Full opinion text
MEMORANDUM OF OPINION AND ORDER LEVI, District Judge. This case concerns the allocation of costs for the cleanup of the Iron Mountain Mine (“the mine”). Plaintiffs are the United States and the State of California, who are suing under CERCLA to recover response costs for cleanup activities at the mine. Defendants are Iron Mountain Mines, Inc. and T.W. Arman (“IMMI/Arman”), the current owners of the mine, and Rhone Poulenc Basic Chemicals Co. (“RP”), the corporate successor to Mountain Copper, Co., Ltd., the previous owner of the mine. RP has filed CERCLA and state law counterclaims and third-party claims against the United States and the State of California. The United States and the State of California now move to dismiss these counterclaims and third party claims under Rule 12(b)(1) for lack of subject matter jurisdiction based on various theories of immunity, and under Rule 12(b)(6) for failure to state a claim. I. Background Iron Mountain is located nine miles northwest of Redding, California. From the late 1800’s until 1962, the mountain was mined for gold, copper, zinc and pyrite. The extensive mining exposed sulfide deposits, which react with rainwater and groundwater to form acid mine drainage (“AMD”), a pollutant harmful to fish. The AMD flows into two creeks, and they in turn flow into Spring Creek. Spring Creek terminates in the Spring Creek reservoir behind the Spring Creek Dam, a project constructed in 1963 by the United States Bureau of Reclamation (“USBR”). The Spring Creek Power Plant was constructed at about the same time by the USBR and is located on Spring Creek just downstream of the dam. The Spring Creek Dam releases AMD-tainted water into the Keswick Reservoir at a point on the Sacramento River between the Shasta Dam (upstream) and the Keswick Dam (downstream). A. Claims against the United States 1. Oumership and Operation of Dams and Power Plant RP contends that the United States is liable for response costs associated with the release of AMD from Iron Mountain Mine because of its construction and operation of the Shasta, Keswick and Spring Creek Dams and the Spring Creek Power Plant. (RP Third Am.Countercl. — U.S. ¶ 9.) RP’s claim rests on the theory that had the United States not dammed the Sacramento River and Spring Creek, the natural flow of the watershed would have diluted the AMD and there would have been no response costs. (Id. ¶ 10.) In its counterclaim, RP paints a picture of failure upon failure by the USBR to control the pollution problems stemming from the construction of the Shasta Dam in the 1940s. According to RP, when the Kes-wick Dam was built in 1950 for the purpose of regulating the flow from the Shasta Dam, it had the effect of catching and impounding hazardous sediments from Spring Creek. (Id. ¶ 12.) The Spring Creek Dam was built in 1963 both to prevent debris from clogging the Spring Creek Power Plant and to remedy the AMD problem created by the mine and the Shasta and Keswick dams. (Id. ¶ 13.) According to RP, the Spring Creek Dam is itself composed partly of sediments containing hazardous substances, and the USBR disposed of hazardous sediments elsewhere in the facility during its construction. (Id. ¶ 14.) RP also alleges that the Spring Creek Dam is too small to sufficiently dilute the AMD runoff without coordinating releases from the Spring Creek Dam with releases from the Shasta Dam, and, according to RP, this coordination has been lacking. In the language of CERCLA, RP contends that the United States is liable as an owner or operator of the dams, behind which AMD concentrates, or as an arranger since the Spring Creek Dam was built in part to dispose of the AMD. (Id. ¶¶ 21-24.) 2. Activities During and After World War II As an alternative basis for liability, RP claims that the United States is liable for response costs because' of the government’s involvement in the mine during and after World War II. According to RP, when the war began, there had been no copper mining at Iron Mountain since 1930, and mining was limited to surface mining for gold and intermittent mining for pyrite. (Id. ¶¶ 27-28.) By directive, the government prohibited gold mining as not essential to the war effort. (Id. ¶ 28.) The government also established the Premium Price Plan as an incentive for the production of copper and zinc; prices were set to encourage maximum production. (Id. ¶32.) Having been told to cease gold mining, Mountain Copper (then owner of the mine) contracted to sell its entire copper and zinc output to the government. The government controlled marketing and pricing of the ore produced. (Id. ¶ 29, 32-33.) Under the contract, a new ore body was opened at the mine, a new mill was built, and pyrite ore removal increased. (Id. ¶¶ 29-31, 34.) If Mountain Copper had not increased pyrite production, the government could have seized the mine to assure production. (Id. ¶34.) RP alleges additional government involvement with the mine during the war affecting the labor force and shipment of the metals. After the war, the mining and milling of copper and zinc ore ceased, but RP contends that the government continued its involvement with the mine. According to RP the government paid for equipment and operations in conducting explorations for new ore and participated in decisions as to the best methods to search for new ore. (Id. ¶39.) 3. RP’s Claims against the United States Based on these allegations, in Counts I and II of its counterclaim, RP claims response costs under CERCLA § 107(a) and contribution under CERCLA § 113(f) and common law. RP contends that the United States is liable as owner and operator of the dams and power plant, which are part of the facility or facilities in which response costs have been incurred, and that the United States is further liable as an arranger and transporter. RP also claims that the United States’ activities during and after World War II render the United States liable as an operator and arranger. (RP Third Am.Counterel. — U.S. ¶¶ 1-2, 41^7.) In Count III, RP avers that if it is liable to the United States, the award should be reduced in recoupment under CERCLA, common law, and equitable indemnification. In Counts I and II of its Amended Third-Party Complaint Against the United States, RP requests contribution and indemnification under CERCLA and common law with respect to any liability RP might incur to the State of California. B. Claims against the State of California RP contends that the State of California is also liable for the response costs associated with the mine. RP claims the State “has actively participated in the operation of the Shasta and Keswick Dams.” (RP Am.Coun-tercl. — Cal. ¶ 6.) In particular, RP alleges that officials from the State’s Central Valley Regional Water Pollution Control Board and Department of Fish and Game met with USBR officials to agree on operating plans for releases from the dams. RP also asserts that the State operates the Spring Creek Dam along with the USBR, and thereby shares responsibility for the release of polluted water from Spring Creek into the Sacramento River. RP’s claim as to the Spring Creek Dam is based upon alleged agreements and contacts between the State and USBR concerning operation of the dam. (Id. ¶¶ 9, 11, 14, and 15.) RP further asserts that the State has not fulfilled its public trust duty of supervising appropriated water. According to RP, the State has: (1) failed to require the USBR to coordinate the operation of the dañas to minimize environmental harm, (id. ¶ 16); (2) failed to repair the louvers on the Spring Creek Dam that would have minimized release of AMD during storms, (id. ¶ 17); and (3) wasted dilution water by diverting it from Spring Creek with the result that in 1992 the Spring Creek Dam overflowed, releasing toxic waters, (id. ¶ 18). Finally, RP alleges that the State owns the streambed of Spring Creek below the Spring Creek Dam and the streambed of the Sacramento River in the Shasta and Keswick areas. Because hazardous substances are located in the streambeds, RP asserts that the State is an owner of a “facility” from which a release of pollution has occurred. (Id. ¶¶ 22-25.) Based on these allegations, RP asserts numerous counterclaims against the State. Counts I and II are CERCLA claims based on the State as an owner and operator, respectively, of facilities — the dams and the streambeds — from which hazardous substances have been released. Count III claims that the State is an arranger under CERCLA; Count IV makes a CERCLA contribution claim. RP also asserts state law claims for negligence (Count V), nuisance (Count VI), creation of a dangerous condition on public property (Count VII), breach of public trust (Count VIII), breach of mandatory duty (Count IX), and equitable indemnification (Count XI). Count X is a recoupment counterclaim which incorporates all other claims. C. Procedural History The United States’ complaint was “filed on June 12, 1991; the State of California’s complaint was filed on August 29, 1991. The two cases were consolidated by Order of November 21, 1991. In United States v. Iron Mountain Mines, Inc., 812 F.Supp. 1528 (E.D.Cal.1992), Judge Schwartz, who was first assigned to this case, held that mining waste was excluded from the coverage of CERCLA under the Bevill Amendment. The CERCLA claims against RP and IMMl/Arman were dismissed on this basis, (see Order of May 11, 1993, and Order of August 23, 1993). However, the plaintiffs’ complaints were reinstated after the Ninth Circuit held that releases of wastes listed in the Bevill Amendment may create liability under CERCLA. See Louisiana-Pacific Corp. v. ASARCO, Inc., 24 F.3d 1565 (9th Cir.1994). Subsequently, RP filed its Amended Third-Party Complaint and Third Amended Counterclaim against the United States, and its Amended Counterclaim against the State of California. The United States and the State now move to dismiss RP’s claims. Relying on sovereign immunity doctrines, the United States moves to dismiss those of RP’s claims that are based on the USBR’s ownership and operation of the three dams and the power plant. The United States advances two main arguments in support of its motion: (1) all claims are barred by section 3 of the Flood Control Act of 1928 (33 U.S.C. § 702e); and (2) the CERCLA claims regarding the Spring Creek Dam are barred by sovereign immunity because the dam serves a remedial purpose and CERCLA exempts the government from liability when it acts in a governmental or remedial capacity. (United States Mem. at 1-2.) The State similarly moves to dismiss RP’s counterclaims concerning its involvement in the operation of the dams. Resting on the Eleventh Amendment and the same case law relied upon by the United States, the State contends that CERCLA does not abrogate its immunity under the Eleventh Amendment when it acts in a remedial or regulatory capacity. The State also moves to dismiss these claims on the basis that they do not state a claim for operator or arranger liability under CERCLA. As to RP’s claims under CERCLA relating to the United States’ activity at the mine during and after the war, the United States seeks dismissal arguing that the complaint fails to state a claim because the allegations, if true, fall short of claiming that the United States had managerial authority over the mine or the disposal of waste. Finally, the United States and the State move for dismissal of RP’s recoupment claims.. II. Immunity Under § 702c Section 3 of the Flood Control Act, 33 U.S.C. § 702c, provides in relevant part that “[n]o liability of any kind shall attach to or rest upon the United States for any damage from or by floods or flood waters at any place.” The United States argues that all of RP’s claims based on the construction and operation of the three dams and the power plant are barred by this provision. There are two major points of dispute between the parties concerning application of § 702c to the circumstances here: (1) whether the waters at the Shasta, Keswick, and Spring Creek Dams, and at the Spring Creek Power Plant, are “flood waters” within the terms of § 702c; and (2) if so, whether CERCLÁ supersedes the § 702e immunity. A. Scope of § 702c The government claims the immunity in § 702e on the basis that all of the waters at issue are either part of flood control projects or are gathered and released in coordination with flood control projects. RP argues that the immunity does not apply because the damage of which it complains was not caused by waters released for reasons of flood control. Specifically, RP argues that the Spring Creek Dam was built to control AMD and to assist the Power Plant and not for flood control. RP also contends that it was the decision to build the Shasta and Keswick dams and the concentration and then release of hazardous sediments that has been the cause of damage, not the release of flood waters. RP further maintains that the Spring Creek Dam was built with AMD tainted soil and that it was this action by the USBR, as opposed to the release of flood waters, that has been the source of damage. Finally, RP contends that uneven releases of water from the Spring Creek Power Plant stir up contaminated sediments, causing their release into the Keswick reservoir. RP’s arguments fail in the face of the broad and emphatic language in § 702c and the broad interpretation consistently given to that language in the Supreme Court and the Ninth Circuit. In United States v. James, 478 U.S. 597, 106 S.Ct. 3116, 92 L.Ed.2d 483 (1986), the Supreme Court examined the scope of the government’s immunity under § 702e. The case involved a claim brought under the Federal Tort Claims Act for injuries sustained by plaintiffs when they were pulled through the underwater gates of a dam while waterskiing at a reservoir. Id. at 598-600, 106 S.Ct. at 3118. The Court found that § 702e was not limited to property damage but applied to any damage including a claim of personal injury caused by negligence. As to the definition of “flood waters” under the section, the Court held that the term was unambiguous: The Act concerns flood control projects designed to carry floodwaters. It is thus clear from § 702(c)’s plain language that the terms “flood” and “flood waters” apply to all waters contained in or carried through a federal flood control project for purposes of or related to flood control, as well as to waters that such projects cannot control. Id. at 605, 106 S.Ct. at 3121. The Court also found that its reading of the statutory language was consistent with the legislative history and purpose: “the sweeping language of § 702c was no drafting inadvertence. Congress clearly sought to ensure beyond doubt that sovereign immunity would protect the government from ‘any’ liability associated with flood control.” Id. at 608, 106 S.Ct. at 3122 (citation omitted). Thus, the Court held that the negligent failure to warn plaintiffs of the danger of boating too close to the dam was part of the management of a flood control project and was within the immunity. Id. In keeping with the sweeping language of § 702c, the Ninth Circuit has also interpreted the immunity broadly. Under Ninth Circuit case law the immunity applies if the damage sustained is “not wholly unrelated” to the use of a flood control project: the immunity provision of § 702c does not apply when the damage or injury is “wholly unrelated to any Act of Congress authorizing expenditures of federal funds for flood control”.... So long as a project is authorized by Congress for flood control purposes, and the damage or injury is related to the use of that project, immunity continues to attach. McCarthy v. United States, 850 F.2d 558, 562 (9th Cir.1988) (quoting Morici Corp. v. United States, 681 F.2d 645, 647 (9th Cir.1982), and Peterson v. United States, 367 F.2d 271, 275 (9th Cir.1966)) (internal citations omitted). Under the “wholly unrelated” test, the fact that a flood control project may have multiple purposes does not affect the government’s immunity. This much is plain from James where the injured parties were using the reservoir for recreation. See McCarthy, 850 F.2d at 562. What may be less certain from James is whether the immunity may only be claimed if the injury stems from a particular release of flood waters for the purpose of flood control. The injuries in James were caused by the Corps of Engineers’ decision to release waters from reservoirs that had reached the flood stage. Thus, the Court had no occasion to consider whether the immunity depended on this factor although it did reject plaintiffs’ parallel argument that the government’s failure to warn could be viewed as a separate activity “wholly unrelated” to flood control. This issue, however, is resolved in the Ninth Circuit; in Morid and then McCarthy the court expressly rejected “the notion that the applicability of the immunity provision in cases involving multi-purpose projects depends upon the particular use to which the project was being put when the negligence occurred.” McCarthy, 850 F.2d at 562. See Fryman v. United States, 901 F.2d 79, 82 (7th Cir.1990) (noting that the Ninth Circuit’s interpretation of § 702c may be broader than that stated by the Supreme Court in James). Under the “wholly unrelated” test if one of the purposes of the project is flood control then immunity applies when “waters contained in a federal flood control project for purposes related to flood control were a substantial factor in bringing about” the injury. McCarthy, 850 F.2d at 561-62. In the context of this case, the “wholly unrelated” test may be stated as follows: if the waters behind the Shasta, Keswick, and Spring Creek dams are not wholly unrelated to a flood control project, immunity will apply if these waters are a substantial factor in bringing about the injury caused by AMD even if the purpose of some of the water releases was for a reason other than flood control. In short, the court rejects RP’s contention that the immunity hinges upon a showing that the damage was caused by a particular release of water for the very purpose of flood control. Once the test is stated in this fashion, RP’s claims readily fall within the immunity provided by § 702c. The Shasta and Keswick Dams are part of the Central Valley Project, which Congress authorized for flood control purposes. See Morici Corp. v. United States, 491 F.Supp. 466, 489 (E.D.Cal.1980) (“The Shasta and Keswick facilities are, indisputably, multi-purpose facilities which are devoted in part to flood control.”), aff'd, 681 F.2d 645, 648 (9th Cir.1982) (Central Valley Project “was authorized by Congress for flood control purposes”). The Spring Creek Dam and Power Plant are part of a different project known as the Trinity River Diversion. Although the Trinity River Diversion was authorized by Congress principally as an irrigation project, it was also to be operated as an integral part of the Central Valley Project. See Morici, 491 F.Supp. at 490-91 (discussing legislative history of the Trinity River Diversion and the Central Valley Project). As the court noted in Morici, “[t]he Central Valley Project is operated as an integrated whole, rather than as a number of separate, isolated parts, because water releases at any one facility must be coordinated with releases at other facilities in that river basin.” Morici, 491 F.Supp. at 490. Because of this interrelationship, which RP does not dispute, the operation of the Spring Creek Dam and Power Plant must be considered part of the Central Valley Project, or at least as integrally related to that project. Indeed, RP’s theory of liability suggests this very interrelation: RP contends that the AMD problem stems ultimately from construction of the Shasta Dam which, because of reduced river flows and increased concentration of pollutants, in turn required construction of the Spring Creek Dam. From this vantage the Spring Creek Dam and reservoir is related to a flood control project just as much as if it had been built as part of the Central Valley Project. All of the injury claimed by RP is subject to the immunity because all of the injury is attributable, or at least not “wholly unrelated,” to waters collected and released as part of a federal flood control project. Thus, the injury which resulted from the decision to build the Shasta and Keswick Dams, with the consequent limitation on river flows and the build up of tainted sediment behind the Kes-wick Dam, is covered by the immunity since the sediments are carried by flood waters and the build up and concentration of the sediments are directly related to the use of a flood control project. Similarly, the alleged use of tainted soils to build the Spring Creek Dam is within the scope of the immunity since the damage is caused by the release of waters that are managed for flood control, as part of the Central Valley Project, and because the dam itself was constructed because of the Central Valley Project. Finally, the injury allegedly caused by the Power Plant— surges of water which in combination with releases from Spring Creek stir up contaminated sediments that travel into Keswick reservoir — is not “wholly unrelated” to a flood control project because it is the combination of water from the Power Plant and from the Dam that causes the problem. Moreover, the Power Plant, like the power plant at the Shasta Dam, is part of the Central Valley project, see Morici, 491 F.Supp. at 490-91, and thus releases of water from the Power Plant are related to a flood control project. Thus, as an initial matter, the United States is entitled to sovereign immunity under § 702c with respect to all of RP’s counterclaims arising from ownership and operation of the dams and power plant. The United States may be liable only if CERCLA supersedes this immunity. B. CERCLA Abrogates § 702c Immunity Section 120(a)(1) of CERCLA provides: Each department, agency, and instrumentality of the United States (including the executive, legislative, and judicial branches of government) shall be subject to, and comply with, this chapter in the same manner and to the same extent, both procedurally and substantively, as any nongovernmental entity, including liability under Section 9607 of this title. Section 9607, cross-referenced in § 9620(a)(1), is the liability section in CERC-LA. For purposes of this discussion, the critical language appears in the very first sentence of § 9607, which establishes liability in certain categories of persons, including governmental entities, “Notwithstanding any other provision or rule of law, and subject only to the defenses set forth in subsection (b) of this section.” Subsection (b) provides limited defenses for releases caused solely by an act of God, war, or of a third party in certain narrow circumstances. There is no immunity expressed for the United States and there is no cross reference to 33 U.S.C. § 702c or to any other immunity provision relating to the United States. In combination, this language appears to expressly waive any immunity that the United States may have under any other statute or common law rule. In reviewing § 9620(a)(1), the Supreme Court noted that it is “doubtless an ‘unequivocal!] express[ion]’ of the Federal Government’s waiver of its own sovereign immunity, since we cannot imagine any other plausible explanation for this unqualified language.” Pennsylvania v. Union Gas Co., 491 U.S. 1, 10, 109 S.Ct. 2273, 2279, 105 L.Ed.2d 1 (1989) (citations omitted). But it is particularly the combination of the language in § 9620(a)(1) with that in § 9607 which appears to waive any immunity that the government might otherwise have. This is because § 9620(a)(1) subjects the United States to liability under § 9607, just like a private party, and § 9607 has a broadly worded waiver of any limitation on liability and a specific limitation of defenses to those expressly stated in the statute. In light of this language, it would appear that the statute does not impliedly, but rather, expressly waives any immunity that the United States could claim, including immunity under § 702c. In response to this statutory language the government relies mainly on caselaw addressing the Federal Tort Claims Act (FTCA), 28 U.S.C. § 2761 et seq., and § 702e immunity. As the government notes, every circuit court to examine the issue has concluded that the FTCA did not impliedly repeal § 702c in the context of tort claims against the United States. See National Mfg. Co. v. United States, 210 F.2d 263, 270-75 (8th Cir.1954); Clark v. United States, 218 F.2d 446, 452 (9th Cir.1954); Stover v. United States, 332 F.2d 204 (9th Cir.1964); Aetna Ins. Co. v. United States, 628 F.2d 1201, 1205 (9th Cir.1980); Taylor v. United States, 590 F.2d 263, 266 n. 4 (8th Cir.1979); Callaway v. United States, 568 F.2d 684, 686 (10th Cir.1978) (collecting cases from other circuits). The government argues that if the FTCA did not repeal § 702e immunity in the context of tort actions neither did CERCLA repeal the immunity in the context of a claim for environmental damage under CERCLA. Although not without force, the government’s reliance on the FTCA case law is not compelling in the context of CERCLA for two reasons. First, there is no language in the FTCA comparable to that contained in § 9607. The FTCA does treat the United States as liable in tort claims “in the same manner and to the same extent as a private individual under like circumstances.” 28 U.S.C. § 2674. This language is certainly similar to that contained in CERCLA at § 120(a)(1), and were there no further waiver the argument based on the FTCA caselaw might prevail. But § 9620(a)(1) specifically subjects the government to liability under § 9607, and that section has a sweeping renunciation of “any other provision or rule of law” that might impair liability and any other defenses other than those specifically listed. This language works an express repeal of all immunity, including § 702c, and there is no such language in the FTCA. Second, the FTCA has a revocation section, and § 702c was not included on it. See 60 Stat. 812, 842, 846-847. The absence of § 702c on the list of repealed statutes suggests that Congress did not intend that the FTCA would override § 702c. The Supreme Court relied on this reasoning in United States v. James, 478 U.S. 597, 598-600, 106 S.Ct. 3116, 3118, 92 L.Ed.2d 483 (1986) in noting that § 702e immunity survived passage of the FTCA. But there is no such comparable provision in CERCLA and thus no basis on which to argue that Congress intended to leave § 702c immunity untouched by CERCLA. For these reasons, the court finds that although the immunity in § 702e applies to the operation and construction of the dams and power plant at issue in this case, this immunity has been waived by the Congress in §§ 107 and 120(a)(1) of CERCLA. The government’s motion to dismiss on the basis that § 702c bars the claims relating to its operation of the dams and power plant is therefore denied. III. Immunity for Regulatory or Remedial Activities The United States and the State of California argue that CERCLA recognizes an immunity from suit for government entities when they act in a regulatory or remedial role. On this theory, they claim entitlement to immunity from RP’s CERCLA counterclaims relating to operation of the dams because they have acted in a regulatory or remedial capacity in their operation of these facilities. The State frames its argument in terms of the Eleventh Amendment while the United States relies on general principles of sovereign immunity, but the arguments are basically the same and rest on the same case law and statutory construction. Both the State and the federal government rely heavily on a policy argument that it would be unfair or unduly burdensome to expose a governmental entity to any liability for acts taken to clean up pollutants not generated by the government and not located on government property even if the government’s remedial response is grossly reckless or negligent. Whatever may be said on behalf of this position as a matter of policy, CERCLA does not incorporate such an immunity by its plain language and the court declines to create such an exemption where Congress has not done so and in the face of clear statutory language subjecting government entities to liability. A. Statutory Treatment of Government Actors The starting point in considering the assertion of an immunity for government remedial actions is the language and structure of the statute. CERCLA has a liability section at 42 U.S.C. § 9607 that places liability on “any person” who owns or operates a “facility” from which a release of hazardous substances occurs. The term “person” is perhaps the most basic building block in the CERCLA edifice. The definition provided in CERCLA is all inclusive and covers government entities: The term .‘person’ means an individual, firm, corporation, association, partnership, consortium, joint venture, commercial entity, United States Government, State, municipality, commission, political subdivision of a State, or any interstate body. 42 U.S.C. § 9601(21). The liability imposed on “any person” extends to all costs of removal or remedial action incurred by the United States, a state, or by any other person so long as these costs are not inconsistent with the national contingency plan. This liability is subject to certain limited and enumerated defenses set forth in subsection (b) — acts of God and of war and acts of third parties in certain narrow circumstances. As noted in the preceding discussion of § 702c immunity, § 9607 begins with the statement that liability applies under the section “notwithstanding any other provision or rule of law, and subject only to the defenses in subsection (b),” and there are no immunity defenses provided in subsection (b). In addition to the liability section, there is also a provision in CERCLA for contribution among persons who share exposure to liability. Section 9613 provides that a person who has incurred response costs may seek contribution from “any other person” who is liable. Government bodies are therefore subject to contribution actions, as well as actions under § 9607, because government entities are within the definition of “person.” In resolving contribution claims, CERCLA reposes broad discretion in the court to “allocate response costs among liable parties using such equitable factors as the court determines are appropriate.” 42 U.S.C. § 9613. In several sections CERCLA specifically addresses the liability of government entities beyond its inclusion of such entities within the definition of “person.” In what the Supreme Court describes as an unequivocal waiver of sovereign immunity, § 9620(a)(1) provides that the United States and all of its subdivisions “shall be subject to, and comply with, this chapter in the same manner and to the same extent, both proeedurally and substantively, as any nongovernmental entity, including liability under section 9607 of this title.” 42 U.S.C. § 9607(a)(1) (emphasis added). Similarly, the states’ Eleventh Amendment immunity is abrogated by the inclusion of the states within the definition of “person” and by § 9601(20)(D), which provides that the states shall have liability as “owners” and “operators” “in the same manner and to the same extent, both proeedurally and substantively, as any nongovernmental entity, including liability under section 9607 of this title.” 42 U.S.C. § 9601(20)(D) (emphasis added). See Pennsylvania v. Union Gas Co., 491 U.S. 1, 109 S.Ct. 2273, 106 L.Ed.2d 1 (1989). Moreover, CERCLA expressly addresses the liability of those who act in a remedial capacity, and who are otherwise not liable as owners or operators, and provides them with protection from strict liability: [N]o person shall be liable ... for costs or damages as a result of action taken or omitted in the course of rendering care, assistance, or advice in accordance with the National Contingency Plan (“NCP”) or at the direction of an onscene coordinator appointed under such plan, with respect to an incident creating a danger to public health or welfare or the environment as a result of any releases of a hazardous substance or the threat thereof. This paragraph shall not preclude liability for costs or damages as the result of negligence on the part of such person. 42 U.S.C. § 9607(d)(1). This provision applies to governmental bodies because they are within the definition of “person.” Indeed, the following paragraph makes it plain that governmental bodies are included within § 9607(d)(1) by providing a special standard of liability for state and local governments acting “in response to an emergency created by the release or threatened release of a hazardous substance generated by or from a facility owned by another person.” In such instances, state and local governments are liable only for costs or damages resulting from their “gross negligence or intentional misconduct.” 42 U.S.C. § 9607(d)(2). There is no comparable provision for the federal government, leaving it liable for its own negligence under § 9607(d)(1). Certain other provisions within CERCLA limit or refine the liability of a governmental entity when the entity acts in a remedial capacity. When a federal or state agency purchases or condemns real property in the course of a remedial action, no liability is incurred by the agency “solely as a result of acquiring an interest in real estate under this subsection,” 42 U.S.C. § 9604(j)(3), and states and local governments are excluded from liability as “owners” and “operators” of a facility when ownership or control was acquired “involuntarily through bankruptcy, tax delinquency, abandonment, or other circumstances in which the government involuntarily acquires title by virtue of its function as sovereign.” 42 U.S.C. § 9601(20)(D). In judicial review “of any issues concerning the adequacy of any response action taken or ordered” by the federal government, the government’s choice of response actions can only be challenged under the “arbitrary and capricious” standard. 42 U.S.C. § 9613(j)(l) & (2). Looked at as a whole, the statutory scheme suggests rather powerfully that there is no overarching, unexpressed remedial immunity for state or federal agencies. Government bodies are persons subject to liability under § 9607 “notwithstanding any other provision or rule of law.” See Union Gas Co., 491 U.S. at 7-9, 109 S.Ct. at 2278 (§ 9607 evinces congressional understanding that “States would be liable in any circumstance described in [§ 9607(a) ] from which they were not expressly excluded”). Section 9607 recognizes only the defenses explicitly set out in subsection (b), and the limited defenses there provided do not include any defense or immunity for a governmental entity acting in a remedial capacity.- The statute reiterates at § 9620(a) and § 9601(20)(D) that the federal and state governments are liable just as private parties. The act provides a limitation on liability for all of those who act in a remedial capacity in certain circumstances, and expressly provides further protection to state agencies acting in an emergency. Where Congress saw fit to provide greater protection to governmental entities for remedial actions, CERCLA expressly provides for such limitation on liability. See, e.g., 42 U.S.C. § 9604(j)(3). Of course, were there an implicit, all inclusive immunity for state and federal government remedial activities, these sections would be unnecessary, and indeed would be in conflict -with the immunity, for far from limiting liability, which is their apparent purpose, they would extend liability beyond the immunity that the United States and the State contend is incorporated into CERCLA. See Union Gas Co., 491 U.S. at 9-11, 109 S.Ct. at 2279 (discussing § 9607(d)(2) and noting that Congress would not need to provide an exemption from liability unless there would otherwise be liability). All of this suggests that there is no unexpressed, residual immunity for the states or the federal government when they act in a regulatory or remedial capacity. Finally, this exposure to liability for negligent, reckless, or willful acts need not lead to intolerable consequences or exposure of the public fisc. Through the contribution section, CERCLA calls upon the court to equitably adjust the relative responsibility of different persons all of whom bear some liability for a release. Presumably where the government responsibility consists only of remedial action the share of liability will be adjusted accordingly. Moreover, the government body would only be liable were its activities such as to make it an “owner” or “operator” of the facility or otherwise bring the government within the terms of § 9607. In short, the clear language of CERCLA, its structure and its evident purpose to treat state and federal agencies no differently from private parties are inconsistent with a judicially created remedial immunity for government agencies. B. Remedial Immunity in the Case Law Notwithstanding the clear language of the statute and its omission of a remedial defense, several federal district courts have recognized the immunity. See United States v. Western Processing Co., 761 F.Supp. 725, 728 (W.D.Wash.1991) (claims based on EPA regulatory and cleanup activities barred by sovereign immunity); United States v. Azrael, 765 F.Supp. 1239, 1244 (D.Md.1991) (counterclaims based on EPA removal action barred by sovereign immunity); Stilloe v. Almy Bros., Inc., 782 F.Supp. 731, 736 (N.D.N.Y.1992) (adopting reasoning of Azra-el and Western Processing)', In re Paoli Railroad Yard PCB Litigation, 790 F.Supp. 94, 97 (E.D.Pa.1992) (no waiver of immunity for claim that United States was liable as operator based on control of site during cleanup activities), aff'd without opinion, 980 F.2d 724 (3rd Cir.1992); United States v. Berks Assocs., 1992 WL 68346 (E.D.Pa.1992); United States v. Hardage, 733 F.Supp. 1424 (W.D.Okla.1989), aff'd in part, rev’d in part on other grounds, 982 F.2d 1436 (10th Cir.1992) (claims arising out of EPA’s cleanup activities not supported by statutory waiver of sovereign immunity, though claims might be proper in recoupment); United States v. Skipper, 781 F.Supp. 1106, 1111-12 (E.D.N.C.1991); see also United States v. American Color and Chem. Corp., 858 F.Supp. 445, 449-50 (M.D.Pa.1994) (counterclaims based on EPA removal action barred where government acts in regulatory, not business, capacity) (citing other cases listed here). The court, with respect, must decline to follow these cases because they are fundamentally at odds with the language of CERCLA. The above cited cases rely on several reasons for finding the immunity. To begin with, it is suggested that government entities should be treated differently from private parties because private parties do not clean up pollution caused by others while governmental bodies will step forward in such circumstances. But this is an unconvincing rationale because CERCLA does encourage private cleanup of pollution caused by others by exposing current owners to liability and at the same time providing them with a cause of action for the recovery of their costs. Nurad, Inc. v. William E. Hooper & Sons Co., 966 F.2d 837, 841 (4th Cir.1992); 3550 Stevens Creek Assoc. v. Barclays Bank, 915 F.2d 1355, 1357 (9th Cir.1990). In many instances private remediators will be current owners of property who had nothing to do with the creation of the hazardous waste problem that they must respond to. These private parties are not different in any significant respect from a government agency acting in a remedial capacity and CERCLA does not acknowledge such a difference but treats them in the same way. It is also suggested in these cases that imposing negligence liability under CERCLA on governmental entities responsible for releasing pollution would shift the cost of the cleanup to the government and thus would thwart “Congress’ intent to ensure that those who benefit financially from a commercial activity should internalize the health and environmental costs of that activity into the costs of doing business.” Azrael, 765 F.Supp. at 1245; see also Western Processing, 761 F.Supp. at 729; Atlas, 797 F.Supp. at 421. One may doubt that this statement accurately expresses the principal purpose of CERCLA. If CERCLA can be said to have a central purpose, that purpose is to achieve cleanup. In doing so- CERCLA may impose liability on parties who had little culpability in creating the contamination problem at issue. This is frequently the case when a later owner becomes responsible for pollution released many years earlier, sometimes under different technology and under different industrial standards. There is no question that CERCLA places liability on new owners, but it does not work to internalize the costs of pollution into current operations in these situations. Indeed, there may be no current business operations, yet the current landowner will be Hable. More importantly, even if it is critical to make polluters internalize the costs of their harmful activities, it does not foHow that they should also internalize the costs of the government’s negHgent response during its cleanup efforts. At any rate, Congress has resolved these quite complex poHey questions in CERCLA and the courts ought not upset the resolution on some other view. Another argument for holding governmental entities immune from HabiHty for their negHgent cleanup efforts is the supposition that Congress intended aH challenges to remedial activities to be “addressed as part of a determination for consistency with the National Contingency Plan.” Western Processing, 761 F.Supp. at 729. Under this theory, if the government’s cleanup is faulty, the potentially responsible person can challenge the government’s attempt to recover costs under § 9607(a)(4)(A), but cannot assert an independent contribution counterclaim under § 9613(f)(1). Id.; Atlas, 797 F.Supp. at 419-20; Skipper, 781 F.Supp. at 1112; Azrael, 765 F.Supp. at 1246; Paoli, 790 F.Supp. at 97. In Skipper, the Coast Guard had allegedly violated the NCP when it buried an oily sand mixture during a remedial action. Years later, the EPA returned to clean up the site and sued private landowners to recover its costs. The court dismissed the defendants’ counterclaim based on the Coast Guard’s negHgent cleanup, finding that the sole remedy was to raise inconsistency with the NCP as an affirmative defense to the government’s claim under § 9607(a)(4)(A), not to file a counterclaim under § 9613(f). Id. at 1112-13. The court noted that it would be “inconceivable” were the private party forced to bear the expense of the government’s dereHction. However, under Skippeds holding, this “inconceivable” result is entirely possible, even likely, were the landowners themselves to have cleaned up the site as CERCLA encourages. If they had, their effort to seek contribution from all responsible parties, as provided by CERC-LA, would be blocked by the remedial exception to HabiHty in the case of the Coast Guard, and the “affirmative defense” of inconsistency with the NCP would provide no remedy at all. CERCLA resolves the matter by providing that government bodies are subject to HabiHty and claims for contribution. The cases recognizing remedial immunity for government agencies also suggest that to permit a claim for governmental negHgence at a cleanup site would be tantamount to creating a “new negHgence defense” to CERCLA HabiHty when the government seeks to recover cleanup costs. See Azrael, 765 F.Supp. at 1245 & n. 12; Western Processing, 761 F.Supp. at 729. But CERCLA expressly provides that the government’s choice of remedies can be challenged when it seeks to recover cleanup costs. See 42 U.S.C. § 9613(j). Further, a claim against another responsible party is not fairly viewed as a defense, certainly not when CERCLA provides for contribution and does not deem one responsible party’s effort to gain contribution in any sense as a defense. Rather, the immunity itself is a new defense to government liability, one not provided for by the statute. Another theme running through the cases recognizing remedial immunity is that since the statute does not clearly authorize a cause of action against the government acting in a remedial or regulatory capacity, and since waivers of sovereign immunity must be strictly construed, sovereign immunity bars such a claim. See Azrael, 765 F.Supp. at 1244-45; Skipper, 781 F.Supp. at 1111-12; Paoli 790 F.Supp. at 96; Western Processing, 761 F.Supp. at 728. This argument founders on its first assertion. CERCLA does provide for a claim against governmental agencies and instructs that they are to be treated like private parties except in a few instances where some greater protection is provided to government bodies acting in a sovereign capacity. One may construe waivers of sovereign immunity narrowly, but the statute must be read fairly and its “cascade of plain language” emphatically waiving sovereign immunity must be acknowledged. See Pennsylvania v. Union Gas Co., 491 U.S. 1, 9-11, 109 S.Ct. 2273, 2279, 105 L.Ed.2d 1 (1989). The remedial cases must overcome the hurdle to absolute immunity for remediators posed by § 9607(d)(1) and its provision for liability by such persons when costs or damages result from their negligence. These cases argue that the provision for liability should not be read to suggest that a claim may be brought under CERCLA against a negligent person who undertakes to act in a remedial capacity but only “clarifies Congress’ intent that the CERCLA remedial scheme not be viewed as occupying the field to the exclusion of tort claims” that could be brought under state law. Skipper, 781 F.Supp. at 1112; Western Processing, 761 F.Supp. at 729-30. Under this view, § 9607(d) would not appear to permit an action under CERCLA against a person who negligently acts in a remedial capacity. This reading is not persuasive. The first sentence of § 9607(d)(1) begins: “Except as provided in paragraph 2, no person shall be liable under this subchapter for costs or damages....” 42 U.S.C. § 9607(d)(1). Thus, the first sentence refers to CERCLA liability for “costs or damages.” The very next sentence provides that “[t]his paragraph shall not preclude liability for costs or damages as the result of negligence on the part of such person.” The remedial cases read this second reference to “liability for costs or damages” as referring to liability under state tort law. This reading is made possible only because Congress chose not to repeat the phrase “under this subchapter” in the latter sentence when it appeared in the immediately preceding sentence. Yet surely the reference in the second sentence is to “costs or damages” under CERCLA and not state law, a subject never mentioned. Moreover, if the statement “[t]his paragraph shall not preclude liability for costs or damages as a result of negligence” were merely a “clarification” of Congress’ intent not to preempt state law, it would be entirely superfluous; the subject of preemption is expressly dealt with elsewhere and comprehensively in § 9614 of CERCLA. Finally, § 9607(d)(2)’s partial exemption of state and local governments from the negligence standard that would otherwise apply demonstrates that § 9607(d)(1) sets a standard of negligence liability for remedial actions, including government remedial actions, because otherwise there would have been no need for the exemption in (d)(2). See 42 U.S.C. § 9207(d)(1) & (2); see also Union Gas Co., 491 U.S. at 9-11, 109 S.Ct. at 2279 (§ 9607(d)(2) “is, needless to say, an explicit recognition of the potential liability of States under this statute”). A last theme that runs through some of the cases endorsing immunity for government remedial activity is reliance on Congress’ rejection of a proposed amendment by Senator Helms “that would have added government misconduct and negligence as a separate defense to CERCLA liability.” Shipper, 781 F.Supp. at 1112; Western Processing, 761 F.Supp. at 730; Atlas, 797 F.Supp. at 415 n. 6. This rejection is said to support the view that Congress did not intend that governmental agencies would be liable for their remedial activities. This argument confuses rejection of a defense to liability with conferral of immunity. That Congress did not wish to provide a polluter with a defense to CERCLA liability because of the government’s negligence does not at all suggest that it intended to exempt the government from all liability for its proportionate share. Indeed, it suggests just the reverse. As discussed above, CERCLA throws a broad liability net and relies on the courts’ sense of equity to apportion the costs of cleanup. Consistent with this scheme it would be surprising if Congress exempted an otherwise responsible person from all liability — whether that person is a governmental body or a private polluter — merely because some other person was also responsible. That is a common situation contemplated by CERCLA, and the rejection of the Helms Amendment does nothing to support the argument that the government should be immune from liability when it negligently causes harm to the environment during a remediation process. In short, Congress had the liability of governmental entities firmly in mind when it enacted CERCLA. The statute sets forth a general rule that governmental entities are to be liable to the same extent as private parties. 42 U.S.C. §§ 9601(20)(D) (states), 9620(a)(1) (federal government). The liability provision, § 9607(a), does not differentiate between governmental and private entities in describing the four classes of responsible persons, although it does differentiate between them in defining recoverable costs. Neither does § 9607(d)(1) provide for special treatment for governmental entities: it provides that “no person” shall be liable for rendering care at a CERCLA site, but that liability is not precluded for costs or damages as the result of negligence on the part of such person. • 42 U.S.C. § 9607(d)(1). Section 9607(d)(2) provides additional protection to state or local governments who respond to emergencies caused by releases from facilities owned by others; they are still liable, and therefore not immune, but only for gross negligence or intentional misconduct. 42 U.S.C. § 9607(d)(2). Based on this “cascade of plain language,” the court holds that there is no “regulatory” or “remedial” exception to CERCLA liability. Where a governmental entity’s “regulatory” or “remedial” activities, of whatever nature, bring the entity within the definition of the terms “owner,” “operator,” “arranger”, or “transporter,” as those terms are applied to private parties, the government will be liable. Further, when a government agency undertakes to remediate a site within the terms of § 9607(d), it will face liability under the standards set forth in that section. The United States’ and State of California’s motions to dismiss RP’s claims as barred by immunity for government acts undertaken in a regulatory or remedial capacity are denied. IV. The United States’ Activities During and After World War II The United States also moves to dismiss the claims related to its activities at Iron Mountain during and after World War II. The United States argues that the activities alleged are insufficient to make it an “operator,” “arranger” or “transporter” under CERCLA. (United States Mem. at 2-3.) RP does not contend that the United States was an owner or transporter at the mine during or after World War II. (RP Opp’n — U.S. at 41, claiming only operator and arranger liability.) Thus, the motion to dismiss is granted to the extent that the counterclaims allege owner or transporter liability based on government activities at the mine during and after World War II. It remains to discuss RP’s claims that the United States was (1) an “operator” and (2) an “arranger” with respect to the mine. A. “Operator” Liability CERCLA defines an “owner or operator” as “any person owning [or] operating” a facility. 42 U.S.C. § 9601(20)(A). RP claims that the United States’ activities at the mine during World War II make the government liable as an “operator” under CERCLA. Although “the circularity of [the] definition [of operator] renders it useless,” it is well settled that “ ‘operator’ Lability under section 9607(a)(2) only attaches if the defendant had authority to control the cause of the contamination at the time the hazardous substances were released into the environment.” Kaiser Aluminum v. Catellus Dev., 976 F.2d 1338, 1341 (9th Cir.1992) (citation omitted); Lincoln Properties, Ltd. v. Higgins, 823 F.Supp. 1528, 1534 (E.D.Cal.1992). See also Long Beach Unified School Dist. v. Dorothy B. Godwin California Living Trust, 32 F.3d 1364, 1367 (9th Cir.1994) (to be liable as an operator one must “play an active role in running the facility, typically involving hands-on, day-to-day participation in the facility’s management”). RP points to several factual allegations in the complaint that it contends support its claim that the United States operated the mine during World War II. First, RP alleges that the United States’ regulatory encouragement of metals mining effectively “directed” Mountain Copper to mine copper, zinc and pyrite ore, “the only business available to it.” (RP Opp’n — U.S. at 43.) The government did this by prohibiting gold mining and by institution of the Premium Price Plan, under which the government contracted to buy zinc and copper from the mine, controlled the marketing and pricing of the ore produced, and set prices to encourage maximum production. (Id. at 43-44.) In response, Mountain Copper built a new mill to process ore, and opened a new ore body at the mine. (Id. at 44.) RP contends that if production of pyrite ore at the mine had not increased, there was a strong likelihood that the mine would have been seized. (Id. at 45.) In addition, RP contends that the government took a variety of steps to assist production, particularly in the areas of labor and transportation. The government built roads on Iron Mountain to facilitate shipments from the mine, hired new employees to work in the mine, granted deferments to mine workers, fixed mine workers’ wages and imposed minimum work weeks, and prevented employees from seeking employment elsewhere without a certificate of separation. (Id. at 46.) Even assuming the historical accuracy of these assertions, the complaint does not state a cause of action for “operator” liability under CERCLA. Although the United States certainly played the role of a very interested consumer during the War, RP’s counterclaim is bereft of any allegations that the United States was involved in the “hands-on, day-today” management of the mine, Long Beach Unified School Dist., 32 F.3d at 1367, or that the United States controlled the cause of contamination, the mining equipment, or made any decisions regarding disposal of the mining waste. Kaiser, 976 F.2d at 1341. Nor are there any allegations that the United States supplied the mine with mining equipment or any other materials. Purchasing a product and encouraging its production are not the same as controlling the cause of the contamination or managing the mine. Thus, the allegations in this case fall far short of the facts in FMC Corp. v. United States Dept. of Commerce, 29 F.3d 833 (3d Cir.1994), the case upon which RP principally relies. In FMC the Third Circuit held the United States liable as an “operator” for its World War II activities at a facility that produced high tenacity rayon for use in airplane and truck tires. But unlike the activity here, amounting to encouragement and regulation, in FMC the government also leased and installed equipment at the facility, built an adjacent factory to supply the rayon plant with sulfuric acid, commissioned the building of another plant to produce carbon bisulfide for the rayon facility, and required the operator of the rayon plant to use those raw materials for the specific purpose of making high tenacity rayon. By its distribution of raw materials, the government “determined the operating level” of the rayon plant. Moreover, the government’s management of the work force and day to day plant operations far exceeded what is alleged here. In the FMC case the government participated in managing and supervising the workers, id. at 837, “appointed a full-time representative to ... address problems at the facility concerning manpower, housing, community services, and other related matters,” and “placed a representative on-site with the authority to promulgate rules governing all operations at the site and to remove workers who were incompetent or guilty of misconduct.” Id. In FMC, “[tjhrough continuous informal contacts and communications, the government was involved directly and substantially with the facility’s production activities and management decisions.” Id. Finally, the government had direct involvement in the generation of waste: “the government knew that generation of waste inhered in the production process; it was aware of the methods for disposal of the waste; and it provided the equipment for the waste disposal.” Id. at 838. Thus, the government’s activities in FMC amounted to “substantial control over the facility” and “active involvement in the activities there,” and were sufficient to create operator liability. Id. at 843 (internal citation and quotation marks omitted). By comparison, RP’s allegations here fall short of placing the government in a position of day to day management with control over waste disposal and mining practices. See United States v. Vertac Chem. Corp, 46 F.3d 803, 809 (8th Cir.1994) (holding United States not liable as operator or arranger of Agent Orange production facility based on regulatory activities despite directing increased production and assisting in procurement of, but not providing, raw materials, because government was not “actively involved on a regular basis in, and thus never exerted substantial control over” the facility). RP’s allegations regarding post-World War II operation of the mine also fall short of establishing operator liability. RP alleges that after the war, the mining and milling of copper and zinc ore ceased, but the government continued its involvement with the mine by paying for equipment and operations in conducting explorations for new ore and by participating in decisions as to the best methods to search for new ore. (RP Countercl.— U.S. ¶ 39.) But RP does not allege that these explorations resulted in any production or release of hazardous wastes, and RP states that the mining had finished at the end of the war. (RP Opp’n — U.S. at 14.) For the foregoing reasons, RP’s claims that the United States is liable under CERC-LA as an “operator” of Iron Mountain Mine based on its activities during and after World War II are dismissed without prejudice for failure to state a claim upon which relief can be granted. B. “Arranger” Liability Nor do RP’s allegations support “arranger” liability under CERCLA. CERCLA “arranger” liability applies to “any person who by contract, agreement, or otherwise arranged for disposal or treatment [or transport] of hazardous substances owned or possessed by such person, by any other party or entity, at any facility_” 42 U.S.C. § 9607(a)(3). Although the wording of this provision is inartful, it predicates arranger liability on ownership or possession of the hazardous substances. See Vertac, 46 F.3d at 810-11 (arranger liability requires ownership or possession). The facts alleged in the complaint fail to establish arranger liability in the United States because of its activities at the mine. RP does not allege that the United States owned the mining wastes from the mine. As discussed above, RP does all