"Establishing Liability Under CERCLA: The Causal Nexus and the Alternative Liability Theory"

ABA Brown Bag Series - Monday, June 05, 2000

Introduction

One of the more heavily litigated issues under the Comprehensive Environmental Response, Compensation and Liability Act of 1980 (CERCLA) is the issue of causation. CERCLA is a strict liability scheme which imposes joint and several liability on parties found to be responsible for contaminating a site. But because CERCLA defines responsibility broadly, a CERCLA plaintiff does not have to show that a defendant caused the release which resulted in the incurrence of response costs. Rather, the plaintiff only needs to show that the defendant is potentially responsible for the release because the defendant falls into at least one of four statutorily-defined categories of covered persons. Once the plaintiff has proven that the defendant is a potentially responsible party (PRP), a presumption of liability arises, and the burden of proof shifts to the defendant, who then may assert an affirmative defense by disproving causation.

Congress created this liability scheme when it chose to enact the Senate's version of the CERCLA bill rather than the House's version, which predicated liability solely on whether the defendant "caused or contributed" to a release or threatened release. See John Copeland Nagle, CERCLA, Causation and Responsibility, 78 Minn. L. Rev. 1493, 1507-08 (June 1994) (citing 1 U.S. Senate Comm. on Env't & Public Works, A Legislative History of the Comprehensive Envtl. Response, Compensation, and Liability Act of 1980 (Superfund), Public Law 96-510, 97th Cong., 2d Sess. 320 (1983)).

The conjecture is that Congress chose this scheme because it allows plaintiffs to circumvent the insurmountable hardship of"fingerprinting" hazardous substances. To prove causation under traditional tort law, a plaintiff would have to trace particular molecules of contamination back to the defendant, who may have deposited the materials at the release site decades ago. Current technology cannot reliably trace, or fingerprint, chemical compounds with this level of specificity, especially after the compounds have mixed and reacted with other substances, as is often the case in hazardous waste facilities. See United States v. Wade, 577 F. Supp. 1326, 1332 (E.D. Pa. 1983) (stating that"to require a plaintiff under CERCLA to >fingerprint' wastes is to eviscerate the statute").

CERCLA requires the plaintiff to prove only (1) that there has been a release or threatened release covered by CERCLA, (2) that the plaintiff has incurred response costs, (3) that the response costs were necessary and consistent with the national contingency plan, and (4) that the defendant is a PRP as defined by CERCLA. See 42 U.S.C. '9607(a); see, e.g., United States v. Alcan Aluminum Corp., 990 F.2d 711, 721 (2d Cir. 1993) (hereinafter Alcan (2d Cir.); Amoco Oil Co. v. Borden, Inc., 889 F.2d 664, 68 (5th Cir. 1989) (citing Ascon Properties, Inc. v. Mobil Oil Co., 866 F.2d 1149, 1152-53 (9th Cir. 1989), among others); Dedham Water Co. v. Cumberland Farms Dairy, Inc., 889 F.2d 1146, 1150 (1st Cir. 1989); see also Nagle, supra, at 1504-05; James M. Sweeney, Comment, Opening the Front Door: The Argument for a Causal Requirement in Multisite CERCLA Litigation, 56 UCLA L. Rev. 1989, 1994 (August 1999). If the plaintiff succeeds in proving these elements by a preponderance of the evidence, a rebuttable presumption of liability arises, and the burden shifts to the defendant to assert an affirmative defense under CERCLA '107(b) by disproving causation. See Amoco, 889 F.2d at 668; United States v. Monsanto Co., 858 F.2d 160, 170 (4th Cir. 1988). This liability scheme eliminates traditional tort causation from the plaintiff's prima facie case. See Nagle, supra, at 1506-08, 1511 n.86.

Nevertheless, the plaintiff's prima facie case must still contain some degree of causation. In proving the four elements of a CERCLA cost recovery or contribution case, and particularly in identifying the defendant as a PRP, the plaintiff is demonstrating a link between the defendant and the release or threatened release. This link is the"causal nexus," which resembles a relaxed version of traditional tort causation. See Amoco, 889 F.2d at 670, n.8 (discussing the relaxed manner of interpreting a causation requirement); see also Sweeney, supra, at 1995-96; Nagle, supra, at 1510-11.

Since CERCLA is somewhat ambiguously worded and lacks a coherent legislative history, case law elucidates the causal nexus requirement. Over the years, litigants and commentators have proposed numerous theories as to what the elements of a causal nexus are. Some have argued that there should be a minimum contamination threshold. See, e.g., Acushnet Co. v. Mohasco Corp., 191 F.3d 69 (1st Cir. 1999); Alcan (2d Cir.), 990 F.2d 711; United States v. Alcan Aluminum Corp., 964 F.2d 252 (3d Cir. 1992) (hereinafter, Alcan (3d Cir.)); Amoco, 889 F.2d 664. Others believe that the courts should read causation back into the statute. See generally Sweeney, supra. The following discussion reviews the elements of a causal nexus and looks at where the related issue of minimum contamination thresholds fits into the overall CERCLA scheme. Focusing particularly on the recent Third Circuit decision in New Jersey Turnpike Authority v. PPG Industries, 1999 WL 1057213 (3d Cir. November 22, 1999), this discussion will also briefly consider the application of the alternative liability theory to CERCLA cost recovery and contribution cases.

CERCLA '107(a) lists the categories of covered persons who are PRPs and provides innocent parties with a cost recovery action against the PRPs. In 1986, the Superfund Amendments and Reauthorization Act added '113(f), which uses '107(a)'s terms to provide PRPs with the right of contribution against each other. See 42 U.S.C. '9613(f). Neither '107(a) nor '113(f) mentions causation as a requirement for recovery. However, '107(a)'s terms have a causal nexus requirement built into them.

The causal nexus has two elements: (1) a link between the defendant and the remediation site (the site nexus) and (2) a link between the release or threatened release and the response costs that have been incurred (the cost nexus). See Sweeney, supra, at 2004-05; Nagle, supra, at 1511. While these elements link the defendant to the site and the release to the response costs, they do not link the defendant to the release. See Nagle, supra, at 1511. Therefore, a defendant only needs to be statutorily linked to the site by virtue of being a PRP in order to be presumed to have caused the release or threatened release and to be held liable for the plaintiff's response costs, if they are objectively reasonable.

Sometimes referred to as simply the"nexus," the first element of the causal nexus requires a link between the defendant and the site of the release or threatened release. See New Jersey Turnpike Authority, 1999 U.S. App. LEXIS 30389 at *19-20 (citing General Elec. Co. v. AAMCO Transmissions, Inc., 962 F.2d 281, 286 (2d Cir. 1992)). If there is such a link, the defendant is a PRP. In single-site cases, where hazardous substances were deposited at the site of the release or threatened release, establishing the site nexus is a one-step process. In multi-site cases, where hazardous substances were deposited at a site other than the site of the release or threatened release, establishing the site nexus is a two step process.

CERCLA defines each category of PRP according to how the PRP is linked to the site of the release or threatened release. See 42 U.S.C. '9607(a); see also Nagle, supra, at 1511. This link constitutes the entire site nexus requirement in single-site cases. Current owners and operators are linked to a site by their ownership and operation of the site. See 42 U.S.C. ''9607(a)(1), 9601(20)(A)(ii). Previous owners and operators are linked to a site if they owned or operated the site at the time a hazardous substance was deposited there. See 42 U.S.C. '9607(a)(2). Transporters are linked to a site if they ever brought a hazardous substance to the site for disposal or treatment and if the site still contains such a substance at the time of release or threatened release. See 42 U.S.C. '9607(a)(4); see also Nagle, supra, at 1511-13. Generators and arrangers are linked to a site if they generated hazardous substances that were delivered, or arranged for hazardous substances to be delivered, to the site for disposal or treatment and if the site still contains such substances at the time of release or threatened release. See 42 U.S.C. '9607(a)(3); see also Nagle, supra, at 1511.

These links are based on CERCLA's justifications for holding each particular PRP group liable. For example, the owner and operator of a release site has theoretically derived benefit from his use of the property and is therefore responsible for its condition, whether or not he caused or was even aware of the release. See Monsanto 858 F.2d at 170. Similarly, the previous owner and operator of a release site and transporters who delivered hazardous substances to the release site have also derived benefit from the use of the property and therefore are also responsible for the condition of the property.

The broad scope of the site nexus is evident in the latitude of these links. The site nexus is concerned only with the delivery of hazardous substances to the release site. The identity of the hazardous substances deposited by the defendant at the site or during the defendant's ownership and operation of the site is not relevant to the establishment of the site nexus, only to the cost nexus. See Nagle, supra, at 1511-13. Section 107(a) defines PRPs according to"the release, or threatened release... of a hazardous substance," indicating that the release or threatened release of any hazardous substance, not just the particular substance deposited by the defendant, is sufficient to trigger a PRP inquiry. 42 U.S.C. '9607(a) (emphasis added).

The statutorily defined link for generators and arrangers is too broad to be justifiable, however. Under a strict reading of '107(a)(3), a defendant who generates or arranges for a hazardous substance to be disposed of or treated at a site but who never actually sends anything to the release site can still be linked to the site and held jointly and severally liable for response costs. See 42 U.S.C. '9607(a)(3); O'Neill v. Picillo, 682 F. Supp. 706, 719 n.2 (D.R.I. 1988). Not surprisingly, the site nexus requirement for generators and arrangers has been a frequent point of contention in court. As a result, case law has created a test for establishing a generator's link to the release site. Although still challenged by litigants, it enjoys standard use.

To identify a generator as a PRP in a particular release site, there must be evidence that the generator's hazardous substances were removed from the generator's property and were deposited at the release site. See, e.g., General Elec., 962 F.2d at 283, 286; Monsanto, 858 F.2d at 169 n.15; Dana Corp. v. American Standard, Inc., 866 F. Supp. 1481, 1493 (N.D. Ind. 1994); see also Developments in the Law, supra, at 1522. In keeping with CERCLA's strict liability scheme, there does not need to be any evidence that the generator intended for the hazardous substance to go to that particular site. See Wade, 577 F. Supp. at 1333 n.3; see also Nagle, supra, at 1511 n.86 (citing Monsanto, 858 F.2d at 169 n.15; United States v. Mottolo, 695 F. Supp. 615 (D.N.H. 1988); Violet v. Picillo 648 F. Supp. 1283 (D.R.I. 1986); United States v. Conservation Chem. Co., 619 F. Supp. 162 (W.D. Mo. 1985)). It is sufficient to show that the generator sent the hazardous substance on its way and that it arrived at the release site.

Since this test permits a broad application of liability, it has often been challenged. On the issue of intent, one generator defendant argued that since it had allowed the transporter to choose where to take the hazardous waste, it should not be held liable for the consequences of the transporter's decision. See Monsanto, 858 F.2d at 169 n.15. Another generator defendant argued that the transporter's misconduct in taking the hazardous substance to a site other than the one chosen by the generator should absolve the generator of all liability. See Violet, 648 F. Supp. 1283. Yet another generator defendant wanted to be dismissed from the cost recovery action because its hazardous substances had been unearthed from the site to which they had been originally sent and had been moved to the release site without the generator's knowledge. See Developments in the Law, supra, 1516 n.19 (citing Missouri v. Indep. Petrochem. Corp., 15 Envtl. L. Rep. 20161 (E.D. Mo. Jan. 8, 1985)). In every case, the courts held that evidence of the generator's intentions is irrelevant. The test for a generator's link to a release site looks only for delivery and is not concerned with proof of intention.

The case of New Jersey Turnpike Authority v. PPG Indus., 16 F. Supp. 2d 460 (D.N.J. 1998), provides a good illustration of the site nexus test in a generator case. In that case, the New Jersey Turnpike sought '113(f) contribution from three industrial defendants for the remediation of several sites contaminated with chromate ore processing residue (COPR). The Turnpike had evidence to demonstrate (1) that the defendants had generated COPR in the vicinity of the release sites, (2) that they had made the COPR available for use as fill in local construction projects, (3) that the release sites were located in areas where there had been major highway and underground construction at the time, and (4) that COPR contamination at the sites had caused the Turnpike to incur response costs. See id. at 464-65. While this evidence proved that the defendants had generated COPR and that COPR had turned up at the release sites, it did not prove that the defendants' COPR had actually been deposited at the sites. See id. at 468. The defendants moved for summary judgment based on the Turnpike's failure to prove a site nexus under CERCLA. See id. at 465 (stating"that NJTA cannot recover from any of the Generator Defendants in this action unless it establishes, at a minimum, that COPR generated by that specific defendant was deposited at one or more of the seven NJTA sites at issue in this litigation.").

In response to the defendants' motion for summary judgment, the Turnpike argued that there was a genuine issue of material fact because the defendants were the only COPR generators in the region. See id. 464-65. The Turnpike reasoned that since the defendants were the only possible sources of COPR, at least one of them had to have generated the COPR which was contaminating the release sites, but since the COPR generated by each defendant was indistinguishable from the other defendants' COPR, it would be impossible to link one or another defendant to each site. Therefore, the Turnpike suggested that the court apply the theory of alternative liability to the case and shift the burden of proof to each defendant to prove that it did not deposit COPR at any of the sites.

The district court refused to apply the theory of alternative liability. First, the court noted that there was no precedent for applying the theory of alternative liability under CERCLA. Second, the court stated that shifting the burden of proof would not accomplish much since the Turnpike was actually in a better position than the defendants to determine the source of the COPR at the release sites; it was the present owner and operator of the release sites, and it was the party who had contracted to receive the COPR at the release sites in the first place. Third, the court explained that alternative liability applies only when the plaintiff is blameless and thus could not apply in this instance because the Turnpike was a jointly and severally liable PRP in all the release sites. See id. at 471.

The district court also found that the Turnpike had"not demonstrated with competent evidence that any of the Generator Defendants acted negligently towards [the Turnpike] by transporting, depositing, or even directing the delivery of COPR onto any of the sites in question." Id."While it is not necessary for [the Turnpike] to...trace the cause of the response costs to each Generator Defendant, it is not enough that it simply prove that each Generator Defendant produced COPR and that COPR was found at each of the sites in question and ask the trier of fact to supply the link." Id. at 469. Therefore, the court held that the Turnpike had failed to demonstrate a causal nexus between the defendants and the release sites, and granted summary judgment to the defendants.

On appeal, the Third Circuit affirmed, holding that the Turnpike's evidence was"insufficient to prove the nexus required for the Turnpike to recover from the appellees under...CERCLA" and insufficient"to show that each of these appellees acted in a tortious manner within the meaning of these statutes toward these sites such that an alternative liability theory would be appropriate." New Jersey Turnpike Authority v. PPG Indus., 1999 WL 1057213, *11. The court criticized the Turnpike's evidence as"present[ing] probabilities rather than proof," and as being"vague and imprecise, of questionable reliability, and therefore not sufficiently probative to create an issue for trial." Id. at *8, *11. The Turnpike had not proved delivery and therefore had"not produced sufficient evidence to survive summary judgment, even if alternative liability were to apply." Id. at *8.

Multi-site cases present an added complexity to proving a site nexus. In multi-site cases, the release of hazardous substances occurs at a location other than the site with respect to which the defendant is a PRP. Therefore, establishing that the defendant is a covered person under CERCLA with respect to one site does not necessarily establish a site nexus between the defendant and the site of the release. For example, if the defendant stored hazardous substances at site A and the hazardous substances leaked into the ground and migrated to site B where the plaintiff discovered them and remediated them, showing that the defendant stored hazardous substances at site A does not automatically link him to site B, particularly if there are other hazardous substance storage facilities near site B.

Proof of a site nexus in multi-site cases is two-fold. First, the plaintiff must establish a link between the defendant and some site where hazardous substances are located. Then, the plaintiff must show that this site was the source of the hazardous substances that were released elsewhere, causing the plaintiff to incur response costs. See Sweeney, supra, at 2017 (Ain a situation where multiple sites are alleged to be sources of contamination, plaintiffs must demonstrate that the release from [the defendants' waste site] contributed to the contamination" on the plaintiff's site, Kelley v. Kysor Indus. Corp., 1994 U.S. Dist. LEXIS 21194 (W.D. Mich. Oct. 27, 1994)). The link between the two sites is particularly important in cases where there is more than one possible source of the hazardous substances. See id. (stating,"[W]here multiple sites may be responsible for releases causing the contamination and that contamination resulted in response costs being incurred, plaintiffs must provide evidence that the contamination resulted from a release from defendant's site.").

Nevertheless, the degree of certainty required to prove this link is not great. The most reliable method of proving this link would be to show that hazardous substances from the defendant's site physically migrated to the plaintiff's property, and at least one commentator has concluded that a plaintiff should perform chemical analyses of samples from both sites and demonstrate that the samples match in order to recover under CERCLA. See generally, Sweeney, supra. However, the Fourth Circuit stated in Dedham Water Company v. Cumberland Farms Dairy, Inc.,"To our knowledge, every court that has addressed this issue, with the exception of the district court in the instant case, has held that it is not necessary to prove actual contamination of plaintiff's property by defendant's waste in order to establish liability under CERCLA. There is nothing in the statute, its legislative history, or the case law, which requires proof that the defendant's hazardous waste actually have migrated to plaintiff's property, before CERCLA liability is triggered." 889 F.2d at 1154. The Dedham court held that CERCLA plaintiffs only need to demonstrate the possibility that hazardous substances migrated from one site to the other to complete the site nexus link. See id.

The Dedham court cited Artesian Water Co. v. New Castle County, 659 F. Supp. 1269 (D. Del. 1987), aff'd, 851 F.2d 643 (3d Cir. 1988), as an illustration of how to establish a multi-site nexus. In that case, the court granted summary judgment to the plaintiff on the issue of liability based on the plaintiff's proffer of evidence demonstrating that hazardous substances were present in the groundwater at the plaintiff's site and in the sediment near the defendant's site. See id. at 1281. The district court found this evidence to be sufficient to show the possibility that hazardous substances had migrated from one site to the other, even though some of the analyses merely assumed that the contaminants detected near the defendant's site had emanated from the defendant's site. See id.

The Artesian court explained that the plaintiff"ha[d] made a showing of a release or threatened release of hazardous substances from the [defendant's site] sufficient to satisfy its burden on summary judgment... [O]nce [the plaintiff] produced facts in support of its summary judgment motion, the [defendant] could not rely on simple contrary assertions to preserve a genuine issue for trial. (citation omitted)." Id."[A]nd most importantly, the policies underlying section 107 conflict[ed] with the [defendant's] demand that [the plaintiff] prove beyond dispute that the contaminants found near the [defendant's site] actually flow[ed] from the [defendant's site]." Id. at 1282.

The second element of the causal nexus requires a link between the release or threatened release and the response costs for which the plaintiff is seeking recovery. Because the site nexus looks only to a link between the defendant and the site and does not consider the identity of the hazardous substances in question, the cost nexus is necessary to ensure that the assessment of liability against the defendant is objectively reasonable. This element of the causal nexus test is tantamount to the damages element of a traditional tort case, but once again, it requires only minimal links. See Nagle, supra, at 1516.

Section 107(a) states that PRPs are liable if"there is a release, or a threatened release which causes the incurrence of response costs, of a hazardous substance." According to the placement of the commas, the phrase"which causes the incurrence of response costs" modifies"a threatened release" only, implying that there is no evidentiary requirement for the link between an actual release and the plaintiff's response costs. See Control Data Corp. v. S.C.S.C. Corp., 53 F.3d 930, 935 n.8 (8th Cir. 1994) (noting that A[t]he absence of a comma after"threatened release" in the phrase...makes it at least questionable that there is a causation element at all in cases where there is an actual...release."); New York v. Shore Realty Corp., 759 F.2d 1032, 1044 n.18 (2d Cir. 1985) (noting that"the absence of [the] comma...leaves it uncertain whether there is liability from a release without the incurrence of >response costs.'"); Sweeney, supra, at 1996. Since all of a plaintiff's response costs would be presumed to have been caused by the release, a plaintiff could theoretically recover for response activities unrelated to the release.

Few courts, if any, apply '107(a) this way. Most require at least an inference that the release and the response costs are connected. See, e.g., Control Data Corp, 53 F.3d 935 n.8 (citing itself in General Elec. Co. v. Litton Indus. Automation Sys., Inc., 920 F.2d 1415, 1417-18 (8th Cir. 1990); Louisiana-Pac. Corp. v. Beazer Materials & Servs., Inc., 811 F. Supp. 1421, 1428 (E.D. Cal. 1993). This provides a measure of protection to defendants, particularly in cases where the response costs are severable and do not relate to the defendant's specific hazardous substance. Thus, if a defendant disposes of substance A at a site, and the plaintiff incurs response costs cleaning up substance B but leaves substance A in situ, no cost nexus exists, and the defendant is not liable for the response costs.

b. Threatened Releases

As for threatened releases, it is unambiguous that '107(a) requires plaintiffs to prove that their response costs were caused by the threat of release. Because no physical harm has occurred in a threatened release, the link between the defendants and the plaintiff's injury is already quite tenuous. It only makes sense that the plaintiff should prove that his investigation expenses and preventive actions were necessitated by a real threat, rather than by mere conjecture or other causes. See generally, Sweeney, supra.

For example, in Lansford-Coaldale Joint Water Authority v. Tonolli Corporation, 4. F.3d 1209 (3d Cir. 1993), the plaintiff, a public water authority, sought to recover the costs of water quality monitoring and treatment allegedly necessitated by the defendant's threatened release of hazardous substances at an adjacent site. At summary judgment, the district court reviewed evidence indicating that the water authority's reservoirs were hydrogeologically separate from the adjacent site's groundwater flows. The court then found that the threatened release from the adjacent site had reasonably caused the water authority to incur the initial response costs of testing its water supplies and performing a hydrogeological study. But since the hydrogeological study had shown that it would be impossible for an actual release to occur, the court held that no future response costs could be reasonably incurred because of the threatened release. The Third Circuit affirmed on these two points.

CERCLA '113(f) states that"[a]ny person may seek contribution from any other person who is liable or potentially liable under section [107(a)] of this title." 42 U.S.C. '9613(f). It does not require contribution plaintiffs to provide any more evidence than '107(a) requires. Since it borrows its evidentiary requirements entirely from '107(a), the causal nexus test does not differ between the two causes of action.

Because the causal nexus requirement is a derivation of CERCLA's statutory wording, courts are loath to add new elements to the test for a causal nexus. One element many defendants have promoted for inclusion in the causal nexus test is a minimum contamination threshold. The minimum contamination threshold would require a plaintiff to prove that a generator or transporter defendant not only had a link to the site of the release or threatened release but also had contributed more than a certain minimum quantity of a hazardous substance.

One particularly well-litigated argument for the minimum contamination threshold is based on the definition of"hazardous." In United States v. Wade, 577 F. Supp. 1326, the defendant claimed that he could not be held liable for response costs because he had sent too small a volume of waste to the release site for it to be considered hazardous. While noting that the defendant had a good point, the district court declined to create a minimum contamination threshold test. Instead, the court considered the defendant's de minimis status in the liability allocation stage of the lawsuit, stating that"the defendant's fears of draconian liability are overstated. Given my ruling on joint and several as opposed to apportioned liability, a defendant whose sole contribution to a hazardous waste dump site was a copper penny would not be responsible for the entire cost of cleaning up the site." Id. at 1341.

In the Alcan Aluminum cases, the Second and Third Circuits agreed with the Wade court's application of the defendant's de minimis status. See Alcan (2d Cir.), 990 F.2d 711; Alcan (3d Cir.), 964 F.2d 252. The Third Circuit's opinion stated that"courts that have addressed this issue have almost uniformly held that CERCLA liability does not depend on the existence of a threshold quantity of a hazardous substance." Alcan (3d Cir.), 964 F.2d at 260. AIf Congress had intended to impose a threshold requirement, it could easily have so indicated. We should not rewrite the statute simply because the definition of one of its terms is broad in scope." Id. at 261. The Second Circuit echoed the Third Circuit's sentiments in holding that"CERCLA does not impose quantitative requirements with respect to hazardous substances." Alcan (2d Cir.), 990 F.2d at 711.

Another argument in favor of a minimum contamination threshold concerns the laxity of the link required between a release and the plaintiff's response costs. In Acushnet v. Mohasco, 191 F.3d 69, a group of defendants argued that they had sent so little waste to the release site, in effect, they could not have caused the plaintiff to incur any response costs. Evidence presented to the court demonstrated that one of these defendants was responsible for only 1/500,000th of the contamination found at the release site and the others were responsible for similarly small amounts of contamination. Id. at 74. The district court granted summary judgment to the defendants on two independent grounds: First, the district court agreed that the defendants had not contributed enough material to the release site to have caused any of the response costs. Id. Second, the district court held that there was not enough evidence to prove that any of the response costs could be apportioned to the defendants. Id.

The First Circuit affirmed the Acushnet judge's holding but disavowed the first of his two reasons."To the extent that the [district] court's ruling may be interpreted to incorporate into CERCLA a causation standard that would require a polluter's waste to meet a minimum quantitative threshold, we disagree." Id. at 72. The First Circuit then placed the issue of minimum contamination thresholds squarely in the realm of cost allocation by stating,"[W]e conclude that the record was insufficient to permit a meaningful equitable allocation of remediation costs against any of these defendants under ['113(f)]." Id.

Perhaps what makes the application of minimum quantity evidence so confusing is the fact that so many courts use it at the summary judgment stage well before the question of liability has been settled. The structure provided in CERCLA's statutory wording indicates that cost recovery or contribution actions should progress as follows:"liability is fixed first and immediately for enforcement purposes; litigation later to sort out what contribution is owed and by whom as a result of the remediation effort." Alcan (2d Cir.), 990 F.2d at 723. But although it seems like the courts are using minimum quantity evidence in the liability phase of the action, they are actually considering cost apportionment early.

In Alcan, the Second Circuit stated that case management and judicial efficiency should dictate when each phase of a CERCLA action should be visited."[T]he choice as to when to address divisibility and apportionment are questions best left to the sound discretion of the trial court in the handling of an individual case." Id. The concurring judge in the Monsanto case went further and explained,"I see great danger in postponing the ultimate apportioning of the damages to a later day. As an example, a small generator which deposited a few gallons of relatively innocuous waste liquid at a site is jointly and severally liable for the entire cost of cleanup...The vagaries of and delays in his subsequent suit for contribution might result in needless financial disaster." Monsanto, 858 F.2d at 176 (Widener, J., concurring and dissenting).

In short, quantity evidence is not relevant to the determination of liability under CERCLA because the causal nexus presents an absolute yes or no question. Depending on what kind of hazardous substances are present, their relative toxicities, their migratory capabilities and their other characteristics, quantity evidence may be relevant at the apportionment phase only. Therefore, if a court does consider quantity evidence at the summary judgment phase of a cost recovery or contribution action, it is performing an apportionment, usually for de minimis PRPs, having determined that doing so would be in the best interest of judicial efficiency and case management.

Because it is sometimes difficult to establish even the minimum facts to satisfy the causal nexus requirement, plaintiffs have recently begun asserting the alternative liability theory in CERCLA cases. The alternative liability theory would lower the burden of proof even further than the causal nexus requirement does by requiring plaintiffs to prove only that the defendants’ hazardous materials were probably deposited at the site of the release or threatened release, rather than that they were actually deposited there.

Although the New Jersey Turnpike requested and vigorously argued that the theory alternative liability should apply in its case against PPG Industries et al., neither the district court nor the Third Circuit applied it. Both courts left open the question of whether alternative liability can be used under CERCLA and decided the Turnpike case on other issues. However, dicta in both courts' opinions indicate that if alternative liability could apply under CERCLA, the threshold test for determining whether to apply it would require proof 1) of the plaintiff's innocence under CERCLA and 2) of the defendants' tortious actions towards the release site. In other words, the plaintiff would have to prove at least that the defendant's hazardous substances had been directed towards the site in order for alternative liability to apply. Evidence proving the mere"probability" of this direction would not be sufficient; the Third Circuit's opinion demanded that the plaintiff would have to present"proof" of the direction. See New Jersey Turnpike Authority, 199 WL 1057213, at *8.

In his concurrence, Chief Judge Becker of the Third Circuit assumed that alternative liability was applicable and disagreed with the majority as to what the threshold test for alternative liability should be. Chief Judge Becker felt that as long as the plaintiff exercised diligence in determining the identity of the responsible parties, the plaintiff should only have to prove 1) that the plaintiff had suffered an actionable harm, 2) that the defendant had acted in a manner that"exposed it to liability," and 3) that the plaintiff had joined all defendants who might be responsible for the harm. To define when a defendant exposes itself to liability, Chief Judge Becker stated that"the requisite action would be generating the relevant kind of hazardous waste in the relevant area regardless of fault." See id. at *12.

Since it has not been determined whether alternative liability can apply under CERCLA, it is not obvious which of these two tests would apply. Nevertheless, the Third Circuit majority opinion appears to lay out a test that is more consistent with CERCLA and with Summers v. Tice, 33 Cal.2d 80, 199 P.2d 1 (Cal. 1948), the seminal case on alternative liability. In Summers, a hunter was allowed to shift the burden of proof to the two defendants because both had negligently fired their guns in his direction and both were in a better position than the plaintiff to determine whose bullet had hit the plaintiff. Like Summers, the test established in the majority opinion requires the defendants to have acted negligently toward the plaintiff. Chief Judge Becker's test, on the other hand, eliminates the requirement of negligence or carelessness entirely. Under such an analysis, the plaintiff in Summers would have been able to hold all the hunters within shooting distance liable for his wound, whether or not they were aiming in his direction.

Plaintiffs seeking contribution or recovery of response costs under CERCLA must prove a causal nexus between the defendant and the response costs. To date, there is no authority allowing plaintiffs to avoid proving the causal nexus. Although the theory of alternative liability has been asserted in a few CERCLA cases, it has not yet been applied and its applicability remains an exercise in academics.

The elements of a causal nexus are derived from CERCLA's statutory wording and represent a relaxed causation standard. To prove a causal nexus, a plaintiff must show that the defendant is associated with the site where the response costs were incurred and that the response costs are associated with the release or threatened release. The plaintiff does not have to show that the defendant caused the release or that the defendant contributed more than a minimum threshold of contaminants to the release site. If available, evidence of these latter issues is relevant to cost allocation and apportionment, which may be considered at the summary judgment stage, if the trial court finds it to be in the interest of efficiency.