Law and the Environment : Environmental Lawyer & Attorney : Foley Hoag Law Firm : Boston, Washington D.C.

My partner Robby Sanoff blogged last week about the “Illusion of Finality in CERCLA.” His post addressed City of Emeryville v. Sherwin-Williams, in which the 9^th Circuit Court of Appeals held that a person who was not a party to a prior settlement could bring a contribution claim against such a settling party, at least where the new claim involved contamination at a downgradient property, rather than the property that was the subject of the first settlement.

City of Emeryville seems to have been rightly decided. The only real lesson it teaches is that a PRP who wants to settle and achieve finality must do everything possible to ensure that all potentially liable parties are brought to the table and that the site is properly defined.

However, a case from earlier this summer just came to my attention. Its lesson is different. Its lesson is that CERCLA is just plain nuts, and judges’ efforts to harmonize its decisions can do more harm than good. Ashland v. Gar Electroplating teaches that there is no finality under CERCLA and that parties who step up to the plate to settle with the government and perform cleanups can be subject to joint and several cost recovery claims brought by recalcitrants. Alice in Wonderland, here we come.

Although Ashland seems wrong, Judge Lisi’s decision is a not surprising consequence of the simplistic CERCLA jurisprudence announced by Justice Thomas in Aviall and Atlantic Research. In holding that contribution claims are limited to those who have been the subject of litigation or certain kinds of settlements with the United States or a state, and that PRPs not subject to suits or administrative orders can bring cost recovery actions under § 107 of CERCLA, the Supreme Court has opened a Pandora’s Box.

What evil genie was loosed in Ashland? The facts are too complicated for a complete recital, but some history is required. Here goes:

1.                   The government sued a number of parties, including United Technologies, under CERCLA. The claims related to the Davis Liquid Waste Site, in Smithfield, Rhode Island.

2.                   The defendants brought third-party claims against a number of other PRPs, including Ashland.

3.                   Some original defendants (including this firm’s client, Ciba-Geigy) settled.

4.                   The United States went to trial against UT and obtained a finding of liability. UTC then settled with the government and agreed to perform a portion of the remedy. It also agreed to allocate sums recovered from the third-party defendants with the government.

5.                   A number of third-party defendants, not including Ashland, settled with UTC – and the government. 

6.                   The case against Ashland (and some other non-settlors) went to trial. Ashland was found liable. The judge allocated liability and imposed a percentage share on Ashland. 

7.                   At some point, EPA directly notified Ashland that it was a PRP and sent Ashland an administrative order on consent, requiring it to perform certain groundwater cleanup.

8.                   It is not clear if Ashland signed the AOC, but, according to an EPA affidavit, Ashland agreed to perform certain cleanup work.

9.                   Ashland spent approximately $2 million performing the work.

10.               It then brought claims under § 107 of CERCLA.  It had no contribution claim, because nothing had happened to bring Ashland under the ambit of § 113.

11.               The United States and the State of Rhode Island both filed amicus briefs on behalf of those sued by Ashland, arguing that their respective settlements with UTC barred Ashland’s claims.

Faced with these facts, what did Judge Lisi do? First, she found that Ashland does have a right to bring claims under § 107. Because it was not sued by the United States or Rhode Island, Ashland was not limited to contribution claims.

Second, Judge Lisi found that the allocation previously reached in the private action was not applicable to this case, precisely because the private case was for contribution and this case is a claim under § 107, which is assumedly one for joint and several liability. While Judge Lisi noted that the defendants in this action can bring counterclaims for contribution, the burden is on them to prove that the harm is divisible. 

As someone might say, OMG. Here’s what Justice Thomas has done to us. The recalcitrant party has a joint and several claim against those who settled with the government, and the burden is on the good guys to prove that liability should not be joint and several. 

If I may be permitted to gloat, I raised this specter following Aviall and Atlantic Research, and was generally told not to worry; the district courts would find ways to reach practical results. Plan B would seem to be the 1^st Circuit Court of Appeals. Plan C? I can’t wait to see what Justice Thomas would make of this fine mess.

• Email This
• Print
• Comments (1)
• Trackbacks
• Share Link

When the Burlington Northern decision was first announced, I concluded that “never has the Supreme Court done so much by doing so little.” On May 5, Judge John Mendez, of he Eastern District of California, proved me at least half right. In United States v. Iron Mountain Mines, joint and several liability was imposed on the defendants in 2002. The 2002 decision stated that “given the nature of pollution at the site, it would be difficult to identify distinct harms.” The court did not analyze whether there was a reasonable basis for apportionment of liability. 

Following the Burlington Northern decision, the defendants moved for reconsideration, arguing that Burlington Northern constituted an intervening change in the law. Defendants argued that “the Supreme Court clearly meant to send a signal to other courts that they must begin evaluating apportionment in a different way.” I think that the defendants in Iron Mountain were right.  Unfortunately, that’s not the standard for a motion for reconsideration.   If Iron Mountain were being decided for the first time today, the defendants might get a better result, but that doesn't mean that they win their motion for reconsideration.

What the Supreme Court really said in Burlington Northern isn’t that the law was wrong; it is that District Courts weren’t applying the law correctly. District court judges had their collective judicial thumbs firmly on the side of the government. The Supreme Court simply told the lower courts to take those thumbs off the scales. I hope that this decision will not encourage lower courts to keep the thumbs on the scales.

• Email This
• Print
• Comments
• Trackbacks
• Share Link

One more note on the Burlington Northern decision.  A client of mine has already noted that one impact of the decision will be to result in more litigation over divisibility, which will be good for private lawyers (ouch!).  She’s right, as my clients always are, but she shouldn’t be.

Litigation should only increase if EPA does not adjust its settlement demands. If EPA responds appropriately, and makes demands which reflect a fair resolution of a divisible liability, then there shouldn’t necessarily be more litigation than there is today.  However, if EPA continues to negotiate with PRPs as though liability is always joint and several, then there will certainly be more litigation – and EPA will start to lose some more cases.  

Anyone care to bet which response by EPA is more likely?

If EPA adjusts its settlement demands downward in response to the decision – or if we start to see litigation in which courts find liability to be divisible – then EPA’s ability to fund Superfund cleanups will be under even more pressure.  This could provide some additional momentum behind the current Congressional effort to reinstitute the Superfund tax.

• Email This
• Print
• Comments
• Trackbacks
• Share Link

Those of us who have practiced in the Superfund arena for some time know that the government, in those rare cases where it has been forced to litigate, has used the same oral argument in every case: “Good morning, your honor. My name is ______. I represent the government in this action and we win.” Today, the Supreme Court made clear that that the government now needs a new oral argument template.

In Burlington Northern v. United States, the Supreme Court issued two important decisions in one. First, the Court held that a defendant must actually intend its waste to be disposed of before it can be found liable as an arranger under § 107(a)(3) of CERCLA. The facts were these. Shell Oil sold pesticides to Brown & Bryant, Inc., which operated a chemical distribution business. As part of the transfer of pesticides from Shell to B&B, some pesticides were released on the property. There was evidence that Shell knew that releases were a regular part of the transfer process. Both the District Court and Appeals Court concluded that Shell’s knowledge that releases occurred was enough to establish arranger liability.

Noting that CERCLA does not define the term “arrange[e] for”, the Court looked the phrase’s ordinary meaning. Doing so, the Court concluded that liability may attach only where the defendant “takes intentional steps to dispose of a hazardous substance.” The government argued that, because the defendant knew that disposal was the inevitable result of its sale of product to the site owner, the defendant had “intended” disposal to occur. The Court rejected this argument. The Court was very clear: The defendant “must have entered into the sale of [the product] with the intention that at least a portion of the product be disposed during the transfer process.”

The direct holding with respect to Shell will be important in a number of cases and is helpful in setting a fairly bright line on arranger liability. Even beyond the immediate holding, however, I wonder what, if anything, this case means for what is known as transshipment liability. Under section 107(a)(3), a person is liable as an arranger if they

arranged for disposal or treatment … of hazardous substance owned or possessed by such person, by any other party or entity, at any facility or incineration vessel owned or operated by another party…

It has always seemed to me that the plain reading of § 107(a)(3) is that the defendant must have “arranged” for the disposal of the hazardous substances at the site where disposal occurred. In those not uncommon situations where the site operator transshipped the waste – without the generator’s knowledge or consent – the generator should not be liable under CERCLA at the transshipment site, because it did not intend for any disposal at the transshipment site. Given the Supreme Court’s emphasis on what the generator intended, I think that, in the right case, a transshipment generator defendant would stand a pretty good chance of winning, if he or she were willing to litigate the case all the way up to the Supreme Court. 

I hope someone will and I hope I’m right. 

The second holding in Burlington Northern may be of even more practical significance. In it, the Court reversed the Court of Appeals and upheld the District Court’s original divisibility finding with respect to the Burlington Northern Railroad. The District Court used a simple formula based on percentage of the site owned by Burlington Northern and the percentage of time that Burlington Northern leased the land as compared to the total duration of site operations. What’s most significant is that the Court did not even require any significant analysis to uphold the District Court; Justice Stevens’ opinion merely stated that there was evidence that contribution from the railroad parcel to the overall contribution was limited, so that, “[w]ith these background facts in mind, we are persuaded that it was reasonable for the court to use the size of the leased parcel and the duration of the lease as the starting point for its analysis.”

This seems obvious, but is probably a game changer in government Superfund litigation. The overwhelming tenor of lower court opinions has been that the defendant’s burden in a divisibility argument is almost overwhelming and that the burden will be satisfied in the rarest of cases and only upon almost perfect evidence of divisibility. The Supreme Court has made clear that that is simply not the case. Superfund cases are no different than other cases and there is no unstated higher burden of proof. 

Thus, while a district court judge might still be affirmed if he or she concludes that the defendant did not meet its burden of proving divisibility, the real import of the decision is that now district court judges need not fear that they will be automatically reversed if they do conclude that the harm is divisible. Given the standard stated in Burlington Northern, it might go too far to say that most cases will be divisible, but divisibility findings should not be at all rare – and that’s definitely news.

• Email This
• Print
• Comments
• Trackbacks
• Share Link