Negotiation of Environmental
Disputes: A New Role for the Courts?

Patricia M. Wald*

Just a decade ago, Judge Harold Leventhal-wrote a landmark
article entitled "Environmental Decisionmaking and the Role of
the Courts,"' in which he pointed out that courts no 'longer for-
mulated and directed the application of pertinent legal rules in
the environmental area as they once had. This was true, he said,
because "[p]rimary responsibility has been vested in executive of-
ficials and independent regulatory agencies" under the new envi-
ronmental legislation. 2 Despite this displacement, Judge
Leventhal continued, the courts still maintained "a role of review
... of major significance." Through the exercise of review,
Judge Leventhal stated, courts have exerted "a pervasive influ-
ence over the legislation's implementation."-3

To fulfill that role, he said, a court was required to "study the
record attentively, even the evidence on technical and specialist
matters, 'to penetrate to the underlying decisions of the agency,
to satisfy itself that the agency has exercised a reasoned discretion
with reasons that do not deviate from or ignore the ascertainable

* Judge, United States Court of Appeals for the District of Columbia Circuit. B.A. Con-
necticut College, LL.B. Yale Law School. The article is based on remarks made at the
Second National Conference on Environmental Dispute Resolution in Washington, D.C.
on October 2, 1984. The author wishes to thank her law clerk, Stephen Presser, for his
assistance and Philip Harter for his good-natured and helpful post-conference comments
on the paper.

The Columbia Journal of Environmental Law would like to thank Norman Silber, Co-
lumbia Law School, Class of 1986, for his editorial assistance on this article.

1. Leventhal, Environmental Decisionmaking and the Role of the Courts, 122 U. PA. L. REV. 509
(1974). For an overview ofJudge Leventhal's views on the relationship between courts and
administrative agencies, see Estreicher, Pragmatic Justice: The Contrtibutions of Judge Harold
Leventhal to Administrative Law, 80 COLUM. L. REV. 894 (1980).

2. Leventhal, supra note 1, at 510.
3. Id.



2 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 10:1

legislative intent.' ,,4 According to Judge Leventhal, the court
needed to insure that the agency had taken a "hard look" at the
salient problems and engaged in reasoned decision-making; the
decision itself must come "within a zone of reasonableness." 5

The objective of judicial review in environmental cases, he said,
was "to provide supervision that emphasizes broad questions of
fairness . . . .[and] to combine supervision with restraint, mak-
ing the court a genuine kind of partner with the agency in the
overall administrative process." 6 The courts had assumed a "cen-
tral role of ensuring the principled integration and balanced as-
sessment of both environmental and nonenvironmental
considerations in federal agency decision-making." 7 In sum,
Judge Leventhal saw the courts and the Environmental Protection
Agency ("EPA") as engaged in a collaborative partnership in wise
decision-making. Judge Leventhal's stance was an activist one: his
hard look philosophy extended by implication from decision-mak-
ing of agencies to the scrutiny of the reviewing courts. Although
the primary duty of a court was to insure compliance with legisla-
tive intent, exploration of all salient facets of the problem and
reasoned analysis were the watchwords. Like an exam grader, the
court would give a pass or fail mark depending on whether the
course material was covered and whether the use made of that
material was acceptably intelligent.

To be sure, even a hard look court would not ordinarily raise
new issues, no matter how glaring the agency's errors might be.
Once issues were presented, however, a court could and often did
become actively involved: it might reject the arguments of both
sides, suggest compromise positions, or remand cases for the
consideration of new solutions. Under Judge Leventhal's vision,
a reviewing court's primary loyalty was to the Congress, which
gave it the power of independent review. The agency, in turn,
perceived the court as a major participant in construing and im-
plemen'ting the statute. An agency's efforts were often directed
toward getting the court to agree with a particular interpretation
or action, and its internal decisions were often predicated on the
likelihood of obtaining that court's approval. The agency might,

4. Id. at 511.
5. Id. (citing Greater Boston TV Corp. v. FCC, 444 F.2d 841, 852 (D.C. Cir.), cert. denied,

403 U.S. 923 (1971)).
6. Id. at 554.
7. Id. at 555.



Environmental Negotiation

at times, actively welcome judicial review to obtain such an en-
dorsement since a court's pronouncements embodied preceden-
tal implications for future agency policies and directions that went
far beyond the immediate case.

That was judge Leventhal writing in 1974; this is 1985. In the
intervening decade, we have witnessed two phenomena that have
significantly altered both the regulatory landscape and the courts'
role in agency processes. First, we have seen the emergence of
massively complex environmental programs whose success essen-
tially depends on some type of collaborative partnership, not be-
tween the courts and the agency, but between the EPA and the
parties it regulates. Second, the ability of the legal system to
oversee effectively multifaceted environmental regulation has in-
creasingly been cast into doubt. To some extent, this doubt is
reflected in the Supreme Court's recent restrictions on the role of
appellate review in the implementation of environmental statutes
and other agency legislation.8

More generally, this skepticism is embodied in the movement
among lawyers and commentators away from traditional judicial
oversight through the adversary system and in the direction of
more collaborative techniques of dispute resolution. As Owen
Fiss recently has pointed out, this movement is so strong that it
has acquired its own acronym: 'ADR' or Alternative Dispute Res-
olution. 9 Proponents of ADR argue that application of ADR tech-
niques to environmental disputes will enable us to overcome a
purported legal and regulatory malaise and to reinvigorate the
business of safeguarding the environment.

Before adopting particular approaches, it might be is helpful to
examine ADR from the perspective of an appellate judge.' 0 This
examination first discusses the latest Supreme Court ruling on the
scope ofjudicial review of agency actions, and then surveys briefly
the current impact of ADR techniques on environmental disputes.
It goes on to explore the tools presently available to district and
appellate courts sympathetic to an expanded use of negotiation
and settlement. Finally, it considers in some detail two "hot top-

8. See Chevron, U.S.A. v. Natural Resources Defense Council, - U.S. -, 104 S. Ct.

2778 (1984).
9. Fiss, Against Settlement, 93 YALE L.J. 1073 (1984).
10. For a discussion of environmental ADR from the perspective of an attorney for the

federal government see Dinkins, Shall We Fight or Shall We Finish: Environmental Dispute Reso-
lution in a Litigious Society, 14 ENVrL. L. REP. (ENvrL. L. INST.) 10398 (1984).

1985]



4 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 10:1

ics" in what might be called "environmental ADR": the possibili-
ties for encouraging settlements in EPA enforcement actions, and
a proposal for negotiated regulations. Both of these topics di-
rectly implicate the sometimes conflicting aims of ADR and judi-
cial review. By giving each a "hard look," the discussion raises
some pre-nuptial concerns about the proposed marriage between
ADR and environmental regulation, concerns that must be at-
tended to if settlement and negotiation are to become mainstays
of the agency process.

I. THE CURRENT ROLE OF APPELLATE REVIEW

Before discussing new roles for the courts in environmental
decisionmaking, it is important to have a realistic sense of what
their currently legitimized role is. During the October 1984 term,
in Chevron v. Natural Resources Defense Council" I the Supreme Court
laid down what seems to me a quite restricted role for reviewing
courts in construing environmental statutes. Chevron involved
Clean Air Act permits for new or modified stationary sources in
"non-attainment" states.' 2 At issue was a 1981 EPA regulation

which allowed a state to adopt a plantwide or so-called "bubble"
definition of "stationary source" and departed from the prior reg-
ulation that required each emission source within a plant to be
assessed independently. Neither the applicable definitions of
"stationary source" in the Clean Air Act' 3 nor the legislative his-
tory of the Act addressed the issue directly.' 4

The opinion emanating from our court sought to fill this gap by
considering the overriding purposes of the Act. It concluded that
a more liberal application of the plantwide bubble concept to
states that had not yet met EPA air quality standards would run
counter to the Act's core purpose of improving air quality.15 The
Supreme Court reversed, however, declaring that the "basic legal
error of the Court of Appeals was to adopt a static judicial defini-
tion of the term 'stationary source' when it had decided that Con-
gress itself had not commanded that definition."' 6 Justice

11. 104 S. Ct. 2778 (1984).
12. 42 U.S.C. § 7411 (1982).
13. 42 U.S.C. § 7411(a)(3) (1982).
14. See Chevron, 104 S. Ct. at 2785-87.
15. Natural Resources Defense Council v. Gorsuch, 685 F.2d 718, 725-28 (D.C. Cir.

1982).
16. Chevron, 104 S.Ct. at 2793.



Environmental Negotiation

Stevens, writing for a unanimous six-justice court, marked in bas
relief the present parameters for judicial review of agency inter-
pretations of statutes:

When a court reviews an agency's construction of the statute
which it administers, it is confronted with two questions. First,
always, is the question whether Congress had directly spoken
to the precise question at issue. If the intent of Congress is
clear, that is the end of the matter; for the court, as well as the
agency, must give effect to the unambiguously expressed intent
of Congress. If, however, the court determines Congress has
not directly addressed the precise question at issue, the court
does not simply impose its own construction on the statute, as
would be necessary in the absence of an administrative inter-
pretation. Rather, if the statute is silent or ambiguous with re-
spect to the specific issue, the question for the court is whether
the agency's answer is based on a permissible construction of
the statute. 17

This view sounds distinctly different from the collaborative
partnership between courts and the EPA Judge Leventhal was
talking about. It is true that any comparison of Judge Leventhal's
article and Justice Stevens' opinion is complicated by the fact that
Judge Leventhal was talking mainly about agency applications of
statutory terms, and Justice Stevens was discussing agency con-
structions of statutory language. But there is no bright line be-
tween agency application and agency interpretation of statutory
terms for the purposes of judicial review. Indeed, both Judge
Leventhal and Justice Stevens used the same sensitive and excru-
ciatingly vague standard of "reasonableness" to judge agency ac-
tion. And it is difficult to see the logic of allowing judges a freer
hand in reviewing the evidence or expertise supporting an agency
application than they are permitted in considering an agency's in-
terpretation of the statutory wording itself. Thus, even in light of
last years' "Airbags" case,' 8 it is impossible for me to conclude
that Judge Leventhal and Justice Stevens envisioned the same de-
gree of judicial inquiry or oversight into agency action.

In Chevron, the litigants were admonished not to "wag[e] in a
judicial forum a specific policy battle which they ultimately lost in
the agency . . . ." Such policy arguments, the Court continued,
"are more properly addressed to legislators or administrators,

17. Id. at 2781-82 (footnotes omitted).
18. Motor Vehicles Mfrs. Ass'n v. State Farm Mut. Auto Ins. Co., - U.S. -, 103 S. Ct.

2856 (1983).

1985]



6 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 10:1

not to judges."1 9 According to the Court, where a statute has con-
flicting policy objectives and "Congress intended to accommo-
date both interests, but did not do so itself on the level of
specificity" required to give a clear answer to the statutory ques-
tion, the agency deserves deference.2 0

Judges are not experts in the field, and are not part of either
political branch of the government. Courts must, in some
cases, reconcile competing political interests, but not on the
basis of the judges' personal policy preferences. In contrast, an
agency to which Congress had delegated policymaking respon-
sibilities may, within the limits of that delegation, properly rely
upon the incumbent administration's views of wise policy to in-
form its judgments. While agencies are not directly accountable
to the people, the Chief Executive is, and it is entirely appropri-
ate for this political branch of the Government to make such
policy choices-resolving the competing interests which Con-
gress itself either inadvertently did not resolve, or intentionally
left to be resolved by the agency charged with the administra-
tion of the statute in light of everyday realities.2 1

It may turn out that the Court will not pursue a role of review
as deferential as its language in Chevron suggests. I personally
conclude, however, that although Chevron reflects no fundamental
departure from several recent pronouncements on statutory con-
struction, 22 it is a mood piece-and the mood is a rather somber
one for proponents of active judicial review of agency decisions.23

The climate for judicial partnership in environmental decision-
making, in other words, has cooled considerably since 1974;
judges are limited partners at best. Perhaps this narrowing ofju-
dicial oversight is a response to the uncertainty and delay that a
broad, far-reaching brand of judicial review introduces into the
regulatory process. If so, it may well be fueled by the same fire
that lights the way toward the application of ADR to environmen-
tal disputes.

Despite the reduced role of judicial review which the Chevron
decision may portend and the growing attempt to divert conflicts

19. Chevron, 104 S. Ct. at 2793.
20. Id.
21. Id.
22. See, e.g., FEC v. Democratic Senatorial Campaign Comm., 454 U.S. 27, 32 (1981);

SEC v. Sloan, 436 U.S. 103, 118 (1978).
23. See also Reed, Three Strikes and the Umpire is Out: The Supreme Court Throws the D.C.

Circuit Out of the Bubble Review Game, 14 ENVTL. L. REP. (ENVrL. L. INST.) 10338 (1984). ("In
Chevron .. .the Supreme Court set a new tone of judicial restraint in reviewing agency
action.").



Environmental Negotiation

from judicial resolution through ADR, environmental ADR in the
form of consent decrees or negotiated regulations may actually
increase rather than decrease judicial influence in environmental
decisionaking; at best, ADR could replace the old problems of
appellate review with a new set. Before discussing those
problems, however, the ways in which courts can help or hinder
ADR's growth should be investigated.

II. THE CURRENT STATUS OF ENVIRONMENTAL ADR

A. The Present Track Record

How big a bite out of environmental litigation ADR techniques
can take is a matter of some dispute. After studying six environ-
mental cases in which some form of settlement or mediation had
been employed, Allan Talbot concluded that, at best, such tech-
niques could be successful only 10% of the time: in those cases
where the issues are clearly defined, where there is a balance of
power between the parties, and where the parties' precise objec-
tives cannot be achieved (at least without great risk) without ne-
gotiation. 24 Timing, Talbot said, is all-important. Some issues
not negotiable at the onset of the dispute become negotiable over
time. 25 On the other hand, the existence of numerous parties or
factions, ideologically based disputes, or non-predictable long
term trends militate against successful negotiation of environ-
mental disputes.2 6

Alan Miller echoed many of the same concerns in his recent
report of successfully negotiated regulations affecting the steel in-
dustry.27 In March, 1983, the Iron and Steel Institute, NRDC and
EPA filed a settlement agreement in the Third Circuit resolving
all legal challenges to water pollution effluent guidelines for the
steel industry. The settlement avoided protracted litigation and
allowed permits to be issued without the threat of subsequent at-
tack and judicial review. All in all, the settlement was just the sort
of regulatory action promised by ADR proponents. Although the

24. A. TALBOT, SETrLING THINGS: SIX CASE STUDIES IN ENVIRONMENTAL MEDIATION 91
(1983).

25. Id. at 91; Harter, Regulatory Negotiation: The Experience So Far, RESOLVE 1 (Winter
1984).

26. TALBOT, supra note 24, at 91.
27. Miller, Steel Effluent Limitations: Success at the Negotiating Table, 13 ENVTL. L. REP.

(ENVTL. L. INST.) 10094 (1983).

1985]



8 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 10:1

steel negotiations were apparently a remarkable success, Miller,
the chief NRDC negotiator, cautioned that:

negotiation can be successful only in special circumstances.
The first requirement is arms' length negotiation with all sides
legally and technically well represented. In this case, each
party had something to lose through lengthy court proceedings
... . More frequently, either the bargaining process is one-
sided or one party has nothing to gain from negotiation. When
former EPA Administrator Anne Burford negotiated modifica-
tions to lead standards for the benefit of an oil refinery, no en-
vironmentalists were involved and an adversarial response was
assured. In the dance of environmental negotiation, it takes
three to tango.2 8

In the steel case, only 13 out of hundreds of pollution standards
were in dispute so the challengers had much to gain by settling in
lieu of an arduous and expensive litigation. The EPA also had a
very real incentive to dance since the steel industry was only the
first of many industries for which such limitations had to be estab-
lished. 29 All sides (environmental groups, industry, and the EPA)
had highly competent and experienced counsel who had un-
doubtedly worked together or at least fought together in the past.
Both Miller's comments and the Talbot study, then, suggest that
successful negotiation may be the exception and not the rule in
environmental disputes.

An area that perhaps most dramatically illustrates the gap be-
tween the potential for environmental ADR and its current per-
formance is the Superfund law for environmental cleanups.a0 It is
obvious to almost everyone that voluntary settlements are the
best and perhaps the only hope for Superfund's success.3 1 The
EPA has a duty to investigate between 18,000 and 22,000 waste
disposal sites for possible inclusion on the National Priorities List
which establishes sites that are eligible for remedial action under
Superfund; currently, there are between five and six hundred
such sites on the list, and the estimated clean-up cost is eight to

28. Id. at 10095.

29. Id. at 10094.

30. Comprehensive Environmental Response, Compensation and Liability Act of 1980

[hereinafter Superfund], 42 U.S.C. § 9601-9657 (1982).
3 1. See e.g., Atkeson, Settlement Uncertainty Persists Under Superfund, Legal Times, July 30,

1984, at 45, col. 3. To the extent that settling contributors pay for the cleanup themselves,

Superfund monies are freed up for those sites involving contributors that cannot be found

or are unable to pay. See Stoll, Superfund Settlement Memo Needs Clarification, Legal Times,
Dec. 24-31, 1984, at 15, col. 1.



Environmental Negotiation

ten million dollars per site.3 2 Until a few months ago, however,
the EPA did not seek Superfund settlements in any substantial
number. Last summer the Administrative Conference of the
United States (ACUS) decried what it termed the "slow and ex-
pensive implementation of Superfund clean-ups and. . . specifi-
cally recommended a far more liberal use of negotiated
settlements." According to ACUS, then, the agency currently
"puts too little stress on negotiations and has adopted a series of
procedural and substantive requirements that unnecessarily con-
strict the number" of such negotiations.33

One commentator suggested that the government did not seek
more Superfund settlements because it won so often in court. 34

Courts have held that the liability of the parties who contributed
to the sites are joint and several;3 5 and the causation standards
they have used to assess liability are, according to some, quite
liberal.36 Probably more important, however, was the fact that
the government reportedly would not contribute to any settle-
ment where the parties did not themselves agree to pay at least
80% of the clean-up costs, and it was reluctant to grant releases
from future governmental actions if the settlement did not ac-
complish the clean-up satisfactorily. 37

There was a further problem of what legal principles governed
the liability of settlers in later contribution suits brought by non-
settlers against them. At least one commentator has attributed a
large part of the general reluctance to settle to the legal uncer-
tainties about such future contribution suits:

Given the importance all parties ascribe to voluntary Superfund
settlements, it is striking that 31/2 years after its initial passage,
the Superfund law applicable to voluntary settlements is so un-
developed . . . . What is clearly needed at this point is clear

32. Id., citing 75 Removal Actions Expected this Year at Superfund Sites, EPA OFficial Says, 15

ENV'T. REP. (BNA) [Current Developments] 374-75 (June 29, 1984)); see also Stoll, supra,
note 31, at 22 n.4.

33. See Regulatory Roundup: ACUS Calls for More Negotiation in Waste Cases, Legal Times,
July 9, 1984, at 10, col. 1.

34. See Atkeson, supra note 31.
35. United States v. South Carolina Recycling and Disposal, Inc., 14 ENv-rL. L. REP.

(ENVTL. L. INST.) 20272 (D.S.C. 1984); United States v. Wade, 577 F. Supp. 1326 (E.D. Pa.
1983)

36. Rogers, Three Years of Superfund, 13 ENVTL. L. REP. (ENvrL. L. INST.) 10361 (1983).

37. Atkeson, supra note 31 (citing Administrative Conference Keeps Stronger Stance on

Superfund Negotiation, Pesticide and Toxic Chemical News, 31 (July 11, 1984)); see also Stoll,
supra note 31.

1985]



COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 10:1

attention to policy; and legal obstacles to voluntary Superfund
settlements. This needs to be followed by some simple well
drafted legislation on contribution. 38

Recently, however, the Government has embarked on a more
aggressive settlement program. EPA has announced that it will
negotiate with contributors even if their contributions will not
add up to 80% of clean-up costs so long as their joint contribu-
tions will pay a "substantial proportion" of the costs. For the first
time, it will consider paying the remainder costs out of Superfund
itself. The EPA will now take a more liberal stance in providing
releases from future liability to settlers and it may even take steps
to protect them from future liability to nonsettling contributors
who are sued by the government. De minimis contributors may
be allowed to make "cash out" payments and escape litigation al-
together. The EPA settlement regulations list ten criteria for
evaluating settlements including the volume and toxicity of the
waste, strength of case, ability to pay, other inequities or aggra-
vating factors. Settlements may be negotiated for specific phases
of the clean-up, i.e., investigation, feasibility study, surface re-
movals, groundwater remedies. 39

Still, the enormous complexity of the issues that must be re-
solved before settlements in Superfund cases can be worked out
suggests how far we have to go. Moreover, it suggests a dynamic
and continuous role for the district judge, even where settlements
are actively encouraged. How are all the potential tortfeasors to
be identified? Must they be brought into the original government
suit for liability? What principles of apportionment will be used?
If settlements are to come about, the parties will clearly have to

38. Atkeson, supra note 31, at 45.
39. The settlement guidelines were published in the Federal Register in February 1985.

See 50 Fed. Reg. 5034 (Feb. 5, 1985). All settlement agreements must contain a "re-
opener" clause to take care of serious but previously unknown conditions that emerge at
the site as well as previously unknown scientific information that develops about known
conditions. See Moore, New EPA Guidelines May Facilitate Waste Cleanups, Legal Times, Dec.
10, 1984 at 1, col. 3. The EPA has also taken the position that settling parties are pro-
tected from third party liability under common law principles akin to those set out in the
Uniform Contribution Among Tortfeasors Act, § 1 (1955), see 12 UNIFORM LAws ANNO-

TATED 57, 63-64 (1975). Not all courts accept that proposition, however. See, e.g., Dono-
van v. Dorfman, No. 84-1287, slip op. at 13-20 (7th Cir.Jan. 3, 1985) (declining to adopt a
rule insulating settling parties from future contribution liability in approving a settlement
of an ERISA action). Accordingly, the EPA says that it will seek legislation specifically
applying those common law principles to Superfund settlements; in the interim, it may
reduce judgments against nonsettling tortfeasors by an amount sufficient to protect the
settlers from third party suits. See Moore, supra; Stoll, supra note 31.



Environmental Negotiation

work out some cost allocation formula. Finally, a bill passed by
the last Congress might complicate matters even further by re-
quiring public participation through notice and comment in some
settlements under the Solid Waste Disposal Act. 40

The Superfund situation thus illustrates some important uncer-
tainties surrounding environmental ADR. Settlements seem the
best way to expeditiously implement the Superfund law, yet legal
perplexity about the principles governing the scope of liability
and rights of contribution currently impede such settlements.
Certainly, from ajudicial observer's point of view, even more spe-
cific statutory guidance for settlements seems desirable.

So right now, environmental mediation or negotiation is a
promising infant with unknown potential and a short track record.
But the Superfund example and some other recent studies under-
score an important and often overlooked aspect of ADR: negotia-
tion rarely eliminates court action altogether. Rather, as will be
discussed later, it only changes the nature of the subsequent judi-
cial proceeding. In most cases the impetus for negotiation
springs from the possibility of full-scale litigation and even those
disputes that undergo some aspect of ADR are normally submit-
ted to the courts for initial approval or later modification or en-
forcement. Accordingly, a sympathetic judicial attitude toward
ADR can often be critical in encouraging the use of negotiation
and settlement in environmental disputes.

B. The Current Judicial Role in Promoting Settlements

The focus of judicial encouragement for environmental negoti-
ation or mediation is primarily in the district courts, where the
majority of settlements occur and where consent decrees are nor-
mally entered. There is ample authority in the Federal Rules of

40. See Hazardous and Solid Waste Disposal Amendments of 1984, Pub. L. No. 98-616,
98 Stat. 3221 (to be codified at 42 U.S.C. § 3251 et seq.). The amendments recognize, at
least implicitly, the role of settlement in superfund litigation. Section 404 provides that
whenever the United States or the EPA proposes to settle a suit brought under the Solid
Waste Disposal Act, "notice, and an opportunity for a public meeting in the affected area,
and a reasonable opportunity to comment on the proposed settlement prior to its final
entry shall be afforded to the public." This public participation provision also provides
that the government's decision to settle "shall not constitute a final agency action subject
to judicial review under this Act or the Administrative Procedure Act." Id. The amend-
ments do not address the issue of apportionment. They do however, absolve transporters
of hazardous waste of contribution liability if the transporter is no longer involved in the
transportation or storage of the waste and if the transportation of the waste was under a
simple contractual relationship. See Section 491(c).

1985]



12 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 10:1

Civil Procedure for trial judges to encourage settlements in ap-
propriate cases. Recently amended Rule 16 expressly authorizes
the court to "direct the attorneys for the parties . . . to appear
before it for a conference or conferences before trial for such pur-
poses as . . . .facilitating the settlement of the case". 4 1 Subsec-
tion (c) of that Rule says the participants at any conference may
consider and take action with respect to "the possibility of settle-
ment or use of extra-judicial procedures to resolve the dispute."'42

The advisory committee note encourages active judicial pretrial
management.

[E]mpirical studies reveal that when a trial judge intervenes
personally at an early stage to assume judicial control over a
case and to schedule dates for completion by the prties of the
principal pretrial steps, the case is disposed of by settlement or
trial more efficiently and with less cost and delay than when the
parties are left to their own devices. 43

The note specifically includes judicially encouraged settlements
within that management framework by stating:

Since it obviously eases crowded court dockets and results in
savings to the litigants and the judicial system, settlement
should be facilitated at as early a stage of the litigation as
possible.

44

While Rule 16 in no way authorizes a trial judge to impose set-
tlements or mandate out of court dispute resolution, a strong ju-
dicial push in that direction is surely tolerated and must surely
have been expected. Some judges, of course, are notorious set-
tlers; others are almost total failures at settlement either because
they are reluctant to "get involved" in negotiating or because
they lack the training needed to do it well.4 5

Environmental lawyers tell me that the sine qua non of success-
ful negotiation is an assertive judge who actively pushes settle-
ment or out-of-court resolution. 46 Such a judge often keeps the
pressure up by insisting that preparation for a trial go on simulta-

41. FED. R. Civ. P. 16.
42. FED. R. Civ. P. 16(c).
43. FED. R. Civ. P. 16 advisory committee note.
44. FED. R. Civ. P. 16 advisory committee note.
45. On the changing role ofjudges as "case managers", see Resnick, Managerial Judges,

96 HARV. L. REV. 374, 378 n.14, 379 (1982). ("[jiudges have begun to experiment with
schemes for speeding the resolution of cases and for persuading litigants to settle rather
than try cases wherever possible.")

46. Judge Weinstein's handling of the "Agent Orange" class action suit in the Eastern
District of New York appears to exemplify this point. See FriedJudge Gives Final Approval to



Environmental Negotiation

neously with the settlement negotiations. Rule 16, which requires
that a scheduling order be issued in each case within 120 days of
filing the complaint, provides the framework for such pressure.47

The active settler also tries to narrow the issues and help the par-
ties to assess the costs and benefits of settling by making interim
legal rulings where possible. He or she will also demand status
reports from the parties on their efforts at resolution. The judge
may even designate a fellow judge to supervise or mediate the
settlement negotiations to avoid the spectre of over-intervention
by the judge who must find the facts if a trial eventuates. This
delegation technique is specifically mentioned in the note.48

There are three other procedural mechanisms that a district
court might be able to employ to promote settlements. I say might
because their status as settlement-promoting tools is very uncer-
tain at the moment. First, there is the controversial Rule 68
which currently allows a defendant to make an offer of settlement
anytime up to ten days before trial.49 If the offeree refuses and
eventually fails to obtain a judgment more favorable than the of-
fer, then the offeree must pay all costs incurred after the offer.
Proposed amendments to Rule 68 would allow the trial judge
broad discretion to impose a variety of sanctions, including attor-
neys' fees, upon a plaintiff or defendant for "unreasonably" refus-
ing an offer, thereby causing unnecessary delay and increased
litigation costs. 50 The current Rule 68, however, has enjoyed lim-

Accord on Agent Orange, N.Y. Times, Jan. 8 1985, at B2, col. 1; Sawyer, Legal Fees Sets in Vets'
Lawsuit, Washington Post, Jan. 8, 1985, at Al, col. 4.

47. FED. R. Civ. P. 16(b).
48. FED. R. Civ. P. 16(c)(7) advisory committee note.

49. FED. R. Civ. P. 68.

50. See Proposed Amendments to the Federal Rules of Civil Procedure, 28 (Prelim. draft
Sept. 6, 1984), reprinted in 590 F. Supp. CXXI, CXLVI (1984). The proposed amendments
would allow all parties, including claimants, to make offers of settlement- in contract to
the present rule which applies only to defendants. The proposed amendments would also
eliminate the current provision requiring the offeror to include the "costs then accrued" in
the offer. This provision has created great uncertainty over whether a Rule 68 offer must
specifically include attorneys' fees in cases brought under statutes that define "costs" to
include attorneys fees. See Advisory Committee Note to Proposed Rule 68, 590 F. Supp.
CXLVII, CXLVIII (1984). Under the proposed amendments, "acceptance of the offer
would amount to a settlement of the entire amount claimed by the offeree, including ac-
crued costs and attorneys fees" in litigation where a prevailing party would otherwise be
entitled to a statutory fee award. Id. at CXLVIII. The proposed modifications in Rule 68
would thereby encourage the controversial practice of simultaneously negotiating the mer-
its and the fee award. See infra text accompanying notes 57-58.

1985]



14 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 10:1

ited-very limited-success. 51 Judges are afraid of its potential
for forcing unfair settlements and lawyers are reluctant to invoke
it on their opponents lest it someday be invoked against them.

Although Rule 68 clearly seems geared to money suits, its lit-
eral language could apply to other actions. The present version
sanctions an offer to allow judgment "for the money or property
or to the effect specified in [the] offer"; 52 the proposed amend-
ment, in turn, speaks of "money, property, or relief specified in
the offer." Although I know of few applications of Rule 68 to
non-monetary suits, and am frankly wary of the implications of so
using it, an adenturous judge might experiment with its use in
some such suits which seem especially worthy of settlement. 53

And its use in Superfund suits which often do involve monetary
settlements would not involve a strained interpretation at all.

Rule 53 enjoys a similarly uncertain status as an authority for
settlement-minded trial judges. The Rule specifically provides
for the appointment of masters and the taxation of resulting costs
to the parties. 54 The Rule defines a master as a "referee, auditor,
examiner, or assessor." Could this include a mediator or a nego-
tiator? In the absence of such an interpretation of Rule 53, courts
apparently do not have the authority to tax parties with the costs
of a mediator or a negotiator. Rule 54 of the district court Rules
and Rule 39 of the Appellate Court Rules generally assess "costs"
to the losing party,55 but is this a legitimate "cost" under the
Rules? Moreover, the U.S. Government can be assessed costs
only "to the extent supported by law." 56 It is still doubtful then,
whether a court has the authority under existing rules to assess or
even apportion costs of mediation or negotiation against or
among the parties to a settlement. Instead, the parties have to
agree ahead of time about who should pay for mediation services.

Finally, there is the crucial matter of attorneys' fees-often crit-
ical in environmental disputes as in so many others where public
interest representation is likely to be involved. Many environ-

51. See Dombroff, Amended Rule 68 Could Lead to Legal Quagmire, Legal Times, Aug. 20,
1984, at 9, col. 1.

52. FED. R. Civ. P. 68.

53. See Liberty Mut. Ins. Co. v. EEOC, 691 F.2d 438 (9th Cir. 1978) (upholding an
award of costs to state employment commission after commission offered to enjoin certain
investigation practices).

54. FED. R. Civ. P. 53(a).
55. FED. R. Civ. P. 54(d); FED. R. APP. P. 39(a).
56. FED. R. APP. P. 39(a).



Environmental Negotiation

mental statutes provide attorneys' fees for prevailing parties.
Although those statutes have been interpreted to include settle-
ments within the concept of "prevailing," just who qualifies in a
negotiated or mediated settlement is not always clear. Moreover,
the prospects of a negotiation can often be affected by the contro-
versial tactic of requesting the waiver or reduction of attorneys'
fees as a term of settlement. This happens frequently in jurisdic-
tions which permit the practice of negotiating attorneys' fees si-
multaneously with the merits of the dispute. Only a few courts so
far have outlawed the practice of simultaneous negotiation in
cases where a judge must approve an award of attorneys' fees,57

although we have a similar request pending before us. 58 Here
again, if there is no general prohibition, the judge's role in decid-
ing whether to permit simultaous negotiation of both merits and
fees may be a critical one in encouraging or discouraging
negotiation.

In general, then, the district courts have at their disposal both
proven and unexplored techniques for promoting negotiation
and settlement. The appellate courts, by contrast, have far fewer
such tools. Direct appeals from EPA orders and rules are, of
course, most often brought in the circuit court of appeal, not the
district courts. But very few courts of appeal, including my own,
admit to any responsibility for promoting settlement. I am told
the Second Circuit and many state appellate courts have a good
track record on settlement in some (chiefly monetary) kinds of
cases, but many of the appellate judges I know shrink from the
notion that they or even staff counsel should actively engage in
case management by pushing settlements, especially in complex
cases involving significant policy issues. Although appellate
judges occasionally approve a stipulated dismissal, they generally
stay far away from the merits of a settlement. This may be a mis-
take, for we see a number of cases, including appeals from agency
actions, that do not belong in the appellate courtroom. There are
cases that settle out after a few cogent observations from the

57. SeeJeff D. v. Evans, 743 F.2d 648 (9th Cir. 1984) (rejecting the settlement of a class
action suit alleging constitutional and statutory violations where the settlement was condi-

tioned upon the plaintiff's waiver of attorneys' fees otherwise available under 42 U.S.C.
§ 1988); Prandini v. National Tea Co., 557 F.2d 1015 (3d Cir. 1977) (rejecting the settle-

ment of a class action employment discrimination suit that included attorneys' fees where

the fees would be available under Title VII).

58. Moore v. National Assoc. Sec. Dealers, 572 F. Supp. 1219 (D.D.C. 1983), appeal

argued, No. 83-2213 (D.C. Cir. Sept. 29, 1984).

1985]



16 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 10: 1

bench, and some that settle as soon as the panel is announced;
both types might better have been straightened out beforehand if
some court official had taken the zealous counsel into a room and
gently bumped heads together.

On the other hand, isolated attempts to introduce settlement
conferences in federal appellate courts have not been notably suc-
cessful. 5 9 This lack of success is not entirely surprising: appellate
courts, particularly those with current dockets, simply do not
threaten parties with the dangers of a protracted trial or of an
interminable, discovery-related delay that might induce a settle-
ment. By the time an appeal has been filed, the parties involved
have already sustained most of the costs associated with litigation.
At the appellate level, the brief is filed, the appeal is argued, and
that is that.

In the regulatory context, furthermore, there is even less op-
portunity for an appellate judge to promote settlement. It is true
that the party affected by agency action may have an incentive to
settle: it may not wish, for example, to have its alleged wrongdo-
ing aired and commented upon by a federal appellate court, or it
may realize that even an appellate victory will mean only a re-
mand followed by more agency consideration and subsequent liti-
gation. But such concerns seldom motivate the agencies. An
agency will have already articulated and justified its position
through the decision below and, unless it committed some egre-
gious fatal flaw, chances are that the agency's position will be af-
firmed. Indeed, almost 90% of agency decisions appealed to in
the D.C. Court of Appeals are affirmed. Unless there are special
circumstances, as in the steel regulations discussed earlier, 60 the
likelihood of an agency conceding important cases at this stage
does not seem great.

Yet, despite reduced incentives and few role models, settlement
at the appeals level is still an area for further study and experi-
mentation-for many of our appeals accomplish so little. Inevita-
bly, appeals are concerned with procedural errors, the adequacy
of an agency's rationale, or the quantum of evidence in the rec-
ord. They rarely result in a clean victory for the appellant or even
a discussion of the underlying merits of the controversy. Lawyers
too often seem the chief beneficiaries of appeals. Whoever profits

59. Bedlin and Nejelski, Unsettling Issues about Settling Civil Litigation: Examining "Doomsday
Machines ", "Quick Looks" and Other Modest Proposals, 68 JUDICATURE 8, 20-21 (1984).

60. See infra text accompanying notes 27-29.



Environmental Negotiation

from delay may also benefit; but since agency action is very rarely
stayed pending appeal, the delay factor is often incidental.

In sum, I see limited opportunity for the courts of appeals to
push the negotiation of environmental disputes within the current
framework. If settlements, as we are told, are often triggered by
sniffs of what lies ahead in the judicial process-"quick looks"
and "doomsdays" 6 '-then the most propitious time to settle is
probably after, not before the appellate court has acted, i.e.,
before the proceeding picks up again on remand. The most we
can do before that-through settlement conferences-is to send
signals that the issues on appeal may or may not be successful, in
order to give impetus to the parties to settle their underlying
dispute.

III. JUDICIAL OVERSIGHT OF NEGOTIATED REGULATIONS AND

AGENCY SETrLEMENTS

Despite their circumscribed role in promoting ADR in environ-
mental controversies, appellate courts are likely to play a more
crucial role in reviewing negotiated rules and settlements. As I
mentioned earlier, negotiation typically does not eliminate court
involvement altogether; instead it changes the nature and scope
of the judicial role. Negotiated regulations, for example, are still
subject to judicial review, and even settlements of environmental
disputes that have been submitted to the courts for approval, en-
forcement or modification are not infrequently challenged on
appeal.

A. The Judicial Role in Negotiated Regulations

Perhaps the most novel and promising application of negotia-
tion is in the initial formulation of proposed regulations. In its
present form, the hybrid rulemaking process often exacerbates
the worst aspects of the adversary system. It invites parties to
take extreme stances in initial negotiations so that they will be
positioned to challenge the rule finally adopted. The agency is
often perceived as only going through the motions of considering
outsider comments, having already made up its mind when it pro-
posed the rule. Once a rule is adopted, if challenges arise, agency
programmers turn its defense over to their legal staff-and the
opportunity for genuine negotiation narrows. Against this back-

61. See generally Bedlin and Nejelski, supra note 59.

19851



18 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 10:1

drop, ADR proponents have suggested that direct negotiations
between the agency and the various affected parties take place
before a regulation is developed and formally proposed. Such ne-
gotiations would involve the affected parties in the regulatory
process from its onset. They would presumably soften the adver-
sary posture that animates the current comment process and re-
duce the inevitability of legal challenges to adopted rules. The
direct participation of a wide range of groups in negotiated
rulemaking would generate a sense of political legitimacy cur-
rently lacking in the talismanic invocations of agency expertise.

The idea of negotiated regulations has been endorsed by ACUS
and has been experimented with by EPA and other agencies. 62 A
protocol for negotiating regulations has been explored at length
by Phillip Harter in a recent article in the Georgetown Law Jour-
nal.63 Harter's ground rules would include publication of the
proposed negotiation in the Federal Register, participation by
those "materially affected" on submitting an application, publica-
tion of those selected for participation, an option to make the ne-
gotiations confidential, including a prohibition on the use of
offers or data in later court proceedings or in response to Free-
dom of Information Act requests, and a "consensus" on the rule
ultimately noticed for comment under regular APA procedures. 64

When so noticed, the statement of basis and purpose would not
have to be as comprehensive it would under ordinary rulemaking;
its main claim to legitimacy would rest on the consensus achieved
by the parties. 65

From the perspective of an appellate judge, the most interest-
ing aspect of Harter's protocol is his suggestion concerning the
"less stringent" scope of review of negotiated regulations. Har-
ter reasons that "if individuals can boycott the negotiation group
and then obtain judicial review under a stringent standard, the
regulatory negotiation process could unravel." 66 Moreover, he

62. See supra note 33; see also 49 Fed. Reg. 17576 (Apr. 24, 1984) (EPA announcement of
an Advisory Committee to negotiate issues leading to a rulemaking on nonconformance
penalties under Section 206(6)(1) (1982)).

63. Harter, Negotiating Regulations: A Cure for Malaise, 71 GEO. L. J. 1 (1982).
64. Id. at 42-110. More recently, Harter has suggested that in most cases it is probably

not worth the price in decreased public confidence to close the meetings of the negotiating
group. (Letter to Patricia M. Wald, October 17, 1984). (Available in ColumbiaJournal of
Environmental Law office).

65. Harter, supra note 63, at 92-97.
66. Id. at 102.



Environmental Negotiation

reasons, the negotiation process will itself alter the backdrop to
judicial review. 67 A need for confidentiality and the goal of iron-
ing out differences in the negotiation process, for example, will
produce a very different record from the kind we now see on ap-
peal. And even though a negotiated rule would be submitted for
general public comment after negotiations by the major, inter-
ested parties, the formal notice and comment process cannot be
expected to reflect the "give and take" of an initial testing ground
for regulations. Accordingly, Harter calls for a narrow scope of
judicial review for negotiated regulations, one tailored to the ne-
gotiating process itself. In essence, he posits that "[a negotiated]
rule should be sustained to the extent that it is within the agency's
jurisdiction and actually reflects a consensus among the inter-
ested parties." 68

Harter's proposal contemplates a bifurcated system of judicial
review. The appellate court would first have to determine "juris-
diction"-a multifaceted word-and then make some kind of em-
pirical "real world" determination of whether a consensus had in
fact been reached in negotiations among the interested parties. If
this threshold was passed, the rule would be upheld. If not, the
normal standard of review-the APA's arbitrary and capricious
standard or another standard provided by the statute authorizing
the agency to promulgate the rule at issue-would govern.

As a rookie/veteran (take your pick) of five and a half years of
sitting on these cases, I have some comments and concerns about
the proposed changes in the scope of review. Consider, for ex-
ample, just exactly how judges would go about determining
whether negotiated rules are within an agency's jurisdiction. So
far as I can tell, courts would still have to ensure that the rule is
within the agency's statutory perimeters and does not conflict
with specific congressional intent. I do not think Harter is sug-
gesting that the jurisdictional inquiry can be more summary than
it is now. An appeals court still has an independent obligation to

67. Id. at 103.
68. Id. at 103. Harter writes that his present view is that each interest that is participating

concurs in or at least does not oppose the result. (Letter, supra note 64). The Administra-
tive Conference Recommendation provides "Consensus in this context means that each
interest represented in the negotiating group concurs in the result, unless all members of
the group agree at the outset on another definition." Procedures for Negotiating Proposed
Regulations (Recommendation No. 82-4), 1 C.F.R. § 305.82-4 (1984).

1985]



COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 10:1

insure that the agency is not thwarting Congressional intent, re-
gardless of how many parties agree with the agency's rule.

The second leg of Harter's test-whether a true consensus ex-
isted-might be met simply by formal signatures to the negoti-
ated proposal. On the other hand, it could in some cases require
a close look at the process by which alleged consensus was
reached in those, perhaps rare, cases where a post-negotiation
dispute about "consensus" arose. This possibility, in turn, could
mean that a careful record must be kept of the negotiation and
that courts be allowed access to that record in cases of doubt.

It is vital as well to be clear on what standards an appellate
court would use to determine whether a consensus has been
reached. To those of us who are veterans of group politics the
word consensus has a subjective overtone, i.e., what will be ac-
cepted without contest. Unless it is objectively defined (e.g., a
vote of 100%, 51% or 2/3 of the participants; on all facets, on
most facets, etc.), however, a mere requirement of consensus
would invite different courts to evaluate similar negotiations quite
differently.

69

A problem might also arise with respect to whether the appro-
priate interests were represented in the consensus. This could
come up by way of someone trying to appeal the rule who had not
participated in the negotiations. Under prevailing law anyone
who can show an adverse interest and, for some agencies, who
participated in the notice and comment stage of rulemaking can
appeal. 70 But Harter seems to suggest that under the new process

69. See generally Harter, supra note 63, at 102-07. Harter points out that it is also impor-
tant that the proposal contemplates that someone will make a relatively extensive effort to
identify the interests and to contact suitable representatives with respect to the issues and
their participation. In that sense, the process has a potential for wider participation than
mere notice and comment. But in the case of consumers, for instance, it may be particu-
larly difficult to secure one or even a few representatives to express the diversity of views
and interests among that group. If that is the case, he says, negotiating regulations may be
inappropriate for such a rule. But he also suggests that no more than fifteen interest
groups should participate in a negotiation. In fact, one of his rules for a satisfactory pro-
cess is that "The interests significantly affected by the subject matter of the regulation
should be such that individuals can be selected who will represent those interests ade-
quately." Harter, The Political Legitimacy and Judicial Review of Consensual Rules, 32 AM. U. L.
REV. 471, 479 (1983).

70. Section 10(a) of the APA, 5 U.S.C. § 702(a), generally provides that any "person
aggrieved" by agency action has standing to seek judicial review of that action. Other
statutes govern standing to seek judicial review of that actions of particular agencies.
Under the Administrative Orders Review (Hobbs) Act, 28 U.S.C. §§ 2341-2352 (1982), for
example, appellants must have been actual parties to agency proceedings, including notice



Environmental Negotiation

a challenger might have to pass an additional test-i.e., show that
his "interest" had not been adequately represented in the negoti-
ations and, if not, that he had himself tried to participate when
the proposed negotiation was announced. I have trouble, how-
ever, seeing the justice or efficiency of a rule that requires every
party potentially affected by a negotiation to either demand per-
sonal representation or take his chances on whether the desig-
nated interest group representative will truly represent his
interests. This sort of standing rule might well encourage unnec-
essary challenges to the original committee if it became necessary
to preserve an opportunity for later, substantive appeals.

Furthermore, only a limited number of interest groups can take
part in the negotiation of any regulation. Although Harter points
to class action suits and unions as models of interest group repre-
sentation, 7' neither analogy provides total comfort. A union rep-
resentative is elected by its members and given statutory
authority to represent those members' interests; no such process
is at work in negotiated regulations. The class action analogy is
somewhat more relevant, since in Rule 23(b) actions all class
members are bound regardless of consent. 72 Yet the class repre-
sentative must be certified by a court, and the class itself is bound
only by a ruling, settlement, or dismissal that is judicially imposed
or approved.

In sum, I am somewhat wary of the intrusion of the "interest"
test into appellate standing. I worry about individuals being
bound by "interest" group surrogates so as to preclude their
challenging a rule on appeal when they can show they are ad-
versely affected by the regulation and they participated in the no-
tice and comment phase. At the same time, I do sympathize with
Harter's concern that parties whose interests are obviously in-
volved not sit out the negotiations and come forward only after
the rule has been hammered out by others. Part of my difficulty,
no doubt, is in conceptualizing an "interest" in this context. One
may be a consumer or a small businessman but not agree with the
compromise made by the consumer or small business advocate in

and comment proceedings, in order to have standing to challenge the actions of six agen-
cies including the Interstate Commerce Commission and the Nuclear Regulatory Commis-
sion. See 28 U.S.C. § 2344; Simmons v. ICC, 716 F.2d 40, 42-44 (D.C. Cir. 1983); Gage v.
AEC, 479 F.2d 1214, 1218 (D.C. Cir. 1973).

71. Harter, supra note 63, at 104 nn.558, 559 & 561.
72. FED. R. Civ. P. 23(b).

1985]



22 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 10:1

the negotiation. Even if an objection is made at the committee
formation stage, how does the agency decide how many consum-
ers or small businessmen to let in before negotiations even begin?
In this sense, negotiatied rulemaking is quite different from a
class action, where the complaint at least sets out the goal of the
plaintiffs' representatives. Will the announcement of a regulatory
committee tell "interest" group members anything about the sub-
stantive positions of their designated representatives? To over-
come this hurdle, agencies could insist that committee members
announce their goals and major positions in advance to discour-
age parties from claiming non-representation later. At a mini-
mum, then, Congressional action would be necessary to convert
our normal individual-oriented APA standing requirements into
"interest"-oriented criteria.

Furthermore, I am not sure it is wise for courts, in determining
standing under the proposed criteria, to get into questions of
whether the challenger's particular "interest" was in fact repre-
sented or represented adequately, or whether a major issue was
left out of the negotiations that concerned him. We now look
only to see if he submitted comments and has made a plausible
argument that he is adversely affected by the result.

Let me move now to what I consider to be the substantive core
of Harter's judicial review schemata: that consensus is a surrogate
for, or functional equivalent of the usual tests of reasonableness
or nonarbitrariness that are used to measure compliance with
APA standards ofjudicial review. Consensus among affected par-
ties strongly suggests reasonableness in most situations, particu-
larly if the court has already determined that there is no clear
congressional intent on the subject. But courts will have to look
beyond consensus, because it "ain't necessarily so" that consen-
sus will serve as a substitute for reasonableness in all cases. The
parties could have good reason to want a particular result even if
it made little or no sense under a given statute; they could, in
short, agree to a result that effectively took the place of a new
statute or an amendment to an old one, but their agreed-upon
rules might not be a reasonable product of the original statute.
The consensus could also be pure political logrolling-I give you
this and you give me that-rather than rational decisionmaking. I
may be speaking more conceptually than practically, but certainly
some inquiry into the reasonableness of a consensus rule will al-
ways be required, unless the APA is substantially changed.



Environmental Negotiation

And even if the negotiations over a proposed rule have reached
a "true" consensus, the agency must surely still provide a con-
vincing rationale and a sufficient factual basis for the final product
that emerged from the negotiation. There is, of course, some
precedent for reviewing negotiated actions differently. The re-
sults of collective bargaining conducted pursuant to the NLRA
enjoy immunity from judicial review; 73 judicial review of arbitra-
tion awards is also very limited. 74 judicial review of settlements in
litigation has a standard all its own; but in the absence of a statu-
tory charge, all agency rules-no matter how they are promul-
gated-are subject to the "arbitrary and capricious" standard.
While consensus certainly may be an important factor in pursuing
the meaning of that standard, it is doubtful that consensus can
displace the traditional judicial gloss which has accumulated over
the past forty years.

On balance, then, I am not at all sure that courts would become
substantially less involved in negotiated regulations under Har-
ter's judicial review standard. Inevitably, each phrase in the pro-
posed new standard ofjudicial review opens up its own Pandora's
box. This is not to deny the value of change, but it does indicate
that blending the old with the new will take time and thought.

Harter's proposal for limiting judicial review of negotiated reg-
ulations has its roots in a theory of administrative law that rejects
both the expertise 75 and non-delegation doctrines 76 as meaning-
ful ways to govern the relationship between agencies and courts.
This tradition views the regulatory process in essentially political
terms and grounds the legitimacy of agency action, indeed of all
government action, on the ability of the government to reconcile
conflicting political and practical interests as expressed by inter-
est group representatives. 77 Injecting the interest group conflict
into the agency's decision-making process promises to make the

73. National Labor Relations (Wagner) Act of 1935, 29 U.S.C. §§ 151-169 (1982); see,

e.g., NLRB v. Insurance Agents' Int'l Union, 361 U.S. 477, 499 (1960); NLRB v. American

Nat'l Ins. Co., 343 U.S. 395, 404 (1952).

74. See e.g., United Steelworkers v. Enterprise Wheel & Car Corp., 363 U.S. 593, 596-98
(1960); United Steelworkers v. American Mfg. Co., 363 U.S. 564, 568 (1960); International

Union of Office and Professional Employees, Local 2 v. Washington Metropolitan Area

Transit Auth., 724 F.2d 133, 137-39 (1983).
75. See, e.g., Cheuron, 104 S. Ct. at 2793.
76. See, e.g. Hampton v. United States, 276 U.S. 394, (1928); Estreicher, supra note 1, at

896-900.
77. See, e.g., R. DAHL, WHO GOVERNS (1961); D. TRUMAN, THE GOVERNMENTAL PROCESS

(1951).

1985]



COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 10:1

government more responsive to the public interest. If agencies
are required to take into account the views of all relevant
groups-so the theory goes-their decisions will more closely re-
semble those made by the people themselves. 78 In such a theory,
judicial oversight is both unnecessary and excessive.

Harter's proposal-and the more general attempt to apply
ADR to the entire regulatory process-represents the latest out-
growth of this pluralist approach to administrative law. By sug-
gesting that affected parties play a direct role in the development
of regulations from day one, it rejects the notion that regulatory
dilemmas can be solved through technological rationality or en-
lightened expertise. By emphasizing the ills of the adversary sys-
tem, this approach also rejects the hope that agencies can be
directed by legislative will as determined by the courts. Perhaps
most importantly, however, proponents of environmental ADR
see in negotiated rulemaking the political legitimacy currently
missing from the regulatory process. As Harter puts it:

The classic way of establishing public confidence. . . is to have
representatives of the people make the policy choices. Thus,
an alternative, more direct way to make these inherently polit-
ical decisions would be to adapt the legislative process itself to
the development of regulations. Such a process would enable
representatives of the competing interests, including the rele-
vant agency itself, to thrash out a consensus on the policy in-
stead of making a pitch to the umpire.79

Historically, however, the attempt to model the regulatory pro-
cess on the political process has been troubled by three thorny
problems. For any particular agency action, interest group plu-
ralists have to determine which interest groups should be repre-
sented before the agency, what kind of participation is
appropriate, and what sort of procedural rules will govern the
ironing out of a consensus among those conflicting group inter-
ests. 8 0 These are essentially the same problems that linger in
Harter's proposed standard for judicial review of negotiated reg-
ulations.si Under Harter's proposal, appellate courts might well

78. For description and criticism of this "pluralist" view, see Frug, Ideology of Bureaucracy

in American Law, 97 HARV. L. REV. 1276, 1355-61, 1368-73. (1984); Stewart, The Reformation

of American Administrative Law, 88 HARV. L. REV. 1667, 1713-22. (1975).
79. Harter, supra note 63, at 27-28.
80. See Frug, supra note 78, at 1368-73.
81. Harter writes that in some negotiations the participants have agreed that interests

would not formally ratify any agreement but rather would engage in an informal agree-



Environmental Negotiation

be called upon to resolve any or all of these questions in review-
ing a negotiated rulemaking: to develop legal standards for iden-
tifying relevant interest groups, and for assessing the extent of
their participation and the "process" of negotiation. This could
be a tall order for judges in many cases. Indeed, there is some
question in my mind how much less complex and adversarial judi-
cial review of the negotiating process would be than it is now
under the current APA mixture of expertise and non-delegation
theories.

B. The Judicial Role in Agency Settlements

The role of courts in reviewing settlements in environmental
litigation is similarly evolving and somewhat uncertain at the mo-
ment. The standards for reviewing settlements involving the EPA
are largely derived from the standards that courts have developed
for reviewing settlements involving the government more gener-
ally. The APA expressly recognizes the power of the court "to
supervise and consider offers of settlement. . .where the public
interest permits."8 2 Settlements in government cases most often
eventuate in consent decrees. Consent decrees, which must be
approved by the court, have the advantage of a status of final
judgment, enforceable by court applied sanctions, even con-
tempt. If the private party does not live up to its obligations, the
government need not institute a new proceeding to prove the
original violation. The decree is res judicata on the issue.83

Judges are not free to reject settlements or consent decrees at
will-they are presumed valid unless they contain provisions that
are "unreasonable, illegal, unconstitutional, or against public in-
terest." 84 Presumably, a court could deny approval if the settle-
ment looked like a real sellout on the part of the government, or

ment to support the consensus in all appropriate administrative fora and before a court in
a subsequent judicial review. Each party recognized that no other party could bind all its
members, be they corporate members of a trade association or individual members of a
labor union. That would permit anybody who did feel aggrieved to challenge the rule.
But the fact that others would be expected to intervene on behalf of the agency would
indicate that the challenger is not necessarily representing a widely held view and that the
decision-maker needs to focus narrowly on the issues raised, as opposed to painting with a
broad brush. "That may, in the end, be all we can ask for." (Letter, supra note 64).

82. Administrative Procedure Act, 5 U.S.C. § 554(c)(1) (1982).
83. United States v. City of Miami, Fla., 664 F.2d 435, 439-40 (5th Cir. 1981).
84. Id. at 440 (citing U.S. v. City of Alexandria, 614 F.2d 1358, 1362 (5th Cir. 1980)); see

also Donovan v. Dorfman, supra note 39.

1985]



26 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 10:1

perhaps even if the terms of a settlement for one violator were
remarkably disproportionate to the terms offered another of like

culpability. Ensuring that settlements do not violate the public in-
terest or a third party's interest is not difficult for courts.
Although civil enforcement suits brought by the government indi-
rectly involve the interests of third parties and the public, courts
usually presume that the government is a surrogate for public in-
terests and that a fairness hearing before the judge can ferret out
any special private ones.

The Supreme Court itself has said that a court's refusal to ap-
prove a consent decree containing permanent injunctive relief in
a Title VII case is an appealable order, in part because settle-
ments support one of the key policies supporting Title VII-vol-
untary compliance.8 5 At the same time, however, settlement
approval is not supposed to be a "perfunctory" matter. In Pro-
fessor Moore's words, "the court is not properly a recorder of
contracts, but is an organ of government constituted to make ju-
dicial decisions and when it has rendered a consent judgment it
has made an adjudication." 86 At least one court has pointed out
that a consent decree requires substantially more scrutiny than a
settlement of a monetary claim:

[T]he consent decree does not merely validate a compromise
but, by virtue of its injunctive provisions, reaches into the fu-
ture and has continuing effect. . . .Even when it affects only
the parties, the court should, therefore, examine it carefully to
ascertain not only that it is a fair settlement but also that it does
not put the court's sanction on and power behind a decree that
violates Constitution, statute, or jurisprudence. If the decree
also affects third parties, the court must be satisfied that the
effect on them is neither unreasonable nor proscribed. 87

Appellate courts, we are told, should overrule a trial court's ap-
proval of a settlement only for a clear abuse of discretion,8 8 espe-

85. Carson v. American Brands, 446 F. Supp. 780 (E.D. Va. 1977), aff'd, 606 F.2d 420
(4th Cir. 1979), rev'd, 450 U.S. 79, 86-88 (1980); see also Dorfman, supra note 39, slip op. at
2-10 (holding that a court's refusal to approve a consent decree containing injunctive relief
should be treated like a denial of a preliminary injunction and is therefore appealable
upon a showing of irreparable harm).

86. lB MOORE'S FEDERAL PRACTICE 0.409[5] at 1030 (2d Ed. 1980).
87. Williams v. City of New Orleans, 543 F. Supp. 662, 671 (E.D. La. 1982) (citing

Miami, 664 F.2d at 441).
88. Patterson v. Newspaper & Mail Deliverers Union, 514 F.2d 767, 771 (2d Cir. 1975),

cert. denied, 427 U.S. 911 (1976); In a comprehensive discussion of the judicial role in re-
viewing agency settlements, the Seventh Circuit has recently held that the abuse of discre-



Environmental Negotiation

cially where the trial judge held fairness hearings on it. And
appellate courts cannot ordinarily modify the terms of a consent
decree, as they may sometimes do with judicially imposed
decrees. 89

Settlements that come directly to the appellate court from an
agency, although in a somewhat different posture, receive basi-
cally the same deferential treatment. Such settlements most often
occur in ratemaking cases. In United Municipal Distributors Group v.
FERC,90 a settlement was negotiated between the pipeline, the
Federal Energy Regulatory Commission staff, state utility com-
missions and the pipeline's customers excepting the municipali-
ties. The Commission approved the settlement as to all parties
except the "munies." The Commission then set only the munies'
part of the rate package down for trial. The munies appealed,
arguing that a valid settlement which failed to include them could
not be implemented. We approved the Commission's action as
preserving a settlement for the vast majority of noncontesting
parties, allowing them to have "the benefit of a settlement deter-
mined by the Commission to be fair and reasonable and in the
public interest" while the contesting party has full due process
rights to a hearing.9' The court also found that the formal ap-
proval of a rate settlement was equivalent to the determination of
a just and reasonable rate, and was therefore reviewable. 92 In
other cases, we have upheld settlements that apply to only some
of the customers at issue in a rate dispute- even when the non-
settling customers are similarly situated and eventually receive
different rates. 93

To be sure, this judicial willingness-indeed eagerness-to ap-
prove agency settlements should encourage an increase in negoti-
ated dispute resolution at the agency level. And in my view, a
limited judicial role in this area rarely entails any abdication of
our duty to ensure that agency action promotes the public inter-
est as determined by Congress. Lately, however, courts are hear-

tion standard should also govern appellate review when a district judge disapproves a
consent decree. See Dorfman, supra note 39, slip op. at 12. The Dorfman court noted, how-

ever, that a district court must "justify any departure from . . . the principle . . . that

settlements are favored and ordinarily should be approved." Id.
89. See, e.g., Dunlop v. Pan Am. World Airways, 672 F.2d 1044 (2d Cir. 1982).
90. 732 F.2d 202 (D.C. Cir. 1984).
91. Municipal Distributors, 732 F.2d at 209.
92. Id. at 206.
93. See, e.g., City of Bethany v. FERC, 727 F.2d 1131, 1138-40 (D.C. Cir. 1984).

1985]



28 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 10:1

ing cases involving the settlement of citizen suits brought against
the agency, and these cases involve a very special set of settle-
ment-related problems.

Several of the environmental laws have citizen suit provisions
which encourage private groups to bring suit when the Adminis-
trator is not enforcing the law. 94 These suits were surely envi-
sioned as a supplementary means to enforce the law. Some
difficulties involved in settling private enforcement actions sur-
faced but were not entirely resolved, in a case decided over a year
ago in the District of Columbia Court of Appeals and recently de-
nied certiorari by the Supreme Court, Citizens for a Better Environ-
ment v. Gorsuch.9 5 Simply put, the issue is whether and to what
degree a consent decree can bind the agency to do something
other than that commanded by law, i.e., to administer the law in a
certain way that by implication means the agency cannot do it
some other permissible way.

In Better Environment, a consortium of environmental organiza-
tions (Natural Resources Defense Council, Environmental De-
fense Fund, Audubon Society, Citizens for a Better Environment)
sued the EPA for nonenforcement of the Clean Water Act. 96 The
parties negotiated an agreement over the objections of the chemi-
cal manufacturers who intervened. The district court held open
hearings on the proposed agreement and received comments
from interested parties including the industry intervenors. No

appeal was taken from the decree itself. The agreement required
the agency to undertake a detailed program for developing regu-
lations for the discharge of 65 toxic pollutants on an industry-by-
industry basis.9 7 The EPA also committed itself to a research pro-
gram based on specific criteria for determining which other pollu-
tants should be regulated. 98  The resulting regulations
themselves would be subjected to notice and comment-and no
substantive constraints were placed on them. The decree was
subsequently modified by the district court to extend deadlines

94. E.g., Federal Water Pollution Control Act (Clean Water Act), 33 U.S.C. § 1365
(1982); Clean Air Act, 42 U.S.C. § 7604 (1982); see generally, Note, Awards of Attorneys'Fees to

Unsuccessful Environmental Litigants, 96 HARV. L. REV. 677, 677 n.2 (collecting additional
statutory provisions).

95. 718 F.2d 1117 (D.C. Cir. 1983), cert. denied, -U.S.-, 104 S. Ct. 2668 (1984).

96. Federal Water Pollution Control Act (Clean Water Act), 33 U.S.C. § 1251 (1982).

97. Better Environment, 718 F.2d at 1120-21.
98. Id.



Environmental Negotiation

for those regulations.9 9 The issue of whether the original decree
constituted an impermissible usurpation of the Administrators'
discretion was raised by the court, sua sponte, when the modifica-
tions were appealed.' 0 0 The case was remanded for the district
court to decide that issue.

On remand, the chemical manufacturers argued that the de-
cree's criteria for deciding whether a pollutant would be regu-
lated, although not inconsistent with those laid down in the
statute, impermissably circumscribed the Administrators' discre-
tion. 1 1 Nonetheless, the district court held that the decree did
not illegitimately interfere with the Administrators' discretion be-
cause the decree was "process oriented" rather than outcome de-
terminative. 0 2 The court also emphasized that all interested
parties had an opportunity to comment on the decree, and that, in
the 1977 Clean Water Act Amendments, Congress had expressly
referred to the settlement and had acted in such a way as to im-
plement it.1°3 A split panel of our court upheld the decree, rely-
ing on EPA's own consent to the decree, the need for government
agencies to be free to settle cases, the underlying allegations of
unlawful action by the agency in not enforcing the law, and the
consistency of the decree with the statute. 0 4 Implicit in the opin-
ion, however, was a recognition that such settlements might not
be permissible if they tied the Administrator's hands too tightly
and of the potential danger of "sweetheart deals" between an
agency and a preselected "opponent."'' 0 5

In dissent, Judge Wilkey argued that the decree was an imper-
missible invasion of the agency's discretionary powers under the
statute. He saw no real difference between the procedural and
substantive requirements in the decree, concluding that the de-
cree bound the Administrator on important choices as to meth-
ods, priorities, and allocations of resources. 0 6 Permitting such

99. Natural Resources Defense Council v. Costle, 12 ENV'T REP. CAS. (BNA) 1833
(D.D.C. 1979).

100. Environmental Defense Fund v. Costle, 636 F.2d 1229, 1258-59 (D.C. Cir. 1980).
101. See Better Environment, 718 F.2d at 1121.
102. Natural Resources Defense Council v. Gorsuch, 16 ENV'T REP. CAS. (BNA) 2084,

2088 (D.D.C. 1982).
103. Id. at 2089 (citing 123 CONG. REC. S. 19647-48 (daily ed. Dec. 15, 1977)).
104. Better Environment, 718 F.2d at 1117.
105. See generally id. at 1127-1130, (concluding that the consent decree "does not in-

fringe on the Administrator's statutory discretion").

106. Id. at 1131-32.

19851



COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 10:1

settlements, he thought, would "weaken . . .democratic control
over agency policy [and strengthen the powers ofl those special
interests who are party to the decree."' 0 7 He also argued that
judicial approval of such decrees amounted to an abandonment
of the court's role in reviewing agency actions. 0 8

Although the Supreme Court denied certiorari in Better Environ-
ment, the debate is not likely to go away. In fact, another panel of
our court composed of Judges Wilkey, Ginsburg and McGowan,
in Womens Equity Action League v. Bell,' 0 9 recently remanded a dis-
trict court injunction based on Adams v. Richardson, 0 which was
originally brought to enforce various agency obligations under
the civil rights provisions in the higher education arena. The gov-
ernment now seeks to vacate the underlying consent decree on
the ground that the decree impermissibly intruded on the govern-
ment's statutory and constitutional authority to manage and su-
pervise the agency's enforcement of civil rights laws.'" The
government argued that the consent decree establishes the dis-
trict court as the "perpetual supervisor" of the appellants, alter-
ing the normal relations between agency and court.1 12 The
government looks for further support from the recent Supreme
Court opinion in Allen v. Wright, which emphasized that courts
have limited powers to hear "suits challenging not specifically
identifiable Government violations of law, but the particular pro-
grams agencies establish to carry out their legal obligations. Such
suits . . . are rarely, if ever appropriate for federal court
adjudication." 1 

3

Tough questions remain about the limits on negotiated settle-
ments between a government agency and private industry or even
quasi-private, public interest groups suing for enforcement. The
settlement effected in Better Environment seems appropriate; but
what if, for example, the EPA agreed to enforce Superfund by
cleaning up ten dumpsites a year?

And where is the logical stopping point in allowing agencies to
agree to complex enforcement decrees as long as they concern

107. Id. at 1136.

108. Id. at 1136-37.

109. 743 F.2d 42 (D.C. Cir. 1984).
110. 480 F.2d 1159 (D.C. Cir. 1973).
111. Women's Equity Action League, 743 F.2d at 42-43.

112. Id. at 43.

113. Allen v. Wright, - U.S. -, 104 S. Ct. 3315, 3329 (1984).



Environmental Negotiation

"procedure" rather than "substance"? Can an Administrator, for
example, commit him or herself-and more importantly succes-
sors- to administering the law in a particular way unless the suc-
cessor can convince a court to later modify the decree? Such a
negotiated commitment would bind successors more tightly than
would procedural rules, which can be changed at the agency's ini-
tiative after notice and comment. Judicial review of rule changes
in enforcement procedures, furthermore, is confined to arbitrary
and capricious oversight. But where enforcement procedures are
incorporated in a consent decree, changes can only result from a
modification hearing. In that context, the court-not the
agency-becomes the central decisionmaker. By what standards
then should the judge decide if modifications are appropriate?
Does deference have any role? At what point does the decree
runout? How does a court handle a settlement in which an
agency and the challenging party agree, but intervenors with im-
portant interests do not? What kind of hearings should the judge
hold on far-reaching decrees-should they be like mini-rulemak-
ing proceedings? Should he or she feel free to rewrite the de-
cree-subject to the parties consent or not-or to send them back
to the bargaining table with instructions on certain topics, or with
the requirement that other parties participate in the negotiations?

What kind of changed circumstances warrant modification-
would a change in administration and regulatory philosophy suf-
fice? Traditionally, the parties' expectations are given paramount
consideration in deciding upon a request for modification of a
settlement agreement, but one might look at consent decrees in
citizen suit enforcement actions a bit differently. If the case had
been litigated, won by the challengers, and mandatory and in-
junctive relief had been granted, how broad or limited would the
courts' powers be so as to provide relief but not unnecessarily
usurp the administrator's duties? Obviously there is no bright-
line answer. Comparing negotiation with litigation for this pur-
pose might, however, lead a judge to assess the original or contin-
ued validity of a settlement differently than if he or she compared
it with ordinary agency policy-making.

These problems may sound familiar. They are quite similar to
the debate that has been going on in the human services area for
the past decade. Have courts overreached themselves in laying
down structural decrees for prison and mental hospital adminis-
trators that are too detailed? Having found violations of constitu-

19851



32 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 10:1

tional rights, how far into the day-to-day workings of the asylum
can courts legitimately go to make their correction? Judges ap-
proving settlements in major environmental enforcement suits
will have to consider at what point they are intruding too deeply
into policy choices and questions reserved for the administrator
when they seek to provide some effective insurance that the law
will be implemented as Congress wished it to be. There are diffi-
cult line drawings ahead. Without subscribing to the view that
the court can go no further than tell the administrator to obey the
law in the future, it should be evident that courts are not naturally
equipped to monitor for years infinitely detailed programs or im-
plementation of a complex law in constantly changing circum-
stances. This is an area that I expect will require real sensitivity
on the part of judges and litigators if we are to preserve the effi-
cacy of citizens' suits and settlements.

Another nettlesome issue is when the judge should approve a
settlement when the challenging party and the agency agree, but
intervenors with important interests at stake, e.g., the industrial
intervenors in Better Environment do not. Almost all environmen-
tal disputes are tri-cornered. How does a court handle a settle-
ment in which only two out of three of the parties agree? The
intervenors can, of course, present their concerns to the judge;
but that puts her in the position of deciding whether they are vital
enough to support a finding of no public interest. When a third
party attacks a settlement in a fairness hearing, what are the ex-
tent of that party's rights, for example, to present evidence or to
take discovery?

Finally, one may question whether settlements should reflect or
be consistent with established policies of the agency, or if they
can depart markedly from those policies. Could the EPA, for the
first time in a settlement (rather than in rule-making or adjudica-
tion) implement a new definition or application of the "bubble"
concept that represented a departure from existing policy or
practice? Because settlements usually involve no articulated ex-
planation, would they therefore constitute violations of the judi-
cial rule that agency departures from existing policies must be
satisfactorily explained to a court?

These are open questions. What they suggest is that the judi-
cial role in negotiating settlements will not and should not be pas-
sive in important, complex cases with significant policy
implications for implementation of environmental laws. The



Environmental Negotiation

partnership will continue. But the partners will play different, in
some ways even more taxing roles than Judge Leventhal envi-
sioned ten years ago.

In closing, I want to note a fact of which ADR proponents are
well aware: that the push for settlements is not a universally ap-
plauded one. In a provocative article, Owen Fiss argues that set-
tlements are not generally preferable to full-scale litigation, and
should not be institutionalized on any wholesale or indiscriminate
basis. 14 Analogizing the settlement to plea bargaining, he finds
it a "highly problematical technique for streamlining dockets"
and a capitulation of the ideal ofjustice to the demands of a mass
society 115 Fiss' objections to settlements are threefold: consent is
too often coerced by imbalances in the bargaining strength of the
parties; the bargain may be struck by someone without the au-
thority of a judge; and the absence of any trial or judgment ren-
ders subsequent judicial involvement troublesome. 116  Fiss
worries about limits on the power ofjudges, as well as their desire
to enforce complex, controversial and expensive decrees when
the going gets rough; and he cites some experiences to prove his
point.' 17 Fiss places great emphasis on the courts as "value artic-
ulators." On a philosophical plane, he believes that settlements
reduce the social function of a lawsuit to one of resolving private
disputes, rather than one of declaring public values."I8

I am not against settlements. They have always been and will
continue to be the only way the system can work, particularly in
making room for those cases which must be litigated to the end.
But the amputation of meaningful judicial review from settle-
ments or negotiated regulations in the environmental field would
make these ADR techniques far less attractive to some of the par-
ties as instruments of justice. And if settlements are to embody
the nearest approximation of the ideals of justice to which Fiss
gives expression (and which he assumes emerge from litigation,
though I am not so sure), we must exercise restraint and fore-
thought about what they contain and how we want judges to treat
them.

114. Fiss, Against Settlement, 93 YALE L.J. at 1073 (1984).

115. Id. at 1075.

116. Id. at 1076-85.

117. Id. at 1084.

118. Id. at 1085.

1985]