Supreme Court Narrows Reach of 'Arranger' Liability, Potential for Joint and Several Liability Under CERCLA

Article

In-Sites

May 8, 2009

In a closely watched pair of consolidated cases involving the federal Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA), the U.S. Supreme Court has made it much more difficult to impose cleanup liability on companies that shipped hazardous substances to other companies’ facilities in circumstances that were not obvious disposal transactions. The Court also endorsed a fairly relaxed approach to apportionment of harm that effectively lightens the burden for CERCLA defendants seeking to escape joint and several liability. The case is Burlington Northern & Santa Fe Ry. Co. v. United States, No. 07-1601 (U.S. May 4, 2009), which had been consolidated with another case arising from the same lower court decisions, Shell Oil Co. v. United States, No. 07-1607.

CERCLA imposes strict liability for cleanup costs on several categories of parties associated with a contaminated site. One such category consists of those who “arranged for disposal or treatment” of hazardous substances at the site. 42 U.S.C. § 9607(a)(3). The consequences of being adjudged liable can be severe: the statute has been read to permit (without requiring) the imposition of joint and several liability, so that only one or a few of all of the liable parties may be ordered to pay 100% of the cleanup costs.

Burlington Northern concerned both of these critical issues. Between 1960 and 1989, Brown & Bryant, Inc. operated a chemical distribution facility in Arvin, California, on a site it owned and, as of 1975, also on a smaller adjacent parcel owned by several railroads. Brown & Bryant purchased pesticides and other chemicals from Shell and other suppliers, and then applied its products to customers’ farms. Over time, the site became contaminated, and beginning in 1989 (after Brown & Bryant had become insolvent and ceased operations), the federal and state governments took steps to clean it up The federal government also ordered the railroads to perform remedial work. Eventually, separate lawsuits brought to recover cleanup costs were consolidated and tried in federal district court.

The district court found both Shell (as an arranger) and the railroads (as owners of part of the site, see 42 U.S.C. § 960 (a)(1), (2)) were liable for cleanup costs. But rather than imposing joint and several liability on Shell and the railroads for all of those costs, the district court found the harm at the site to be divisible and therefore capable of apportionment. Considering (1) the percentage (19%) of the total site area owned by the railroads, (2) the percentage (45%) of the site’s time of operation during which the railroads leased their parcel to Brown & Bryant, and (3) its determination that chemicals spilled in the railroads’ parcel were responsible for roughly two-thirds of the contamination requiring remediation, and applying a 50% margin to allow for calculation errors, the court assigned a 9% share to the railroads ([0.19 x 0.45 x 0.66] x 150% = 9%). On appeals by Shell and the governments, the Ninth Circuit affirmed as to Shell’s liability, but reversed on the district court’s apportionment and held Shell and the railroads jointly and severally liable. The Supreme Court then agreed to hear the appeals of Shell and the railroads.

“Arranger” Liability Requires an Intent to Dispose

Writing for an 8-1 majority (only Justice Ginsburg dissented, Justice Stevens noted that some transactions clearly are arrangements for disposal — such as a manufacturer’s shipment of waste to a landfill — and some clearly are not — such as the same manufacturer’s sale of a “new and useful product” that the buyer, without the manufacturer’s knowledge, later discards). Slip op. at 9-10. What about cases that fall in between these extremes? One such situation is a tolling arrangement, in which Company A ships a chemical to Company B’s plant for processing, with the processed material subsequently returned to Company A. Who may be held responsible when Company B’s facility is contaminated by spills from the processing of Company A’s material? Beginning with the seminal case of United States v Aceto Agricultural Chemicals Corp., 872 F.2d 1373 (8th Cir. 1989), some circuit courts held that in such circumstances, Company A could be liable for cleanup costs at Company B’s facility because it “arranged for disposal” of hazardous substances there, even if the purpose of the transaction was not the disposal of waste at Company B’s plant.

Rejecting all such interpretations of non-traditional “arranger” liability, and relying on the plain, ordinary meaning of “arrange” — and citing the one appellate opinion that had read the statute that way, see Amcast Industrial Corp. v. Detrex Corp., 2 F.3d 746 (7th Cir. 1993) — the Court squarely imposed an intent requirement for “arranger” liability. “[A]n entity may qualify as an arranger under § 9607(a)(3),” wrote Justice Stevens, “when it takes intentional steps to dispose of a hazardous substance.” Slip op. at 11. While a company’s knowledge that its product will be spilled “may provide evidence” of such intent, “knowledge alone is insufficient to prove that an entity ‘planned for’ the disposal, particularly when the disposal occurs as a peripheral result of the legitimate sale of an unused, useful product.” Slip op. at 12. Thus, even though Shell knew that the chemicals it shipped to Brown & Bryant were being spilled during transfer operations, and even though (and indeed because) it took steps to minimize such spills, Shell could not be held liable as an “arranger.”

Harm May Be Apportioned on the Basis of Estimates and Incomplete Information

The Court next took up the issue in the railroad’s appeal: did the district court appropriately apportion the harm at the site, or was the Ninth Circuit correct in rejecting that decision and imposing joint and several liability? The holding here concerned not the standard to be applied — all parties agreed on that — but in how it should be applied to given facts.

Following the approach established just three years after CERCLA’s enactment by United States v. Chem-Dyne Corp., 572 F. Supp. 802 (S.D. Ohio 1983), federal courts have relied on Section 433A of the Restatement (Second) of Torts to determine whether liability for a given harm should be apportioned among multiple defendants. Under the Restatement approach, if the harm is indivisible, each defendant is subject to liability for the entire harm; but if there is a “reasonable basis for division according to the contribution of each” defendant, then each is subject to liability only for that portion of the harm that defendant caused. Restatement (Second) of Torts § 875. In other words, apportionment of a single harm is proper when “there is a reasonable basis for determining the contribution of each cause.” Id. § 433A(1)(b). Defendants bear the burden of showing a reasonable basis for apportionment. If they are successful, they can drastically reduce their potential liability. Thus, significant dollars can ride on that determination.

What, then, qualifies as a “reasonable basis” for defendants to prevail on apportionment? How much evidence must the record contain? How reliable must that evidence be? The evidence in a typical CERCLA case is far from perfect, and generally characterized by data gaps, estimates, and an incomplete understanding of the links between each defendant’s activities and the precise amount of harm that they caused. In Burlington Northern, the Ninth Circuit had found that such routine shortcomings in the evidentiary record stood in the way of establishing a reasonable basis for apportionment of the harm at the Brown & Bryant site, finding insufficient data to establish the “precise proportion of contamination” that occurred on each portion of the site or the rate of contamination prior to 1975, concluding that neither the duration of the railroads’ lease to Brown & Bryant nor the size of their parcel alone was a “reliable measure of the harm” caused by activities there, and noting that the district court had relied on “estimates rather than specific and detailed records as a basis for its conclusions.” Slip op. at 17.

None of these shortcomings gave Justice Stevens pause. (Nor did the fact that the district court decided to apportion harm without any meaningful help from the railroads, who adopted a “scorched earth, all-or-nothing approach to liability” and failed to acknowledge any responsibility.) “Despite these criticisms,” he wrote, “we conclude that the facts contained in the record reasonably supported the apportionment of liability.” Slip op. at 17. And despite the Ninth Circuit’s faulting of the district court for relying on the “simplest of considerations,” the appellate court itself had acknowledged that the very factors considered by the district court were “relevant to the apportionment analysis.” Slip op. at 17.

On the critical issue of joint and several liability, then, Burlington Northern articulates more of an attitude toward apportionment than a new rule of law. It endorses a flexible, non-exhaustive approach to the question whether there is a “reasonable basis” for apportionment, that does not insist on mathematical precision or exhaustive evidentiary support but instead is free to employ estimates and less-than-complete data and may be based on a subset of all potentially relevant factors.

Conclusion

For some unknown percentage of potential CERCLA defendants — those who shipped chemicals elsewhere without intending that they be disposed of — Burlington Northern is a major victory. Conversely, governmental and private plaintiffs who perform or pay for cleanups at such sites will now have fewer parties to whom they can spread those costs.

Similarly, the Court’s endorsement of a flexible approach to apportionment of harm may well mean that fewer parties will face joint and several liability in CERCLA cases. This approach has the potential to dramatically change the dynamics of CERCLA litigation, as what had become a presumption in favor of joint and several liability is replaced by the very real prospect of apportionment, and the possibility that the defendants with the greatest financial resources may be assigned a relatively small share. The effects will be especially great in cases involving significant “orphan shares” (those attributable to defunct or insolvent parties).