In the teeth of two decades of contrary precedent from four circuits, defendant-appellant Mallinckrodt, Inc. asks us to restrict the role of private citizens in the abatement of imminent and substantial threats to the environment and public health. In support of this entreaty, Mallinckrodt presents a gallimaufry of new, hitherto unconsidered arguments. After careful consideration of this asseverational array, we conclude that our sister circuits have adroitly distilled the meaning of section 7002(a)(1)(B) of the Resource Conservation and Recovery Act (RCRA), 42 U.S.C. § 6972(a)(1)(B) — the so-called citizen suit provision. Correctly interpreted, this provision allows citizen suits when there is a reasonable prospect that a serious, near-term threat to human health or the environment exists.1 In such situations, the provision permits remedies consistent with the scope of a district court's equitable discretion.

The district court read the statute in this manner and faithfully applied the law to the facts. Its supportable liability finding, coupled with a choice of remedy that comes within the encincture of its discretion, leads us to reject Mallinckrodt's appeal.

We rehearse here only those facts that are directly relevant to the issues on appeal, referring readers who hunger for more information to the district court's initial opinion. See Me. People's Alliance v. HoltraChem Mfg. Co., 211 F. Supp. 2d 237 (D.Me.2002). This narrative credits the factual findings of the district court to the extent that those findings are not clearly erroneous. See Fed. Refinance Co. v. Klock, 352 F.3d 16, 27 (1st Cir.2003).

From 1967 to 1982, Mallinckrodt, then called International Minerals and Chemicals Corporation, owned and operated a chlor-alkali plant (the Plant) situated on the banks of the Penobscot River in Orrington, Maine. Thereafter, the Plant continued operations under other owners, namely, Hanlin Group, Inc. and HoltraChem Manufacturing Co.,2 until it closed in 2000. During the period of its operation, the Plant deposited tons of mercury-laden waste into the Penobscot River. See Me. People's Alliance, 211 F.Supp.2d at 253. While there have been a number of other significant contributors to mercury in the Penobscot, "Mallinckrodt has been a dominant source." Id. at 255.

In 1986, the Plant's continuous release of mercury led the Environmental Protection Agency (EPA) to file an administrative RCRA action against Hanlin (the Plant's quondam owner). That action resulted in an agreement for corrective measures. Deeming turnabout fair play, Hanlin sued Mallinckrodt for contribution. In a 1991 settlement, Mallinckrodt agreed to pay a portion of the compliance costs imposed by the agreement.

A subsequent enforcement action led to a 1993 consent decree that superseded the earlier agreement. Although not a party to this consent decree, Mallinckrodt, consistent with the Hanlin settlement, paid its share of the compliance costs and participated in ongoing negotiations with government regulators.3 That included working with both EPA and Maine's Department of Environmental Protection (MDEP).

The 1993 consent decree contemplated a tripartite process comprising site investigation, evaluation of possible corrective measures, and remediation. In line with the first phase of this process, Mallinckrodt compiled and submitted a site investigation report. In March of 1997, EPA and MDEP, acting in concert, issued a draft notice of disapproval. Mallinckrodt countered with a supplemental site investigation report but, in 2000, EPA and MDEP again disapproved. Among other things, the regulators instructed Mallinckrodt to study the effects of mercury downriver from the Plant.

Within a matter of months, Mallinckrodt commissioned a study aimed at examining downriver mercury contamination. It conducted a second downriver study during the summer of 2001. Notwithstanding the submission of these studies, however, the district court supportably found that Mallinckrodt made only minimal efforts to pursue the designated line of inquiry and that the decision to forgo more vigorous efforts was deliberate. Id. at 244 & n. 9.

In the midst of this sparring, two environmental groups — the National Resources Defense Council and the Maine People's Alliance — joined forces to commence a citizen suit under RCRA § 7002(a)(1)(B). The plaintiffs alleged that mercury contamination downriver from the Plant "may present an imminent and substantial endangerment to health or the environment." Acknowledging the possibility that remediation might eventually prove to be either unnecessary or infeasible, their principal prayer for relief was that Mallinckrodt be ordered to fund an "independent, comprehensive, scientific study to determine the precise nature and extent of the endangerment."

Mallinckrodt tried on several occasions to derail the suit on the ground that EPA, not the courts, had primary jurisdiction. The district court demurred, holding that the suit would not present any conflict with agency action due to EPA's apparent lack of interest in the lower Penobscot.

The case was reached for trial in March of 2002. By that time, EPA and MDEP had made public, but had not adopted, preliminary media protection standards, potentially applicable to the lower Penobscot. Had those standards gone into effect, no remediation would have been required for the region with which this litigation is concerned.

During a nine-day bench trial,4 one of the plaintiffs' principal experts was Dr. Robert Livingston. The district court found Livingston, an aquatic biologist, to be "particularly credible and persuasive." Id. at 251. Drawing on three main sources — the data gathered under the EPA-ordered site studies, some limited field work, and the scientific literature concerning mercury in aquatic systems — Livingston opined that there might be a serious endangerment to both human health and the environment resulting from mercury contamination in the lower Penobscot. Although believing it "highly likely" that these harms would prove to be both real and severe, he cautioned that he had not yet "done the right research to determine that." Due to the absence of sufficient research, no one could know with certitude "if there is a problem" or "what the problem is." When all was said and done, however, he thought it "highly likely" that localized and targeted remediation would be both necessary and desirable.

The plaintiffs also adduced testimony from other experts, from individuals within their respective memberships, and from EPA and MDEP representatives. A number of defense experts testified as well. After both sides had rested and submitted briefs, the district court issued a thoughtful rescript.

The court found that mercury in aquatic systems is susceptible to being transformed by microscopic organisms into its organic form (known as methylmercury). Id. at 244. Methylmercury is a highly toxic substance which, even in low dosages, is inimical to human health; for example, it "attacks the nervous system, the kidneys, the immune system, and the reproductive system" and is especially damaging to a developing fetus. Id. at 245. Methylmercury is especially pernicious because it is the most bioavailable form of mercury and therefore, is readily accumulated in humans and animals alike. Id. at 244.

Next, the court found that mercury concentration in sediments extracted from the lower Penobscot runs five times higher than in the Kennebec River (which Mallinckrodt's expert identified as an appropriate comparator). Id. at 248. Despite this high concentration, the court wisely recognized that "the mere presence of mercury contaminated sediments is alone not enough to constitute an imminent and substantial endangerment," id., so it proceeded to examine the available data concerning mercury contamination in various species in the lower Penobscot, including benthos, killfish (minnows), lobsters, blue mussels, cormorants, osprey, and eagles. This examination led the court to conclude that "mercury is methylating downriver" and that "methylmercury is bioavailable, entering biota, and biomagnifying throughout the food web." Id. at 251.

The court expressed heightened concern about a region known as Frankfort Flats, which displayed extraordinarily high mercury readings in both sediments and biota. Id. at 252. Frankfort Flats receives drainage from a marsh system, and marshes are considered to be hotbeds of methylation. See id.

When the district court turned to the legal standard for citizen suits under RCRA § 7002(a)(1)(B), it characterized that standard as "lenient." 211 F. Supp. 2d at 246. It cited with approval case law emphasizing that RCRA allows such a suit when the putative polluter "may" have caused an imminent and substantial endangerment. Id. at 246-47 (collecting cases). On this basis, the court concluded that the statute's "imminent and substantial endangerment" standard would be satisfied by a "reasonable medical concern for public health and a reasonable scientific concern for the environment." Id. at 252.

Applying this interpretation of the statute to the facts as found, the court determined that Mallinckrodt's disposal activities may have created an imminent and substantial danger and that, therefore, the plaintiffs had carried their burden of proof anent liability. Id. at 251-52. Then, having found liability, the court directed the parties to make a good-faith effort to agree on a study plan. Id. at 256. The parties complied and, on August 10, 2005, the court approved a plan which, if carried out, probably will require Mallinckrodt to spend around $4,000,000 for laboratory analyses, independent of all other costs. The purpose of the study is to learn whether, in actuality, mercury contamination in the lower Penobscot adversely affects either human health or the environment, and if so, to devise a feasible remedial approach.

Mallinckrodt now appeals, asserting that the plaintiffs lacked standing to sue in the first place; that the lower court set the bar too low for RCRA citizen suits; and that, in all events, the court abused its discretion in fashioning relief. We address these assertions one by one.

As a threshold matter, Mallinckrodt alleges that the plaintiffs lack standing to sue because they have not suffered an injury in fact.5 The existence vel non of standing is a legal question and, therefore, engenders de novo review. See N.H. Right to Life Political Action Comm. v. Gardner, 99 F.3d 8, 12 (1st Cir.1996). When, however, the trial court's standing determination rests on findings of fact, we must honor those factual findings unless they are clearly erroneous. See Rivera v. Wyeth-Ayerst Labs., 283 F.3d 315, 319 (5th Cir.2002).

We start our inquiry into standing with the undisputed fact that both of the plaintiffs are associations. In order to ground a claim of associational standing (that is, standing to bring suit on behalf of its membership), an association must show three things: (i) that individual members would have standing to sue in their own right; (ii) that the interests at stake are related to the organization's core purposes; and (iii) that both the asserted claim and the requested relief can be adjudicated without the participation of individual members as named plaintiffs. Friends of the Earth, Inc. v. Laidlaw Envtl. Servs. (TOC), Inc., 528 U.S. 167, 181, 120 S. Ct. 693, 145 L. Ed. 2d 610 (2000). The plaintiffs in this case plainly have made the latter two showings. Hence, the inquiry reduces to whether the organizations' individual members would have had standing to proceed in their own right.

Because there is nothing in RCRA's text or history that suggests a congressional intent to erect statutory standing barriers beyond those imposed by Article III of the Constitution and because Mallinckrodt has not identified any prudential standing concerns, we focus on what is essential to establish Article III standing. Those requirements are expressed in a familiar three-part algorithm: a would-be plaintiff must demonstrate a concrete and particularized injury in fact, a causal connection that permits tracing the claimed injury to the defendant's actions, and a likelihood that prevailing in the action will afford some redress for the injury. Lujan v. Defenders of Wildlife, 504 U.S. 555, 560-61, 112 S. Ct. 2130, 119 L. Ed. 2d 351 (1992); Pagán v. Calderón, 448 F.3d 16, 27 (1st Cir.2006). These requisites must be proved "with the manner and degree of evidence required at the successive stages of the litigation." Lujan, 504 U.S. at 561, 112 S. Ct. 2130. When, as now, standing is reviewed after trial, the facts establishing standing "must be supported adequately by the evidence adduced at trial." Id. (internal quotation marks omitted). The ultimate quotient of proof is a preponderance of the evidence. See Perry v. Vill. of Arlington Heights, 186 F.3d 826, 829 (7th Cir.1999).

We first dispose of an argument that verges on the specious. Mallinckrodt protests that the plaintiffs cannot have established a cognizable injury since the district court thought it appropriate to order a remedy — the study — that would determine whether mercury in the Penobscot is "having significant adverse effects" on the environment or "posing an unacceptable risk to human health." This protestation conflates the district court's finding of liability with its choice of remedy. As we shortly shall explain, probabilistic harms are legally cognizable, and the district court made a supportable finding that a sufficient probability of harm exists to satisfy the Article III standing inquiry. See Me. People's Alliance, 211 F.Supp.2d at 253. The fact that the court chose a remedy that aspires to furnish a degree of determinacy before fashioning further relief speaks only to the court's cautious use of discretion in selecting remedies; it does not speak to the plaintiffs' standing as of the present time.

Having dispatched this attempted sleight of hand, we proceed to more serious matters. Mallinckrodt concentrates its standing attack on the plaintiffs' ostensible failure to prove the injury-in-fact component. At trial, the plaintiffs called four witnesses from within their respective memberships, all of whom reside on or near the banks of the Penobscot River. All four vouchsafed that they have modified their behavior due to fear of mercury contamination. Although eager to do so, none of them will eat fish or shellfish from the river nor recreate on or near it. One witness added that, but for the mercury contamination, she would harvest mussels and sell them to supplement her income. The district court credited this testimony. Id.

Plaintiffs in environmental suits may predicate claims of injury on aesthetic or recreational harms. See Laidlaw, 528 U.S. at 183, 120 S. Ct. 693; Sierra Club v. Morton, 405 U.S. 727, 735, 92 S. Ct. 1361, 31 L. Ed. 2d 636 (1972); Save Our Heritage, Inc. v. FAA, 269 F.3d 49, 55 (1st Cir.2001). Still, neither a bald assertion of such a harm nor a purely subjective fear that an environmental hazard may have been created is enough to ground standing. Rather, an individual's decision to deny herself aesthetic or recreational pleasures based on concern about pollution will constitute a cognizable injury only when the concern is premised upon a realistic threat. See Laidlaw, 528 U.S. at 184, 120 S. Ct. 693; see also City of Los Angeles v. Lyons, 461 U.S. 95, 107 n. 8, 103 S. Ct. 1660, 75 L. Ed. 2d 675 (1983) (explaining that "the reality of the threat . . ., not the plaintiff's subjective apprehensions," constitutes the cognizable injury).

In this instance, Mallinckrodt exhorts us to find that the witnesses' stated fears are unreasonable. It points out — and the record confirms — that waterways throughout Maine suffer to some extent from mercury pollution and that, therefore, it is responsible at most for a probabilistic increase in a risk that the witnesses would in any event have had to run.

There is some confusion as to the thrust of Mallinckrodt's argument. The plaintiffs read its brief, not implausibly, as arguing that these facts necessarily limit cognizable injuries to identifiable medical concerns. See Appellees' Br. at 54. Mallinckrodt's reply brief, however, acknowledges that the plaintiffs' principal claim of injury is for diminished enjoyment of their environment and joins issue on the sufficiency of the proof in that regard. See Appellant's Reply Br. at 25. At bottom, this argument suggests that the plaintiffs must show that Mallinckrodt's activities created a significantly increased risk of harm to health or the environment so as to make it objectively reasonable for the plaintiffs' members to deny themselves aesthetic and recreational use of the river.

To establish an injury in fact based on a probabilistic harm, a plaintiff must show that there is a substantial probability that harm will occur. See Warth v. Seldin, 422 U.S. 490, 504, 95 S. Ct. 2197, 45 L. Ed. 2d 343 (1975); see also Adams v. Watson, 10 F.3d 915, 923 (1st Cir.1993). Mallinckrodt suggests that the instant plaintiffs have not demonstrated a sufficiently probable increase in harm because of (i) Dr. Livingston's admitted uncertainty about whether any problem exists and (ii) the notion that politically accountable branches of government, not courts, are the appropriate entities to make judgments as to what risks are acceptable in modern society. Neither suggestion is convincing.

Mallinckrodt's first suggestion relies ultimately on a single snippet of Dr. Livingston's testimony, wrested from its contextual moorings. But a trial court, confronted with a complex and highly ramified factual situation, is fully entitled to consider the import of a witness's testimony as a whole. This case is a good example: Dr. Livingston testified, in effect, that the presence of a great deal of smoke justified looking for a fire. Mallinckrodt excerpts only his isolated statement that he had not actually seen a fire yet. This plucking of the record overlooks the obvious fact that the district court, drawing on the whole of Dr. Livingston's testimony as well as a plethora of other evidence, supportably concluded that "mercury is methylating downriver, and that such methylmercury is bioavailable, entering biota, and biomagnifying throughout the food web" in sufficient quantity that it may well present an imminent and substantial danger to the environment. Me. People's Alliance, 211 F.Supp.2d at 251 (citation omitted). Relatedly, the court found that "the effects resulting from methylmercury exposure . . . clearly endanger reproduction, development, and overall health of the public and the environment," id. at 252, and that "Mallinckrodt has been a dominant source of mercury in the Penobscot River," id. at 255.

In other words — to return to our metaphor — the lower court discerned telltale signs that a fire might already be smoldering. In light of its warrantable findings, the court had ample reason to conclude that Mallinckrodt has created a substantial probability of increased harm to the environment. That increased risk, in turn, rendered reasonable the actions of the plaintiffs' members in abstaining from their desired enjoyment of the Penobscot.

Mallinckrodt's second suggestion is nothing less than a no-holds-barred assault on the federal courts' institutional competency. It emphasizes that the plaintiffs allege no violation of any federally prescribed discharge limits, and then treats this omission as dispositive. See Appellant's Br. at 27 (asserting that "in the absence of any finding by [EPA]," courts are not equipped to determine whether "the medical and scientific uncertainties created by mercury in the Penobscot create an unacceptable increased risk of harm" (emphasis in original)).

This postulate proceeds from a two-part premise. First, in terms of environmental regulation — where important policy tradeoffs must be made between protection and progress — the only injuries that satisfy the criteria for Article III standing are those injuries that are "unacceptable"; and second, the political branches alone, not courts, have the expertise and accountability to determine which injuries society must be prepared to tolerate. In Mallinckrodt's view, this need to restrict cognizable injuries to "unacceptable" injuries is especially acute where, as in this case, the asserted harms are probabilistic.

Mallinckrodt provides scant authority for an argument that has such breathtaking ramifications for the scope of judicial power. Its notion of acceptability is apparently derived from a single, quarter-century-old opinion, in which a respected court cautioned that judges cannot "formulate policy with respect to what [environmental] risks are acceptable." Envtl. Def. Fund v. EPA, 598 F.2d 62, 83-84 (D.C.Cir. 1978). That court, however, was reviewing a challenge to an EPA regulation based on an alleged lack of substantial evidence. There is no discussion either of standing or of Article III's limits on judicial power. The opinion, therefore, affords no solid foundation for the proposition that Mallinckrodt asserts.6

In a related vein, Mallinckrodt argues that a grant of standing in this case would be tantamount to judicial usurpation of regulatory authority because it would permit "private parties to attack EPA risk assessments collaterally, using the very risks EPA deemed acceptable to establish standing and liability, with no deference afforded to EPA's policy judgment." Appellant's Reply Br. at 27-28. But the ingredients that comprise what Mallinckrodt sees as a recipe for disaster — the relatively broad scope of citizens' rights to sue polluters, the existence of liability in such suits, and the less-than-total deference afforded to agency inaction — all turn on legislative choices. Congress has elected to create a cause of action for affected citizens notwithstanding the absence of any EPA-sponsored standard. While we share Mallinckrodt's belief that it would be a usurpation of legislative prerogative for a court to assume policymaking control over environmental regulation, it would be no less offensive a usurpation for a court to refuse to undertake a task validly entrusted to it by Congress. In the last analysis, Article III requires a cognizable injury; it does not speak to the wisdom of the legislature's actions in providing redress for that injury.

We add an eschatocol of sorts. In rejecting Mallinckrodt's arguments as to standing, we remain confident that Congress has not asked federal courts to perform tasks that are beyond their institutional competency. In our view, courts are capable of assessing probabilistic injuries. Moreover, nuisance principles contribute heavily to the doctrinal template that underbraces statutes like RCRA, see, e.g., Cox v. City of Dallas, 256 F.3d 281, 289 (5th Cir.2001), and the tasks involved in adjudicating environmental cases are well within the federal courts' accustomed domain. While courts can (and do) benefit from available agency expertise, it is an insupportable leap of logic to maintain that, in the absence of such input, claims of injury are not cognizable at all.

That ends this aspect of the matter. For the reasons discussed above, we hold that the plaintiffs have standing to sue under RCRA § 7002(a)(1)(B). We turn, then, to the meat of the appeal.

This case revolves around the meaning and purport of RCRA § 7002(a)(1)(B), a statute that, as described above, allows citizens to sue persons or firms whose handling of solid or hazardous waste "may present an imminent and substantial endangerment to health or the environment." Id. The district court read this language as meaning that such suits could be brought to alleviate reasonable medical or scientific concerns. Me. People's Alliance, 211 F.Supp.2d at 252. Mallinckrodt urges a more circumscribed interpretation. To the extent that this interpretive controversy presents a question of statutory construction, we afford de novo review.7 See Lattab v. Ashcroft, 384 F.3d 8, 21 (1st Cir. 2004).

An historical perspective illustrates the strength of the current against which Mallinckrodt is swimming. Congress enacted RCRA in 1976, Pub.L. No. 94-580, 90 Stat. 2795, with the avowed intention of closing "the last remaining loophole in environmental law, that of unregulated land disposal of discarded materials and hazardous waste." H.R.Rep. No. 94-1491, pt. 1, at 4, reprinted in 1976 U.S.C.C.A.N. 6238, 6241. In its original iteration, RCRA § 7002 (now codified in pertinent part at 42 U.S.C. § 6972(a)(1)(A)) offered citizens the opportunity to bring suit against a polluter only when the polluter was alleged to be in violation of a permit, standard, regulation, condition, requirement, or order issued by EPA. At the same time, RCRA created a cause of action, available exclusively to the EPA Administrator, for cases in which the "disposal of any solid waste or hazardous waste is presenting an imminent and substantial endangerment to health or the environment." RCRA § 7003 (codified as amended at 42 U.S.C. § 6973(a)). In suits brought under this latter provision, federal district courts were granted broad remedial authority to "restrain" polluters and take "such other action as may be necessary." Id.

Pertinently we think, Congress later loosened the standard for liability under section 7003. This transpired four years later when Congress passed the Solid Waste Disposal Act Amendments of 1980. That legislation amended section 7003 by substituting the words "may present" for the words "is presenting." Pub.L. No. 96-482, § 25, 94 Stat. 2334, 2348.

In United States v. Price, 688 F.2d 204 (3d Cir.1982), a seminal RCRA § 7003 case, a district court found that a landfill leaking toxic substances posed an imminent and substantial danger to Atlantic City's water supply but refused to order a study of this hazard at the preliminary injunction stage. Id. at 209. The Third Circuit reversed, ruling that the use of the word "may" in RCRA § 7003 was intended to make the provision "expansive." Id. at 213. In its view, "Congress, by enacting section 7003, intended to confer upon courts the authority to grant affirmative equitable relief to the extent necessary to eliminate any risks posed by toxic waste." Id. at 214.

Soon thereafter, Congress passed the Hazardous and Solid Waste Amendments of 1984 (1984 amendments), Pub.L. No. 98-616, 98 Stat. 3221. The 1984 amendments introduced a new provision, RCRA § 7002(a)(1)(B), into the statutory scheme. Using language that tracked the post-1980 text of RCRA § 7003, this new provision extended to citizens the right to sue a polluter who may be causing an imminent and substantial endangerment to public health or the environment. Id. § 401, 98 Stat. at 3268-69. The Senate Report that accompanied the 1984 amendments approvingly cited and quoted Price on several occasions, specifically endorsing that court's conclusion that section 7003 is intended to give courts the tools to "eliminate any risks posed by toxic waste." S.Rep. No. 98-284, at 59 (1983).

Around this same time, the Fourth Circuit decided another RCRA § 7003 case, in which it flatly rejected the proposition that "section 7003 was designed to control pollution only in emergency situations." United States v. Waste Indus., Inc., 734 F.2d 159, 165 (4th Cir.1984). The court emphasized the statute's use of the word "may" and cited Price approvingly.

Price and Waste Industries have become guideposts for courts endeavoring to interpret the counterpart language contained in RCRA § 7002(a)(1)(B). To date, at least four of our sister circuits have construed that provision expansively. See Interfaith Cmty. Org. v. Honeywell Int'l., Inc., 399 F.3d 248, 258-59 (3d Cir.2005); Parker v. Scrap Metal Processors, Inc., 386 F.3d 993, 1015 (11th Cir.2004); Cox, 256 F.3d at 299; Dague v. City of Burlington, 935 F.2d 1343, 1355 (2d Cir.1991), rev'd in part on other grounds, 505 U.S. 557, 112 S. Ct. 2638, 120 L. Ed. 2d 449 (1992). In taking this position, all four courts have emphasized the preeminence of the word "may" in defining the degree of risk needed to support RCRA § 7002(a)(1)(B)'s liability standard.

This expansiveness in construing the requisite degree of risk has largely been matched in the courts' assessment of the gravity and immediacy of the threatened harm. With one possible exception, the courts have agreed that the word "substantial" implies serious harm.8 See, e.g., Parker, 386 F.3d at 1015. There has, however, been some reluctance to quantify the needed level of harm more precisely. See, e.g., Honeywell, 399 F.3d at 259. Imminence generally has been read to require only that the harm is of a kind that poses a near-term threat; there is no corollary requirement that the harm necessarily will occur or that the actual damage will manifest itself immediately. See Cox, 256 F.3d at 299-300.

Mallinckrodt argues that this long line of cases has consistently misread Price (which, Mallinckrodt says, only concerned remedial power, not scope of liability) and, in the bargain, has disregarded the strictures imposed by the adjectives "imminent" and "substantial." We have not had occasion to construe the reach of the citizen suit provision contained in RCRA § 7002(a)(1)(B), and we are obligated to offer our independent judgment on an issue of first impression here (which, as we shortly shall explain, coincides with the result reached by the other courts of appeals that have confronted the question). Accordingly, we proceed to test the mettle of the conventional construction of RCRA § 7002(a)(1)(B) against Mallinckrodt's challenge.

The district court, following the interpretive trail blazed by the four above-mentioned courts of appeals, employed the conventional construction. It began by noting that the word "endangerment" does not imply actual harm but, rather, implies only potential harm. Me. People's Alliance, 211 F.Supp.2d at 246. The court then noted that the statutory standard is further relaxed because of Congress's use of the word "may." Id. As a result, the court found RCRA § 7002(a)(1)(B) to be a "sweeping provision indicat[ing] Congress's intent `to confer upon the courts the authority to grant affirmative equitable relief to the extent necessary to eliminate any risk posed by toxic waste.'" Id. at 246-47 (quoting Dague, 935 F.2d at 1355 (quoting Price, 688 F.2d at 214) (emphasis supplied in Dague)).

The court read the statute's adjectival elements in line with the breadth of the authority granted. It found that the word "imminent" connotes only that the "factors giving rise to [the harm] are present, even though the harm may not be realized for some time." Id. at 247. Relatedly, the court found that the word "substantial" connotes no more than "reasonable cause for concern that someone or something may be exposed" to harm. Id. (quoting Raymond K. Hoxsie Real Estate Trust v. Exxon Educ. Found., 81 F. Supp. 2d 359, 366 (D.R.I.2000) (internal quotation marks omitted)). A "reasonable medical concern" would, the court said, be sufficient to ground liability under the statute. Id. at 252. The court's only explicit limitation on the availability of relief was when "the risk of harm is remote in time, completely speculative in nature, or de minimis in degree." Id. at 247 (quoting United States v. Reilly Tar & Chem. Corp., 546 F. Supp. 1100, 1109 (D.Minn.1982)). Mallinckrodt balks at this interpretation of RCRA § 7002(a)(1)(B) for a host of reasons. It argues that this construction ignores the plain meaning of the phrase "imminent and substantial endangerment," does violence to the internal coherence of RCRA by overlooking the hierarchy of standards contained within that statute, tampers with Congress's wise delegation of policymaking to an expert agency, and frustrates congressional intent. We address each of these criticisms in turn.

Mallinckrodt's textual argument rests on the premise that the courts that heretofore have explicated section 7002(a)(1)(B)'s liability standard have been blinded by the glare of the word "may" and have lost sight of the plain meaning of the words "imminent and substantial." In Mallinckrodt's view, the phrase "may present an imminent and substantial endangerment," when read as a whole, requires a risk of grave harm that is more likely than not to occur. Mallinckrodt deduces this construction from a Rosetta Stone that is part case law and part lexicography.

The word "endangerment," Mallinckrodt says somewhat tautologically, is "the state of being placed in danger." Webster's Third New International Dictionary 748 (1993). In that connection, it defines danger as "exposed to harm" or "peril." Id. at 573 (excess capitalization omitted). To elucidate the meaning of "may," Mallinckrodt points to a sixty-year old Supreme Court decision that defines "may," as used in section 2(a) of the Clayton Act, as "probably." Corn Prods. Refining Co. v. FTC, 324 U.S. 726, 738, 65 S. Ct. 961, 89 L. Ed. 1320 (1945). Finally, to give content to the phrase "imminent and substantial," Mallinckrodt invokes case law suggesting that, in other environmental contexts, the unadorned word "endanger" implies a lower standard than that denoted by the phrase "imminent and substantial endangerment." See Ethyl Corp. v. EPA, 541 F.2d 1, 20 n. 36 (D.C.Cir.1976) (en banc); Reserve Mining Co. v. EPA, 514 F.2d 492, 528 (8th Cir.1975) (en banc).

Mallinckrodt's textual argument makes sense — but only to a point. While the decisions in Ethyl Corp. and Reserve Mining are some evidence that, as of 1976, the phrase "imminent and substantial endangerment" was thought to denote a heightened standard,9 the relevant question is how that term was understood in 1984 (when section 7002(a)(1)(B) was enacted). The Senate Report on the 1984 amendments defines the word "endangerment" separately from the phrase "imminent and substantial." See S.Rep. No. 98-284, supra, at 59. Mallinckrodt's criticism fails to account either for that circumstance or for Congress's insertion, in 1980, of the word "may" into section 7003 and its subsequent use in section 7002. That word does not appear at all in the statute considered in Reserve Mining, 514 F.2d at 528 & n. 70 (citing 33 U.S.C. § 1364). In Ethyl Corp., one of the referenced statutes does use the word "may," but that court did not parse the entire sentence and looked only to the words "imminent and substantial." See Ethyl Corp., 541 F.2d at 20 n. 36 (citing 42 U.S.C. § 300i). We also note that statutes referenced in Ethyl Corp. and Reserve Mining bore the subtitle "Emergency Powers"; in contrast, neither section 7002 nor section 7003 carry such a label.

The sockdolager, we think, is that "words are like chameleons; they frequently have different shades of meaning depending upon the circumstances." United States v. Romain, 393 F.3d 63, 74 (1st Cir.2004). The terrain of the Clayton Act and provisions specifically classified as conferring emergency powers are sufficiently distinct in subject matter from RCRA's terrain. Thus, we are unprepared to say that the nuances attached to the phrase "may present an imminent and substantial endangerment" by judicial decisions in those other contexts should be transplanted root and branch into the differently textured soil of RCRA § 7002(a)(1)(B)-a statute enacted by a different Congress at a different time for a different purpose. We conclude, therefore, that Mallinckrodt's textual argument does not carry the day: the interpretive question before us cannot be resolved favorably to Mallinckrodt on the basis of plain meaning alone.

Mallinckrodt next argues that the conventional interpretation of section 7002(a)(1)(B) drains the phrase "imminent and substantial" of any meaning because RCRA already defines "hazardous waste" as material that will cause an increase in mortality or serious illness or "pose a substantial present or potential hazard to human health or the environment." 42 U.S.C. § 6903(5). Thus, any release of hazardous waste would satisfy the conventional construction of RCRA § 7002(a)(1)(B), and the requirement that such a release create an "imminent and substantial endangerment" would be superfluous.

This argument is easily dispatched. In terms, section 7002(a)(1)(B) applies to both solid waste and hazardous waste. RCRA's definition of "solid waste" does not share the same characteristics as its definition of "hazardous waste," so to that extent the phrase "imminent and substantial" retains an independent meaning.

Relatedly, Mallinckrodt asserts that the conventional interpretation of "imminent and substantial endangerment" overlooks RCRA's commitment to a hierarchy of risks in which "imminent and substantial endangerment" ranks at or near the top. This argument builds on the idea that Congress employed relatively lenient risk standards elsewhere in RCRA. See, e.g., RCRA § 4004(a), codified at 42 U.S.C. § 6944(a) (providing that sanitary landfills must have "no reasonable probability of adverse effects on health or the environment"). Correspondingly, the word "endangerment" is used in other contexts in connection with "imminent danger of death or serious bodily injury." See, e.g., RCRA § 3008(e), codified at 42 U.S.C. § 6928(e) (defining the crime of "knowing endangerment"). Thus, the decision to require an "imminent and substantial endangerment" must signify a special, harder-to-achieve benchmark.

This argument has some superficial appeal. After all, it is a cardinal rule that courts should strive to interpret statutes as a whole and to give effect to every word and phrase. King v. St. Vincent Hosp., 502 U.S. 215, 221, 112 S. Ct. 570, 116 L. Ed. 2d 578 (1991); United States v. Ven-Fuel, Inc., 758 F.2d 741, 751-52 (1st Cir. 1985). But the conventional interpretation of section 7002(a)(1)(B) gives full effect to the "imminent and substantial endangerment" language; it merely eschews the mechanical cross-referencing, not mandated by Congress, that Mallinckrodt advocates. That, in itself, should not raise eyebrows: where the various parts of a complicated and multifaceted statutory scheme discuss significantly different topics and function within different paradigms, mechanical cross-referencing, not mandated by Congress, can lead to confusion rather than clarity.

This is such a case. There is no meaningful parallelism between section 7002(a)(1)(B) and the provisions that Mallinckrodt seeks to use as comparators. For example, RCRA § 3008(e) is a criminal provision. Given the divergent concerns that drive criminal statutes as opposed to civil remedial statutes, it should not be surprising that the same word may vary in meaning as the context shifts. Indeed, if "endangerment" as defined in section 3008(e) were to mean precisely what it means in section 7002(a)(1)(B), the word "imminent" would be rendered utterly redundant — the statute would, in effect, require an imminent imminence — thus transgressing the very canon of construction that Mallinckrodt labors to invoke.

Mallinckrodt offers a more sophisticated version of the argument that RCRA creates a hierarchy of risks when it posits that the statute's most expansive remedies are reserved for the greatest risks. In this regard, it points out that under RCRA § 3019(b), EPA can commission a health assessment whenever it believes that a landfill "poses a substantial potential risk to human health," 42 U.S.C. § 6939a(b); under RCRA § 3013, EPA can order the owner or operator of a facility to conduct monitoring, testing, and analysis as long as the facility "may present a substantial hazard to human health or the environment," id. § 6934(a); and under RCRA § 7003, which has a risk standard identical to RCRA § 7002(a)(1)(B), EPA can act to protect the public health, see id. § 6973(a). From this statutory melange, Mallinckrodt concludes that only the immediate threat of grave harm is sufficient to trigger such far-reaching remedies.

The notion that a hierarchy of risks and remedies exists within RCRA is not easily disproved. What is fairly clear, however, is that any such hierarchy is well-disguised; if one exists, it seems more a product of fortuity than a product of a purposeful legislative initiative. We explain briefly.

Section 3019(b) was first enacted in 1984, Pub.L. No. 98-616, § 247, 98 Stat. 3221, 3265; section 3013 in 1980, Pub.L. No. 96-482, § 17(a), 94 Stat. 2334, 2344; and section 7003 (previously cited) in 1976 (though that provision was amended both in 1980 and 1984). Given that the changes were made to different subtitles of the statute at different times, such incrementalism weakens (perhaps to the vanishing point) any inference that Congress specifically intended a strict hierarchy of harms. If Congress actually wants a delineated hierarchy of risks and remedies, it will have to say so more distinctly. Cf. King, 502 U.S. at 222, 112 S. Ct. 570 (construing that "differences do not necessarily make hierarchies").

This brings us to Mallinckrodt's separation-of-powers argument. It asseverates that RCRA's allocation of policymaking authority to EPA is such that the citizen suit provision must be viewed as an "interstitial, emergency-type remedy." Appellant's Br. at 45. Mallinckrodt asserts that this policymaking authority necessarily includes responsibility for setting pollution standards and that, in setting such standards, EPA, consistent with the tradeoffs inherent in setting virtually any standard, does not aspire to eliminate all risks.

This assertion is founded on an indisputable verity: the principal responsibility for implementing and enforcing RCRA resides with EPA, not with citizens acting as private attorneys general. See Meghrig v. KFC Western, Inc., 516 U.S. 479, 483-84, 116 S. Ct. 1251, 134 L. Ed. 2d 121 (1996). We disagree, however, with Mallinckrodt's contention that the conventional interpretation of section 7002(a)(1)(B) — the interpretation espoused by our sister circuits and by the court below — is inconsistent with this scheme because it allows courts to second-guess EPA's judgments too freely and affords relief based upon harms that EPA has found acceptable.

Mallinckrodt attempts to bolster this argument by citing a plethora of cases for the proposition that environmental standards should be set by EPA, not the courts. See, e.g., Amoco Oil Co. v. EPA, 501 F.2d 722, 735 (D.C.Cir.1974). This proposition is unremarkable, and we take no issue with it. We agree, moreover, that courts must show appropriate respect for EPA's judgments. See Envtl. Def. Fund, 598 F.2d at 83-84. But allowing citizen suits to proceed is not the functional equivalent of allowing courts to hijack EPA's regulatory authority and weave safety standards out of whole cloth.

The case at hand illustrates this point. Although EPA (acting in concert with MDEP) was leaning toward the adoption of media protection standards that would not have required downriver remediation, no standards had been adopted at the time the district court acted (nor, for that matter, does the record suggest that any have been adopted up to the present time). Furthermore, EPA has never taken the position — or even so much as hinted — that correction of the Plant's effects on downriver pollution is bad policy. Thus, this is not a situation in which a court has presumed to grant relief that flies in the face of an express EPA authorization of certain conduct.

We add, moreover, that the district court has been sensitive to separation-of-powers concerns. On three different occasions, it considered and thoughtfully rejected primary jurisdiction challenges.10 And the court sensibly left open the possibility that "primary jurisdiction concerns could arise in the future." Me. People's Alliance, 211 F.Supp.2d at 255. That presumably would cover the contingency of changed circumstances that would eventuate should EPA, some day, actually engage in an enforcement action pertinent to the condition of the lower Penobscot.

The fact that courts retain some latitude in this area is not in any sense incompatible with the statutory scheme. There are four different ways that EPA can preempt a citizen suit — and all four require that EPA itself take diligent steps to remedy looming environmental harm. See RCRA § 7002(b)(2)(B) (codified at 42 U.S.C. § 6972(b)(2)(B)). That same provision narrowly circumscribes EPA's preemptive power; it states that, when preemption is premised on an EPA order, citizen suits are "prohibited only as to the scope and duration of the administrative order." The short of it is that Congress has told the federal courts that they are not required to steer clear of an area simply because that area might be a focus of future EPA activity.11

Mallinckrodt tries to embellish its separation-of-powers argument in another way as well; it maintains that section 7002(a) embodies a strong preference for permit-violation suits as opposed to imminent and substantial endangerment suits. This attempted embellishment does not withstand scrutiny.

When not premised on the title related to hazardous waste, a permit-violation suit may be brought after a 60-day waiting period while an imminent and substantial endangerment suit requires observance of a 90-day waiting period. Compare RCRA § 7002(b)(1) (codified at 42 U.S.C. § 6972(b)(1)), with RCRA § 7002(b)(2) (codified at § 6972(b)(2)). This is the only substantive difference between permit-violation suits and suits alleging imminent and substantial endangerment. Given the relatively minor nature of that lone disparity, we conclude, without serious question, that the structure of the statutory scheme does not offer any persuasive evidence that Congress sought to disfavor suits alleging imminent and substantial endangerment.

Mallinckrodt's emphasis on statutory structure includes an assertion that giving section 7002(a)(1)(B) a broad reading will render nugatory the right to bring permit-violation actions and the like under RCRA § 7002(a)(1)(A). See Appellant's Reply Br. at 15-16 (querying whether, if "there is a big hole in a fence for the big cat, need there be a small one for the small one?") (citation and internal quotation marks omitted). The flaw in this argument, as it pertains to this case, is that the two holes were drilled at different times. Congress placed the "imminent and substantial endangerment" gloss on RCRA's citizen suit provision eight years after making provision for permit-violation suits. Thus, the more appropriate question is: "If Congress deliberately cut a second hole in the fence, is there any reason not to accept the obvious premise that Congress wanted to increase the ease with which cats of all sizes could come through the fence?"

Insofar as Mallinckrodt theorizes that courts lack the competence to function under the conventional interpretation of RCRA § 7002(a)(1)(B), we reject its thesis. To be sure, Mallinckrodt cites case after case for the proposition that forging policy is a task that non-expert, non-accountable judges should not undertake. See, e.g., Chevron U.S.A., Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837, 866, 104 S. Ct. 2778, 81 L. Ed. 2d 694 (1984); United States v. Gilman, 347 U.S. 507, 511-13, 74 S. Ct. 695, 98 L. Ed. 898 (1954). Some of these cases suggest that judicial policymaking in the environmental sphere is especially inappropriate because judges lack special competence to interpret complex scientific, technical, and medical data. See, e.g., Safe Food & Fertil. v. EPA, 365 F.3d 46, 49 (D.C.Cir.2004).

This view sells the federal judiciary short: federal courts have proven, over time, that they are equipped to adjudicate individual cases, regardless of the complexity of the issues involved. Federal courts are often called upon to make evaluative judgments in highly technical areas (patent litigation is an excellent example).12 Performing that quintessentially judicial function in the environmental sphere is not tantamount to rewriting environmental policy. To the contrary, what the lower court did here — listening to the testimony of expert witnesses, assessing their credibility, and determining whether or not a litigant has carried the devoir of persuasion — is very much within the core competency of a federal district court.

As a last-ditch measure, Mallinckrodt strives to convince us that Congress's revealed intent, as reflected in the legislative history, counsels in favor of a narrow reading of section 7002(a)(1)(B). We are not persuaded.

The argument shapes up along the following lines. Citing S.Rep. No. 98-284, supra, at 56, Mallinckrodt reads this archival information as confirming that section 7002(a)(1)(B) was intended to operate "exactly the same" as section 7003. Building on that foundation, Mallinckrodt notes that section 7003, when enacted in 1976, was envisioned as a means of providing "emergency authority." S.Rep. No. 94-988, at 16 (1976). This gloss, though artful, mixes plums and pomegranates.

What is relevant to a congressional statement, in 1984, that section 7002(a)(1)(B) is intended to operate in the same way as section 7003, is not how Congress viewed section 7003 at the time of its original passage but, rather, how Congress understood section 7003 in 1984. This is especially significant because section 7003 was not worded the same in 1984 as it was in 1976. We think it is clear that the 1980 amendment to the provision, substituting "may present" for "is presenting," fundamentally altered how this provision was understood.

The proof of the pudding is in the legislative archives. We refer particularly to the way in which the provision was discussed during testimony taken in anticipation of the 1984 amendments.

At that time, the House of Representatives was told authoritatively that section 7003 contained "very broad and general statutory language" and that the government, "over the last several years," had been advocating "as broad an interpretation [of it] as possible." Solid Waste Disposal Act Amendments of 1983: Hearing on S. 757 Before the Subcomm. on Envtl. Pollution of the S. Comm. on Env't and Public Works, 98th Cong. 17, 29 (1983) (statement of Carol Dinkins, Asst. Atty. Gen., Land and Nat. Res. Div., Dep't of Justice). It also learned that the Department of Justice viewed section 7003 as "loosely worded," so that it conveyed "extraordinary, broad law enforcement powers." Id. at 110, 120.

The Senate's discussion of section 7003 in the course of considering the 1984 amendments likewise suggests an expansive view of the provision. The Senate Report enthusiastically quotes Price for the proposition that section 7003 is "intended to confer upon the courts the authority to grant affirmative equitable relief to the extent necessary to eliminate any risks posed by toxic wastes."13 S.Rep. No. 98-284, supra, at 59 (quoting Price, 688 F.2d at 214). It went on to observe that the "primary intent of the provision is to protect human health and the environment." Id. Thus, there is good reason to believe that Congress, intending to create a provision modeled along the lines of section 7003, understood that section as offering much more than emergency authority.

The legislative history is also at odds with Mallinckrodt's argument that the conventional interpretation of section 7002(a)(1)(B) usurps EPA's policymaking role. In parsing this legislative history, it is important to recognize that Congress, in 1984, was acting against a background finding that there were "serious gaps" in RCRA, that EPA's enforcement actions were characterized by "inadequate effort," and that EPA "ha[d] not been diligent in vigorously pursuing a tough enforcement program." H.R.Rep. No. 98-198, pt. 1, at 20 (1983), reprinted in 1984 U.S.C.C.A.N. 5576, 5578-79. Congress had heard, and apparently credited, testimony from a former Assistant Attorney General about "an astonishing two years of mismanagement at EPA in which the enforcement of the hazardous waste disposal laws suffered greatly." Id., pt. 3, at 6, reprinted in 1984 U.S.C.C.A.N. 5636, 5641. Given this miseen-scène, it seems counterintuitive to suggest that Congress intended to erect an enforcement structure built on exaggerated deference to EPA.

To be sure, the legislative history reflects the anticipation that "courts will accord some deference to [EPA's] technical findings concerning the nature and extent of endangerment." S.Rep. No. 98-284, supra, at 56. But we do not think that "some deference" means either total obeisance or blind allegiance. Congress desired a "tough enforcement program" and found that EPA had not been "diligent in vigorously pursuing" one. Citizen suits were meant to fill the resultant void.14

We are unimpressed by Mallinckrodt's citation to language in the House Report to the effect that section 7002(a)(1)(B) was designed to create a "limited right" for citizens to sue. H.R.Rep. No. 98-198, supra, pt. 1, at 53, 1984 U.S.C.C.A.N. at 5612. The very next sentence in that report explains the nature of the limitation, namely, that this "right can only be exercised if the Administrator (following notice of the intended litigation) fails to file an action under 7003." Id. The conventional interpretation of the statute does not in any way offend this directive.

Finally, Mallinckrodt spotlights a comment from the floor debate on section 7002, in which the bill's sponsor characterized the provision as giving citizens the power "to abate the most serious kinds of hazardous waste situations: Those that may present an imminent and substantial endangerment." 130 Cong. Rec.2081, 2815 (1984) (statement of Sen. Mitchell). This one remark cannot outweigh the substantial countervailing evidence that the citizen suit provision was intended to empower private citizens by granting them relatively broad authority to litigate when EPA had not acted in the face of a reasonable prospect of serious, near-term harm. See Weinberger v. Rossi, 456 U.S. 25, 35, 102 S. Ct. 1510, 71 L. Ed. 2d 715 (1982).

Legislative history is often a mixed bag. Parties frequently are able to mine nuggets from it selectively, picking and choosing isolated statements that serve particular (sometimes conflicting) ends. Here, the legislative history is less than pellucid. On the whole, however, it tends to support an expansive reading of the "imminent and substantial endangerment" standard for liability under RCRA § 7002(a)(1)(B).

To sum up, the combination of the word "may" with the word "endanger," both of which are probabilistic, leads us to conclude that a reasonable prospect of future harm is adequate to engage the gears of RCRA § 7002(a)(1)(B) so long as the threat is near-term and involves potentially serious harm. The language, structure, purpose, and legislative history of the provision will not comfortably accommodate the more restricted reading that Mallinckrodt espouses. While there may be good and wise reasons to adopt a regime in which EPA determinations of environmental liability are exclusive, it is Congress's place, not ours, to construct such a regime. To this date, Congress has not done so.

Given our conclusion that the conventional interpretation of RCRA § 7002(a)(1)(B) is correct, the remainder of the liability inquiry falls neatly into place. None of Mallinckrodt's arguments persuade us that the district court either misconstrued this standard or misapplied it to the facts of this case. While an imminent and substantial endangerment requires a reasonable prospect of a near-term threat of serious potential harm, the court below made supportable findings that suffice to bring this case within the compass of that standard. See, e.g., Me. People's Alliance, 211 F.Supp.2d at 245 (concluding that methylmercury is a "highly toxic substance"); id. at 251 (concluding that, in an aquatic system, "methylation is a continuous process that can go on for decades"). Based on these and other well-founded findings, the plaintiffs established that the potential risk from mercury is serious and likely to be present here and now. In turn, these findings support a conclusion that, as the district court held, there may be an imminent and substantial endangerment to the lower Penobscot River. No more is exigible.

Our environmental odyssey is not yet finished. Mallinckrodt contends that the district court committed an abuse of discretion in ordering it to fund a study of the lower Penobscot.

This contention is rooted in the notion that a court must balance the relevant harms before granting injunctive relief under an environmental statute — even if the statute specifically authorizes that type of relief. See United States v. Bethlehem Steel Corp., 38 F.3d 862, 867 (7th Cir.1994). Mallinckrodt would have us achieve this balance by applying the familiar four-part framework for determining the appropriateness of injunctive relief. Under that framework, the injunction-seeker "must demonstrate: (1) that it has suffered an irreparable injury; (2) that remedies available at law, such as monetary damages, are inadequate to compensate for that injury; (3) that, considering the balance of hardships between the [parties], a remedy in equity is warranted; and (4) that the public interest would not be disserved by a permanent injunction." eBay Inc. v. MercExchange, L.L.C., ___ U.S. ___, ___, 126 S. Ct. 1837, 1839, 164 L. Ed. 2d 641 (2006).

Conceptually, we agree with Mallinckrodt that a trial court, in an environmental case, should consider the balance of relevant harms before granting injunctive relief, even though the statute itself authorizes such relief. The familiar four-part framework for injunctive relief is a suitable guide in such situations. We caution, however, that the operation of that framework is inevitably colored by the nature of the case and the purposes of the underlying environmental statute (here, RCRA).

Against this backdrop, Mallinckrodt insists that the district court mishandled the third of the four enumerated elements by failing adequately to account for the onerous hardships that the study would impose and contrast those hardships with the meager benefits that the study might generate. In a nutshell, Mallinckrodt's position is that the study will saddle it with a staggering economic burden — a burden so open-ended that the overall cost cannot be predicted with any assurance — yet will yield supposed benefits to the plaintiffs' members that are at best speculative and at worst ephemeral. In this regard, Mallinckrodt stresses that it is an open question whether the lower Penobscot is actually in need of any remediation.15

One difficulty with Mallinckrodt's position is that it treats this injunction like a garden-variety injunction and, in the bargain, undervalues the extent of the equitable discretion possessed by a district court after a finding of liability under RCRA § 7002(a)(1)(B). While Mallinckrodt's hardships are relevant to the acceptability of a RCRA remedy, the argument that hardship must always be outweighed by deliverable benefits offends the logic of Price. Even on Mallinckrodt's restrictive reading of that decision, Price holds that "Congress sought to invoke the broad and flexible equity powers of the federal courts in instances where hazardous wastes threaten [] human health." 688 F.2d at 211.

Ironically, Mallinckrodt dresses this argument in the raiment of a defense of judicial prerogatives. It cites Weinberger v. Romero-Barcelo, 456 U.S. 305, 313, 102 S. Ct. 1798, 72 L. Ed. 2d 91 (1982), for the proposition that judges should not lightly assume that Congress has stripped courts of their equitable discretion. But that is a red herring: while it is true that a district court is not commanded, regardless of the circumstances, to issue an injunction after a finding of liability, the court below was under no misapprehension on this score. In reality, Mallinckrodt wishes to restrict, not preserve, the equitable discretion of the district courts, and the case law that it cites is therefore beside the point.

In reviewing a district courts exercise of its equitable discretion under such circumstances, we think it fully appropriate to acknowledge the statutory scheme under which liability was found. Given the strong statement in Price, embraced by the Senate Report, advocating the exercise of equitable remediation of environmental hazards, we perceive a congressional thumb on the scale in favor of remediation. With this backdrop in place, we are unwilling to say that the district court abused its discretion either by starting with the proposition that its primary concern ought to be how best to remedy a potentially serious near-term environmental hazard or by granting relief notwithstanding the absence of a showing that the remedy's demonstrable benefits exceeded its probable costs.

This is not to say that the costs associated with injunctive relief are immaterial; we can imagine circumstances in which the expense entailed in carrying out a particular remedial plan might dwarf the potential benefits to the environment or to human health. Here, however, the anodyne chosen by the district court does not seem so vastly disproportionate to the threatened harm as to warrant, from the vantage point of a cold appellate record, a recalibration of the balance.

This conclusion is reinforced by the fact that Mallinckrodt has not identified any less burdensome, more cost-effective remedy that the court could have imposed to address the perceived environmental harm.16 Where, as here, a polluter has failed to articulate a "nice adjustment and reconciliation between the public interest and private needs," Hecht Co. v. Bowles, 321 U.S. 321, 329, 64 S. Ct. 587, 88 L. Ed. 754 (1944), a burdensomeness argument rarely will gain much traction.

Taking a slightly different tack, Mallinckrodt characterizes the study-plan order as unfair. It should not be forced to pay the entire cost of the study, it complains, because it is not the only entity to have contributed to the pollution of the lower Penobscot. That plaint rings hollow. While Mallinckrodt was not the sole source of mercury contamination, it was a dominant one, so in that sense its claim of inequitable treatment comprises more cry than wool.

At any rate, RCRA liability, generally speaking, is joint and several. See Cox, 256 F.3d at 301 n. 37. The joint and several nature of environmental liability makes it fitting to hold a single polluter responsible for the totality of the damage where, as here, the harm is indivisible. See id.

We have said enough on this score. Once liability has been found, equitable relief in RCRA citizen suits is largely in the informed discretion of the trial court. For aught that appears, the court below did not abuse this discretion in its choice of a condign remedy.

We need go no further. Despite an impressive array of arguments, skillfully presented by extraordinarily able counsel, Mallinckrodt has not persuaded us that the plaintiffs lack standing to sue, that its cramped interpretation of RCRA § 7002(a)(1)(B) is what Congress had in mind, or that the district court acted outside the realm of its discretion in fashioning a remedy for the threatened harm. Consequently, we uphold the district court's rulings in all respects.

Affirmed.