DECISION AND ORDER DENYING RECONSIDERATION AND ADDRESSING INSURANCE SETOFF ISSUE

WILLIAM C. GRIESBACH, Chief Judge.

Before this action may be concluded, two issues remain outstanding. First, Appleton Papers Inc. has filed a motion for reconsideration. Second, NCR and P.H. Glatfelter present an unresolved issue regarding a setoff from Glatfelter's recovery against NCR. I address both of these issues below and direct entry of a final judgment pursuant to Dkt. Nos. 1480 and 1481.

I. Appleton Papers' Motion for Reconsideration

Plaintiff Appleton Papers Inc. (API) has moved for reconsideration of this court's August 2008 order dismissing the claims it brought under CERCLA § 107, 42 U.S.C. § 9607. The delay in bringing the present motion is due to the fact that in API's view, things changed with this court's 2012 ruling that it had no CERCLA liability, a conclusion that means it should now be allowed to press the § 107 claims that had been dismissed in 2008. For the reasons given below, its motion for reconsideration will be denied.

In short, the premise underlying the motion for reconsideration is that the finding of no CERCLA liability means API has no viable § 113 claim for contribution because it does not share common liability with the other PRPs. See 42 U.S.C. § 9613(f). API thus argues that with the § 113 claim out of the picture, it must be allowed to pursue a § 107 claim or else it would find itself with no CERCLA remedy at all. The ability to pursue its claim for recovery of response costs under § 107, as opposed to § 113 would, of course, be a great advantage to API because the equitable considerations that the court held barred NCR's right to contribution play no role in determining a party's right to recover under § 107. See California v. Neville Chemical Co., 358 F.3d 661, 672 (9th Cir.2004) ("Every court of appeals that has considered the precise question whether § 9607 permits equitable defenses has concluded that it does not, as the statutory defenses are exclusive.").

The government and several of the other Defendants argue that API is simply an indemnitor who stands in the shoes of NCR, in essence no different than the numerous insurance companies that owed indemnity payments on behalf of the PRPs. In fact, that was the essence of API's argument for no liability: although API had agreed by contract to indemnify NCR for NCR's liability, it had not agreed to assume CERCLA liability in its own right. It was, in short, nothing but an indemnitor. In finding API not independently liable under CERCLA, this court accepted that logic and concluded that API was indeed little more than NCR's indemnitor.

Courts have held that parties who are merely indemnitors do not incur costs under CERCLA and thus have no independent standing to pursue CERCLA relief under § 107. For example, in a recent Ninth Circuit case the court found that Chubb, an insurer, "has no standing to bring suit under CERCLA section 107(a) because it did not incur any `costs of response' related to the removal or remediation of a polluted site." Chubb Custom Ins. Co. v. Space Systems/Loral, Inc., 710 F.3d 946, 952-53 (9th Cir.2013). There, pursuant to a policy, Chubb paid some $2.4 million to its insured as a reimbursement for the insured's incurrence of cleanup costs on its property. The court found that Chubb had not "incurred" response costs because it was not itself liable, under CERCLA, for the environmental harm. "Chubb only alleges that by virtue of reimbursing Taube — Koret under its Policy, it became subrogated to Taube — Koret's right to pursue a section 107(a) claim. But a subrogee — simply by stepping into the shoes of the insured via a reimbursement — cannot be liable for response costs under CERCLA, and thus cannot itself incur response costs." Id. at 962.

Id.

API does not dispute that as a general principle indemnitors do not have § 107 claims, but it argues that it was much more than a simple indemnitor. API did not merely reimburse response costs paid by NCR; it paid response costs directly. In addition, unlike a typical insurer-indemnitor, API was subject to numerous governmental orders to clean up the Site. Moreover, the government sued API, brought two preliminary injunction motions against it, and strenuously argued (and still apparently believes) that API is liable under CERCLA and that it should be deemed a PRP just like the other Defendants. API also participated in and even controlled the LLC that was performing the cleanup work, and it funded the project as though it were a PRP, not merely an indemnitor. As this court observed in a previous order, "parties on all sides appeared to be operating under the assumption that Appleton Papers Inc. was liable under CERCLA." (ECF No. 193 at 4-5.) This level of involvement stands in sharp contrast to the indemnitor whose obligation obviously arises solely out of an insurance policy. Thus, because API acted and was treated as though it were a liable party for several years, it believes it incurred cleanup costs under CERCLA and did not merely make indemnity payments. As such, it should be allowed to recover those costs through its own § 107 claim.

The fact that API paid response costs directly is not dispositive. Whether it paid the response costs directly or reimbursed NCR for the response costs, the fact remains that API's obligation to pay them arose out of its contract with NCR. It thus follows that API is subject to the same limitations and defenses as NCR. To hold otherwise would allow any insurer or contractual indemnitor to avoid any defense that might apply to its insured or indemnitee by the simple expedient of directly paying the costs for which its insured or indemnitee is liable. However API may have been viewed or viewed itself previously, it has now convinced the court that it's only liability in the case is as NCR's partial indemnitor. As such, it stands in NCR's shoes and enjoys no greater rights than NCR does. Just as NCR may not seek recovery under § 107, neither can API.

This does not leave API without a remedy, to the extent it is entitled to one. API was subrogated to NCR's claim for contribution by virtue of its payments and thus had a subrogation claim. See 42 U.S.C. § 9612(c)(2). In this connection, it should be noted that API asserted a viable § 113 claim in this lawsuit — and lost. Because § 113 was available to it, § 107 was not, a point the Seventh Circuit recently made clear. Bernstein v. Bankert, 702 F.3d 964, 980 (7th Cir.2012) ("[W]e agree with our sister circuits that a plaintiff is limited to a contribution remedy when one is available."). Accordingly, API's motion for reconsideration will be denied.

II. Insurance Setoff

The final remaining issue involves a dispute between P.H. Glatfelter and NCR over Glatfelter's insurance recoveries. In July 2012 I concluded that Glatfelter may recover some $4.3 million from NCR for Glatfelter's payments to conduct natural resource damage assessment and restoration activities. (Although since NCR was not deemed an arranger, Glatfelter could not recover its much larger OU1 expenses.) I had previously ruled, however, that because double recoveries are not permitted under CERCLA, any such recoveries against NCR would be subject to setoff for any amounts Glatfelter had received from its insurer covering the same costs it was seeking from NCR. Glatfelter received substantial amounts in settlements with nine insurers. The proceeds were not payments for specific claims but instead were policy buy-outs or releases that ended each insurer's liability for any past and future costs (including defense costs) relating to the Fox River Site. In other words, each insurer paid Glatfelter a sum of money to eliminate all of the insurer's obligations under the applicable policy. Because these funds were for global releases of all liability rather than for specific claims, the funds were not "earmarked" or otherwise identified as being "for" specific types of claims or expenses. Thus, the central question is what amount of these settlements (if any) were payments for defense costs or future liability (which are not recoverable in contribution and thus not subject to setoff), and which portion was attributable to the expenses Glatfelter now seeks to recover from NCR (which would pose an impermissible double recovery).

In a previous decision, I ruled that the Defendants would not be entitled to contribution for amounts that had been (or would be) covered by insurance. Many cases, including Friedland v. TIC-The Indus. Co., 566 F.3d 1203, 1210 (10th Cir. 2009), stand for the principle that CERCLA contribution actions do not follow the collateral source rule, meaning that windfalls due to double recoveries are not to be distributed to "innocent" parties. Instead, amounts covered by insurance act to reduce the amount that party may recover in contribution. (One objection to this is that some parties become the beneficiaries of insurance purchased by others, which can result in more culpable parties paying less than they otherwise might.)

The more difficult question is how to determine what amounts were actually "covered" by insurance or, to put it another way, what given insurance payments were "for." Money is fungible, and when payment is received for a number of different things (things that do not have price tags or itemized bills of sale), it can be difficult to parse the lump sum payment and assign value to each thing in the group. For example, suppose A brings a lawsuit against B alleging four claims. A and B enter into a settlement for ten thousand dollars, under which B is released from any liability for the four claims already brought as well as any other claims that A might bring in the future. In such a situation, it is difficult to determine how much of the settlement money (if any) was "for" Claim one or two, for example, and how much was "for" the release of any other future liability. The settlement agreement is silent, and so it is impossible to know exactly what value each party placed on each claim or how they arrived at the ten-thousand-dollar figure.

Owing in part to these difficulties, Friedland says that a party seeking to assign value to certain discrete settlement proceeds is out of luck unless the settlement documents themselves provide evidence of that value. Both the district court and the Tenth Circuit in Friedland concluded that an entire insurance settlement was subject to setoff because the settlement did not explicitly allocate funds to anything other than the kinds of expenses that were recoverable in contribution. Relying on that holding, NCR argues that the entirety of the insurance settlements must be subject to setoff because none of them explicitly indicated that they were "for" any expenses other than those that are recoverable in contribution. Both the district court and Tenth Circuit found it crucial that the insurance settlements did not explicitly allocate expenses:

566 F.3d at 1210.

In addition, the Tenth Circuit approved of the district court's exclusion of expert testimony. The district court had rejected the expert testimony as to allocation of settlement amounts because the settlement documents themselves were silent on the question. "Given the silence of the settlement agreements regarding allocation, this testimony is far too speculative to be admissible and consequently, does not create a genuine issue of material fact for trial." Friedland v. The Indus. Co., 2008 WL 185693, at *2 n. 4 (D.Colo.2008).

For obvious reasons, Friedland has been the centerpiece of NCR's setoff argument: applying its holding here would mean that all of Glatfelter's insurance settlements would redound to NCR's benefit by virtue of a setoff because under Friedland the assumption is that the amounts are setoff unless explicitly carved out in the documents themselves. Friedland's holding has an undeniable attraction. It stands for the ostensibly sensible principle that if a party wants to allocate funds in a particular way, it should document that fact at the time of settlement. Doing so will obviate the need to reconstruct the settlements to try to discern intent or purpose long after the fact, an undertaking (as this action has demonstrated) fraught with arcane and opportunistic arguments and battling experts.

But this is an action in equity, which is based at least in part on common sense, fairness, and, at its core, a recognition that in many cases an exact, "scientific" figure will not be attainable. The insistence on documentation in Friedland thus seems out of place, at least in a case like this where everyone concedes that the question of allocation is extremely difficult and when most of the relevant settlements were reached long before this case was even filed. Even if the parties in Friedland had sat down to earmark certain amounts as being "for" defense costs or "for" cleanup expenses, would that even be an accurate reflection of the reality of the situation? Parties to a settlement often have different motives, and it is not hard to imagine that one side might easily accede to the other's preference that most of the expenses be deemed "for" defense costs, for example. In other words, although the attractions of a bright-line rule are obvious, the contemporary allocation of funds called for in Friedland could easily produce artificial results. Accordingly, I am not persuaded that I should mandate a complete setoff simply because the parties to long-ago transactions did not anticipate the need to negotiate and document how they wanted to allocate their settlement funds.

Fundamental to my rejection of Friedland's insistence on documentation is the fact that in this case both sides concede that some of the settlement money is obviously attributable to certain kinds of costs. Even NCR's expert agrees that millions of dollars' worth of settlement proceeds were for attorney's fees and other defense costs — amounts that are not subject to recovery in contribution. As such, denying Glatfelter any benefit from its insurance settlements would produce a result that neither insurance expert believes to be correct. In my view, equity demands more: specifically, it requires that a serious effort be made to reach a reasonable solution, even if that solution is difficult and only an approximation not subject to "proof" in the traditional, scientific (or scientistic) sense. Here, in equity, I conclude that applying Friedland's holding under these circumstances would produce an inequitable result by denying Glatfelter the benefit of its decision to purchase insurance for millions of dollars' worth of expenses that are not subject to setoff. Ultimately, the fact that a figure is difficult to reach does not mean that it should not be attempted.

That said, Friedland is important for its cautionary value. The effort to reconstruct settlements and determine allocations years after the settlements were reached can easily become bogged down by twenty-twenty hindsight arguments tailored to specific outcomes. Thus, when there is little contemporaneous evidence of allocation, a court should be wary of ex post facto attempts (on both sides) to earmark funds. Although I agree with the Friedland court that expert testimony on the subject could be too speculative, I am satisfied that the highly competent expert testimony in evidence here is useful to the extent it does not attempt to opine about the subjective intent of the parties entering into the settlements. Subjective intent is outside the experts' area of expertise and is inherently so speculative as to be irrelevant. Instead, I found the expert testimony helpful in identifying what is typical of environmental settlements generally; it was also helpful in explaining what would have made economic sense in each of the settlements at issue here. Since all sides seem to agree that an exact figure is not attainable, the above considerations provide a benchmark and a framework for arriving at a solution that approximates the reality of the settlements and achieves an equitable result.

Glatfelter's expert, Ross Mishkin, had years of experience in analyzing insurance issues relating to mass tort claims, including long-tail asbestos and environmental liability matters. In his view, none of the insurance proceeds were subject to setoff because Glatfelter has not yet been made whole. Glatfelter has paid some $44 million required by the OU1 consent decree. Its settlements with insurers total less than the $44 million it has paid due to its CERCLA liability (it has disclosed the exact figure to the court under seal). Because its settlement amount is lower than its overall liability expenditures, there has been no double recovery. And the $4.2 million it now seeks from NCR will not bring its recoveries over the $44 million level either.

NCR argues that things are more complicated than that. For example, the policy issued by the insurer known as Carrier X had a property damage limit as well as an unlimited obligation to pay for Glatfelter's defense. Carrier X settled with Glatfelter for a large amount that was a multiple of the policy's liability limit, presumably because the costs of defense that Carrier X owed were expected to be much higher than the actual liability limit. That means that although most of the Carrier X settlement would not be subject to setoff (because those funds are not "for" Glatfelter's liability), the amount received to extinguish the policy limit would arguably be a payment "for" Glatfelter's liability. And that amount would be subject to setoff.

Glatfelter's response is that we can calculate a maximum possible amount that it received to indemnify it for its liability. Just looking at the Carrier X settlement, we know that a substantial portion of that payment was for defense costs and only a small portion was for Glatfelter's liability. We know this because the policy limit provides a demonstrable upper bound for what the liability portion of the settlement payment would be. Suppose that Glatfelter received $30 million from an insurer known as Carrier Z even though its policy limit was only $5 million. We could say that the $5 million was clearly on account of its liability, but the $25 million clearly was not. We only "count" the $5 million for liability in calculating whether Glatfelter has been made whole for its liability obligations. Looked at another way, we would subtract the $25 million not for liability in calculating how much might be subject to setoff.

Here, subtracting the much larger non-liability portion of its Carrier X settlement from Glatfelter's total insurance recoveries gets us to a figure that is substantially below the $44.5 million amount it has paid for OU1 cleanup costs. That is, if we add up all the insurance settlements that could possibly be attributable Glatfelter's liability, that figure would be less than $44.5 million. This means there has been no double recovery yet, and there will not be one even if the $4.2 million Glatfelter now seeks is added to those totals. In other words, the amount Glatfelter has spent to extinguish its liability exceeds the amount of insurance recoveries it has received on account of that same liability by more than $4.2 million. This means there is no double recovery in any sense of the word.

I am satisfied that Glatfelter's approach is sound. Things were different during the arranger liability trial and briefing. Then, the setoff discussion was premised largely on Glatfelter being able to recoup its extensive OU1 expenses. In such a scenario, its recovery would have been much greater and a setoff would have made more sense because otherwise it would have been recovering from NCR and its own insurers for the same liability expenses. What's changed is that, in finding that Glatfelter could not recover its OU1 expenses, the amount it is actually out-of-pocket is dramatically higher. We are no longer analyzing the setoff issue with the assumption that Glatfelter would receive its millions in OU1 expenses from NCR; instead, given that it has not been made whole for these expenses, it would not make sense to redistribute any of its insurance recoveries to NCR in the form of a setoff. There is, in short, no way to look at this problem and conclude that Glatfelter has yet been made whole for the liability it has incurred.

Accordingly, I conclude that Glatfelter is entitled to judgment against NCR for the entire $4,283,114.13.

III. Conclusion and Order for Judgment

For the reasons given above, API's motion for reconsideration is DENIED. I further conclude that no setoff is appropriate and that Glatfelter is entitled to judgment against NCR in the amount of $4,283,114.13.

The parties have submitted a status report (ECF No. 1480) detailing the issues remaining in this action. This order appears to resolve all remaining issues. Accordingly, I will direct entry of judgment as set forth in the parties' proposed final judgment (ECF No. 1481-1), as modified by this order.