Full opinion text
ORDER HAND, Senior District Judge. The United States filed this action under the Comprehensive Environmental Response, Compensation, and Liability Act (“CERCLA”), 42 U.S.C. §§ 9601, et seq., against defendant Olin Corporation (“Olin”), a Virginia corporation which owns and operates a chemical plant in McIntosh, Alabama, in southwest Washington County, approximately 25 miles east of the Mississippi state line and 75 miles north of the Gulf of Mexico. With the complaint, the parties filed a proposed consent degree. This court requested the parties to brief certain statutory and constitutional issues relating to CERCLA. For the reasons stated below, this court has determined that rather than signing the consent decree, it must dismiss the action both because 1) Congress did not clearly express its intent that the liability provision of CERCLA be retroactive, as required by Landgraf v. USI Film Products, 511 U.S. 244, 114 S.Ct. 1483, 128 L.Ed.2d 229 (1994) and 2) the application of CERCLA, at least on the facts of this case, violates the Commerce Clause as interpreted in United States v. Lopez, 514 U.S. —, 115 S.Ct. 1624, 131 L.Ed.2d 626 (1995). I. THE COMPLAINT AND PROPOSED CONSENT DECREE A. The Olin Property in the Complaint In its complaint, the Justice Department alleges that the Olin plant site in McIntosh, comprising 1,500 acres, contains two actionable sites. This action concerns only the first of two sites at the Olin property. “Site 1” includes 20 acres on the southern edge of the property, on which an active chemical-production facility operates. This site contains a number of “solid waste-management units,” both active and inactive, many of which have been closed and treated for the removal of hazardous substances. To the north and northwest of the active chemical production facility are undeveloped tracts of land. To the west of the facility is a brine-well field. Four thousand feet east of the main plant area is a steep bluff which defines the edge of Site 2, the natural basin. The government alleges that in 1952 Olin Mathieson began operating a mercury-cell chloralkali plant on Site 1 which generated and released wastewater containing mercury into Site 2 until 1974. This plant ceased operating in 1982. In 1955, on Site 1, Olin Mathieson built a “crop-protection-chemicals” plant which discharged wastewater into Site 2 until 1974. This plant also ceased operating in 1982. As a result of the operation of these two plants (1950’s — late 1982), mercury and chloroform, which are alleged to be hazardous substances under 42 U.S.C. § 9601(14), were released into Site 1. Most of the alleged contamination resulting from the operation of these two plants occurred prior to the effective date of CERCLA, December 11, 1980. However, to the extent that both of these plants also operated after CERCLA’s enactment, i.e., from December 11, 1980, until late 1982, and to the extent that a threat of continuing releases at and from Site 1 continues, as the government alleges, the government also seeks to recover CERCLA cleanup costs from Olin for post-enactment conduct. In 1978, the company (whose name changed that year to Olin Corporation) built a diaphragm-cell caustic-soda/chlorine plant which it currently operates. There is no allegation in the complaint that the operation of this plant has any connection with the contaminants found on Site 1. The Environmental Protection Agency (“EPA”) listed the Olin plant site on the “national priorities list” in September 1984. Cf. 42 U.S.C. § 9605; 40 C.F.R. Part 300. The complaint indicates that Olin conducted a “remedial investigation” of Site 1, completed in July 1993, and then a “feasibility study” of proposed responses to contamination at Site 1, completed in February 1994. In December 1994, the EPA executed its “record of decision” based on the remedial investigation and the feasibility study. The record of decision provides a remedy for the contamination of the alluvial aquifer which the agency contends is consistent with CERCLA and the national contingency plan. For present purposes, the court accepts the allegations of the complaint as true and notes, in fairness, that Olin’s answer admits many, while denying some, of them. For example, the defendant admits that there have been releases in the past, but denies that there is any threat of future release of hazardous substances. The defendant also denies any imminent and substantial danger to the public health or welfare, or the environment, because of actual or threatened releases from Site 1. B. The Proposed Consent Decree and Remedial-Investigation Report With the complaint, the Justice department filed a proposed consent decree, which both parties had signed. The government published notice thereof in the Federal Register to afford the public opportunity for comment. 603 Fed.Reg. 36309, 36309-10 (1995). When the government received no comment within 30 days, the parties jointly moved for the court to enter the. consent decree. See 42 U.S.C. § 9622(d)(2). The consent decree runs to 71 pages. It makes the defendant, its officers, directors, and anyone else associated with the defendant, liable for everything even remotely associated with the clean-up of Site 1, including insuring the automobiles the government uses in fulfilling and supervising the consent decree. Although the EPA estimates the present value of the cost of compliance to Olin at $10,339,000, the EPA has an almost unconstrained right to amend the consent decree. Of course, the defendant may appeal, but the defendant must’ direct the appeal to the EPA. The propriety of such procedures is beyond the scope of this inquiry, yet defense counsel has acknowledged that the consent decree gives the EPA carte blanche over the company treasury. Olin has committed itself to performing the actions which the consent decree specifically requires, regardless of whether the court approves of the decree; however, it has sought to do so under the supervision of the Alabama Department of Environmental Management (“ADEM”) rather than under the EPA. The ADEM has sought the same. Twice in May 1994, the ADEM’s director wrote to the EPA to request that the EPA allow the ADEM to implement the record of decision for Site 1 under the Resource Conservation and Recovery Act, 42 U.S.C. §§ 6901 et seq., and the Alabama Hazardous Waste Management and Minimization Act, Ala.Code §§ 22-30-1 et seq., rather than CERCLA. In December 1994, the EPA’s regional administrator in Atlanta, Georgia, wrote the ADEM director to convey the EPA’s belief that it would be “premature at this time to defer” to the ADEM. The regional administrator stated that accepting the state’s proposal would require modifying a permit under the Resource Conservation and Recovery Act, which would “add this additional layer of involvement and federal resources to the already complex situation.” He further asserted that “continued implementation under CERCLA insures consistency and decreases the potential for conflict.” Be that as it may, consistency and the decreased potential for conflict are not necessarily synonymous with constitutionality. Four days after rejecting the ADEM’s request, the EPA issued its record of decision, the major portions of which require Olin to (1) pump and treat additional ground water, (2) upgrade and extend the existing caps over closed portions of Site 1, (3) increase monitoring of ground water, and (4) establish certain institutional controls. When the court inquired at oral argument why the defendant negotiated a settlement and entered into this “consent” decree, counsel for the defendant responded that it was “a pragmatic business judgment.” It believed that “the fastest and most expedient way to get the work performed would be to simply go along with what the EPA sought here.” Olin wanted to complete the work and went along despite reservations about its legality and despite the government’s rigid dictation of the terms and conditions it would accept. In addition to the consent decree, the government filed a remedial-investigation report, prepared by Woodward-Clyde Consultants of Baton Rouge, Louisiana, in July 1993. The report documents remedial-investigation activities under an amended work plan and was completed under EPA supervision and ultimately with EPA approval. The objectives of the remedial investigation were (1) to investigate the nature, lateral extent, and vertical extent of contamination in order to ascertain the types, concentrations, and distributions of waste in all potentially affected media, such as the air, groundwater, soil, surface water, and sediment; (2) to determine the potential of the wastes to migrate beyond Site 1 and Site 2; and (3) to assess the current and potential risk to public health, welfare, and the environment. Shortly after the government filed the remedial-investigation report, the parties jointly informed the court that they neither needed nor desired further discovery in this action. According to the remedial-investigation report, any contaminants still at Site 1 affect groundwater there mostly by migrating through the alluvial aquifer. This aquifer lies atop a miocene aquifer. The remedial-investigation report indicates that there is little or no migration between the aquifers. The potential does exist for minimal migration along the casing of an enormous well on Site 1; e.g., if the well were to develop a leak in the casing and/or a rupture in the annular seal. If so, the potential concentration of contaminants not only would be localized around the well, but also negligible because the amount of water removed by this well is so great. Wells established around the site to contain any outward flow of contaminants monitor the alluvial aquifer. Nothing shows that the alluvial aquifer carries water across any state lines or into any river or other transportation source that would take it across any state lines, nor is there any indication of transport of contaminants through the air across state lines. Indeed, the record reflects that any contamination at Site 1 is of such minimal proportions as not to constitute any hazard to the public. As stated in the EPA Record of Decision: The risk assessment indicates that only risks associated with potential future on-site resident exposures to contaminated groundwater and surface soils would not be within EPA’s acceptable risk range. The chemicals of concern would pose unacceptable risks if the on-site groundwater were used as a source of potable water or if children living on the site were exposed to contaminated surface soils. Future use of this site as a residential area is considered unlikely and thus the proposed remedial goals are directed at protecting the groundwater for its maximum beneficial use. C. Review of the Consent Decree Notwithstanding the initial willingness of the defendant to enter into a “consent” decree, this court has a duty to examine a consent decree not only to determine whether its factual and legal determinations are reasonable, but also to ensure that the decree does not violate the Constitution, a federal statute, or the controlling jurisprudence. E.g., Howard v. McLucas, 871 F.2d 1000, 1008 (11th Cir.) (citing United States v. City of Miami, 664 F.2d 435, 441 (5th Cir.1981) (Rubin, J., concurring); Cotton v. Hinton, 559 F.2d 1326, 1330-31 (5th Cir.1977)), cert. denied, 493 U.S. 1002, 110 S.Ct. 560, 107 L.Ed.2d 555 (1989). The court, therefore, has made two requests for briefs, the first of which directed the parties to brief whether CERCLA, as applied in this case, is consistent with the Supreme Court’s view of the Commerce Clause as recently explained in United States v. Lopez, 514 U.S. —, 115 S.Ct. 1624, 131 L.Ed.2d 626 (1995). While both parties have responded to this issue, the defendant has also raised the issue of CERCLA’s retroactivity. The defendant contends that Congress did not intend for CERCLA to be retroactive and that if it did, CERCLA violates the Due Process Clause and unconstitutionally delegates legislative power to the EPA. As a result of further direction from the court by order of December 29, 1995, the Justice Department has responded to the defendant’s claims concerning retroactivity. A frequently reaffirmed principle dictates that a court avoid deciding a constitutional issue if a case can be disposed of on non-constitutional grounds. See, e.g., Ashwander v. TVA 297 U.S. 288, 347, 56 S.Ct. 466, 483, 80 L.Ed. 688, 711 (1936) (Brandeis, J., concurring). Therefore, before considering the several constitutional issues, this court first addresses the statutory question of whether Congress intended to make CERCLA retroactive. II. CERCLA AND RETROACTIVITY Although the retroactivity of CERCLA has been the subject of holdings in other circuits, the issue has not been squarely addressed by the Eleventh Circuit. In dicta, a panel of the Eleventh Circuit recently referred to CERCLA as being retroactive. Virginia Properties Inc. v. Home Ins. Co., 74 F.3d 1131, 1132 (11th Cir., 1996). The issue of retroactivity, however, was not before that court. In another CERCLA case before this court recently, the parties failed to raise the issue of CERCLA’s retroactivity, see Redwing Carriers, Inc. v. Saraland Apartments, Ltd., 875 F.Supp. 1545 (1995); in its decision, however, this court refused to apply certain regulations retroactively. 875 F.Supp. at 1565 (“Since there is no such clear intent in FI-FRA, see, e.g., 7 U.S.C. Secs. 136a(a), 136a(d)(2), 1361, the court declines to apply registration cancellation retroactively.”). Of those federal decisions which have directly addressed the issue of CERCLA’s retroactivity, none have declined to apply CERCLA on retroactivity grounds. Nevertheless, all of those eases were decided before the Supreme Court’s decision in Landgraf v. USI Film Prods., 511 U.S. 244, 114 S.Ct. 1483 (1994), on which the defendant rests its argument against retroactivity. A. The Justice Department has responded somewhat cavalierly to the issue of CERCLA’s retroactivity, contending the matter is “well settled” and unaffected by Landgraf: Every court to face CERCLA retroactivity challenges has rejected the arguments advanced here. Indeed, courts have uniformly held that (1) Congress clearly and unequivocally intended retroactive application of CERCLA and; (2) such a liability scheme is rationally related to a legitimate governmental interest. The Supreme Court’s decision in Landgraf v. USI Film Prods., 511 U.S. 244, 114 S.Ct. 1483, 128 L.Ed.2d 229 (1994) announces no new constitutional rules, and in no way impacts this case law. See Pi’s Mem. on the Retroactivity of CERCLA and Due Process Issues, p. 2. To begin with, Justice errs in claiming that the lower federal courts have “uniformly held that (1) Congress clearly and unequivocally intended retroactive application of CERCLA.” In fact, not all the courts which have applied CERCLA to pre-enactment conduct have agreed that it is retroactive, and those which have actually analyzed the retroactivity issue have differed on whether Congress was acting “clearly and unequivocally.” More importantly, Justice’s observation that Landgraf “announces no new constitutional rules” fails to assist this court with what is a significant non-constitutional decision. Written against the backdrop of several constitutional issues, the Landgraf holding addresses a rule of statutory construction. Justice cannot credibly contend that “[t]he result in Landgraf is unremarkable.” It is understandable that Justice would attempt to dismiss Landgraf so casually. Landgraf certainly demolishes the interpretive premises on which prior eases had concluded CERCLA is retroactive. Landgraf does so in the course of attempting to clarify confusion regarding the interpretive rules applicable to retroactivity: Our precedents on retroactivity left doubts about what default rule would apply in the absence of congressional guidance, and suggested that some provisions might apply to cases arising before enactment while others might not. Compare Bowen v. Georgetown Univ. Hospital, 488 U.S. 204, 109 S.Ct. 468, 102 L.Ed.2d 493 (1988) with Bradley v. Richmond School Bd., 416 U.S. 696, 94 S.Ct. 2006, 40 L.Ed.2d 476 (1974). Landgraf, 511 U.S. at —, 114 S.Ct. at 1495. In reaffirming the traditional presumption against retroactive legislation, Landgraf disapproves language in Bradley which had appeared to reverse that traditional presumption. Bradley allowed an award of attorneys’ fees for work done prior to the effective date of the statute providing for attorneys’ fees. The Supreme Court stated it was doing so “on the principle that a court is to apply the law in effect at the time it renders its decision, unless doing so would result in manifest injustice or there is a statutory direction or legislative history to the contrary.” 416 U.S., at 711, 94 S.Ct., at 2016. The applicability of this language was severely limited by Landgraf: Although the language suggests a categorical presumption in favor of application of all new rules of law, we now make it clear that Bradley did not alter the well-settled presumption against application of the class of new statutes that would have genuinely “retroactive” effect. Landgraf, 511 U.S. at-, 114 S.Ct. at 1503 (emphasis in the original). An early and influential case finding CERCLA retroactive, Ohio v. Georgeoff, 562 F.Supp. 1300 (N.D.Ohio, 1983), did exactly what Landgraf disapproves. Georgeoff began quite appropriately by “initially deter-min[ing] the standard to be applied in determining whether a statute should be applied retroactively.” 562 F.Supp. at 1306. While it acknowledged an “historieal[ ] ... presumption favoring a prospective only application of a statute”, the court otherwise approved of and seemed to apply a presumption in favor of retroactivity. The premises for the decision in Georgeoff were disapproved in and are no longer tenable after Landgraf. As a result, Georgeoff and the cases which rely on its analysis, — and which do not do their own analysis — cannot be considered persuasive. Besides Georgeoff, only two other cases actually do analyze the issue of retroactivity; the rest of the cases basically rely on one or more of these three cases and other cases which cite these cases. Those other two cases which do their own analyses are United States v. Shell Oil Co., 605 F.Supp. 1064 (D.Colo., 1985) and United, States v. Northeastern Pharmaceutical & Chemical Co., Inc., 810 F.2d 726 (8th Cir., 1986), both of which also cite Georgeoff with approval. See 605 F.Supp. at 1072 and 810 F.2d at 733. Unlike Georgeoff, neither case explains how it is applying the presumption against retro-activity; but like Georgeoff, both cases demonstrate little regard for the presumption. Following an analysis unlike that of Landgraf and after quoting Georgeoff, Id. at 1072, Shell Oil concludes that CERCLA is “unavoidably retroactive.” Id. at 1073. It does so on the basis of CERCLA’s “general purpose and scheme.” Under Landgraf, however, the fact “that retroactive application of a new statute would vindicate its purpose more fully ... is not sufficient to rebut the presumption against retroactivity.” 511 U.S. at-, 114 S.Ct. at 1507-8. Other than its discussion of “general purpose and scheme,” Shell Oil does not explain precisely what overrides the presumption against retroactivity. As for Northeastern, on which Justice relies, the case treats the presumption itself rather lightly, devotes only one sentence to the statutory language, relies on Shell Oil and Georgeoff among other eases, and offers one paragraph about the “statutory scheme.” It is understandable that, prior to Landgraf, lower federal courts would have tended to minimize the importance of the presumption against retroactivity given some Supreme Court precedents. Landgraf reflects the different views on the Court about whether some of its precedents, in particular Bradley v. Richmond School Bd., 416 U.S. 696, 94 S.Ct. 2006, 40 L.Ed.2d 476 (1969) and a ease on which it relies, Thorpe v. Housing Authority of Durham, 393 U.S. 268, 89 S.Ct. 518, 21 L.Ed.2d 474 (1969), have been consistent with other Supreme Court precedents. The dissenting opinion in Landgraf remarks on the majority’s “eagerness to resolve the ‘apparent tension’ ... between Bradley v. Richmond School Bd. ... and Bowen v. Georgetown University Hospital”, 511 U.S. at —, 114 S.Ct. at 1508 (Blackmun, J., dissenting) (citations omitted). The concurring opinion simply says that Bradley and Thorpe “invented an utterly new and erroneous rule.” Id. at -, 114 S.Ct. at 1496 (Scalia, J., concurring). On the one hand, the majority “begin[s] by noting that there is no tension between the holdings in Bradley and Bowen.” Id. at-, 114 S.Ct. at 1496 (emphasis in the original). Later, however, the majority isolates and limits the troublesome language in Bradley, quoted above, which created the confusion about the strength of the presumption against retroactivity. Whether Bradley actually is or is not at odds with pre and post-precedent need not concern this court. It suffices to recognize that, contrary to the assertion by Justice, Landgraf does at least clarify the analysis of retroactivity and, therefore, does “impact this ease.” B. The majority opinion in Landgraf sets forth an analysis which, as here summarized, requires a court 1) to determine a) whether Congress has expressly stated the statutes reach and b) if not, whether the text and legislative history have “clearly prescribed” Congress’ intent to apply the provision retroactively; ■ 2) if not, to determine whether the provision actually has “retroactive effect on the party or parties in the litigation;” and 3) if so, to apply the traditional presumption against retroactivity — absent a clear congressional intent to the contrary. The majority opinion’s approach breaks down as follows: 1) In the last sentence of Part II of the opinion, the majority says: “Our first question, then, is whether the statutory text ... manifests an intent that the 1991 Act should be applied to cases that arose and went to trial before its enactment.” 511 U.S. at-, 114 S.Ct. at 1492. What follows in Part III of the opinion, 511 U.S. at-, 114 S.Ct. at 1493-96, is an analysis of the arguments based on the text and legislative history. At the end of this section, the majority determines: “[i]n the absence of the kind of unambiguous directive found in See. 15 of the 1990 bill [which explicitly provided for retroactivity], we must look elsewhere for guidance on whether Sec. 102 applies to this case.” 511 U.S. at-, 114 S.Ct. at 1496. 2) Part IV of the opinion, — U.S. at ---, 114 S.Ct. at 1496-1505, discusses the presumption against retroactivity. The majority notes the “ ‘apparent tension’ ” between the different canons of construction found in Bradley as opposed to other Supreme Court precedents. Id. at -, 114 S.Ct. at 1496. The Court concludes this section with the following framework for analysis: When a case implicates a federal statute enacted after the events in suit, the court’s first task is to determine whether Congress has expressly prescribed the statutes’ reach. If Congress has done so, of course, there is no need to resort to judicial default rules. When, however, the statute contains no such express command, the court must determine whether the new statute would have retroactive effect, i.e., whether it would impair rights a party possessed when he acted, increase a party’s liability for past conduct, or impose new duties with respect to transactions already completed. If the statute would operate retroactively, our traditional presumption teaches that it does not govern absent clear congressional intent favoring such a result. Id. at -, 114 S.Ct. at 1505. (emphasis added). 3) In Part V of the opinion, 511 U.S. at ---, 114 S.Ct. at 1505-1508, the majority “ask[s] whether, given the absence of guiding instructions from Congress, Sec. 102 of the Civil Rights Act of 1991 is the type of provision that should govern cases arising before its enactment.” 511 U.S. at-, 114 S.Ct. at 1505. They find the new damages remedy to have a retroactive effect and conclude it “is the kind of provision that does not apply to events antedating its enactment in the absence of clear congressional intent,” 511 U.S. at -, 114 S.Ct. at 1506. The section proceeds without reconsidering its earlier discussion of congressional intent— other than to reject as insufficient to overcome the presumption against retroactivity the argument that “retroactive application of a new statute would vindicate its purpose more fully” 511 U.S. at-, 114 S.Ct. at 1507 (footnote omitted) — to conclude that they “have found no clear evidence of congressional intent that Sec. 102 of the Civil Rights Act of 1991 should apply to eases arising before its enactment.” 511 U.S. at -, 114 S.Ct. at 1508. In the framework of analysis as quoted above in 2), the first step involves only a determination of whether Congress has made an express statement concerning retroactivity; other evidence of “clear congressional intent” is left to step three. In the opinion itself, however, the Court considers both express and other evidence of congressional intent in the first step. As reflected from the quotes from Parts II and III of the opinion, see 1), the Court does not clearly differentiate between express language and other clear language of intent from the statute or legislative history. While it does not seem to matter whether the discussion of intent is confined to step one or divided between step one and step three, the analysis here follows what Landgraf in fact 'does. 1. Has Congress Expressed Its Intent On CERCLA Retroactivity? CERCLA contains no language explicitly stating it is retroactive. Under language in some Supreme Court opinions, that failure would be fatal for retroactivity. United States v. Heth, 3 Cranch 399, 414, 2 L.Ed. 479 (1806). (Cushing, J.) (“[I]t [is] unreasonable in my opinion, to give the law a construction which would have such a retrospective effect, unless it contained express words to that purpose.”) While Landgraf demonstrates a preference for express language regarding retroactivity and seems to warn Congress about the danger of not being explicit, its discussion of other (i.e., non-express) statutory language and legislative history establishes that these should be considered in determining congressional intent. a) Non-Express Statutory Language. In resolving retroactivity issues, Landgraf instructs that the answer may vary among provisions within an act. 511 U.S. at -and-, 114 S.Ct. at 1494 and 1505. Accordingly, the analysis focuses on the particular provisions of CERCLA, at issue in the complaint, Sections 106(a) and 107(a). Under Section 106(a) the government seeks injunctive relief in the first claim of the complaint. Although injunctive relief is ordinarily prospective, when it requires a party to spend funds related to actions taken prior to CERCLA’s enactment, such relief is nevertheless retroactive. See Georgeoff, 562 F.Supp. at 1303-05. Thus to the extent that the government’s claims under 106(a) and 107(a) relate to actions taken prior to the effective date of CERCLA, they involve the issue of retroactivity. The Justice Department devotes little attention to statutory language as evidence of congressional intent. To the extent it does, Justice’s brief relies on Northeastern’s observation that “[t]he language used in the key liability provision, CERCLA § 107 ... refers to actions and conditions in the past tense.” Northeastern, 810 F.2d at 733. Georgeojf, after a more extensive analysis, finds the use of the past tense not dispositive. Even while finding other terms supportive of retro-activity, Georgeojf states that the statutory provisions are insufficient to overcome the presumption of retroactivity. Despite these statutory arguments, the Court is unable to declare that the statute evidences the “imperative character” required to overcome the presumption against retroactivity. Regardless, these provisions provide some evidence that Congress intended CERCLA to apply retroactively. The Court, therefore, will consider these statutory terms as indicia, but not dispositive indicia, of a Congressional intent to allow retroactive application of CERCLA. 562 F.Supp. at 1311. Shell Oil agrees with Georgeojf,! Georgeojf's and Shell Oil’s conclusion that the statutory language in CERCLA is not sufficient to establish retroactivity is persuasive. Although Northeastern arguably reaches a contrary conclusion, it does so without conducting nearly as extensive an analysis as done in Georgeojf. This court concludes, therefore, that the language of section 107 provides “no clear evidence of Congressional intent,” as required by Landgraf, that CERCLA’s liability provisions be given retroactive effect. Section 106 contains no language indicating congressional intent to authorize relief that is retroactive. If anything, the language suggests the very argument, later made by Justice, that CERCLA liability is not “retroactive.” The pertinent language of § 106 borrows from law on the abatement of nuisance; it refers to “an imminent and substantial endangerment to the public health or welfare or the environment” and provides “such relief as may be necessary to abate such danger or threat.” Georgeojf records Justice’s unsuccessful attempt to argue that, “although it reaches pre-enactment conduct, legislation designed to alleviate a continuing public nuisance does not act retroactively.” 562 F.Supp. at 1304. United States v. So. Carolina Recycling and Disposal Inc., 653 F.Supp. 984 (D.S.C., 1984), on the other hand, adopts that argument As discussed below, Landgraf now precludes such attempts to avoid the issue of retroactivity. The fact that in the early cases Justice thought it necessary to argue, at least in the alternative, that CERCLA is not retroactive certainly indicates Justice’s initial concern that courts might not find a clear congressional intent that CERCLA be applied “retroactively.” b) Legislative History. CERCLA itself has almost no legislative history. Although Congress had worked on “Superfund” cleanup of toxic and hazardous waste bills, and on parallel oil spill bills for over three years, the actual bill which ■became Public Law No. 96-510 had virtually no legislative history at all, because the bill which became law was hurriedly put together by a bipartisan leadership group of Senators — with some assistance of their House counterparts — introduced and passed by the Senate in lieu of all pending measures on the subject. It was then placed before the House, in the form of a Senate amendment of the earlier House bill. It was considered on December 3, 1980, in the closing days of the lame duck session of an outgoing Congress. It was considered and passed, after very limited debate, under a suspension of the rules, in a situation which allowed for no amendments. Faced with a complicated bill on a take-it-or-leave-it basis, the House took it, groaning all the way. Frank P. Grad, Treatise on Environmental Law Sec. 4A.02[2][a], at 4A-51 (1994). (footnote omitted) (Hereinafter, “Grad”) (emphasis added). The unusual lack of legislative history, as reflected by the failure even to produce a House — Senate Conference Report, is attributable to “[t]he delicate nature of the compromise” which led to passage, according to the Preface to the Congressional Research Service Legislative History. Much of what passes for legislative history of CERCLA, therefore, comes from “bills introduced which contributed to some extent to the final act.” In Landgraf, the Supreme Court does consider a prior bill as part of its review of the legislative history. The Court places some weight on the fact that a bill vetoed in the previous year had explicitly provided for retroactivity. The legislative history considered in Landgraf comes not from committee reports, but from the language of the prior bill itself. The fact that the later enacted legislation had no such provision prompts the Court to infer: “it seems likely that one of the compromises that made it possible to enact the 1991 version was an agreement not to include the kind of explicit retroactivity command found in the 1990 bill,” 511 U.S. at -, 114 S.Ct. at 1495 (emphasis in the original). As acknowledged in Georgeoff, “the precise issue of retroactivity ... was not addressed in Congressional debates,” 562 F.Supp. at 1311, including debate on prior bills. Id. Under Landgraf, that fact would seem to be nearly fatal to any attempt to overcome the presumption against retroactivity. The Justice Department attempts to distinguish CERCLA from the civil rights statute considered in Landgraf because no bill prior to CERCLA explicitly provided for retroactivity. See “Pi’s Mem. on Retroactivity” at 4-5. That fact, however, does not strengthen the case for retroactivity. It only means that what Justice and other courts have labeled the legislative history of CERCLA may not be as clear as was the legislative history of the Civil Rights Act considered in Landgraf. It still remains for Justice to demonstrate that the legislative history does contain clear intent of CERCLA’s retroactivity. In attempting to do so, Justice contends that the “history, as analyzed by the courts, demonstrates unequivocally that Congress was concerned about past, pre-enactment acts of disposal.” “Pi’s Mem.” at 8. The argument of Justice, relying as it does on past cases, fails to overcome the presumption against retroactivity because those prior cases do not follow the analysis of Landgraf and because they find clarity in legislative history which does not exist. Many of the past cases are unclear about two things which are distinguished in Landgraf: congressional intent and retroactive effect. As discussed below, Landgraf struggles with the term “retroactive.” The majority excludes certain statutes from the presumption against retroactivity, specifically procedural and jurisdictional statutes. See 511 U.S. at ---, 114 S.Ct. at 1501-02. Approving of Justice Story’s discussion in Society for Propagation of the Gospel v. Wheeler, 22 F.Cas. 756 (No. 12,156) (CCDNH 1814), the majority says: A statute does not operate “retroactively” merely because it is applied in a case arising from conduct antedating the statute’s enactment... 511 U.S. at-, 114 S.Ct. at 1499. In other words, the fact that legislation might have retroactive effect does not necessarily mean that Congress clearly intended that it be so applied. In searching the legislative history, some courts have placed weight on sources that have little no value. As two commentators write: “[t]he available reports pertaining to prior versions of the legislation are of little value in interpreting the statute — CERCLA, as enacted, differed significantly from previous proposals.” Allan J. Topol and Rebecca Snow, Superfund Law and Procedure § 1.1. at 5 (St. Paul, West; 1992). The most that can be said from the legislative history is that Congress left many questions, including retroactivity, as open ones to be decided later. Like Landgraf, the circumstances surrounding passage of CERCLA strongly suggest that the failure expressly to prescribe the reach of the statute was deliberate on Congress’ part. It would have been a simple matter for Congress to have included a provision within the Act providing that liability would be imposed retroactively. Given the undoubted Congressional awareness of an existing problem, this omission takes on special importance. There can be no question that Congress was aware that the issue of retroactivity could arise. Yet, Congress failed to make this statement. Georgeoff, 562 F.Supp. at 1309. The political circumstances suggest why there was little actual legislative history and why Congress left many issues deliberately ambiguous. In 1980, CERCLA emerged from the lame duck session of Congress after President Carter had been defeated and former Governor Reagan had yet to assume the presidency. Its language was derived mainly from S. 1480, 99th Cong., 2d Sess. (1980), a bill that had been reported from the Senate Committee on Environment and Public Works but that had little hope for passage in its complete form. A group of approximately 25 to 30 senators, with assistance from House members and the outgoing Administration (at one point including President Carter himself), and crafted new legislative language to paper over policy disputes on which there was no consensus. Because many such disputes remained after these negotiators finished their deliberations, floor debate in the Senate often contain rival explanations of the same provisions of CERCLA. Alfred R. Light, CERCLA Law and Procedure Compendium (BNA; Washington, D.C., 1992) at 1-1. The circumstances explain why Congress decided not to decide many issues, leaving them for others — the EPA and the courts— to decide. The fact that CERCLA includes a legislative veto, see 42 U.S.C. § 9655, since determined to be unconstitutional, see INS v. Chadha, 462 U.S. 919, 103 S.Ct. 2764, 77 L.Ed.2d 317 (1983), suggests that members of Congress may not have been as concerned as they should have been about a number of issues, including retroactivity, because they assumed incorrectly that Congress has the power to “fix it” later without concern about a presidential veto. Whatever the reasons, the legislative history lacks the clear eongressional intent to make CERCLA liability retroactive. Given that the language — express or otherwise — and the legislative history— broadly and narrowly understood — fail to demonstrate a clear congressional intent for retroactivity, Landgraf requires that the presumption against retroactivity be applied if the statute is one to which that presumption applies. 2. Does CERCLA Have Retroactive Effect? Landgraf does not simply ask: is the statute retroactive? Framing the question that way, as have a number of the lower federal courts, reduces the analysis to a matter of labeling rather than distinguishing between the effect of the statute and the intent of Congress as does Landgraf. A few of the early cases manipulate the label “retroactive” apparently as a way of avoiding the issues of congressional intent and constitutionality. Thus United States v. South Carolina Recycling Disposal Co., 653 F.Supp. 984 (D.S.C., 1984), aff'd in part and vac. in part sub nom. United States v. Monsanto Co., 858 F.2d 160 (4th Cir., 1988), cert. denied, 490 U.S. 1106, 109 S.Ct. 3156, 104 L.Ed.2d 1019 (1989), concludes that considering pre-enactment conduct does not make CERCLA liability “retroactive” because the statute actually imposes liability for the post enactment releases or threatened releases. Although the Justice Department now contends that CERCLA is clearly retroactive, early cases reflect that Justice made the argument that CERCLA is not retroactive. See e.g., Georgeoff, 562 F.Supp. at 1304. Under the Landgraf analysis, CERCLA liability certainly has “retroactive effect” because as applied in this and many other cases it easily falls within the explanatory language of that term. That is to say, the Justice Department’s attempt in this ease to impose liability under § 107(a) largely on actions occurring prior to the statute’s effective date “would impair rights a party possessed when he acted, increase a party’s liability for past conduct, or impose new duties with respect to transactions already completed.” 511 U.S. at-, 114 S.Ct. at 1505. What Landgraf said about compensatory damages can be said about the financial liabilities under CERCLA for pre-enactment conduct: “[t]he new damages remedy in See. 102, we conclude, is the kind of provision that does not apply to events antedating its enactment in the absence of clear congressional intent.” 511 U.S. at-, 114 S.Ct. at 1506. 3. Should the Presumption Against Retroactivity be Applied? In the third step of the analysis, Landgraf asks “whether, given the absence of guiding instruction from Congress, the [particular section of the act] is the type of provision that should govern eases arising before its enactment.” The Court does not ask generally whether the act is retroactive, but focuses on the particular section. The opinion then distinguishes between a procedural provision of that section (jury trial right) which would “presumably apply to eases ... regardless of when the underlying conduct occurred,” and its punitive and compensatory damages provisions. Recognizing that “[r]etrospective imposition of punitive damages would raise a serious constitutional question”, Id. at -, 114 S.Ct. at 1505, the Court avoids that constitutional question by interpreting the punitive damages provision not to be retroactive. The provision “authorizing recovery of compensatory damages is not easily classified,” id., according to the Court, because the conduct itself was already unlawful — only the remedy was new — and also because compensatory damages are not punitive. Despite these differences, Landgraf applies the presumption against retroactivity even to the compensatory damages provision. The CERCLA provision, § 107(a), as applied in this case falls somewhere between the punitive damage and the compensatory damage provisions considered in Landgraf. In this case, punitive damages are not sought. Nevertheless, § 106 does provide for fines for failure to comply with an executive branch abatement order; such fines are clearly punitive. Section 107(c)(3) also authorizes punitive, treble damages. The EPA uses the threat of punitive damages as a negotiating tool. Given the very real threat of punitive damages, CERCLA retroactivity poses very nearly the same “ex post facto ” danger referred to in Landgraf. 511 U.S. at ---, 114 S.Ct. at 1505-06. According to Landgraf a provision for punitive damages should not be construed as retroactive unless the language forces that conclusion because the court must then confront substantial constitutional questions which follow. Regardless of the threat of punitive damages, retroactive CERCLA liability is more egregious than the compensatory relief which the Court refused to apply retroactively in Landgraf. In the instant and many other cases, liability under CERCLA would require compensation for actions which when taken violated no federal or state law. At least as to the compensatory damages remedy sought in Landgraf, it could be said that retroactivity involved no new legal standard concerning the conduct itself. Nevertheless, even on the compensatory damages issue Landgraf says, “it is the kind of provision that does not apply in the absence of clear congressional intent.” Certainly, under Landgraf principles, CERCLA liability is the kind that does not apply retroactivity without clear congressional intent. As discussed above, pre-Landgraf cases addressing CERCLA retroactivity are unreliable. Among these cases, Justice places particular reliance on Northeastern which characterizes CERCLA as “overwhelmingly remedial and retroactive” and as having a “backward focus.” Such labels only confuse the issue as explained by Landgraf. Legislation cannot be remedial if the conduct being “remedied” was lawful at the time of its occurrence. As discussed above, determining that a statute has “retroactive effect” is only part of the analysis; by itself, it does not indicate Congress’ intent. Northeast-em’s statement about CERCLA having a “backward focus,” is no more dispositive than a similar characterization of the compensatory damage provision discussed in Landgraf. Unlike certain other forms of relief, compensatory damages are quintessentially backward-looking. Compensatory damages may be intended less to sanction wrongdoers than to make victims whole, but they do so by a mechanism that affects the liabilities of defendants. They do not “compensate” by distributing funds from the public coffers, but by requiring particular employers to pay for harms they caused. The introduction of a right to compensatory damages is also the type of legal change that would have an impact on private parties’ planning. 511 U.S. at-, 114 S.Ct. at 1506. (emphasis added in first sentence only). Landgraf insists that in no case “in which Congress had not clearly spoken, have we read a statute substantially increasing the monetary liability of a private party to apply to conduct occurring before the statute’s enactment.” 511 U.S. at-, 114 S.Ct. at 1507. Liability under CERCLA certainly must be in accord with this principle. The fact that “retroactive application of [this] statute would vindicate its purpose more fully”, as in Landgraf, “is not sufficient to rebut the presumption against retroactivity.” Id. at---, 114 S.Ct. at 1507-08. Only one sentence in Landgraf, if taken in isolation, lends any support at all to Justice’s position. Near the end of the opinion, Landgraf says: “Section 102 is plainly not the sort of provision that must be understood to operate retroactively because a contrary reading would render it ineffective.” Id. at-, 114 S.Ct. at 1508 (emphasis in original). Justiee’s argument boils down to a claim that CERCLA must be read to be retroactive. Without any doubt, in CERCLA Congress is addressing an existing situation as well as future problems. In this respect, CERCLA certainly has a “backward focus.” It does not follow, however, that the liability provision “must be understood to operate retroactively because a contrary reading would render it ineffective.” Early on, the Justice Department itself argued, at least in the alternative, that CERCLA’s liability provision need not be interpreted to be retroactive. See Georgeoff at 1304-1306. Quickly, however, that argument failed, but ultimately became unnecessary as courts approved of applying CERCLA retroactively. Citing certain “main policy objectives underlying CERCLA ... and the need for a liberal interpretation of the Statute, the courts have expanded the scope of CERCLA liability well beyond anything that Congress itself could have envisioned.” Allan J. Topol and Rebecca Snow, 1 Superfimd Law and Procedure, § 1.1 at 6 (West Pub.; St. Paul, Minn. 1992). Justice confuses what it, the EPA, and a number of courts consider desirable with what it can be said Congress clearly intended. Insofar as pre-enactment releases are concerned, the purpose of CERCLA can be covered through the Superfimd. The EPA, however, has chosen to recover as much as possible from private parties, no doubt in part due to Congress’ failure to provide sufficient resources to pay for cleaning all the sites, even as the need was thought to be in 1980. See Georgeoff, 562 F.Supp. at 1312-13. While Georgeoff takes the lack of funding as an indication of congressional intent to make CERCLA retroactive, Id. at 1313, lack of funding does not render the operation of the statute itself ineffective in the sense used in Landgraf. Landgraf mentions accepted means such as certain taxes and zoning regulations by which governments “impose burdens on past conduct.” 511 U.S. at-n. 24, 114 S.Ct. at 1499, n. 24. Such means, if not otherwise unconstitutional, allow legislators to “remedy” past conduct. Legislators can address environmental damage; however, under the “Takings Clause,” “confiscatory regulations ... cannot be newly legislated or decreed (without compensation), but must inhere in the title itself, in the restriction that background principles of the State’s law of property and nuisance already place upon land ownership.” Lucas v. So. Carolina Coastal Council, 505 U.S. 1003, 1029, 112 S.Ct. 2886, 2900-01, 120 L.Ed.2d 798 (1992). As Landgraf observes, the “Takings Clause” is one of several constitutional provisions which incorporate the anti-retroactivity provisions. 511 U.S. at-, 114 S.Ct. at 1497. Of course, agencies and officials at all levels of government generally would wish to induce or compel particular parties to pay for government initiatives rather than having to impose a general tax. However, the general presumption against retroactivity and particular constitutional provisions are designed to limit, and in some instances to prevent, governments from doing so. Ironically, the EPA has resisted the retro-activity principle when it would have worked in favor of private parties. Under Section 106(b)(2)(A), added in 1986 by Superfund Amendments and Reauthorization Act (“SARA”), private parties can seek certain reimbursement from the Superfund. The EPA has “adopted the position that orders issued before the passage of SARA do not provide the petitioner with a right to reimbursement against the Fund.” 2 Superfund Law and Procedure supra § 11.1 at 272 (footnote omitted). Courts have generally agreed with the EPA position. See cases cited in id., n. 7-10. Nothing presented in the Justice Department brief or pre-Landgraf cases concerning the statutory language of CERCLA or its legislative history demonstrates that Section 107(a) (and/or Section 106(a) as related to it in this case) is “the sort of provision that must be understood to operate retroactively because a contrary reading would render it ineffective.” 511 U.S. at -, 114 S.Ct. at 1508. Accordingly, this court holds that Section 107(a), and Section 106(a) as linked to it in this case are not retroactive. III. CONSTITUTIONAL ISSUES The holding that Sections 106(a) and 107(a) cannot be given retroactive effect resolves much, but not all, of this case. Moreover, because of the court’s statutory interpretation holding, the court need not rule on the due process constitutional issue, which applies only to pre-enactment conduct. The complaint, however, covers conduct which, while largely occurring prior to the effective date of CERCLA in 1980, includes some post-enactment conduct. As the defendant mentions in its brief, the provisions of the consent decree respond “mostly with respect to waste disposal activities that preceded the enactment of CERCLA in 1980” (emphasis added). After briefly addressing one constitutional issue which has been mooted by the court’s holding on pre-enactment conduct, the court considers more fully those constitutional issues which are not limited to preenactment conduct. A. Retroactivity and Due Process Landgraf recognizes the due process and other constitutional concerns lurking in the background of the interpretive issue of whether Congress intended retroactive application. See 511 U.S. at---, 114 S.Ct. at 1497-98. Such constitutional considerations seem to have contributed to the Court’s confirmation of the presumption against retroactivity. Yet, with the possible exception of a punitive damages provision, the Court seems to accept that Congress has the constitutional power to give a statute retroactive effect if it clearly intends to do so. Landgraf, 511 U.S. at-, 114 S.Ct. at 1499 (“Since the early days of this Court, we have declined to give retroactive effect to statutes burdening private rights unless Congress has made its intent clear.” (emphasis added)). Given the court’s statutory interpretation of CERCLA, there is no need to reach the constitutional validity of retroactive application of CERCLA under the due process clause. B. Delegation of Legislative Power The defendant contends that CERCLA unconstitutionally “delegates to EPA the basic policy decisions” with “no ‘intelligible standard’ to guide EPA’s decisions about which contaminated sites pose significant risks and how much money should be spent to reduce or eliminate such risks.” Def.’s Mem. of Law in Resp. to Ct.’s Ord. of Sept. 19, 1995 at 20-21. This constitutional challenge attacks CERCLA as generally administered by the EPA. This broader challenge based on the non-delegation doctrine has not been mooted by the holding that CERCLA liability cannot be applied retroactively. The EPA’s decision to apply CERCLA retroactively against private parties represents one example of what the defendant contends is the unauthorized exercise of leg-' islative power. One way of understanding the decision in Landgraf is to view it as applying a specific rule of non-delegation, namely that Congress cannot leave such a basic legislative decision as a statute’s possible retroactivity to the executive and judicial branches. Both the presumption against retroactivity and the non-delegation doctrine derive from the more fundamental doctrine of separation of powers. Thus, Landgraf cites INS v. Chadha, 462 U.S. 919, 103 S.Ct. 2764, 77 L.Ed.2d 317 (1983), in support of the proposition that “even the will of the majority does not become law unless it follows the path charted in Article I, § 7, cl. 2 of the Constitution.” 511 U.S. at-, 114 S.Ct. at 1496. Landgraf could also have cited Chadha as to what is “essentially legislative”, namely Congressional “action that [has] the purpose and effect of altering the legal rights, duties, and relations of persons.” 462 U.S. at 952,103 S.Ct. at 2784-85. Chadha, which is the first modern ease to declare a congressional statute unconstitutional on clear separation-of-powers grounds, illustrates the relationship between the general doctrine and particular constitutional provisions. Chadha declares the one-house legislative veto void because it violates the bicameralism and the presentment clauses. Following Chadha, a line of separation-of-powers cases raise various issues, including delegation of powers and removal from office. See Metro. Washington Airports Author. v. Citizens for the Abatement of Aircraft Noise, Inc., 501 U.S. 252, 111 S.Ct. 2298, 115 L.Ed.2d 236 (1991) (invalidating a provision in a federal statute approving an agreement for the transfer of two Washington-area airports from the U.S. Department of Transportation to an “Airports Authority,” which created a “Board of Review” consisting of nine members of Congress because it violated separation of powers either by allowing members of Congress to exercise executive power or by allowing members of Congress to enact legislation without complying with the bicameralism and presentment requirements of the Constitution.); Mistretta v. United States, 488 U.S. 361, 109 S.Ct. 647, 102 L.Ed.2d 714 (1989) (finding no separation-of-powers violation in a statute establishing the United States Sentencing Commission as “an independent commission in the judicial branch of the United States” with removal by the President of commission members.); Morrison v. Olson, 487 U.S. 654, 108 S.Ct. 2597, 101 L.Ed.2d 569 (1988) (finding no separation-of-powers violation in a statute providing for the appointment of “Independent Counsel” by a panel of Article III judges and removal only for cause by the Attorney General.); Bowsher v. Synar, 478 U.S. 714, 106 S.Ct. 3181, 92 L.Ed.2d 583 (1986) (holding that the provision in the Balanced Budget and Emergency Deficit Control Act giving executive power to the Comptroller General, who was removable by Congress, violates separation of powers). In Chadha the dissent worried that the Court’s decision threatened the “modern administrative state.” See 462 U.S. at 984-89, 103 S.Ct. at 2801-04. As the post-CTmdha cases summarized above reflect, the Court has not consistently pursued a rigorous separation-of-powers approach. Consequently, the modern administrative state has not been much disturbed by the line of cases cited above. A rigorous separation-of-powers approach, which would include a stricter non-delegation doctrine, would greatly curtail the rule-making authority of administrative agencies, including the EPA. By itself, however, a broad non-delegation constitutional challenge to CERCLA cannot currently succeed. Under present Supreme Court jurisprudence, “there is no significant ‘anti-delegation’ principle which restricts the exercise of the commerce power.” John E. Nowak and Ronald D. Rotunda, Constitutional Law (St. Paul, West; 1995) at 156. A year ago, of course, it could have been said that almost nothing restricts Congress’ exercise of the commerce power. In the interim, the jurisprudence has become unsettled due to the Supreme Court’s decision in United States v. Lopez, 514 U.S. —, 115 S.Ct. 1624, 131 L.Ed.2d 626 (1995), which, for the first time in more than fifty years, struck down a congressional statute which regulates non-state activity on grounds that it exceeded Congress’ power under the Commerce Clause. Some of the concerns raised by the defendant under the non-delegation doctrine overlap with the issues in Lopez, to which this court now turns. C. The Commerce Clause The Supreme Court’s interpretation of the Commerce Clause has expanded greatly over the better part of the last two centuries. In one pre-Lopez case, a federal district court described the expansion of that clause as follows: To be within the ambit of the Commerce Clause, an activity originally had to be commerce that moved across state lines. With time, courts began to hold that the commerce need only affect movement across state lines. Finally, courts dropped the commerce requirement and applied the Clause to any activity that affects movement across state lines. The sole remaining limitations on Congress’s power under the Commerce Clause is the requirement that the effect on commerce be substantial. Michigan Protection & Advocacy Service v. Babin, 799 F.Supp. 695, 735 (E.D.Mich.1992), aff'd, 18 F.3d 337 (6th Cir.1994). Since the 1930’s the Supreme Court has found very little in the Constitution that limits Congress’ power to justify legislation under the Commerce Clause. Prior to United States v. Lopez, 514 U.S. —, 115 S.Ct. 1624, 131 L.Ed.2d 626 (1995), the only unsettled area appeared to involve the ability of Congress to regulate states themselves. In National League of Cities v. Usery, 426 U.S. 833, 96 S.Ct. 2465, 49 L.Ed.2d 245 (1976) the Court held that the Commerce Clause did not give Congress power to regulate “the state as a state.” A few years later, the Court overturned National League of Cities by its decision in Garcia v. San Antonio, 469 U.S. 528, 105 S.Ct. 1005, 83 L.Ed.2d 1016 (1985), holding that the sovereignty of the states does not limit Congress’ power under the Commerce Clause. Then, somewhat inconsistently, the Court decided in New York v. United States, 505 U.S. 144, 112 S.Ct. 2408, 120 L.Ed.2d 120 (1992), that the Constitution does not give Congress power to “[C]ommandeer the legislative procedure of the states by directly compelling them to enact and enforce a federal regulatory program.” Id. at 175-76, 112 S.Ct. at 2428 (quotation omitted). While the holding in New York did not rest on the Tenth Amendment, the Court noted that “the Tenth Amendment confirms that the power of the Federal Government is subject to limits that may, in a given instance, reserve power to the states.” Id. at 156-58, 112 S.Ct. at 2418. The application of the federalism principles found in New York has remained somewhat uncertain as reflected in recent federal circuit cases ruling on the Brady Bill. At the same time, however, the Court has recently reaffirmed the traditional principles of state sovereignty in a case from this circuit, Seminole Tribe of Fla. v. Florida, — U.S. —, 116 S.Ct. 1114, 134 L.Ed.2d 252 (1996), which rejected the principle that Congress’ Commerce Clause power could be used to overrule a state’s sovereign immunity. United States v. Lopez did not involve federal regulation of a state itself. Rather, the federal gun statute at issue attempted to exercise police power over matters historically falling within the jurisdiction of local government. In declaring Congress had exceeded its powers under the Commerce Clause in regulating non-state action, the Court seemed to open up many questions. The dissenting Justices warned that “the Court’s holding threatens legal uncertainty in an area of law that, until [Lopez ], seemed reasonably well settled.” Id. at 514 U.S.-, 115 S.Ct. at 1664 (Breyer,