*4 ALITO, and Before FUENTES BECKER, Judges. Circuit ensuring precise THE for the correct and mea- OF COURT OPINION (which they surements their waste BECKER, Judge. Circuit certify), we do not that a obliged believe from an order appeal This is laboratory error defense-where the error judgment for the granting District Court in overreporting-is resulted inconsistent against defen- plaintiff United States Rather, regime. with this inasmuch as the Corporation Ludlum Allegheny dant penalty imposed is for an unlawful dis- (“ALC”) brought in an for violations action faulty charge reporting, (“CWA” Act or the Clean Water deprivation think that the defense would “Act”) Pennsyl- at five of ALC’s Western the CWA and pin-pose advance the judg- manufacturing vania facilities. unfair, especially grossly that it would be (1) multifaceted, flowing ment from: the presence companion pro- view of Court; legal by the pretrial determinations (2) imposing liability visions CWA jury liability verdict on a number (3) issues; monitoring reporting determinations violations. We following penalty hearing. the judgment part will therefore vacate *5 mixed; jury pre- verdict was each side laboratory and remand so that error issues, on a number of and ALC’s vailed adjudicated defense can be considered and appeal unchallenged significant por- leaves respect claims. to affected against tions it. judgment of the Howev- requires to appeal The also us determine er, major appeal challenge does as- whether District Court made either pects judgment of the and also of the civil of mistake law or abused its discretion against assessment leveled ALC calculating the economic benefit that ALC alleged for the violations in the sum of obtained from those violations that un- are $8,244,670. 1319(d) challenged appeal. on Section important question presented first Court, requires CWA the District by viability appeal concerns the of the determining when the amount of a civil “laboratory so-called error defense.” The CWA, penalty under the consider “the eco- operates CWA a self-monitoring under (if any) resulting nomic benefit from the reporting system whereby dis- violation,” playing so as to “level the field.” charger of toxic waste measures and re- The District calculation was Court’s here ports Environmental Protection (“EPA”) an agglomeration, based on a number Agency the volume of its dis- largest factor charge. single ALC maintains that the EPA factors. was the predicated aspects certain the violation by govern- 12.73% interest rate used by upon reports submitted ALC were compute ment and the District by laboratory tainted error caused interest the date of violation to the contaminated reagent resulting overre- judgment date of the so as to calculate the porting of the amount of toxic zinc total economic benefit to ALC. This rate discharge. The District Court declined predicated largely on a calculation of laboratory allow the error defense weighted average capital cost of grounds recognized that it had been (“WACC”). Noting that it was uncontest- Circuit, the Third and that to allow such a ed at trial that ALC had an actual rate of “new defense” would contravene CWA. capital that return on was less than half Court, Although operates 12.73%rate used the District the CWA under a regime liability, designed ALC contends that the 12.73% rate is ex- strict to en- polluters sure that responsibility will take cessive. application conclude that the of the Court counting
We erred violations of may vastly monthly averages 12.73% rate so overstate the as violations for each We, day course, of the month. improper economic benefit to of its follow precedent our in Natural discharges, play- that it does not “level the Resources De- Council, field,” Inc. v. Refining Texaco ing and that it constitutes an & abuse fense Inc., (3d Marketing Cir.1993), F.3d 493 prelude of discretion. As a to making this daily average computed by limit is determination we explore potential averaging effluent only days levels ramifications of the notion of economic facility which the operated. Although 1319(d). § under benefit conclude We some notably Courts-most the Fourth Cir possible approaches there two Foundation, cuit in Chesapeake Bay Inc. (1) calculation economic benefit: Ltd., v. Gwaltney Smithfield, 791 F.2d i.e., cost of capital, what it would cost the (4th Cir.1986), vacated on other polluter necessary to obtain the funds grounds, 484 U.S. 108 S.Ct. equipment necessary install the to correct (1987)-have L.Ed.2d 306 held that a viola- (2) violation; the actual return on monthly tion of a average parameter con- i.e., capital, what the polluter earned on stitutes a violation for day each capital it declined to divert for month, we find approach incomplete. equipment. installation of the Because adopt Gwaltney We insofar estab- variable, these factors are depending so lishes an absolute upper bound on the upon market conditions and the financial penalty that can be a monthly assessed for polluter, soundness of the we leave it to average However, violation. permit limits Court, in the sound exercise of *6 can be in many ways, exceeded different discretion, its approach decide which by very both large, discharges isolated and (there apply apply and how to are a discharges. moderate continuous Fur- models). variety However, explain we thermore, daily monthly average and limits why application the District Court’s designed are to avoid distinct environmen- was, minimum, WACC this case at a result, tal harms. As a in some cases a evidence, unsupported by the and needs to wrongful violator’s conduct pun- will merit be recalculated should the District Court daily monthly ishment for and both viola- on remand elect pursue approach. tions, others, while the conduct will have contrast, In we conclude that the Dis- sufficiently punished been by penalties trict application legal- Court’s of the other daily violations alone. hold that We dis- ly required factors to calculate ALC’s eco- determine, trict courts have discretion to nomic costly benefit-the least method of case, many the facts of each how viola- compliance periods non-eompli- and the days tion should be assessed for supported by ance-were the record. In purposes for the a monthly violation of determination, the course of this clarify limit, average based on whether violations proper for determining method already sufficiently are sanctioned as viola- economic benefit is to base the calculation daily tions of a maximum In limit. this costly on the least compliance. method of case, the District Court did not have the benefit, On the issue economic we there- standard, benefit of this so we will vacate respect fore vacate and remand with to the its penalty assessment and remand for fur- interest rate issue. proceedings. ther
Finally whether, we must decide in com- part, We will therefore affirm in vacate piling the number of violations for the in part, proceed- and remand for further purpose assessing penalty, ings consistent with opinion. (“PADEP”). History The two WWTPs at Procedural tection
I. Facts and facility discharge Leechburg West steel and owns and manufactures ALC pursuant to anoth- the Kiskiminetas River three comprising spe- plants five operates permit. Vandergrift The er NPDES facilities in manufacturing cialty steel discharges process WWTP treated waters Brackenridge Pennsylvania: Western Valley Pollution Con- to the Kiski Water (the and Natrona Facility Brackenridge (“KVWPCA”) Authority pursuant trol (the Facility Leechburg plants); West applying it. After further permits with plants); and Leechburg Bagdad West treatment, to the Ki- discharges KVWPCA Facility. The Bracken- Vandergrift River. skiminetas melting, continuous Facility conducts ridge filed this action The United States operations. finishing casting, rolling, against ALC on 1995. The Com June Vandergrift Fa- Leechburg and The West amended, alleged types three plaint, as finishing operations. cilities (1) discharges at each of ALC’s violations: steelmaking process generates a containing discharges five in ex facilities pollution. considerable amount permits cess of ALC’s as shown steel-making water from ad- process uses (“DMRs”) Reports Discharge Monitoring jacent pro- The water is used rivers. (2) EPA; discharges submitted to the cooling wa- cess water and as non-contact Vandergrift facility interfered directly in water is used ter. Process operations Valley with the of the Kiski steel, makes contact process making (3) WPCA; report ALC’s failure to equipment. steel-making with steel or required by its permits. violations as cooling Non-contact water cools steel- parties summary filed motions cross actually making without touch- equipment judgment. response, District operates six on-site ing steel. ruled that ALC could raise several (‘WWTPs”) plants treatment wastewater violations, the reported defenses to includ at these facilities. The three WWTPs at ing “laboratory error” defense facility Brackenridge discharge to the reported which ALC contended that River, Allegheny pursuant to a National from erroneous labora violations resulted *7 System Discharge Pollution Elimination tory analyses-later to discovered (“NPDES”) permit by Pennsyl- by reagenb-which issued caused contaminated Department of zinc pollutant vania Environmental Pro- overstated levels.1 The again split samples overreporting ALC's zinc with the two outside labo- 1. of exceedences analyses ratories, upon sample per- was based effluent provided this but time each laborato- by Laboratory sets, formed ALC's Technical which ry sample undigested with two one allegedly turned out to be flawed. ALC tried predigested by Laboratory. one ALC's Ac- to determine the cause the zinc excee- of ALC, cording generated by to values dences, 1996, February success. without In digested analyzed samples that were both laboratory’s per- to started examine its own signifi- outside laboratories showed samples took formance. ALC effluent cantly correspond- than lower zinc results part analyzed Laboratory at the had ALC However, laboratory. ing from results ALC's laboratories, analyzed part by two outside sample analyzed sets the outside labo- protocol "split sampling.” known as Accord- samples predigested by were ratories after ALC, ing to zinc results obtained its high laboratory ALC's were as in zinc as the laboratory higher significantly were than laboratory, leading from ALC results ALC's laboratories, those obtained at the outside digestion pro- the conclusion that it was while outside laboratories' results were laboratory causing in ALC's that was cess split each this consistent with other. In sam- values to zinc be overstated. Once ALC pling, laboratory performed each di- own March, 1996, gestion samples. In ALC
171 28, 2000, opined Court the defense had not vember the final judgment of Circuit, recognized in 20, 2002, been this and that it February 59(e) and the Rule defense, adopt would not such a new “es- 8, order of October 2002.
pecially interpreted since the Act can be as jurisdiction The District Court had creating obligation to insure that the pursuant § to 28 U.S.C. ap 1331. ALC’s accurate, self-monitoring pollutants is peal timely is R.App. under Fed. P. assigning inaccuracy the risk of to the 4(a)(1)(B), and appellate jurisdic we have company.” granted The thus par- pursuant tion § to 28 U.S.C. 1291. Our summary tial judgment to the United grant review the of summary judgment States on that issue. The Court denied plenary. See Shelton v. Univ. Med. & summary motion for N.J., (3d Dentistry 220, 223 F.3d judgment on the in- reporting failure and Cir.2000). We imposition review the of a claims, terference finding that ALC had 1319(d) penalty under Section of the CWA provided sufficient evidence to tri- create discretion, for abuse of see Tull v. able of fact. United issues States, 412, 1831, 481 U.S. 107 S.Ct. jury The District Court held a trial on (1987), L.Ed.2d 365 but our review of the liability January February 1319(d) legal construction of ple Section jury 2001. found favor of ALC nary, see Public Group, Interest Research all reporting of the interference and failure Terminals, Inc., Duffryn Inc. Powell claims, but favor of the (3d Cir.1990). 913 F.2d remaining half of the reported violations total, claims. In the violations for which II. Laboratory Error Defense stipulated liability, those which granted summary
the court judgment, and A. Overview of the Clean Water Act jury those for which the returned a verdict (“CWA”) 1,122 against The Clean Water Act up days ALC added was en- July through by Congress violations from acted in 1972 Febru- “to restore and ary chemical, maintain the physical, and bio- logical integrity of the Nation’s waters.” 8, 2001, February From 5 to the Court 1251(a). § 33 U.S.C. In order to achieve conducted a bench trial on penalty goal, prohibits discharge CWA time, amount. To save the Court allowed any pollutant into waters of the United give the experts testimony their direct except expressly States authorized un- in the form of proffers, written and al- 1311(a). § der Act. See 33 U.S.C. lowed live cross-examination. Following discharge order to pollutants naviga- into trial, parties submitted *8 waters, ble one must obtain a National (judicial) proposed opinions, and on Febru- Discharge System Pollution Elimination 20, 2002, ary the opinion Court filed an and (“NPDES”) permit. § 33 U.S.C. 1342. judgment against entered the Discharges comply with the limits and $8,244,670. amount of ALC filed a motion 59(e) conditions in an permit NPDES under Fed.R.Civ.P. to alter or amend comply deemed to with Act. judgment, which the District the 33 U.S.C. Court 1342(k). 26, 2002, § requires denied. On November The ALC filed CWA NPDES a appeal permittees notice of from the District to test their report Court’s effluent and summary judgment September order of EPA in Discharge the results to the Moni- 28, 2000, (“DMRs”). the reconsideration order toring Reports of No- 33 U.S.C. reagent, longer reported
switched to a different it no zinc exceedences.
172 “necessary adjunct 122.41(j), authority §§ 122.48. an is a 1318(a); 40 C.F.R.
§ water pollution establishment of effective of the CWA authorizes 307 Section and the of such requirements enforcement regulations prohibiting promulgate EPA to (cit- Br. requirements.” Government at 16 a Pub- discharge any pollutant into (1971)). 92M14, (“POTW”) at ing Rep. S. 62 Further- licly Owned Treatment Works more, with, points Congress it out that intended or passes through, that “interferes requirements” to “these new “avoid incompatible with” the POTW. otherwise 1371(b)(1). necessity finding, investiga- lengthy fact prohibits § The Act 33 U.S.C. tions, en- negotiations at time of discharges to that are in excess POTWs 92-414, Rep. at (citing forcement.” Id. S. pretreatment those standards. 33 U.S.C. (1971)). 1317(d). government argues then § The 62 The general EPA has issued that, requirement with the Act’s and national cate- consistent pretreatment standards self-reporting, for accurate courts should gorical pretreatment standards DMRs, industry. treat which must be certified manufacturing iron and steel 403, as admissions that are suf- discharger, Pts. 420. See C.F.R. liability ficient to establish under bring the EPA The Act authorizes CWA. injunctive re- civil enforcement actions for in this relevant, respect relies up penalties, lief and at times Co., F.2d $25,000 Sierra Club v. Union Oil day per for each violation. See (9th 1480, Cir.1987), 1319(d). 1491-92 vacated on § A violation of U.S.C. 931, grounds, 485 other U.S. 108 S.Ct. can showing Act be established (1988), 1102, 99 L.Ed.2d where person discharged pol- a defendant is who of Appeals held that a defen point CWA navigable lutants from source into liability escape dant could not based waters in violation of the terms of alleged sampling violations. The Court applicable NPDES or permit into a POTW program noted funda pretreatment the “NPDES in violation of a standard. 1317(d). mentally self-monitoring” relies on §§ See 33 In assess- U.S.C. Congress deemed accurate DMRs ing § penalty civil for a violation of “critical Act.” operation § to effective or the court must “the consider: violations, opined It that allowing permit- Id. CWA or seriousness violation (if to impeach tees their own DMRs “would any) resulting the economic benefit violation, sanctioning countless additional hours any history such new, litigation violations, creating of NPDES any good-faith comply efforts to complicated factual questions for district applicable requirements, the eco- courts to resolve.” nomic Id. impact on the viola- if tor, permittees Court further reasoned that justice may other such matters 1319(d). impeach reported could their own viola require.” § 33 U.S.C. error, laboratory
tions with claims of B. The Government’s Contentions perverse would “create result of re warding laboratory permittees sloppy argues The government that the CWA practices” efficacy and “undermine the liability establishes scheme of strict *9 Id.; self-monitoring program.” accord aimed facilitating at enforcement. It first Env’t, Inc. Upjohn Conn. Fund v. Congress notes that gave the EPA Co., (D.Conn.1987). 660 1397 F.Supp. information, data, “authority require reports, Relying reasoning, well as establish monitor- on this govern- ing requirements,” reject recognizing that such ment submits that we should
173 (sometimes laboratory error defense. discrepancies Because the defendant’s regulations require dischargers higher, to amend results were they and sometimes they their sworn DMRs whenever were lower than discover the outside laboratories’ results). reporting, an error in their and because
failure to do so constitutes a criminal viola- explained The Court “if a defendant itself, tion in and of con- wishes to contest accuracy of its allowing tends that dischargers to contest DMRs, it ‘has a heavy burden to establish their own DMRs conflicts with the statute faulty analysis.’” Id. at (quoting 1178 applicable regulations. and the 40 See Student Pub. Interest Research Group, (1)(8). (k)(2), §§ C.F.R. 122.41 gov- Georgia-Pacific Inc. v. Corp., 615 F.Supp. ernment further argues allowing (D.N.J.1985)). 1429 “The ‘defendant laboratory error defense would frustrate must present direct evidence of reporting intent, “congressional would com- reward ‘may rely inaccuracies’ and unsup- on panies monitoring practices, for inaccurate ported “speculation” of measurement er- ” give and would them an incentive to wait (quoting ror.’ Id. Georgia-Pacific Corp., they until accuracy are sued to ensure the 1429). F.Supp. The fact that “no of their DMRs.” court in this district far ha[d] thus found a heavy burden,” defendant to have met this upon by C. The Authorities Relied ALC id., however, preclude did not possibili- ty ALC first counters the ar- of the defense as a matter of law. guments by citing number of cases from Court, Atochem in discussing Elf district courts within this Circuit that have reasoning in Upjohn, quotes Chesa- recognized-either explicitly implicitly- or peake Bay Foundation v. Bethlehem Steel availability laboratory error de- Corp., F.Supp. 440, (D.Md.1985), “ fense. While no defendant these cases which ‘[gjiven heavy stated em- actually past summary has made it phasis accuracy the Act and the clear judgment stage laboratory based on the Congressional policy that DMRs should be defense, error lack of success has purposes, used for enforcement the court been due to finding district courts accept will not claims of inaccurate moni- ” defendants genuine failed to raise a issue Atochem, toring aas defense.’ Elf of material fact as to the existence of a F.Supp. at 1179. The Atochem Court Elf error, laboratory and not because the de- agreed that places “the Act the burden of precluded raising fendants were from accurately monitoring pollu- the levels of defense as a matter law. squarely tants their effluent on the holders, permit shoulders of and that we Group, Public Interest Research Inc. obligation,” must hold them to that but it America, Inc., Atochem North Elf ultimately held that (D.N.J.1993), F.Supp. 1164 a similar case of potential overreporting came agree [Upjohn before while we with the court ] District Court. The defendant claimed that it is inconsistent with the structure that errors in laboratory testing purpose had Act permit to allow overreporting resulted in the escape liability altogether of toxic dis- holders to charges. Split error, sampling laboratory over a six-month basis find accurate, period large revealed discrepancies laboratory be- more where error shown, tween the defendant’s results and those has been to hold a defendant laboratories, though outside no con- monitoring liable violation rather pattern sistent could discharge be detected those than a violation.
174 Atochem, (empha- summary judgment a motion based F.Supp. at 1179 vived Elf added). laboratory a error defense. See id. sis Interest Research Similarly, in Public D. Discussion Industries, Inc., 757 v. Inc. Yates Group, reasoning the of the We find Elf (D.N.J.1991), 438, the Court F.Supp. The persuasive. Atochem Court violations laboratory the error expressly recognized discharged alleged that at issue here defense, may that “DMRs noting of in violation of the terms pollutants lia- establishing when admissions deemed violations, prove to these permit. order motions,” summary judgment but bility necessary for the to government it was liability. of proof are conclusive did fact violate the establish that ALC circumstances, held that under some Court permit permittee reports If a that terms. may liability at the “defendant avoid a limit, report permit it has a the violated of summary judgment stage on the basis discharge sufficient data in DMRs.” Id. While inaccurate but neither production, burden heavy bur- recognized the Yates Court also any regulation itself of which CWA nor we prove laboratory on the defendant to den aware such a conclusive.2 report are makes “ error, showing that a ‘errors stated The trier of fact must still be convinced performed tests which in the actual permit was in fact violated. Evi ” may de- permit violations [sic]’ showed inaccurately that reports dence overre judgment Id. summary feat a motion. ported discharge certainly level of (quoting Student Pub. Interest Research that relevant show no violation occ Inc., Polymers, Group, Inc. v. Tenneco urred.3 (D.N.J.1985)). Thus, F.Supp. ultimately the Court sum- granted while government stresses fact mary judgment against defendants liability provisions of the civil on the fact that the letters the based cover regime liability, create of strict CWA but submitted to the were too defendant argument misses the mark. Strict they merely asserted speculative liability government relieves rea, the defendant “felt” and “believed” obligation show mens not the actus Fuels-Utah, See, laboratory errors had occurred-it Inc. e.g., reus. W. clearly implied Safety that had the factual situa- Fed. Mine & Health Review (D.C.Cir. Comm’n, different, tion been Yates have sur- F.2d could 713-14 any question permit- 2. The before us is whether a call to our attention administrative inter- permit discharges tee violates its if its in fact pretation to she that deference which claimed comply with the terms of CWA but its government brought was owed. has the Nor erroneously reports permit indicate was any interpretation to such our attention after law, pure question violated. This is a argument. brief, gov- plenary. our review is In its that, argue interpreting did ernment "laboratory 3. We use the term error defense” provisions any the relevant rele- Act or opinion in this because the term has been regulations, give degree any vant should prior parties used in cases and is used any deference to formal or adminis- informal here, important but it note laborato- interpretations regula- trative Act or ry error is not an affirmative defense liabil- By any argument failing to make tions. such Instead, ity. laboratory evidence of error is brief, government any in its waived con- simply ques- that is to the evidence relevant Moreover, based when tention on deference. require- permit tion whether a violation questioned counsel ment in fact occurred. point argument, on this at oral she did not *11 1989). case, present In the context of the the archetypal violation, Clean Act Water liability means that the is vio subjects strict CWA discharger to strict liabili- permittee discharges pollutants if a ty.” lated Pozsgai, But in liability strict was permit, regardless in violation of its of the imposed upon based discharge, unlawful permittee’s liability mens rea. Strict does not the report mistaken of a discharge. permittee may not mean that a be held seems to be aware of this violating permit liable for its even if it does difference when it argues that strict liabili- in fact do so. ty imposed should be on reporting require- ments, conjunction as it writes about government’s policy argu While reporting requirements “CWA’s certainly ments are forceful in the case of imposition liability permit of strict vio- permittee reporting a under levels of toxic added). So, lations.” (emphasis while the claiming laboratory waste and then a error CWA unambiguously imposes strict liabili- defense, unpersuaded they we are ty for unlawful discharges, it is no in a prove compelling case like this where means obvious that a liability similar strict permittee alleges laboratory regime imposed has been faulty report- reporting error resulted in the over ing. public poli levels of toxic waste.4 From a a cy perspective, polluter should not be fact, In the existence of a mechanism to given opportunity underreport lev correct erroneous suggests op- DMRs waste, thereby els of toxic in dumping 122.41(1)(8) (re- posite. § See 40 C.F.R. then, permit, excess of its when quiring permittee a who becomes aware of cry “laboratory caught, kept error me any inaccuracy in a promptly DMR to knowing from that I was violation!” EPA). notify the regulation pro- That bar, opposite appar But the case at mulgated pursuant to the Administrator’s ently conducting occurred: ALC was 1318(a) authority § under 33 to im- U.S.C. sampling reagent but a contaminated used pose reporting requirements. Since 33 laboratory’s analysis the ALC was caus administrative, § U.S.C. 1319 authorizes ing laboratory systematically to ov- civil, penalties and even criminal for viola- erreport the amount of toxic zinc that was § tions of the failure to correct an dumped into the water. We fail to see independent inaccurate DMR is an viola- what incentive could have had to regulations tion of the CWA thereun- overreport dumping how much zinc it was very der. Moreover the circumstances into the river when it knew that such that support laboratory would error de- would result in amounts fines. We do not likely support finding fense would also assigning believe that a scheme lia strict monitoring of a violation. See 40 C.F.R. bility discharge violations the case of § (j). 122.41 In light of these direct sanc- sense, overreporting errors makes nor do DMRs, wanting tions inaccurate we find infer the CWA that such was government’s argument that CWA Congress’s intent. provisions addressed to actual discharges ought surrogate
In
citing
Pozsgai,
United States v.
be made
enforcers of
(3d
Cir.1993),
sum,
reporting requirements.
F.2d
govern-
bar-
correctly
ring
laboratory
ment
discharge
asserts
the assertion of
error
in compliance
permit
is not
with a
“is defense seems unfair and at odds with the
situation,
underreporting
permit-
liability provisions
In the
irrelevant under the civil
attempting
laboratory
tee would be
to use
of the CWA.
rea,
error to show that it lacked mens
which is
*12
CWA,
did
especially
specifically,
in a
while the District Court
plan of the
overall
alleged
laboratory
the
have
of ALC’s
such
this where
labora-
available some
case
overreporting
having
rather
in
tory
penalty phase,
error caused
error
the
evidence
underreporting.
than
that ALC
liable
already determined
was
claims,
discharge
for
this after-the-fact
arguments
the
We have considered
penalty
of the evidence for
consideration
Oil
and the Union
government
the
pre-
not
the
in
purposes does
cure
error
laboratory error
defense
recognizing
that
in the
cluding
laboratory
the
error defense
practices and under-
sloppy
would reward
trial.
liability jury
Arguably the District
self-monitoring program by giv-
mine the
laboratory
ALC’s
er-
Court’s evaluation of
an invitation to wait until
ing companies
phase strength-
in the penalty
ror evidence
they
arguments
But these
do
sued.
it was
to
argument
ens ALC’s
that
entitled
is
overreporting, which
almost
apply
not
to
jury
have the
evaluate such evidence be-
certainly involuntary.
suspect
alsoWe
cause,
doing
what the District Court was
rare, for
overreporting
only
that
the
credibility
the
to that evi-
to assess
pound-foolish
per-
penny-wise
most
(“[I]t
laboratory
is not
that
dence
credible
expose
mittees
to the
of a
would
itself
cost
....”),
persist
normally
error
would
(as here)
any
if it
litigation
decade of
had
jury function.
up
im-
clearing
chance of
matter
with
proved laboratory testing and amended
did
Since
District Court
not consider
Concomitantly,
reports.
NPDES
we are
sufficiency
laboratory
error defense
by
government’s argu-
underwhelmed
it,
argument
light,
in the
this
proper
will
permitting
ment that
the defense
add Court,
in the
should consider
defense
bottom,
litigation.
to
At
time NPDES
we
first
instance. We will therefore vacate
efficiency
should
do
believe
over-
laboratory
so
remand
error
Thus,
in
ride fairness
administration.
adjudicated
can
defense
be considered
validity
we
gainsay
while
do not
respect
with
to the
claims
it affected.
that,
government’s argument
consistent
requirement
Act’s
accurate
Penalty
III. The
Calculation—
DMRs,
courts
treat
self-reporting,
should
Economic Benefit
by
which
the discharger,
must be certified
to
Objections
A. ALC’s
as admissions that are
estab-
sufficient
Penalty Assessment
CWA,
liability
lish
under
hold that
presence
DMRs
of certified
does not
penalties
The assessment of civil
preclude
laboratory
error defense
these
as sought
violations
the United
overreporting.
cases of
1319(d).
governed
§
States
33 U.S.C.
1319(d) provides that
Section
the violator
argued
has
of a
permit
pursuant
issued
the Act
if
laboratory
even
is recog
defense
subject
shall
penalty
a civil
not to
nized, there is
evidence in
insufficient
this
$25,000 per day for each
exceed
violation.
it.
support
record to
The District Court
provision
This
further
states
appear
position
did
in a
endorse
assessing
penalty,
the court shall
post-trial opinion: “Nothing
prof
factors:
following
consider
testimony
or
persuades
fer
on this issue
solely
court that these violations arise
violation
viola-
the seriousness of the
or
(if
laboratory
tions,
But that
any)
error.”
state
the economic benefit
re-
violation,
sulting
any history
ment followed a trial at
from the
which
laborato
violations,
ry
good-faith
error defense had
excluded.
any
been
More
such
efforts
comply
applicable require-
by weighting
with the
the cost of each source of
ments,
penal-
impact
the economic
proportion
funds
of the total market
violator,
ty on the
and such other mat-
Ross,
value of the firm.” Stephen A.
Ran
justice may require.
ters as
Jaffe,
dolph
Jeffrey
W. Westerfield &
Cor
(6th ed.2002).5
porate Finance 932
considered each of
Id. The District Court
penal-
these factors
connection with the
*13
ALC
that
asserts
the District Court’s
ty determination. The Court found ALC’s
economic benefit calculation did not “level
It
violations of the CWA to be serious.
field,”
law,
playing
required by
the
but
questioned the level of ALC’s commitment
imposed
penalty.
rather
a severe
ALC
obligations imposed by
to the
the Act. It
also contends that the District Court’s cal-
the economic benefit to
to be
found
ALC
apply
culations failed to
principles
other
considerable,
primarily
terms
the
(1)
law,
required by
including that
expendi-
(inad-
stemming
avoided cost
from reduced
tures made and
included
the economic
equate) staffing at its wastewater
treat-
calculation
directly
benefit
must relate
plants,
delay
plant upgrade
its
in a
at
ment
(2)
violations;
costly
least
method
Vandergrift facility,
and a number of
compliance
should
in calculating
be used
projects.
other
totaled
smaller
The Court
(3)
benefit;
economic
and
economic benefit
$4,122,335,
the economic benefit at
and
only
calculations must be based
periods
$8,244,670
ultimately doubled it to
as the
reject
of non-compliance.
argu-
We
penalty.
final
See
note 6.
infra
apply
ment that the District Court did not
imposition
penalty
The
of a
under
proper
legal precepts.
Rather
1319(d)
subject
§
is
to the exercise of a question
application,
is the manner of
and
district court’s discretion. See Tull v. whether the District
clearly
Court made
States,
412, 426-27,
United
481 U.S.
107
fact findings
erroneous
which skewed the
(1987).
1831,
Many findings of the District Court’s argues purpose of the eco supported, unchallenged appeal. component penalty nomic benefit primary issue contested here relates playing to “level the economic field.” We to economic benefit-i.e. the Court’s use of agree. States v. Mun. Auth. See United government’s experts’ computation (3d 259, Township, Union 150 F.3d 263-64 weighted average capital cost of Cir.1998) (“WACC”) Dairy]. Dean [hereinafter as the interest rate to use to words, bring money purpose prevent other is to forward to the judgment party violating gaining the CWA from an date. WACC is defined as “the average capital advantage against competitors, cost of unfair on the firm’s exist- activities, ing projects prevent profiting .... calculated and to it from from its course, funds") general only extremely 5. Of so of ... definition is "cost can useful; moving subject dispute complex from the definition to as this case so broad (in particular establishing aptly the actual numbers demonstrates. 178 established, Duffryn appro- an figure Termi- Once that See Powell
wrongdoing.
government, on
nals,
benefit
priate
F.2d
calculation
economic
hand,
penal-
CWA
other
submits
of mon-
reflect the time value
should also
“promote
immediate
are intended
calculation,
ties
ey.
In order to make that
future violations”
compliance”
“deter
an
rate to de-
“apply
court must
interest
regulated
and other
enti-
by the defendant
present
of the avoided
termine the
value
Earth, Inc. v.
Laid
ties. Friends of
delayed
or
costs.” United States
Smith
Servs., Inc.,
167, 185,
528 U.S.
law Envtl.
(4th
Foods, Inc.,
F.3d
field
(2000).
L.Ed.2d 610
S.Ct.
Cir.1999).
Herein lies the crux
Therefore,
government agrees
while the
Dis-
disagreement: ALC contends
analysis is de
the economic benefit
high
rate so
trict Court used
interest
much money
to calculate how
signed
rather
than
punitive
effect was
*14
gained by failing
spend
illegitimately
con-
government
whereas the
“leveling,”
on
appropriate amounts
environmental
the
tends that the interest rate used
the
agree
does not
the as
safeguards,
entirely appropriate
was
District Court
stop
a
penalty
of
need
at
sessment
yielded
a result
well within
that was
view,
point
In our
the latter
figure.
ad
the Court’s discretion.
Act,
aspect
a different
as
dresses
in the
explained
margin.6
Adopted
The Interest
C.
Rate
Putting
way
ultimate
in
aside the
by the District Court
the result of the economic benefit
which
Court,
arriving
in
The District
at
might
employed,
calculation
such a cal
assessment,
penalty
adopted
the eco
intended,
base,
identify
culation is
at its
analysis proffered by
govern
nomic
the
by a
the benefit realized
violator from
submission,
alleged
ment.
In that
the
delayed expenditures
comply
with the
economic
from each vio
stemming
benefit
The economic benefit calculation
CWA.
lation
computed
or that
was
forward
the
spent
starts with
costs
should
spent,
February
date of
compliance.
have been
to achieve
violation
that,
1319(d)
govern-
§
prescribe
specific
6. The CWA does
a
found
while
determining appropriate
pen-
method for
civil
trebling
ment
a
advocated
of the economic
Dairy,
alties for
In Dean
we noted
violations.
benefit,
doubling
appropri-
a
more
would be
“top
ap-
that some courts use the
down”
ate
case for
under the circumstances of the
proach
penalty
which
in
the maximum
is set
$8,244,670.
Dairy,
penalty
total
of
In Dean
($25,000
day
per
at
of violation
the times
doubling
approved
we
economic
of
benefit
here),
appropriate
as
relevant
and reduced
possible
assessing
penalty
as a
method
considering the six enumerated elements of
that,
doubling
stating
even after
of eco-
factors,
1319(d) mitigating
§
as
while other
benefit,
barely
"penalty
nomic
was
of
9%
employ
up” approach,
courts
the "bottom
statutory penalty
maximum
which
established,
which economic benefit is
Dairy
subject.”
Daily,
Dean
was
Dean
1319(d)
remaining
§
five elements of
bar,
statutory
F.3d
In the case
at 265.
at
adjust
upward
figure
used to
or down-
penalty
maximum
that could have been lev-
Dairy,
ward. Dean
F.3d at 265.
Dean
million,
against
$28.05
eled
was
count-
Dairy, we held that the method used in assess-
$25,000
1,122
ing
days
for each
of
ing
penalty
the civil
is best left to the trial
$8,244,670
approximate-
violations. While
bar,
court’s discretion. See id. In the case at
million,
ly
larger pro-
$28.05
a much
29%
up”
followed the "bottom
portion
approved
Daily,
Dean
than
9%
approach. Having
figure
at a
eco-
arrived
$4,122,335,
are satisfied
District Court was
totaling
nomic benefit
District
well within its discretion to assess such a
analysis
Court then conducted a detailed
remaining
penalty
factors
enumerated
in this case.
judgment)
date of the
at a
1. The
(roughly the
Contentions of the Parties
annually,
of 12.73%
to arrive at a
rate
ALC characterizes the 12.73%rate as “a
$4,122,335
at
total economic benefit
theoretical,
(denom-
risk-adjusted
rate
judgment.
time of
inated
EPA
weighted average
as the
The District Court derived this rate
WACC’),
capital
cost of
or
based on broad
proffer
from the
witnesses
averages
capital
across the U.S.
markets.”
Gary Amendola and Robert Harris who As the foregoing explanation suggests, this
explained
steps they
the three
took to
characterization
generally
accurate.
First,
they
calculate the WACC.
deter
using
ALC contends that
such a hypotheti-
“A,”
rating
mined
ALC had
debt
cal rate of interest
error of law
(“S
assigned by
& Poor’s
&
Standard
because ALC had presented evidence of its
P”). Then, they
typi
researched what the
which,
actual rate of return
capital
monthly
cal
interest rate was for A-rated
trial,
time
showed that
year
computed
bonds
each relevant
the average rate of return on capital for
yearly
adjusted
averages. This rate was
parent
ALC and its
company between
advantageous
to account for the
tax treat
1990-2000 was 5.7%. This fact was uncon-
payments
corporate
ment of interest
tested, and thus ALC submits that
Second, they
debt.
calculated the cost of 12.73% rate did not achieve
legal pur-
*15
equity
They
as follows:
started with a 30- pose
“leveling
playing
economic
year treasury
They
a
bond as
baseline.
field,” but rather was used to exact a
“beta,”
up
company’s
next
looked
penalty “reflecting
severe
not the time val-
a
which is measure used to evaluate the
money
ue of
nor ALC’s benefits from re-
particular
relative risk of a
for an
stock
funds,”
taining
rather
in-
theoretical
but
equity
Finally, they
investor.
assumed a
averages
vestment
that indisputably were
generic
premi
value for the market-risk
by
not achieved ALC.
premium that a person
um-the
would de
that,
ALC submits
instead of the 12.73%
than in
mand
invest
stock rather
a
rate, one of four alternative rates should
free)
(risk
treasury
At
instrument.
have been used:
point, they multiplied
by
mar
beta
(1)
(6%)
premium,
ket-risk
and added an “interme
statutory
interest rate
premium”
years
diate stock
for the
before
(2)
represented by
the risk-free rate
merged
entity
ALC
with another
and be
treasury
during
short-term U.S.
rates
bigger,
company. They
came a
“safer”
period
the relevant time
30-year treasury
then added this to the
(3)
average
the actual
rate of ALC’s
equity
bond rate to
at an
by
arrive
cost
capital
return of
from 1990-2000
Third,
year.
they combined these cost of
(4)
average
the actual
rate of ALC’s
by
equity
taking
debt and cost of
measures
capital
return of
from 1990-2001
them,
weighted average
based
proportions
equity
suggested by
relative
of debt and
Each of the rates
re-
ALC
capital
year.7
structure for that
in approximately
sults
the same interest
brief,
government
specific
actually
In its
mischaracter-
than was
the case. As we
experts’ testimony
explained,
experts
izes its own
and states that
have
seem to have re-
primarily
general
the WACC was calculated
"first determin-
lied
market numbers
ing
similarly
companies
the rate at which
over a
ALC borrowed
funds
situated
ALC
during
period.”
repre-
long period.
the relevant
time
We
This
do not know
reasons for
mischaracterization,
implies
government
experts
sentation
we
but
figures
disapproval.
relied on
that were much more ALC
do note our
6%,
$8,244,670.
pay
government
The
5.2% and
rate, hovering between
it
discrepancy
to this
between what
points
half the rate that
than
is less
which
actually
for and
the Court
did
actually used.8 ALC adds
asked
what
Court
District
really
Court
does
proof
has
re
the District
WACC
been
the “theoretical
have,
have,
great
amount of
applied
compa
should
consistently when
jected
in determining
types of
achieving
discretion
these
and industries
nies
penalties.
rates of return.”9
theoretical
such
responds
a num-
government
government
points
The
also
to the deci-
The
First,
government
arguments.
approved
other
that have
ber of
sions of
courts
economic benefit
correctly notes
the use of
to discount economic
WACC
In Dean
precise.
calculating
penalties,
not be
need
benefit when
CWA
calculation
Smithfield,
that economic bene-
where the
Dairy,
recognized
particularly
Foods
Court,
expert testimony,
capable
ready
crediting
determi-
“may
fit
District
nation,”
discount the defen-
gave
Court
“the district
used
WACC to
discretion,
economic benefit.
United
penalty
award
wide
dant’s
See
court’s
Foods,
Inc., 972
though
represents
approxima-
States
even
Smithfield
(E.D.Va.1997),
Tull,
n. 17
F.Supp.
&
(citing
tion.”
In of this we think that view problem government’s first is the cal inappropriate it would be for us to decree culation of the WACC. That methodology which should be used since in calculation situation, any given “leveling playing ALC-specif relied on values that were not might readily field” be more achieved with ic. using yield Instead of the actual Therefore, one or the other. we think that issued, govern bonds that ALC had methodology the choice of should be left to experts ment computed the WACC us sound discretion of ing yield on Standard & Poors A-rated case, however, In Court.10 it is not bonds. P using figure While the S & clear that the District Court was aware might well have been a approx reasonable range options or considered the avail yield, imation of ALC’s bonds’ a more ac able. easily curate calculation could have been by using figures specific achieved to ALC’s a. Economic Benefit as Measured bonds.11 Capital the Cost of above, problem The second is the As noted economic can benefit entity’s capital. application averages measured cost of of the WACC. WACC provision Gwaltney, F.Supp. 10. We note a from the field of trusts As in at 1558-59 & n. that enables the District Court to exercise its the choice is within the discretion of the *17 Court, choosing District and we are confident that it appropriate discretion in mea- give equities will due consideration to the assessing liability sure for a trustee's in the selecting appropriate involved in an measure case of a breach of trust. The choice to make Indeed, of economic benefit. we do not even (i.e., here both cost measures and actual re- permis- hold that economic benefit is the sole possible ways valuing turns are of economic assessing approach penalty; to a there sible benefit, adopted?) so which should be resem- may ways. well be other Given this variabili- (Second) bles choice. The Restatement ty, disagree with the dissent’s contention (in provides disjunctive): § of Trusts holding saps that our too much discretion trust, a If trustee commits breach of he from district courts in cases under the CWA. chargeable is with: (a) any depreciation loss or in value of contrast, tell, as far as we can the cost resulting the trust estate from the of breach ALC-specific equity was as calculation trust; or reasonably value could be achieved: The for (b) any profit by through made him ALC-specific, beta seems to have been and the trust; breach of or figures computa- other that entered into (c) (the any profit premium which would have accrued and the “inter- tion market-risk premium”) to the trust estate if there had been no mediate stock are not their company specific. breach of trust. nature a processing industry of a is stable compa- the basis food are constructed on (that is, industry attract existing capital companies probably structure ny’s where equity). debt and proportions capital new on terms similar to their exist relative a company’s figure based industry, A WACC ing capital structure. The steel given at a time is capital existing structure contrast, in highly has been volatile not, support, necessarily further without foreign rife with disloca competition, stiff marginal or cur- company’s the same as tions, Indeed, bankruptcies. as the (i.e., capital that time what rent cost of at noted, industry going District is capital) it would cost obtain additional and more through restructuring,” a “brutal come might new in a dif- capital because than steelmak twenty-five United States equity. mix of See Aswath ferent debt and sought protection ers bankruptcy have Damodaran, Corporate Applied Finance Thus, a in ALC’s company since 1997. (“In (1999) estimating [the current cost have, in position may not at the times WACC], in using we have capital on the question, capital been able raise in quo conceded the status terms of sense struc existing capital same terms as its mix, financing since we have estimated the cannot) (indeed, ture. need not re We existing capital at the mix. It is cost this; solve for our it is purposes, but firm, entirely possible changing that a enough there was evi insufficient mix, capital.”). could lower its cost of say dence for the District Court to good to be a ap- WACC is shown Unless existing capital rep structure was marginal for or proximation current of the terms which new resentative capital, light little cost it sheds on how Thus, capital be raised. if the eco would expensive it com- would have been nomic is benefit to ALC to be established pany go capital, to the market for its measure, eost-of-eapital measure diverting instead of funds that should have marginal to use current cost is ALC’s or gone improving pollution controls. years question.12 of new in the in capital above, As noted and the this appear Some courts have endorsed District Court relied on Foods. Smithfield See, approach. e.g., Gwaltney Smith- But, upon analysis, closer Smithfield (“[T]he field, F.Supp. actual help government. Foods does not Gwaltney paid rate itself on borrowed suspect There are reasons to funds a more ... accurate basis (in processing industry food which Smith- Gwaltney’s determining benefit economic operated), may field the WACC have been delay.”). entirely approximation appropriate possible It benefit, ap- of course Smithfield’s economic whereas (in proach make an off might offender worse industry conditions steel which pro- than operates) radically under the WACC different. example, a precisely, may posal. company More For dire have been that *18 may financial marginal that the a straits well have a Foods WACC was Smithfield (offered by cost of good approximation capital for the terms on which lenders who see investment) raised; money currently could have been it as a high-risk that exceeds to, by looking 12. This could be can the established for establish to District Court's satisfac- example, any capital actually the cost of particular tion that WACC is-in this case-a times, by ALC by raised at the relevant or the good capi- approximation marginal for cost of expert opinion of an investment banker re- tal, accepted could be a surro- then WACC as garding the terms on which ALC could have gate marginal capital. measure of the cost of course, capital. expert testimony raised Of if not, anomaly. industry com- ALC the steel This no For were is its WACC. relevant, capital enjoying are for up hard times stellar returns. panies market, Indeed, in the it is above, afford to raise it cannot it uncontested as noted the tempting forego all too doubtless a capital at trial that ALC had return on costly pollu- improvements sometimes half that was less than the 12.73% rate required by are tion controls that view, by used District Court. On this regulations. and EPA But such CWA any advantage enjoyed that ALC over its law. must still be held companies competitors avoiding by the cost of CWA essence, is, in allow otherwise To do compliance by measured is the return to take out low- polluters capital-strapped actually realized on retained funds against loans the environment. interest it en- might or risk-free return have joyed using those funds. think that We no view on what a of course intimate We is capital quite the return on a viable may respecting ALC’s sit- develop remand field, government’s leveling along means the playing in the 1990s. The uation (then) costs for P proffer marginal shows debt S & with the or current cost of experts’ in the 6.68%—10.06% A-rated bonds were capital. significantly in the That is
range 1990s. figure than the relied lower 12.73%WACC Observations About the Other Moreover, in by on District Court. Analysis Court’s District also years, figure which would recent potential with problems There are other calculations, rates been interest have calculation, which re District Court’s very low. The record does reflect methodology by lied on provided strength, may ALC’s actual financial government’s experts. It not) (or had) appears (or may have have a also government’s experts annual es computed good rating throughout credit relevant years timates for of WACC each of period. 1990-1998, up and came the 12.73% with b. Economic Benefit as Measured figure taking the mean.13 by arithmetic
by Return Actual savings from violations Since the different accrued different over a several dates talking have so far about meas- We been an year period, questionable it is whether economic of additional uring benefit when average appropriate, interest rate is capital capital cost to obtain that year-to-year interest rate estimates option But to use elsewhere. the other is mini used only known and could be capital rates of to com- actual return experts.14 mal effort additional pute economic benefit. practice using potentially problematic importance leveling cited experts span time large “in mean rate over field same interest playing the economic obvious, government’s experts’ in the present It is industry.” example, clear, slight, referring any we are 14. While correction will be 13. To be "mean” might precision the (which, interest of District "weighted average" not to WACC whether, average to be if also consider sorts) to the is a mean but rather further used, procedure to use a would be the correct taking step a whole the mean of series geometric (computed th root of mean as the n (one figures years each of the WACC items), product percentages of n since the period). objection, no We have the relevant multipli- applied in involved are consecutive *19 suggests, dissent to the use of the as the Damodaran, Corporate Applied cations. See economic WACC formula to assess benefit. at 69-70. Finance up opinions, this wound a district court has discretion to happens, As it report.15 fig theoretical WACC credit one over the other. But this discre- hurting ALC: (15.85% in 1990 early 1990s adopt opinion ures from tion is not a license to 1992) are the and 13.95% methodology, based on unsound whatever really but have no group, highest its source. benefit conferred bearing on the economic upon analysis govern Based our Thus, the aver by post-1992 violations. expert’s methodology, ment’s we are un toward the less- age WACC was biased convinced that the use of the 12.73%inter estimates from the higher relevant WACC purpose est rate achieves the stated early 1990s. field,” “leveling the playing economic nor that the Finally, we note we sure that it much bears connection correct its assertion at unquestionably meaningful to a measure ALC’s cost of any computation argument oral must (much capital capital). less return on its use the same discount rate for both for- We therefore must set aside the computations during
ward and backward pro calculation and remand for further period. example, the same For would ceedings rate, respect to the interest clearly inappropriate to discount all eco- fully open possibilities to the that the rec nomic benefit backwards to a uniform date support higher, ord on remand will low rate, using one then use different er, substantially penalty. or similar We carry rate to the value forward to the date among will not choose the alternatives we judgment. (or suggested suggested by have those ALC) above; in the discussion rather we 4. Conclusion Court, shall leave it to the District after are, course, acutely We aware that we receipt par of further submissions both review the District Court’s interest rate ties, to decide what alternative rate is best discretion, determination for abuse applied to developed the circumstances that its determination need not be exact. the record on remand.16 Dairy, See Dean F.3d 264-65. Our mean, scope deferential of review does not D. The District Court’s Determination however, that we cannot intervene when Compliance of Costs of finding District Court makes a 1. Costly Introduction-The Least flawed, methodologically if, even under Compliance Method of theory, penalty figure such it ultimate- ly plausible. arrives at is The second basis on which ALC asks us to overturn the District calculation
In the Court’s “given highly dissent’s view our review, of economic benefit is its contention that deferential standard of clearly did not the District Court erred in crediting calculating err money amount of it would have witness over ALC’s witness cost ALC adopting changes the WACC to calculate eco- institute the that would have course, nomic benefit.” present- compliance requirements Of when led to with the brief, ed with two conflicting expert permits. sound but argues Moreover, 16. The District practice unique Judge originally is not to the who heard resigned use and decided this case has of WACCas a measure of economic from the bene- fit; Accordingly, parties bench. will doubt- regardless it is an issue of the method develop less have to a record for the edifica- used to derive the interest rate. newly assigned judge. tion of the *20 (which permit that pretreatment issued for up EPA with came numbers however, that, Court) ALC, kinds claims Octo- and the site. adopted were 1993, shortly considerably Vandergrift ber of before proposed it were of solutions ceased, that, it began ac- and to considering violations installed especially overpriced, ALC, a diversion to the already operate it fixed the tank connected had cording to money discharge piping leading much and could to the Vander- less problems diver- facility WWTP outfall. This implemented grift the solutions it had that show $150,000 tank cost no than sion more already worked. ALC, According install. buy and is wheth question The threshold shortly af- pretreatment stopped violations law, Court er, matter of the District as a tank ter the installation of the diversion using the economic benefit must calculate average although monthly there were two This costly compliance. method of least in No- daily four maximum violations and not to have been appear does question which ALC vember December any Appeals. addressed Court ALC “start-up problems.” attributed addressed Courts that have Those District 15, 1993, starting contends that December that calculations should hold the issue single violation ALC then not occurred. costly method of on the least be based that, government’s argues adopting the See, e.g., Smith Gwaltney compliance. $600,000 proposed project to solve (holding field, F.Supp. at 1563 n. 25 compliance, bring into problem ALC could economic benefit calculations clearly Court erroneous the District made “perma- expensive, based on more not be findings of fact. “in- expensive when a less nent solution” already had achieved com- solution” terim is, however, another side There Inc., Steel, United States v. WCI pliance); above, were story. As noted there (N.D.Ohio 1999) (finding F.Supp.2d 810 after “start-up violations” several so-called testimony expert the defendant’s credible installed, tank the diversion measures possible compliance regarding Decem- claim a clean record until cannot calculating economic based benefit 15,1993. ALC claims that District ber of com- expensive less method significantly have December should used the govern- proposed than that pliance date because compliance 1993 date as decisions expert). We find these ment’s vio- reported last pretreatment that is the and hold that economic benefit persuasive, in Au- upgrade lation before WWTP least cost- should be based on the analysis However, gust of 1994. However, compliance. contra
ly method of Amendola, Gary explained he expert, contentions, to us appear does August compli- to use 1994 as chose District Court took a different Court) (as did be- ance date approach. installed Octo- cause the diversion tank to address the 1993 was sufficient ber
2. The Plant Vandergrift ex- Amendola Vandergrift. problem report- the fact that ALC had plained instances ALC cites two main half first during no violations miscalculation of benefit. ed alleged the diversion plant, not establish that first, relating Vandergrift to the 1994 did compliance a sufficient measure calculation tank was from the District Court’s stems facility operat- $600,000 previously had project because incorporated which reporting at a for months time without posited would ed government’s expert chose The District Court compliance any violations. brought have ALC into *21 testimony govern- only of the fact that ALC identifies one internal to credit 107,” monitoring point, tank expert the diversion “Outfall associated ment’s $476,090 adequate prevent re-lining project. The not have been “would Amendola, however, government expert, A decision to credit the all violations.” opined relining project testimony expert of one witness was re- expert quired to to deference. cure violations associated with over another is entitled See Joiner, 136, 143, Line, Number 90 Anneal and Pickle which Elec. 522 U.S. Gen. Co. (1997) 512, discharged through L.Ed.2d 508 Outfall 007. Outfalls 118 S.Ct. monitoring points that a assessment 107 and 207 are internal (holding District Court’s testimony discharged through accorded “the Outfall 007. expert be reported hallmark Since ALC violations at deference that is the of abuse-of- Outfall review”). through long discretion Under these circum- December after alleges it findings stances the District of fact ALC cured those violations with Court’s erroneous, clearly they maintenance efforts in were not must October therefore be left to stand. contends that the maintenance enough
efforts Outfall 107 are not Outfall carry day. next issue concerns It is clear that the District Court decid- $476,090project adopt government’s ex ed to framework necessary pert posited bring regarding monitoring ALC links be- compliance into cooling non-contact wa tween these different outfalls. view of ter violations at evidentiary dispute, Outfall 107. ALC con the bona its find- fide tends that ings clearly District Court’s economic were not erroneous and must benefit calculation which adopted fig upheld. premised clearly
ure was erroneous E. Non-Compliance Periods of findings of fact which misapplication led to costly of the least method of compliance complaint ALC’s final relating to the legal principle. argues ALC that viola analysis economic benefit undertaken tions at the outfall were limited to June the District Court is the identification of though October 1994 and that those viola period non-compliance. Obviously, tions were through resolved various main length period of non-compliance efforts, tenance including repairing cracks a direct relationship has to the calculation in certain trenches sumps. gov of economic longer period benefit: The ernment, however, points out that ALC’s non-compliance, greater the amount any brief does not contain record citation benefit, of economic higher and the indicating presented that it factual materi penalty. govern- ALC contends that the al to the District relating Court at trial to ment period miscalculated the of non-com- efforts, its maintenance and that to the that, pliance in adopting govern- extent that there is such evidence in the calculations, ment’s the Court calculated record, the evidence is limited to ALC’s purported economic benefit on Furthermore, pleadings. own govern lengths of time that bore no semblance to reported ment contends that viola reality. long alleges
tions after it it cured do not it necessary engage We find them October 1994. lengthy analysis of the various conten- This opinion difference of as to whether regarding periods tions of noncompli- more violations occurred stems from the ance and will set forth some of the factual holding of Having portion that case is that only margin.17 disputes record, daily average hold violations of the limits result reviewed the thoroughly *22 penalties only days in for the findings as to the number of the District Court’s that facility within month that the supported operated. non-compliance periods not, record, erroneous, however, clearly That does resolve decision by the were this case. must be left to stand. in leading authority area is the this Monthly Average Violations IV. Court of for the Fourth Circuit’s Appeals ruling the pretrial In a District that opinion Gwaltney. Gwaltney in held Court held: monthly a average parameter violation of a day Counting in Limine on constitutes a violation each
Plaintiffs Motion Violation, 242, month. No. The Court reasoned: Days of Doc. All violations
GRANTED.
does
di-
While the statute
not address
of defen-
monthly average parameters
rectly
average
matter of monthly
shall be counted
permits
limitations,
speak
dant’s NPDES
it does
in terms of
violation,
in number to all the
equal
violations
penalties per day
rather
in
month.
At-
days
per
the monitored
See
than
This lan-
penalties
violation.
Tyson
suggests
a
Legal
guage strongly
lantic States
Found’n
that where
(11th
Inc.,
Foods,
1128,
in
defined
a time
897 F.2d
violation is
terms of
Cir.1990).
a
period longer
day,
than
the maximum
for
penalty assessable
that violation
District
that
Court
ALC maintains
in
num-
should
defined
terms of the
ruling,
particular
in so
erred
days
ber of
in that time period.
that actual
improperly excluding evidence
(footnote omitted). The
days.
on fewer
Our modified however, I opinion. disagree, however, accord with Texaco. crafted applied only days. penalties de- nominated into a statute authorizes approach offered a conclusion that the District the WACC reasonable majority’s po- when it credit- an represented average its discretion because it Court abused and used expert during economist EPA’s tential investments made ed the rate to calculate interest expert’s it had use of the funds that it did the time rather than the benefit economic spend compliance. The District expert. The ma- by ALC’s presented Dairy’s rate also followed Dean endorse- Court the District Court com- jority WACC, writes ment used Smithfield because, in applying the error mitted clear Foods. rate, the so EPA’s 12.73% discount majority finds fault with the Dis- the economic benefit to vastly overstated analysis, noting gov- trict Court’s (“CWA”) Act Water vio- ALC of Clean ernment’s calculation of the WACC “relied it failed to level the economic lations that ALC-specific.” on values that were not view, my selecting field. playing Maj. majority Op. at 181. The also be- rate, the District Court acted 12.73% that, using average lieves rather than discretionary authority. within its squarely WACC, such as the should applied the actual rate it have have would I. years capital cost ALC to raise for the discussing the discount rate issue Before diverting it was when funds should discretion, I and the Court’s exercise Therefore, gone pollution have control. commenting proceed- think it worth on the majority concludes that ing by the District Court. The conducted *25 in using Court erred the 12.73% discount $8,244,670 penalty imposed on ALC came rate. three-day penalty hearing during after testimony
which the District Court heard
witnesses,
from
11
2 through
13
live and
II.
depositions. These witnesses included ex-
it,
IAs
see
the central
issue here is
benefit,
perts on economic
cost avoidance
whether the District Court abused its dis-
aquatic toxicology,
ALC’s Director of
in crediting
expert
cretion
one
over anoth-
Affairs,
Environmental
and officials from
er when it determined the interest rate.
Guard,
the United States Coast
the Penn-
many
We have noted
times that abuse of
sylvania Fish and Boat Commission and
highly
discretion is a
deferential standard
Pennsylvania Department
of Environ-
And,
stated,
of review.
we have
on numer-
testimony
mental
Expert
Protection.
was
occasions,
ous
that a decision to credit the
proffer
submitted
written
with live
testimony
expert
of one
witness over an-
subject
cross-examination. On the
of eco-
other is entitled to deference. See United
benefit,
nomic
EPA presented
testimo-
v.
States
Universal Rehabilitation Ser-
Harris,
ny
economist,
from Robert
an
who
(PA), Inc.,
(3d
657,
vices
205 F.3d
665
explained how he calculated the 12.73%
Cir.2000), quoting General Elec. Co. v.
presented testimony
WACC. ALC
Joiner,
136, 143,
512,
522
Pifer,
U.S.
118 S.Ct.
Dr.
proposed using
Howard
who
(1997),
191 working including and as a consultant to be accorded “the deference mony is major compa- re- to the EPA and several steel the hallmark abuse-of-discretion Joiner, 437; 41- F.Supp.2d App. 522 187 at I at Elec. v. nies. General Co. view.” 143, 118 required court 42. The District Court not S.Ct. A district U.S. every possibility. explore it “bases its As Su- its discretion when abuses stated, clearly finding preme Court has a district court on a erroneous opinion conclusion, fact, or an need not have an “exhaustive legal conducted erroneous See, possible of all law to fact.” La search” alternatives. improper application of Pharms., v. Holding e.g., Bank v. Daubert Merrell Dow Nat’l First Conn. Salle (3d XXIII, Inc., 579, 2786, 279, 597, 125 509 U.S. 113 S.Ct. Group, L.L.C. 287 F.3d (1993); Cir.2002). Indeed, L.Ed.2d 469 Kumho Tire “[i]n have said see also Co., Carmichael, 137, 147, reversal, justify a district court’s Ltd. order to U.S. (1999) 1167, 143 resulting must be 119 S.Ct. L.Ed.2d 238 analysis and conclusion ” (extending gatekeeping States v. Daubert’s obli- ‘arbitrary or irrational.’ United (PA), gation expert testimony). to all Rehabilitation Services Universal (3d Cir.2000), Inc., quot- 205 F.3d Still, majority protracted conducts Litig., R.R. ing In re Paoli Yard PCB theories, survey of economic considers (3d Cir.1997) (internal quota- F.3d specifically presented by ex- treatises omitted). re- of discretion tions Abuse Court, and de- perts before District error, inap- showing of clear quires disagrees cides view, our my given propriateness. discretionary Of Court’s determination. review, the standard of highly deferential course, always disagreement there will clearly did not err credit- District Court scientific, or in among experts concerning government’s witness over ing the economic, However, it case theories. to calcu- adopting the WACC witness fact Judge, for the District benefit. late economic finder, disagreements to resolve those expert wit- judging credibility Here, testi- considering after all of the *26 evidence, nesses, resolving conflicting the testi- mony, the District Court credited expert’s assessing weight of the and expert mony government’s economic testimony. nothing There is in the record WACC, stating “rep- that it concerning the ex- government’s indicate that here to company rate of a must resents the return and methodology not pert did use sound annually to to attract its earn continue support and noth- adequately opinion, his its current current investors maintain District Court ing to show operations. It a rate which is levels of is crediting opinion. clearly erroneous making commonly companies used majority’s as to which budgeting disagreement capital decisions.” 440, is not “appropriate” Proffer rate is more F.Supp.2d quoting at Harris interest (internal omitted); es justify I a This is quotations App. enough remand.19 at 6 light Dairy, where credited the true of Dean pecially at 47. The Court also District “may expert on that economic benefit testimony we stressed determination,” costs, noting years capable ready avoided that he had 30 not be field, court’s in the we accorded “the district experience environmental states, cost-of-capital appropriate majority example, that an "[i]n constitutes 19. commenting upon cost-of-capital measure Maj. Op. at of economic benefit.” measure [x.e., WACC], adopted by the District Court provide what hope guidance as to some discretion, capital current cost of for the relevant wide even of a award approximation.” years, majority suggests. as the He stat- represents though 264, Tull v. citing at United ed: 150 F.3d 412, 426-27,
States, 107 S.Ct. 481 U.S. that I is rate consider [The WACC] (1987). 1831, Surely the L.Ed.2d 365 a rate that falls proper represents government’s expert choice to credit the equi- rate and the between the risk free falls within this wide discre- over ALC’s I ty rate. The reason that believe that Indeed, Dairy went tion. the Dean is be- appropriate the WACC rate is say “[p]reeise economic ben- on to fungible. company’s cause cash may prove” be difficult to polluter is, efit to segregated That funds are not approximations and that specific purposes. used for Funds are “Reasonable will suffice.” 150 F.3d at economic benefit many ways and the used different Research quoting Public Interest company receives different returns for Duffryn Ter Group of N.J., Inc. Powell projects high each use. Some earn a (3d Cir.1990). minals, Inc., 913 F.2d rate of return. earn a low or no Others As when the District Court credited here impossible say It rate return. expert’s approximation one reasonable exactly how the funds should have another’s, it the economic benefit over act- spent example in this were used. been ought its ed well within discretion. We Therefore, appropri- I believe the most opinion our for that of substitute own average ate rate to use is the return the the District Court’s. all In company projects. earns on of its essence, average this is the return for attempt to fault District company. Court’s calculation of the WACC for App. IV 1009. The record evidence “rely[ing] on values were ALC- clearly shows that the District Court’s de- [ie., specifíc” using yields theoretical cision to use the supported WACC was yields], bonds issued rather than actual considerations, including, various as testi- fact, majority, in concedes that the District government’s expert, fungi- fied analysis ap- Court’s contained reasonable bility company’s of a funds and the varia- Maj. proximations. Op. at 181. It states a company ble rates return receives bond-yield “figure might that while the depending how it uses those funds. approximation well have been a reasonable yields, of ALC’s bonds’ a more accurate Further, the Court’s exercise of discre- easily calculation could have been achieved supported by tion is the case law. Dean *27 by using figures specific to bonds.” Dairy cites the Foods Court’s Smithfield However, Id. long as as the District favorably, use of the indicating WACC that “reasonable,” Court’s calculation was perfectly acceptable is a WACC inter- cannot find the Court to have abused its a in est rate for district court this circuit to Relying discretion. on theoretical values adopt calculating when economic benefit. rather than actual values to calculate the 266, citing 150 F.3d at United States v. does not WACC render the District Foods, Inc., 338, 972 F.Supp. Smithfield Court’s decision ma- “unsupported,” (E.D.Va.1997). as the 349 jority contends. majority’s The failure to find clear error government’s combing
The record shows that the after in record is evident sev- expert gave satisfactory explanation places. for eral For example, majority his decision to use the in this government’s expert’s WACC case criticizes the use of of, (instead example, marginal instead for or the arithmetic geome- mean of the enjoyed by spending money less mean) of the ALC compute an estimate trie facilities, Although 1990-1998. the District Court noted years for the staff its WACC “any correction that majority expert admits government’s comes to the WACC slight,” will assumptions that were favor- made two 12.73%, it is 12.71%, when opposed as First, in he included able to defendant. mean. using geometric calculated staffing costs time billed ALC’s actual Surely, a discretion- at 183 n. 14. Maj. Op. by by stopped maintenance workers who that results by judge a district ary choice having a main- facility, though even an higher than in an interest rate .02% stop tenance worker is not the same clearly viewed as alternative cannot be Second, having staffing. full-time erroneous. money calculations do not include [his] majority govern criticizes the also Leechburg at its saved ALC West all, at interest rate use of a mean ment’s facilities to en- Brackenridge prior and up hurting ALC.” asserting that it “wound Pa- try agreements of the consent with agree I do not Maj. Op. at 184. En- [Pennsylvania Department DEP Taking unduly punished ALC. calculation Protection], vironmental for all of interest rates average of the 7, citing at n. Amendola F.Supp.2d non-compliant in was years which ALC 17; Also, I at 41. in App. Proffer at perfectly acceptable a common costly calculating upgrade the least single figure arriving at method into brought compliance have would ben calculating ALC’s economic use when 1994, facility Vandergrift before at its disagree I years. those during efit noted theo District Court majority’s contention that “[t]he early 1990s figures from the retical WACC might pointed have States United the economic really bearing ... have no upgrade million considered a $1.8 by post-1992 violations” conferred benefit cost or the entire ALC 1988 they highest fig because simply upgrade million of the Van- of the $5.7 Maj. at group. Op. 184. ures of Treatment dergrift [Wastewater WWTP equally as from 1990 to 1992 are figures Plants], money argued that should 1998, as those from 1993 to relevant as than spent rather have been occurred in each violations CWA favor- approach But in an that is 1994. no to 1998. There is years from 1990 ALC, expert] government’s [the able to majority’s assertion support for the record costly upgrade the least calculated early from the figures WACC likely eliminated that would have than those 1990s are “less-relevant” $600,000 violations, provided Therefore, majority’s Id. years. later alternative. average WACC suggestion that 439, citing Amendola F.Supp.2d high towards numbers unduly biased 12-13; I App. at As with Proffer at Further, inaccurate.20 calculation, District Court ex- instances, WACC that, in it cred out some pointed supported here and its discretion ercised ways expert ited the *28 explanations. acceptable with its decision example, benefitting ALC. For up wound that however, figure at a Here, it arrived that calculating in the economic benefit average, particu- computing an degree when majority the to outlier also overstates 20. The appears and larly figure twice the same highest figures when deviate from which the range figures from figure 10.53% A the rest of in the calculation. rest of those not be considered a statistical 13.95%. would 15.85% 194 competitive ... in indirect benefits majority fails to ex- result ALC.
benefited in the compared companies decision falls within when how plain while its WACC comply discretion with the same field that do District Court’s 436; not. does at I at F.Supp.2d App. calculation 187 [CWA].” hypothesizes also majority appropriate have been as may WACC that Finally, majority asserts of economic benefit approximation an in its discretion District Court abused in company Smith- it was ALC as of a lower choosing the WACC instead in the of differences Foods because field using interest rate because alternative in each industries which volatility of the punish evidenced an effort to WACC Again, the standard operated. company calculating when the economic bene- deter discretion, abuse review is However, clearly fit. the District Court might have been decision another whether steps that there are two to the recognized Further, majori- “appropriate.” more approach penalty assess- up” “bottom proposition for the ty cites no authority step it second that ment and is the interest rate as using a theoretical deterring the geared punishing toward particular one in a to an actual opposed violator. The District Court stated: majori- industry clearly erroneous. The deterrence, goal To achieve Found., Bay Inc. v. ty quotes Chesapeake encompass must both Ltd., appropriate penalty Smithfield, F.Supp. Gwaltney of (E.D.Va.1985), the economic benefit defendant stating that through noncompliance, its Gwaltney itself obtained actual interest rate “[t]he ... punitive component funds is a more an additional paid on borrowed determining Gwaltney’s penalty accurate basis for into account the factors list- takes Maj. 1319(d). delay.” Op. from at economic benefit the second ed Section Without put into (ellipsis original). When by component, regulated those the CWA context, however, sup- this case does not nothing violating would have to lose it. In port majority’s position. Gwaltney, 439; App. at I at 54. F.Supp.2d Gwalt- plaintiffs computed calculation clearly The District Court was mindful ney’s delay using from economic benefit two-step process to be used when as- ten-year equity “the rate of return on sessing penalty, calculating first the eco- Foods, earned Smithfield Inc.—Gwalt- considering nomic and then benefit ney’s F.Supp. parent corporation.” penalty figure. factors to increase that that “[a]t 1559. The Court went on hold analysis, The Court followed the correct circumstances, least these the Court only punitive in the sec- taking measures actual interest rate believes 13%—the step ond when doubled economic Gwaltney paid itself on borrowed funds—is figure. Despite majority’s benefit con- determining a more accurate basis for contrary, tention to the the District Court Gwaltney’s delay.” economic benefit application proper demonstrated Court, therefore, Gwaltney Id. The held and, assessing law in there- against parent corporation’s the use of a fore, did not abuse its discretion. rate, interest but not the use of theoreti- short, nothing In in the record addition, there is per cal interest rate se. committed show that the District Court record shows that the District Court did choice of the interest rate calculation clear error consider the economic benefit context, industry-specific stating in an to calculate economic benefit. After care- comply fully weighing presented with the can the evidence “[failures [CWA] *29 during three-day on both sides experts trial, exercised the District Court and cred- as the trier of fact
its discretion testimony of one witness over
ited the supported by the The decision is
another. as our case law. testimony as well
expert I do not believe that the District
Because erroneous, fact-finding clearly
Court’s to deference under
its decision is entitled of discretion review.
abuse would, therefore, affirm the District
I the interest rate decision as to
Court’s to calculate economic benefit.
used BAMBA, Appellant
Karim RILEY, Jr.,* F. Interim District
William Immigra
Director of the Bureau of Enforcement.
tion and Customs
No. 03-2275. Appeals,
United States Court of
Third Circuit. March 2004.
Submitted
April
* AmendedPer Clerk's Order dated 07/02/03