# Amicus Curiae Brief — Utility Water Water Act Group Group v. Riverkeeper, Inc. (Nos. 07-597, 07-588, 07-589)

> Briefs, arguments, decisions, and more.

URL: https://www.frixlaw.com/law-library/documents/brief%3Amicro_IA40385015_0048%3A35

## Record

- **Collection:** Supreme Court brief
- **Document type:** Amicus Curiae Brief
- **Published:** January 1, 2009

## Text

~\ af =
NBDE asa,
Nos. 07-588, 07-589 & 07-597 (Consolidated) _ .

2o6§
In The nnn
Supreme Court of the Gnited States
ia

ENTERGY CORP., Petitioner,
V.
ENVIRONMENTAL PROTECTION AGENCY, ET AL.,
Respondents.

¢

PSEG FOSSIL LLC AND PSEG NUCLEAR LLC,

Petitioners,

V.

RIVERKEEPER, INC., ET AL., Respondents.
¢

UTILITY WATER ACT GROUP, Petitioner,

V.

RIVERKEEPER, INC., ET AL., Respondents.
.

On Writ Of Certiorari To The
United States Court Of Appeals
For The Second Circuit

+

BRIEF OF AMICUS CURIAE OMB WATCH
IN SUPPORT OF RESPONDENTS

+

PROFESSOR AMY SINDEN

Counsel of Record

Temple University Beasley School of Law #
1719 N. Broad Street

Philadelphia, PA 19122

(215) 204-4969

October 6, 2008

TABLE OF CONTENTS

es CO FAAF INE COI caicicssccccccscesacessccossersesecene iv
INTEREST OF AMICUS CURIAE.............0.. ee. 1
APPR S CE AIRC IRIBIC SE 6.0.0 ccccscccscssnccsscesescssecneses 1
i ici icc alien itil dimnaindtidanaaianinnsenneseens 4

I. FORMAL CBA, LIKE EPA RELIED ON
HERE, IS NOT “COMMON SENSE”; IT
IS A SPECIFIC AND CONTROVERSIAL
DECISION PROCEDURE THAT
CONGRESS PROHIBITED IN THE
AE a ee oot rn nf ee OO 4

Il WHEN THE CWA WAS _ PASSED,
LAWMAKERS WERE HIGHLY
SKEPTICAL OF FORMAL CBA. ............. cee eeeeee 5

A. Congress Eschewed Formal CBA in
Every Federal Environmental Statute
Passed in the Barly 19706. ...........c0.cccccoscccscccese 6

B. In Most’ Environmental Statutes,
Congress Directed Agencies to Employ
Technology-based Standards, Rather
than Formal CBA or Limited Cost-
I a as ieeiliadneconin 9

C. Inthe Few Instances when Congress did
Authorize Cost-Benefit Balancing, It
Intended a Limited Qualitative
Balancing, Not Formal CBA......................0000 10

D. No Consensus Favors Formal CBA; It
has Always been Highly Controversial. ......... 12

i
Ill. CONGRESS ESCHEWED FORMAL CBA
FOR GOOD REASON: [IT PRODUCES

UNRELIABLE AND IRRATIONAL
a es niensitiissisnsaiinsnvaeniniadeinsibtesitceppnandeaiitappaiinliiviannts 14

A. The Data and Scientific Understandings
Necessary to Quantify Many
Environmental and Health Benefits are
ERTL ER NLA De AD ee 16

1. The CBA of the Phase Il Rule was
Grossly Incomplete Because EPA did
not even Quantify the Vast Majority of
IT EE LAPS TAN DAS Res le ORE 16

2. Even for those Benefits EPA did
Attempt to Quantify, the Estimate was
I FI ia iichitccsienniperietbeitioeriicinncius 18

B. Techniques for Assigning Monetary
Values to Non-Market Goods _ are
Unreliable and Highly Contestable................ 21

1. The CBA for the Phase II Rule was
Grossly Incomplete Because EPA was
Unable to Find Non-Controversial
Methods for Monetizing Non-Market

a) EPA Used a Method for Estimating
the Benefits of Increased Forage
Species Survival that Measured
“Only a Small Share of these
I sidiedcedanibidensdniniadennddadnensnactiienalasinnins 24

b) Although EPA Believed that “Non-
Use Benefits may be Significant,” It
Ultimately Assigned them a Value of
TUTTI -icinisdicieudedcanehchiaiiciandaiakpacebcealiiegdabidbedtadiininiedale 25

C. EPA’S Estimate of Benefits for the
Phase II Rule was Grossly Incomplete,
Making Any Comparison with Costs
Irrational and Misleading..........................06008 27

1. OIRA Pressured EPA into Weakening
its Proposed Rule on the Basis of the
Highly Flawed, Incomplete, and
Irrational Results of its CBA, .2....2... 00200000000 28

2. Although EPA Cited Its Formal CBA
as the Basis for Rejecting the Closed
Cycle Alternative, the Proposed Rule
Preamble went on to Cite Dozens of
Reasons the Benefits were Under-
IETS PRS Sto IVE ME PO SCE TOE TNE OS RO 31

3. The Phase ll Rulemaking Jllustrates
How CBA Undermines’ Rational
Agency DecistOnima KING. ......2.02ccccceceeececeseeees 33

TE Glin: cktialenisemtininanteonenatenndnsibidienindidesasannes 35

IV
TABLE OF AUTHORITIES

CASES
American Textile Mfrs. Inst. Inc. Vv.

Donovan, 452 U.S. 490 (1981) «0.0... eee 6,8
Ass'n of Pacific Fisheries v. EPA, 615 F.2d

, TER nae 10, 12
BASF Wyandotte Corp. v. Costle, 598

I 11
EPA v. California, 426 U.S. 200 (1976)................. 9,10
EPA v. Natl Crushed Stone Ass’n, 449

RR SO CET La a 11
Indus. Union Dept v. Am. Petroleum

Pn GR I ID GID cccectnuedevcesenenniaytorsevcovsnesein 7
Tenn. Valley Auth. v. Hill, 437 U.S. 153

STENTS RTE ESS eae ae err 6
Weyerhaeuser v. Costle, 590 F.2d 1011

| ____RAURTS RRP 10, 11, 12
Whitman v. American Trucking Ass'ns,

I colonise 6,8
LEGISLATIVE MATERIALS
Clean Water Act, 33 U.S.C.

tii ccsinsnseciiiidanoeininnnaiiisoubiiadipes 11
ee en . sscendliounicedssiebenneetnle 7
ee Io cscucecmenesnsicnosawannoucsesaies 11,12
Oe os scscnnenpiasonsuesboconsonsestubsus 9
National Environmental Policy Act

UE aR, SE UIE... © GRBIED .nccc cscs ceneccccessccsscccecesce 10
Clean Air Act, 42 U.S.C. § 7411 (a)Q1) ooo. eee )
1 A Legislative History of the Water

Pollution Control Act Amendments of

Re es BU IE IID cccescccccenesscccsnsnsenctcccessesies 11

Vv

2 A Legislative History of the Water
Pollution Control Act Amendments of
1972, Ser. No. 93-1 (1973); S. Rep. No.

es eo ee te ciecnitsnsidatersdiedesiniiasabbaconiinbiniads

EPA: Investigation of Superfund and
Agency Abuses (Part 3): Hearings
Before the Subcomm. on Oversight and
Investigations of the Comm. on Energy

and Commerce, 98"» Cong. (1983)..............

Subcommittee on Oversight and
Investigations of the Committee on
Interstate and Foreign Commerce,
House of Representatives, 94% Cong.,
Federal Regulation and Regulatory

Reform, H. Rep. No. 75-931 (1976)............

U.S. Senate Committee on Governmental
Affairs, 96% Cong., Study on Federal

Regulation, S. Doc. No. 96-13 (1978).........

ADMINISTRATIVE MATERIALS
Exec. Order 12,291, 46 Fed. Reg. 13,193

RARER Re Ot ee cake eae mere mrah Seer
66 Fed. Reg. 6,976 (2001) .............ccccccceeeeeeeeee
Be BG, BRO, ETI GI esc cvcnscnccoscececoscceecs
69 Fed. Reg. 41,576 (2004) .....cecccccscessseesesees

EPA, Case Study Analysis for the
Proposed Section 316(b) Phase II
Existing Facilities Rule, EPA-821-R-02-

OE Crewe BOD oa ccevcesccvceesnsccessswvesssnenes

EPA, Economic and Benefits Analysis
(EBA) for the Final Section 316th)
Phase Il Existing Facilities Rule, EPA-

821-R-04-005 (February 2004)..............00..

EPA, Economic and Benefits Analysis
(EBA) for the Proposed Section 316(b)

.... passim
.... passim

pinning 26

19, 28, 34

vi

Phase II Existing Facilities Rule, EPA-

821-R-02-001 (February 2002)..............00... 18, 19, 27
EPA, OMB Review Draft for the Proposed

Section 316(b) Rule for Large Cooling

Water Intake Structures at Existing

Power Generating Facilities, Docket

W-00-32, DCN # 4-4005 (Dec. 28, 2001).......... 29, 30
EPA, Regional Analysis Document for the

Final Section 316(b) Phase IT Existing

Facilities Rule, EPA-821-R-02-003

I a citing 17, 23, 24, 25
EPA, Summary of Mayor Changes During

Interagency Review, Docket W-00-32,

re a eeveuensounecssoreoncane 30
Exec. Order 12,866, 58 Fed. Reg. 51,735

EEN EL ec a ee 13
NEPA Implementing Regulations, 40

0 EE 11
OMB, Circular A-4 (Sept. 17, 2003) ........ cece 14
BOOKS

Cass R. Sunstein, Risk and Reason (2002).....6, 12, 35
Elhzabeth Anderson, Value in Ethics and
a sinssoensoesenild 5
Frank Ackerman & Lisa Heinzerling,
Priceless: On Knowing the Price of
Everything and the Value of Nothing

os siddeuucusvesveuveeesse 5
Hon. Stephen Breyer, SBreaking the

I cenhmneniiil 16
Richard L. Revesz, et al., Retaking

RT 12,13

Sidney A. Shapiro & Robert L.
Glicksman, Fisk Regulation at Risk-
Restoring a Pragmatic Approach (2003) ................. 5

vil
ARTICLES

Amy Sinden, Cass Sunstein’s Cost-Benefit
Lite: Economics for Liberals, 29 Col. J.
I capnodubin 15
C. Edwin Baker, VThe Ideology of the
Economic Analysis of Law, 5 Phil. &

I seuss 5
Cass R. Sunstein, Cost-Benefit Default
Principles, 99 Mich. L. Rev. 1651 (2001)................. 8

Cass R. Sunstein, J/ncommensurability
and Valuation in Law, 92 Mich. L. Rev.

0 TEASER REIS ae EEE Ss. SSID 5
Cass R. Sunstein, Vhe Arithmetic of

Arsenic, 90 Geo. L. J. 2255 (2002) ........00...... 15,16
Daniel C. Esty, Revitalizing

Environmental Federalism, 95 Mich. L.

IRE arene i a Ped RE nE en LP aNO Noe Ee 22

David M. Driesen, Js Cost-Benefit
Analysis Neutral?, 77 U. Colo. L. Rev.
ER See near A Wb t 13
Frank Ackerman & Lisa Heinzerling,
Pricing the Priceless: Cost-Benefit
Analysis of Environmental Protection,
150 U. Penn. L. Rev. 1553 (2002) .......................... 34
Lawrence H. Tribe, Ways Not to Think
About Plastic Trees: New Foundations
for Environmental Law, 83 Yale L. J.
| RATE ane aD ke ee D
Richard Parker, Grading the
Government, 70 U. Chi. L. Rev. 1345
RSS een aD TE en TO eon FeO a TR 34
Thomas H. Stevens, et al., Measuring the
Existence Value of Wildlife: What Do
CVM Estimates Really Show? 67 Land
Rs TI ED. oo easadsasaninecaninnnaies 21, 22

INTEREST OF AMICUS CURIAE

OMB Watch is a nonprofit research and
advocacy organization dedicated to promoting
government accountability, citizen participation in
public policy decisions, and the use of fiscal and
regulatory policy to serve the public interest. OMB
Watch was founded in 1983 to shed light on the
activities and operations of the Office of Management
and Budget (OMB). The organization also focuses on
the substantive areas that OMB and an agency
within it, the Office of Information and Regulatory
Affairs (OIRA), oversee: federal regulatory policy, the
federal budget, taxation and government
performance, information and access, and nonprofit
action, advocacy, policy and technology.!

SUMMARY OF ARGUMENT

In developing the Phase II rule, EPA relied in
part on a formal, economic cost-benefit analysis
(“formal CBA”), i.e., an analysis that estimated both
the costs and benefits of the rule in monetary terms
and compared them. Formal CBA is not a “common
sense” tool for “rational” decision making, as
petitioners contend. It is a highly controversial
decision procedure that often produces irrational
results. Congress was extremely skeptical of formal
CBA when it passed the Clean Water Act (CWA) and
thus declined to provide fur it anywhere in the

| The parties have consented to the filing of this brief. No
counsel for a party authored this brief in whole or 1n part, and
no counsel or party made a monetary contribution intended to
fund the preparation or submission of this brief. No person
other than amicus curiae, their institutions, or their counsel
made a monetary contribution to its preparation or submission.

2

statute, including in § 316(b). Indeed, Congress
eschewed formal CBA in every environmental statute
it passed in the early 1970s.

Rather than directing agencies to engage in
formal CBA, or even a limited cost-benefit balancing,
most environmental statutes of that era, including
the CWA, directed agencies to employ technology-
based standards. Technology-based standard-setting
requires the agency to evaluate likely costs of a
proposed standard in order to determine whether it is
economically feasible (i.e., “available”). But it does
not require agencies to delve into the far more
problematic task of attempting to quantify and
monetize environmental benefits.

The technology-based approach was
particularly important to Congress’ overhaul of the
CWA in 1972 because previous versions of that
statute had required agencies to engage in a
cumbersome and unworkable evaluation of water
quality benefits. In those few instances when
Congress did authorize cost-benefit balancing—as it
did under the Best Practicable Technology
standard— it did not authorize formal CBA. Rather,
it directed agencies to perform a limited, qualitative
balancing.

The use of formal CBA to. evaluate
environmental regulation has always’ been
controversial. The practice first became widespread
in 1981, when President Reagan signed an executive
order requiring all major federal regulations to
undergo formal CBA (to be overseen by the OMB’s
Office of Information and Regulatory Affairs (OIRA)).
The executive order did not enjoy bipartisan support.
Indeed, many viewed it as a highly partisan effort to
delay and weaken regulation.

The application of formal CBA to
environmental regulation rests on the untenable
assumption that complex effects on ecological and
human health can be quantified and expressed in
dollar terms. In practice, scientific understandings
are rarely fine-grained enough to predict such
impacts in quantifiable terms. Even where they are,
data are inevitably incomplete. And even for those
data that do exist, the process of converting them into
dollars raises intractable difficulties that render the
results endlessly contestable.

The Phase II rulemaking illustrates these
shortcomings. The formal CBA that accompanied it
was hopelessly incomplete because EPA was unable
to quantify the vast majority of the broad and
complex ecological benefits associated with the rule.
EPA admitted, for example, that its estimate
accounted for less than two percent of the fish subject
to impingement and entrainment from cooling water
intake structures. Moreover, EPA was unable to find
non-controversial methods for monetizing the non-
market goods at issue. Finally, after receiving
extensive criticism, EPA threw up its hands and
simply left most benefits out of the analysis entirely.

The CBA never produced a _ rehable or
meaningful number. The benefits estimate
fluctuated wildly, changing by nearly a factor of ten,
from $735 million under the proposed rule to $83
million under the final rule. Even the larger number
was, by EPA’s own frequent admission, hopelessly
incomplete. The documents describing the CBA were
peppered with caveats and disclaimers, describing
the myriad reasons why the benefits figure grossly
under-estimated the rule’s true benefits.

Yet, despite the innumerable caveats, EPA still
resorted to the raw numbers when it came time to

4

make a decision. At OIRA’s urging, it rejected a
closed cycle cooling alternative because “the
incremental costs of this option relative to the
proposed option ($413 million) significantly outweigh
the incremental benefits ($146 million).” 67 Fed. Reg.
17,122, 17,158 (2002). Thus, EPA based its decision
on a comparison of an incomplete estimate of benefits
with a relatively complete estimate of costs. This is
the kind of irrational decisionmaking formal CBA
often fosters. Congress rejected formal CBA in §
316(b) for that reason.

ARGUMENT

I. FORMAL CBA, LIKE EPA RELIED ON
HERE, IS NOT “COMMON SENSE”; ITIS A
SPECIFIC AND CONTROVERSIAL
DECISION PROCEDURE THAT CONGRESS
PROHIBITED IN THE CWA.

In deciding which of several alternatives to
adopt for the Phase II rule—setting the standard
under CWA § 316(b) for “minimizing adverse
environmental impact” from cooling watcr intake
structures at existing power plants—EPA relied in
part on a formal, economic cost-benefit analysis. See
pp. 28-31, infra. By “formal economic cost-benefit
analysis” (hereinafter “formal CBA”), we refer to an
analysis that estimates a regulation’s costs and
benefits to society in quantified, monetary terms and
then compares them.? See Economists Frank
Ackerman, et al. Amicus Br. 10.

2 EPA appears to also contemplate that formal CBA will be used
in connection with the site-specific compliance alternative. See
67 Fed. Reg. at 17,149 (“EPA believes that a rigorous...

5

This is not, as petitioners contend, simply a
“common sense... imperative of basic rationality to
ensure that actions do more good than harm.”
Entergy Br. 29. It is a specific and highly
controversial decision procedure that Congress
considered and rejected in drafting the CWA as a
whole, and in § 316(b) in particular. Riverkeeper Br.
6-14, 47-50. Accordingly, the EPA violated § 316(b)
by using formal CBA to decide on the Phase II rule,
and the judgment of the court of appeals should be
affirmed.

Il WHEN THE CWA WAS _ PASSED,
LAWMAKERS WERE HIGHLY SKEPTICAL
OF FORMAL CBA.

Congress passed the Clean Water Act in 1972,
during a time when many members of Congress
viewed formal CBA with considerable skepticism.? In
negotiations over the wording of § 316(b), the CBA
skeptics won out, keeping any provision for formal

economic analysis should be performed when a facility seeks a
site-specific determination of best technology available due to
significantly greater cost as compared to the benefits of
compliance.”).

3 Congress’ skepticism reflected an extensive academic
literature critiquing formal CBA. See, e.g., Lawrence H. Tribe,
Ways Not to Think About Plastic Trees’ New Foundations for
Environmental Law, 83 Yale L. J. 1315 (1974); C. Edwin Baker,
The Ideology of the Economic Analysis of Law, 5 Phil. & Pub.
Aff. 3 (1975). A robust academic critique of CBA continues to
this day. See, eg, Frank Ackerman & Lisa Heinzerling,
Priceless: On Knowing the Price of Everything and the Value of
Nothing (2004); Sidney A. Shapiro & Robert L. Glicksman, Risk
Regulation at Risk’ Restoring a Pragmatic Approach (2003):
Cass R. Sunstein, Jncommensurability and Valuation in Law, 92
Mich. L. Rev. 779 (1994); Elizabeth Anderson, Value in Ethics
and Economics (1993).

6

CBA or cost-benefit balancing of any kind out of that
section. Riverkeeper Br. 12-14, 49-50. Indeed, the
CBA skeptics were successful in keeping formal CBA
out of the statute entirely. /d. at 6-12, 48-49. Whilea
few sections of the CWA call for a limited balancing of
costs and benefits, see pp. 11-12, supra, none allow
for formal CBA. Riverkeeper Br. at 12.

A. Congress Eschewed Formal CBA in Every
Federal Environmental Statute Passed in the
Early 1970s.

Indeed, in every major environmental, health,
and safety statute passed during the first half of
1970s, Congress eschewed formal CBA. See Cass R.
Sunstein, Risk and Reason 21-22 (2002) (“[Mlost
federal statutes do not call for cost-benefit balancing,
and many of them seem to preclude it.”); American
Textile Mfrs. Inst. Inc. v. Donovan, 452 U.S. 490, 508-
522 (1981) (Occupational Safety and Health Act of
1970 requires feasibility analysis rather than CBA);
Whitman v. American Trucking Ass‘ns, 531 U.S. 457,
467 (2001) (Clean Air Act of 1970 prohibits weighing
costs against benefits in promulgation of national
ambient air quality standards); Tenn. Valley Auth. v.
Hill, 437 U.S. 153, 173, 184 (1978) (§7 of Endangered
Species Act of 1973 prohibits weighing costs against
benefits).

Members of Congress were concerned that
efforts to quantify and monetize the values associated
with human health and the environment would be so
controversial and so plagued by uncertainties that
agencies would waste valuable time and resources
without producing meaningful results. A 1976 House
report found these limitations “so severe that they
militate against [the] use [of benefit/cost analysis]

7

altogether.” Subcommittee on Oversight and
Investigations of the Committee on Interstate and
Foreign Commerce, House of Representatives,
94th Cong., Federal Regulation and Regulatory
Reform, H. Rep. No. 75-931, at 515 (1976). A 1978
Senate study took a similar view:

Where economic regulation is concerned,
[cost-benefit] analysis can be more easily
applied, since there the consequences
are usually capable of being reduced to
dollar and cent terms. Such is not
always the case with health, safety and
environmental regulation. Here it is
extremely difficult to quantify benefits
since they are. subject to great
uncertainty and often become apparent
only with the passage of time. In
addition, some important benefits—such
as recreational or aesthetic values—are
difficult if not impossible to quantify in
any meaningful way ... . Therefore
there are serious limitations to the use
of economic impact analysis in the
health and_ safety area ;
[D]ecisionmaking to protect the public
from serious hazards should not be
reduced to those terms.

U.S. Senate Committee on Governmental Affairs, 96t»
Cong., Study on Federal Regulation, S$. Doc.
No. 96-13, at xxiv (1978). See also Indus. Union Dep't
v. Am. Petroleum Inst., 448 U.S. 607, 719 n. 32 (1980)
(Marshall, J., dissenting) (“Congress’s antipathy
toward cost-benefit balancing is evident throughout

8

the legislative history of the [Occupational Safety and
Health] Act.”).4

Thus, despite petitioners’ protestation that
Congress “could not possibly have intended” such
“absurd” results (Entergy Br. 22), the fact that, in
1972, Congress drafted § 316(b) of the CWA to
preclude the use of formal CBA (or indeed any kind of
cost-benefit balancing) is not at all surprising.

4 In light of Congress’ antipathy toward formal CBA, petitioners’
suggestion that, unless Congress has explicitly prohibited CBA,
“it may be arbitrary and capricious” for an agency not to do it, is
specious. Entergy Br. 30. The only authority petitioners cite for
this proposition is an article by Professor Cass R. Sunstein,
Cost-Benefit Default Principles, 99 Mich. L. Rev. 1651 (2001), in
which he calls on the federal courts to adopt a presumption
favoring formal CBA where a statute is silent. /d: see also PLF
Amicus Br. 5-6 (advocating Sunstein’s “default principle”);
Entergy Br. 26, n. 8 (same). Here the statute is mot silent; §
316(>) clearly requires EPA to apply a technology based
standard rather than CBA. See Riverkeeper Br. 23-33. But
even if it were, petitioners’ call for the application of a “cost-
benefit default principle” would be unavailing. While Professor
Sunstein is a widely-respected scholar, as shown above, his
proposal is clearly out of step with Congressional intent. This
perhaps explains why this Court has repeatedly declined to
adopt his default principle. See Whitman, 531 U.S. at 467 (“We
have .. . refused to find implicit in ambiguous sections of the
[Clean Air Act] an authorization to consider costs that has
elsewhere, and so often, been expressly granted.”); Donavan, 452
U.S. at 510 (‘When Congress has intended that an agency
engage in cost-benefit analysis, it has clearly indicated such
intent on the face of the statute.”). See generally Amy Sinden,
Cass Sunstein’s Cost-Benefit Lite: Economics for Liberals, 29
Col. J. Envtl. L. 191, 228-240 (2004) (critiquing Sunstein’s
proposal).

B. In Most Environmental Statutes, Congress
Directed Agencies to Employ Technology-based

Standards, Rather than Formal CBA or
Limited Cost-Benefit Balancing.

Rather than directing agencies to engage in
formal CBA or even a limited cost-benefit balancing,
most environmental statutes of the 1970s, including
the CWA, direct agencies to employ technology-based
standards. See, e.g., CWA, 33 U.S.C. §§ 1311(b)(2),
1314(b)(2) (effluent limits for point sources based on
the “best available technology”); Clean Air Act, 42
U.S.C. § 7411(a)(1) (emissions limits for new sources
based on “the best system of emission reduction .. .
adequately demonstrated”). Technology-based
standard-setting requires the agency to evaluate the
likely costs of a proposed standard in order to
determine whether it is economically feasible (i.e.,
“available”). It does not require agencies to delve into
the far more problematic task of attempting to
quantify and monetize the environmental benefits of
regulation in order to compare them to costs. See
Economists Frank Ackerman, et al. Amicus Br. 16-26.

The technology-based approach was
particularly important to Congress’ overhaul of the
CWA in 1972. Previous versions of the Act had
required standard-setting and enforcement to be
based on an evaluation of the benefits of regulation—
1.e., on assessments of the quality of the receiving
waters. HPA v. California, 426 U.S. 200, 202-205
(1976). This approach proved to be entirely
unworkable—in the words of the Senate Committee
on Public Works—“inadequate in every vital aspect.”
2 A Legislative History of the Water Pollution

10

Centrol Act Amendments of 1972, Ser. No. 93-1, at
1423 (1973); S. Rep. No. 92-414, at 7 (1971).
Evaluating the benefits of water pollution
reduction was extremely difficult. California, 426
U.S. at 202. It required tedious and costly site-
specific measurements, as well as assessments of
complicated and inadequately understood ecological
chains of causation. Technology-based standard
setting, on the other hand, allowed EPA to simply set
uniform national standards for each industry based
on the maximum technologically achievable level of
pollution reduction. Weyerhaeuser v. Costle, 590 F.2d
1011, 1042 (D.C.Cir. 1978). This only required the
agency to evaluate technologies and costs. See
Economists Frank Ackerman, et al. Amicus Br. 16-26.
Thus, one of Congress’ primary goals in overhauling
the CWA in 1972 was to relieve permitting agencies
of the burden of attempting to measure regulatory
benefits. See Ass'n of Pacific Fisheries v. EPA, 615
F.2d 794, 805 (9t Cir. 1980); Riverkeeper Br. 6-7.

C. In the Few Instances when Congress did
Authorize Cost-Benefit Balancing, It Intended
a Limited Qualitative Balancing, Not Formal

CBA.

Even in the few instances in the early 1970s in
which Congress did instruct agencies to compare
costs and benefits, it did not expect them to perform a
full-fledged, formal CBA. Rather, it directed them
simply to perform a rough, qualitative comparison to
ensure that costs were not wholly disproportionate to
benefits. See, e.g., National Environmental Policy
Act (NEPA), 42 U.S.C. § 4332(B) (“insure that
presently unquantified environmental... values...
be given appropriate consideration in decisionmaking

11

along with economic . . . considerations”); NEPA
Implementing Regulations, 40 C.F.R. § 1502.23
(NEPA analysis “need not be displayed in a monetary
cost-benefit analysis and should not be when there
are important qualitative considerations.’ ‘.

In the CWA itself, Congress called for a
balancing of costs and benefits in conjunction with
the Best Practicable Technology standard. In
promulgating these interim standards, which were to
remain in effect for only a few years, see 33 U.S.C. §
1311(b)(1)(A); Riverkeeper Br. 34-35, the statute
directed EPA to consider the costs that would be
imposed on industry “in relation to the effluent
reduction benefits to be achieved.” Id. at §
1314(b)(1)(B).

Congress did not intend by this language to
authorize EPA to engage in formal CBA. This
language contemplated no more than a “limited cost->
benefit analysis,” intended to cull out the most
extreme cases where “the additional degree of
effluent reduction is wholly out of proportion to the
costs.” 1A Legislative History of the Water Pollution
Control Act Amendments of 1972, Ser. No 93-1, at
170 (1973); EPA v. Nat? Crushed Stone Ass'n, 449
U.S. 64, 71, n. 10 (1980)(quoting legislative history);
Weverhaeuser, 590 F.2d at 1045 n. 52 (same); BASF
Wyandotte Corp. v. Costle, 598 F.2d 637, 656-57 (1s
Cir. 1979)(CWA imposes obligation on EPA only to
perform “limited cost-benefit balancing” which “is a
relatively subsidiary task and need not be precise”).

Concerned that EPA not be “boglged] down in
burdensome proceedings” the courts held that “cost
need not be balanced against benefits with pinpoint
precision.” Weyerhaeuser, 590 F.2d at 1048. Indeed,
EPA was not supposed to monetize the benefits at all.
Nor did Congress intend the agency to investigate the

12

full social benefits of the regulation in the welfare
economic sense—that is, to attempt to measure the
overall ecological improvement in the quality of the
receiving waters. Pacific Fisheries, 615 F.2d at 805.
Instead, the statute directed EPA to balance only a
very simplified version of the benefits: “the effluent
reduction benefits,” 33 U.S.C. § 1314(b)(1)(B), i.e., the
pounds or tons of a pollutant that would be prevented
from entering the water. Thus, EPA was simply
supposed to hold up the pounds of pollutant reduced
next to the dollar costs of the regulation and perform
a limited balancing. Weyerhaeuser, 590 F.2d at
1047.

D. No Consensus Favors Formal CBA; It has
Always been Highly Controversial.

There is snot a_ long-standing consensus
favoring the use of formal CBA in agency
decisionmaking, as _ petitioners and their amici
contend. See AEI-Brookings Amicus Br. 8. The
widespread use of formal CBA by federal agencies
began in 1981, when President Reagan signed
Executive Order 12,291, requiring formal CBAs to be
prepared for all major federal regulations. 46 Fed.
Reg. 13,193 (1981). The executive order was seen at
the time as a highly partisan effort to slow regulatory
activity. See Richard L. Revesz, et al., Retaking
Rationality 24-29 (2008); see also Sunstcin, Aisk,
supra, at 20 (“Executive Order 12,291 proved
extremely controversial.”). Indeed, it explicitly stated
that its purpose was “to reduce the burdens of
existing and future regulations.” 46 Fed. Reg. at
13,193.

The executive order gave authority to oversee
agency compliance with the new CBA mandate to the

13

newly created Office of Information and Regulatory
Affairs (OIRA) within the White House Office of
Management and Budget. !6 Fed. Reg. at 13,193 (§
2(b)). This review. process, including’ the
relationships and communications between OIRA
officials and regulated industries, was “shrouded in
secrecy,” which only served to heighten’ the
skepticism and suspicion with which many members
of Congress and environmental groups viewed OIRA
and the CBA mandate. See Revesz, supra, at 25;
EPA: Investigation of Superfund and Agency Abuses
(Part 3): Hearings Before the Subcomm. on Oversight
and Investigations of the Comm. on Energy and
Commerce, 98 Cong. 2 (1983) (statement of Rep.
John Dingell) (“OMB acts as a conduit for promoting
the views of industry affected by proposed regulations
through secret, undisclosed, and unreviewable
contacts.”).

It was widely assumed that OIRA review of
regulations under the executive order’s CBA mandate
would have the effect of delaying and weakening
rather than spurring regulation, and that, indeed,
has been the effect. See David M. Driesen, Js Cost-
Benefit Analysis Neutral?, 77 U. Colo. L. Rev. 335,
365, 380 (2006) (Of 25 OIRA reviews studied, 24
resulted in changes that weakened environmental,
health, and safety protections.).

Although the Clinton administration pulled
back from the explicitly anti-regulatory stance of the
Reagan and Bush I administrations, it continued to
implement the same basic cost-benefit mandate,
replacing Reagan’s executive order with a similar one
that softened a few edges but kept the essential
mandate intact. See Exec. Order 12,866, 58 Fed. Reg.
51,735 (1993)(replacing Reagan-era requirement that
benefits “outweigh” costs with “reasoned

14

determination” that benefits “justify” costs); OMB,
Circular A-4, at 10 (Sept. 17, 2003)(“A distinctive
feature of [CBA] is that both benefits and costs are
expressed in monetary’ units.”), avail. at
http://www.whitehouse.gov/omb/circulars/a004/a-
4.pdf. Despite the embrace of CBA by the Clinton
administration, however, debate among scholars and
policymakers continued unabated and continues
today. Seep. 5, note 3, supra.

Ill. CONGRESS ESCHEWED FORMAL CBA FOR
GOOD REASON: IT PRODUCES
UNRELIABLE AND IRRATIONAL RESULTS

Congress had good reason to be skeptical of the
use of formal CBA to evaluate environmental health
and safety regulation in the 1970s, and those reasons
remain valid today. The application of formal CBA to
environmental regulation rests on the untenable
assumption that the complex ecological and human
health processes that such regulation seeks to protect
can be quantified and expressed in dollar terms. In
practice, scientific understandings are rarely fine-
grained enough to predict impacts in quantifiable
terms. Even where they are, data are inevitably
vastly incomplete. And even for those quantifiable
data that do exist, the process of converting such data
into dollar terms raises intractable practical and
theoretical difficulties that make most monetized
estimates of impacts endlessly contestable. As a
result, formal CBA fails miserably at its appointed
task. Rather than providing a “common sense” tool
for insuring “rational” regulation (Entergy Br. 4, 10,
29), formal CBA _ produces Alice-in-Wonderland
results that are so incomplete and so unreliable, they

15

provide endless opportunity for manipulation,
effectively vesting in agency officials virtually
unlimited and unreviewable discretion.

Formal CBA need not pinpoint costs or benefits
with a single number. Some range of uncertainty is
to be expected from any decisionmaking standard.
Results could have some margin of error and still be
quite useful. But the difficulties inherent in trying to
attach a dollar figure to environmental values—like a
human life or clean water or a species or an
ecosystem—create uncertainties of such enormous
magnitude that no meaningful comparison of costs
and benefits is possible.

EPA’s CBA of its regulation limiting arsenic in
drinking water, cited by amicus Pacific Legal
Foundation (PLF Amicus Br. 7), provides a telling
example. EPA estimated the costs of that rule to be
approximately $206 million, see 66 Fed. Reg. 6,976,
7,010 (2001), but a study by Professor Cass Sunstein
concluded that reasonable people making reasonable
assumptions could peg the benefits of the rule
anywhere between a low if $13 million or a high of
$789 million. See Cass R. Sunstein, The Arithmetic
of Arsenic, 90 Geo. L. J. 2255, 2288 (2002). EPA
presented the benefits as falling somewhere within
the modest range of $140 to $200 million. 66 Fed.
Reg. at 7017. This presented a false and misleading
picture that failed to capture the true magnitude of
the uncertainties involved. See Sinden, supra, at
212-29.

A CBA with this level of uncertainty does not
help anyone identify “whether the regulation does
more harm than good.” (Entergy Br. 21). Rather than
“rationalizing” agency decision making (Ud), it makes
it more vulnerable to political distortion. When
uncertainties are so large, interest groups can easily

16

manipulate the numbers to reach their desired
outcome.

A. The Data and Scientific Understandings
Necessary to Quantify Many Environmental
and Health Benefits are Simply Unavailable.

In many instances, we lack the data and/or
scientific understanding necessary to make definitive,
non-controversial estimates of the environmental
benefits of regulation. 7 See, e.g., Hon. Stephen
Breyer, Breaking the Vicious Circle 45 (1993)
(attempt to assess risks dssociated with aflotoxin in
peanuts yielded “[t]wo scientifically plausible models
. . . showling] risk levels differing by a factor of
40,000.”). Quantifying the health benefits associated
with the arsenic rule, for example, was no easy task.
Various studies linked arsenic to seven different
types of cancer as well as. to other health effects, but
only the studies linking arsenic to lung and bladder
cancer provided data in quantifiable form. See
Sunstein, Arsenic, supra, at 2272-74. Even for those
cancers, there were no hard data
at low doses. /d. at 2279-82. From the outset, then,
even before any attempt at monetization, the benefits
estimate for EPA’s arsenic CBA contained a large
range of uncertainty and covered only a fraction of
the relevant health effects.

1. The CBA of the Phase ll Rule was Grossly

Incomplete Because EPA did not even
Quantity the Vast Majority of the Benefits.

For rules with primarily ecological benefits,
the informational problems are even more acute,
because so many aspects of ecological processes are

17

still so poorly understood. Accordingly, EPA has
expressed concern that formal CBAs under the CWA
have generally “been limited in the range of benefits
assessed,” thus “hinderling] EPA’s ability to compare
... benefits and costs . . . comprehensively.” 67 Fed.
Reg. at 17,191. EPA found the process of assessing
benefits for the Phase II rule similarly “challenging.”
69 Fed. Reg. 41,576, 41,655 (2004). Ultimately, its
formal CBA failed to include the vasi majority of the
rule’s benefits.

EPA began by estimating the number of fish
impinged or entrained at intake structures. See 67
Fed. Reg. at 17,190. These impacts vary widely from
site to site, depending on the type of aquatic
ecosystem that exists there, the size and location of
the intake structure, and so on. See id. Accordingly,
EPA did a series of case studies of facilities in
different regions and on different water body types,
and used those results to extrapolate estimates of
impingement and entrainment for all the facilities in
the country. See 69 Fed. Reg. at 41,655.

From the outset, of course, estimating the
number of fish impinged or entrained leaves out a
whole set of other ecological harms caused by cooling
water intake, including the loss of non-fish species
and impacts on the functioning of the ecosystem as a
whole. See EPA, Hegional Analysis Document for the
Final Section 316(b) Phase Il Existing Facilities Rule,
EPA-821-R-02-003, at A9-1 (February 12, 2004)
(available at http://www.epa.gov/waterscience/
316b/phase2/casestudy/final. htm). Phytoplankton
and zooplankton trapped in cooling water intake
structures can also be of crucial importance to the
aquatic food chain but these losses were not
calculated by EPA. Indeed, EPA did not even include

the loss of certain commercially valuable large

18

invertebrate species like lobsters, mussels, crabs, and
shrimp because they lacked adequate data. See EPA,
Economic and Benefits Analysis (EBA) for the
Proposed Section 316(b) Phase II Existing Facilities
Rule, EPA-821-R-02-001, at Cil-7 (February 2002)
(available at http://www.epa.gov/waterscience/316b/
phase2/econbenefits). Nor did EPA include effects on
fish-eating birds or on sea turtles and other
endangered species. See 69 Fed. Reg. at 41,624
(“3,200 threatened or endangered sea turtles entered
enclosed cooling water intake canals at the St. Lucie
Nuclear Generating Plant in Florida.”). Moreover,
the agency made no attempt to account for
cumulative impacts, see EPA, EBA Proposed Rule,
supra, at Cl1-6, or the “stresses to _ overall
communities and ecosystems as_ evidenced by
reductions in diversity or other changes in system
structure and function,” 69 Fed. Reg. at 41,586. EPA
acknowledged that while all of these impacts are very
important, measuring them is _ particularly
challenging, because “[plopulation dynamics and the
physical, chemical, and biological processes of
ecosystems are extremely complex.” Jd.

2. Even for those Benefits EPA did Attempt to
Quantify, the Estimate was _ Grossly
Incomplete.

Even for the narrow slice of benefits that EPA
did attempt to quantify, its estimates were highly
inaccurate and incomplete, as EPA repeatedly
acknowledged.

EPA derived its estimate of the magnitude of
impingement and entrainment impacts on fish from
facility-reported data. 69 Fed. Reg. at 41,656. This
introduced yet another source of uncertainty and

19

inaccuracy. See EPA, EBA Proposed Rule, supra, at
C1-6. Most facilities had never even conducted such
studies. See EPA, Economic and Benefits Analysis
(EBA) for the Final Section 316(b) Phase II Existing
Facilities Rule, EPA-821-R-04-005, at A2-1
(February 2004) (available at http://www.epa.gov/
waterscience/316b/phase2/econbenefits/final.htm).
Among those that had, sampling methods and
equipment were “highly variable.” 69 Fed. Reg. at
41,656. The “data [were] often limited to a subset of
species,” and were in some instances as much as
three decades old. Jd. EPA observed that as a result
of these problems, “the magnitude of impingement
and entrainment is often underestimated.” 67 Fed.
Reg. at 17,190: see also EPA, EBA Proposed Rule,
supra, at C1-6 (“EPA believes that its analysis is
likely to lead to _ potentially significant
underestimates .. . of regulatory benefits.”).

Once it had estimated the number of fish that
are otherwise lost through impingement and
entrainment but would be saved by the rule (1.4
billion annually), 69 Fed. Reg. at 41,657, EPA then
estimat~u the number of those fish that would likely
be caught by commercial and recreational fishermen.
Most: fish species are not caught at all. And even
among those species that are, harvest rates are
usually less than 20 percent. Accordingly, the
number of affected fish that EPA actually counted
was vanishingly small:

Of the organisms which are anticipated
to be protected by the section 316(b)
Phase II rule, it is projected that
approximately 1.8 percent will
eventually be harvested by commercial
and recreational fishers and therefore

20

can be valued with direct use valuation
techniques. The Agency’s direct use
valuation does not account for the
benefits from the remaining 98.2% of the
age 1 equivalent aquatic organisms
estimated to be protected nationally
under today’s rule.

Id. at 41,660-61.

This estimation was, as EPA acknowledged,
grossly incomplete, since the fish that survive
uncaught have an obvious ecological value. They will
reproduce to sustain the population that will be
fished on in future years, they will provide food for
other fish, and undoubtedly play other important
roles in the larger ecosystem as well. See id. at
41,657. This methodology generated considerable
controversy and illustrates the high degree of
indeterminacy that attempts to quantify natural
resource values can create. For example, natural
resource economist Frank Ackerman, Ph.D., of Tufts
University, commented on the _ proposed rule,
criticizing EPA’s incomplete accounting of fish
mortality. See J.A. 211- 244. He argued that a more
reasonable methodology that took into account the
value of the fish that were not caught would have the
effect of doubling EPA’s benefits estimate. J.A. 223.

In sum, EPA’s benefits estimate was hopelessly
incomplete, encompassing only a tiny fraction of the
total benefits of the Phase II rule. Many aspects of
the rule’s benefits were left off the ledger sheet
entirely because EPA simply lacked the data
necessary to assess them. Even with respect to the
benefits for which EPA did have some limited data—
avoidance of impingement and entrainment impacts

21

on fish—EPA’s estimate accounted for less than two
percent of the impacted fish.

B. Techniques for Assigning Monetary Values to
Non-Market Goods are Unreliable and Highly

Contestable.

Formal CBA requires the analyst to attach
dollar figures to values that have no price—human
lives, healthy ecosystems, endangered species. For
many, the very thought of attempting to assign a
dollar price to such intangibles defies common sense.
At a minimum, the practice proves extremely
controversial. While economists have developed a
number of methods for deriving “shadow prices” for
non-market goods, the numbers they produce are
inevitably soft and highly contestable. .

Often, in order to attach a dollar figure to
unpriced natural resources, analysts resort to public
opinion polls. In what is called the “contingent
valuation method” (CVM), analysts give members of
the public information about a particular good—the
preservation of an endangered species, or the
maintenance of clean air or clean water—and then
ask how much they would be willing to pay for it.
Thomas H. Stevens, et al., Measuring the Existence
Value of Wildlife’ What Do CVM Estimates Really
Show? 67 Land Economics 390 (1991); see 67 Fed.
Reg. at 17,150 (suggesting use of contingent
valuation in CBAs for _ site-specific BTA
determinations). One such survey found the average
person willing to pay $19.28 per year to prevent the
extinction of bald cagles. See Stevens, supra, at 396.

Studies like these never produce hard
numbers. They are inevitably subject to challenge on
a variety of methodological grounds. The survey may

22

be criticized for the amount or objectivity of the
background information provided, the way questions
are phrased, the fact that respondents are not subject
to actua] budget constraints, or the geographic scope
of the survey. See id. at 396-99. If EPA were to
conduct a contingent valuation survey on the
endangered sea turtles lost in cooling water intake
structures in Florida, for example, would it be
appropriate to count the willingness-to-pay of all
people in Florida, everyone in the U.S., or everyone in
the world? There is no consensus on how to handle
such problems, yet outcomes can vary widely—by
orders of magnitude—depending on how they are
resolved. See Daniel C. Esty, Revitalizing
Environmental Federalism, 95 Mich. L. Rev. 570, 640
(1996). Accordingly, the results of any such study are
inevitably vulnerable to reasonable arguments that
they should be adjusted significantly up or down.
Alternatively, agencies put off by the expense and
controversy may forego such studies and simply leave
non-market goods out of the calculation altogether.

1. The CBA for the Phase II Rule was Grossly
Incomplete Because EPA was Unable to Find
Non-Controversial Methods for Monetizing

Non-Market Goods.

The vast majority of the benefits associated
with the Phase II rule involved ecological vaiues that
EPA was simply unable to quantify or monetize. The
only aspects of the benefits that directly involved a
market good were the commercial fish species that
would be saved from impingement and entrainment.
EPA estimated this value by simply calculating the
market price of the additional commercial fish that

23

would be caught under the rule. 69 Fed. Reg. at
41,659-60.

To account for the value of the recreationally
caught fish, EPA used a “random utility model,”
which inferred anglers’ willingness-to-pay for
recreational fishing based on their travel costs for
visiting particular fishing sites and then used a
mathematical model to estimate how that
willingness-to-pay would likely increase in response
to increased catch levels. EPA, Regional Analysis,
supra, at All1-1 to A11-13; 69 Fed. Reg. at 41658-57.

Even putting aside the uncertainties and
questionable assumptions associated with EPA’s
valuation of recreational fishing, the fishing benefits
that EPA actually counted represented only a very
small slice of the full spectrum of environmental
benefits associated with the rule. First, EPA’s
estimate of commercial and recreational fishing
benefits accounted for less than two percent of the
total number of aquatic organisms. subject to
impingement and entrainment. See pp. 19-20, supra.
Second, even had they accounted for one-hundred
percent of the fish killed by impingement and
entrainment, this would have covered only a very
narrow slice of the broad and complex harms caused
by cooling water intake structures.

In the documents supporting its CBA, EPA
acknowledged this significant limitation in its
analysis and itemized the wider ecological harms its
CBA omitted. EPA, fegional/l Analysis, supra, at
A9-1. The lst is too long to reproduce here, but
included, for example, the fact that “Iflish are
essential for energy transfer in aquatic food webs,
regulation of food web structure, nutrient cycling,
maintenance of sediment processes, redistribution of
bottom substrates, the regulation of carbon fluxes

ao

from water to the atmosphere, and the maintenance
of aquatic biodiversity.” Jd.

EPA had no way of valuing most of these
broader ecological impacts, both because they involve
processes that are only dimly understood by science,
and because they involve goods and services not
traded in markets. Accordingly, EPA simply left
most of these values off the balance sheet altogether.
The agency did, however, make an attempt to include
two aspects of these non-market benefits in its
quantified estimate: 1) the benefits of increased
survival of “forage” species (fish that other species
prey on), and 2) non-use benefits of the aquatic
ecosystem. As the next two sections explain,
however, even these attempts were largely
unsuccessful.

a) EPA Used a Method for Estimating the
Benefits of Increased Forage Species
Survival that Measured “Only a Small
Share of these Losses.”

EPA initially used one of two methods to
attempt to estimate the benefits of increased forage
species survival, depending on the region. In most
regions, it used “estimates of trophic transfer
efficiency to relate foregone forage production to
foregone commercial and recreational fishery yields.”
67 Fed. Reg. at 17,191. EPA acknowledged that this
method was highly incomplete in that it measured
“only the small share of these losses, . . . namely the
contribution of the forage species to the increased
biomass of landed recreational and commercial
species.” Jd. at 17,193. As such, this method left out
the myriad other ecological functions such species
undoubtedly serve.

25

In a few regions, EPA initially used a Habitat
Replacement Cost analysis instead, which used the
projected costs of habitat replacement programs for
various species suffering impingement = and
entrainment losses as a proxy for the value of forage
species losses. Jd. at 17,191. EPA viewed this
method as considerably more comprehensive, in part
because of its “recognition that impingement and
entrainment losses have impacts on all components of
the aquatic ecosystem . . . beyond that estimated by
reduced commercial and recreational fish catches.”
Id. This method also posed problems, however. EPA
acknowledged that it did “not take into consideration
ecological problems associated with introducing
hatchery fish into wild populations.” /d And it
generated significant controversy. In comments on
the proposed rule, natural resource economist, Robert
Stavins, Ph.D. of Harvard University, criticized the
Habitat Replacement Cost method as “completely
illegitimate” and “fatally flawed.” J.A. 234.

In the final rule, EPA abandoned the Habitat
Replacement Cost analysis altogether. Instead, it
used the far lower (and less complete) numbers
generated by the trophic transfer model, noting that
estimating the value of forage species. is
“challenging.” 69 Fed. Reg. at 41,657: see EPA,
Regional Analysis, supra, at Al15-1. Ultimately, then,
EPA's estimate of the benefits of increased survival of
forage species was, by its own _ admission,
significantly incomplete.

b) Although EPA Believed that “Non-Use
Benefits may be Significant,” It Ultimately
Assigned them a Value of Zero.

26

“Non-use values” are the values’. people
attribute to resources they have no intention of ever
using or experiencing directly. These are sometimes
referred to as “existence values” or “bequest values.”
EPA, Case Study Analysis for the Proposed Section
316(b) Phase Il Existing Facilities Rule, EPA-821-R-
02-001, at A9-10 (February 2002) (available at
http://www.epa.gov/waterscience/3 16b/phase2/
casestudy). Often such values are estimated using
contingent valuation studies, which can be quite
controversial. See pp. 21-22, supra. But in this
instance, EPA lacked the resources to even conduct
such studies. See EPA, Case Study, supra, at A9-10.

Instead, EPA applied a “rule of thumb” based
on a 1977 study, which inferred that non-use values
associated with water-quality improvements were
typically 50 percent of recreational use values. See
1d. 67 Fed. Reg. at 17,149. As described above, EPA’s
estimate of recreational use values was itself based
on shadow pricing methods that were far from
precise. See id. at 17,193 (“[Nlonuse benefits are
most likely understated using the 50 percent rule
because the recreational values used are likely to be
understated.”). Perhaps predictably, then, EPA’s 50-
percent rule of thumb came under considerable fire
during the comment period. Economist Robert
Stavins argued that the 50-percent rule was far too
high. J A. 231, 233. Economist Frank Ackerman, on
the other hand, argued that it was too low and that
based on more recent literature, non-use value should
be estimated at two-times use value. J.A. 220.

Ultimately, when it came to the final rule, EPA
threw up its hands and attached no monetary value
to non-use benefits at all. It explained in the
preamble that while there was “evidence that non-use
values could occur as a result of this rule, ... EPA

27

was unable, by the time of publication . . . to estimate
reliable valuations . . . lor] to fully develop and
analyze ... non-use benefit approaches.” 69 Fed.
Reg. at 41,657; id. at 41,666 (EPA believes omitted
“non-use benefits may be significant.”); id. at 41,662
(same).

By EPA’s own admission, then, even the CBA
that accompanied the proposed rule vastly under-
estimated the benefits. But by the time it issued the
final rule, EPA had backed away from so many of the
methods it had originally used to try to value non-
market goods that its benefits estimate had shrunk
by nearly ten-fold. Thus, while the total benefits
estimate for the proposed rule was $735 million, EPA,
EBA Proposed Rule, supra, at D1-4, the benefits
estimate associated with the final rule was just $83
million. 69 Fed. Reg. at 41,666.

C. EPA’S Estimate of Benefits for the Phase II
Rule was Grossly Incomplete, Making Any
Comparison with Costs. Irrational and
Misleading.

The story of the Phase II rulemaking provides
a helpful illustration of how formal CBA so often
produces results that actually serve to undermine
“common sense” and “rational” decision making.
Entergy Br. 4, 29. EPA’s estimate of the benefits of
the rule was so hopelessly unreliable and incomplete
that any attempt to compare it with the costs was
patently irrational. Thus, rather than ciarifying the
relevant issues and helping to identify “whether the
regulation did more harm than good” (id. at 21),
formal CBA in this instance actually served to
obscure relevant issues and mislead decisionmakers.

28

Even EPA’s own estimates of the benefits of its
regulation varied wildly—by a factor of nearly ten—
between the proposed and final rules. This disparity
resulted largely from the fact that the methods for
monetizing non-market ecological values are so
inadequate and so contestable. In the end, after
receiving extensive criticism on its CBA for the
proposed rule, EPA threw up its hands and simply
left most benefits out of the analysis entirely.

In reporting the costs and benefits of the final
rule, EPA flatly acknowledged that its benefits
estimate was incomplete, making a meaningful
comparison with costs impossible: “EPA notes that
these analyses are based on a comparison of a partial
measure of benefits with a complete measure of costs;
therefore, the results must be interpreted with
caution.” 69 Fed. Reg. at 41,666; see EPA, EBA Final
Rule, supra, at D1-5 (“A comparison of complete costs
and incomplete benefits does not provide an accurate
picture of net benefits to society.”). But even the far
larger estimate of benefits that accompanied the
proposed rule was, by EPA’s own admission, grossly
incomplete. Throughout the preamble to the
proposed rule, EPA repeatedly pointed out the
weaknesses in its analysis that likely caused it to
underestimate benefits.

I. OIRA Pressured EPA into Weakening its
Proposed Rule on the Basis of the Highly
Flawed, Incomplete, and Irrational Results of

its CBA.

Nonetheless, it appears to have been on the
basis of this highly flawed, incomplete, and irrational
CBA that OIRA pressured EPA into changing its
proposed rule. EPA’s originally proposed rule

29

identified closed cycle cooling as the best technology
available for the 59 largest and most environmentally
damaging facilities of the 539 subject to the rule.
EPA, OMB Review Draft for the Proposed Section
316(b) Rule for Large Cooling Water Intake
Structures at Existing Power Generating Facilities,
Docket W-00-32, DCN # 4-4005, p. 72 (Dec. 28, 2001).
These were facilities that either were “located on a
tidal river or an estuary” and had “an intake flow
that exceeds one percent of the tidal excursion” or
were “located on an ocean” and had “an intake flow
that exceeds 500 [million gallons per day].” /d.

EPA based this proposal, on its findings that
closed-cycle cooling “is the most effective technology
for reducing both entrainment and impingement,” /d.
at 75, that closed-cycle cooling is “commercially
available and economically achievable” and already
in use at 21 percent of existing facilities, 7d at 74-75,
and that “facilities can and have installed these
technologies years after the facility began operation,”
id at 75. Additionally, EPA found that “the
compliance costs for [this proposal] would . . . be low
compared to firm-level revenues.” See 67 Fed. Reg. at
17,158 (82% of firms would incur compliance costs of
less than 0.5 % of revenues). Accordingly, this
proposal met the “best available technology” test of
section 316(b) with flying colors. See Economists
Frank Ackerman, et al. Amicus Br. 16-23.

When it came to formal CBA, on the other
hand, EPA was_ prescient. Anticipating the
conclusion it would ultimately reach two-and-a-half
years later after completion of the final rule, the
agency stated that “EPA cannot perform a complete
benefit-cost comparison because not all of the benefits
resulting from the proposed regulatory alternative
can be valved in dollar terms.” EPA, OMB Review

30

Draft, supra, at 211. At this stage, EPA’s formal
CBA was not nearly complete: In many places the
draft contained “X’s” in leu of numbers that were to
be filled in later. /d. at 205, 211-12. But EPA could
tell, even then, that the analysis would not ultimately
be able to quantify enough aspects of the benefits of
the proposed rule to provide a rational comparison
with costs.

On December 28, 2001, after years spent
researching and preparing the proposed rule, EPA
submitted it to OIRA for review. 67 Fed. Reg. at
17,208. The rule that emerged 60 days later was
drastically changed. The OIRA review process
resulted in 58 “major changes” to the rule, including
the remeval of the closed cycle cooling requirement
for the 59 most damaging facilities and the addition
of a provision allowing a site-specific determination of
permit requirements based on a cost-benefit analysis.
See EPA, Summary of Major Changes During
Interagency Review, Docket W-00-32, DCN # 4-4005,
at 1 (2002).

Although records of the communications
between EPA and OMB during the review process
have not been made public, we can surmise the
rationale for the change by comparing the preamble
to the proposed rule that was ultimately published in
the Federal Register with the preliminary draft EPA
submitted to OLRA. In the proposed rule, the closed
cycle cooling requirement was demoted from the
preferred option to an alternative option. 67 Fed.
Reg. at 17,156-58. In describing this option in the
preamble to the proposed rule, EPA reiterated all the
reasons it would be a good option, including its
economic practicality. Jd. at 17,158. The only
rationale EPA offered for why this option might
ultimately be rejected was formal CBA. “EPA notes

31

that the incremental costs of this option relative to
the proposed option ($413 million) significantly
outweigh the incremental benefits ($146 million).”
Id. We can only surmise that this was the rationale
OIRA used to push EPA to change the rule, since
OIRA’s mandate, after all, is to ensure
implementation of the mandate for formal CBA in
Executive Order 12,866. See pp. 12-14, supra.

2. Although EPA Cited Its Formal CBA as the
Basis for Rejecting the Closed Cycle
Alternative, the Proposed Rule Preamble
went on to Cite Dozens of Reasons the
Benefits were Under-Estimated.

Thus, the EPA rejected the more stringent
closed cycle alternative on the grounds that its formal
CBA showed the margifial costs of that alternative
outweighing its marginal benefits. 67 Fed. Reg. at
17,158. Nonetheless, the remainder of the preamble
to the proposed rule went on to systematically
undermine that conclusion by repeatedly citing the
myriad reasons EPA believed its analysis had under-
estimated the real magnitude of the rule’s benefits.
See, e.g. id. at 17,194-96 (repeatedly noting “several
critical caveats and limitations of the analysis”). The
preamble is peppered with literally dozens of such
disclaimers. See, e.g., id. at 17,190 (“the magnitude
of impingement and _ entrainment is_ often
underestimated.”); id. at 17,192 (“Secondary impacts .

[like] effects on marinas, bait sales, property
values, and so forth are not included, even though
they may be significant.”); sd. (“Current fishing
mortality rates (and resulting estimates of yield)
often reflect depleted fisheries, not what the fisheries
should or could be if not adversely impacted by

32

impingement and entrainment and other stressors.”);
id. at 17,193 (“Forage losses are often valued at only
a fraction of their potential full value.”); zd. (“Impacts
on threatened and endangered species are not fully
captured.”).

Indeed, references to the benefits being
“underestimated” or “understated” appear 23 times in
the preamble. See id. at 17,190-203. Nowhere does it
state that the benefits have been either overstated or
overestimated. At one point, EPA entertained that
possibility, see id. at 17,192 (“the Agency’s benefits
estimates could be either over- or under-estimated”),
but then quickly dismissed it:

However, because of the many factors
omitted from the analysis (typically
because of data limitations) and the
manner in- which several key
uncertainties were addressed, EPA
believes that its analysis is likely to lead
to a potentially significant
underestimate of baseline losses and,
therefore lead to understated estimates
of regulatory benefits.

Id.

Natural resources economist Frank Ackerman,
Ph.D., in comments submitted on the proposed rule,
suggested that even just correcting for a few of the
many inaccuracies in EPA’s benefits estimate would
yield an estimate 4-6 times as high. J.A. 223. If that
were true, then the abandoned closed cycle option

would probably actually pass an incremental cost-
benefit test. See 67 Fed. Reg. at 17,158.

33

3. The Phase Il Rulemaking Illustrates How
CBA Undermines Rational Agency
Decisionmaking.

The story of the Phase II CBA illustrates why
formal CBA so often leads to irrational results that
defy common sense and actually obscure the factors
relevant to discerning whether a regulation does
more harm than good. The benefits estimate for the
Phase II rulemaking never produced anything even
approaching a reliable or meaningful number.
Indeed, the numerical estimate jumped wildly
between the proposed and final rules, changing by
nearly a factor of ten, from $735 million under the
proposed rule to $83 million under the final rule. See
p. 27, supra.

Even the larger number was, by EPA’s own
frequent admission, grossly and _ hopelessly
incomplete. The documents describing the CBA were
peppered with caveats and disclaimers, describing
the myriad reasons why the benefits figure grossly
underestimated the true benefits of the rule.

Yet, despite the innumerable caveats, EPA still
resorted to the raw numbers when it came time to
make a decision. It rejected the closed cycle cooling
alternative because “the incremental costs of this
option relative to the proposed option ($413 million)
significantly outweigh the incremental benefits ($146
million).” 67 Fed. Reg. at 17,158. This time there
were no caveats or disclaimers.

This is not an unusual story. One of the
problems with formal CBA is that, no matter how
many narrative explanations, caveats and qualitative
descriptions are included in the analysis, regulators
face an irresistible temptation to reduce formal CBA
to its raw numbers. In this way, they can make

34

decisions look easy and definitive rather than
difficult and ambiguous. Jee Frank Ackerman &
Lisa Heinzerling, Pricing the Priceless: Cost-Benefit
Analysis of Environmental Protection, 150 U. Penn.
L. Rev. 1553, 1579-80 (2002) (unquantifiable benefits
often given lip service in CBA but ultimately ignored;
citing arsenic CBA as example where “{s]ubsequent
public discussion [of the CBA] inevitably referred
only to the EPA’s numerical analysis and forgot about
the cases of avoided illness that could not be
quantified”); Richard Parker, Grading the
Government, 70 U. Chi. L. Rev. 1345, 1348-49,
1404-06 (2003) (observing increasingly prevalent
phenomenon of “regulatory score cards,” which
“reduce... hundreds of pages lin a CBA] to a few
summary statistics”).

But, as the Phase II rule illustrates, when
monetary estimates of benefits and costs are grossly
incomplete and highly contestable, as they so often
are, attempting to compare them “does not provide an
accurate picture of net benefits to society.” EPA,
EBA Final Rule, supra, at D1-5. On the contrary, it
leads to irrational results that actually obscure the
relevant issues behind a false veneer of scientific
objectivity and accuracy. In this way CBA actually
reduces the transparency and _ “rationality of
government decisionmaking.

Rather than rationalizing and disciplining
agency decision making, formal CBA actually renders
agency decision making more vulnerable to interest
group pressure. Because the numbers it produces are
so unreliable and contestable, an interest group
unhappy with the result can almost always find an
economist who can make a credible argument for the
opposite outcome. Indeed, Professor Cass Sunstein

35

made this observation in connection with his careful
analysis of EPA’s formal CBA of the arsenic rule:

We are now in a position to see the
multiple possible challenges to any
agency decision that involves cost-

benefit balancing .... We can see how
creative citizens and lawyers,
representing water systems or

environmentalists, might be able to
mount reasonable challenges to EPA’s
decisions, regardless (almost) of the
content of those decisions.

Sunstein, Aysk, supra, at 179. Thus, rulemaking
devolves into a contest over which interest group has
the resources to hire the best credentialed economists
rather than which side has the better argument.

In sum, formal CBA often undermines rational
agency decisionmaking. That occurred in this case,
when EPA, at OIRA’s urging, rejected the closed cycle
alternative on the grounds of a comparison of the raw
numbers produced by its formal CBA. EPA compared
a monetary estimate of benefits that was hopelessly
unrehable and incomplete to a relatively compete
estimate of costs and used the results of this fuzzy
math to justify its decision.

CONCLUSION

Formal CBA of the type used by EPA in this
case undermines rational decisionmaking because it
inevitably produces incomplete, unreliable, and
contestable results. It requires analysts to attach a
dollar figure to values like human lives, healthy
ecosystems, and endangered species that cannot be

36

meaningfully or reliably quantified and monetized.
Formal CBA is not a common sense tool for rational
decision making, but a very specific and highly
controversial method for evaluating regulations.
Congress had good reason to prohibit its use under
the CWA in general and under § 316(b) in particular.

Respectfully submitted.

PROFESSOR AMY SINDEN

Counsel of Record

Temple University Beasley School of Law
1719 North Broad Street

Philadelphia, PA 19122

(215) 204-4969

October 6, 2008

---

Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/brief%3Amicro_IA40385015_0048%3A35. Public record. Not legal advice.