United States of America et al v. NCR Corporation et al

Filing 1205

DECISION AND ORDER APPROVING REVISED CONSENT DECREE signed by Chief Judge William C Griesbach on 8/22/2017 denying as moot 1125 Motion for Partial Summary Judgment; denying as moot 1134 Motion for Summary Judgment; denying as moot 1141 Moti on for Partial Summary Judgment; denying as moot 1171 Motion for Reconsideration ; granting 1188 Motion for Settlement; granting 1202 Motion for Leave to File. The parties should file a status report concerning the remaining claims in both this action and the Whiting case within 14 days of the date of this order. (This opinion may be subject to further editing) (cc: all counsel) (Griesbach, William)

UNITED STATES DISTRICT COURT
EASTERN DISTRICT OF WISCONSIN
UNITED STATES OF AMERICA and
THE STATE OF WISCONSIN,
Plaintiffs,
v.

Case No. 10-C-910

NCR CORPORATION, et al.,
Defendants.

DECISION AND ORDER APPROVING REVISED CONSENT DECREE

The United States and State of Wisconsin lodged a proposed consent decree they negotiated
with NCR Corporation and Appvion, Inc., with the Court and then published notice of the proposed
settlement in the Federal Register. 82 Fed. Reg. 7862 (Jan. 23, 2017). In the governments’ view,
the proposed consent decree, as recently revised, will settle the overwhelming majority of
outstanding issues relating to the cleanup of the Lower Fox River. Following a public comment
period, the plaintiff governments, joined by NCR and Appvion, have now moved for court approval
of the settlement. Only one of the remaining defendants, P.H. Glatfelter Company, opposes
approval of the revised consent decree. For the reasons given below, the motion will be granted.
I. BACKGROUND
This case is one of two cases before the Court involving the effort to remediate the pollution
of the Lower Fox River caused by the discharge of polychlorinated biphenyls (PCBs), a highly toxic
organic compound which was used from 1954 to 1971 in the production of carbonless copy paper
(CCP), a product developed and sold exclusively by NCR Corporation. The cleanup, which is

ongoing, was anticipated to be the largest environmental dredging project ever undertaken. The
first case, the Whiting case or contribution action (Case No. 08-C-016), was commenced by NCR
and Appleton Papers, Inc., (now Appvion, Inc.) in 2008 in an effort to force other entities the EPA
had identified as potentially responsible parties (PRPs) to contribute to the cost of the clean-up.
This second action, the enforcement action, was commenced by the United States and the State of
Wisconsin against NPR, Appleton Papers and other PRPs to enforce the EPA’s Administrative
Order for Remedial Action after work on the clean-up had slowed in response to certain decisions
the Court had rendered in the contribution action. The following facts, taken largely from the
Court’s December 16, 2009 decision in the contribution action dismissing on summary judgment
the plaintiffs’ contribution claims, Whiting, Dkt. 795 at 1–7; 2009 WL 5064049, at *1–11, provide
the context in which the current dispute between the parties arises.
In 1954 the National Cash Register Corporation introduced “No Carbon Required” copy
paper, the name being a play on NCR's corporate name. A key component of NCR paper, also
known as carbonless copy paper (CCP), was an emulsion (a mixture of liquids that do not mix)
containing Aroclor 1242, a solvent manufactured by the Monsanto Corporation. Aroclor 1242 is
a type of polychlorinated biphenyl (PCB), a stable compound that does not easily degrade. PCBs
have been linked to illness and death in fish, birds and other wildlife, and studies suggest that their
persistence in the environment has caused numerous health problems, including cancer, in humans
who have consumed fish or had other contact with PCBs.
Although NCR developed and sold its carbonless paper product and created the
PCB-containing emulsion, the paper was actually manufactured by the Appleton Coated Paper
Company, which coated sheets of paper with NCR's emulsion. This manufacturing process resulted

2

in significant discharges of PCBs into the Fox River. But PCBs also escaped into the river in other
ways. The Lower Fox River area boasts the largest concentration of paper mills in the world, and
prior to 1972 many of these mills recycled NCR-brand carbonless copy paper. Some purchased the
waste paper and trimmings called “broke” that inevitably resulted from the CCP manufacturing
process from NCR/Appleton Coated Paper and used it to manufacture their own paper products.
As CCP became commonly used throughout the business world, and the paper recycling industry
grew, CCP also became a significant part of the recycled paper that the paper mills purchased on
the after-use market. The recycling and de-inking processes several of the mills developed to
prepare the broke and recycled paper for use in their own manufacturing process likewise led to the
release of PCBs into the river. As a result, the EPA named NCR and Appleton Papers, both
successors to Appleton Coated Papers; the paper mills; and various other entities, such as cities,
utilities and sewerage districts that treated and/or released wastewater containing PCBs from NCR’s
CCP and/or broke into the river as PRPs for the clean-up. Finally, even after NCR stopped using
PCBs in its emulsion in 1971, stocks of pre-1971 paper still remained in offices for years. As this
paper was used up, it was sent for recycling along with other paper, and the recycling process led
to smaller quantities of PCBs releasing into the river throughout the 1970s and possibly later.
In November 2007, the EPA issued a unilateral administrative order pursuant to 42 U.S.C.
§ 9606(a) requiring that NCR, Appleton Papers, and six other respondents conduct certain remedial
actions in the downriver part of the Lower Fox River. The EPA’s order followed years of activity
by various PRPs to address the Lower Fox River contamination. Some of the activity was pursuant
to Consent Decrees the EPA and the Wisconsin Department of Natural Resources (WDNR) reached
with various PRPs and much of it was the result of voluntary cooperation among and between the

3

parties. With the bulk of the cleanup work yet to be done and the parties unable to reach agreement
on how the costs of the remediation were to be allocated, NCR and Appleton Paper commenced the
contribution action “to put in place a procedure for determining each unsettled PRP’s allocable
share of the cleanup costs and natural resource damages.” Whiting, Dkt. 9 at 2. Though initially
naming only George Whiting Company as a defendant because of its refusal to execute a tolling and
standstill agreement, the case expanded through a series of amended complaints. The plaintiffs
eventually brought more than twenty additional parties into the suit, including the various paper
companies along the 39-mile stretch of the River who had allegedly discharged PCBs in the course
of their recycling processes, various municipalities and/or their sewage treatment plants that had
discharged PCBs in treated waste water from the paper mills into the River and perhaps other
sources, and the City of Green Bay and the Army Corps of Engineers whose involvement consisted
of dredging the bottom of the portion of the River within the City limits to allow large ships access
to the industrial plants and warehouses along the River and in the process moving PCBs from one
area of the River to another.
Early motion practice in the case resulted in the dismissal of the plaintiffs’ cost recovery
claim, leaving only a contribution action for past and future cleanup costs and allocation for natural
resource damages. The parties then met in an effort to reach agreement on a case management plan
for what all believed would be a massive undertaking. This was especially true if the key focus
became the amount and location of PCBs each of the parties had discharged into the Lower Fox
River from 1954, when NCR first introduced its CCP, through 1971, when it ceased using PCBs
in the manufacturing process, and beyond as the then existing stock cycled through the economy.
As the Court noted in a case management order entered at that time, “[a]ll parties agree that full

4

preparation of the case for trial would require extensive document discovery, deposition of several
hundred witnesses, and retention of experts at substantial expense to each party. Trial of all issues
at once could take months.” Whiting, Dkt. 252 at 1.
A number of the parties the plaintiffs brought into the case had not been named as
respondents in the EPA’s November 2007 Unilateral Administrative Order (“UAO”) for Remedial
Action (the non-UAO parties) and, viewing their own PCB contributions as minimal to nonexistent, argued they did not belong in the lawsuit. They requested and were granted a 4-month
stay during which they would attempt to work out settlements with the EPA that would bar the
plaintiffs’ contribution claims and thereby allow them to avoid the massive litigation costs required
for a defense. The remaining defendants—primarily the paper companies—did not deny that they
belonged in the lawsuit but argued that discovery should focus first on two issues: (1) when each
party knew, or should have known, that recycling NCR-brand carbonless paper would result in the
discharge of PCBs to a waterway, thereby risking environmental damages; and (2) what, if any,
action each party took upon acquiring such knowledge to avoid the risk of further PCB
contamination. Id. at 2–3. These defendants argued that resolution of these issues first would
materially improve the likelihood that the parties would be able to settle most if not all of the case
without the protracted discovery and trial that would otherwise be required.
NCR and Appleton Paper strongly opposed the defendants’ suggestion, noting that
knowledge and action or inaction by a party are only two of the multiple factors a court should
consider in apportioning liability under CERCLA. The plaintiffs argued that the case would more
likely settle with a firm trial date and discovery proceeding on all issues at once. The plaintiffs also
argued that staying discovery as to certain issues would be unfair and prejudicial because it would

5

risk the further loss of evidence, as key witnesses to events that transpired over three decades ago
pass away, and unduly emphasize only two of the myriad of factors the Court must ultimately
consider.
In a decision that was intended to further the goal of arriving at a “just, speedy, and
inexpensive determination of [the] action,” Fed. R. Civ. P. 1, but that in hindsight is difficult not
to question, the Court adopted the defendants’ recommendation, explaining:
Having considered the arguments of counsel, I now conclude that the defendants’
proposal, as modified, is the more efficient and just way to proceed with this action.
I am satisfied, without prejudging the issue, that knowledge and fault could
constitute the principal bases on which contribution may be allocated. See
Environmental Transp. Systems, Inc. v. ENSCO, Inc., 969 F.2d 503, 509 (7th Cir.
1992). The cases make clear that there are no hard and fast rules for an equitable
claim based on contribution, but certainly fairness would seem to dictate that those
who were in the best position to know about possible contamination (or those who
did know) should bear more responsibility than parties further down the stream (so
to speak). Indeed, under some circumstances, joint tortfeasors having substantially
greater culpability can be denied contribution altogether. See Restatement (Second)
of Torts § 886B (1979). If a ruling can be reached on this issue in advance of
proceeding with the rest of the case, it could obviate the need for numerous parties
to engage in extensive discovery, the cost of which could exceed several parties’
actual liability. Accordingly, defendants’ proposal, as modified to include
consideration of the knowledge and actions of the defendants, and to allow
sufficient time for the court to decide potential motions for summary judgment on
the issues raised, will be adopted herein.
Id. at 4.
The parties undertook discovery in accordance with this plan and then filed motions for
summary judgment. In a decision dated December 16, 2009, the Court concluded from the
undisputed evidence before it that while none of the defendants knew they were discharging PCBs
into the Fox River during the 1960s, “by the late 1960s, NCR had access to the vanguard of data
suggesting an appreciable risk of serious and long-lasting environmental damage resulting from the
production and recycling of NCR paper.” Whiting, Dkt. 795 at 26, 2009 WL 50640949, at *14.
6

This fact, combined with the fact that NCR actually increased production in the final years as the
risk should have been more apparent, while the defendants neither knew of the presence of PCBs
nor were in a position to know of their presence and the environmental risk posed by PCBs, led the
Court to conclude that the plaintiffs’ contribution claims against the other defendants were barred.
Whiting, Dkt. 795 at 27–46; 2009 WL 5064049, at *14–25.
It would be an understatement to say that this decision did not induce the parties to settle
the case. NCR and Appleton Papers, the latter of which was later determined to have no direct
liability due to the terms of the asset purchase agreement by which it acquired the coating plant
from NCR, were convinced that the Court had erred in its equitable apportionment, and slowed and
eventually ceased what had been their voluntary clean-up measures. The remaining defendants,
armed with what they may have viewed as a shield to further liability and a vehicle for recovering
some or all of what they had already spent on the clean-up, were even less receptive to NCR’s
insistence that they contribute.
While the dispute continued between the PRPs, the EPA and the WDNR commenced this
action to enforce the EPA’s UAO against twelve parties it had identified as PRPs to insure that the
clean-up continued. Although the plaintiffs’ complaint asserted six separate claims, including
claims for recovery of past and future response costs of the federal and state governments, and for
natural resources damages (NRD), the governments sought immediate relief on only their fifth
claim for relief which sought enforcement of the November 2007 UAO. On April 27, 2012, the
Court entered an order granting the EPA’s motion for a preliminary injunction requiring NCR to
continue the clean-up work called for by the remedial action plan. Dkt. 370, 2012 WL 1490200.
That order was later affirmed on appeal. United States v. NCR Corporation, 688 F.3d 833 (7th Cir.
2012).
7

In the meantime, discovery and further motion practice in the contribution action continued,
revolving around such issues as the defendants’ contribution counterclaims, insurance set-offs, and
NCR’s potential liability for upstream discharges as an arranger under 42 U.S.C. § 9607, the last
of which issues resulted in a seven-day court trial. Whiting, Dkt. 1405, 2012 WL 2704920. The
cases continued to proceed parallel to each other with the principal focus in the enforcement action
on (1) the propriety of the remedy called for in the EPA’s 2007 amendment of the record of
decision (ROD) and (2) whether the harm to the River was divisible and capable of apportionment.
Following years of investigation, field studies and analysis, the EPA, in partnership with
the WDNR, had selected dredging and removal of PCBs from the riverbed as the default remedial
method, but allowing for capping and sand covering where certain design criteria were met. The
ROD set a remedial action level (RAL) of one part per million (1 ppm), meaning that remediation
was required in all areas of the River where the concentration of PCBs exceeded 1 ppm. The Court
ultimately concluded on summary judgment that the EPA’s decision selecting a combination of
dredging and capping was not arbitrary and capricious, a decision that the Seventh Circuit affirmed
on appeal. United States v. NCR Corp., 911 F. Supp. 2d 767 (E.D. Wis. 2012), aff’d., 768 F.3d 662,
(7th Cir. 2014). The question of whether the harm was divisible and thus capable of apportionment,
however, became the primary defense to the government’s enforcement action of NCR and, later,
P.H. Glatfelter Company, the furthest upstream of the paper companies. It was also the central
issue of an eleven-day bench trial.
Divisibility had taken on fresh significance as a defense following the Supreme Court’s 2009
decision in Burlington Northern & Santa Fe Railroad Company v. United States, 556 U.S. 599
(2009). Section 107(a) of CERCLA imposes strict liability for contamination on the owner of any

8

facility that discharges hazardous substances into the environment, and where owners of two or
more facilities discharge hazardous substances into a body of water such as a river, they are jointly
and severally liable for the entire harm unless they can show that a reasonable basis for
apportionment of the harm exists. Where a reasonable basis for apportionment can be shown,
however, each party is only liable for that portion of the harm that is attributable to that party and
the equitable factors governing contribution claims play no role. This rule is consistent with the
Restatement of the Law of Torts: “Where the tortious conduct of two or more actors has combined
to bring about harm to the plaintiff, and one or more of the actors seeks to limit his liability on the
ground that the harm is capable of apportionment among them, the burden of proof as to the
apportionment is upon each such actor.” Restatement (Second) of Torts, § 433B(2). The reason for
the rule placing the burden of proof as to apportionment on the defendants is to avoid the injustice
of allowing multiple defendants who have combined to cause the plaintiff harm to further burden
the plaintiff by requiring that he or she present such evidence or, even worse, allow the defendants
to escape liability where the nature of the harm makes apportionment difficult or impossible. “As
between the proved tortfeasor who has clearly caused some harm, and the entirely innocent plaintiff,
any hardship due to lack of evidence as to the extent of the harm caused should fall upon the
former.” Id., cmt. d; United States v. NCR Corp., 960 F. Supp. 2d 793, 796–97 (E.D. Wis. 2013).
Prior to Burlington Northern, exceptions to joint and several CERCLA liability were
considered rare and the EPA could usually recover its costs in full from any responsible party,
regardless of that party's relative fault. Metro. Water Reclamation Dist. of Greater Chicago v. N.
Am. Galvanizing & Coatings, Inc., 473 F.3d 824, 827 n.3 (7th Cir. 2007) (“The only exception to
joint liability is when the harm is divisible, but this is a rare scenario.”). Thus, in cases where there

9

were multiple PRPs, they were left to resolve their differences over how much each should pay in
contribution actions if they could not otherwise reach agreement either with the government or
among themselves. That was what NCR was attempting to accomplish here in filing its contribution
action. But in Burlington Northern, decided the year after NCR commenced its contribution action,
the Court upheld a district court finding that the chemical contamination of soil and ground water
on land where a defunct agricultural chemical distribution business had operated was divisible. The
district court in that case had apportioned the liability of the two solvent PRPs: Burlington Northern
Railroad, which had leased a portion of land where spills had occurred to the defunct distributor,
and Shell Oil Company, which had sold one of the chemicals that over time had been spilled onto
the ground. The district court calculated the railroad’s liability at 9% and Shell’s at 6%, thereby
leaving the remaining 85% of the government’s response costs unrecoverable. 556 U.S. at 602–07.
On appeal, the Supreme Court absolved Shell of any liability and upheld the district court’s finding
that the railroad’s liability was limited to 9% of the total clean-up costs, based on the length of time
the railroad had leased its property to the distributor, its estimate of the percentage of spills that
occurred on its property, and a calculation error of up to 50%. Id. at 616–17.
Given the Supreme Court’s decision upholding the district court’s finding that the
contaminated soil and ground water in Burlington constituted divisible harm, NCR focused its
efforts in the enforcement action on establishing just such a defense here in an effort to substantially
limit its ultimate liability. To do so, NCR undertook the very type of detailed and costly discovery
and expert analysis the Court’s initial case management plan was intended to avoid, and the other
parties responded in kind. Experts, along with technicians working under their supervision, were
retained to carefully map the part of the River designated Operating Unit (OU) 4, from the De Pere

10

dam to its termination into the lake waters of Green Bay and analyze sediment samples collected
and analyzed, along with historical documents relating to recycling and production of CCP and
other paper products during the fifteen-year production period that had ended some thirty-five years
earlier. After one expert arrived at an estimate of the percentage of the total PCBs each PRP
discharged into the River over the past fifty-five years, another expert input that data into a
fate-and-transport computer model designed to simulate the movements of sediment and PCB
transport in the river. A third expert then took the conclusions arrived at by the first two experts
and linked those with actual remediation costs within OU4, divided that section of the river up into
73 apportionment polygons, and arrived at a calculation of each party's share of the remediation cost
in each polygon.
Over the course of a two-week trial in December 2012, NCR offered the testimony of these
experts and other witnesses, and the government and remaining defendants countered with experts
and analysis of their own. On May 1, 2013, after full briefing on the issues, the Court issued its
written findings of fact and conclusions of law in a decision holding that NCR and Glatfelter had
failed to prove that the harm caused by the discharge of PCBs into the river was divisible and, thus,
all of the responsible parties were jointly and severally liable. United States v. NCR Corp., 960 F.
Supp. 2d 793 (E.D. Wis. 2013). That same day, the Court entered its order for declaratory and
injunctive relief in favor of the government. In the meantime, having resolved the counterclaims
and insurance set-off issues in the contribution action, the Court entered final judgment in that case
on June 27, 2013. Whiting, Dkt. 1504. Both decisions were appealed.
On September 25, 2014, the Court of Appeals issued decisions in both cases partially
affirming, vacating, reversing and remanding the Court’s judgments. United States v. P. H.

11

Glatfelter, 768 F.3d 662 (7th Cir. 2014); NCR Corp. v. George Whiting Paper, 768 F.3d 682 (7th
Cir. 2014). In the Whiting contribution case, as pertinent here, the Court of Appeals held that the
Court abused its discretion in limiting discovery to knowledge and culpability and then determining
that NCR had primary liability and dismissing its contribution claims upon consideration of that
factor alone. 768 F.3d at 700–03. In the enforcement action, again as pertinent here, the Court of
Appeals held that the Court had failed to explain why the mass-percentage estimates of Dr. John
Wolfe, one of Georgia-Pacific’s experts that NCR used in its analysis, were unreliable, and
therefore vacated the Court’s finding that the harm was not divisible. Both cases were remanded
for further proceedings consistent with the decisions.
Following remand, the Court initially entered an order analyzing the evidence bearing on
NCR’s use of Dr. Wolfe’s estimates and concluding that NCR had established that the harm was
divisible. The Court later vacated that decision on reconsideration, however, and reaffirmed its
original finding that liability remained joint and several. Dkt. 1000, 1033. NCR then moved for
certification of an interlocutory appeal, which the Court denied by order dated January 25, 2016.
Dkt. 1051. After further consultation with the parties, the Court entered a new scheduling order on
February 29, 2016, setting both cases for full trial commencing on March 27, 2017, later changed
to May 8, 2017. Dkt. 1058, 1152.
On January 17, 2017, the governments filed a notice of settlement advising the Court that
they had entered into a proposed consent decree with NCR and Appvion that if approved would
substantially simplify the enforcement action and put an end to most if not all of the litigation in the
contribution case. Because entry of the consent decree, which could only follow a public comment
period, would significantly change the contours of the two actions, NCR, Appvion and the plaintiffs

12

asked for an immediate stay of the litigation in both cases and an adjournment of the trial. The
Court granted the request for a stay and cancelled the pretrial and trial on January 20, 2017.
On March 29, 2017, after expiration of the comment period and further consultation with
the parties, the governments filed their joint motion for Court approval of a revised consent decree
(hereinafter “the Proposed Consent Decree”) under which NCR would take on sole responsibility
for completing all remaining work at the Lower Fox River and Green Bay Superfund Site at an
estimated cost of over $200 million over the next two to three years, over and above the
approximately $668 million that NCR and Appvion have already paid for response actions and
natural resource restoration at the site. The Proposed Consent Decree includes a waiver of claims
by NCR and Appvion, including Appvion’s cost recovery claim against Georgia-Pacific and
Glatfelter for up to $200 million, and a corresponding release of potential claims of their
indemnitors. The agreement sharply constrains the ability of NCR and Appvion to seek reallocation
of their own costs to Georgia-Pacific and Glatfelter, provided Georgia-Pacific and Glatfelter do not
try to strike back against NCR or Appvion. NCR and Appvion also agree to waive their rights to
appeal prior orders in the case and in the related contribution action.
The agreement envisions that Georgia-Pacific and Glatfelter will bear primary responsibility
for long-term monitoring and maintenance activities of the sediment caps installed under the UAO
and prior orders of the Court, at estimated future costs of up to $40 million, with NCR providing
the governments a “back-up” commitment for that work. In addition, if the Proposed Consent
Decree is approved, Glatfelter will be the sole remaining defendant in the litigation of the
governments’ claims for recovery of their own past costs incurred in connection with the clean-up,
approximately $34 million, and up to $30 million in future oversight costs. The precise amounts
would be determined by further motion practice or trial.
13

Although Georgia-Pacific and Glatfelter initially opposed the Proposed Consent Decree,
only Glatfelter has continued its opposition. Glatfalter contends that the Proposed Consent Decree
should not be approved because it departs substantially from previous decisions by the Court and
the Court of Appeals in that it upends the Court’s previous reliance on the knowledge and
culpability factors for equitable allocation of costs and does not take into consideration collateral
source payments. Glatfelter also argues that the agreement is procedurally and substantively unfair
because its method of allocation lacks evidentiary support and ignores evidence presented in the two
cases before the Court.
II. ANALYSIS
A. Standard of Review
“[A] district court must approve a consent decree if it is reasonable, consistent with
CERCLA's goals, and substantively and procedurally fair.” United States v. George A. Whiting
Paper Co., 644 F.3d 368, 372 (7th Cir. 2011) (citing In re Tutu Water Wells CERCLA Litigation,
326 F.3d 201, 207 (3d Cir. 2003)). In reviewing a request for approval, “the trial court must defer
to the expertise of the agency and to the federal policy encouraging settlement.” Id. “That policy
[of deference] has particular force where, as here, a government actor committed to the protection
of the public interest has pulled the laboring oar in constructing the proposed settlement.” United
States v. Cannons Engineering Corp., 899 F.2d 79, 84 (1st Cir. 1990) (citing F.T.C. v. Standard Fin.
Mgmt. Corp., 830 F.2d 404, 408 (1st Cir.1987)). Of course, the court’s decision ultimately depends
on the persuasive power of the agency's proposal and rationale, taking into consideration whatever
practical considerations may impinge and the full panoply of the attendant circumstances. But “the
district court must refrain from second-guessing the Executive Branch.” Id. This is especially true
where the proposed consent decree has the approval of a multiple well-represented parties:
14

Respect for the agency's role is heightened in a situation where the cards have been
dealt face up and a crew of sophisticated players, with sharply conflicting interests,
sit at the table. That so many affected parties, themselves knowledgeable and
represented by experienced lawyers, have hammered out an agreement at arm's
length and advocate its embodiment in a judicial decree, itself deserves weight in the
ensuing balance. . . . The relevant standard, after all, is not whether the settlement
is one which the court itself might have fashioned, or considers as ideal, but whether
the proposed decree is fair, reasonable, and faithful to the objectives of the
governing statute. . . . While the district court should not mechanistically
rubberstamp the agency's suggestions, neither should it approach the merits of the
contemplated settlement de novo.
Id. (internal citations omitted). All of these factors warranting deference apply here with special
force given the overall expertise of counsel, the amount of discovery undertaken, and the procedural
history of the two cases.
B. The Proposed Consent Decree and NCR’s Existing Obligations
Glatfelter’s principal objection to the Proposed Consent Decree is that, in its view, it does
little more than relieve NCR of its obligations under CERCLA. With the Court having reaffirmed
its pre-remand finding that the harm is indivisible, Glatfelter notes that NCR is jointly and severally
liable, with Glatfelter and Georgia-Pacific, for all of the work NCR is agreeing to perform, as well
as the additional monitoring and maintenance costs, and the governments’ past or future response
costs that NCR will not be required to pay, under the Proposed Consent Decree. Glatfelter further
notes that under the Court’s pre-remand holding in the Whiting case that NCR and Appleton Papers
bear primary liability for the harm, NCR would be liable for 100% of the costs with no right of
recovery against the other defendants. Thus, in Glatfelter’s view, the Proposed Consent Decree
frees NCR from liability that it would otherwise have in the case and, given the consent decrees
with the other PRPs, leaves Glatfelter facing that liability alone with NCR serving as back-up.
Although the Proposed Consent Decree also provides that Appvion would waive its own $200

15

million cost recovery claim, Glatfelter contends that evidence obtained during discovery indicates
Appvion’s claim is worth far less.
Glatfelter’s objection is not persuasive. What Glatfelter fails to adequately appreciate is that
the Court’s original holding in Whiting that NCR bears primary responsibility for the clean-up was
vacated by the Court of Appeals. The Court of Appeals remanded the case for a full trial so the
parties could present evidence on all of the factors bearing on equitable allocation of liability and
for the Court to decide which factors would guide its decision on a more complete record. Whiting,
768 F.3d at 703. Glatfelter notes that the Court of Appeals “took no issue with the Court’s factual
findings on knowledge, nor did it find that the Court’s allocation would necessarily be altered on
remand.” Dkt. 1196, at 13 (citing Whiting, 768 F.3d at 703). But while it is true that the Court of
Appeals acknowledged that “a party’s knowledge that it is either doing something wrong or running
an undue risk that it is doing something wrong can be a compelling basis for deciding who should
contribute to the costs of clean-up,” the Court explicitly declined to “privilege that factor above all
others.” Id. “Knowledge may be the point that tips the scale, the Court observed, “but only after
other plausible relevant factors have been considered and either added to the balance or discarded
as inappropriate for the case at hand.” Id.
The Court of Appeals also noted that the reasons given for allocating all costs to NCR were
inconsistent with the period of “general ignorance” of the dangers of PCBs. Recall that this Court
had concluded for purposes of summary judgment only that the undisputed evidence showed NCR
was aware of the potential PCB problem by the late 1960s when production ramped up
significantly. Prior to that, the court concluded that NCR was in the best position to appreciate the
risks since it was the producer of the product that introduced PCBs to the market in the form of

16

CCP. 2009 WL 5064049, at *16–18. The Court of Appeals found these reasons unpersuasive,
noting “[r]egardless of whether NCR produced the carbonless copy paper and the defendant mills
merely recycled it, the fact remains that it was used in all of their businesses. At least for that
period, the court gave no reason to fault one party more than another for producing something that
it did not realize was environmentally toxic.” Whiting, 768 F.3d at 703.
In light of the Court of Appeals’ decision in Whiting vacating this Court’s grant of summary
judgment, Glatfelter’s argument that the Proposed Consent Decree upends nine years of decisions
emphasizing NCR’s knowledge and culpability simply fails. When a district court’s decision is
vacated, its reasoning does not survive as the law of the case. See Johnson v. Bd. of Educ. of
Chicago, 457 U.S. 52, 53–54 (1982) (“Because we have vacated the Court of Appeals' judgments
in this case, the doctrine of the law of the case does not constrain either the District Court or, should
an appeal subsequently be taken, the Court of Appeals.”). It is for that reason that the case was set
for a full trial on all of the factors bearing on apportionment of liability. Nothing in the Court of
Appeals decision required this Court to impose 100% or even a majority of the clean-up costs on
NCR. It was a new ball game. In sum, Glatfelter’s assumption that NCR remains primarily liable
for 100% of the clean-up costs and the Proposed Consent Decree relieves it of liability already
determined is wrong.
Galtfelter’s objection also fails to take into consideration the risk that the harm to OU4
would ultimately be found divisible and NCR’s liability limited to its divisible share. As noted
above, this Court’s finding that the harm in OU4 was divisible was reversed by the Court of
Appeals in the enforcement action and the case remanded for further consideration of the issue. The
Court of Appeals held that this Court had erred in treating the harm caused by PCB contamination

17

and remediation costs as binary and concluded from its consideration of the evidence that the harm
may be divisible: “As a result, we think the harm would be theoretically capable of apportionment
if NCR could show the extent to which it contributed to PCB concentrations in OU4. And if NCR
cleared that hurdle, we think a reasonable basis for apportionment could be found in the remediation
costs necessitated by each party.” Glatfelter, 768 F.3d at 678. Although the Court of Appeals
concluded that the Court had adequately explained why mass-percentage estimates of PCBs by
NCR’s experts likely understated the contribution of NCR and its predecessors, it noted that the
Court had failed to explain why Dr. Wolfe’s estimates were unreliable and remanded the case for
an explanation. Id. While the Court ultimately reaffirmed its original finding that the harm was
divisible on remand, there is no guarantee that this finding would withstand further appeal. By
agreeing to waive its right to appeal in the Proposed Consent Decree, NCR is giving up the
possibility that it could prevail on this issue and limit its liability even further. This, of course,
benefits Glatfelter, as well as the governments and Georgia-Pacific, by resolving the issue in their
favor without the uncertainty and expense of continued litigation.
Even aside from these considerations, Glatfelter is also wrong in its contention that the
Proposed Consent Decree ignores this Court’s previous determinations concerning NCR’s
knowledge and culpability. Notwithstanding the fact that this Court’s decision holding NCR
primarily liable based on its knowledge and culpability was vacated by the Court of Appeals, the
governments’ allocation analysis underlying the Proposed Consent Decree mirrors this Court’s
previous assessment of those factors and places substantial, though not exclusive, weight on them.
This is clear from the allocation scheme the governments devised to which the Court now turns.

18

C. The Governments’ Allocation Scheme
For the purpose of facilitating and supporting settlements, the governments devised a casespecific multi-factor allocation scheme to assess fair shares for NCR/Appvion, Georgia-Pacific, and
Glatfelter. The scheme developed by the governments assigned equitable responsibility based on
four broad allocation categories or factors. The governments then used a 1000 point scoring system
for each PRP, with the possible points divided among the four equitable allocation factor scoring
categories. The four allocation categories and the maximum points assigned were: (1) Knowledge,
Culpability, and Benefit—maximum of 400 allocation points (representing 40% of total points); (2)
PCB Mass Discharge—maximum of 400 allocation points (representing 40% of the total points);
(3) Geographic Considerations and Litigation Exposure—maximum of 100 points (representing
10% of the total points; and (4) Non-Cooperation with the Government—maximum of 100 points
(representing the remaining 10% of the total points). Dkt. 1189 at 10–11. After assigning
NCR/Appvion, Georgia-Pacific, and Glatfelter allocation point scores in each of these categories,
their point scores were summed, compared and converted to equitable share percentages. Because
prior settlers had already agreed to bear 11% of the estimated total costs and damages, the
governments used its allocation calculus to allocate the remaining 89% to NCR/Appvion, GeorgiaPacific, and Glatfelter.
Consistent with this Court’s pre-remand rulings in the Whiting case, the governments
awarded NCR/Appvion the maximum 400 points, and Georgia-Pacific and Glatfelter 0 points, in
the Knowledge, Culpability, and Benefit category, notwithstanding evidence that the presence of
PCBs in CCP was publicly available in the 1950s and that Georgia-Pacific and Glatfelter acquired
knowledge of the risk at the latest after the production period in the 1970s, continued to recycle

19

CCP, and benefitted from use of the less expensive stock. Dkt. 1189 at 12–13. The governments’
assignment of points for PCB Mass Discharged was also unfavorable to NCR/Appvion.
NCR/Appvion received a two to three times larger share of the points allocated for PCB Mass
Discharged (267-360) than the number of points (120-133) awarded to the prior settlers, GeorgiaPacific and Glatfelter, notwithstanding evidence offered at trial that the discharge from NCR’s
production facilities was much smaller. Id. at 13–14.
The governments’ assessment of the third factor, Geographic Considerations and Litigation
Exposure, resulted in a more balanced point distribution, though Glatfelter’s location in the
uppermost river segment meant that it contributed to the PCB pollution in all segments of the river
and in Green Bay and was subject to contribution claims by many of the downstream dischargers,
especially since it had contributed relatively little to the downstream clean-up and, in particular, the
Natural Resource Damages (NRD). NCR/Appvion’s location in OU2 likewise meant that it
contributed to the pollution downstream in OU4 and Green Bay, and Georgia-Pacific’s location and
settlement with the governments limited its ability to obtain contribution from it. On the other hand,
the possibility that NCR could prevail on its divisibility defense and Appvion’s potential recovery
on its cost recovery claim mitigated its risk somewhat. Georgia-Pacific’s location meant its
substantial discharge was limited to the lower segment of OU4. Based on their assessment, the
governments awarded Glatfelter 70 points, NCR/Appvion 50 points, and Geogia-Pacific 30 points.
Id. at 14–15.
On the fourth factor, Non-Cooperation, the governments again awarded the most points (75)
to Glatfelter, noting that apart from its cooperation in the clean-up of OU1, it had been the most
recalcitrant and uncooperative PRP involved in the site. The fewest points in this category (30)

20

were awarded to Georgia-Pacific since it had been the first PRP to settle its NRD liability and had
resolved nearly all government claims for the site through a series of settlements, though it had
contributed to the friction and delays over the last several years. NCR/Appvion was awarded 50
points in this category in recognition of the fact that they had helped advance the clean-up and
restoration work at the site through several partial settlements and as the lead performing parties
under the UAO. The governments recognized that NCR’s initiative had diminished after this
Court’s Whiting ruling, however, and it was at that point that the governments filed this
enforcement action and sought a preliminary injunction to insure that the work continued. Id. at
16–18.
In the final step in the governments’ allocation calculus, the point awards in each of the four
categories were added together for each party and then the totals used to compute corresponding
equitable share percentages.

The resulting equitable share percentages are 55-59% for

NCR/Appvion, 12-14% for Georgia-Pacific, and 18-20% for Glatfelter. Id. at 17. The governments
note that under the Proposed Consent Decree, the approximate total commitments for response costs
and NRD by each party are: (1) for NCR/Appvion $876 million, or 65%; (2) for Georgia-Pacific
$164 million, or 12%; (3) for Glatfelter $162 million, or 12%; and (4) for the prior settlers $144
million, or 11%.

Based on the governments’ allocation analysis, it is thus apparent that

NCR/Appvion are paying significantly more than their fair share, and Glatfelter is paying
significantly less. In reality, the governments contend, the allocation could turn out to be even more
favorable to Glatfelter, given the potential additional back-up responsibilities for long-term
monitoring and maintenance that NCR/Appvion are taking on and the high-end estimates of
Glatfelter’s liability for those same costs and future oversight costs of the governments.

21

Glatfelter disputes the governments’ contention that its allocation scheme gave appropriate
consideration to NCR’s knowledge and culpability. Glatfelter argues that the governments have
concocted their allocation scheme out of thin air as a post hoc justification for the Proposed Consent
Decree and that it is unsupported by any evidence. The scheme calls for assigning points, subject
to a maximum for each category, and then adding up all the points to arrive at a percentage share
of the 89% of the costs not covered by the settling parties. Because the total number of points
assigned to any party can vary, Glatfelter contends that the value of each point can vary as well and
the Court can never determine how much weight the governments actually think a factor should be
given.
Glatfelter also challenges the governments’ choice of the factors they used to perform their
analysis and the point awards they made for those factors. For example, Glatfelter claims that the
governments used the PCB Mass Discharge factor to skew their allocation unfairly in NCR’s favor
and argues that no reason is offered why PCB Mass Discharged is an appropriate equitable factor
in any event. Even assuming PCB Mass Discharge should play a role, Glatfelter argues that there
is no reason why it should be given as much weight as knowledge and culpability, especially in light
of this Court’s previous rulings in the Whiting case. Moreover, Glatfelter contends, discharge
estimates forty to fifty years after the fact are too unreliable to be given significant weight as part
of an equitable allocation, and those estimates, deficient as they are, fail to take into account NCR’s
equitable share of PCBs discharged by the recycling mills.
Glatfelter is similarly critical of the governments’ selection and application of the other
equitable factors it selected to conduct their analysis. Glatfelter contends that the governments
arbitrarily included “Geographic Considerations and Litigation Exposure,” a factor Glatfelter argues

22

is incomprehensible, confusing and not reproducible, and that the assignment of points was made
using a methodology that is not identified. With respect to Cooperation, Glatfelter notes that it fully
cooperated with the clean-up in OU1, the area of the river where it directly discharged, under a
consent decree without the need for a unilateral order. It notes that the OU1 clean-up was not
complete until 2010, and for that reason, the 2007 UAO did not require it to take any action until
the time came for the 2009 dredging season. By that time, NCR and Appvion had commenced the
Whiting case and the matter has been in litigation since. Glatfelter contends it has at all times made
a good faith effort to contribute its fair share toward compliance with the UAO, taking guidance
from the rulings of this court and the Seventh Circuit, particularly, this court’s ruling in 2012 that
although Glatfelter was liable for costs of complying with the UAO, Glatfelter’s liability was
secondary to that of NCR. Given this history, Glatfelter argues that the governments assessment
of the degree of its cooperation is unreasonable and unfair. For all of these reasons, Glatfelter
argues that the governments’ allocation analysis should be rejected.
There is some truth to Glatfelter’s criticism. Dressing an analysis up in numbers does not
transform what is essentially a qualitative judgment of what is a fair and just allocation of
responsibility for the PCB contamination of the Lower Fox River into a quantitative judgment of
what number is largest. Some quantitative measures, certainly on the question of the PCB Mass
Discharge of each party, do bear on the question. But ultimately, the question for the Court is not
reducible to math. Yet, this does not mean that the governments’ overall analysis is unreasonable,
especially in view of the directions of the Court of Appeals in remanding the Whiting case. The fact
that the outcome may not have the precision suggested by the numbers used in the governments’
allocation scheme does prevent the Court from considering the analysis more generally.

23

Glatfelter again argues that, to approve the Proposed Consent Decree, the Court would have
to "reverse" its previous rulings and now find, contrary to its earlier statements, that factors other
than culpability are two or even three times more important. To repeat, Glatfelter fails to
adequately appreciate that the Court’s decision in Whiting was vacated by the Court of Appeals.
For the reasons explained above, the Court’s previous rulings are no longer controlling, and the
Court’s role at this point is significantly different than it was when it issued those rulings. In
approving a settlement, a court must determine whether it is within the bounds of
reasonableness—not whether it is exactly, or even nearly, how the court would allocate
responsibility itself. What is evident from this action and other CERCLA cases is that different
judges might reach different apportionment decisions based on the same set of facts. Glatfelter
fundamentally misapprehends the Court's role at present. The Court is not making a judgment that
the settlement is how it would go about resolving the case (supposing the Court had that power),
but whether the result is within the bounds of reasonableness that any judge reviewing the same
facts might readily deem reasonable. Given this limited role, the Court's previous rulings as to fault
are merely informative rather than binding; they are at most vestiges of a different process and
different role for the Court. Viewed from this perspective, the Proposed Consent Decree is fair and
reasonable.
The Proposed Consent Decree and the governments’ allocation scheme offered in support
of it are consistent with the remand directions provided by the Court of Appeals in the Whiting case.
While acknowledging in that case the “compelling” picture this Court had painted on the knowledge
issue, the Court of Appeals noted “it does not explain why other factors, such as cooperation in the
cleanup or relative volumes dumped, are so wholly irrelevant to the equitable determination that

24

information about them need not even be gathered or presented by the parties.” Whiting, 768 F.3d
at 703. Given the Court of Appeals’ explicit reference to them, the governments’ inclusion of “PCB
Mass Discharged” and “Cooperation With The Government” as allocation factors in its analysis was
unquestionably reasonable. The governments’ decision to reduce the overall weight accorded
knowledge and culpability was also reasonable in light of the Court of Appeals’ concern over the
manner this Court had treated the “period of ‘general ignorance’ preceding the existence of any
serious evidence about the dangers of PCBs.” Whiting, 768 F.3d at 703.
Recall again that the Court’s summary judgment determination of NCR’s actual knowledge
was limited to the period after the late 1960s and relied more on conclusions concerning NCR’s
constructive knowledge—what it knew or should have known—for the earlier period of time during
which PCBs were released into the River. True, it was NCR that introduced CCP, paper coated in
a PCB emulsion, into the economy and profited most from it, as the Court noted in its 2009
summary judgment order. Dkt. 795 at 33–35. But it is also true, as NCR argued and the Court noted
in the same decision, that Glatfelter and the other paper companies had been discharging pollutants
into the River for decades. Id. at 29–30. The Court had concluded then that since the paper
companies had no knowledge of the presence of PCBs in the paper stock they were recycling or the
magnitude of the risk PCBs posed, NCR’s “arguments based the existence of other kinds of
pollution do not speak to the equities of whether any of the Defendants should pay for the cleanup
of PCBs in the Lower Fox River.” Id. at 30. Unfortunately for Glatfelter, the Court of Appeals was
unpersuaded by this reasoning, and it was certainly conceivable that an expanded record at a full
trial could provide support for an inference of some degree of constructive knowledge on the paper
companies as well. Given these considerations, the decision of the governments to elevate other

25

factors and thereby reduce the significance of knowledge and culpability in an effort to settle with
NCR and Appvion was reasonable.
It was also reasonable for the governments to treat PCB Mass Discharge as a principal factor
in their allocation analysis in addition to Knowledge, Culpability and Benefit. CERCLA is a strict
liability statute, and in the absence of clear evidence of culpable intent by a party for the harm to
the entire site, it is logical that the amount of a pollutant a party discharges would be the most
important factor in allocating responsibility for the clean-up; the more a party discharged, the more
that party should pay to clean it up. See United States v. Davis, 31 F. Supp. 2d 45, 63 (D.R.I. 1998)
(citing as first of four categories in which critical factors for allocation can be grouped: “The extent
to which cleanup costs are attributable to wastes for which a party is responsible.”). Because the
Court of Appeals rejected this Court’s conclusion that NCR’s culpability for the PCB contamination
of the entire site justified imposing on it primary liability, the relative contributions of each party
clearly became a crucial factor.
Glatfelter’s objection to the governments’ allocation of points for the PCB Mass Discharge
category is even weaker.

As noted above, the governments devised a scenario based on

consideration of all of the expert opinions provided in the litigation that was unfavorable to
NCR/Appvion and gave both Glatfelter and Georgia-Pacific the benefit of the doubt in awarding
points in this category. NCR/Appvion was given two and three times the number of points that
either Glatfelter or Georgia-Pacific received, meaning that NCR/Appvion’s discharge was between
40% and 50% of the total PCB mass discharged. Dkt. 1189 at 13. No technical analyst had opined
that NCR discharged more than 50% of the PCB mass released into the River, and most had offered
opinions that it was significantly less. Dkt. 1198-2. The only report that included a number above

26

50% for NCR was the 2000 report of Amendola Engineering, which used 54% as a high range in
a sensitivity analysis that assumed a maximum loss rate for the PCB-coating emulsion used in
making CCP that was double the expected loss rate offered by Glatfelter’s expert on the subject,
Dkt. 1197-2 at 7, and also double the rate adopted by the Court based on the evidence offered at
trial. Dkt. 794 at 35–36.
Despite this fact, Glatfelter argues that the unreliability of calculating PCB Mass Discharged
mandates that at most light weight be given to this factor. But as other courts have recognized, “in
most cases of this type, assessing relative responsibility is an imperfect process because it requires
subjective judgments based on evidence that is not completely developed and may be disputed.”
United States v. Davis, 11 F. Supp. 2d 183, 191 (D.R.I. 1998), aff’d, 261 F.3d 1 (1st Cir. 2001). To
require greater exactitude “would be impractical and would frustrate CERCLA’s objective of
encouraging early settlement and prompt remediation of hazardous waste sites.” Id. at 191–92. To
avoid this result, the requirement of substantive fairness is considered satisfied “when a proposed
settlement reflects a rational method of allocating liability in a manner that reasonably approximates
each party's share of responsibility; the method is applied evenhandedly with respect to all PRP's
and sufficient information is presented to enable the Court to determine whether that has been
done.” Id. at 192. “As long as the data the EPA uses to apportion liability for purposes of a consent
decree falls along the broad spectrum of plausible approximations, judicial intrusion is
unwarranted—regardless of whether the court would have opted to employ the same data in the
same way.” Cannons Eng’g, 899 F.2d at 88.
Here, after two trials, three appeals, and almost ten years of litigation, the evidence bearing
on volumetric discharge has been far more developed than in most CERCLA cases. Estimates

27

offered by experts retained by the parties relied on a vast number of historical documents relating
to paper production, recycling and water treatment, as well as chemical analysis of sediment in
sludge ponds and in samples taken over the entire length of the River. While it is true that this long
after the fact estimates of which companies were responsible for what quantities of PCBs released
into the River and where they are located can only be rough estimates, that is hardly a reason for
ignoring volume altogether. The ranges relied on by the governments in awarding points under this
category fall well within “the broad spectrum of plausible approximations.” Given the fact that the
governments adopted a worst-case scenario from NCR/Appvion’s perspective in assessing this
factor, Glatfelter has no cause for complaint.
Finally as to volume, the governments correctly note that Glatfelter’s argument that the
analysis of PCB Mass Discharged category should take into account NCR’s equitable share of PCBs
discharged by the recycling mills ignores the fact that NCR’s responsibility as the source of PCBs
is taken into account in the Knowledge, Culpability, and Benefit factor. The Court found that in
selling its broke to the recyclers, NCR had not arranged for the disposal of hazardous wastes so as
to warrant the imposition of direct liability under CERCLA § 107. Whiting, Dkt. 1405. In other
words, Glatfelter’s discharge of PCBs into the River was not part of an effort to dispose of NCR’s
hazardous waste. To accept Glatfelter’s argument would undermine this finding and essentially
double-count the assessment of NCR’s culpability.
Geographic Considerations and Litigation Exposure were also appropriately considered in
the governments’ allocation scheme. Unlike the first two categories, the governments limited the
number of points that could be awarded for this factor, as for cooperation, to 100, thereby giving
them less weight in the overall analysis. Still, in terms of geography, Glatfelter’s location at the

28

uppermost segment of the River necessarily meant that it contributed to PCB pollution in the entire
River and in Green Bay. This fact, combined with the holding, affirmed on appeal, that Glatfelter
is jointly and severally liable for OU2-5, exposed it to more claims for contribution than, for
example, Georgia-Pacific, which was located in the lower section of OU4, a short distance from
where the River flowed into the Bay, where no dredging was ordered. Glatfelter was also exposed
to liability for a significant share of future costs if NCR were to prevail on its divisibility defense.
Other than reargue its own divisibility defense, which both this Court and the Court of Appeals
rejected, Glatfelter, 768 F.3d at 678–80, Glatfelter offers little reason to question either the
appropriateness of this factor or the relative point assessment included in the governments’ analysis.
Cooperation with the government on the cleanup is also a well recognized factor that courts
have considered in apportioning liability in CERCLA cases. It is also a factor the Court of Appeals
expressly cited as warranting consideration by the Court on remand. Glatfelter’s objection here is
primarily to the governments’ assessment of points for this category. Glatfelter’s objection takes
the form, to a large extent, of explaining why, given this Court’s previous rulings, it was justified
in withholding contributions to the cleanup in OU4 after the Whiting litigation began. Glatfelter
also points to the cooperation it provided in the cleanup effort in OU1.
The governments acknowledged Glatfelter’s cooperation in the cleanup of OU1, however,
and emphasized its lack of cooperation since the Court of Appeals’ decisions vacating the Court’s
Whiting decision and affirming Glatfelter’s joint and several liability. The governments noted that
“[a]t this point, Glatfelter is the only remaining PRP not to have entered into a settlement addressing
its responsibilities for OU2-5.” Dkt. 1198 at 12. The governments also noted that “[e]ven by its
own estimation, Glatfelter only contributed ‘about 8.5% of the work done’ in 2015 and 2016 (Dkt.

29

1196 at 33), far less than the one-third share of the work EPA assigned to Glatfelter in the 2015
Remedial Action Work Plan alone (Dkt. 996-3; Dkt. 1068 at 4.).” Id. To be sure, Glatfelter may
have had good reasons for declining to comply with the EPA’s assignment. It may have been
convinced that ultimately, after a full trial, it would again prevail on its argument that NCR should
have primary liability because of its knowledge. Regardless, the fact remains that by its own
admission, Glatfelter’s cooperation beyond OU1 was significantly more limited than NCR’s. For
this reason, the court cannot say that the governments’ apportionment scheme was unreasonable.
Overall, Glatfelter argues that courts maintain an obligation to involve themselves more
fully in the complex evidentiary details when a case is highly technical, like this one is. Yet in the
context of this action, the opposite would appear true. First, the governments and the settling
defendants have been at loggerheads for years, scarcely agreeing on anything throughout this case's
lengthy and tortuous history. To call this an arms' length settlement would be putting it mildly.
That they have finally sat down and resolved their differences is evidence that the agreement's terms
are hard-fought, both procedurally and substantively, and fair. Second, the governments have
shown a dogged determination to ensure the cleaning up of the River in the most expeditious
fashion, and they and their agencies may be presumed to have not only the will to get the job done,
but the expertise that warrants a court's deference. At this stage, no further evidentiary hearing is
required.
D. Collateral Source Payments
Finally, Glatfelter contends that the Proposed Consent Decree upends the Court’s ruling in
the Whiting action, affirmed by the Court of Appeals, that an equitable allocation must consider
collateral source payments by insurers and others who have agreed to indemnify NCR for portions

30

of its liability. Dkt. 1196 at 10–16. But that is not what this Court held or what the Court of
Appeals affirmed. The Court’s holding, affirmed by the Court of Appeals, was that “the collateral
source rule does not apply in CERCLA § 113(f) contribution actions.” Whiting, 768 F.3d at 707.
The collateral source rule generally bars the reduction of a claimant's recovery by the amount of
compensation received from other sources, i.e., sources ‘collateral’ to the defendant. Molzof v.
United States, 6 F.3d 461, 464 (7th Cir. 1993). Because the collateral source rule is inapplicable in
CERCLA cases, courts may, in principle, take insurance payments into account when deciding
contribution shares. Whiting, 768 F.3d at 707. But there is no requirement that they do so and,
more importantly, NCR is not seeking contribution from Glatfelter for amounts paid by its insurers
or indemnitors.
Before the Court of Appeals vacated this Court’s decision in Whiting holding NCR primarily
liable for the cleanup, the Court held that Glatfelter and the other PRPs could not recover on their
contribution counter claims against NCR cleanup costs they had already covered through insurance.
Relying on the Tenth Circuit’s decision in Friedland v. TIC-The Indus. Co., 566 F.3d 1203, 1207
(10th Cir. 2009), the Court held that “permitting a CERCLA contribution-action plaintiff to recoup
more than the response costs he paid out of pocket flies in the face of CERCLA’s mandate to
apportion those costs equitably among the liable parties.” Whiting, Dkt. 1080 at 24–25; see also
Whiting, Dkt. 1502 at 5. In other words, the Court’s ruling was intended to prevent Glatfelter and
the other contribution counter claimants from receiving a double recovery, one from their insurers
and another from NCR, for the same costs. Glatfelter contends that this ruling in the Whiting case
requires that in determining whether the $876 million NCR and Appvion will ultimately pay under
the Proposed Consent Decree is fair and reasonable, the Court must offset against that amount all

31

reimbursements received from their insurers and indemnitors, and even the tax benefits they may
have received from the government for business deductions and write-offs.
Glatfelter is wrong. The Court’s ruling in the Whiting contribution action has little, if any,
relevance to the Proposed Consent Decree offered to resolve NCR’s liability in this enforcement
action, especially given the Court of Appeals’ decision vacating the Court’s finding that NCR had
primary liability. To the extent the ruling survives, the proposed settlement does not violate it.
Under the Proposed Consent Decree, NCR will not profit by recovering from its indemnitors more
than it has paid, and has yet to pay, for its share of the clean-up. Were the Court to adopt
Glatfelter’s position, a time-consuming and costly analysis of not just NCR’s insurance settlement,
indemnity agreements and tax returns, but those of every other responsible party, including
Glatfelter, would be required to offset their contributions as well. This makes no sense, and nothing
either this Court or the Court of Appeals said in Whiting requires such a result.
CONCLUSION
In sum, the Proposed Consent Decree is substantively fair and reasonable. It allocates
NCR/Appvion 65% of the estimated total response costs and NRD, 12% each to Georgia-Pacific
and Glatfelter, and 11% to the prior settling parties. Appvion waives its cost recovery claim, NCR
foregoes its divisibility defense, and both waive, along with their indemnitors, appeals of any
previous rulings, thus providing finality to the litigation with the assurance to the other parties that
their liability will not be even greater unless they elect to continue to pursue their own claims
against NCR or Appvion.
The Proposed Consent Decree is also procedurally fair and reasonable. It is the result of
arms-length negotiations among counsel who have been involved in the litigation surrounding the

32

site for many years, representing sophisticated clients. The settlement relies on well-documented
past cost figures, informed estimates of future costs, and a well-developed record regarding relevant
allocation factors, including knowledge/culpability, mass discharged, cooperation, geographic
considerations and litigation exposure. The agreement also includes revisions made in response to
concerns expressed by Georgia-Pacific and Glatfelter at an in-person meeting with the governments.
Finally, the Proposed Consent Decree also advances the purposes of CERCLA by assuring
completion of the remedial action work at the site in an expeditious manner, with financial
guarantees. It also streamlines the few issues remaining in the enforcement case and will likely put
an end to most (if not all) of the Whiting case.
For all of these reasons, the motion to approve settlement and enter the Proposed Consent
Decree (Dkt. 1188) is GRANTED. The Clerk is directed to enter the Consent Decree, along with
its appendices, forthwith. Glatfelter’s motion to file a surreply memorandum in opposition to the
motion to approve (Dkt. 1202) is likewise GRANTED. All other motions now pending in the case
(Dkt. 1125, 1134, 1141, and 1171) are denied as moot. The parties should file a status report
concerning the remaining claims in both this action and the Whiting case within 14 days of the date
of this order.
SO ORDERED at Green Bay, Wisconsin this 22nd day of August, 2017.

s/ William C. Griesbach
William C. Griesbach, Chief Judge
United States District Court

33