 122 S.Ct. 708, 151 L.Ed.2d 635 (2002) (quoting Mertens v. Hewitt Associates, 508 U.S. 248, 251, 254, 113 S.Ct. 2063, 124 L.Ed.2d 161 (1993)) (internal quotations omitted) (emphasis in original). RICO already provides for a comprehensive set of remedies. When Congress intended to award remedies that addressed past harms as well as those that offered prospective relief, it said as much. In a criminal RICO action the defendant must forfeit his interest in the RICO enterprise and unlawfully acquired proceeds, and may be punished with fines, imprisonment for up to twenty years, or both. 18 U.S.C. § 1963(a). In a civil case the Government may request limited equitable relief under § 1964(a). Individual plaintiffs are made whole and defendants punished through treble damages under 18 U.S.C. § 1964(c). This "comprehensive and reticulated" scheme, along with the plain meaning of the words themselves, serves to raise a "necessary and inescapable inference," sufficient under Porter, 328 U.S. at 398, 66 S.Ct. 1086, that Congress intended to limit relief under § 1964(a) to forward-looking orders, ruling out disgorgement.

Congress' intent when it drafted RICO's remedies would be circumvented by the Government's broad reading of its § 1964(a) remedies. The disgorgement requested here is similar in effect to the relief mandated under the criminal forfeiture provision, § 1963(a), without requiring the inconvenience of meeting the additional procedural safeguards that attend criminal charges, including a five-year statute of limitations, 18 U.S.C. § 3282, notice requirements, 18 U.S.C. § 1963( l), and general criminal procedural protections including proof beyond a reasonable doubt. Further, on the Government's view it can collect sums paralleling — perhaps exactly — the damages available to individual victims under § 1964(c). Not only would the resulting overlap allow the Government to escape a statute of limitations that would restrict private parties seeking essentially identical remedies, see Agency Holding Corp. v. Malley-Duff Assoc., Inc., 483 U.S. 143, 156, 107 S.Ct. 2759, 97 L.Ed.2d 121 (1987), but it raises issues of duplicative recovery of exactly the sort that the Supreme Court said in Holmes v. Securities Investor Protection Corp., 503 U.S. 258, 269, 112 S.Ct. 1311, 117 L.Ed.2d 532 (1992), constituted a basis for refusing to infer a cause of action not specified by the statute. Permitting disgorgement under § 1964(a) would therefore thwart Congress' intent in creating RICO's elaborate remedial scheme.

A note appended to the statute stating that RICO "shall be liberally construed to effectuate its remedial purposes" does not effect this structural inference. Organized Crime Control Act of 1970, Pub.L. No. 91-452, § 904(a), 84 Stat. 947 (codified in a note following 18 U.S.C. § 1961). This clause may warn us against taking an overly narrow view of the statute, but "it is not an invitation to apply RICO to new purposes that Congress never intended." Reves v. Ernst Young, 507 U.S. 170, 183, 113 S.Ct. 1163, 122 L.Ed.2d 525 (1993). The text and structure of RICO indicate that those remedial purposes do not extend to disgorgement in civil cases.

The Second Circuit in Carson has interpreted "prevent and restrain" not to eliminate the possibility of disgorgement altogether, but to limit it to cases where there is a finding "that the gains are being used to fund or promote the illegal conduct, or constitute capital available for that purpose." Carson, 52 F.3d at 1182. The Fifth Circuit adopted this interpretation in a case holding that disgorgement after the defendant had ceased production of an allegedly defective product would be inappropriately punitive rather than directed toward future violations. See Richard v. Hoechst Celanese Chemical Group, 355 F.3d 345, 355 (5th Cir. 2003). While we avoid creating circuit splits when possible, in this case we can find no justification for considering any order of disgorgement to be forward-looking as required by § 1964(a). The language of the statute explicitly provides three alternative ways to deprive RICO defendants of control over the enterprise and protect against future violations: divestment, injunction, and dissolution. We need not twist the language to create a new remedy not contemplated by the statute.

Our colleague reminds us that the Supreme Court has instructed "[i]f a precedent of this Court has direct application in a case, yet appears to rest on reasons rejected in some other line of decisions, the Court of Appeals should follow the case which directly controls, leaving to this Court the prerogative of overruling its own decisions." Dissent at 1220 (quoting Rodriguez de Quijas v. Shearson/American Express, Inc., 490 U.S. 477, 484, 109 S.Ct. 1917, 104 L.Ed.2d 526 (1989)). This would be most devastating to one side of the case or the other if we were in fact attempting to overrule a Supreme Court precedent. That is, if there were a Supreme Court case that had direct application to the facts before us, we would be required to follow it, and that would be the end of the matter. We would not need to consider any other line of cases. However, the Rodriguez de Quijas language is not particularly helpful when no precedent of the Supreme Court "has direct application," as in the present case. There is not a Supreme Court case dealing with the jurisdiction of a district court to order disgorgement under RICO § 1964(a). There is not a Supreme Court case discussing that question. There is, in short, no Supreme Court case having direct application. With no Supreme Court case having direct application, it is our duty to construe the statute. That is what we have done.

III. Conclusion

Because we hold that the District Court erred when it found that disgorgement was an available remedy under 18 U.S.C. § 1964(a), we reverse the District Court and grant summary judgment in favor of Appellants as to the Government's disgorgement claim.

WILLIAMS, Senior Circuit Judge, concurring.

I join the opinion for the court. I write separately to emphasize problems with the government's fallback interpretation of 18 U.S.C. § 1964(a), under which the government could obtain disgorgement for purposes of reducing the defendant's ability to commit future RICO violations, with the amount accordingly limited to assets "being used to fund or promote the illegal conduct, or [that] constitute capital available for that purpose." United States v. Carson, 52 F.3d 1173, 1182 (2d Cir. 1995). This superficially appealing interpretation in fact creates a kind of pushmi-pullyu, a beast that Congress is most unlikely to have ordained.

I.

The statute gives district courts "jurisdiction to prevent and restrain [RICO] violations." 18 U.S.C. § 1964(a). Reasoning that pure deterrence was an impermissible objective of orders under § 1964(a), the Second Circuit went on to find that disgorgement could "prevent and restrain" if limited to the amount of ill-gotten gains that were "being used to fund or promote the illegal conduct, or constitute capital available for that purpose." Id. at 1182. Because money is fungible, as indeed are virtually all resources when viewed as enablers of future criminal conduct, the government here refines its Carson-derived fallback position, quite sensibly rejecting any limitation to "ill-gotten gains" in the form of specific money or resources so gained. Such a limit, we have said (applying a different statute), would lead to absurd results. SEC v. Banner Fund International, 211 F.3d 602, 617 (D.C. Cir. 2000). There the defendant proposed to confine disgorgement to the "actual assets" unjustly received. We said that what mattered was not the specific assets but the amount by which the defendant was unjustly enriched; the alternative would allow a defendant to escape liability by spending ill-gotten gains while husbanding other assets. Id. at 617. Thus the government's proposal is that the amount of the ill-gotten gains should set a ceiling on the disgorgement recovery, subject to the further limit mentioned above — essentially purporting to limit the disgorgement to crime-enabling resources, broadly construed.

In Carson itself the court ruled that this prevented the government from forcing disgorgement of funds, ill-gotten in the distant past, from a RICO defendant by then retired from the RICO enterprise itself (a union). In the context of corporate defendants such as those before us, a possible limit would be the entire net worth of the companies (a good deal less than the $280 billion that the government claims to have been ill-gotten gains). But perhaps not. Even that limit is arbitrary, as resources can be used for criminal purposes even if offset by company debt. Subject to the bankruptcy laws, nothing in the logic of the crime-enablement theory clearly calls for stopping at confiscation of the shareholders' interests; why not the bondholders' as well?

On the other side, it might be plausible under the Carson theory to exempt firm resources now devoted to non-tobacco enterprises. It is probably about as difficult for these defendants to re-allocate resources from the businesses of cheese and crackers, for example, to criminality in the sale of cigarettes, as for the union in Carson to lure Carson and his funds back from retirement to union criminality.

In short, Carson and the government's fallback position send the court off on a virtually metaphysical quest to draw lines based on the likelihood that particular resources will be devoted to crime.

II.

It is hardly surprising that there are only gossamer lines between drastic disgorgement (destruction of bondholder as well as shareholder wealth) and relatively mild disgorgement (cordoning off resources in non-tobacco subsidiaries). The plain fact is that wealth deprivation is an extremely crude device for "prevent[ing]" criminal behavior. Granted, a criminal miscreant with a billion dollars is potentially more dangerous than an impoverished criminal miscreant. But ordinarily the forces most affecting the likelihood of criminal action are, besides the actors' ethical standards and sense of shame, truly forward-looking conditions: the returns to crime versus the possible costs, all adjusted for risk (such as the risk of getting caught).

Confusion arises from an ambiguity in our understanding that, in the civil context, such remedies as damage awards and restitution "deter," and thus in a sense "prevent" commission of torts, breaches of contract, and other civil wrongs. It is quite true that a rule or practice of awarding such remedies deters, and thus prevents, such wrongs. Indeed, under one viewpoint that is the primary or even sole purpose of awarding such remedies. See William M. Landes Richard A. Posner, The Economic Structure of Tort Law (1987). But it is the rule or practice that creates the incentive. To make the rule credible, of course, the awards must be made; but no individual award has a material deterrent effect.

To evaluate that last statement consider a society that empowered some deus ex machina to randomly excuse one damage judgment in a million. Such an exception to the rules would have no detectable effect on the commission of torts or breaching of contracts. Even the lucky defendant who enjoyed the benefit of the pardon wouldn't — unless a complete fool — materially alter his future conduct because of that manna from heaven.

The equity court, empowered under § 1964(a) to "prevent and restrain" future violations, has before it the history of the defendant, including his past wrongs. It can decree relief targeted to his plausible future behavior. It can define the conditions bearing directly on that behavior. It can, for example, establish schedules of draconian contempt penalties for future violations, and impose transparency requirements so that future violations will be quickly and easily identified.

In assessing the likelihood that Congress intended an additional disgorgement remedy, it makes sense to inquire into the tendency of such an implied remedy to "prevent and restrain" future violations by the defendant. Of course the rule the government seeks here would be a rule, not merely a random extra penalty. But the question would be its incremental effect, on top of (1) RICO's explicit provisions for criminal penalties (including disgorgement and imprisonment under § 1963(a)) and for victim recoveries (trebled) under § 1964(c), and (2) the whole available panoply of genuinely forward-looking remedies — express controls over substantive conduct, transparency-enhancing orders, and contempt penalties for violations. It seems almost inconceivable that many aspiring criminals would find the incremental risk decisive. I find it hard to imagine a waffling villain — already in court for RICO violations — saying to himself: "Well, my chances of escaping § 1963(a) forfeiture and imprisonment because of the statute of limitations and the burden of proof, and of escaping treble damages under § 1964(c), and contempt penalties for violating the court's orders, still leave RICO violations attractive on a net basis; but that implied disgorgement under § 1964(a) — wow! Too much. It tilts me over the line."

The weakness of that scenario supports the inference that for the defendant who winds up before the equity court, Congress intended the words "prevent and restrain" to authorize only a tailored, forward-looking remedy. Penalties for violations of the court's decree, and transparency-enhancing measures meet that standard. A purported § 1964(a) disgorgement remedy, on top of those explicitly authorized, would provide only a trivial incremental effect (the reverse of the pardon granted once in a million), and would not qualify. Nor would disgorgement aimed at reducing the defendant's crime-enabling resources, a factor linked only crudely to his future tendency toward criminality.

Once we (1) accept the proposition that § 1964(a) limits the equity court to forward-looking remedies, as even the dissent appears to do with respect to the government's narrower argument, see Dissent at 1224-25 ("I also share the Second Circuit's apparent conclusion . . . that disgorgement may be ordered only to prevent and restrain a defendant from future RICO violations."), and (2) reject the supposition that "whatever hurts a civil RICO violator necessarily serves to `prevent and restrain' future violations," Carson, 52 F.3d at 1182, the court must try to draw lines between equitable remedies that merely "hurt" the defendant and ones that have a genuine tendency to "prevent and restrain" his future violations.

Because disgorgement under § 1964(a) so evidently lacks that tendency, the dissent relies on Porter and on the government's experts. Porter indeed includes the twice cited phrase suggesting that "[f]uture compliance may be more definitely assured if one is compelled to restore one's ill-gotten gains." Dissent at 1223, 1224. But the statute at issue in Porter gave district courts power to issue orders "enforcing compliance" and thus didn't seem to narrow the grant to forward-looking remedies. Indeed the Porter dissent never suggests such a limit; nor, so far as appears, did the defendant firm. For construing § 1964(a), Porter is of remarkably little help.

The expert testimony offered by the government for the proposition that backward-looking disgorgement will "`prevent and restrain' defendants from committing future RICO violations," see Dissent at 1226, serves no better. Obviously such testimony cannot alone resolve the issue, turning legal analysis of the statute into a fact battle among experts. Thus the experts' testimony is valuable for its analytic quality, not its utterance by a PhD.

The dissent's genuflection before the experts leaves the reader to imagine some supporting analysis. Lest the imagination run riot, I attach an appendix containing all of the expert testimony that the government saw fit to offer on the point in the summary judgment motion. The crux is Dr. Franklin Fisher's statement:

[Defendants' experts] have also suggested that enjoining Defendants from future illegal behavior and threatening them with the possibility of financial penalties would be more effective as future deterrents than would be disgorgement. Professor Weil, for example, suggests that `the Court could establish now a schedule of fines or punishments that it would levy should the Defendants engage in prohibited behavior.' These experts forget that laws prohibiting this behavior already exist and that, despite these laws and their associated remedies, the Defendants allegedly chose to engage in the illegal behavior. In this context, it is important to note that requiring Defendants to pay proceeds would strengthen the credibility of existing laws and thus provide additional economic incentives to deter future misconduct.

While it is a nice rhetorical move to point out that the defendants violated RICO (as we must assume) despite existing sanctions, Fisher offers no analysis as to why the presence of a civil disgorgement remedy in favor of the government would have reduced the likelihood of violations. (Indeed, on the government's theory — that the statute actually creates such a remedy — the defendants would have taken that into account in deciding to proceed with violations.) More important, Fisher looks at the wrong setting. Before this (or any) RICO litigation against a particular defendant, that defendant would have operated without the spotlight of the lawsuit itself. (That may explain why the government let the statute of limitations run for decades, and why the victims failed to seek treble damages.) Now the spotlight is on, and the plausible explanations for non-application of the explicit remedies (other than § 1964(a) equitable relief) have disappeared. And the district court can amplify the spotlight with transparency-enhancing and prior-approval measures. The real question is whether the imposition of this extra remedy on the defendants before the court — backward-looking civil disgorgement in favor of the government — would materially alter their readiness to persist in violations, in the face of all RICO's explicit remedies, and a forward-looking schedule of penalties for even minute infractions, made doubly effective by compulsory disclosure and approval measures. The government's experts simply did not address that question. This court's own analysis provides a clear answer that the extra "remedy" would not do so.

The dissent's use of the government's experts is part of its effort (in its qualified endorsement of the government's fallback position) to transform an issue of statutory interpretation into one of fact. See Dissent at 1222-23, 1227-28; see also id. at 1223 (noting that in Meghrig v. KFC Western, Inc., 516 U.S. 479, 116 S.Ct. 1251, 134 L.Ed.2d 121 (1996), there was no affirmative evidence that the defendants were likely to commit future RCRA violations, and thus suggesting that the case was something other than pure statutory interpretation). But the "facts" hypothesized by the dissent are unrelated to the real world faced by RICO defendants — already arraigned for their past offenses and subject to a battery of new disincentives on top of all RICO's conventional explicit remedies. Statutory interpretation shouldn't turn on factual hypotheticals such as, "What if pigs had wings."

III.

The above analysis seems to me to confirm what intuition suggests about the jurisdictional issue in this case. Even the most narrowly formulated question about the validity of the district court's order — the choice between the government's primary position (that § 1964(a) creates unlimited discretion to order disgorgement) and its fallback position (that it provides authority to award crime-enabling disgorgement) — requires the court to plumb the meaning of § 1964(a). The issues in this case, all turning on the interpretation of § 1964(a)'s lone sentence, are so thoroughly enmeshed that we needn't explore the court's language limiting § 1292(b) jurisdiction to issues "logically interwoven" with the explicitly identified issue. Maj. Op. at 1196. The dissent's hypotheticals as to what might be covered, see Dissent at 1212, plainly depend on an astonishingly broad notion of either logic or weaving. Having analyzed § 1964(a) and having found the order in conflict with its terms, the court must reverse.

One final note. The dissent chides the court for creating a circuit split. See Dissent at 1208. But if we confined ourselves to what the dissent acknowledges to be properly before us, and adopted the dissent's preferred position (that disgorgement is available like any other equitable remedy, regardless of its likely effects on a defendant's future behavior, simply because RICO doesn't explicitly preclude it), we would create no less of a split between this circuit and the Second.

Appendix

Excerpt from United States Memorandum in Opposition to Defendants' Motion for Partial Summary Judgment Dismissing the Government's Disgorgement Claim, Appellee's Appendix at 812-14.

B. Disgorgement Provides Economic Incentives That Will Prevent Further RICO Violations

172. Despite the fact that it is not necessary for the United States to prove this, disgorgement will prevent and restrain further bad acts.

173. Drs. Fisher and Kothari have both stated in their expert reports and deposition testimony, that disgorgement of the proceeds calculated by Dr. Fisher would in fact act to prevent and restrain future RICO violations. Dr. Fisher directly addressed this point in his rebuttal report in which he states:

Defendants' experts have suggested that disgorgement of ill-gotten gains such as the proceeds sought in this matter will not serve the goal of preventing or restraining the defendants from engaging in similar bad acts in the future. For example, Professor Carlton argues, "Having to disgorge past proceeds, by itself, would not affect a defendant's incentives to engage in misconduct in the future because it would not affect the returns (if any) from future misconduct." I address these criticisms with well-known economic principles. What Professor Carlton and the other defendants' experts who espouse this view fail to recognize is that requiring defendants to pay proceeds will affect their expectations (and those of others contemplating malfeasance) about the returns from future misconduct. As a matter of economic principle, the higher the proceeds amount, the lower the expected returns from future misconduct and the greater the desired effect of deterrence.

Expert Rebuttal Report of Franklin Fisher, United States v. Philip Morris, (R. 1450; filed July 24, 2002) at 4-5 ¶ 12.

174. Dr. Kothari's expert report confirms Dr. Fisher's conclusion:

Requiring the defendants to pay ill-gotten proceeds is relevant. The economic incentive for illegal behavior is higher (for defendants and onlookers) if defendants are not required to pay the proceeds. While payment of proceeds has some of the features of sunk cost, it is not identical to a sunk cost because it will affect future decisions or behavior. The higher the proceeds paid the greater the economic incentive to avoid illegal behavior in the future.

Expert Report of S.P. Kothari, United States v. Philip Morris, (R. 1451; filed July 24, 2002) at 3-4, ¶ 8.

175. Dr. Fisher expressly states in his expert report:

[Defendants' experts] have also suggested that enjoining Defendants from future illegal behavior and threatening them with the possibility of financial penalties would be more effective as future deterrents than would be disgorgement. Professor Weil, for example, suggests that `the Court could establish now a schedule of fines or punishments that it would levy should the Defendants engage in prohibited behavior.' These experts forget that laws prohibiting this behavior already exist and that, despite these laws and their associated remedies, the Defendants allegedly chose to engage in the illegal behavior. In this context, it is important to note that requiring Defendants to pay proceeds would strengthen the credibility of existing laws and thus provide additional economic incentives to deter future misconduct.

Expert Rebuttal Report of Franklin Fisher, United States v. Philip Morris, (R. 1450; filed July 24, 2002) at 5-6, ¶ 14.

176. Dr. Fisher has repeatedly confirmed the preventative benefit of disgorgement. At his deposition he stated:

Q. . . . the idea is that disgorgement prevents and restrains future violations by altering the defendants' expectations about the returns they might receive from future misconduct. Is that right?

A. . . .I believe that to be correct.

Q. Does disgorgement prevent and restrain future RICO violations in any other way?

A. Well, it removes at least some, and possibly all, of the assets with which to engage in future illegal activities.

Deposition of Franklin Fisher, United States v. Philip Morris, September 12, 2002, 828:4-19 (Exhibit 77).

177. "[A]s I have repeatedly and clearly stated in my report and deposition testimony, disgorgement of Defendants' proceeds, as I have calculated them, would in fact act to prevent and restrain future RICO violations." Declaration of Franklin Fisher, United States v. Philip Morris, at 7, ¶ 16 (Master Rule 7.1/56.1 St. Exhibit 5)

TATEL, Circuit Judge, dissenting.

Congress passed the Organized Crime Control Act of 1970, which included RICO, "to seek the eradication of organized crime in the United States . . . by providing enhanced sanctions and new remedies to deal with the unlawful activities of those engaged in organized crime." United States v. Turkette, 452 U.S. 576, 589, 101 S.Ct. 2524, 69 L.Ed.2d 246 (1981) (quoting Pub.L. No. 91-452, 84 Stat. 922, 923 (1970)). Through this lawsuit, the United States seeks to end what it perceives as rampant racketeering violations within the tobacco industry. Specifically, the government offers voluminous evidence, which we must view in the light most favorable to it, see Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 255, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986) (stating that at summary judgment the "evidence of the non-movant is to be believed, and all justifiable inferences are to be drawn in [its] favor"), that Philip Morris, Altria Group, R.J. Reynolds, Brown Williamson, Lorillard, BATCo, and Liggett have engaged in a half century of deceptive practices to the detriment of the health — and lives — of their customers. Acting both individually and in concert through collective agreements and jointly funded organizations like the Council for Tobacco Research and the Tobacco Institute (also defendants), these companies publicly defended smoking as both harmless and nonaddictive despite knowing from internal research that it was neither. In their advertising campaigns the companies targeted young people, who "often lack the experience, perspective, and judgment to recognize and avoid choices that could be detrimental to them," Bellotti v. Baird, 443 U.S. 622, 635, 99 S.Ct. 3035, 61 L.Ed.2d 797 (1979), despite publicly claiming otherwise.

The government alleges that during the course of this behavior, the defendants committed over ninety racketeering violations between RICO's 1970 effective date and the government's 1999 complaint. Significantly for this appeal, the government further claims that absent court intervention and despite the master settlement agreement between the tobacco companies and the states, the companies are likely to continue their deceptive practices and commit further racketeering violations in the future. The government's claim regarding likely future conduct rests not only on the companies' alleged history of deceptive activities, but also on record evidence that the companies continue making their misleading statements about both the health consequences of smoking and the addictive nature of nicotine, as well as persisting in their marketing efforts aimed at young people. The government asks the district court to enjoin the tobacco companies from future unlawful conduct and to order them to disgorge the profits they have earned due to their racketeering violations since RICO's effective date — profits the government estimates amount to $280 billion.

In now holding that district courts may never order disgorgement as a remedy for RICO violations, this court ignores controlling Supreme Court precedent, disregards Congress's plain language, and creates a circuit split — all in deciding an issue not properly before us. Because the tobacco companies ask us to address an issue not fairly included in the certified order and not presented at that time to the district court, I would dismiss this interlocutory appeal. Were it appropriate to reach the merits, I would uphold the district court's denial of summary judgment on either of two grounds. First, unless "a statute in so many words, or by a necessary and inescapable inference, restricts the court's jurisdiction in equity," district courts may grant any equitable relief. Porter v. Warner Holding Co., 328 U.S. 395, 398, 66 S.Ct. 1086, 90 L.Ed. 1332 (1946). Because under a fair application of Supreme Court precedent, see id. at 398-403, 66 S.Ct. 1086, no such inference can be drawn about RICO, I would conclude that the district court has authority to order disgorgement. Alternatively, even if RICO's phrase "prevent and restrain violations," 18 U.S.C. § 1964(a), limits the district court's equitable jurisdiction, I would still uphold the denial of summary judgment because the government has presented evidence that disgorgement will accomplish just that purpose in this case.

I.

Under 28 U.S.C. § 1292(b), if a district court "shall be of the opinion that [an] order involves a controlling question of law as to which there is substantial ground for difference of opinion and that an immediate appeal from the order may materially advance the ultimate termination of the litigation," it may certify the order for interlocutory review, and the court of appeals "may thereupon, in its discretion, permit an appeal to be taken from such order." Section 1292(b) establishes a "two-tiered arrangement." Swint v. Chambers County Comm'n, 514 U.S. 35, 47, 115 S.Ct. 1203, 131 L.Ed.2d 60 (1995). Congress "chose to confer on district courts first line discretion to allow interlocutory appeals," id., and "even if the district judge certifies the order under § 1292(b), the appellant still has the burden of persuading the court of appeals that exceptional circumstances justify a departure from the basic policy of postponing appellate review until after the entry of a final judgment," Coopers Lybrand v. Livesay, 437 U.S. 463, 475, 98 S.Ct. 2454, 57 L.Ed.2d 351 (1978) (internal quotation marks and citation omitted). In accepting this interlocutory appeal, this court not only (at the least) pushes the bounds of its jurisdiction, but also exercises its discretion on behalf of defendants whose litigating tactics leave much to be desired.

A.

In 2000, the tobacco companies — usually referred to in this opinion as "Philip Morris" — filed a motion to dismiss, arguing (among other things) that "disgorgement . . . is never available under a civil RICO count." See United States v. Philip Morris Inc., 116 F.Supp.2d 131, 150 (D.D.C. 2000). Denying that motion, the district court held that disgorgement could be available under 18 U.S.C. § 1964(a), but did not address whether disgorgement would be available in this particular case. See id. at 150-52. Philip Morris never sought certification of that order, though it could have done so at any time after the order's issuance. See Fed.R.App.P. 5(a)(3) (providing that the time for filing an appeal runs from when the district court amends the order to include certification, not from the issuance of the actual order); 16 Wright, Miller Cooper, Federal Practice and Procedure § 3929 (2d. ed. 1996) ("This latitude [in Rule 5(a)] makes it possible to employ § 1292(b) with some precision, deferring the question of appeal until it is clear that prompt appeal is apt to be useful.").

In 2004, Philip Morris sought summary judgment regarding the government's request for disgorgement in this case. Contrary to the court's statement, see majority op. at 1193, Philip Morris neither reargued the position it took in 2000 nor asked the district court to revisit its 2000 decision. Philip Morris's only reference to its prior position came in a one-sentence footnote: "As noted previously, Defendants respectfully disagree with the Court and maintain that disgorgement in any fashion is unavailable to the Government in a civil RICO action." Defs.' Br. in Supp. Mot. Partial Summ. J. at 6 n. 4. Instead, Philip Morris urged the court to grant its motion for summary judgment for two primary reasons. First, relying on United States v. Carson, where the Second Circuit held that district courts may order disgorgement as a RICO remedy only where the gains "are being used to fund or promote the illegal conduct, or constitute capital available for that purpose," id. at 20 (quoting United States v. Carson, 52 F.3d 1173, 1182 (2d Cir. 1995)), Philip Morris claimed that 18 U.S.C. § 1964(a) "limits disgorgement to the amount of ill-gotten gains that remain available to defendants to fund future RICO violations," id. Philip Morris further argued that "the Government deliberately has refused to develop the proof properly required under Carson" and this in turn "requires dismissal of the Government's disgorgement claim." Id. at 25. Second, Philip Morris asserted that the government's disgorgement model fails as a matter of law to reasonably approximate the defendants' ill-gotten gains.

The district court rejected both arguments and denied summary judgment to Philip Morris. United States v. Philip Morris USA, Inc., 321 F.Supp.2d 72 (D.D.C. 2004). Interpreting section 1964(a) more broadly than had the Second Circuit, the court concluded that it could order disgorgement in situations besides those identified in Carson. Id. at 77-79. Unsurprisingly, the district court did not revisit its 2000 decision, observing only (in a footnote) that this decision had held "that disgorgement is a permissible remedy under Section 1964(a)." Id. at 76 n. 7. The district court also rejected Philip Morris's contention regarding the government's disgorgement model. Id. at 81-82.

Philip Morris then asked the district court to certify its 2004 order under section 1292(b). In its certification request, Philip Morris did not reassert its legal argument from 2000. Instead, it stated that "[w]hether the Carson standard applies to the Government's disgorgement claim is clearly a controlling question of law. . . . If the Government is wrong, and Carson applies, nothing is left of its claim in this case." Def's Br. Supp. Mot. Certify Order # 550 for Interloc. App. at 4.

The district court agreed that a controlling question of law existed as to whether "the disgorgement allowed under 18 U.S.C. § 1964(a) is limited to those ill-gotten gains which are `being used to fund or promote the illegal conduct or constitute capital available for that purpose.'" United States v. Philip Morris USA, Inc., No. 99-2496, slip op. at 2-4, 2004 WL 1514215 (D.D.C. June 25, 2004) (quoting Carson, 52 F.3d at 1182). Although in its 2004 order the district court had rejected Carson's interpretation of section 1964(a), it found substantial ground for difference of opinion on this issue, explaining that "it is obvious that the arguments to the contrary in Carson are neither insubstantial nor frivolous," and certified the 2004 order. Id. at 4, 7.

In its initial petition urging this court to accept the interlocutory appeal, Philip Morris never raised the broader question the district court had addressed in 2000, i.e., whether disgorgement is ever available under section 1964(a). Instead, Philip Morris focused on the narrower issue actually raised in its 2004 motion for summary judgment, arguing that the district court had erred in rejecting Carson's interpretation of section 1964(a) and claiming that "[i]f this Court agrees with the Second Circuit in Carson, its decision on appeal would dispose of the Government's disgorgement claim." Emergency Pet. for Permission to Appeal an Order at 9. The government opposed Philip Morris's section 1292(b) petition, arguing that a host of factual issues would require resolution regardless of whether this court adopted Carson's or the district court's interpretation of section 1964(a) and thus that "interlocutory appeal would not materially advance the termination of this litigation." Resp. in Opp'n to Emergency Pet. at 15.

Responding to the government's opposition, Philip Morris suddenly changed tack and brought in play the issue decided in 2000. Philip Morris wrote:

The district court rejected [the government's] argument [that an interlocutory appeal would not materially advance the litigation's termination] as a reason not to permit an appeal, and this Court should as well.

First, and most obviously, if this Court reverses the district court's ruling that `disgorgement is a permissible remedy under section 1964(a),' (Summary Judgment Order at 8 n. 7), then the Government's $280 billion claim is precluded as a matter of law.

Reply to Emergency Pet. for Permission to Appeal an Order at 5. This entirely disingenuous statement conveyed the impression that the district court had ruled on this broader issue in the certified 2004 order rather than simply mentioning its 2000 decision. Moreover, by placing this statement under the heading "The District Court Properly Determined That an Appeal From Its Order Would Materially Advance This Litigation," id., Philip Morris insinuated that the district court had certified this issue to this court as opposed to the narrower question actually resolved in the 2004 order. The government, of course, had no opportunity to correct these misrepresentations, and a motions panel accepted Philip Morris's appeal, expressly leaving the merits panel free to reconsider and dismiss the appeal. In re Philip Morris USA, Inc., No. 04-8005 (D.C. Cir. July 15, 2004).

Philip Morris's opening brief on the merits reveals the scope of its bait and switch. The brief devotes forty pages to the issue decided in the 2000 order and only seven to the issues decided in the certified 2004 order. In response, the government urges us to dismiss the appeal entirely, suggesting that we lack jurisdiction over the issue decided in the 2000 order and observing that "Defendants' tactics subvert the mechanism for appeal established by section 1292(b)." Appellee's Br. at 45-46.

B.

As the foregoing discussion indicates, Philip Morris asks us — and the court now agrees — to decide an issue (1) not briefed in the motion leading up to the certified order, (2) not decided in the district court's opinion accompanying the certified order, (3) not raised by Philip Morris in its request for certification, (4) not discussed in the order granting certification, (5) not raised by Philip Morris in its section 1292(b) petition before this court, and (6) decided in an entirely different order which Philip Morris could at any time have asked the district court to certify. This presents serious questions on two separate fronts: our jurisdiction over this appeal under section 1292(b), and our general policy of declining to consider arguments not made to the district court in the motion leading to the order under appeal. Unlike the court, I cannot brush these concerns aside.

Regarding our jurisdiction under section 1292(b), the Supreme Court has made clear that an appellate court can review "any issue fairly included within the certified order" because "[a]s the text of § 1292(b) indicates, appellate jurisdiction applies to the order certified to the court of appeals, and is not tied to the particular question formulated by the district court." Yamaha Motor Corp., USA v. Calhoun, 516 U.S. 199, 205, 116 S.Ct. 619, 133 L.Ed.2d 578 (1996) (holding that where the district court decided two issues in the certified order but identified only the damages issue as the controlling question of law, the court of appeals could nonetheless address the other issue). But the "court of appeals may not reach beyond the certified order to address other orders made in the case." Id.; see also United States v. Stanley, 483 U.S. 669, 677, 107 S.Ct. 3054, 97 L.Ed.2d 550 (1987) (holding that the court of appeals erred in addressing a claim not raised in the certified order though closely related to it). Both "[c]ommentators and courts have consistently observed that `the scope of the issues open to the court of appeals is closely limited to the order appealed from [and][t]he court of appeals will not consider matters that were ruled upon in other orders.'" Stanley, 483 U.S. at 677, 107 S.Ct. 3054 (quoting 16 Wright, Miller, Cooper Gressman, Federal Practice and Procedure § 3929 (1977)) (second and third alterations in original).

This case falls near the intersection of these commands. For all intents and purposes, Philip Morris asks us to address the 2000 order. Today's decision overturns that order. This court has jurisdiction to do this under Yamaha only if the issue addressed in the 2000 order is "fairly included within the certified order." Taking a broad view of "fairly included," the court concludes that because the 2004 order denies dismissal of the government's disgorgement claim, we may review (at a minimum) any basis for summary judgment that is "logically interwoven with the explicitly identified issue." See majority op. at 1196. This approach not only gives us jurisdiction over the issue decided by the district court in the 2000 order, but also over the district court's 2002 determination, made in denying Philip Morris's motion for a jury trial, that disgorgement is an equitable remedy rather than a legal one, United States v. Philip Morris, Inc., 273 F.Supp.2d 3, 8-11 (D.D.C. 2002). Indeed, although the concurrence apparently does not share this approach, see sep. op. at 1206 (Williams, J., concurring), the majority opinion suggests that any issue which would result in " complete dismissal of the Government's claim for disgorgement with prejudice" lies within our jurisdiction "regardless of the grounds the District Court gave for its decision," see majority op. at 1194. By this logic, we may also have interlocutory jurisdiction to review the district court's denial of the tobacco companies' 2000 motion to dismiss, where they claimed that the government has not "adequately alleged that Defendants' racketeering activity will continue into the future," 116 F.Supp.2d at 147-50, and even the district court's denial of Liggett's 2000 motion to dismiss, where the company argued that (as to it) the government could not show two elements required for a RICO claim, id. at 152-53. Because victory for the tobacco companies on the first issue (and, for Liggett, victory on the second) could also trigger dismissal of the government's disgorgement claims, under the court's theory our interlocutory jurisdiction may extend to these issues as well.

The court's approach is problematic in several respects. Most significantly, it curtails the district court's section 1292(b) certification role. In this case, the district court had neither an opportunity to exercise "first line discretion to allow interlocutory appeal ," Swint, 514 U.S. at 47, 115 S.Ct. 1203, on the broader issue resolved in its 2000 order nor notice that Philip Morris would raise this issue with us. In future cases, district courts will lose their flexibility to certify discrete issues for review, since the certification of one order may give this court jurisdiction over all sorts of prior orders. Today's situation illustrates this: under the court's theory, we have jurisdiction in this interlocutory appeal to review at a minimum two prior orders, neither of which Philip Morris sought to certify. Moreover, by reducing the opportunity for tailored review, the court's jurisdictional theory threatens this circuit with interlocutory overload. Parties who persuade us to accept an interlocutory appeal may feel encouraged to raise any or even all issues decided in prior orders that fall within our newfound jurisdiction especially since, according to the court, issues raised in prior orders are "preserved" for section 1292(b) purposes, see majority op. at 1196, and not simply for the purpose of appeal after final judgment.

By contrast, no harm of consequence would result from holding, as I would, that the only issues "fairly included" within a certified order are those decided in the district court's accompanying memorandum — exactly the situation with the issue reached by the Supreme Court in Yamaha, 516 U.S. at 203-05, 116 S.Ct. 619. There, the Court found "fairly included" an issue that the district court had resolved in the same opinion in which it decided the issue identified as the controlling question of law, see Calhoun v. Yamaha Motor Corp., USA, No. 90-4295, 1993 WL 216238 (E.D.Pa. June 22, 1993). While the Court did not explicitly rely on this point, it is relevant to determining whether Yamaha's "fairly included" language stands for the proposition that appellate courts have interlocutory jurisdiction over all possible bases for reversing a summary judgment denial (as my colleagues read it) or only over bases which the district court considered and resolved in this denial (as I read it).

My approach, moreover, respects the Court's instruction in Stanley that we should "not consider matters that were ruled upon in other orders." 483 U.S. at 677, 107 S.Ct. 3054 (citation omitted); cf. Briggs v. Goodwin, 569 F.2d 10, 25 (D.C. Cir. 1977) (noting that any possible justification for addressing "all other issues relevant to the result reached by [a certified] order" would "be substantially diminished . . . where the order certified for appeal is a separate order from the one [containing the other issues]"); Dinsmore v. Squadron, Ellenoff, Plesent, Sheinfeld Sorkin, 135 F.3d 837, 840 (2d. Cir. 1998) (finding that the certified order referred to rather than incorporated a prior order and concluding that no interlocutory jurisdiction existed over the issue decided in the prior order). It is thus hardly surprising that the court today points to no case in which an appellate court has exercised interlocutory jurisdiction over an issue decided in a different order from the one under certification. True, under my approach a party seeking an interlocutory appeal on a matter split across two orders would need to seek certification of both orders to bring the matter fully to this court. But that seems a small burden. If the party fails to make this effort (as in this case) and we conclude that it would be inappropriate to address only the issues raised in the certified order (as I would here), then we have discretion under section 1292(b) to refuse to permit the interlocutory appeal altogether — a point this court overlooks.

In addition to resting on a dubious interpretation of section 1292(b), the court's decision to review the broader issue runs counter to this circuit's general rules regarding waiver. Parties may raise here only those arguments they presented to the district court in their papers seeking (and opposing) the order under review, since only in exceptional circumstances will we consider an argument not made to the district court. See United States v. British Am. Tobacco (Invs.) Ltd., 387 F.3d 884, 887-88 (D.C. Cir. 2004) (finding waiver based on a party's failure to appear and defend a privilege claim in the proceedings resulting in the interlocutory appeal, even though the party had asserted the privilege in a related proceeding in the same case); see also id. at 892 (refusing to consider argument not raised below) (citing United States v. Hylton, 294 F.3d 130, 135-36 (D.C. Cir. 2002)). Here, as discussed earlier, Philip Morris never argued the broader issue in the relevant pleadings; a sentence-long footnote stating "respectful disagreement" is not an argument, particularly when offered in such a cursory fashion. Cf., e.g., Cement Kiln Recycling Coalition v. EPA, 255 F.3d 855, 869 (D.C. Cir. 2001) (per curiam) (observing that a "litigant does not properly raise an issue by addressing it in a `cursory fashion' with only `bare-bones arguments'"); Wash. Legal Clinic for the Homeless v. Barry, 107 F.3d 32, 39 (D.C. Cir. 1997) (declining to address argument made in a footnote). Although it is true, as the court points out, that in the two just-cited cases the issues were apparently never raised at an earlier stage, here we are reviewing not the entire case but only the certified 2004 order, which sets the bounds of both our jurisdiction and waiver doctrine. Moreover, while we sometimes make exceptions to our waiver rules, I would not do so here given Philip Morris's questionable tactics. Even under my colleagues' jurisdictional theory, only by exercising our discretion to accept an argument not raised in the district court — and further exercising our discretion to accept the interlocutory appeal — does the broader issue stand before us.

In sum, whether viewed in terms of jurisdiction or waiver, only Philip Morris's narrower challenge is properly before us. True, this means we should dismiss the appeal altogether, as it makes little sense to decide the narrower question at this time when the broader question might be appealed later. But Philip Morris itself created this problem. It had several ways it could properly have brought the broader issue to our attention. In its 2004 motion for summary judgment, it could have reargued the broader question and asked the district court to reconsider its decision; the district court's denial of reconsideration would have brought the issue fairly into the challenged order. Even more appropriately, Philip Morris could have asked the district court to certify both the 2000 and 2004 orders and candidly explained that it wished this court to review the earlier order as well. Either way, the district court, having fair notice that Philip Morris wanted to raise both issues with us, could have performed its section 1292(b) gatekeeping function. Taking neither approach, Philip Morris instead not only jumped the fence at the district court level, but also circumvented our own screening process by waiting until after the government's opposition to raise the broader issue with the motions panel. This court should not be rewarding such tactics by exercising its discretion to hear this appeal.

I would therefore dismiss the interlocutory appeal. I reach this conclusion reluctantly because I certainly understand how hearing this interlocutory appeal could be helpful to Judge Kessler, who is presiding over a long and difficult trial. In my view, however, preserving section 1292(b)'s integrity and discouraging the kind of litigating tactics reflected in this record far outweigh the efficiency that hearing this interlocutory appeal might produce in this concededly complex case.

But the court disagrees with my position. The appeal stands before us, so in the following sections I exercise a dissenter's prerogative to address the merits. See, e.g., Gratz v. Bollinger, 539 U.S. 244, 291, 123 S.Ct. 2411, 156 L.Ed.2d 257 (2003) (Souter, J., dissenting); Arizona v. Evans, 514 U.S. 1, 18, 115 S.Ct. 1185, 131 L.Ed.2d 34 (1995) (Stevens, J., dissenting); Larson v. Valente, 456 U.S. 228, 258, 102 S.Ct. 1673, 72 L.Ed.2d 33 (1982) (White, J., dissenting).

II.

Like my colleagues, I begin with the structure and language of RICO's remedial provisions. RICO authorizes criminal penalties and civil remedies against those engaging in patterns of racketeering behavior. 18 U.S.C. § 1963 sets out the criminal penalties: guilty persons shall "be fined under this title or imprisoned . . . or both, and shall forfeit to the United States" any illegally acquired interest. Section 1964 provides for the civil remedies. At issue in this case is subsection (a), which states:

The district courts of the United States shall have jurisdiction to prevent and restrain violations of section 1962 of this chapter by issuing appropriate orders, including, but not limited to: ordering any person to divest himself of any interest, direct or indirect, in any enterprise; imposing reasonable restrictions on the future activities or investments of any person, including, but not limited to, prohibiting any person from engaging in the same type of endeavor as the enterprise engaged in, the activities of which affect interstate or foreign commerce; or ordering dissolution or reorganization of any enterprise, making due provision for the rights of innocent persons.

Another subsection, § 1964(c), authorizes injured persons to sue RICO violators for treble damages and to recover attorneys' fees. Finally, Congress directed that RICO "shall be liberally construed to effectuate its remedial purposes," Pub.L. No. 91-452, § 904(a), 84 Stat. 922, 947 (1970) (codified in a note following 18 U.S.C. § 1961) — a provision that, if it "is to be applied anywhere, [should be applied] in § 1964, where RICO's remedial purposes are most evident," Sedima, S.P.R.L. v. Imrex Co., 473 U.S. 479, 491 n. 10, 105 S.Ct. 3275, 87 L.Ed.2d 346 (1985).

The government argues that district courts have authority to order any remedy, including disgorgement, within their inherent equitable powers. More narrowly, the government argues that assuming the district courts may only impose equitable remedies for the purpose of keeping defendants from committing RICO violations, disgorgement — by reducing the incentives for the tobacco companies to violate RICO in the future — will accomplish that purpose in this case. These two distinct arguments present very different consequences for district courts: under the first theory, courts may order disgorgement any time they find the remedy necessary to ensure complete relief, while under the second theory courts may order disgorgement only to prevent ongoing or future violations. In this case, the district court accepted only the second argument. See 321 F.Supp.2d at 74-80. The court today rejects both.

A.

In dismissing the argument that district courts may impose any equitable remedy for RICO violations, the court distinguishes — unconvincingly, in my view — the two Supreme Court cases relied on by the government, Porter v. Warner Holding Co., 328 U.S. 395, 66 S.Ct. 1086, 90 L.Ed. 1332 (1946), and Mitchell v. Robert DeMario Jewelry, Inc., 361 U.S. 288, 80 S.Ct. 332, 4 L.Ed.2d 323 (1960). I believe these two cases control this case and compel the conclusion that district courts may impose any equitable remedy for RICO violations.

In Porter, the Supreme Court considered whether a district court had authority to order restitution in a suit brought by the Price Control Administrator against a landlord who had violated the Emergency Price Control Act (EPCA) by charging too much rent. The act contained no specific provision for restitution or disgorgement, but — like RICO — authorized a broad array of other remedies, both criminal and civil. On the criminal side, offenders could be fined and imprisoned. EPCA, § 205(b)-(c), 56 Stat. 23, 33 (1942). On the civil side, injured individuals could sue for treble damages plus attorneys' fees, and if they were not entitled to sue or the statutory period for their suit had passed, the Administrator could sue for the same remedy on behalf of the United States. Id. § 205(e), 56 Stat. at 34, as amended by Stabilization Extension Act of 1944, § 108(b), 58 Stat. 632, 640-41. The Administrator could also sue to suspend a violator's license. Id. § 205(f)(2), 56 Stat. at 35.

In the section most at issue in Porter, the act further provided that

[w]henever in the judgment of the Administrator any person has engaged or is about to engage in [violations of the act], he may make application to the appropriate court for an order enjoining such acts or practices, or for an order enforcing compliance with such provision, and upon a showing by the Administrator that such person has engaged or is about to engage in any such acts or practices a permanent or temporary injunction, restraining order, or other order shall be granted without bond.

Id. § 205(a), 56 Stat. at 33. Although this section clearly authorized injunctions aimed at future behavior, it made no express provision for restitution and did not, contrary to my colleagues' suggestion, explicitly "grant general equitable jurisdiction" to the district courts, see majority op. at 1197. Indeed, in Porter, the Eighth Circuit had held that district courts were without authority to order restitution as a remedy for violations of the EPCA. Bowles v. Warner Holding Co., 151 F.2d 529, 532 (8th Cir. 1945) (concluding that the district court had no authority to order restitution because "[i]t is well settled `That where a statute creates a right and provides a special remedy, that remedy is exclusive'") (citations omitted).

The Supreme Court reversed. Discussing "the jurisdiction of the District Court to enjoin acts and practices made illegal by the Act and to enforce compliance with the Act," 328 U.S. at 397-98, 66 S.Ct. 1086, the Court concluded — and I quote at length since the language is so critical to the disposition of this case — that

[s]uch a jurisdiction is an equitable one. Unless otherwise provided by statute, all the inherent equitable powers of the District Court are available for the proper and complete exercise of that jurisdiction. And since the public interest is involved in a proceeding of this nature, those equitable powers assume an even broader and more flexible character than when only a private controversy is at stake. . . . [T]he court may go beyond the matters immediately underlying its equitable jurisdiction and decide whatever other issues and give whatever other relief may be necessary under the circumstances. Only in that way can equity do complete rather than truncated justice.

Moreover, the comprehensiveness of this equitable jurisdiction is not to be denied or limited in the absence of a clear and valid legislative command. Unless a statute in so many words, or by a necessary and inescapable inference, restricts the court's jurisdiction in equity, the full scope of that jurisdiction is to be recognized and applied.

Id. at 398, 66 S.Ct. 1086 (citations omitted). The Court concluded that because the EPCA, despite the very detailed and specific nature of the authorized remedies, did not rule out restitution by a "necessary and inescapable inference," the district court could order restitution even if not expressly authorized by the statute. See id. at 398-400, 66 S.Ct. 1086; see also Mitchell, 361 U.S. at 291, 80 S.Ct. 332 (discussing Porter).

Indeed, the Court further suggested that restitution could be considered an "other order" to enjoin or enforce compliance within section 205(a) in either of two ways. First, it could be "considered as an equitable adjunct to an injunction decree" since "where, as here, the equitable jurisdiction of the district court has properly been invoked for injunctive purposes, the court has the power to decide all relevant matters in dispute and to award complete relief even though the decree includes that which might be conferred by a court of law." 328 U.S. at 399, 66 S.Ct. 1086. Second, restitution could "be considered as an order appropriate and necessary to enforce compliance with the Act" since "[f]uture compliance may be more definitely assured if one is compelled to restore one's illegal gains." Id. at 400, 66 S.Ct. 1086. The Court then remanded for the district court to "exercise the discretion that belongs to it" and decide whether to order restitution. Id. at 403, 66 S.Ct. 1086.

Porter was not unanimous. "It is not excessive to say that perhaps no other legislation in our history has equaled the Price Control Acts in the wealth, detail, precision and completeness of its jurisdictional, procedural and remedial provisions," id. at 404, 66 S.Ct. 1086, wrote Justice Rutledge in dissent. "The scheme of enforcement was highly integrated, with the parts precisely tooled and minutely geared." Id. "Congress could not have been ignorant of the remedy of restitution. It knew how to give remedies it wished to confer." Id. at 405, 66 S.Ct. 1086. "[E]ven courts of equity may not grant relief in disregard of the remedies specifically defined by Congress." Id. at 408, 66 S.Ct. 1086.

The court's opinion today sounds a lot like the Porter dissent. The court observes that the language of section 1964(a) — a court has "jurisdiction to prevent and restrain violations" — does not explicitly open the door to all of equity, but neither did EPCA section 205(a) (a court may issue orders "enjoining" violations or "enforcing compliance"). The court asserts that reading full equitable jurisdiction into RICO will render section 1964(a)'s language largely meaningless, but Porter rejected just this concern with regard to EPCA section 205(a). The court emphasizes that RICO "already provides for a comprehensive set of remedies," majority op. at 1200, but the EPCA had at least as comprehensive a remedial structure. The court further points out that should restitution be available, the government could obtain duplicative recovery (given RICO's criminal forfeiture provisions) and also escape the applicable statutes of limitations, but the Porter majority dismissed similar concerns, 328 U.S. at 401-02, 66 S.Ct. 1086; see also id. at 406-08, 66 S.Ct. 1086 (Rutledge, J., dissenting). Finally, the court attempts to distinguish Porter on the grounds that the EPCA had a different policy goal than RICO (preventing inflation rather than seeking to eradicate organized crime), but this has no effect on Porter's essential holding that "the court may go beyond the matters immediately underlying its equitable jurisdiction . . . and give whatever other relief may be necessary under the circumstances," see id. at 398, 66 S.Ct. 1086. In sum, the court offers no basis for concluding that RICO's structure and language get the statute past Porter's high bar for finding by a "necessary and inescapable inference" that Congress intended to empower district courts to order only limited equitable relief.

Nor does Philip Morris point to anything in RICO's legislative history that creates such a "necessary and inescapable inference." Only one remark even gives me pause. The Senate Committee report stated, "Subsection [1964](a) contains broad remedial provisions for reform of corrupted organizations. Although certain remedies are set out, the list is not exhaustive, and the only limit on remedies is that they accomplish the aim set out of removing the corrupting influence and make due provision for the rights of innocent persons." S.Rep. No. 91-617, at 160 (1969); accord H. Rep. No. 91-1549, at 57 (1970). The second part of this "limit" — requiring due provision for the rights of innocent persons — poses no concern, for it describes equity rather than constricts it. See, e.g., Holly v. Domestic Foreign Missionary Soc'y, 180 U.S. 284, 295, 21 S.Ct. 395, 45 L.Ed. 531 (1901) ("[A] court of equity will not transfer a loss that has already fallen upon one innocent party to another party equally innocent."). But the first part of this "limit" — that remedies should accomplish the aim of removing the corrupting influence — does more than simply restate an equitable principle. Suggesting that the remedies must remove the corrupting influence, it allows one to infer that remedies may accomplish only this aim. But that inference is, to use Porter's words, neither "necessary" nor "inescapable." One could also infer that remedies must accomplish this aim as a lower limit (i.e., no corrupting influence may remain), but may also accomplish other aims — just as remedies must make due provision for the rights of the innocent, but may presumably do much more. Indeed, this reading comports with how RICO's sponsor, Senator McClellan, described the bill when he introduced it: the "ability of our chancery courts to formulate a remedy to fit the wrong is one of the greatest benefits of our system of justice. This ability is not hindered by the bill." 115 Cong. Rec. 9567 (1969).

Mitchell, the second Supreme Court decision the government relies on, considered whether district courts could order restitution of wages lost from unlawful discharge in suits brought by the Secretary of Labor under section 17 of the Fair Labor Standards Act (FLSA), 29 U.S.C. § 217 (1960). Relying on Porter, the Court concluded that where the statute provided that "the district courts are given jurisdiction . . . for cause shown, to restrain violations" of the act, 29 U.S.C. § 217, district courts had full equitable powers, 361 U.S. at 291-95, 80 S.Ct. 332; see also id. at 289, 80 S.Ct. 332. Reaffirming Porter's strong presumption in favor of finding equitable relief fully available, the Court stated: "When Congress entrusts to an equity court the enforcement of prohibitions contained in a regulatory enactment, it must be taken to have acted cognizant of the historic power of equity to provide complete relief in the light of statutory purposes. As this Court long ago recognized, `there is inherent in the Courts of Equity a jurisdiction to . . . give effect to the policy of the legislature.'" Id. at 291-92, 80 S.Ct. 332 (quoting Clark v. Smith, 38 U.S. (13 Pet.) 195, 203, 10 L.Ed. 123 (1839)) (omission in original); see also Califano v. Yamasaki, 442 U.S. 682, 704-06, 99 S.Ct. 2545, 61 L.Ed.2d 176 (1979) (using the Porter presumption to conclude that district courts could order injunctive relief not explicitly authorized by the Social Security Act). The Mitchell Court thought it insignificant that because both the aggrieved employees and the Secretary could seek lost wages in actions at law under FLSA section 16, 29 U.S.C. § 216 (1960), duplicative recovery might occur. 361 U.S. at 292-93, 80 S.Ct. 332. But see id. at 303, 80 S.Ct. 332 (Whittaker, J., dissenting) (concluding that the statutory scheme "seems plainly to show that Congress intended by s 16(c) to allow recovery of unpaid minimum wages and overtime compensation at the instance of the Secretary only in an action at law, brought under that subsection, and triable by a jury"). Mitchell reinforces the proposition that district courts may order any equitable relief in civil RICO suits brought by the government. My colleagues suggest that in "the RICO Act, Congress provided a statute granting jurisdiction defined with the sort of limitations not present in the FLSA." Majority op. at 1199. The only jurisdictional hook in the FLSA's text, however, was its language: "the district courts are given jurisdiction . . . for cause shown, to restrain violations" of the act, 29 U.S.C. § 217. If this language opens the door to all equitable relief, then RICO's language — "[t]he district courts . . . shall have jurisdiction to prevent and restrain violations" — certainly does the same. And if the possibility of duplicative recovery did not circumscribe the district court's equitable authority under the FLSA, then neither should that possibility under RICO do so.

Not surprisingly, in the wake of Mitchell and Porter, circuit courts including this one have read general equitable jurisdiction into a variety of statutes that fail to provide explicitly for it. In SEC v. First City Financial Corp., 890 F.2d 1215 (D.C. Cir. 1989), we held that district courts may order disgorgement under the Security Exchange Act's sections 21(d) and (e), 15 U.S.C. § 78u(d)-(e) (1989), which provide that the district courts "shall have jurisdiction to issue writs of mandamus, injunctions, and orders commanding" compliance with the act and regulations made under it. See 890 F.2d at 1230 (relying on Porter and Mitchell). "Disgorgement, then, is available simply because the relevant provisions of the Securities Exchange Act of 1934, sections 21(d) and (e) . . . vest jurisdiction in the federal courts." Id.; see also SEC v. Tome, 833 F.2d 1086, 1096 (2d Cir. 1987); SEC v. Wash. County Util. Dist., 676 F.2d 218, 227 (6th Cir. 1982). Other circuits have reasoned similarly in interpreting other acts. See, e.g., FTC v. Gem Merch. Corp., 87 F.3d 466, 468-70 (11th Cir. 1996) (applying Porter in holding that courts may order restitution as a remedy for violations of the Federal Trade Commission Act); ICC v. B T Transp. Co., 613 F.2d 1182, 1183-86 (1st Cir. 1980) (applying Porter in holding that courts may order restitution as a remedy for violations of the Motor Carrier Act, though noting that "[i]f we were writing on a blank slate, we might agree with the district court that the language of the Motor Carrier Act cannot justify" the remedy of restitution); CFTC v. Hunt, 591 F.2d 1211, 1221-23 (7th Cir. 1979) (applying Porter in holding that courts may order disgorgement as a remedy for violations of the Commodity Exchange Act).

Instead of following Porter and Mitchell, the court relies on a later Supreme Court decision, Meghrig v. KFC Western, Inc., 516 U.S. 479, 116 S.Ct. 1251, 134 L.Ed.2d 121 (1996). In Meghrig, the Supreme Court considered whether private citizens could seek restitution under the Resource Conservation and Recovery Act (RCRA) for the cost of having cleaned up a prior landowner's toxic waste. The statute provided that the "district court shall have jurisdiction . . . to restrain any person who has contributed or who is contributing" to waste problems, "to order such person to take such other action as may be necessary, or both." Id. at 482 n. *, 116 S.Ct. 1251 (quoting 42 U.S.C. § 6972(a)). The Court held that it was "apparent from the two remedies described . . . that RCRA's citizen suit provision is not directed at providing compensation for past cleanup efforts." Id. at 484, 116 S.Ct. 1251. While not explicitly defining the limits of the two remedies described, the court suggested that these remedies should be equated with prohibitory and mandatory injunctions. Id. Moreover, relying in part on the fact that an analogous statute expressly authorized damages, the Court concluded that "neither remedy . . . contemplates the award of past cleanup costs, whether these are denominated `damages' or `equitable restitution.'" Id. at 484-85, 116 S.Ct. 1251. According to the Court, it "`is an elemental canon of statutory construction that where a statute expressly provides a particular remedy or remedies, a court must be chary of reading others into it.'" Id. at 488, 116 S.Ct. 1251 (quoting Middlesex County Sewerage Auth. v. Nat'l Sea Clammers Ass'n, 453 U.S. 1, 14-15, 101 S.Ct. 2615, 69 L.Ed.2d 435 (1981)).

The Meghrig Court noted that in arguing that the district court had inherent authority to award equitable remedies, the plaintiffs relied on Porter and its progeny. Id. at 487, 116 S.Ct. 1251. Without expressly distinguishing those cases, the Court explained that "the limited remedies described in [RCRA], along with the stark differences between the language of that section and the cost recovery provisions [of the analogous statute], amply demonstrate that Congress did not intend for a private citizen to be able to undertake a cleanup and then proceed to recover its costs under RCRA." Id. at 487, 116 S.Ct. 1251. Notably for our purposes, Meghrig did not overrule Porter. Indeed, even after Meghrig, the Supreme Court has cited Porter for the proposition that "we should not construe a statute to displace courts' traditional equitable authority absent . . . an `inescapable inference' to the contrary." Miller v. French, 530 U.S. 327, 340, 120 S.Ct. 2246, 147 L.Ed.2d 326 (2000); see also United States v. Oakland Cannabis Buyers' Co-op., 532 U.S. 483, 496, 121 S.Ct. 1711, 149 L.Ed.2d 722 (2001).

At one level, reconciling Meghrig with Porter and Mitchell is difficult. Meghrig suggests that "to restrain" only authorizes prohibitory injunctions. By contrast, Mitchell holds that this language imposes no limit on the district court's full equitable powers. Meghrig, relying on a version of the canon expressio unius est exclusio alterius, observes that courts should be "chary" in reading remedies into a statute which expressly provides for other remedies. By contrast, Porter indicates that in the context of equity jurisdiction, the general expressio unius canon gets inverted, meaning that district courts possess all equitable powers unless the statute "inescapabl[y]" provides to the contrary. Cf. Renegotiation Bd. v. Bannercraft Clothing Co., 415 U.S. 1, 18-20, 94 S.Ct. 1028, 39 L.Ed.2d 123 (1974) (discussing these competing canons).

These tensions cannot be dealt with simply by dismissing Porter and Mitchell. Meghrig not only left both cases intact, but also suggested that the "limited remedies" in RCRA, together with the "stark differences" between RCRA and the analogous statute, explain the different outcomes. Given this, our responsibility is to follow the Supreme Court's oft-cited instruction that "[i]f a precedent of this Court has direct application in a case, yet appears to rest on reasons rejected in some other line of decisions, the Court of Appeals should follow the case which directly controls, leaving to this Court the prerogative of overruling its own decisions." Rodriguez de Quijas v. Shearson/Am. Express, Inc., 490 U.S. 477, 484, 109 S.Ct. 1917, 104 L.Ed.2d 526 (1989); see also Agostini v. Felton, 521 U.S. 203, 237, 117 S.Ct. 1997, 138 L.Ed.2d 391 (1997) (reaffirming this requirement).

In my view, Porter and Mitchell, not Meghrig, "directly control" this case. Several reasons support this conclusion, and nothing points the other way. First, RICO's statutory scheme resembles the EPCA more than the RCRA. Both RICO and the EPCA stand alone in grappling with a broad social issue, whereas the RCRA had a closely related statute on which the Court in Meghrig relied heavily. Second, as in both Porter and Mitchell, the government brought the suit rather than a private party like the Meghrig plaintiff, and Porter makes clear that district courts may have "even broader and more flexible" equitable powers where the public interest is involved, 328 U.S. at 398, 66 S.Ct. 1086. This point has particular traction if the government is the only party that may seek equitable relief under RICO. See Religious Tech. Ctr. v. Wollersheim, 796 F.2d 1076, 1083-89 (9th Cir. 1986) (holding that equitable relief under RICO is available only to the government). But see Nat'l Org. for Women, Inc. v. Scheidler, 267 F.3d 687, 695-700 (7th Cir. 2001) (holding that private plaintiffs can seek equitable relief under RICO), rev'd on other grounds, 537 U.S. 393, 123 S.Ct. 1057, 154 L.Ed.2d 991 (2003). Finally, Meghrig's suggestion that "restrain" in the RCRA refers only to prohibitory injunctions cannot apply to section 1964(a), since that section explicitly authorizes other remedies — e.g., divestment — to "prevent and restrain" RICO violations. For these reasons, in determining whether the phrase "prevent and restrain" limits the district court's equitable powers, I think it makes more sense to look to Porter and Mitchell, not Meghrig.

The court "[r]ead[s] Porter in light of" the statement in Kokkonen v. Guardian Life Insurance Co., 511 U.S. 375, 377, 114 S.Ct. 1673, 128 L.Ed.2d 391 (1994), that "`[f]ederal courts are courts of limited jurisdiction'" and "`possess only that power authorized by Constitution and statute, which is not to be expanded by judicial decree.'" Majority op. at 1197-98. But "`[j]urisdiction,' it has been observed, `is a word of many, too many, meanings.'" Steel Co. v. Citizens for a Better Env't, 523 U.S. 83, 90, 118 S.Ct. 1003, 140 L.Ed.2d 210 (1998) (citation omitted). Kokkonen simply makes the unremarkable point that federal courts have subject-matter jurisdiction over cases only if the Constitution or Congress so provides, 511 U.S. at 377, 114 S.Ct. 1673, and the Supreme Court has since clarified that it is "unreasonable" to apply subject-matter jurisdiction principles where a statute uses the term jurisdiction "merely [in] specifying the remedial powers of the court," 523 U.S. at 90, 118 S.Ct. 1003.

Finally, while Congress modeled section 1964(a) on the antitrust laws, see 115 Cong. Rec. 9567 (1969) (statement of Sen. McClellan); see also 15 U.S.C. § 4 (the "district courts . . . are invested with jurisdiction to prevent and restrain violations"); accord 15 U.S.C. § 25, I disagree with Philip Morris that the Supreme Court's antitrust decisions provide useful guidance as to whether the phrase "prevent and restrain" limits the equitable remedies available to district courts. On the one hand, the Court once ignored, though did not explicitly reject, an invitation by Justice Douglas to apply Porter to antitrust actions. See United States v. Nat'l Lead Co., 332 U.S. 319, 366-67, 67 S.Ct. 1634, 91 L.Ed. 2077 (1947) (Douglas, J., dissenting in part); cf. United States v. Oregon State Med. Soc'y, 343 U.S. 326, 333, 72 S.Ct. 690, 96 L.Ed. 978 (1952) (emphasizing that in antitrust actions the purpose of injunctive relief is to "forestall future violations"); Texas Indus., Inc. v. Radcliff Materials, Inc., 451 U.S. 630, 639-47, 101 S.Ct. 2061, 68 L.Ed.2d 500 (1981) (declining to fashion and apply a common law right of contribution in the antitrust context). On the other hand, some antitrust cases suggest that courts may impose equitable remedies beyond those intended merely to stop future violations from occurring. E.g., United States v. Crescent Amusement Co., 323 U.S. 173, 189, 65 S.Ct. 254, 89 L.Ed. 160 (1944) (although the district court ordered a remedy said to "exceed any reasonable requirement for prevention of future violations," the "Court has quite consistently recognized in this type of Sherman Act case that the government should not be confined to an injunction against further violations. . . . Those who violate the Act may not reap the benefits of their violations"); cf. United States v. U.S. Steel Corp., 251 U.S. 417, 452, 40 S.Ct. 293, 64 L.Ed. 343 (1920) (observing that the Sherman Act is "clear in its direction that the courts of the nation shall prevent and restrain [monopolies] (its language is `to prevent and restrain violations of' the act); but the command is necessarily submissive to the conditions which may exist and the usual powers of a court of equity to adapt its remedies to those conditions"); Schine Chain Theatres v. United States, 334 U.S. 110, 128, 68 S.Ct. 947, 92 L.Ed. 1245 (1948) (suggesting that "[l]ike restitution," divestment "merely deprives a defendant of the gains from his wrongful conduct" and upholding it as a remedy under the Sherman Act), overruled on other grounds by Copperweld Corp. v. Indep. Tube Corp., 467 U.S. 752, 763 n. 8, 777, 104 S.Ct. 2731, 81 L.Ed.2d 628 (1984). As these cases illustrate, antitrust precedent offers little reason to doubt the applicability of Porter and Mitchell to the case at hand.

To sum up, Porter and Mitchell rather than Meghrig control this case, and no "necessary and inescapable inference" limits the district court's jurisdiction in equity. If the district court concludes that the government has shown that the tobacco companies have committed RICO violations by advertising to youth despite assertions to the contrary and by falsely disputing smoking's addictive, unhealthy effects, then it may order whatever equitable relief it deems appropriate. Of course, the court must work within the bounds of equitable doctrines, recognizing defenses like laches and unclean hands, paying due regard for the rights of the innocent, and generally exercising its discretion. With these principles in mind, the district court can "do complete rather than truncated justice," Porter, 328 U.S. at 398, 66 S.Ct. 1086.

B.

In addition to rejecting the government's argument that district courts may impose any equitable remedy on RICO violators, the court rejects the government's alternative, narrower argument — that even if district courts may order only remedies that "prevent and restrain" RICO violations, disgorgement can appropriately accomplish that purpose. Because the court's analysis of this argument is as flawed as its analysis of the government's broader argument, I add this discussion of the issue. In my view, the court transforms what should be a question of fact — what remedies appropriately prevent and restrain future violations — into a question of statutory interpretation in a way that disregards section 1964(a)'s plain language and ignores Supreme Court precedent recognizing the equitable flexibility of district courts.

Under section 1964(a), district courts may issue "appropriate orders" to prevent and restrain" RICO violations. "Prevent" has many meanings. The first nonarchaic one listed in Webster's Third New International Dictionary (1961) is "to deprive of power or hope of acting, operating, or succeeding in a purpose." "Restrain" can mean "to hold (as a person) back from some action, procedure, or course: prevent from doing something (as by physical or moral force or social pressure)" and "to limit or restrict to or in respect to a particular action or course: keep within bounds or under control." Webster's Third New International Dictionary (1961).

The government offers expert testimony to the effect that a disgorgement order will deter the tobacco companies from violating RICO in the future — in the dictionary's language, it will deprive them of the hope of succeeding in benefiting from future RICO violations and hold them back from committing such violations. In essence, the government claims that the tobacco companies, having engaged in a persistent pattern of deceptive representations over decades, will be less likely to continue this illegal behavior if they must surrender their past ill-gotten profits. Treating the government's expert testimony as correct, as we must at this stage of the litigation, see Anderson, 477 U.S. at 255, 106 S.Ct. 2505, I think it enough to forestall summary judgment in Philip Morris's favor. Indeed, the Supreme Court has accepted just this theory of deterrence, stating in Porter that restitution "could be considered as an order appropriate and necessary to enforce compliance with the Act" since "[f]uture compliance may be more definitely assured if one is compelled to restore one's illegal gains." 328 U.S. at 400, 66 S.Ct. 1086. If restitution helps enforce compliance, then we should have little doubt that disgorgement helps prevent and restrain violations.

This court does not conclude that disgorgement can never have a restraining effect on future conduct of the defendants — the only conclusion that could justify a holding that district courts can never order disgorgement under section 1964(a). Instead, the court offers several unpersuasive reasons for its conclusion that as a matter of statutory interpretation disgorgement is not a permissible remedy under section 1964(a).

First, the court states that disgorgement "is a quintessentially backward-looking remedy." Majority op. at 1198. Although I agree that a court sitting in equity cannot order disgorgement that exceeds a defendant's past ill-gotten profits, see Tull v. United States, 481 U.S. 412, 424, 107 S.Ct. 1831, 95 L.Ed.2d 365 (1987) (observing that "[r]estitution is limited to `restoring the status quo and ordering the return of that which rightfully belongs to the purchaser or tenant'") (quoting Porter, 328 U.S. at 402, 66 S.Ct. 1086), this does not mean disgorgement is always backward-looking and can never have a forward-looking effect on the defendants. The Supreme Court made this clear in Porter, 328 U.S. at 400, 66 S.Ct. 1086, and Meghrig nowhere rejects Porter's conclusion that a disgorgement order can impact future conduct — indeed, there was no evidence in Meghrig that the defendants were likely to commit future RCRA violations, and in any event, as discussed supra at 1220-21, Porter and Mitchell are the cases most directly on point for our purposes.

Second, the court concludes that district courts are limited not merely by the words "prevent and restrain," but also "by those [three remedies] explicitly included in the statute" by application of the canons noscitur a sociis and ejusdem generis. See majority op. at 1200; cf. United States v. Thomas, 361 F.3d 653, 659 (D.C. Cir. 2004) (defining these canons). Even assuming we should apply these canons, however, they spell out nothing more than what everyone agrees on: that the only "appropriate" orders under this section are equitable ones. See West v. Gibson, 527 U.S. 212, 225-26, 119 S.Ct. 1906, 144 L.Ed.2d 196 (1999) (Kennedy, J., dissenting) (observing that these canons "suggest the appropriate remedies authorized by [a statute using the word `including'] are remedies of the same nature as reinstatement, hiring, and backpay — i.e., equitable remedies" and noting that "the phrase `appropriate remedies,' furthermore, connotes the remedial discretion which is the hallmark of equity").

More important, I doubt the canons apply here at all. While the canons can prove useful where there is otherwise "no general principle in sight," Dong v. Smithsonian Inst., 125 F.3d 877, 880 (D.C. Cir. 1997); see also Wash. State Dep't of Health Servs. v. Guardianship Estate of Keffeler, 537 U.S. 371, 384, 123 S.Ct. 1017, 154 L.Ed.2d 972 (2003) (applying the canons in interpreting the last listed term of "execution, levy, attachment, garnishment, or other legal process"), here the statute provides the general principle of preventing and restraining violations. Indeed, the Supreme Court declined to use these canons altogether in interpreting a statute which gave the EEOC the power of enforcement "through appropriate remedies, including reinstatement or hiring of employees with or without back pay," 42 U.S.C. § 2000e-16(b). See West, 527 U.S. at 218, 119 S.Ct. 1906 (stating that the "word `including' makes clear that `appropriate remedies' are not limited to the examples that follow that word"); cf. Harrison v. PPG Indus., Inc., 446 U.S. 578, 588-89, 100 S.Ct. 1889, 64 L.Ed.2d 525 (1980) (declining to apply ejusdem generis canon where Congress used "expansive language"). I see no reason why we should do otherwise here, especially since section 1964(a) uses the even more expansive language: "including, but not limited to." Finally, noscitur a sociis and ejusdem generis should not be used to limit the types of equitable relief available to district courts given Congress's instruction that RICO "shall be liberally construed to effectuate its remedial purposes," see supra at 1215, one of which is preventing and restraining future violations — an aim that, far from being a "new purpose that Congress never intended," see majority op. at 1201 (quoting Reves v. Ernst Young, 507 U.S. 170, 183, 113 S.Ct. 1163, 122 L.Ed.2d 525 (1993)), expressly appears in the statute's text. If an equitable remedy achieves this goal, then the statute authorizes it.

Third, the court suggests that disgorgement should be unavailable because it allows the government to achieve relief "similar in effect" to criminal forfeiture, raising concerns that the government can achieve duplicative recovery and evade the procedural safeguards girding the forfeiture provision. See majority op. at 1200-01. To be sure, such concerns are relevant in considering whether to infer additional causes of action. As discussed earlier, supra at 1217, however, given the Supreme Court's explicit rejection of similar concerns in Porter and Mitchell, they cannot carry the day. Nor should such concerns stop a court from issuing equitable orders that accomplish the express statutory purpose of preventing and restraining RICO violations, whether the remedies are specifically listed in section 1964(a), e.g., divestment, or available as other "appropriate orders." Discussing RICO, the Supreme Court has observed that "Congress has provided civil remedies for use when the circumstances so warrant. It is untenable to argue that their existence limits the scope of the criminal provisions." United States v. Turkette, 452 U.S. 576, 585, 101 S.Ct. 2524, 69 L.Ed.2d 246 (1981). The converse should hold as well. If an equitable remedy prevents and restrains RICO violations — one of the remedial purposes which we should liberally construe the statute to effectuate — it is untenable to claim that the existence of criminal provisions renders this remedy nonetheless beyond the scope of district court authority.

Of course, that disgorgement may sometimes serve to prevent and restrain defendants from committing RICO violations does not mean that it will always accomplish that purpose. As the district court here recognized, a court must first find that the defendants are likely to commit future RICO violations. 321 F.Supp.2d at 75-76. This is not a foregone conclusion. In Carson, for example, while the Second Circuit recognized that disgorgement can sometimes serve to prevent and restrain RICO violations, it was rightly skeptical that disgorgement of the "gains ill-gotten long ago by a retiree" who had long since left the union position that he had abused in accepting kickbacks would accomplish this purpose. 52 F.3d at 1182. Assuming district courts are limited to remedies that prevent and restrain, but see supra Part II.A, I also share the Second Circuit's apparent conclusion that disgorgement may be ordered only to prevent and restrain a defendant from future RICO violations, see 52 F.3d at 1182. But see Richard v. Hoechst Celanese Chem. Group, 355 F.3d 345, 355 (5th Cir. 2003) (leaving open the possibility that disgorgement might be ordered solely to deter other possible offenders). Because any remedy imposed for a solely exemplary purpose (i.e., to dissuade others from committing RICO violations) would amount to punishment, it goes beyond what Congress intended, see S.Rep. No. 91-617, at 81, as well as pushing the boundaries of what equity permits, cf. Tull, 481 U.S. at 422, 107 S.Ct. 1831. In this case, however, the government offers evidence that the defendant companies themselves are likely to commit future RICO violations by misleading the public about the health consequences of smoking and the addictive effects of nicotine, as well as by persisting in marketing to young people.

According to Philip Morris, only injunctions are "appropriate orders" under section 1964(a) because, in its view, they will always adequately prevent past lawbreakers from committing future violations, particularly given the threat of heavy contempt penalties. Refining this point, the concurrence finds it "almost inconceivable" that disgorgement can change the incentives governing a defendant's future behavior given RICO's other provisions. See sep. op. at 1203-04 (Williams, J., concurring). The concurrence thus concludes that as a matter of law, Congress intended to exclude disgorgement from those remedies appropriate to prevent and restrain RICO violations. See id. at 1204-05. I think this approach is flawed in several respects.

To begin with, as noted above, Porter indicated that disgorgement may encourage guilty defendants to obey the law in the future. Interpreting a statute replete (like RICO) with other remedies, the Court concluded that "[f]uture compliance may be more definitely assured if one is compelled to restore one's illegal gains." 328 U.S. at 400, 66 S.Ct. 1086. We are without license to ignore the Supreme Court's views on this point.

Moreover, Philip Morris's suggestion that only injunctions provide "appropriate" relief under section 1964(a) not only cuts against the statute's plain language — Congress would hardly have included divestment in its list of sample remedies if it thought injunctions alone would be adequate — but also ignores the equitable flexibility the statute was designed to preserve, see, e.g., 115 Cong. Rec. 9567 (1969) (statement of Sen. McClellan). Indeed, nothing in the statute requires courts to prefer contempt penalties (not explicitly named in section 1964(a)) to disgorgement (also not explicitly named). Rather, no single remedy is always appropriate. "The essence of equity jurisdiction has been the power of the Chancellor to do equity and to mold each decree to the necessities of the particular case. Flexibility rather than rigidity has distinguished it." Swann v. Charlotte-Mecklenburg Bd. of Educ., 402 U.S. 1, 15, 91 S.Ct. 1267, 28 L.Ed.2d 554 (1971) (quoting Hecht Co. v. Bowles, 321 U.S. 321, 329-30, 64 S.Ct. 587, 88 L.Ed. 754 (1944)). Sometimes injunctive relief alone will make the most sense; other times, different equitable remedies or combinations of equitable remedies, perhaps including disgorgement, might prove as or more effective.

To be sure, given RICO's comprehensive remedial scheme, disgorgement orders may prove appropriate in preventing and restraining future violations only in rare circumstances. But "[i]n equity, as nowhere else, courts [should] eschew rigid absolutes," Franks v. Bowman Transp. Co., 424 U.S. 747, 777 n. 39, 96 S.Ct. 1251, 47 L.Ed.2d 444 (1976) (internal quotation marks and citation omitted), and precisely what remedy or combination of remedies, within the bounds of the equitable doctrines discussed earlier, will serve to prevent and restrain defendants from committing RICO violations, is an issue of fact, not statutory interpretation. For these determinations, we must rely in the first instance not on what we appellate judges can or cannot imagine will "prevent or restrain," but on tried and true methods of fact-finding before district courts — including cross-examination and presentation of contrary evidence. Cf. id. at 780, 96 S.Ct. 1251 (noting district courts' "`keener appreciation' of peculiar facts and circumstances") (citation omitted).

Finally, and again as noted earlier, record evidence in this case suggests that disgorgement will in fact "prevent and restrain" defendants from committing future RICO violations. As one of the government's experts stated, "[R]equiring defendants to pay proceeds will affect their expectations . . . about the returns from future misconduct." Appellee's App. at 813. The expert added that, even if coupled with an injunction laden with contempt penalties, disgorgement will "provide additional economic incentives to deter future misconduct" by "strengthen[ing] the credibility of existing laws" which the defendants have allegedly violated in the past. Id. at 814. Disagreeing, the concurrence offers its own "expert opinion" of the incentives driving the behavior of past RICO violators. See sep. op. at 1203-05, 1205-06. According to the concurrence, the most appropriate deterrence will stem from the "spotlight of the lawsuit," if properly "amplif[ied]" by "transparency-enhancing and prior-approval measures." Id. at 1205-06. Perhaps so, but "on summary judgment, the evidence should be viewed in favor of the nonmoving party, not," as the concurrence would have it, "the other way around." Langon v. Dep't Health Human Servs., 959 F.2d 1053, 1059 (D.C. Cir. 1992) (reversing district court grant of summary judgment where that court disregarded admissible expert testimony); see also Sears, Roebuck Co. v. Gen. Servs. Admin., 553 F.2d 1378, 1381-83 (D.C. Cir. 1977) (holding that district court inappropriately granted summary judgment where experts disagreed about whether certain data constituted a "trade secret" from which an intelligent competitor could gain information). At this stage of the litigation, then, we must assume that the government expert is correct and that disgorgement will "prevent and restrain" future RICO violations. Should Philip Morris offer expert testimony along the lines suggested by the concurrence, then it will be up to the district court to evaluate the competing evidence and make appropriate findings of fact. Should either party appeal, this court, unrestrained by the inferences required at summary judgment, would then review that factual determination pursuant to Rule 52's clear error standard. See Fed.R.Civ.P. 52 advisory committee's note (observing that judgment under this standard "differs from a summary judgment under Rule 56 in the nature of the evaluation made by the court"); see also 9A Wright Miller, Federal Practice and Procedure § 2585 (2d. ed. 1994) (noting that under Rule 52 a reviewing court need not view the evidence in the light most favorable to the appellee).

C.

In sum, were this case properly before us, I would hold, in accordance with Porter and Mitchell, that district courts have authority to order any remedy, including disgorgement, necessary to ensure complete relief. As the concurrence points out, sep. op. at 1206 (Williams, J., concurring), my approach would create a circuit split, since Carson did not apply Porter and Mitchell to RICO (and, indeed, the parties do not appear to have brought these cases to the Second Circuit's attention). Even if, as Carson holds, district courts may only impose equitable remedies for the purpose of keeping defendants from committing RICO violations, I would still affirm the denial of summary judgment, leaving it to the district court to determine, on the basis of a fully developed record, whether disgorgement will help accomplish this purpose. I disagree with my colleagues' conclusions not because they have created a circuit split of their own by rejecting Carson's holding that disgorgement may prevent and restrain RICO violations, but because they have done so by accepting an interlocutory appeal that we should not hear and by disregarding both Supreme Court precedent and section 1964(a)'s plain language.

III.

This leaves one final, distinct issue. Philip Morris claims that the government's disgorgement model fails as a matter of law to measure the tobacco companies' ill-gotten profits. Because the district court decided this issue in the certified order, it is — unlike the issue the court does resolve — properly before us. See Yamaha, 516 U.S. at 205, 116 S.Ct. 619.

In calculating disgorgement, the government first identifies what it calls the "Youth Addicted Population" (YAP), namely, all people who were smoking an average of at least 5 cigarettes a day at the time they turned 21. The government next calculates that from RICO's effective date in 1970 to 2001, the tobacco companies earned profits of $280 billion through sales to these people. The government arrives at this calculation by (1) determining the gross revenue from these total sales minus the direct costs (excluding overhead and taxes) and (2) adjusting for the time value of money. Philip Morris asserts that the government has failed to show that these profits are attributable to the companies' alleged RICO violations, relying on admissions by government experts that it would be "highly unlikely" to say that "nobody under the age of 21 would have ever smoked regularly . . . but for the defendants' alleged RICO violations."

Philip Morris cannot prevail on this issue at summary judgment because the government need not show that nobody under 21 would have smoked but for the RICO violations. As we held in First City Financial, 890 F.2d at 1229, "disgorgement need only be a reasonable approximation of profits causally connected to the violation." In First City Financial, we found that the district court appropriately ordered disgorgement of all profits on a stock sale where the defendants failed to make a material disclosure, purchased stock whose value would likely have already risen had the disclosure been made, and then sold the stock for a killing after the undisclosed news broke. See id. at 1229-32. Although the government never proved that all increases in the stock's value stemmed from the violation, we rejected the defendants' argument that because the increase in price may have depended on other factors, disgorgement of all profits was "simplistic, quite unrealistic, and so de facto punitive." See id. at 1231. Noting that "[r]ules for calculating disgorgement must recognize that separating legal from illegal profits exactly may at times be a near-impossible task," we held that "the government's showing of appellants' actual profits on the tainted transactions at least presumptively satisfied" its "burden of persuasion that its disgorgement figure reasonably approximates the amount of unjust enrichment." Id. at 1231-32. Although recognizing that this might result in "actual profits becoming the typical disgorgement measure," we observed that "the risk of uncertainty should fall on the wrongdoer whose illegal conduct created that uncertainty." Id. at 1232; see also SEC v. Banner Fund Int'l, 211 F.3d 602, 617 (D.C. Cir. 2000).

Disentangling the tobacco companies' legal and illegal profits might also be a "near-impossible task." The government offers evidence that the tobacco companies not only fraudulently suggested that smoking was harmless and nonaddictive, but did so through a comprehensive, decades-long pattern of deliberate behavior. The government further offers evidence that advertising is a "very substantial influence on young people starting to smoke," see Appellee's App. at 783, and that the tobacco companies committed RICO violations in advertising to young people while publicly denying that they were doing so. Under First City Financial, then, the government's calculations serve as a reasonable approximation: just as we permit actual profits in insider trading cases to serve as a proxy for ill-gotten gains, so too can actual profits from sales to the YAP meet the government's initial burden of reasonably approximating the tobacco companies' unlawful gains. The burden would thus shift to Philip Morris to "demonstrate that the disgorgement figure was not a reasonable approximation," 890 F.2d at 1232, and the district court would have to sort out who is right.