Module 3
Review of Agency Decisions and Chevron
A. Reasoned Decisionmaking
As we have seen in Chapter 4’s examination of the arbitrary, capricious standard,
the ideal in administrative law is for agency policymaking to be the result of reasoned
decisionmaking, understood as application of agency expertise to the factors made
relevant by governing law, including statutes and regulations. The normal push and pull
of the political process, changing circumstances, and the differing perspectives of various
actors including Congress and the President often divert agency policymaking from the
ideal. What follows elaborates and expands on that ideal and the issues that have arisen
surrounding it.
Agency policymaking should be the product of agency expertise as applied to
governing statutes. The primary matters that agencies are permitted to take into account
are derived from applicable statutes including the agency’s organic statute, and other
generally applicable statutes such as NEPA. Agencies are also generally permitted to take
into account the views of the President and the administration, but only insofar as those
views are consistent with governing statutes. Under normal principles of judicial review,
agencies are required to consider the factors made relevant by statute and should not
consider extraneous matters not statutorily relevant. Recent developments have
reinforced the prohibition against agencies taking into account factors that, while they
may be reasonable, are not contemplated by governing statutes.
The Supreme Court’s decision in Massachusetts v. EPA, 549 U.S. 497 (2007),
supports the notion that agencies should make policy by applying their expertise to the
factors made statutorily relevant. The Court rejected agency consideration of factors that
had long been thought appropriate for agencies to take into account, such as scientific
uncertainty and the President’s views on the best way to proceed in an area with
international implications. That decision reviewed the EPA’s decision not to engage in
rulemaking, not the substance of a final rule, but it still appears to have implications for
judicial review of agency policymaking generally.
Courts have, in some circumstances,imposed obligations of clarity and
consistency upon agencies. Briefly stated, agencies are sometimes required to operate
under clearly stated substantive criteria, and they are sometimes held to a relatively
strong obligation to treat like cases alike unless they explicitly disavow the substantive
rule governing prior decisions. These requirements are closely related to Administrative
ProcedureAct (APA) standards of judicial review and should be understood in
conjunction with them. In short, in some cases agency action that appears to be
inconsistent with prior decisions or that does not proceed from a discernible standard has
been held to violate APA §706’s arbitrary, capricious standard of review.
Courts have held that certain agency action may be taken only pursuant to clear
criteria. This is termed the ‘‘clarity’’ requirement. The clarity cases are in tension with
decisions allowing agencies a great deal of discretion under the APA over whether to
promulgate rules or decide issues on a case-by-case basis. Three reasons support the
clarity requirements: notice to the party of the standards for government action,
prevention of arbitrary agency action, and facilitation of review of agency action (both at
higher agency levels and in the courts).
The D.C. Circuit agreed with the manufacturers, holding that ‘‘the APA requires
the agency to explain why it rejects their proposed health claims— to do so adequately
necessarily implies giving some definitional content to the phrase ‘significant scientific
agreement.’ We think this proposition is squarely rooted in the prohibition under the APA
that an agency not engage in arbitrary and capricious action.’’ Pearson, 164 F.3d at 661.
This seems to be a relatively stringent clarity requirement, and it is not clear that this
approach would be followed by courts with a more deferential attitude toward agency
action than that often displayed by the D.C. Circuit. ‘‘Significant scientific agreement’’
on its own is not so hopelessly vague that a court cannot intelligently evaluate the
agency’s reasons for granting or denying permission. It is not a precise standard, but it
has sufficient content to ensure that the agency has applied its expertise, considered the
relevant factors, and taken a ‘‘hard look’’ at the issues involved. Most courts, including
the Supreme Court, do not require more.
The clarity requirement has arisen mainly in cases challenging state agency action
as inconsistent with federal due process requirements. In light of more recent
developments in due process norms, including the Roth line of cases defining property
interests and the development of the Mathews v. Eldridge balancing test (see Chapter 8),
it is unclear whether courts would adhere to these decisions today. It is also unclear
whether the APA imposes a similar, non-constitutional requirement on federal agencies.
The following cases are examples of situations in which clarity requirements have been
invoked.
Related to the clarity requirement is the well-established Chenery rule that agency
decisions are evaluated, on judicial review, based on the reasons given by the agency at
the time the decision was made. This is different from the rule that normally prevails on
appeal of a judgment by a court where a judgment may be affirmed on any basis, not
merely the lower court’s reasoning in support of its decision. Under the Chenery rule, an
agency may not, on judicial review, defend its decision on a basis that was not relied
upon by the agency at the time it made the decision. However, if the court remands a
matter to the agency because its explanation is inadequate, the agency may adhere to its
original decision if it constructs an acceptable alternative explanation. The Chenery case
is the most famous example of the operation of this rule. The Chenery family managed
and owned shares in the Federal Water Service Corporation.
The corporation was in a reorganization under which holders of preferred stock
would ultimately control the corporation. During the reorganization process, the
Chenerys purchased a large block of preferred stock on the open market. The Securities
and Exchange Commission (SEC) conditioned its approval of the reorganization on the
Chenerys selling their stock back to the corporation. In support of this decision, the SEC
relied solely upon traditional equity rules governing fiduciary relationships for its finding
that the Chenerys’ purchase of the preferred stock while the reorganization was pending
was improper. On judicial review, the SEC defended its action as based not only on
traditional equity principles but also with arguments based on the policies underlying the
securities laws and its expertise in applying those laws.
It is often stated that agencies are required to be consistent, which means that they
must treat like cases alike. As a corollary to this, agencies must offer an explanation
when they treat apparently like cases differently. However, agencies are free to change
their policies so that later cases are treated differently from earlier ones, but they must
explain any changed policy, and the new policy must, of course, be consistent with
governing statutes and acceptable under the applicable standard of judicial review. This
purported obligation to treat like cases alike is in tension with the general rejection of
discriminatory enforcement claims.
In discriminatory enforcement claims, which are discussed in more detail in
Chapter 9, the subject of an agency enforcement action claims that others, usually
competitors, are violating the same provision and that the agency should not be able to
enforce an order against it unless and until it also issues an order against the other
violators. These claims often involve competitors because the subject of the enforcement
action claims that without enforcement against competitors, they will be disadvantaged in
competition if they have to obey an order while their competitors do not. The Supreme
Court has not been sympathetic to claims of discriminatory enforcement, holding that
agencies have great discretion to choose their subjects of enforcement unless there is a
‘‘patent abuse of discretion.’’ This reasoning undercuts the argument that there is an
enforceable general obligation to treat like cases alike.
Agencies are normally required to follow their own rules, whether those rules
have been adopted in a rulemaking proceeding or announced in the course of agency
adjudication. If an agency wishes to change a rule, it must do so in a procedurally valid
fashion and must explain the basis for the change. It cannot simply ignore the rule when
the rule does not suit the agency’s purposes, even if it would be reasonable for the agency
to change the rule. If an agency adopts a rule in a rulemaking or a formal adjudication,
the agency may change that rule in a subsequent proceeding — a rule adopted in
rulemaking may only be revised in a rulemaking proceeding, while a rule adopted in
adjudication may be altered either in a subsequent adjudication or in a rulemaking
proceeding. The new rule will be reviewed to determine whether it is within the agency’s
statutory authority and whether the record provides adequate support for it. In a change
situation, judicial review is somewhat heightened because in addition to the usual factors,
the court will ask whether the agency has adequately justified the change from the prior
rule.
Related to the question whether agencies are always bound to follow their own
procedural rules is the issue of estoppel of agencies by the conduct or statements of
agency employees. ‘‘Estoppel’’ is a situation in which a party cannot assert a legal
argument, usually a defense, because of conduct that would make assertion of the defense
inequitable. Estoppel against the government arises when an agency official provides
erroneous advice to a member of the public that leads the member of the public to follow
an incorrect procedure or take actions that lead to substantively bad consequences. For
example, a claimant is instructed to use Form A to apply for a benefit when the correct
form is Form B. When the error is discovered, the claimant may argue that because the
agency told the claimant to file Form A, the government must provide benefits dating
back to when that application was filed even though the law states that benefits are paid
only beginning from the date that Form B is filed. The usual rule is that agencies are
normally not estopped by the conduct or statements of agency officials. Erroneous advice
given by an agency official (for example, by misstating the eligibility requirements for a
government program) does not estop an agency from relying upon the program’s actual
requirements and denying claims even if the claimant relied upon erroneous advice.
B. Cost-Benefit Analysis
One method for improving and evaluating the rationality of agency
decisionmaking is cost-benefit analysis. Under cost-benefit analysis, a policy is evaluated
by comparing the policy’s costs with its benefits. It has been argued that agencies should
be required to conduct cost-benefit analyses regarding their major policy decisions and
should not adopt a policy unless the benefits outweigh the costs. However, while it is
useful in some contexts, cost-benefit analysis has serious limitations that prevent it from
providing an overarching standard against which to judge every exercise of agency
discretion.
The most prominent appearance of cost-benefit analysis in administrative law is
in the process of centralized review of agency regulations that has existed since President
Reagan put it in place in 1981. As discussed above in Chapter 2, in Executive Order
12,291, President Reagan ordered all executive branch agencies to submit their proposed
rules to the Office of Management and Budget (OMB) for review. He also ordered
agencies, ‘‘to the extent permitted by law’’ not to take any regulatory action ‘‘unless the
potential benefits to society for the regulation outweigh the potential costs to society.’’
The order also required agencies to prepare a Regulatory Impact Analysis including cost-
benefit analysis on all major rules, and, if permitted by law, to include cost-benefit
analysis in its decisionmaking process. The Executive Orders have given the OMB the
authority to review the cost-benefit analysis and require further analysis before the
agency is permitted to finalize its rulemaking proposal. To some, this gave OMB an
inordinate amount of power over the regulatory process, in contravention of statutory
delegations of power to particular agencies.
Agencies often include cost-benefit analysis in their decisionmaking process, but
they may do so only if the statutes governing agency action permit it. Under the arbitrary
and capricious standard of judicial review, agencies must consider those factors, and only
those factors, made relevant by statute. If an agency’s organic statute requires an agency
to perform cost-benefit analysis, it must do so. If the governing statute prohibits the
agency from considering the costs of its actions, it may not do so. For example, the
Supreme Court in the American Trucking case approved of the D.C. Circuit’s
longstanding view that the EPA may not consider costs when establishing National
Ambient Air Quality Standards (NAAQS). Whitman v. American Trucking Associations,
521 U.S. 457 (2001). This makes sense because NAAQS are like a definition of clean air.
Costs are more likely to be relevant at the implementation stage.
It is impossible in this book to analyze comprehensively the pros and cons of the
use of cost-benefit analysis in agency decisionmaking. In most situations, it makes
perfect sense to consider costs and to not promulgate regulations when the costs outweigh
the benefits. There are, however, reasons to be cautious about cost-benefit analysis when,
for example, the costs or benefits of regulation are difficult to quantify or if there is
concern that the analysis is likely to be biased for political or other reasons. What follows
is a brief discussion of the promise and the pitfalls of costbenefit analysis. One strong
point in favor of agency use of cost-benefit analysis is that it would be a vast
improvement over the vague and standardless decisionmaking that seems to prevail at
some agencies.
Cost-benefit analysis would force agencies to consider the consequences of their
policies in a concrete, rigorous, and material way. Agencies would have to quantify the
costs and benefits as much as possible, and compare those costs and benefits to the costs
and benefits of potential alternatives. Costbenefit analysis would at least appear to be
more systematic than current practice at some agencies and it would provide a clear basis
for comparison with other potential policies. Further, a cost-benefit analysis requirement
would limit the potential for arbitrary agency action by specifying, in advance, a set of
relevant considerations. Finally, cost-benefit analysis may reveal (in some cases) that
when all the costs and benefits are taken into account, the proposed regulatory action will
actually cause a decrease in social welfare
C. Impact Statements
A common method for controlling or channeling the exercise of administrative
discretion is to require the agency to prepare an impact statement that contains a detailed
discussion of the likely effects of the proposed regulation, either comprehensively or
focused on a particular type of impact. Impact statements tend to be used to force
agencies to consider issues that are outside the scope of their mission. For example, the
most commonly known impact statement is the Environmental Impact Statement (EIS),
which is required for federal actions by the National Environmental Policy Act (NEPA).
The EIS forces agencies in diverse areas such as commerce, finance, and land
management to consider environmental effects that their own organic statutes might not
include. Impact statements improve decisionmaking by forcing agencies to focus on the
effects of their actions and by providing an opportunity for public scrutiny of the
agency’s plans.
The use of impact statements has increased in recent decades in line with greater
appreciation and awareness of the effects of agency action on issues outside the agencies’
core missions. Courts on judicial review have enforced impact statement requirements
procedurally, by requiring that the impact statement include the required information and
that it is included in the record of agency action. As we shall see, reviewing courts have
also required agencies to consider the issues addressed in the impact statement but they
have generally not required agencies to take any particular substantive action based on
the contents of an impact statement. In other words, if an agency prepares a proper EIS, it
can go ahead with its proposed action even at great environmental costs, as long as the
record shows that the agency was aware of and considered those costs.
As noted above and in Chapter 2, President Reagan instituted, in E.O. 12,291, a
requirement that agencies prepare a Regulatory Impact Analysis of their proposed rules,
which would include a cost-benefit analysis of the proposal. This Executive Order and its
successor, E.O. 12,866, have required agencies, as part of all major regulatory initiatives,
to prepare detailed regulatory impact statements that comprehensively address the
impacts of the proposed regulatory initiative. This and other required or proposed impact
statements address economic effects, social effects, effects on other government
programs or institutions, and/or environmental effects. While impact statement
requirements and proposals have merit, they also have their problems. Requiring agencies
to prepare detailed, comprehensive regulatory impact statements on every major initiative
adds significantly to the substantial legal, political, and technical burdens agencies
already face in formulating and implementing policy. A glance at the Federal Register
reveals that the process of proposing a regulation or taking many other important actions
has become very complicated and expensive, adding to the burden under which agencies
operate and contributing to regulatory delay.
NEPA does not explicitly state that the agency must actually consider the EIS
once it is prepared and placed in the record, although the Supreme Court has found in the
statute an implicit obligation to consider the environmental effects detailed in the
statement. See Strycker’s Bay Neighborhood Council v. Karlen, 444 U.S. 223, 227
(1980). NEPA certainly does not state that an agency must abandon a proposal if the EIS
reveals that the environmental costs are very high. The Court has stated with regard to
NEPA that ‘‘the only role for a court is to insure that the agency has considered the
environmental consequences.’’ Strycker’s Bay Neighborhood Council v. Karlen, 444
U.S. 223, 227 (1980).
D. Jurisdiction
Federal court jurisdiction over any claim requires a statutory grant. The federal
courts generally have jurisdiction over petitions for review of federal agency action under
two statutory sources. First, some agency enabling acts often grant a right of judicial
review and explicitly create federal court jurisdiction over petitions for review. Most such
statutes grant jurisdiction over petitions for review to the U.S. Courts of Appeals. See,
e.g., The Federal Trade Commission Act of 1914, 15 U.S.C. §45(c). A few grant
jurisdiction to the district court. See, e.g., 42 U.S.C. §405(g), granting jurisdiction over
review of social security determinations to the U.S. District Courts. Second, the general
grant of federal question jurisdiction, 28 U.S.C. §1331, grants jurisdiction over petitions
for review to the federal district courts. Claims for review of agency action arise under
federal law, easily meeting the test for jurisdiction under §1331.
In most cases the choice between the court of appeals and the district court is
simple because either the agency’s governing statute identifies the proper forum for
judicial review or, in the absence of a statute stating otherwise, review is available in the
district court. Most statutes that specify the forum provide for review of administrative
action in the court of appeals. The most notable exception is the Social Security Act, 42
U.S.C. §405(g), which provides for district court review of orders denying or terminating
benefits even though the agency has conducted a formal adjudication. In the absence of
statutory authority for choosing the court of appeals, the federal question jurisdiction
provision of 28 U.S.C. §1331 directs review to the district court.
The APA’s judicial review provisions, together with a tradition based on the ideal
of the rule of law, create a presumption in favor of reviewability of agency action. This
presumption is consistent with the judiciary’s role in protecting individuals from arbitrary
exercises of government power and ensuring that government acts only when it has legal
authority. Judicial review is thought to be necessary to keep agencies within the bounds
established by Congress. The presumption in favor of judicial review of agency action
did not exist before the APA was enacted, and many agency decisions were not subject to
judicial review. Common law writs, such as mandamus, were available to challenge only
nondiscretionary (or ministerial) agency action. Any hint of executive discretion would
sometimes lead federal courts to deny reviewability. For example, in Decatur v. Paulding,
39 U.S. (14 Pet.) 497 (1840), the Court refused to review a decision of the Secretary of
the Navy to deny Mrs. Decatur a widow’s pension.
The Secretary denied the pension because on the same day that Congress passed
the general pension law, it also passed a private bill granting Mrs. Decatur a pension. In
the Secretary’s judgment, Mrs. Decatur was entitled to one pension or the other, but not
both. The Court denied review even though the determination seemed to be a matter of
statutory interpretation, i.e., did Congress intend for Mrs. Decatur to have both pensions.
The Court denied review, using very deferential language: ‘‘The Court could not
entertain an appeal from the decision of one of the Secretaries, nor revise his judgment in
any case where the law authorized him to exercise discretion, or judgment. Nor can it by
mandamus, act directly upon the officer, and guide and control his judgment or discretion
in the matters committed to his care, in the ordinary discharge of his official duties.’’ 39
U.S. at 515. The Court also stated that ‘‘[t]he interference of the courts with the
performance of the ordinary duties of the executive departments of the government would
be productive of nothing but mischief, and we are quite satisfied that such a power was
never intended to be given to them.’’ Id. at 516.
E. . Reviewability
APA §704 provides for judicial review of ‘‘[a]gency action made reviewable by
statute and final agency action for which there is no adequate remedy in a court.’’ This
statute creates a strong presumption of reviewability of final agency action. In essence it
provides a cause of action for judicial review of final agency action when no other statute
provides for review. The key provisions of §704 may be understood as follows: The
‘‘agency action made reviewable by statute’’ provision is redundant. It merely states that
if a statute other than the APA, such as an agency enabling act, provides for judicial
review of a particular agency action, the action is reviewable under that provision. The
only real substantive effect of this provision is to foreclose any argument that upon its
passage the APA became the exclusive basis for judicial review. It also suggests that,
absent contrary statutory provisions, the requirements of Chapter 7 of the APA, including
the standards of judicial review, govern the remaining issues in judicial review even
when another statute provides the basis for judicial review. APA standards of review are
discussed in Chapter 4. The ‘‘final agency action for which there is no adequate remedy
in a court’’ provision creates a general entitlement to judicial review of final agency
action. The standards for determining when an agency action is final are discussed below
in the finality and ripeness section.
The APA’s references to ‘‘agency action’’ mean that review under the APA is
available only for ‘‘agency action.’’ The judicial review chapter of the APA contains its
own definition of ‘‘agency,’’ see APA §701(b)(2), but that definition is almost identical
to the one contained in the definitions section of the APA, §551(1). Agency means ‘‘each
authority of the government of the United States,’’ except, inter alia, Congress, civil and
military courts, and the governments of territories and possessions of the United States.
Despite the absence of an explicit exemption, the Supreme Court has decided that the
President is not an agency within the meaning of the APA. Franklin v. Massachusetts,
505 U.S. 788 (1992). This means that personal action by the President is not subject to
judicial review under the APA.
The APA explicitly preserves review under the pre-APA procedures such as
petitions for mandamus, general federal question equity actions and actions for
declaratory relief under the federal Declaratory Judgment Act. This non-APA category of
review has been denominated ‘‘nonstatutory review,’’ under which agency action that is
covered by neither a specific review provision nor the APA is reviewed. The term ‘‘non-
statutory review’’ is a misnomer, since these forms of non-statutory review depend at
least to some extent on various statutes including APA §703, which provides that if the
action for judicial review under the APA is inadequate or unavailable, the challenger may
employ ‘‘any applicable form of legal action, including actions for declaratory judgments
or writs of prohibitory or mandatory injunction or habeas corpus, in a court of competent
jurisdiction.’’ APA §559 also explicitly preserves ‘‘additional requirements imposed by
statute or otherwise recognized by law,’’ which has been interpreted to mean that
Congress did not intend for the APA to displace the federal courts’ traditional common
law powers in administrative law. See Kenneth Culp Davis, Administrative Common
Law and the Vermont Yankee Opinion, 1980 Utah L. Rev. 3, 10.
APA §701(a)(1) provides that judicial review is not available when ‘‘statutes
preclude judicial review.’’ Statutory preclusion often involves agencies with their own
elaborate internal review mechanisms. Normally, for a statute to preclude judicial review,
it should explicitly mention judicial review. The best example of such a statute was old
§211(a) of the Veterans’ Administration Act, which provided that the ‘‘decisions of the
[Veterans’ Administration] on any question of law or fact under any law administered by
the Veterans’ Administration [VA] . . . shall be final and conclusive and no other official
or any court of the United States shall have power or jurisdiction to review any such
decision[.]’’ A statute like this precludes review of the administrative action specified —
decisions of law and fact by the agency under any law it administers.
APA §701(a)(2) bars judicial review of agency action ‘‘committed to agency
discretion by law.’’ That provision, a descendant of the now-abandoned notion that
discretionary administrative action was never reviewable, can be understood in three
different ways — all of which reinforce the idea that certain decisions have been left to
agency discretion and are free from judicial review. These are the ‘‘no law to apply’’
approach, the ‘‘deeming clause’’ approach, and the ‘‘traditionally unreviewable’’
approach. In the Overton Park decision, the Court explained that agency action is
‘‘committed to agency discretion by law’’ when the governing ‘‘statutes are drawn in
such broad terms that in a given case there is no law to apply.’’ See Citizens to Preserve
Overton Park v. Volpe, 401 U.S. at 410, quoting S. Rep. No. 752, 79th Cong., 1st Sess.,
26 (1945). Judicial review is not possible in such cases because there is no discernible
statutory standard against which to judge the legality of agency action. Because Congress
normally attempts to give agencies statutory guidance, this exception to reviewability of
final agency action is rarely met and even in those cases in which the exception applies,
there were special reasons in addition to the vagueness of the statutory standard to think
that precluding review is appropriate. These circumstances are explained below.
F. Standing to Seek Judicial Review
A party seeking judicial review must have standing to sue. In general, as
government regulation became more pervasive during the twentieth century, the Supreme
Court has expanded the category of parties with standing to seek review of government
action. Standing problems arise most often in actions seeking injunctive (or similar) relief
regarding an agency’s treatment of a party other than the one seeking review. It may be
unclear whether the agency’s action has injured the party seeking judicial review and
whether holding the agency action unlawful (and setting it aside) will alter the plaintiff’s
situation.
Early standing doctrine generally denied standing to third parties, prohibiting
anyone from challenging the regulatory treatment of someone else. Standing was
recognized only for parties whose own legal rights had been allegedly violated by agency
action. Parties injured by agency treatment of others lacked standing to seek judicial
review. Cases involving regulation of competitors best illustrate the limits on standing
implicit in the legal right test: A business entity’s claim that regulation of a competitor
(or lack of a legally required regulation) gives the competitor an unfair advantage would
be rejected on the ground that the plaintiff’s own legal rights were not at stake. For
example, in Alexander Sprunt & Son, Inc. v. United States, 281 U.S. 249 (1930), the
Court rejected a claim that the Interstate Commerce Commission (ICC) had set railroad
shipping rates for Sprunt’s competitors too low, thereby injuring Sprunt’s business.
APA §702 states, in part, that ‘‘[a] person suffering legal wrong because of
agency action, or adversely affected or aggrieved by agency action within the meaning of
a relevant statute, is entitled to judicial review thereof.’’ The first clause of APA §702 is
the legal right test. Thus, the APA grants standing to anyone whose legal rights are
allegedly violated by agency action. The second clause of §702 appears to liberalize
standing beyond the legal right test to all those injured by the agency action, but is
limited by the phrase ‘‘within the meaning of a relevant statute.’’ Because this language
is somewhat obscure, it is unclear how much APA §702 expands upon the legal right test.
Article III of the Constitution grants the federal courts jurisdiction over various
classes of ‘‘cases’’ and ‘‘controversies.’’ This grant can also be understood as a
limitation— the jurisdiction of the federal courts extends only to genuine cases or
controversies. The federal courts may not render advisory opinions or adjudicate cases
without a real dispute in which a party may benefit from a judicial remedy. This
limitation of jurisdiction to cases and controversies has given rise to the basic
constitutional requirements for standing— namely that the plaintiff has suffered an
injury-in-fact that is fairly traceable to the challenged conduct and redressable by a
favorable judgment. These requirements ensure that the plaintiff has a real legal
controversy with the defendant, not just an abstract disagreement. While the
constitutional requirements for standing are relatively simple and straightforward, they
have been applied so inconsistently over the years that it is very difficult to generalize
from the decisions. The most one can hope for is to understand the arguments in the cases
and try to make sense of their reasoning.
G. The Timing of Judicial Review: Finality, Ripeness, Exhaustion of Administrative
Remedies, and Mootness
The ripeness and mootness doctrines are related to the constitutional requirements
for a case or controversy, namely that the plaintiff is injured by the challenged conduct
and the injury will be remedied by a favorable judgment. When a case is brought too
early or too late, either there is not yet an injury sufficient for standing or the injury has
ended, and a favorable judgment will not remedy it. For example, if a person brings a
petition for judicial review of a regulation before the agency has attempted to enforce the
regulation against the petitioner, the action may not be ripe because the petitioner has not
yet been injured by the regulation. If the agency repeals the challenged regulation before
or during judicial review, or if the challenger is no longer subject to the regulation, the
case may be moot because the petitioner is no longer injured. However, the ripeness and
mootness doctrines are more flexible than basic standing requirements and thus may not,
in all circumstances, be constitutionally based.
APA §704 grants a right to judicial review of ‘‘final agency action for which there
is no adequate remedy in a court.’’ The ‘‘final agency action’’ provision is essentially a
ripeness requirement, which excludes from review any agency action that is not yet final.
Courts also apply ripeness requirements in addition to the APA’s finality requirement,
although, as discussed below, there is a question whether this is proper, or whether courts
should find an action ripe for review whenever the action meets statutory (including
APA) standards of finality.
The finality and ripeness issues are a bit more complicated when it comes to
rulemaking. Technically speaking, a rule is final once the rulemaking process is
completed and the agency issues the final rule by publishing it in the Federal Register.
However, the Supreme Court has held that rules are not necessarily ripe for review when
issued. Rather, some rules are ripe upon promulgation while others may not be reviewed
until they are enforced, with judicial review essentially providing a defense to an
enforcement action brought by the agency.
Sometimes it is difficult to discern whether informal agency action is final and
reviewable. If further proceedings are contemplated to determine whether a violation has
actually occurred, the issuance of a citation alleging a regulatory violation is not final. If,
however, the agency makes a determination that a regulated party must pay a fine or
change its conduct, no matter how informally that determination is made, if no further
agency proceedings are contemplated, the determination is final and likely to be ripe for
judicial review.
One of the oldest, most established doctrines in administrative law is that
challengers must exhaust remedies within the agency before seeking judicial review.
Courts have applied this doctrine most strongly in cases of agency adjudication where
there are normally one or two appeals available within the agency. However, in APA
cases, there is no general exhaustion requirement beyond APA §704’s finality
requirement. Only when a statute or rule requires exhaustion does the requirement apply
in a case subject to the APA. Darby v. Cisneros, 509 U.S. 137 (1993).
A case is moot if there is no longer a live controversy between the parties. If a
party is no longer subject to an agency rule, or if the agency repeals the rule, a claim for
judicial review of the rule may be moot. For example, if a prisoner seeks injunctive relief
to improve prison conditions and is released from prison, the prisoner’s claim is highly
likely to be moot. The released prisoner may still have a damages claim for past
treatment, but no longer has a live controversy with the prison over treatment in the
future. Mootness has sometimes been characterized as standing over time— a plaintiff in
essence loses her standing when she is no longer subject to the challenged agency
conduct.
H. Choosing Standards of Judicial Review Under the APA
The scope of judicial review is established in virtually every case by a statute that
specifies a standard against which agency action is measured by the reviewing court. The
most common statutory source is the APA, which lists several standards of review. Most
of the APA’s standards apply to all agency action while some, namely the substantial
evidence test and de novo review, apply only in specified circumstances. However, if an
agency’s enabling act contains a provision establishing a standard of review that differs
from the applicable APA standard, which many do, the enabling act provision takes
precedence over the APA standard. The following table contains a simplified model of
the applicability of APA standards of judicial review.
APA §706 thus directs courts to ‘‘hold unlawful and set aside’’ agency action that
fails to meet the applicable standard of judicial review. Some of the standards of review
listed in §706 incorporate legal standards found elsewhere such as the Constitution
(‘‘contrary to constitutional right, power, privilege, or immunity’’), substantive statutes
(‘‘in excess of statutory jurisdiction, authority, or limitations, or short of statutory
right’’), and procedural statutes and rules (‘‘without observance of procedure required by
law’’). Our focus here is largely on the three standards that relate directly to the
substantive wisdom and legality of agency action, namely §706(2)(A), (E) and (F), the
provisions that specify, respectively, the arbitrary, capricious standard, the substantial
evidence test and de novo review
Those provisions of §706(2) with no textual guidance on when they apply are
applicable to all reviewable administrative action. Specifically, subsections (2)(A)
(arbitrary, capricious), (2)(B) (contrary to the Constitution), (2)(C) (without statutory
authority), and (2)(D) (contrary to procedural requirements) apply to all agency action.
Subsections (2)(B), (2)(C), and (2)(D) are rarely mentioned because it seems to go
without saying that courts have the authority to ensure that all procedural requirements
are followed and that agency action is taken only with statutory authority and within
constitutional limits. Subsection (2)(A), however, which provides that agency action
should be set aside if it is ‘‘arbitrary, capricious, an abuse of discretion, or otherwise not
in accordance with law,’’ is adverted to frequently. It is the substantive standard of
review that is applied to the results of most agency rulemaking proceedings and also to
most informal agency decisions. Because it has no limitations on its application, the
arbitrary, capricious test is the standard that is applied when no other substantive standard
of review applies.
I. Defining and Applying the Standards of Review
The most important standard of judicial review for our purposes is §706(2)(A)’s
‘‘arbitrary, capricious, abuse of discretion or otherwise not in accordance with law’’
standard, often referred to as the ‘‘arbitrary and capricious’’ test. This standard applies to
most informal rulemakings and to other informal agency action. Although in pure volume
of cases the substantial evidence test may apply more often (because it applies to
thousands of formal agency benefits hearings), review of informal rulemaking and other
less formal agency action is generally more important and garners more attention than
review of individual benefits hearings. As elaborated in the pages that follow, the
arbitrary, capricious standard requires that agencies base their decisions on consideration
of the factors relevant to the regulatory scheme, that they consider alternatives, that their
policy conclusions make sense, that the logical path to their conclusions is discernible
and that there be a rational connection between the facts found and the policy decisions
made. Reviewing courts are not supposed to substitute their judgment for that of the
agency but still must conduct a ‘‘thorough, probing, in-depth review.’’
Agencies are required to entertain and answer petitions for rulemaking. APA §553
states that ‘‘[e]ach agency shall give an interested person the right to petition for the
issuance, amendment, or repeal of a rule.’’ Further, APA §555(e) requires agencies to
answer petitions and provide reasons for any denial. For a long while, it was an open
question whether decisions not to regulate, such as the denial of a petition for rulemaking,
were reviewable at all, and if they were reviewable there remained the question of the
proper standard of review for such decisions. Massachusetts v. EPA provided answers to
both of these questions.
Another setting in which it has been argued that the usual rigors of the arbitrary,
capricious test should be relaxed is deregulation, when an agency revokes, rescinds, or
relaxes an existing rule. The argument in favor of a more lenient standard of review is
that deregulation is more like a refusal to regulate than a decision to impose regulation
because it moves toward a state of less regulation rather than more regulation. Judicial
review is most important, the argument goes, when regulatory burdens are imposed and is
less important when regulatory burdens are reduced.
The final special case asks whether review under the arbitrary, capricious standard
should be enhanced when an agency changes its view. There is a long tradition of courts
scrutinizing agency decisions more carefully when they represent a change from past
practice. Perhaps this is due to suspicion that politics rather than expertise are behind
agency changes to longstanding policies. By demanding reasons beyond mere policy
disagreement for abandoning the prior administration’s decision to require passive
restraints, the Airbags opinion made it more difficult for agencies to change their
policies. More recently, however, the Court made it clear that the arbitrary, capricious
standard is not heightened merely because the agency has altered its policy. In FCC v.
Fox Television Stations, Inc., 129 S. Ct. 1800, 1810 (2009), the Court reviewed the
FCC’s decision to tighten up on regulation of indecency on television by abandoning its
prior policy that it would not bring enforcement action based on ‘‘fleeting’’ use of
indecent language during live broadcasts. Fox argued that review should be more
searching because the FCC had changed its longstanding policy. The Court’s response
was that ‘‘[w]e find no basis in the Administrative Procedure Act or in our opinions for a
requirement that all agency change be subjected to more searching review.’’ Id. The
Court acknowledged that agencies must provide reasons for changes in policy, but held
that review of those reasons would be no more stringent than review of the reasons given
when an agency adopts a policy initially.
J. Chevron Cases
Agencies must often interpret the statutes from which they derive their authority
to act. There has long been disagreement and confusion over whether courts or agencies
are primarily responsible for interpreting regulatory statutes. One of the most vexing and
controversial questions in judicial review of agency action has been the degree to which
courts should defer to agency statutory interpretation. The issue muddled along for
decades until 1984 when the Supreme Court, in Chevron U.S.A., Inc. v. Natural
Resources Defense Council, Inc., 467 U.S. 837 (1984), appeared to announce a new and
apparently very deferential standard for reviewing agency interpretation. However,
deeper analysis reveals that Chevron has not resolved any of the preexisting difficulties in
this area and has added its own layer of difficulty and confusion.
Competing Traditional Standards of Review of Agency Conclusions of Law.
Before Chevron, there were competing traditions regarding judicial review of agency
conclusions of law. Under one tradition, questions of law were reviewed de novo by
courts on the theory that it is the judicial role to declare the law. The competing tradition
held that courts defer to reasonable agency interpretations of law because agency
expertise assists in understanding Congress’s statutory commands and other legal issues
within the agency’s jurisdiction. A middle position held that deference to agency
statutory interpretation decisions varied depending on numerous factors. See Skidmore v.
Swift & Co., 323 U.S. 134 (1944). Today, these questions are analyzed under the
Chevron framework, which is detailed below. Application of Law to Particular Facts.
Courts have traditionally shown the greatest deference to agency decisions involving the
application of law to particular facts. Such decisions are affirmed if they enjoy ‘‘warrant
in the record’’ and a ‘‘reasonable basis in law.’’ NLRB v. Hearst Publications, Inc., 322
U.S. 111 (1944). In Hearst, the Court reviewed the NLRB’s determination that people
selling newspapers on the street were employees of the newspaper company entitled to
the protections of federal labor laws, rather than independent contractors who would not
be protected. In the first part of its Hearst opinion, applying traditional methods of
statutory interpretation, the Court upheld the Board’s decision not to apply the tort law
definition of ‘‘employee’’ in determinations of employee status under the labor laws. The
Court did not appear to defer to the Board at all. Rather, it appeared to decide de novo
that Congress did not intend for the tort law definition to apply.
The Administrator ruled that canning operations with more than seven employees
would not be exempt even if they were in the ‘‘area’’ of agricultural production as
previously defined by the Administrator. The Supreme Court, using traditional statutory
interpretation methods, held that once the Administrator determined the ‘‘area’’ of
agricultural production, there was no statutory authority to exclude canning operations
from the exemption based on the size of the operation. The Court held that there was no
occasion for deference to the agency since this was an issue of statutory authority. The
Court no longer distinguishes between issues of statutory authority and other statutory
issues. Issues of statutory authority are analyzed, like other issues of statutory
interpretation, under the Chevron doctrine outlined below.
The first difficulty is that it is unclear whether Chevron applies to pure questions
of statutory interpretation or only when the case involves application of a statutory
standard to a particular situation or policy questions bound up with statutory meaning.
Pure questions of statutory interpretation are those issues that involve only the meaning
of the words of the statute. They do not involve applying those words to a particular
situation. In a footnote in the Chevron opinion that has been quoted in subsequent cases,
the Court stated that ‘‘[t]he judiciary is the final authority on issues of statutory
construction and must reject administrative constructions which are contrary to clear
congressional intent.’’ 467 U.S. at 843 n. 9. In later cases, the Court added the word
‘‘pure’’ to this phrase, rendering it as ‘‘The question . . . is a pure question of statutory
construction for the courts to decide.’’ INS v. Cardoza-Fonseca, 480 U.S. 421, 446
(1987). This issue is elaborated upon below. The Clean Air Act provision involved in
Chevron regulated permits for the discharge of air pollution from a ‘‘stationary source.’’
In Chevron, the Court reviewed the Environmental Protection Agency’s (EPA) definition,
contained in a regulation, of the statutory term ‘‘stationary source.’’ The EPA adopted a
new definition of ‘‘stationary source,’’ which provided that all of the pollution-emitting
devises in an entire plant could be treated as a single stationary source, as if the plant was
encased in a giant bubble. This definition allowed greater flexibility for polluters who
could, for example, increase pollution from one smokestack at a plant while decreasing
pollution from another smokestack without needing a new permit for the increase at the
former smokestack. The challengers argued that the statutory language required that each
smokestack must be regulated as a separate stationary source.
467 U.S. at 833-3, this passage from the Chevron opinion has been understood as
establishing a two-step process for judicial review of agency interpretations of statutes.
Step one of Chevron asks whether Congress has directly spoken to the precise question at
issue. If so, Congress’s intent prevails. If the agency’s interpretation conflicts with
Congress’s intent, the court should overrule the agency and replace the agency’s
interpretation with the interpretation required by Congress’s intent. If Congress’s intent is
unclear, or if Congress explicitly left a gap for the agency to fill, the analysis moves to
step two of Chevron, which is very deferential to the agency’s view, upholding
reasonable or permissible agency statutory construction.
In Chevron itself, the Court held that Congress had not directly spoken to the
precise issue of whether each smokestack must be regulated as a stationary source, that
the term ‘‘stationary source’’ was ambiguous, and that the agency’s construction,
employing the bubble concept, was permissible because it fell within the range of
meanings that ‘‘stationary source’’ could bear. The theory underlying deference to
agency interpretations is that when Congress delegates authority to administer a statute to
an agency, it also delegates power to the agency to fill gaps and clarify ambiguities in the
statute. In other words, Chevron understands statutory silence or ambiguity to entail
congressional intent to delegate interpretive authority to the agency, not to the reviewing
court. Chevron distinguished between two different types of interpretive delegations
(implicit and explicit) and established different standards for reviewing agency
interpretations under step two, depending on whether the interpretive delegation to the
agency is implicit or explicit.
There is also some confusion over whether step two is about the reasonableness of
the agency’s interpretation as a matter of the meaning of the words of the statute or is
more about the reasonableness of the agency’s policy decision to adopt the particular
meaning. As part of the step two analysis in the Chevron opinion, in reviewing the
agency’s plant-wide definition of ‘‘stationary source,’’ the Court concluded that ‘‘the
EPA’s use of that concept here is a reasonable policy choice for the agency to make.’’
Chevron at 845 (emphasis supplied). See also National Cable & Telecommunications
Ass’n v. Brand X Internet Services, 545 U.S. 967 (2005). This analysis appears to call for
review of factors that would be relevant under ordinary arbitrary, capricious review of the
substance of an agency policy. The Ninth Circuit recently rejected an EPA rule on the
ground that the agency’s construction of the statute was ‘‘arbitrary and capricious,’’ and
the analysis focused more on the wisdom of the agency’s interpretation of the Clean
Water Act than on the linguistic plausibility of the agency’s construction. See Natural
Resources Defense Council v. United States Environmental Protection Agency, 526 F.3d
591, 605-08 (9th Cir. 2008).
The analysis of whether Chevron applies begins with a question of congressional
intent — did Congress delegate to the agency the power to make authoritative decisions
on the meaning of the statute? Or as the Mead Court put it, does it appear that Congress
intended for agency interpretations to have the ‘‘force of law’’? The Mead Court
concluded that the most important indication that Congress intended for agency
interpretations to have the force of law is when Congress authorized the agency to
employ relatively formal procedures such as notice and comment rulemaking and formal
adjudication. When Congress authorizes an agency to act informally, for example
through opinion letters or informal guidance documents, that indicates that Congress did
not intend for agency interpretations to have the force of law, and thus they would not
receive Chevron deference.
Although authorization to employ rulemaking or formal adjudicatory procedures
are good indications of congressional intent to delegate the power to make rulings with
the force of law, the Mead Court was careful not to make any single factor decisive. With
regard to the lack of rulemaking procedures, the Court stated that ‘‘the want of that
procedure here does not decide the case, for we have sometimes found reasons for
Chevron deference even when no such administrative formality was required and none
was afforded.’’ The Court relied upon additional factors to support its decision not to
grant Chevron deference to the Customs Ruling in Mead, such as the fact that thousands
of Customs Rulings are issued every year by numerous offices and on their terms cannot
be relied upon by anyone other than the importer whose goods were the subject of the
ruling. Perhaps this means that an agency making fewer decisions in a more centralized
manner at a high level within the agency hierarchy might receive Chevron deference even
without rulemaking or formal adjudication.
The Supreme Court, in an opinion by Justice Breyer, has suggested the following
factors that might lead toward Chevron deference in a less formal setting: ‘‘the interstitial
nature of the legal question, the related expertise of the agency, the importance of the
question to the administration of the statute, the complexity of that administration, and
the careful consideration the Agency has given the question over a long period of
time[.]’’ Barnhart v. Walton, 535 U.S. 212, 222 (2002). In Walton, the Court approved an
interpretation that was embodied in regulations but that had initially been arrived at less
formally.
The applicability of Chevron may be affected by the lack of consistency of the
agency’s current interpretation with prior agency interpretations. Although the Court has
stated that agency interpretations are not ‘‘carved in stone’’ and thus may be changed
when appropriate, the existence of a longstanding contrary interpretation may lead courts
to question the validity of any reinterpretation, especially if the prior interpretation
received prior court approval, see Maislin Industries, U.S. Inc. v. Primary Steel, Inc., 497
U.S. 116 (1990), or appears to be more consistent with other action taken by Congress
and the agency. See Food and Drug Administration v. Brown & Williamson Tobacco
Corp., 529 U.S. 120 (2000) (characterizing longstanding FDA view that statute does not
give the FDA authority to regulate tobacco as evidence that Congress did not intend to
grant the FDA general regulatory power over tobacco). See also Barnhart v. Thomas, 540
U.S. 20 (2003) (longstanding agency interpretation entitled to greater deference). The
Court has, on more than one occasion, stated that deference is reduced when an agency
interpretation goes against the agency’s prior interpretation. See, e.g., INS v. Cardoza-
Fonseca, 480 U.S. 421, 447 n. 30 (1987). See also Natural Resources Defense Council v.
United States Environmental Protection Agency, 526 F.3d 591, 605 (9th Cir. 2008). On
the other hand, the Court has applied full Chevron deference to an interpretation of a
statute that was rendered 130 years after the statute being interpreted was enacted. The
Court stated that the delay was not relevant as long as ambiguity in the statute left a gap
for the agency to fill. See Smiley v. Citibank, 517 U.S. 735 (1996).
K. Review of Questions of Fact After Agency Adjudication: The Substantial Evidence
Test
The substantial evidence test is the standard of review for formal agency
adjudication and formal rulemaking conducted under APA §§556 and 557 (or conducted
on the record after a hearing pursuant to the agency’s enabling act). In addition, some
enabling acts specify that the substantial evidence test applies to that particular agency’s
informal rulemaking. In this section, we look at the meaning and application of the
substantial evidence test. By contrast with other areas of administrative law, the meaning
and application of the substantial evidence test have both been relatively stable over time,
with little in the way of controversy.
A court performing substantial evidence review examines the evidence that was
before the agency and determines, in a rather uncomplicated way, whether the agency’s
decision was reasonable in light of the evidence on the record. Two situations merit
special attention. First, when an agency’s decision is based (in whole or in part) on the
credibility of witnesses, the agency’s decision is entitled to great deference because the
reviewing court reviews only the paper record without the opportunity to observe the
demeanor of the witnesses. If the agency’s decision relies on witness credibility, it takes a
great deal of contrary evidence to convince a court that the agency’s decision lacks
substantial evidence.
L. De Novo Review of Questions of Fact
De novo review means review without regard to the agency decision — no
deference at all. Trial de novo means that the facts are retried in the reviewing court. The
possibility of de novo review arises from APA §706(2)(F), which states that the court
should reverse agency action on review to the extent that the agency decision is
‘‘unwarranted by the facts to the extent that the facts are subject to trial de novo by the
reviewing court.’’ As discussed above, the statute does not tell us when de novo review
applies, but the Supreme Court has adopted the language of a legislative committee report
that explains that de novo review applies only in two rare circumstances: first when
agency adjudicatory factfinding procedures are inadequate, and second when new factual
issues arise in an action to enforce nonadjudicatory agency action. SeeOverton Park, 402
U.S. at 415. De novo review means that the reviewing court considers the facts and
reaches its own decision without deferring to the conclusions of the agency. Usually, in
de novo review of facts, the facts are retried in the trial court.
Reviewing courts have the power under APA §706 to ‘‘hold unlawful and set
aside’’ agency action found not to meet the applicable standard of review. In many
situations, courts do just that: They hold regulations null and void or order agencies to
pay benefits or award permits unlawfully withheld. However, especially in rulemaking
that involves policy questions, courts often choose to remand matters to the agency for
further consideration without ordering the agency to change its decision. One of the
reasons that reviewing courts may prefer remand to outright reversal is that courts seem
to prefer to rest the rejection of agency policy decisions on procedural grounds such as
inadequate notice or inadequate opportunity to comment. Courts shy away from direct
substantive rejection of an agency’s policy decision because agency decisions on matters
of policy implicate agency expertise. Courts do not wish to appear to be substituting their
judgment for that of the experts. Another reason for remand rather than outright reversal
is that reviewing courts may not have the authority to step into the agency’s shoes and
make the discretionary decisions necessary to choose the course the agency will take after
its initial decision has been rejected on judicial review.
