Content uploaded by Kristin E. Hickman
Author content
All content in this area was uploaded by Kristin E. Hickman on Jan 06, 2014
Content may be subject to copyright.
\\server05\productn\C\COL\107-6\toc1076.txt unknown Seq: 1 11-OCT-07 14:53
COLUMBIA LAW REVIEW
VOL. 107 OCTOBER 2007 NO. 6
CONTENTS
ARTICLES
I
N
S
EARCH OF THE
M
ODERN
S
KIDMORE
S
TANDARD
Kristin E. Hickman 1235
Matthew D. Krueger
A
BOLITION OF THE
C
ORPORATE
D
UTY TO
C
REDITORS
Henry T. C. Hu 1321
Jay Lawrence Westbrook
NOTES
S
TATUS ON
T
RIAL
: T
HE
R
ACIAL
R
AMIFICATIONS
OF
A
DMITTING
P
ROSTITUTION
E
VIDENCE
U
NDER
S
TATE
R
APE
S
HIELD
L
EGISLATION
Karin S. Portlock 1404
E
DITING
D
IRECT
D
EMOCRACY
: D
OES
L
IMITING
THE
S
UBJECT
M
ATTER OF
B
ALLOT
I
NITIATIVES
O
FFEND THE
F
IRST
A
MENDMENT
?John Gildersleeve 1437
ESSAY
C
OMMON
L
AW
C
ONSTITUTIONALISM
AND THE
L
IMITS OF
R
EASON
Adrian Vermeule 1482
\\server05\productn\C\COL\107-6\boe1076.txt unknown Seq: 8 11-OCT-07 14:53
COLUMBIA LAW REVIEW
VOL. 107 OCTOBER 2007 NO. 6
K
ARIN
S. P
ORTLOCK
Editor-in-Chief
N
EIL
M. S
NYDER
S
HANNON
R
EBHOLZ
J
ENNIFER
S. N
AM
J
OHN
G
ILDERSLEEVE
Executive Articles Executive Managing
Executive Essay & Review
Executive Notes
Editor Editor
Editor
Editor
A
NDREW
A
MEND
D
AVID
L
IEBERMAN
J
AMES
W. D
OGGETT
T
RACY
A
PPLETON
A
DRIAN
J. R
ODR´
IGUEZ
C
ARI
F
AIS
R. S
ETH
D
AVIS
J
OSEPH
M. R
USCHELL
C
HRISTOPHER
H
OGAN
D
AVID
G
RINGER
V
IVIAN
H. W. W
ANG
B
RIAN
P. L
ARKIN
E
MILY
M. M
EEKER
Managing Editors Notes Editors
P
ETER
S
CHWINGLER
Articles Editors K
IMBERLY
D. J
AIMEZ
M
ATTHEW
M
ILIKOWSKY
A
RIANA
C
OOPER
J
EFFREY
V
ERNON
H
ELEN
O
GBARA
Librarian & Senior Editor Essay & Review Editors Milligan Fellowship Coordinator
K
ATHLEEN
M. C
OCHRANE
E
LIJAH
M. A
LPER
J
OSEPH
D
WYER
M
URPHY
Alumni Liaison & Operations Staff Development Editor &
Senior Editor Editor Senior Editor
C
OLIN
S. K
ELLY
J
ENNA
L
EVINE
S
TUART
M
C
P
HAIL
Online Editors
J
ORDAN
C
ONNORS
R
YAN
W. P
AKTER
D
OUG
S
NODGRASS
S
ARA
F
ROIKIN
J
ULIAN
M. P
ETRIN
B
RIAN
S
NYDER
A
NDREW
G
OODRICH
A
NDY
P
RATT
D
ANIEL
S
TEWART
A
LICE
Y. L
EE
A
KIVA
S
HAPIRO
Z
HIYUAN
Z
HU
Senior Editors Senior Editors Senior Editors
Staff
D
AVID
A
BRAMOWICZ
J
ESSICA
K
ARP
P
AWAN
N
ELSON
K
EVIN
A
NGLE
S
HIRA
K
IEVAL
J
ENNIFER
P
HILBRICK
S
AMEER
B
AJAJ
J
OHN
K
OERNER
E
ILEEN
P
LAZA
B
ENJAMIN
J. B
EATON
A
NDREW
L
E
G
RAND
T
HANE
R
EHN
A
NDREW
B
RANTINGHAM
D
ANIELLA
L
ICHTMAN
R
YAN
R
EYNOLDS
S
COTT
B
UELL
K
AREN
L
IN
E
MILY
R
OSSI
C
ORY
B
ULAND
C
HANG
D. L
IU
B
ENJAMIN
R
OTHSTEIN
E
MILY
B
USSIGEL
A
DRIANA
L
UCIANO
S
HELBY
R. S
CHWARTZ
M
ICHAEL
C
ABIN
K
ARA
M
AGUIRE
N
OAH
S
OLOWIEJCZYK
Z
E
-
WEN
J
ULIUS
C
HEN
B
RANDON
C. M
ASON
T. A
LORA
T
HOMAS
J
ESSIE
C
HENG
K
ABIR
M
ASSON
J
ASON
V
ITULLO
E
THAN
F
RECHETTE
B
RYAN
M
C
A
RDLE
R
ONNI
W
EINSTEIN
M
ATTHEW
G
URGEL
A
MY
M
C
C
AMPHILL
R
OBERT
W
EINSTOCK
K
IRSTEN
J
ACKSON
C
HRISTOPHER
A. M
ULLER
E
MILY
W
EISS
K
ATHLEEN
C
HOJNICKI
G
UILLERMO
C
´
ACERES
I
N´
ES
D
UBBELS
Business Manager Project Coordinator Assistant Business Manager
Directors of the Columbia Law Review Association, Inc.
P
ETER
B
USCEMI
L
OUIS
L
OWENSTEIN
W
ILLIAM
S
AVITT
P
ETER
C
ANELLOS
G
ERARD
L
YNCH
D
AVID
S
CHIZER
G
RANT
R. M
AINLAND
, ex officio G
ILLIAN
M
ETZGER
L
EWIS
Y
ELIN
N
ORY
M
ILLER
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 1 11-OCT-07 15:18
COLUMBIA LAW REVIEW
VOL. 107 OCTOBER 2007 NO. 6
ARTICLES
IN SEARCH OF THE MODERN SKIDMORE STANDARD
Kristin E. Hickman*
Matthew D. Krueger**
This Article offers a comprehensive examination of the Skidmore stan-
dard of judicial review as applied by the courts in the period since the
Supreme Court revitalized Skidmore in United States v. Mead Corp.
The Article documents an empirical study of five years worth of Skidmore
applications in the federal courts of appeals. In the study, we evaluate two
competing conceptions of Skidmore review that are apparent from the Su-
preme Court’s post-Mead jurisprudence—the independent judgment model
and the theoretically more deferential sliding-scale model—and demonstrate
that the appellate courts overwhelmingly follow the sliding-scale approach.
Also, we document that Skidmore review is much more deferential to agency
legal interpretations than indicated by two other, significantly more limited
studies, with agency interpretations prevailing in more than sixty percent of
Skidmore applications. Drawing from the Skidmore applications stud-
ied, we analyze qualitatively how the appellate courts apply the Skidmore
standard as a sliding scale and identify where those courts are struggling to
make sense of Skidmore’s dictates within that model. To resolve the lower
courts’ difficulties, we propose reconceptualizing Skidmore’s sliding scale as
balancing comparative agency expertise against the potential for agency arbi-
trariness across three attitudinal zones. Finally, we note several burgeoning
issues concerning the scope of Skidmore’s applicability and offer prelimi-
nary thoughts for addressing those questions.
I
NTRODUCTION
.................................................. 1236
R
I.
S
KIDMORE
D
EFERENCE AND
J
UDICIAL
R
EVIEW OF
A
DMINISTRATIVE
I
NTERPRETATIONS
......................... 1239
R
A. The Evolution of Judicial Deference Doctrine ........ 1239
R
1. Skidmore in the Pre-Chevron Period ................ 1240
R
* Associate Professor of Law, University of Minnesota Law School.
** Bristow Fellow, Office of the Solicitor General, U.S. Department of Justice. All of
the views expressed in this Article are our own, and none represent the positions or views
of the United States or the Department of Justice. We would like to thank Ellen Aprill,
Tino Cuellar, Bill Eskridge, Brad Karkkainen, Ron Levin, Brett McDonnell, Tom Merrill,
Jim Rossi, and David Zaring for helpful comments and suggestions, and Lindsey Tonsager
for excellent research assistance.
1235
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 2 11-OCT-07 15:18
1236 COLUMBIA LAW REVIEW [Vol. 107:1235
2. The Chevron Revolution and the Uncertain Status
of Skidmore ....................................... 1241
R
3. The Mead Counterrevolution and Revitalization of
Skidmore .......................................... 1245
R
B. The Deference Framework ........................... 1246
R
II. W
HAT
I
S
S
KIDMORE
D
EFERENCE
?........................... 1250
R
A. Competing Conceptions of Skidmore Review ........... 1251
R
1. Distinguishing Independent Judgment from
Judicial Deference ............................... 1251
R
2. The Independent Judgment Model of Skidmore
Review ........................................... 1252
R
3. Skidmore’s Review as Deference Varying Along a
Sliding Scale ..................................... 1255
R
B. Skidmore Deference in Practice: An Empirical Study . . 1259
R
1. Identifying Skidmore Applications ................. 1259
R
2. Evaluating the Skidmore Applications .............. 1267
R
3. Results of Study of Skidmore in Courts of Appeals . 1271
R
a. Establishing Dominance of the Sliding-Scale
Model ....................................... 1271
R
b. Documenting Skidmore’s Deferential
Character .................................... 1275
R
c. One Final Note: Skidmore “Step One” ........ 1280
R
4. Summary ........................................ 1280
R
C. Skidmore’s Sliding Scale in Practice ................... 1281
R
1. Thoroughness of Consideration .................. 1281
R
2. Formality of the Agency’s Procedure and
Interpretation .................................... 1283
R
3. Validity of the Agency’s Reasoning ............... 1285
R
4. Consistency of the Agency’s Interpretation........ 1286
R
5. Agency Expertise ................................. 1288
R
6. Longstanding or Contemporaneous
Interpretations ................................... 1289
R
7. Summary ........................................ 1291
R
III. R
ECONCEPTUALIZING
S
KIDMORE
R
EVIEW
.................... 1291
R
IV.
S
KIDMORE
’
S
D
OMAIN
...................................... 1299
R
C
ONCLUSION
.................................................... 1309
R
I
NTRODUCTION
Skidmore deference is back. For forty years, the Supreme Court’s
opinion in Skidmore v. Swift & Co.
1
enjoyed prominence as perhaps the
Supreme Court’s best expression of its policy of judicial deference toward
many if not most agency interpretations of law.
2
Skidmore called upon
1. 323 U.S. 134 (1944).
2. See, e.g., Batterton v. Francis, 432 U.S. 416, 425 n.9 (1977) (citing Skidmore for
guiding standard of review for agency interpretation); Gen. Elec. Co. v. Gilbert, 429 U.S.
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 3 11-OCT-07 15:18
2007] MODERN SKIDMORE 1237
reviewing courts to assess multiple factors to decide on a case-by-case basis
what deference, if any, to afford agency legal interpretations.
3
With its
well-known 1984 decision in Chevron U.S.A. Inc. v. Natural Resources Defense
Council, Inc. and its emphasis therein on mandatory deference toward
reasonable agency interpretations of ambiguous statutes,
4
the Supreme
Court threw the viability of Skidmore into doubt. Then, with Christensen v.
Harris County in 2000 and United States v. Mead Corp. the following year,
the Court clarified that Chevron’s scope is not limitless and that Skidmore
governs a wide range of administrative interpretations that do not carry
congressionally authorized legal force.
5
Thus, in 2001 the “modern” Skidmore era began. Courts again regu-
larly invoke Skidmore—as well as Christensen and Mead—as providing the
guiding standards for judicial review of administrative interpretations.
6
However, while Christensen and Mead resurrected Skidmore’s now boiler-
plate recitation of factors, the Court has been substantially less clear in
explaining how lower courts should apply the Skidmore standard. Indeed,
the Court’s discussions of Skidmore in Christensen and Mead reflect surpris-
ingly different conceptions of Skidmore’s standard for evaluating adminis-
trative interpretations. All agree that Skidmore is less deferential than
Chevron, but how much less and in what way remain open questions. Fur-
thermore, just as the boundaries of Chevron’s domain were substantially
less certain pre-Mead, the scope of Skidmore’s applicability in the post-
Mead era is still unclear.
While Skidmore has reemerged as a unique and frequently used stan-
dard of review, contemporary scholarship contains little discussion of
these unsettled questions regarding the “modern” Skidmore doctrine.
7
125, 141–42 (1976) (same); Fed. Mar. Bd. v. Isbrandtsen Co., 356 U.S. 481, 499–500 (1958)
(same); Mabee v. White Plains Publ’g Co., 327 U.S. 178, 182 (1946) (same); 2 Kenneth
Culp Davis, Administrative Law Treatise § 7:10, at 50 (2d ed. 1979) [hereinafter 2 Davis]
(describing Skidmore in such terms).
3. See Skidmore, 323 U.S. at 140.
4. 467 U.S. 837, 843–45 (1984).
5. See United States v. Mead Corp., 533 U.S. 218, 230–33 (2001); Christensen v.
Harris County, 529 U.S. 576, 587 (2000).
6. See, e.g., Gonzales v. Oregon, 546 U.S. 243, 268–69 (2006); Raymond B. Yates,
M.D., P.C. Profit Sharing Plan v. Hendon, 541 U.S. 1, 18 (2004); Alaska Dep’t of Envtl.
Conservation v. EPA, 540 U.S. 461, 487–88 (2004); Clackamas Gastroenterology Assocs.,
P.C. v. Wells, 538 U.S. 440, 449–50 (2003); Wash. State Dep’t of Soc. & Health Servs. v.
Guardianship Estate of Keffeler, 537 U.S. 371, 385–86 (2003); Nat’l R.R. Passenger Corp. v.
Morgan, 536 U.S. 101, 110 n.6 (2002).
7. The whole of scholarly discussion of Skidmore in its “modern” form can be found in
these works: William N. Eskridge, Jr. & Lauren E. Baer, The Supreme Court’s Deference
Continuum, An Empirical Analysis (from Chevron to Hamdan), 96 Geo. L.J. (forthcoming
Apr. 2008) (manuscript at 92–94, on file with authors); Michael Herz, Judicial Review of
Statutory Issues Outside of the Chevron Doctrine, in A Guide to Judicial and Political
Review of Federal Agencies 125 (John F. Duffy & Michael Herz eds., 2005) [hereinafter
Herz, Judicial Review]; Ronald J. Krotoszynski, Jr., Why Deference?: Implied Delegations,
Agency Expertise, and the Misplaced Legacy of Skidmore, 54 Admin. L. Rev. 735, 750–56
(2002); Richard W. Murphy, A “New” Counter-Marbury: Reconciling Skidmore Deference
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 4 11-OCT-07 15:18
1238 COLUMBIA LAW REVIEW [Vol. 107:1235
This Article begins to fill this void by evaluating the Skidmore standard
from several angles and synthesizing the results into a single, practicable
Skidmore framework.
Part I of this Article provides a brief overview of Skidmore’s history
and place in the judicial deference framework that includes Chevron and
Mead. Part II then empirically assesses the “modern” Skidmore standard
that courts apply daily as a doctrine of judicial deference. In Part II.A, we
identify two conceptions of Skidmore that compete in both the Supreme
Court’s and lower courts’ opinions: the independent judgment model,
which effectively denies any deference to agencies, and the sliding-scale
model, which tailors deference in accordance with Skidmore’s factors. In
Part II.B, we then document our empirical study of federal courts of ap-
peals’ post-Mead applications of Skidmore’s standard. We conclude that
the sliding-scale model predominates over the independent judgment
model and, thus, that most judges perceive Skidmore as an actual restraint
on their decisionmaking. In the same vein, our study shows that Skid-
more’s standard is, as a whole, surprisingly deferential, with courts apply-
ing Skidmore’s standard to accept agencies’ views at a higher rate than was
previously assumed by some scholars.
8
However, our study also shows that the sliding-scale model lacks uni-
formity and that courts have varying views of what Skidmore’s factors
mean. Therefore, in Part II.C we delve deeper into the sliding-scale ap-
proach to Skidmore deference by analyzing the various factors as applied.
Based on our analysis of appellate practice, in Part III we propose
reconceptualizing Skidmore’s sliding-scale approach in two ways. We posit
that Skidmore’s sliding scale encompasses three zones or “moods” reflect-
ing strong, intermediate, and weak or no deference. To determine which
of these moods to adopt in evaluating an agency interpretation, we also
suggest that courts focus on Skidmore’s underlying goal of respecting
agency expertise while guarding against agency arbitrariness, employing
Skidmore’s factors in pursuit of that end rather than for their own sake.
Finally, our study uncovered burgeoning questions over when Skid-
more’s deference standard should apply at all, rather than pure de novo
review. We suspect that, as with Chevron deference, it is only a matter of
and Agency Interpretive Freedom, 56 Admin. L. Rev. 1, 46–51 (2004) [hereinafter
Murphy, Counter-Marbury]; Christopher M. Pietruszkiewicz, Discarded Deference: Judicial
Independence in Informal Agency Guidance, 74 Tenn. L. Rev. 1, 45 (2006); Jim Rossi,
Respecting Deference: Conceptualizing Skidmore Within the Architecture of Chevron, 42
Wm. & Mary L. Rev. 1105, 1137–46 (2001); Amy J. Wildermuth, Solving the Puzzle of Mead
and Christensen: What Would Justice Stevens Do?, 74 Fordham L. Rev. 1877, 1896–1906
(2006); Eric R. Womack, Into the Third Era of Administrative Law: An Empirical Study of
the Supreme Court’s Retreat from Chevron Principles in United States v. Mead, 107 Dick. L.
Rev. 289, 323–33 (2002). Of course, many other articles discuss Skidmore, but principally in
the context of analyzing Chevron, Mead, or other related issues.
8. See infra notes 223–234 and accompanying text (summarizing and analyzing
R
studies by Eric Womack and Amy Wildermuth finding low rates of agency success under
Skidmore).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 5 11-OCT-07 15:18
2007] MODERN SKIDMORE 1239
time before some of these questions divide the lower courts outright. In
Part IV, we bring these underlying issues to light and offer some prelimi-
nary thoughts on the proper extent of Skidmore’s domain.
I.
S
KIDMORE
D
EFERENCE AND
J
UDICIAL
R
EVIEW OF
A
DMINISTRATIVE
I
NTERPRETATIONS
This Article focuses on Skidmore, not Chevron or Mead. Nevertheless,
the Supreme Court’s judicial deference doctrine has evolved over time,
through each of these cases, so that they now function collectively as parts
of a comprehensive framework for judicial review of administrative inter-
pretations. Evaluating the modern Skidmore standard thus calls for a brief
treatment of the deference doctrine’s evolution and the Court’s current
framework for judicial deference. This Part summarizes Skidmore’s his-
tory, particularly as it relates to Chevron and Mead. It also situates the
recently revitalized Skidmore standard within the analytical framework for
judicial review of administrative interpretations.
A. The Evolution of Judicial Deference Doctrine
Judicial deference to agency interpretations of law predates any of
Skidmore, Chevron, or Mead. In cases such as AT&T v. United States and
Atchison, Topeka & Santa Fe Railway v. Scarlett, the Supreme Court in-
structed that reviewing courts should uphold regulations adopted pursu-
ant to a specific grant of legislative power unless the promulgating agency
exceeded the scope of its statutory authority.
9
The same principle of con-
trolling deference also applied where an agency exercised a specific au-
thority grant through formal adjudication.
10
Nevertheless, such specific
authority grants from Congress were few; and for forty years before Chev-
ron was decided, the Supreme Court’s opinion in Skidmore v. Swift & Co.
11
was a leading expression of the Court’s policy toward judicial review of
most other administrative interpretations.
12
9. Atchison, Topeka & Santa Fe Ry. v. Scarlett, 300 U.S. 471, 474 (1937) (“The
regulation having been made by the commission in pursuance of constitutional statutory
authority, it has the same force as though prescribed in terms by the statute.”); AT&T v.
United States, 299 U.S. 232, 236–37 (1936) (“This court is not at liberty to substitute its
own discretion for that of administrative officers who have kept within the bounds of their
administrative powers.”); see also, e.g., Norfolk & W. Ry. v. United States, 287 U.S. 134, 141
(1932); Fawcus Mach. Co. v. United States, 282 U.S. 375, 378 (1931); Kan. City S. Ry. v.
United States, 231 U.S. 423, 447 (1913); United States v. Moore, 95 U.S. 760, 763 (1877).
10. See, e.g., NLRB v. Hearst Publ’ns, Inc., 322 U.S. 111, 131 (1944) (“But where the
question is one of specific application of a broad statutory term in a proceeding in which
the agency administering the statute must determine it initially, the reviewing court’s
function is limited.”); Gray v. Powell, 314 U.S. 402, 412 (1941) (“Where, as here, a
determination has been left to an administrative body, this delegation will be respected
and the administrative conclusion left untouched.”).
11. 323 U.S. 134 (1944).
12. See supra note 2 and accompanying text.
R
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 6 11-OCT-07 15:18
1240 COLUMBIA LAW REVIEW [Vol. 107:1235
1. Skidmore in the Pre-Chevron Period. — The interpretive question
in Skidmore was whether the time that fire-fighting employees of a packing
plant spent on call for fire alarms constituted “working time,” for which
the employees were owed overtime pay under the Fair Labor Standards
Act (FLSA).
13
The Administrator of the Department of Labor’s Wage
and Hour Division had issued informal rulings applying the statute in
various scenarios and advocating a case-by-case approach to interpreta-
tion.
14
Though none of those rulings addressed the circumstances at bar,
the Administrator applied the rulings in an amicus brief to conclude that
only some of the time in question was compensable.
15
Nevertheless, the
lower courts in Skidmore had ignored the Administrator’s interpretation
and decided as a matter of law that such “waiting time” could not be
working time.
16
Congress had expressly given the courts, rather than the
Administrator, primary interpretive responsibility over the FLSA,
17
but
the Skidmore Court recognized its own past practice of giving weight to
interpretations by executive agencies of statutes they administered.
18
The Court therefore neither accepted nor explicitly rejected the lower
courts’ interpretation of the FLSA.
19
Instead, the Court remanded the
case for reconsideration, and in so doing articulated the standard by
which the lower courts should evaluate such cases:
[T]he rulings, interpretations and opinions of [the agency],
while not controlling upon the courts by reason of their author-
ity, do constitute a body of experience and informed judgment
to which courts and litigants may properly resort for guidance.
The weight of such a judgment in a particular case will depend
upon the thoroughness evident in its consideration, the validity
of its reasoning, its consistency with earlier and later pronounce-
ments, and all those factors which give it power to persuade, if
lacking power to control.
20
Skidmore carefully disclaims an administrative interpretation’s power
to “control” the court’s decision, but it also suggests that the interpreta-
13. See Skidmore, 323 U.S. at 135–36. The FLSA requires employers to provide
overtime pay for hours worked in excess of forty per week. See 29 U.S.C. § 207(a)(1)
(2000). The firefighters were required to be on duty at or near the firehouse to respond to
incoming fire alarms but typically spent such time sleeping or playing pool or dominos.
See Skidmore, 323 U.S. at 136. In a companion case, Armour & Co. v. Wantock, the Court
analyzed the relevant statutory provisions and found that the statute did not expressly
preclude treating such time as working time subject to overtime pay. See 323 U.S. 126,
132–34 (1944); see also Skidmore, 323 U.S. at 136 (relying on this finding in Armour to
reach its own holding).
14. Skidmore, 323 U.S. at 138.
15. Id. at 139.
16. See id. at 136.
17. Id. at 137 (citing Kirschbaum v. Walling, 316 U.S. 517, 523 (1942)).
18. See id. at 140.
19. See id.
20. Id.
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 7 11-OCT-07 15:18
2007] MODERN SKIDMORE 1241
tion should have “weight” in some cases.
21
Thus, several scholars have
described Skidmore as prescribing a kind of weak deference, falling some-
where between the poles of independent judgment and controlling
deference.
22
2. The Chevron Revolution and the Uncertain Status of Skidmore.—
The dichotomy of strong deference for exercises of specific authority
grants and weaker Skidmore deference for other administrative interpreta-
tions prevailed until 1984,
23
when the Court changed the deference land-
scape with its decision in Chevron.
24
The Chevron decision is best known
for articulating the Court’s two-part test for evaluating agency interpreta-
tions of law: whether the meaning of the statutory language in question
is clear and unambiguous; and if not, whether the agency’s interpretation
of that statutory language is a permissible one.
25
In fact, Chevron’s two
steps merely reflect pre-Chevron deference principles.
26
Even before
Chevron, if the meaning of a statute was clear, there was no opportunity
for an agency to claim judicial deference.
27
Unambiguous statutes are
not susceptible of multiple interpretations; and absent constitutional is-
sues, courts are bound to follow the clearly expressed intent of
21. Id.
22. See, e.g., John F. Manning, Constitutional Structure and Judicial Deference to
Agency Interpretations of Agency Rules, 96 Colum. L. Rev. 612, 686–88 (1996) (describing
Skidmore as “a nonbinding version of deference” from “a court exercising independent
judgment”); Richard J. Pierce, Jr., Democratizing the Administrative State, 48 Wm. & Mary
L. Rev. 559, 568–69 (2006) (describing Skidmore standard as “weaker and more contingent
type of deference” than Chevron); Rossi, supra note 7, at 1116–18 (describing Skidmore as
R
“weak deference” and as “a lesser degree of deference” than Chevron).
23. See, e.g., Chrysler Corp. v. Brown, 441 U.S. 281, 295–316 (1979) (explaining that
only “properly promulgated, substantive agency regulations have the ‘force and effect of
law’” (citing Batterton v. Francis, 432 U.S. 416, 425 n.9 (1977))); Foti v. INS, 375 U.S. 217,
223 (1963); United States v. Mersky, 361 U.S. 431, 437–38 (1960); Atchison, Topeka &
Santa Fe Ry. v. Scarlett, 300 U.S. 471, 474 (1937); see also Batterton, 432 U.S. at 425–26 &
n.9 (1977) (summarizing then-prevailing deference doctrine); 2 Davis, supra note 2,
R
§ 7:10, at 50–54 (discussing same).
24. 467 U.S. 837 (1984).
25. See id. at 842–43.
26. See, e.g., Kenneth Culp Davis, Administrative Law of the Eighties: 1989
Supplement to Administrative Law Treatise 2d § 29:16, at 505–07 (1989) [hereinafter
Davis, 1989 Supplement] (opining that Chevron’s command of controlling deference to
reasonable legislative regulations reflected longstanding doctrine). But see, e.g., E.
Donald Elliott, Chevron Matters: How the Chevron Doctrine Redefined the Roles of
Congress, Courts and Agencies in Environmental Law, 16 Vill. Envtl. L.J. 1, 6–8 (2005)
(describing Chevron’s two steps as major doctrinal shift).
27. See, e.g., NLRB v. Brown, 380 U.S. 278, 291 (1965) (“Reviewing courts are not
obliged to stand aside and rubber-stamp their affirmance of administrative decisions that
they deem inconsistent with a statutory mandate or that frustrate the congressional policy
underlying a statute.”); Norwegian Nitrogen Prod. Co. v. United States, 288 U.S. 294, 315
(1933) (“True indeed it is that administrative practice does not avail to overcome a statute
so plain in its commands as to leave nothing for construction.”).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 8 11-OCT-07 15:18
1242 COLUMBIA LAW REVIEW [Vol. 107:1235
Congress.
28
For ambiguous statutes, the standard of strong, mandatory
deference for reasonable agency interpretations issued pursuant to ex-
press congressional commands pre-existed Chevron.
29
Standing alone,
therefore, Chevron’s two-part test is notable more as a tool for organizing
judicial analysis than as a unique doctrinal statement.
Chevron is more significant but less often appreciated for expanding
the applicability of the strong form of judicial deference.
30
The interpre-
tive question in Chevron involved the Environmental Protection Agency’s
exercise of general rather than specific rulemaking authority to redefine
a term in the Clean Air Act.
31
The Court characterized the ambiguous
statute, coupled with the general rulemaking grant under which the EPA
interpreted it, as an “implicit” delegation of legislative authority over the
instant question.
32
Thus, the Court counseled mandatory, controlling
deference not only where Congress specifically calls for regulatory elabo-
ration or formal adjudication, but also where Congress implicitly dele-
gates interpretive power through the combination of statutory ambiguity
and administrative responsibility.
33
This application of compulsory judi-
cial deference to so-called implicit delegations, more than the two-part
test, is what made Chevron revolutionary.
34
However, Chevron did not make clear when exactly courts should pre-
sume that Congress delegated interpretive authority to the agency, or
28. There is extensive debate, however, over how clear a statute must be to qualify as
unambiguous in this context. See infra notes 168–170 and accompanying text.
R
29. See supra notes 10–11 and accompanying text; see also Davis, 1989 Supplement,
R
supra note 26, § 29:16, at 505–07 (suggesting that Chevron reflects longstanding doctrine);
R
Bernard Schwartz, Administrative Law § 10.34, at 703 (3d ed. 1991) (citing Inv. Co. Inst. v.
Conover, 790 F.2d 925, 931–32 (D.C. Cir. 1986), for view that Chevron deference is more
evolutionary than revolutionary).
30. See Chevron, 467 U.S. at 844 (“Sometimes the legislative delegation to an agency
on a particular question is implicit rather than explicit. In such a case, a court may not
substitute its own construction of a statutory provision for a reasonable interpretation
made by the administrator of an agency.” (emphasis added)); Davis, 1989 Supplement,
supra note 26, § 29:16-1, at 508, § 29:16-10, at 525 (noting that Chevron expanded
R
applicability of strong deference).
31. See Chevron, 467 U.S. at 840–41 (citing 46 Fed. Reg. 50,766 (Oct. 14, 1981)). The
term in question, “stationary source,” was loaded with policy implications, and a change in
presidential administrations prompted the EPA to alter its definition to include an entire
pollution-emitting plant rather than an individual pollution-emitting piece of equipment.
See id. at 841–42.
32. See id. at 844.
33. See id.
34. See, e.g., Michael Herz, Deference Running Riot: Separating Interpretation and
Lawmaking Under Chevron, 6 Admin. L.J. Am. U. 187, 203 (1992) [hereinafter Herz,
Running Riot] (“If Chevron is a revolutionary case, what makes it so is its apparent
hospitality to implied delegations generally, and delegations by ambiguity in particular.”);
Jonathan T. Molot, The Judicial Perspective in the Administrative State: Reconciling
Modern Doctrines of Deference with the Judiciary’s Structural Role, 53 Stan. L. Rev. 1,
71–73 (2000) (labeling Chevron’s call for deference to implied delegations a “transfer of
interpretive authority . . . to administrators [that] alters the traditional relationship
between judges and legislators”).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 9 11-OCT-07 15:18
2007] MODERN SKIDMORE 1243
concomitantly, when Chevron’s framework of controlling deference was
appropriate. Some, most notably Justice Antonin Scalia, read Chevron
broadly to govern any authoritative administrative interpretation of a stat-
ute the agency was charged with implementing.
35
By this view, Chevron’s
scope was vast, completely replacing the pre-Chevron multifactor ap-
proach—including Skidmore.
36
Others, however, attempted to reconcile
Chevron with the pre-Chevron case law. One group, including now-Justice
Stephen Breyer, contended that the pre-Chevron factors remained rele-
vant as part of Chevron analysis (or vice versa).
37
Another camp advocated
separate spheres for Chevron and Skidmore review.
38
Under these concep-
tions, Skidmore retained some vitality, although just how much was not
precisely clear.
Chevron seemed to abandon many of the factors previously consid-
ered relevant in determining the court’s proper level of deference. Skid-
more’s focus on the context of the agency’s interpretation seemed particu-
larly out of place in Chevron’s regime.
39
Indeed, the Chevron Court
35. See Antonin Scalia, Judicial Deference to Administrative Interpretations of Law,
1989 Duke L.J. 511, 519; see also Peter H. Schuck & E. Donald Elliott, To the Chevron
Station: An Empirical Study of Federal Administrative Law, 1990 Duke L.J. 984, 1024
(contending that Chevron “swept aside all of these [Skidmore] criteria for determining the
extent of deference and set forth a dramatic reformulation of the grounds for deferring to
agency constructions of statutes”).
36. See Christensen v. Harris County, 529 U.S. 576, 589–91 & n.* (2000) (Scalia, J.,
concurring) (asserting Chevron deference is inapplicable only when “the statute is
unambiguous,” “no interpretation has been made by personnel . . . responsible for
administering the statute,” or that “interpretation . . . was not authoritative,” and calling
Skidmore an “anachronism”); EEOC v. Arabian Am. Oil Co. (ARAMCO), 499 U.S. 244,
259–60 (1991) (Scalia, J., concurring) (characterizing Skidmore’s standard, as represented
in Gen. Elec. Co. v. Gilbert, 429 U.S. 125, 141 (1976), again as an “anachronism”); see also
Kenneth W. Starr, Judicial Review in the Post-Chevron Era, 3 Yale J. on Reg. 283, 297 (1986)
(declaring that Chevron “cast doubt upon” continuing validity of Skidmore’s multifactor
approach to deference).
37. See, e.g., Stephen Breyer, Judicial Review of Questions of Law and Policy, 38
Admin. L. Rev. 363, 379–81 (1986) (arguing for loose reading of Chevron incorporating
pre-Chevron considerations); Herz, Running Riot, supra note 34, at 208–09 (supporting
R
“continuing role” for Skidmore within Chevron analysis); see also Christensen, 529 U.S. at
596–97 (Breyer, J., dissenting) (contending that “Chevron made no relevant change” to
Skidmore analysis but rather “simply focused upon an additional, separate legal reason for
deferring to certain agency determinations”).
38. See, e.g., Robert A. Anthony, Which Agency Interpretations Should Bind Citizens
and the Courts?, 7 Yale J. on Reg. 1, 40–63 (1990) [hereinafter Anthony, Agency
Interpretations] (suggesting that Chevron should apply in some cases and Skidmore in
others).
39. See, e.g., Smiley v. Citibank (S.D.), N.A., 517 U.S. 735, 740–41 (1996) (denying
relevance of interpretation’s contemporaneity for Chevron analysis); Rust v. Sullivan, 500
U.S. 173, 186–87 (1991) (rejecting argument that Chevron deference is unavailable for
revised interpretation, but also noting that agency “amply justified [its] change of
interpretation”). In other cases, the Court invoked various Skidmore factors in applying
Chevron step two, although the factors played a diminished role. See, e.g., Good Samaritan
Hosp. v. Shalala, 508 U.S. 402, 416–20 (1993) (noting inconsistency of agency’s
interpretations over time but nonetheless accepting agency’s current view as reasonable);
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 10 11-OCT-07 15:18
1244 COLUMBIA LAW REVIEW [Vol. 107:1235
emphatically rejected the argument that the EPA’s interpretation should
fail because it was inconsistent with previous interpretations.
40
Some ini-
tial post-Chevron decisions correspondingly suggested that Chevron super-
seded Skidmore.
41
By the early 1990s, however, the Court began hinting that Skidmore
persisted as a separate standard of deference.
42
In the 1991 case of EEOC
v. Arabian American Oil Co. (ARAMCO), the Court rejected the view em-
bodied in an EEOC interpretive guideline, which construed Title VII of
the Civil Rights Act to cover U.S. citizens working for U.S. companies
outside the U.S.
43
The majority opinion concluded that the EEOC’s
guideline merited only Skidmore rather than Chevron deference because
the Court’s precedents established that Congress “‘did not confer upon
the EEOC authority to promulgate rules or regulations’” interpreting Ti-
tle VII.
44
In the years both before and after ARAMCO, the Court similarly
suggested in several other cases that Chevron would not apply to some
administrative interpretations and that some lesser deference standard
existed.
45
Yet Justice Scalia, concurring in ARAMCO, called Skidmore def-
erence “an anachronism” in the post-Chevron era.
46
Furthermore, the
Court in the 1990s extended Chevron deference to a number of agency
Arkansas v. Oklahoma, 503 U.S. 91, 110 (1992) (noting agency’s consistency in accepting
its interpretation as reasonable).
40. See Chevron U.S.A. Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837, 863–64
(1984); see also Richard W. Murphy, Judicial Deference, Agency Commitment, and Force
of Law, 66 Ohio St. L.J. 1013, 1033–35 (2005) [hereinafter Murphy, Judicial Deference]
(describing consistency as irrelevant under Chevron analysis).
41. See, e.g., Adams Fruit Co. v. Barrett, 494 U.S. 638, 649–50 (1990) (rejecting
Chevron’s applicability in absence of delegated authority but failing to discuss Skidmore as
available alternative standard).
42. See Thomas W. Merrill & Kristin E. Hickman, Chevron’s Domain, 89 Geo. L.J. 833,
856–58 (2001) (documenting Court’s recognition of Skidmore’s utility alongside Chevron
beginning in early 1990s).
43. See EEOC v. Arabian Am. Oil Co. (ARAMCO), 499 U.S. 244, 256–58 (1991).
44. Id. at 257 (quoting Gen. Elec. Co. v. Gilbert, 429 U.S. 125, 141 (1976)).
45. See, e.g., Bragdon v. Abbott, 524 U.S. 624, 642 (1998) (noting but avoiding issue
of whether Chevron applied to administrative interpretations of Rehabilitation Act, for
which Congress did not delegate administrative authority to any single agency, and instead
citing Skidmore to justify examination of agencies’ views which “ ‘constitute a body of
experience and informed judgment’” (quoting Skidmore v. Swift & Co., 323 U.S. 134,
139–40 (1944))); Metro. Stevedore Co. v. Rambo, 521 U.S. 121, 136 (1997) (applying
Skidmore’s standard to opinion of the Director of the Office of Workers’ Compensation);
Reno v. Koray, 515 U.S. 50, 61 (1995) (affording Bureau of Prisons internal guidelines
“some deference”). A forthcoming study of Supreme Court deference cases by Eskridge
and Baer identified fifty-four cases between Chevron and Mead in which the Court relied on
the Skidmore review standard. See Eskridge & Baer, supra note 7 (manuscript at 23).
R
46. See ARAMCO, 499 U.S. at 259–60 (Scalia, J., concurring in part and concurring in
the judgment). Justice Scalia cites General Electric Co. v. Gilbert, 429 U.S. 125, rather
than Skidmore, for the deference standard at issue. See ARAMCO, 499 U.S. at 259–60. As
the majority opinion notes, however, the cited passage from Gilbert in turn cites Skidmore;
and the standard as described is that of Skidmore review. See id. at 257 (quoting Gilbert, 429
U.S. at 141–42 (quoting Skidmore, 323 U.S. at 140)).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 11 11-OCT-07 15:18
2007] MODERN SKIDMORE 1245
actions that clearly would not be Chevron-eligible post-Mead.
47
Thus,
Chevron left Skidmore in doctrinal limbo.
48
3. The Mead Counterrevolution and Revitalization of Skidmore. — With
Christensen
49
in 2000 and Mead
50
in 2001, the Court significantly con-
stricted Chevron’s scope. In so doing, these cases reaffirmed Skidmore as
the deference standard for most administrative interpretations. In Chris-
tensen, the Court considered a dispute reminiscent of Skidmore: The
United States as amici urged Chevron deference for a nonbinding Depart-
ment of Labor opinion letter that interpreted the FLSA to bar employers
from mandating employees to take compensatory time instead of over-
time pay.
51
The Court declined to defer and held further that the inter-
pretations expressed in the letter were “‘entitled to respect’ under our
decision in Skidmore v. Swift & Co., but only to the extent that those inter-
pretations ha[d] the ‘power to persuade.’”
52
The Court’s explanation of
its holding on this point was exceptionally brief; the Court stated, in total,
that Chevron is appropriate for those agency interpretations
arrived at after, for example, a formal adjudication or notice-
and-comment rulemaking. Interpretations such as those in
opinion letters—like interpretations contained in policy state-
ments, agency manuals, and enforcement guidelines, all of
which lack the force of law—do not warrant Chevron-style defer-
ence. Instead, interpretations contained in formats such as
opinion letters are [governed by Skidmore].
53
Thus, the Court held that Chevron did not apply to a broad array of ad-
ministrative interpretations that lacked the force of law and resulted from
47. See, e.g., Your Home Visiting Nurse Servs., Inc. v. Shalala, 525 U.S. 449, 452–53
(1999) (evaluating Medicare Provider Reimbursement Manual under Chevron); Smiley v.
Citibank (S.D.), N.A., 517 U.S. 735, 740–43 (1996) (affording Chevron deference to
proposed agency rule adopted in response to litigation); cf. Holly Farms Corp. v. NLRB,
517 U.S. 392, 398–99 (1996) (citing Chevron and using Chevron ’s two-step inquiry, though
not expressly endorsing Chevron’s mandatory deference, in reviewing NLRB adjudication
lacking legal force). Before the Court decided Mead, the circuits disagreed over whether
agencies that lacked rulemaking authority could be eligible for Chevron deference.
Compare, e.g., Atchison, Topeka & Santa Fe Ry. v. Pe˜
na, 44 F.3d 437, 441–42 (7th Cir.
1994) (“[O]nly statutory interpretations by agencies with rulemaking powers deserve
substantial deference [under Chevron].”), with OSG Bulk Ships, Inc. v. United States, 132
F.3d 808, 812 n.7 (D.C. Cir. 1998) (“But where, as here, Congress has not explicitly
delegated rulemaking authority to the agency charged with administering the statute, the
Chevron analysis is the appropriate means by which to evaluate the agency’s interpretation
of the statute.”).
48. For a more comprehensive analysis of the confusion over Chevron’s scope, see
Merrill & Hickman, supra note 42, at 848–52 (2001) (identifying fourteen areas of
R
confusion, including several circuit splits, over Chevron’s scope within lower court
jurisprudence).
49. 529 U.S. 576 (2000).
50. 533 U.S. 218 (2001).
51. See Christensen, 529 U.S. at 586–87.
52. Id. at 587 (citation omitted) (quoting Skidmore v. Swift & Co., 323 U.S. 134, 140
(1944)).
53. Id. (citations omitted).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 12 11-OCT-07 15:18
1246 COLUMBIA LAW REVIEW [Vol. 107:1235
relatively informal procedures—or perhaps those that fulfilled just one of
these criteria. In either case, Skidmore’s domain would be vast.
The following term, in Mead, the Court elaborated Christensen and
further contracted Chevron’s scope in favor of Skidmore. This time, the
majority held that Chevron’s framework was inapplicable to a Customs Of-
fice’s ruling letter that classified Mead’s day planners as “bound” diaries
and therefore subject to tariffs.
54
The Court characterized Skidmore as the
baseline deference standard and Chevron as applying to a subset of “inter-
pretive choices distinguished by an additional reason for judicial defer-
ence.”
55
This “additional reason” that defined Chevron’s scope was the
reason Chevron gave for deference: the existence of Congress’s delega-
tion of legislative authority to the agency.
56
Yet the Mead Court affirmed
Chevron’s principle that some delegations are “ ‘implicit.’”
57
Thus, review-
ing courts must consider all circumstances surrounding the statutory
scheme and agency action to ascertain whether “Congress would expect
the agency to be able to speak with the force of law” on the matter at
hand.
58
As in Christensen, the Mead Court contended that agency posi-
tions reached through relatively formal procedures qualified for Chev-
ron’s approach, since it is more plausible that Congress would expect the
agency’s action to carry the force of law when the agency engages in a
deliberative, interpretive process.
59
Chevron deference was inapposite for
the tariff rulings because the statutory scheme and informality of the let-
ters did not suggest that Congress would have intended the letters to hold
the force of law.
60
Hence, after the Chevron revolution cast doubt on Skidmore’s vitality,
Christensen and Mead confirmed that Skidmore’s standard continues to gov-
ern many if not most administrative interpretations. However, Christensen
and Mead did not resolve all of the questions surrounding the Court’s
deference doctrine.
B. The Deference Framework
The Court’s decisions in Christensen and Mead make clear that the
current regime for judicial review of agency legal interpretations includes
both Chevron and Skidmore as separate standards of review.
61
As outlined
54. Mead, 533 U.S. at 225, 231–34.
55. Id. at 228–29.
56. See id. at 229.
57. See id. (quoting Chevron U.S.A. Inc. v. Natural Res. Def. Council, Inc., 467 U.S.
837, 844 (1984)).
58. Id.
59. See id. at 230; see also Christensen v. Harris County, 529 U.S. 576, 587 (2000);
Cass R. Sunstein, Chevron Step Zero, 92 Va. L. Rev. 187, 224–26 (2006) [hereinafter
Sunstein, Step Zero] (discussing Mead’s emphasis on agency procedure).
60. See Mead, 533 U.S. at 231–34.
61. For a more recent case articulating the role of Chevron and Skidmore in modern
review of agency interpretations, see Gonzales v. Oregon, 546 U.S. 243, 255–56 (2006)
(summarizing Supreme Court’s current deference framework). Though we describe the
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 13 11-OCT-07 15:18
2007] MODERN SKIDMORE 1247
above, the Chevron deference standard contemplates the two steps of eval-
uating whether a statute is ambiguous, and if so, deferring to any reasona-
ble or permissible agency interpretation.
62
The Skidmore deference stan-
dard, by contrast, calls upon reviewing courts to evaluate an
interpretation’s persuasiveness by weighing various factors including the
agency’s thoroughness and consistency.
63
Mead, in turn, articulates its own two-part inquiry for discerning
which of these two standards of review applies in any given case: whether
Congress gave the agency in question the authority to bind regulated par-
ties with “the force of law” and, if so, whether the agency “exercise[d] . . .
that authority.”
64
Some have described Mead’s inquiry as a “step zero” in
the overall analytical framework, coming before the application of either
Chevron’s two steps or Skidmore’s multiple factors.
65
Others view Mead as
“sort of a Chevron step one-and-one-half,” relevant only if the reviewing
court first concludes that the statute’s meaning is ambiguous.
66
Both
conceptualizations are technically correct. Mead’s two steps provide a
threshold inquiry to determine which of two potential evaluative stan-
dards, Chevron or Skidmore, applies to a given case. Yet because a review-
ing court will not defer to an agency under either doctrine if the statute’s
meaning is clear, the Skidmore standard implicitly replicates Chevron’s first
step.
67
Thus, a court can engage in step one analysis before having to use
Mead to make the choice between Chevron and Skidmore.
current framework as consisting of Mead, Chevron, and Skidmore, the forthcoming study by
Eskridge and Baer identifies seven different deference regimes, including Chevron and
Skidmore, within the Court’s jurisprudence since 1984. See Eskridge & Baer, supra note 7
R
(manuscript at 11–12). However, most of the other deference doctrines they identify—
such as Seminole Rock deference for agency interpretations of their own regulations, see
Bowles v. Seminole Rock & Sand Co., 325 U.S. 410, 414 (1945); anti-deference doctrines
such as the rule of lenity; or Curtiss-Wright deference for national security and foreign
affairs matters, see United States v. Curtiss-Wright Export Corp., 299 U.S. 304 (1936)—
represent narrow exceptions or apply in different situations from the norm of the
framework consisting of Mead, Chevron, and Skidmore. See Eskridge & Baer, supra note 7
R
(manuscript at 11–33) (discussing different deference regimes).
62. See Chevron, 467 U.S. at 842–43.
63. Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944). Citing Skidmore, the Mead
Court paraphrased these factors in saying that agency interpretations not entitled to
Chevron deference should be evaluated based upon “the degree of the agency’s care, its
consistency, formality, and relative expertness, and . . . the persuasiveness of the agency’s
position.” Mead, 533 U.S. at 228 (footnotes omitted).
64. Mead, 533 U.S. at 226–27.
65. See, e.g., Merrill & Hickman, supra note 42, at 836; Sunstein, Step Zero, supra
R
note 59, at 191.
R
66. See Joseph Cordaro, Note, Who Defers to Whom? The Attorney General Targets
Oregon’s Death With Dignity Act, 70 Fordham L. Rev. 2477, 2506 (2002); see also, e.g.,
Scott H. Angstreich, Shoring Up Chevron: A Defense of Seminole Rock Deference to Agency
Regulatory Interpretations, 34 U.C. Davis L. Rev. 49, 64 (2000) (describing Christensen’s
force-of-law inquiry similarly as an intermediate step after Chevron step one inquiry).
67. See infra notes 165–173 and accompanying text and Part II.B.3.c (elaborating
R
relationship between Skidmore and Chevron’s first step inquiry).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 14 11-OCT-07 15:18
1248 COLUMBIA LAW REVIEW [Vol. 107:1235
Of course, whether a given statute is clear is often a close call. In
applying the interpretive process of Chevron’s first step, even the most
unbiased judge may find herself preferring a particular view of the stat-
ute. Considering the Mead questions first requires a judge at least to con-
sider what her attitude should be toward an agency’s interpretation
before becoming too committed to her own statutory analysis. Yet the
Mead analysis suffers from its own lack of clarity.
68
Many cases may be
resolved either by deciding that the statute’s meaning is clear or that the
court would or would not defer to the agency’s interpretation under ei-
ther standard.
69
Hence, judges with heavy dockets may reasonably prefer
to assess cases in such terms before entering Mead’s thicket to discern
whether Skidmore or Chevron is applicable.
Regardless of how a reviewing court chooses to order the various in-
quiries, Christensen and Mead present Skidmore deference as distinct from
Chevron deference. Just how different the two doctrines are remains a
matter of some debate.
Justice Breyer has long adopted the view that Chevron and Skidmore
are functionally similar, with Chevron’s emphasis on delegation represent-
ing merely another factor for a reviewing court to evaluate in deciding
whether to defer to an administrative interpretation.
70
Justice Breyer
premises his view in large part on his observations that the courts have
always considered congressional intent in resolving deference questions,
and yet that Chevron’s notion of implicit congressional delegation of law-
making power is mere legal fiction.
71
Justice Breyer demonstrates his ap-
proach most concisely in his opinion for the Court in Barnhart v. Walton,
in which he incorporates Skidmore-like factors into his analysis of whether
Chevron applies, even as he cites Mead in support of his analysis:
[T]he interstitial nature of the legal question, the related exper-
tise of the Agency, the importance of the question to the admin-
istration of the statute, the complexity of that administration,
and the careful consideration the Agency has given the question
over a long period of time all indicate that Chevron provides the
68. See, e.g., William S. Jordan, III, Judicial Review of Informal Statutory
Interpretations: The Answer is Chevron Step Two, Not Christensen or Mead, 54 Admin. L.
Rev. 719, 719 (2002) (describing Mead test as “a cumbersome, unworkable regime under
which courts must draw increasingly fine distinctions using impossibly vague standards”);
Ronald M. Levin, Mead and the Prospective Exercise of Discretion, 54 Admin. L. Rev. 771,
780–806 (2002) [hereinafter Levin, Exercise of Discretion] (analyzing problems with Mead
analysis); Adrian Vermeule, Introduction: Mead in the Trenches, 71 Geo. Wash. L. Rev.
347, 353–58 (2003) (criticizing Mead as overly abstract, “producing a great deal of
confusion and error”).
69. See Herz, Judicial Review, supra note 7, at 144 (highlighting these two possibilities
R
in which Chevron and Skidmore would lead to same outcome).
70. See, e.g., Breyer, supra note 37, at 379–81; see also Sunstein, Step Zero, supra
R
note 59, at 198–202 (describing Justice Breyer’s conception of relationship between
R
Chevron and Skidmore).
71. See Breyer, supra note 37, at 379–81.
R
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 15 11-OCT-07 15:18
2007] MODERN SKIDMORE 1249
appropriate legal lens through which to view the legality of the
Agency interpretation here at issue.
72
Alternatively, one can view Chevron and Skidmore as fundamentally
distinct, arising from different premises and serving different purposes.
73
Chevron relies on an admittedly fictional presumption that Congress
chose an agency rather than the courts to be the primary interpreter of a
given statutory scheme.
74
Thus, Mead limits Chevron’s scope to cases in
which a court affirmatively finds that Congress implicitly delegated pri-
mary interpretive power and that the agency exercised that power in tak-
ing the action in question.
75
Chevron does not require the courts to abdi-
cate their responsibility for interpreting the law altogether;
76
but where it
applies, Chevron deference is mandatory.
77
By contrast, Skidmore merely reflects a policy of judicial prudence.
Unlike Chevron, Skidmore envisions the courts rather than the agencies as
the primary interpreters of statutes. Nevertheless, as the Skidmore Court
acknowledged, courts often lack the resources and expertise to under-
stand and evaluate fully the consequences of complex statutory
schemes.
78
Sometimes agencies are simply better at assessing and apply-
ing alternative statutory interpretations. Thus, unless circumstances oth-
erwise suggest arbitrary or unreasonable agency behavior, reviewing
courts are often wise to defer to an agency’s greater expertise and, some-
times, extensive interpretive efforts. Other evaluative standards such as
hard look review, as well as Skidmore’s emphasis on factors such as thor-
oughness and consistency, allow the courts to guard against arbitrariness
while simultaneously deferring to administrative interpretations.
79
72. 535 U.S. 212, 222 (2002) (citing United States v. Mead Corp., 533 U.S. 218
(2001)); see also Krzalic v. Republic Title Co., 314 F.3d 875, 879 (7th Cir. 2002) (reading
Barnhart v. Walton as merging the Chevron and Skidmore standards).
73. See, e.g., Krzalic, 314 F.3d at 882 (Easterbrook, J., concurring) (rejecting merger
of Chevron and Skidmore standards as inconsistent with Mead).
74. See, e.g., Mead, 533 U.S. at 229–30 & n.11.
75. See id. at 226–27; see also Thomas W. Merrill, The Mead Doctrine: Rules and
Standards, Meta-Rules and Meta-Standards, 54 Admin. L. Rev. 807, 813 (2002) [hereinafter
Merrill, The Mead Doctrine] (describing Mead’s holding).
76. See, e.g., Whitman v. Am. Trucking Ass’n, 531 U.S. 457, 481 (2001) (rejecting
agency interpretation at Chevron step two); FDA v. Brown & Williamson Tobacco Corp.,
529 U.S. 120, 132–33 (2000) (rejecting agency interpretation at Chevron step one); AT&T
Corp. v. Iowa Utils. Bd., 525 U.S. 366, 389–92 (1999) (rejecting agency interpretation at
Chevron step two).
77. See Nat’l Cable & Telecomm. Ass’n v. Brand X Internet Servs., 545 U.S. 967,
982–86 (2005) (requiring courts to defer under Chevron—even when prior judicial
construction of ambiguous statute differs from agency’s subsequent interpretation).
78. See Skidmore v. Swift & Co., 323 U.S. 134, 139–40 (1944).
79. Where Skidmore or even Chevron deference applies to an agency’s legal
interpretation, the courts still evaluate agency action for adequacy of process under the
arbitrary and capricious review standard of the Administrative Procedure Act, 5 U.S.C.
§ 706(2)(A) (2000), also known as “hard look review.”
[T]he agency must examine the relevant data and articulate a satisfactory
explanation for its action including a “rational connection between the facts
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 16 11-OCT-07 15:18
1250 COLUMBIA LAW REVIEW [Vol. 107:1235
However one construes the relationship between Chevron and Skid-
more deference, a few things are certain. Post-Mead, the Court clearly
considers Skidmore to be an important component of its deference doc-
trine. Also, Skidmore is less deferential than Chevron. What remains un-
clear, at least from the Supreme Court’s opinions, is precisely how much
less deferential Skidmore is and in what way this is so.
II. W
HAT
I
S
S
KIDMORE
D
EFERENCE
?
Drawing fine distinctions among deference standards may seem a
purely academic exercise. Legal realists contend that such an effort is
pointless, as courts only invoke deference standards to justify their pre-
ferred outcome.
80
Although we acknowledge that this critique may be
true in some instances, we nevertheless submit to the contrary that defer-
ence standards matter. We accept that courts feel constrained by defer-
ence standards and speak sincerely when they discuss the application of
those standards.
It is easy enough to recognize the consensus view that Skidmore gives
judges more discretion than Chevron’s command of mandatory defer-
ence. Similarly, from the Court’s articulation of the two standards, one
can readily discern that Chevron deference involves two binary inquiries,
while Skidmore requires courts to evaluate several factors. Nevertheless,
once a reviewing court finds itself in Skidmore’s realm of discretionary def-
erence, elucidating the appropriate degree of deference is not so simple
as plotting a point on a line. Standards of review are not precision instru-
ments. Rather, to paraphrase Justice Frankfurter, standards of review are
more accurately described in terms of the “mood” a reviewing court
should possess in evaluating the issue at bar.
81
The question to be an-
swered, therefore, is what sort of mood Skidmore analysis contemplates.
found and the choice made.” In reviewing that explanation, we must “consider
whether the decision was based on a consideration of the relevant factors and
whether there has been a clear error of judgment.” Normally, an agency rule
would be arbitrary and capricious if the agency has relied on factors which
Congress has not intended it to consider, entirely failed to consider an important
aspect of the problem, offered an explanation for its decision that runs counter to
the evidence before the agency, or is so implausible that it could not be ascribed
to a difference in view or the product of agency expertise. The reviewing court
should not attempt itself to make up for such deficiencies; we may not supply a
reasoned basis for the agency’s action that the agency itself has not given.
Motor Vehicle Mfrs. Ass’n of the U.S. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43
(1983) (citations omitted); see also Brand X, 545 U.S. at 981–82 (discussing relationship
between Chevron and hard look review); Martin v. Occupational Safety & Health Review
Comm’n, 499 U.S. 144, 157–58 (1991) (noting applicability of both Skidmore and hard look
review); Gary Lawson, Outcome, Procedure and Process: Agency Duties of Explanation for
Legal Conclusions, 48 Rutgers L. Rev. 313, 325–31 (1996) (contrasting Chevron analysis
with hard look review).
80. See Colin S. Diver, Statutory Interpretation in the Administrative State, 133 U. Pa.
L. Rev. 549, 564 (1985) (acknowledging and dismissing legal realist critique).
81. Universal Camera Corp. v. NLRB, 340 U.S. 474, 487 (1951).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 17 11-OCT-07 15:18
2007] MODERN SKIDMORE 1251
A. Competing Conceptions of Skidmore Review
Our examination of how Skidmore functions begins with an old de-
bate. Pre-Mead, some suggested that Skidmore actually requires no defer-
ence at all, but instead prescribes nothing more than independent judg-
ment by the reviewing court. Others described Skidmore as a type of
deference that varies in extent from case to case along a sliding scale.
The following sections explain these competing conceptions of Skidmore.
1. Distinguishing Independent Judgment from Judicial Deference. — Many
scholars have described standards of review as falling along a spectrum
with independent judgment and deference at opposite poles.
82
But inde-
pendent judgment and deference differ not only in degree but also in
kind.
The key difference between independent judgment and deference is
whether a court is restrained to give an agency’s interpretation special
consideration that the court need not give to other litigants.
83
A court
applying deference must at least consider whether to give weight to the
agency’s point of view, even if it is not required to give such weight. Defer-
ence to an administrative interpretation is triggered by the interpreta-
tion’s “pedigree”—i.e., the fact that an agency holds the view.
84
In con-
trast, a court exercising independent judgment is free to consider the
merits of the agency’s interpretation alone, or even to ignore the
agency’s interpretation altogether.
85
For a court exercising independent
judgment, “the pedigree of an interpretation—that is, the identity of its
sponsor or author”—has no impact on the court’s decision.
86
In inde-
pendent judgment mode, a court may still examine the agency’s view of
the statute, but the court considers it on “an equal a priori footing” with
all other arguments advanced by litigants in the case.
87
Once a court finds itself in a deferential mode, the question of how
much deference to afford the agency interpretation arises, since defer-
ence may occur in varying degrees.
88
The bare synopsis of Chevron and
82. See, e.g., Cynthia R. Farina, Statutory Interpretation and the Balance of Power in
the Administrative State, 89 Colum. L. Rev. 452, 453–54 & n.10 (1989); Thomas W. Merrill,
Judicial Deference to Executive Precedent, 101 Yale L.J. 969, 971 & n.6 (1992) [hereinafter
Merrill, Judicial Deference]; Henry P. Monaghan, Marbury and the Administrative State, 83
Colum. L. Rev. 1, 6–7 (1983). See generally Diver, supra note 80, at 552–67 (1985)
R
(providing useful overview of courts’ modes of review).
83. See Merrill & Hickman, supra note 42, at 855 (“Skidmore is properly regarded as a
R
deference doctrine because the court cannot ignore the agency interpretation—the court
must assess that interpretation against multiple factors and determine what weight they
should be given.”).
84. See Diver, supra note 80, at 559.
R
85. See id.
86. Id. (emphasis omitted).
87. See id.
88. See, e.g., Anthony, Agency Interpretations, supra note 38, at 6–7 (describing
R
courts’ deliberations regarding “degrees of ‘deference’ or ‘weight’ to be granted the
agency interpretations” (quoting Bureau of Alcohol, Tobacco & Firearms v. Fed. Labor
Relations Auth., 464 U.S. 89, 98 n.8 (1983); Zuber v. Allen, 396 U.S. 168, 192 (1969)));
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 18 11-OCT-07 15:18
1252 COLUMBIA LAW REVIEW [Vol. 107:1235
Skidmore deference offered above touches upon this notion of degrees of
deference. In general, scholars agree that Chevron’s step two nears the
fully deferential end of the spectrum: Courts employing this standard
retain little discretion and are required to defer to the agency’s view un-
less it is unreasonable.
89
Thus, it is unsurprising that most agency inter-
pretations survive Chevron’s second step.
90
Commentators also generally
agree that Skidmore is less deferential than Chevron, falling somewhere fur-
ther away from the deference pole.
91
This is all well and good, but it
offers little guidance for the application of Skidmore as a stand-alone
doctrine.
2. The Independent Judgment Model of Skidmore Review. — One might
conceptualize Skidmore as directing courts to engage in independent judg-
ment when reviewing administrative interpretations. Most closely associ-
ated with work by Colin Diver before the Chevron doctrine seriously took
hold, this conception of Skidmore does not require courts to consider giv-
ing weight to an agency’s view on the basis of contextual factors, but in-
stead leaves courts free to do what they will in evaluating an administra-
tive interpretation. Justice Jackson in Skidmore concluded that the
“weight” the administrative interpretation should receive depends upon
its “power to persuade.”
92
As Diver suggested, such deference really rep-
resents no deference at all: “Of course, the ‘weight’ assigned to any advo-
cate’s position is presumably dependent upon the ‘thoroughness evident
in its consideration’ and the ‘validity of its reasoning.’”
93
The independent judgment model of Skidmore deference thus under-
stands the “persuasiveness” of an administrative interpretation to depend
ultimately on the interpretation’s merits or rightness.
94
This conception
discounts Skidmore’s contextual factors and does not require courts to re-
Diver, supra note 80, at 564–67 (discussing different kinds of deference that courts may
R
allocate to administrative interpretations); Merrill, Judicial Deference, supra note 82, at
R
972 (noting that, prior to Chevron, Supreme Court recognized deference as “exist[ing]
along a sliding scale”).
89. See, e.g., Rossi, supra note 7, at 1112–14.
R
90. A study of all federal courts of appeals’ applications of Chevron over a two-year
period found that at Chevron step two, courts upheld the administrative interpretation
eighty-nine percent of the time. Orin S. Kerr, Shedding Light on Chevron: An Empirical
Study of the Chevron Doctrine in the U.S. Courts of Appeals, 15 Yale J. on Reg. 1, 30–31
(1998).
91. See, e.g., Murphy, Counter-Marbury, supra note 7, at 46 (noting that “Skidmore
R
deference is considerably weaker than the strong stuff of Chevron”); Rossi, supra note 7, at
R
1109 (noting that Skidmore is often called “weak deference”); Wildermuth, supra note 7, at
R
1898–99.
92. See Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944).
93. Diver, supra note 80, at 565 (quoting Skidmore, 323 U.S. at 140).
R
94. See id. at 559, 565 (describing independent mode of statutory interpretation as
evaluation only of whether agency’s interpretation is “more plausible or convincing than”
others); Murphy, Judicial Deference, supra note 40, at 1015 (explaining that Skidmore
R
“basically instructs courts to exercise independent judgment regarding statutory meaning
subject to the weak requirement that they carefully consider agency views for
persuasiveness”).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 19 11-OCT-07 15:18
2007] MODERN SKIDMORE 1253
gard the presence or absence of those factors as particularly relevant. At
most, this view understands Skidmore to require “due regard” be given to
the agency’s view, while “instruct[ing] courts to adopt the statutory inter-
pretations that they themselves deem best.”
95
In effect, then, Skidmore
directs courts to treat the agency’s view just as it would the view of any
litigant.
96
The independent judgment conception of Skidmore finds support in
more recent case law. Most notably, the majority opinion in Christensen v.
Harris County—a case which helped revitalize Skidmore—applies Skidmore
in this fashion.
97
As discussed above,
98
in Christensen, the Court encoun-
tered a Department of Labor opinion letter interpreting the FLSA as pre-
cluding employers from mandating compensatory time.
99
The Court first
engaged in independent review of the statute and determined that the
best interpretation of the statute permitted mandatory compensatory
time.
100
Only then did the Court address the contrary opinion letter,
explaining that under Skidmore, the Court owed the letter respect “only to
the extent that those interpretations have ‘the power to persuade.’”
101
The Court ignored Skidmore’s contextual factors and dismissed the opin-
ion letter, declaring simply that it was “unpersuasive” in comparison to
the Court’s preferred interpretation.
102
Christensen’s application of
Skidmore demonstrates how the independent judgment approach does
not ask a court to assess the proper weight to give the agency’s interpreta-
tion on the basis of contextual factors; instead, this approach permits a
court to enforce its preferred interpretation.
103
The Christensen majority’s approach to Skidmore analysis is not unu-
sual, particularly not within the Supreme Court’s jurisprudence. Since
deciding Christensen and Mead, the Court has had several opportunities to
apply Skidmore deference. In most of those cases, the Court’s analysis
closely resembles the above description of the Christensen opinion.
104
In
Raymond B. Yates, M.D., P.C. Profit Sharing Plan v. Hendon, for example,
95. Murphy, Judicial Deference, supra note 40, at 1015 n.13.
R
96. Cf. Diver, supra note 80, at 565. Diver labels Skidmore an extremely low level of
R
deference. See id. However, his view of Skidmore is similar to the independent judgment
conception, since he concludes that under Skidmore, the interpretation’s “pedigree adds
nothing” to the court’s consideration, and that courts treat the administrative view on par
with any litigant’s position. See id.
97. See 529 U.S. 576, 587–88 (2000); see also Rossi, supra note 7, at 1131–34
R
(analyzing Christensen’s majority opinion as independent judgment approach to Skidmore).
98. See supra notes 51–53 and accompanying text.
R
99. See 529 U.S. at 586.
100. See id. at 582–86.
101. See id. at 587.
102. See id.
103. Cf. Rossi, supra note 7, at 1127 (describing Christensen majority’s approach as
R
essentially insisting upon what it believed to be the “better” interpretation).
104. See Gonzales v. Oregon, 546 U.S. 243, 268–69 (2006); Raymond B. Yates, M.D.,
P.C. Profit Sharing Plan v. Hendon, 541 U.S. 1, 17–18 (2004); Clackamas Gastroenterology
Assocs. v. Wells, 538 U.S. 440, 448–50 (2003); Wash. State Dep’t of Soc. & Health Servs. v.
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 20 11-OCT-07 15:18
1254 COLUMBIA LAW REVIEW [Vol. 107:1235
the Court analyzed the relevant statutory text, history, and purpose to
reach a conclusion about the statute’s meaning.
105
The Court then “fi-
nally” noted a Department of Labor advisory opinion’s concurrence with
the Court’s interpretation and described the agency’s interpretation as
reflecting “‘a body of experience and informed judgment to which courts
and litigants may properly resort for guidance.’”
106
The Court cited
Skidmore for that proposition but never discussed any of the factors ordi-
narily identified as relevant to Skidmore analysis. Similarly, in Clackamas
Gastroenterology Associates v. Wells, the Court compared its own precedents
with the interpretations the parties suggested to conclude that its own
view of the statute was best.
107
Only after reaching that conclusion did
the Court note that an EEOC amicus brief and guidelines adopted a simi-
lar position, and that it was “persuaded” by the EEOC’s approach.
108
Again, the Court cited Skidmore but failed to mention any of the Skidmore
factors.
The independent judgment conception of Skidmore should be distin-
guished from a court’s finding that a statute’s meaning is unambiguous,
as in Chevron step one.
109
The two modes of review seem similar on their
surface: In both, the reviewing court independently finds and enforces
what it believes to be the statute’s meaning.
110
These modes of review
differ, however, by the court’s ability to exercise discretion. When a stat-
ute’s meaning is clear, the court has no discretion and the administrative
interpretation can be of no import—regardless of whether the interpreta-
tion concurs or differs from the statute, the court must enforce the con-
gressional intent plainly embodied in the statute.
111
In such cases, it mat-
ters not whether the court applies Skidmore, Chevron, or some other
standard, because deference is not available. By contrast, when a court
applies Skidmore, the statute is typically ambiguous, or at least not so clear
as to prevent the court’s exercise of discretion. In Christensen, the Court
observed the statute’s total silence on the relevant question.
112
In Yates,
the Court described the statute’s definitions of the terms in question as
Guardianship Estate of Keffeler, 537 U.S. 371, 385–86 (2003); Nat’l R.R. Passenger Corp. v.
Morgan, 536 U.S. 101, 111 (2002).
105. Yates, 541 U.S. at 12–17.
106. Id. at 17–18 (quoting Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944)).
107. Clackamas, 538 U.S. at 448.
108. Id. at 448–49.
109. Cf. Rossi, supra note 7, at 1131–32 (comparing independent judgment model of
R
Skidmore to Packard Motor Car Co. v. NLRB, 330 U.S. 485 (1947), even though Packard
Court confronted an “unambiguous” statute).
110. Compare Packard, 330 U.S. at 492–93 (upholding administrative determination
that aligned with statute found to have “no ambiguity” without reference to agency’s
views), with Christensen v. Harris County, 529 U.S. 576, 587 (2000) (rejecting
administrative interpretation after determining best view of silent statute and dismissing
agency’s view as “unpersuasive”).
111. See Chevron U.S.A. Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837, 842–43
(1984).
112. See 529 U.S. at 582, 588.
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 21 11-OCT-07 15:18
2007] MODERN SKIDMORE 1255
“uninformative.”
113
In Clackamas, the Court noted the total inadequacy
of the relevant statutory definition and turned to its own precedents to
“fill the gap in the statutory text.”
114
In short, cases in which the reviewing court employs the indepen-
dent judgment model of Skidmore review are distinguishable from those
in which a court finds the statute to have a clear meaning. Independent
judgment cases purport to follow Skidmore deference but do so in a man-
ner that is not at all deferential, instead imposing the courts’ determina-
tion of the best, better, or preferred interpretation of the statute in
question.
3. Skidmore’s Review as Deference Varying Along a Sliding Scale. — A
competing conception characterizes Skidmore as prescribing deference
along a continuum or sliding scale, with the degree of deference varying
according to the reviewing court’s evaluation of Skidmore’s contextual fac-
tors.
115
Most closely associated with Kenneth Culp Davis and Thomas
Merrill,
116
this view of Skidmore represents a type of “deference” because a
court is not free to ignore the administrative interpretation or to reject it
solely because it differs from the court’s preferred interpretation.
117
In-
stead, Skidmore review intrudes upon courts’ judgment by requiring courts
to apply multiple factors to the agency’s interpretation to decide how
much weight to assign to the interpretation.
118
In effect, under this
model, Skidmore prescribes a method by which a reviewing court should
113. Raymond B. Yates, M.D., P.C. Profit Sharing Plan v. Hendon, 541 U.S. 1, 12
(2004).
114. Clackamas Gastroenterology Assocs. v. Wells, 538 U.S. 440, 444 (2003).
115. See, e.g., Merrill & Hickman, supra note 42, at 855.
R
116. Davis did not use the sliding-scale descriptor but consistently described pre-
Chevron deference in similar terms. See, e.g., 5 Kenneth Culp Davis, Administrative Law
Treatise § 29:16, at 400 (2d ed. 1984) [hereinafter 5 Davis] (describing pre-Chevron
deference as “variable; it can be stronger or weaker”); Kenneth Culp Davis, Administrative
Rules—Interpretative, Legislative, and Retroactive, 57 Yale L.J. 919, 934 (1948)
(“Legislative rules normally have greater authoritative weight than interpretative rules, but
the authoritative weight of interpretative rules varies considerably.”). Merrill regularly
utilizes the sliding-scale phraseology to describe Skidmore deference. See, e.g., Merrill,
Judicial Deference, supra note 82, at 972 (describing pre-Chevron deference as sliding
R
scale, “from ‘great’ to ‘some’ to ‘little’” (citing 5 Davis, supra, § 29:16, at 400)); Merrill,
The Mead Doctrine, supra note 75, at 810 (quoting Justice Scalia’s use of the sliding-scale
R
term in United States v. Mead Corp., 533 U.S. 218, 250 (2001) (Scalia, J., dissenting));
Merrill & Hickman, supra note 42, at 855 (describing Skidmore as sliding scale in which
R
“agency interpretations receive various degrees of deference, ranging from none, to slight,
to great, depending on the court’s assessment of the strength of the agency interpretation
under consideration”).
117. Murphy, Counter-Marbury, supra note 7, at 46 (“[Skidmore] does intrude on
R
judicial independence by requiring courts to give serious consideration to agency views. . . .
[A] court is free to reject those with which it disagrees after fair consideration; but it is
equally true that courts are not free to ignore them.”).
118. See Merrill & Hickman, supra note 42, at 855 (describing sliding-scale
R
conception of Skidmore); Rossi, supra note 7, at 1134–37 (same).
R
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 22 11-OCT-07 15:18
1256 COLUMBIA LAW REVIEW [Vol. 107:1235
determine how much deference to give an agency’s interpretation but
does not mandate the outcome of that determination.
The sliding-scale model of Skidmore counsels special consideration of
agency interpretations that courts do not necessarily afford to the views of
other litigants. Skidmore justified giving agency interpretations such
“weight” on two grounds. First, agencies typically hold specialized exper-
tise and experience related to their respective regulatory schemes.
119
In
light of this reality, Skidmore directs courts to assess to what extent the
interpretation reflects an exercise of the agency’s potentially superior in-
terpretive competency.
120
Second, courts can promote uniformity of the
law and thereby promote the public good by harmonizing judicial inter-
pretations with administrative interpretations.
121
The sliding-scale conception of Skidmore can be found in some of the
Court’s recent decisions, most notably in the Court’s recent reaffirmation
of Skidmore in Mead.
122
As discussed above, the Mead Court held that
Skidmore’s standard (rather than Chevron’s) governed a Customs tariff
classification letter.
123
The Court outlined how lower courts should re-
view such administrative opinions: “The fair measure of deference to an
agency administering its own statute has been understood to vary with
circumstances . . . . The approach has produced a spectrum of judicial
responses, from great respect at one end, to near indifference at the
other.”
124
In this and other sections of Mead, the Court affirmed that
lower courts have latitude to decide how much deference to give any par-
ticular administrative interpretation.
125
Although most of the Court’s post-Mead applications of Skidmore re-
view reflect the independent judgment model described above, the Court
has also employed the sliding-scale model since deciding Mead. In Alaska
Department of Environmental Conservation v. EPA, the Court was called upon
to evaluate an interpretation of the Clean Air Act advanced in interpre-
tive guidelines published in 1983, 1988, 1993, and 1998.
126
Acknowledg-
ing that such guidelines lack the force of law and thus are ineligible for
Chevron deference, the Court nevertheless counseled deference under
119. See Skidmore v. Swift & Co., 323 U.S. 134, 139–40 (1944) (noting that agency’s
policies were “based upon more specialized experience and broader investigations and
information than is likely to come to a judge in a particular case,” and agency’s opinions
“constitute a body of experience and informed judgment to which courts and litigants may
properly resort for guidance”).
120. See Rossi, supra note 7, at 1136–37 (discussing Skidmore’s rationale).
R
121. See Skidmore, 323 U.S. at 139–40.
122. See United States v. Mead Corp., 533 U.S. 218, 228–29 (2001). For other
examples of the sliding-scale approach to Skidmore, see EEOC v. Arabian Am. Oil Co.
(ARAMCO), 499 U.S. 244, 256–58 (1991) and Gen. Elec. Co. v. Gilbert, 429 U.S. 125,
141–43 (1976).
123. See Mead, 533 U.S. at 231–32.
124. Id. at 228 (footnotes and citations omitted).
125. For example, the Court explained that Skidmore stood for the proposition “that
an agency’s interpretation may merit some deference whatever its form.” Id. at 234.
126. 540 U.S. 461, 487–88 (2004).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 23 11-OCT-07 15:18
2007] MODERN SKIDMORE 1257
Skidmore.
127
Rather than interpreting the statute for itself, the Court ex-
amined the agency’s interpretation in conjunction with the statute’s lan-
guage and history and found the agency’s approach to be reasonable.
The Court also emphasized the agency’s expertise and the longstanding
duration of its interpretation.
128
Even if one assumes that the sliding-scale model of Skidmore is the
correct one, it is not altogether clear exactly how the sliding scale oper-
ates. The Court has not offered firm rules, either in Skidmore or else-
where, for how courts should calibrate their level of deference along the
sliding scale. For example, the Court’s conclusion in Alaska Department of
Environmental Conservation that the agency’s interpretation was “reasona-
ble” and not “impermissible” prompted Justice Kennedy to accuse the
majority of applying Chevron-style analysis under the Skidmore label.
129
Justice Kennedy’s accusation raises this question: Does Skidmore defer-
ence allow a court to defer to an interpretation that it considers merely
reasonable but not the only or even best option, given the presence of
other contextual factors?
Moreover, the Court has not precisely delineated which contextual
factors the courts should evaluate in applying the sliding scale. Neither
Skidmore nor Mead purports to provide a conclusive list of factors. Both
invite courts to consider any fact speaking to an interpretation’s persua-
siveness.
130
And neither Skidmore nor Mead explain how these factors re-
late to each other or whether certain factors are more important than
others.
131
Prior to Chevron, the courts relied upon a host of factors to deter-
mine the appropriate level of deference owed to an agency’s interpreta-
tion in any given case.
132
Thomas Merrill usefully groups these “pre-
Chevron deference factors” into three categories.
133
The first category of
factors appraises whether Congress intended courts to defer to the
agency’s construction, with a focus on the distinction between “legislative
rules” and “interpretative rules.”
134
The second category of factors fo-
127. See id.
128. See id. at 487.
129. See id. at 517–18 (Kennedy, J., dissenting).
130. See Mead, 533 U.S. at 235 (permitting courts to consider “any other sources of
weight”); Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944) (directing courts to consider
“all those factors which give [the agency view] power to persuade”).
131. See Mead, 533 U.S. at 228–29, 234–35; Skidmore, 323 U.S. at 140.
132. See Anthony, Agency Interpretations, supra note 38, at 14–15 (noting that pre-
R
Chevron opinions did not explain “which ‘factors’ were to be heeded, and how they were to
be used”); David R. Woodward & Ronald M. Levin, In Defense of Deference: Judicial
Review of Agency Action, 31 Admin. L. Rev. 329, 332–35 (1979).
133. See Merrill, Judicial Deference, supra note 82, at 973–75. For other helpful
R
collections of these factors, see Diver, supra note 80, at 562 n.95; Woodward & Levin, supra
R
note 132, at 332–35.
R
134. Merrill, Judicial Deference, supra note 82, at 973. Here courts examined the
R
underlying statutory scheme to determine whether Congress delegated lawmaking
authority to the agency, such that Congress intended the agency, rather than courts, to set
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 24 11-OCT-07 15:18
1258 COLUMBIA LAW REVIEW [Vol. 107:1235
cuses on the context of the particular agency interpretation, such as
whether it embodied a longstanding and consistent agency position; the
extent to which the agency drew upon specialized expertise; how thor-
oughly the agency considered the interpretation; and whether the agency
supported the interpretation with a well-reasoned explanation.
135
A final
category of factors potentially shows that the agency’s view aligned with
Congress’s preference on the substantive question at hand.
136
In this
vein, courts favor an agency interpretation that was adopted contempora-
neously to the enactment of the underlying statute
137
or an interpreta-
tion of a statute that Congress reenacted while aware of the agency’s
interpretation.
138
The relevance of all of these factors to modern Skidmore analysis is
unclear. For example, while Mead relies on congressional intent to deter-
mine the applicability of Chevron as opposed to Skidmore review, contem-
porary Court opinions do not incorporate that factor again into Skidmore
analysis. Similarly, neither Mead nor Christensen mentions longevity or
contemporaneity as a component of Skidmore analysis; yet a few post-
Chevron opinions by the Court mention such considerations in conjunc-
tion with Skidmore analysis.
139
Nevertheless, the Court has identified certain factors as particularly
informative. Since Mead, the Court has often reiterated the factors articu-
lated in the Skidmore opinion itself: “the thoroughness evident in [the
interpretation’s] consideration, the validity of its reasoning, [and] its con-
sistency with earlier and later pronouncements.”
140
The Skidmore Court’s
repeated emphasis on the agency’s expertise renders that an important
factor as well.
141
In Mead, the Court lists a slightly different set of factors:
“the degree of the agency’s care, its consistency, formality, and relative
expertness, and . . . the persuasiveness of the agency’s position.”
142
Else-
where, the Mead Court produces yet another list: “thoroughness, logic,
the law. Courts tended to give greater deference to “legislative rules”—those adopted
pursuant to a specific lawmaking delegation—and lesser deference to “interpretive
rules”—those adopted without such delegation. See, e.g., Batterton v. Francis, 432 U.S.
416, 425–26 (1977).
135. Merrill, Judicial Deference, supra note 82, at 973–74.
R
136. Id. at 974.
137. See Woodward & Levin, supra note 132, at 333 (“Special weight is given to a
R
construction which the agency has followed since its governing statute was adopted,
especially if the agency participated in the drafting of the legislation.” (footnote omitted)).
138. See, e.g., Red Lion Broad. Co. v. FCC, 395 U.S. 367, 380–81 (1969); Woodward &
Levin, supra note 132, at 334.
R
139. See, e.g., Alaska Dep’t of Envtl. Conservation v. EPA, 540 U.S. 461, 487 (2004).
140. Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944); see Gonzales v. Oregon, 546
U.S. 243, 268–69 (2006); Edelman v. Lynchburg Coll., 535 U.S. 106, 122–23 (2002)
(O’Connor, J., concurring); Wis. Dep’t of Health & Family Servs. v. Blumer, 534 U.S. 473,
505 (2002).
141. See Skidmore, 323 U.S. at 139–40.
142. United States v. Mead Corp., 533 U.S. 218, 228 (2001) (footnotes omitted).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 25 11-OCT-07 15:18
2007] MODERN SKIDMORE 1259
and expertness, [and] its fit with prior interpretations.”
143
These articu-
lations can be distilled to five key factors in modern Skidmore analysis:
thoroughness, formality, validity, consistency, and agency expertise. As
noted above, longevity and contemporaneity may together comprise a
sixth key factor.
To summarize, Skidmore’s standard for reviewing administrative inter-
pretations has ascended in importance in the wake of Christensen and
Mead. An incredible diversity of administrative interpretations now falls
under Skidmore’s scope. Yet, it is not clear whether Skidmore is best under-
stood as allowing reviewing courts to exercise independent judgment or
requiring them to apply deference along a sliding scale based on an anal-
ysis of contextual factors. Another way of thinking about this question is
to ask whether the Skidmore doctrine is really deferential at all, or whether
the courts merely employ the doctrine to reinforce their own indepen-
dent judgment. Moreover, to the extent that a court applies the sliding-
scale approach, the relevant contextual factors and mechanics for their
application remain less than clear. Part II.B explains how we undertook
to study which of these conceptions of Skidmore deference the federal
courts of appeals actually employ as well as how they employ them.
B. Skidmore Deference in Practice: An Empirical Study
The goals of this study were twofold. First, we sought to determine
the extent to which the appellate courts apply Skidmore using the inde-
pendent judgment model versus the sliding-scale model. Second, to the
extent that the appellate courts use the sliding-scale approach, we wanted
to evaluate how the sliding scale operates. This section explains the set of
cases considered in this study and then lays out the tests we applied to the
cases and factors we tracked within that set of cases. Finally, this section
details our findings. Specifically, we conclude that the sliding-scale
model of Skidmore deference dominates in practice. Relatedly, we find
that, while Skidmore is indeed less deferential than Chevron, Skidmore nev-
ertheless represents a highly deferential standard of judicial review.
Because the sliding-scale model of Skidmore review prevails in most
day-to-day judicial decisionmaking, Skidmore’s contextual factors and the
analysis they represent are all the more important. While the courts’ ap-
plication of these factors does not lend itself particularly well to empirical
study, in the course of reviewing 106 identified Skidmore applications, we
observed certain patterns in the courts’ consideration of those factors. In
this section, we also discuss these observations.
1. Identifying Skidmore Applications. — To evaluate the courts’ con-
temporary application of Skidmore, we turned to opinions issued by the
federal courts of appeals during the five-year period following the Court’s
143. Id. at 235.
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 26 11-OCT-07 15:18
1260 COLUMBIA LAW REVIEW [Vol. 107:1235
opinion in Mead on June 18, 2001.
144
Recognizing that the courts occa-
sionally cite Mead or Christensen rather than Skidmore when applying the
Skidmore standard, for thoroughness we collected federal appellate court
opinions that cited any of the three.
145
Structuring our study around judicial citations to these three cases
presented certain limitations. We did not review cases that cited Chevron
alone.
146
We also did not seek out cases in which courts employed a
Skidmore-like review without citing any of Skidmore, Christensen, or Mead.
For example, courts sometimes cite other cases that applied Skidmore’s
standard, such as circuit precedent, to support their own Skidmore analy-
sis.
147
We see little reason to think that these applications of Skidmore
analysis would differ substantially from cases in which the reviewing court
does cite Skidmore, Christensen, or Mead. Further, courts often speak
loosely of deference, agency expertise, and the like while citing as prece-
dents other cases with similarly loose deference rhetoric, making it diffi-
cult if not impossible to discern precisely which standard of review the
court means to invoke.
148
Such cases, while perhaps related to Skidmore’s
144. We limited the study to federal courts of appeals because of the role that those
courts play in reviewing administrative interpretations. Because the Supreme Court hears
so few cases, the federal courts of appeals usually give the final word on administrative
interpretations. Although the federal district courts also apply Skidmore routinely, by virtue
of the judiciary’s structure, the appellate courts set the pattern for how district courts apply
Skidmore. Finally, the federal courts of appeals often give the only review of administrative
interpretations, since some statutes permit appeals from administrative interpretations to
go straight to a federal appellate court. See William F. Fox, Jr., Understanding
Administrative Law § 62[A] (3d ed. 1997).
145. Unsurprisingly, the vast majority of cases that cited only Mead or Christensen and
not Skidmore did not include an application of Skidmore . Most cited Mead or Christensen for
a different reason, such as to justify the court’s application of Chevron to an administrative
interpretation. See, e.g., Metrophones Telecomms., Inc. v. Global Crossing Telecomms.,
Inc., 423 F.3d 1056, 1064–66 (9th Cir. 2005) (holding that Chevron applies to FCC
interpretation). Nonetheless, this search yielded twenty-two cases that applied Skidmore’s
test after citing Mead or Christensen alone. See infra Appendix.
146. A Westlaw search indicates that the federal courts of appeals issued opinions
fitting this description in 832 cases during the five-year period studied.
147. See, e.g., Morenz v. Wilson-Coker, 415 F.3d 230, 235 (2d Cir. 2005) (failing to
cite Skidmore, Christensen, or Mead, and instead relying on Community Health Center v. Wilson-
Coker—which in turn discusses Skidmore ’s factors and cites Mead—in concluding that
agency’s interpretations “warrant ‘respectful consideration’” (quoting Cmty. Health Ctr. v.
Wilson-Coker, 311 F.3d 132, 138 (2d Cir. 2002))).
148. In Watters v. Wachovia Bank, N.A., for example, Justice Stevens, writing in dissent,
rejected Chevron as the standard for reviewing an agency regulation preempting state law.
See 127 S. Ct. 1559, 1584 (2007) (Stevens, J., dissenting). Instead, noting the agency’s
expertise and the statute’s complexity, Justice Stevens advocated giving the regulation, or
at least the agency’s interpretation thereof, “some weight.” See id. Justice Stevens cited
Geier v. Am. Honda Motor Co., 529 U.S. 861, 883 (2000), and Medtronic, Inc. v. Lohr, 518
U.S. 470, 512 (1996) (O’Connor, J., concurring in part and dissenting in part), rather than
Skidmore as support for that proposition. See Watters, 127 S. Ct. at 1584. Geier in turn cites
several cases, including Medtronic, 518 U.S. at 496, 506 (Breyer, J., concurring in part and
concurring in judgment), and Auer v. Robbins, 519 U.S. 452, 461–62 (1997), in deferring
to an agency’s interpretation of its own regulations. See Geier, 529 U.S. at 883–84. Justice
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 27 11-OCT-07 15:18
2007] MODERN SKIDMORE 1261
standard, are simply too vague and indefinite to illuminate any court’s
understanding and application of Skidmore as such.
Also, because we relied on judicial citations to Skidmore, Christensen,
and Mead, we did not evaluate cases in which courts exercised indepen-
dent judgment without reference to deference principles even though
the Skidmore standard arguably applied. As noted above, our goals with
this study emphasize how federal appellate courts purporting to apply
Skidmore did so, rather than when such courts deemed Skidmore appropri-
ate or whether they utilized Skidmore in all cases where we think they
should. Because we regard applying independent judgment in lieu of
deference as distinguishable from employing the independent judgment
model of the Skidmore standard, opinions that fail to mention Skidmore at
all are unlikely to provide much insight. Thus, when this study discusses
the independent judgment conception of Skidmore, it refers to instances
in which a court cites Skidmore, Mead, and/or Christensen yet treats those
cases as prescribing independent judgment.
We thus began our inquiry with a total population of 450 federal
appellate court opinions that cited Skidmore, Christensen, or Mead. From
the initial group of 450 cases, we attempted to identify instances in which
courts applied Skidmore to review an administrative interpretation. In de-
fining a Skidmore application, we prioritized two criteria. First, the defini-
tion needed to be as objective as possible to limit our own biases in apply-
ing the definition. Second, the definition needed to exclude cases in
which the court mentioned the Skidmore standard but did not actually
apply it.
149
With these goals in mind, we considered a case to involve a
Skidmore application if it exhibited the following four elements:
(1) The case addressed a federal administrative agency’s interpreta-
tion of a statute in the course of resolving a dispute;
(2) The majority opinion cited Skidmore or the analogous parts of
Mead or Christensen for its standard of review of the administrative
interpretation;
(3) The majority opinion did not find the underlying statute that the
agency interpreted to be plain, clear, or unambiguous; and
(4) The majority opinion accepted or rejected the agency’s interpre-
tation upon application of Skidmore’s standard of review.
150
O’Connor’s opinion in Medtronic denied the applicability of Chevron deference to agency
regulations concerning federal preemption of state law. See Medtronic, 518 U.S. at 512
(O’Connor, J., concurring in part and dissenting in part). The doctrinal basis for Justice
Stevens’s call for deference is thus unclear. In their forthcoming study of Supreme Court
deference cases, Eskridge and Baer acknowledge a large number of such “Skidmore-Lite”
cases, where the Court arguably engaged in Skidmore analysis without citing Skidmore. See
Eskridge & Baer, supra note 7 (manuscript at 25–28) (discussing this category of cases).
R
149. In a study such as this one, such goals are easier said than met. Courts are not
known for articulating their conclusions clearly and precisely. Accordingly, some
subjectivity of analysis was inevitable, despite our best efforts.
150. We did not specifically track how many cases were excluded by each of the
individual elements. We note, however, that 229 of the initial 450 cases cited Mead or
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 28 11-OCT-07 15:18
1262 COLUMBIA LAW REVIEW [Vol. 107:1235
The first element is exclusive in important ways. First, it excludes
instances of courts considering agencies’ interpretations of their own reg-
ulations. As we discuss further in Part IV, the law regarding the appropri-
ate standard for such cases is unsettled.
151
Courts occasionally apply Skid-
more
152
but often rely upon the highly deferential standard of Bowles v.
Seminole Rock & Sand Co.
153
instead to evaluate such interpretations.
154
This uncertainty and inconsistency suggests that Skidmore applications in
such contexts may not represent the heartland of the Skidmore doctrine.
More importantly, reviewing regulations strikes us as sufficiently different
from reviewing statutes as to color the courts’ application of Skidmore in
the former context.
155
Accordingly, we chose to cordon off these applica-
tions for future study rather than artificially conflate the analysis here.
156
Christensen without also citing Skidmore. Although we ultimately retained twenty-two such
cases as Skidmore applications, see supra note 145, a vast majority of the others cited Mead
R
or Christensen for a purpose other than invoking the Skidmore standard. See, e.g.,
Metrophones, 423 F.3d at 1065–67 (citing Mead in recognizing Chevron as relevant standard
of review). Accordingly, most of the cases from the initial population of 450 were excluded
from the study under the second element. See infra note 162 and accompanying text.
R
151. See Angstreich, supra note 66, at 56–58 (noting “th[e] longstanding need for a
R
better understanding of why, and when, courts ought to defer to an agency’s interpretation
of its own regulation”); Manning, supra note 22, at 680–96 (arguing against mandatory
R
deference to agency interpretations of their regulations); see also infra notes 385–398 and
R
accompanying text (discussing relationship between Seminole Rock and Skidmore review).
152. See, e.g., Mercy Catholic Med. Ctr. v. Thompson, 380 F.3d 142, 155 (3d Cir.
2004) (applying Skidmore to Secretary of Health and Human Service’s interpretation of
departmental regulations); Ind. Family & Soc. Servs. v. Thompson, 286 F.3d 476, 480–83
(7th Cir. 2002) (applying Mead’s affirmation of Skidmore to Health Care Financing
Administration interpretation of its own manual and regulations); see also Gonzales v.
Oregon, 546 U.S. 243, 268–69 (2006) (identifying Skidmore as standard for evaluating
agency’s interpretation of regulations that it does not administer).
153. 325 U.S. 410, 414 (1945) (explaining that, in reviewing administrative
interpretations of regulations, “the ultimate criterion is the administrative interpretation,
which becomes of controlling weight unless it is plainly erroneous or inconsistent with the
regulation”).
154. In addition, Auer v. Robbins, 519 U.S. 452, 461–63 (1997)—which endorsed the
standard of Seminole Rock—often draws citations along with or in lieu of Seminole Rock . See,
e.g., Humanoids Group v. Rogan, 375 F.3d 301, 305–06 (4th Cir. 2004) (distinguishing
Christensen as applying only to statutory interpretations and applying Auer deference to
“agency’s interpretation of its own regulation”); Eli Lilly & Co. v. Bd. of Regents, 334 F.3d
1264, 1266 (Fed. Cir. 2003) (citing both Seminole Rock, 325 U.S. at 414, and Auer, 519 U.S.
at 461–62, for evaluating such interpretations); see also Gonzales, 546 U.S. at 255
(identifying Auer deference in dicta as applying to agency’s interpretation of its “own
ambiguous regulation”).
155. See Harold J. Krent, Judicial Review of Nonstatutory Legal Issues, in A Guide to
Judicial and Political Review of Federal Agencies, supra note 7, at 151–58 (discussing
R
unique concerns related to agency’s interpretations of their own regulations).
156. In some contexts, it is difficult to say whether an agency has construed a statute
or a regulation; for instance, an agency opinion letter may address a specific factual
question that requires reference to both statutes and regulations. See, e.g., Beck v. City of
Cleveland, 390 F.3d 912, 919–26 (6th Cir. 2004) (determining whether city’s compensatory
time policy complied with “unduly disrupt” standard of 29 U.S.C. § 207(o)(5) (2000) by
referencing both Department of Labor regulations and opinion letters). Only where it was
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 29 11-OCT-07 15:18
2007] MODERN SKIDMORE 1263
Separately, while Skidmore’s attention to agency expertise might suggest
that it should apply to state agencies,
157
the Court has never suggested
that Skidmore extends that far. Accordingly, we did not treat cases involv-
ing state agency interpretations as Skidmore applications.
This first element is also inclusive in certain ways. It includes inter-
pretations of statutes that multiple agencies are charged with enforcing,
since the Court has applied Skidmore in such instances.
158
This element
also sweeps broadly by including disputes between private parties that
nevertheless implicate agency interpretations and thus Skidmore review.
This element requires only that a court address an agency’s interpreta-
tion, not that the agency be a party to the litigation or that the litigation
involve a direct challenge to the interpretation.
159
We chose to include
these cases because often a private party relies on an agency’s interpreta-
tion and forces the court to pass judgment on the interpretation, creating
a situation hardly different from one in which the agency is a party.
160
Also, this element does not attempt to weed out applications of Skidmore
that arguably occur in dicta.
161
Attempting to separate necessary ele-
ments of a case’s holding from mere dicta would have introduced an-
other level of subjectivity in the case coding.
The second element requires that the majority opinion cite Skidmore
or the analogous sections of Mead or Christensen as providing the appro-
priate standard of review of the agency’s statement. The effect of this
decision was to limit this study to cases in which courts subjectively under-
stand themselves to be applying Skidmore as a standard of review. Courts
occasionally cite Skidmore, Christensen, or Mead for legal propositions
clear that the interpretation to which Skidmore was applied construed a regulation did we
exclude the case. See, e.g., Southco, Inc. v. Kanebridge Corp., 390 F.3d 276, 285–87 (3d
Cir. 2004) (deferring under Skidmore to Copyright Office’s conclusion that a part number
falls within the “short phrases” provision in 37 C.F.R. § 202.1(a) (2004)).
157. See, e.g., Ace Elec. Contractors, Inc. v. Int’l Bhd. of Elec. Workers, 414 F.3d 896,
903 (8th Cir. 2005) (citing Skidmore to support deference to state agency interpretation of
Minnesota statute barring age discrimination).
158. See, e.g., Bragdon v. Abbott, 524 U.S. 624, 642 (1998) (citing Skidmore to justify
deference to agencies’ reading of Rehabilitation Act, for which Congress delegated
administrative authority to multiple agencies); cf. In re New Times Sec. Servs., Inc., 371
F.3d 68, 77–80 (2d Cir. 2004) (discussing decision to afford Skidmore deference to SEC’s
interpretation, which conflicted with interpretation of Securities Investor Protection
Corporation).
159. Cf. Kerr, supra note 90, at 18–20 (limiting study of Chevron applications to
R
“appeals from adverse agency adjudications and direct challenges to agency regulations”).
160. Skidmore itself involved such a situation—the Administrator was not a party to the
dispute, but his views were taken into account by the Court. See 323 U.S. 134, 137–40
(1944).
161. See, e.g., Kort v. Diversified Collection Servs., Inc., 394 F.3d 530, 539 (7th Cir.
2005) (approving of Department of Education’s interpretation of statute even though
court decided case on different grounds); Glover v. Standard Fed. Bank, 283 F.3d 953,
962–63 (8th Cir. 2002) (applying Skidmore as alternate basis for holding, in case Christensen
rendered Seminole Rock deference inappropriate for regulation that merely parrots the
statutory term).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 30 11-OCT-07 15:18
1264 COLUMBIA LAW REVIEW [Vol. 107:1235
other than judicial deference, then proceed either to exercise indepen-
dent judgment without reference to deference doctrine or to extend def-
erence without explicitly invoking Skidmore or any other deference stan-
dard.
162
Thus, as this study discusses the independent judgment and
sliding-scale conceptions of Skidmore, it refers to instances in which a
court actually purports to be applying Skidmore or the relevant sections of
Mead or Christensen.
In applying this element, we included cases that cited Skidmore, Mead,
or Christensen in a string citation of several cases.
163
However, we ex-
cluded Skidmore applications that occur only in concurring or dissenting
opinions. We adopted this approach because the empirical aspect of this
study seeks to describe the “law” regarding Skidmore and, of course, only
majority opinions are characterized as law in a given circuit.
164
Nonethe-
less, we did consider the dissenting and concurring opinions for substan-
tive arguments raised in their theoretical discussions of Skidmore’s test.
The third element excludes cases in which a court finds the statute’s
meaning plain, clear, or unambiguous. Each of these terms stands for
the same idea: Because the court discerns statutory clarity, deference to
an administrative interpretation is not an option. The Supreme Court
has not explicitly held that Skidmore contains a “step one” analogous to
Chevron’s first step.
165
However, the logic of Chevron’s step one applies
with equal force to interpretations governed by Skidmore. If the statute
clearly expresses Congress’s intent as to the matter at hand, no contrary
administrative interpretation can stand, no matter what standard of re-
view a court employs.
166
Thus, Skidmore does not come into play in a
162. For example, Skidmore is often cited for the highly fact-bound inquiry into
whether waiting time in a particular case represents working time for purposes of the
FLSA. See, e.g., Brigham v. Eugene Water & Elec. Bd., 357 F.3d 931, 935–36 (9th Cir.
2004).
163. See, e.g., Coal. for Gov’t Procurement v. Fed. Prison Indus., Inc., 365 F.3d 435,
477 (6th Cir. 2004) (citing first Wash. State Dep’t of Soc. & Health Servs. v. Guardianship
Estate of Keffeler, 537 U.S. 371, 385 (2003), and then Christensen v. Harris County, 529
U.S. 576, 587 (2000), and United States v. Mead Corp., 533 U.S. 218, 234 (2001), to
support proposition that agency’s interpretation “is entitled to deference to the extent that
it is reasonable”).
164. More cynically, a judge writing a concurring or dissenting opinion may be less
thorough or precise in applying Skidmore, knowing that the opinion carries no precedential
weight.
165. See, e.g., Mead, 533 U.S. at 227–28 (describing Skidmore’s place within judicial
review of agency’s interpretations).
166. See, e.g., Nat’l Ass’n of Home Builders v. Defenders of Wildlife, 127 S. Ct. 2518,
2534 (2007) (“‘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.’” (quoting Chevron U.S.A. Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837,
842–43 (1984))); Fed. Election Comm’n v. Democratic Senatorial Campaign Comm., 454
U.S. 27, 31–32 (1981) (describing lower court’s discussion of “whether and to what extent
it should defer to” agency’s interpretation as “pointless if the court was correct that the
agency agreements violated the plain language of the Act”).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 31 11-OCT-07 15:18
2007] MODERN SKIDMORE 1265
meaningful way when a court concludes that the statute’s meaning is
clear or unambiguous.
167
Assessing statutory meaning is not a precise inquiry, of course. The
scope of Chevron step one analysis is a matter of extensive debate over,
among other things, how clear is clear enough and which methods and
tools of statutory construction are permissible in the inquiry.
168
The
mere existence of litigation over a statute’s meaning is prima facie evi-
dence of some level of ambiguity, and courts rarely require absolute clar-
ity to resolve a case at Chevron step one.
169
Consequently, there may be at
least marginal overlap between a court’s evaluation of statutory clarity
and its consideration of an agency interpretation. Some commentators
go so far as to assert that Chevron step one and Skidmore are synonymous,
with both requiring a court to make an independent assessment of con-
gressional intent through textual analysis as well as reference to extrinsic
interpretive aids, including the agency’s construction.
170
Although we would resist conflating Chevron step one and Skidmore
review, it suffices for present purposes to explain that we excluded opin-
ions that found the statute’s meaning to be clear because they offer little
insight into the nature of Skidmore review. When a court concludes that
the statute allows for only one interpretation, the agency’s view cannot
play a meaningful role in the court’s decisionmaking. If the agency inter-
pretation aligns with the court’s view of the statute’s clear meaning, that
congruity may increase the court’s confidence in its decision.
171
But the
167. See, e.g., Murphy, Counter-Marbury, supra note 7, at 43–44 (“[T]he Skidmore
R
framework should include a step one at which a court uses the ‘traditional tools of
statutory construction’ to check for clear meaning.” (quoting Chevron, 467 U.S. at 843
n.9)); Rossi, supra note 7, at 1139 (“[E]ven when courts review the types of agency
R
statements for which Skidmore would normally apply, they presumably engage in the
Chevron step-one inquiry.”); Wildermuth, supra note 7, at 1906 (observing that Chevron’s
R
first step “is the beginning point for every analysis of statutory interpretation involving an
agency regardless of the type of agency action that is taken when interpreting the statute”).
But see Herz, Running Riot, supra note 34, at 203–04, 208–09 (arguing that Skidmore ’s
R
analysis should inform Chevron step one).
168. See, e.g., Thomas W. Merrill, Textualism and the Future of the Chevron Doctrine,
72 Wash. U. L.Q. 351, 351–63 (1994) (analyzing relationship between textualism and
Chevron deference); Scalia, supra note 35, at 520–21 (highlighting debate over Chevron step
R
one); Cass R. Sunstein, Law and Administration After Chevron, 90 Colum. L. Rev. 2071,
2105–19 (1990) (discussing interpretative norms and Chevron analysis).
169. See Levin, Exercise of Discretion, supra note 68, at 779 (describing Chevron step
R
one as encompassing “a range of potential judicial responses” and noting disparate
Supreme Court cases).
170. See Herz, Running Riot, supra note 34, at 209 (“Within step one, the court
R
attempts to determine congressional intent ‘employing the traditional tools of statutory
construction;’ one of those tools, of course, is the interpretation of the agency charged
with administering the statute.” (footnote omitted) (quoting Chevron, 467 U.S. at 843
n.9)); Herz, Judicial Review, supra note 7, at 142–45 (discussing overlap between Skidmore
R
and Chevron doctrines); Levin, Exercise of Discretion, supra note 68, at 778–84 (discussing
R
role of deference in evaluating plain meaning at Chevron step one).
171. In this vein, even after concluding that a statute’s meaning is certain, some
courts still proceed to note the agency’s concurring interpretation, stating that it merited
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 32 11-OCT-07 15:18
1266 COLUMBIA LAW REVIEW [Vol. 107:1235
court would reach the same decision without the agency’s interpretation,
and any reference to the deference purportedly due under Skidmore is
superfluous.
172
Alternatively, if the agency interpretation differs from the
court’s view of the statute’s clear meaning, the agency interpretation
must be rejected, and Skidmore cannot direct a different result.
173
By ex-
cluding circumstances in which the court believed itself bound by the
statute’s terms, we were left with a set of applications in which the
agency’s interpretation, considered under Skidmore’s standard, had the
potential to influence the court’s decision significantly. In these in-
stances, Skidmore’s true nature is revealed, whether that be independent
judgment or deference varying along a sliding scale.
That said, we did not presume that Skidmore review necessarily en-
compasses a step one. This element does not require that a court affirma-
tively declare the statute to be unclear or ambiguous and thus worthy of
deferential review. Instead, an opinion that did not engage in a step one
analysis, making no finding of statutory clarity or ambiguity, remained in
the data set.
174
As discussed below, however, we found that many courts
did, in fact, engage in a step one analysis before applying Skidmore.
175
Evaluating whether a particular judicial opinion engaged in this step
one inquiry was at times subjective. To limit bias in applying this element
of our definition, we relied on signals from the reviewing court rather
than our own evaluation of statutory clarity. Thus, only when a court
expressly stated that a statute’s meaning was “plain,” “clear,” or “unambig-
“deference.” See, e.g., Fairhurst v. Hagener, 422 F.3d 1146, 1149–51 (9th Cir. 2005)
(determining “plain meaning” of term “chemical waste” in 33 U.S.C. § 1362(6) (2000) and
then noting that EPA’s corroborating interpretation was “entitled to some deference”);
Kaspar Wire Works, Inc. v. Sec’y of Labor, 268 F.3d 1123, 1130–31 (D.C. Cir. 2001)
(declaring “[t]he plain language of the Act could hardly be clearer,” then also stating that
“even had Congress had [sic] not spoken directly to the question [at hand], the Secretary’s
interpretation would be entitled to deference” under Skidmore). We do not consider such
references to Skidmore to represent an application of its standard of review.
172. See, e.g., Russ Berrie & Co. v. United States, 381 F.3d 1334, 1336, 1338 (Fed. Cir.
2004) (concluding that statutory provisions’ “meaning and intent are clear” and therefore
that “we need not consider” whether deference was due to Customs’s classification
rulings); Marcella v. Capital Dist. Physicians’ Health Plan, Inc., 293 F.3d 42, 47–48 & n.1
(2d Cir. 2002) (interpreting “plain language of the statute,” then noting that agency’s
opinion letters supported this interpretation, but ultimately declining to “decide whether
we would be obligated to defer to these opinion letters” because they were “in accord with
our interpretation” of statute).
173. See, e.g., Catskill Mountains Chapter of Trout Unlimited, Inc. v. City of New
York, 273 F.3d 481, 491–94 (2d Cir. 2001) (noting Skidmore’s standard but ultimately
ignoring EPA’s interpretation because it contradicted “the plain meaning of the text”). In
Skidmore parlance, a court in this posture often deems the agency interpretation
“unpersuasive.” See, e.g., Kai v. Ross, 336 F.3d 650, 654–55 (8th Cir. 2003) (determining
“[t]he plain language of th[e] provision,” and then concluding that agency’s opinion letter
is not “persuasive” because “plain language of the statute is otherwise”).
174. See, e.g., Rocknel Fastener, Inc. v. United States, 267 F.3d 1354, 1358–61 (Fed.
Cir. 2001) (engaging in no explicit “step one” inquiry and considering Customs’s
interpretation throughout discussion of statutory provision).
175. See infra Part II.B.3.c.
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 33 11-OCT-07 15:18
2007] MODERN SKIDMORE 1267
uous,” or used some other cognate language indicating that its hands
were tied, did this element operate to exclude the case.
176
Finally, the fourth element requires that a court actually apply the
Skidmore standard to reach a conclusion regarding the administrative in-
terpretation. This element weeds out situations in which a court dis-
cusses Skidmore as a standard of review, but then decides the issue on
some other ground.
Ultimately, in the first five years of the modern Skidmore era, 104
cases applied Skidmore as the controlling standard of review for an admin-
istrative interpretation. Two of these cases applied Skidmore to two sepa-
rate agency interpretations,
177
so the total number of Skidmore applica-
tions in the time period studied was 106.
2. Evaluating the Skidmore Applications. — Having identified our
group of Skidmore applications, we evaluated the courts’ opinions with re-
spect to two different questions. First, we asked whether the courts of
appeals follow the independent judgment model or the sliding-scale
model of Skidmore deference. Second, to evaluate how deferential
Skidmore actually is, we tracked whether or not courts applying Skidmore
deference sided with or against the agency’s interpretation. The latter
question is entirely objective and easy to track—either the court upholds
the agency’s interpretation or it does not—and, thus, it requires no elab-
oration as to methodology. The former entails some degree of subjective
analysis that necessitates further explanation, however.
We categorized each Skidmore application as exemplifying indepen-
dent judgment or sliding-scale deference, or as indeterminate. In gen-
eral, we categorized an opinion as applying the sliding-scale model of
Skidmore deference if the court discussed at least one of the Skidmore fac-
tors identified above—thoroughness of consideration, agency expertise,
validity of the reasoning, consistency of application, longevity of the inter-
176. See, e.g., Metro Leasing & Dev. Corp. v. Comm’r, 376 F.3d 1015, 1024 (9th Cir.
2004) (“Because the statutory language is not ambiguous, there is no need to rely on the
Department of Treasury’s regulations or administrative rulings that interpret [26 U.S.C.]
§ 535.”); Bullcreek v. Nuclear Regulatory Comm’n, 359 F.3d 536, 541 (D.C. Cir. 2004)
(declining to determine what deference is due NRC’s interpretation “because the result is
the same whether the court applies de novo review, deference under [Skidmore], or Chevron
deference,” as “the text of [42 U.S.C.] § 10155(h) as well as the statutory structure and
legislative history” of statute support agency’s view).
177. See Dabit v. Merrill Lynch, Pierce, Fenner & Smith, Inc., 395 F.3d 25, 34–36 &
n.6 (2d Cir. 2005) (applying Skidmore in accepting SEC’s view that phrase “in connection
with the purchase or sale of a . . . security” means the same thing in Securities Litigation
Uniform Standards Act as it does under section 10(b) of the Securities Exchange Act of
1934); id. at 39–44 & n.9 (applying Skidmore, by way of Cmty. Health Ctr. v. Wilson-Coker,
311 F.3d 132, 138 (2d Cir. 2002), in rejecting SEC’s view that purchaser-seller rule of Blue
Chip Stamps v. Manor Drug Stores, 421 U.S. 723 (1975), does not apply to SLUSA); La.
Envtl. Action Network v. EPA, 382 F.3d 575, 582–84 (5th Cir. 2004) (applying Skidmore to
EPA’s construction of Clean Air Act’s requirement that state antipollution plans include
measures “to take effect” in certain instances); id. at 584–86 (applying Skidmore to EPA’s
view of Act’s measurement of baseline emissions “in the area”).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 34 11-OCT-07 15:18
1268 COLUMBIA LAW REVIEW [Vol. 107:1235
pretation, and formality of format—in evaluating the administrative inter-
pretation. Some courts applied Skidmore factors to the agency interpreta-
tion only briefly in the midst of a larger analysis of the statute’s text, its
legislative history, and its interpretation by other courts. In such cases, it
is difficult to say that the court’s evaluation of that factor was determina-
tive or even a key factor in the court’s decision. Nevertheless, we contend
that a court that bothers to discuss the presence or absence of a Skidmore
factor feels constrained by the need to justify its conclusion in those terms
and, correspondingly, applies that factor to assign some degree of weight
(or no weight) to the agency’s interpretation. Thus, a court’s active dis-
cussion of one or more Skidmore factors, however brief, places its opinion
within the sliding-scale conception of Skidmore.
The Tenth Circuit’s opinion in McGraw v. Barnhart offers a good ex-
ample of sliding-scale analysis in action.
178
McGraw was a social security
case involving the availability of attorney’s fees incurred in agency-level
proceedings under section 406(b)(1) of the Social Security Act.
179
As in
many of the cases we reviewed, the court considered the statute’s text,
legislative history, and purpose, to conclude that a broad interpretation
allowing fee awards under such circumstances was likely, though not nec-
essarily conclusively, the “more appropriate reading.”
180
Citing Skidmore,
the court then recognized that the Social Security Administration had a
policy of not opposing attorney’s fee awards in like cases and observed
that the agency’s practice was consistently applied over a number of years,
thoroughly considered, and supported by valid reasoning.
181
The court
held in favor of the agency’s interpretation of the statute.
182
By contrast, a court following the independent judgment model typi-
cally accepted or rejected the agency’s interpretation based upon its own
independent review of the statute and resulting conclusion that the
agency’s interpretation is or is not the right or best one. In these applica-
tions, courts do not apply any Skidmore factors to determine whether the
agency’s interpretation merits deference. Courts following the indepen-
dent judgment model generally do not even evaluate the agency’s inter-
pretation on its own terms. Instead, courts in such cases tend to evaluate
the interpretive question using a variety of methods of statutory construc-
tion to reach a conclusion that happens to correspond or not with the
agency’s view. Such courts then cite Skidmore, Christensen, or Mead almost
as an afterthought, without discussing or even mentioning the relevant
contextual factors.
The Second Circuit’s opinion in Mack v. Otis Elevator Co. is illustra-
tive.
183
Mack was an elevator mechanic’s helper who brought a sexual
178. 450 F.3d 493, 501 (10th Cir. 2006).
179. Id. at 497–98.
180. See id. at 498–500.
181. See id. at 500–01.
182. See id. at 503.
183. 326 F.3d 116 (2d Cir. 2003).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 35 11-OCT-07 15:18
2007] MODERN SKIDMORE 1269
harassment claim against her employer, Otis, based upon the behavior of
the employee designated “mechanic in charge” of the group of elevator
mechanics at the office building where Mack worked.
184
The interpretive
question in Mack was whether such a mechanic in charge was a “supervi-
sor” for the purposes of Title VII of the Civil Rights Act of 1964.
185
With-
out mentioning any agency interpretation at all, the court considered the
statute’s purposes and policies, related agency law principles, and rele-
vant judicial precedents in adopting a broad definition of supervisor that
included the mechanic in charge.
186
Only after reaching its own conclu-
sion did the court note that enforcement guidelines issued by the EEOC
corresponded with the court’s own interpretation and were thus “persua-
sive” and entitled to “respect” under Skidmore.
187
In referencing the
EEOC’s guidelines, the court mentioned none of the Skidmore factors.
The most subjective classifications involved distinguishing whether a
court was interpreting the statute independently or evaluating the validity
of the agency’s reasoning and thus the permissibility of the agency’s inter-
pretation.
188
We evaluated such cases for the tone of the court’s rhetoric.
Opinions that focused on the agency’s analysis of the statute’s language,
history, and purpose or described the agency’s interpretation as “reasona-
ble” or “permissible” struck us as considering the validity of the agency’s
reasoning as a Skidmore factor.
189
By contrast, opinions were more consis-
tent with the independent judgment model if they examined the statute’s
language, history, and purpose without reference to the agency’s analysis,
accepted an agency’s interpretation because it was the “right” or “best” or
“better” one, or rejected an agency’s interpretation as “unpersuasive” or
184. Id. at 120–22.
185. See id. at 123.
186. See id. at 123–27.
187. See id. at 127 (citing Christensen v. Harris County, 529 U.S. 576, 587 (2000),
which quotes Skidmore for its authority).
188. Twelve Skidmore applications discussed only the validity factor. As discussed
below, while they were challenging to categorize, these cases do not negate our findings
regarding the predominance of the sliding-scale approach. See infra text accompanying
notes 201–207.
R
189. Four such applications that we categorized as applying Skidmore’s validity factor
rather than employing independent judgment were particularly close calls. See St. Mary’s
Hosp. v. Leavitt, 416 F.3d 906, 914–15 (8th Cir. 2005) (accepting agency’s interpretation
because it “makes sense” and was “logical” way to fill statutory gap); United States v. City of
New York, 359 F.3d 83, 93–94 (2d Cir. 2004) (examining circuit precedent, then noting
that EEOC opinion was “entitled to respect” because it was “reasonable and completely
consistent with the Second Circuit’s employee test”); IA 80 Group, Inc. v. United States,
347 F.3d 1067, 1072–73 (8th Cir. 2003) (construing 26 U.S.C. § 168(e)(3)(E)(iii) (2000)
first independently and then “accept[ing] the IRS’s construction” because it “is consistent
with the statute’s legislative history and is not unreasonable”); Bolen v. Dengel (In re
Dengel), 340 F.3d 300, 310 (5th Cir. 2003) (finding agency handbook’s interpretation
“persuasive” in light of statute’s broader context and legislative history, concluding finally
that the interpretation “is not prohibited by” statute).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 36 11-OCT-07 15:18
1270 COLUMBIA LAW REVIEW [Vol. 107:1235
“incorrect.” As discussed above, Justice Thomas’s majority opinion in
Christensen is a prototypical example of this approach.
190
Finally, seven Skidmore applications defied categorization alto-
gether.
191
This indeterminate group of applications resembled neither
the independent judgment nor the sliding-scale model as described
above, nor did their analysis suggest additional conceptions of Skidmore to
compete with those two theories. Rather, in such instances, the court
accepted the agency interpretation and, invoking Skidmore, adopted a def-
erential tone but failed to consider the context of the interpretation to
measure whether or to what extent such deference was due. A court ap-
plying deference in this unconditional manner would seem to fit neither
within the sliding-scale model nor the independent judgment model of
Skidmore, and no scholar has suggested that Skidmore operates in this way.
Yet, a review of the study’s dataset quickly revealed that courts occasion-
ally cite Skidmore to justify deferring to the agency without explaining
whether or why deference is merited.
For example, in Noviello v. City of Boston, the First Circuit considered
whether an employer’s tolerance of a hostile work environment consti-
tutes “discrimination” for the purposes of a retaliation claim under Title
VII.
192
The court scanned Title VII’s text and legislative history, and then
noted that it owed “Skidmore deference” to a compliance manual of the
EEOC, which favored a broad reading of “discrimination.”
193
In deciding
to afford deference to the manual, however, the court did not so much as
mention any of Skidmore’s factors, much less assess the manual under
them. Instead, the court seemed persuaded principally by the fact that
most other circuits facing the question had reached the same
conclusion.
194
In theory, this category could represent a model of Skidmore that is
distinct from the independent judgment and sliding-scale models—a
model in which Skidmore prescribes automatic deference without regard
190. See Christensen, 529 U.S. at 586–87; supra notes 97–103 and accompanying text
R
(summarizing Christensen majority opinion); see also Rossi, supra note 7, at 1125–27
R
(describing Christensen majority as rejecting agency interpretation because it was not the
“better” view).
191. See Appoloni v. United States, 450 F.3d 185, 190–94 (6th Cir. 2006) (deferring
to IRS revenue ruling); Fujitsu Am., Inc. v. United States, 422 F.3d 1364, 1367–69 (Fed.
Cir. 2005) (deferring to Customs tariff classification ruling); Noviello v. City of Boston, 398
F.3d 76, 90 & n.3 (1st Cir. 2005) (deferring to EEOC compliance manual); Goswami v. Am.
Collections Enter., Inc., 377 F.3d 488, 493–94 (5th Cir. 2004) (deferring to FTC staff
commentary); Hecht v. Barnhart, 68 F. App’x 244, 246–47 (2d Cir. 2003) (deferring to
social security program manual); Jewelpak Corp. v. United States, 297 F.3d 1326, 1331–37
(Fed. Cir. 2002) (deferring to Customs tariff classification ruling); Forrester v. Am.
Dieselelectric, Inc., 255 F.3d 1205, 1207–09 (9th Cir. 2001) (deferring to Federal Railroad
Administration memorandum).
192. See Noviello, 398 F.3d at 88–89.
193. See id. at 90 & n.3.
194. See id. at 89 (“Although this court has never fully analyzed the question, our case
law tilts noticeably toward the majority view.”).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 37 11-OCT-07 15:18
2007] MODERN SKIDMORE 1271
to contextual factors. We hesitate, however, to so characterize this group
of applications for several reasons. First, as the “indeterminate” label sug-
gests, classifying these applications was difficult primarily because they re-
sembled both rather than neither of the independent judgment and slid-
ing-scale models.
195
Second, the courts’ analyses in these cases are rather
cursory, suggesting not that the courts understood Skidmore to entail a
blunt directive to defer but that the courts simply declined to elaborate
their application of Skidmore.
196
3. Results of Study of Skidmore in Courts of Appeals. — Our analysis of
Skidmore applications in the federal courts of appeals supports these con-
clusions: First, the sliding-scale model of Skidmore deference dominates
the independent judgment model among the federal circuit courts of ap-
peals. Second, Skidmore deference, while less deferential than Chevron, is
nevertheless highly deferential to administrative interpretations as
applied.
a. Establishing Dominance of the Sliding-Scale Model. — We found that
the sliding-scale conception of Skidmore most accurately describes the pre-
vailing appellate court practice. In 79 of 106, or 74.5%, of Skidmore appli-
cations, the reviewing court assessed at least one Skidmore factor in evalu-
ating the administrative interpretation. By contrast, only 20 of 106, or
18.9%, of Skidmore applications reflected independent judgment. Only 7
applications, or 6.6% of the total, were indeterminate—deferring to the
agency without explanation.
T
ABLE
1: I
NDEPENDENT
J
UDGMENT
V
ERSUS
S
LIDING
-S
CALE
R
EVIEW
Number of Percent of
Description of Skidmore Application Applications Applications
Independent Judgment Model 20 18.9%
Sliding-Scale Model 79 74.5%
Indeterminate 7 6.6%
106 100.0%
This study thus shows that courts do not, on the whole, understand
Skidmore to sanction independent judgment of a statute in the face of an
195. These applications and the independent judgment applications shared the
common feature of the court accepting the agency’s interpretation without applying any
Skidmore factors. Yet these applications differed from the independent judgment
applications since the courts, both in word and in tone, purported to assign weight to the
agency’s interpretation rather than merely using the agency’s interpretation as affirmation
of the courts’ own views. In that sense, they more closely resembled the sliding-scale
applications in which the courts explicitly articulated reasons for deferring.
196. E.g., Fujitsu Am., 422 F.3d at 1366–69 (quoting Mead’s articulation of sliding-
scale standard but then neglecting to assess any factors and instead accepting agency’s view
because plaintiff gave “no[ ] reason for not giving deference”); Goswami, 377 F.3d at
493–94 (accepting FTC interpretation as “persuasive” with no consideration of Skidmore’s
factors and then noting that statute’s legislative history and other judicial decisions were in
agreement).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 38 11-OCT-07 15:18
1272 COLUMBIA LAW REVIEW [Vol. 107:1235
administrative interpretation. Instead, most courts that cite Skidmore be-
lieve themselves bound to afford agencies special consideration that is
not due to ordinary litigants. In a sense, Mead’s vision of Skidmore has
prevailed over Christensen’s approach of citing Skidmore in order to justify
independent review.
Some judicial opinions reviewed in this study discussed forthrightly
the tension between the competing models.
197
For example, in Cathedral
Candle Co. v. United States International Trade Commission, the Federal Cir-
cuit explained:
At times, the [Supreme] Court has characterized the degree of
deference to particular agency interpretations of statutes as de-
pending on “the extent that the interpretations have the ‘power
to persuade’” [citing Christensen]. We are confident that the
Court did not mean for that standard to reduce to the proposi-
tion that “we defer if we agree.” If that were the guiding princi-
ple, Skidmore deference would entail no deference at all. In-
stead, we believe the Supreme Court intends for us to defer to
an agency interpretation of the statute that it administers if the
agency has conducted a careful analysis of the statutory issue, if
the agency’s position has been consistent and reflects agency-
wide policy, and if the agency’s position constitutes a reasonable
conclusion as to the proper construction of the statute, even if
we might not have adopted that construction without the bene-
fit of the agency’s analysis.
198
The Cathedral Candle court thus comprehended how Christensen’s ap-
proach could boil down to independent judgment, accepting the
agency’s view only if the court concludes it is the best interpretation.
199
The court also read Mead to compel sliding-scale deference that is condi-
tioned on the context surrounding the agency’s interpretation.
200
This study’s finding that the sliding-scale approach is more prevalent
than the independent judgment model is robust. However, the courts
employing the sliding-scale approach varied significantly in the extent to
197. See, e.g., Cathedral Candle Co. v. U.S. Int’l Trade Comm’n, 400 F.3d 1352, 1366
(Fed. Cir. 2005); Collins v. Nat’l Transp. Safety Bd., 351 F.3d 1246, 1254 (D.C. Cir. 2003)
(describing Skidmore deference as less than Chevron “but more than acknowledgement that
the agency’s position is more convincing than its adversaries’, as would be true any time it
submitted the more convincing brief”).
198. Cathedral Candle Co., 400 F.3d at 1366 (citation omitted) (quoting Christensen v.
Harris County, 529 U.S. 576, 587 (2000)).
199. See id. For an even more stark example of a court reading Christensen’s
formulation of Skidmore as prescribing independent judgment, see Forest Park II v. Hadley,
336 F.3d 724, 732 n.6 (2003) (“[W]e owe no deference to an opinion letter purporting to
interpret a statute that is not the result of the agency’s rulemaking procedures.” (citing
Christensen, 529 U.S. at 587)).
200. See Cathedral Candle Co., 400 F.3d at 1365–66 (relying on Mead to explain that
“[w]hile the Skidmore standard does not entail the same degree of deference to
administrative decisionmaking as the Chevron standard, it nonetheless requires courts to
give some deference to informal agency interpretations of ambiguous statutory dictates,
with the degree of deference depending on the circumstances”).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 39 11-OCT-07 15:18
2007] MODERN SKIDMORE 1273
which they applied the Skidmore factors. The next subsection explores the
sliding-scale approach in finer detail, but one subset of applications—
those in which the court applied only the “validity” factor—bears men-
tioning here because it is potentially distinct from other applications of
the sliding-scale approach to Skidmore.
Recall that we defined “validity” as testing the merits of the agency’s
interpretation for its reasonableness or permissibility.
201
In some of the
instances that we identified as Skidmore applications, a court purporting to
engage in Skidmore analysis only considered whether the agency’s inter-
pretation was reasonable and ignored whether the agency’s consideration
was thorough, whether its position was consistent or the product of for-
mal procedures, etc.
202
Such an approach neglects the assessment of
whether the agency has exercised care and expertise in forming its inter-
pretation—an assessment that goes to the core of Skidmore’s rationale for
affording deference to administrative interpretations.
203
Instead, be-
cause these courts evaluate only the merits of the interpretation, their
decision to defer (or not) is disconnected from Skidmore’s underlying pre-
mise and is motivated by other concerns. To the extent that a court ac-
cepts an agency’s interpretation solely because it is “valid,” the court po-
tentially extends deference beyond what Mead envisioned.
204
Indeed,
such cases resemble Chevron’s step two, which at least conceptually evalu-
ates the permissibility of the agency’s conclusion without regard to the
agency’s interpretive process or procedures.
205
201. See supra notes 188–190 and accompanying text.
R
202. See, e.g., M. Fortunoff of Westbury Corp. v. Peerless Ins. Co., 432 F.3d 127,
139–40 (2d Cir. 2005) (accepting Federal Motor Carrier Safety Administration’s
interpretation because it “makes economic sense” and is “consistent with the [underlying
statute]”); Cmty. Bank of Ariz. v. G.V.M. Trust, 366 F.3d 982, 987–89 (9th Cir. 2004)
(accepting Office of Comptroller of the Currency’s view because it was “reasonable” and
“consistent with the statutory text”); O’Brien v. Town of Agawam, 350 F.3d 279, 298 (1st
Cir. 2003) (upholding Secretary of Labor’s construction because it “provides a clear and
useful” test for undefined statutory term); see also St. Mary’s Hosp. v. Leavitt, 416 F.3d 906,
915 (8th Cir. 2005); La. Envtl. Action Network v. EPA, 382 F.3d 575, 583–84 (5th Cir.
2004); George Harms Constr. Co. v. Chao, 371 F.3d 156, 160–63 (3d Cir. 2004); Coal. for
Gov’t Procurement v. Fed. Prison Indus., Inc., 365 F.3d 435, 477–78 (6th Cir. 2004); Tum
v. Barber Foods, Inc., 360 F.3d 274, 280–81 (1st Cir. 2004); United States v. City of New
York, 359 F.3d 83, 93 (2d Cir. 2004); Malacara v. Garber, 353 F.3d 393, 401 (5th Cir. 2003);
IA 80 Group, Inc. v. United States, 347 F.3d 1067, 1072–73 (8th Cir. 2003); Bolen v. Dengel
(In re Dengel), 340 F.3d 300, 310 (5th Cir. 2003).
203. Skidmore v. Swift & Co., 323 U.S. 134, 139–40 (1944) (justifying deference to
administrative interpretations by their status as “a body of experience and informed
judgment to which courts and litigants may properly resort for guidance”); see also United
States v. Mead Corp., 533 U.S. 218, 227–28 (2001).
204. See Mead, 533 U.S. at 235–38 (recalling that Court has recognized “more than
one variety of judicial deference,” and emphasizing that “judicial responses to
administrative action must continue to differentiate between Chevron and Skidmore”).
205. See, e.g., Lawson, supra note 79, at 325–31 (contending that Chevron step two
R
considers only interpretive outcomes, while questions of agency procedure and process
represent independent inquiries). For a contrary view of Chevron step two that
incorporates process considerations, see, for example, Ronald Levin, The Anatomy of
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 40 11-OCT-07 15:18
1274 COLUMBIA LAW REVIEW [Vol. 107:1235
On the other hand, courts that assess only an interpretation’s validity
may be implicitly signaling that application of the other Skidmore factors is
unnecessary. This could be because the court independently agrees with
the agency’s view and would accept its interpretation regardless of how
the factors turn out.
206
Similarly, a court might find the agency’s view
impermissible and so refuse to accept it no matter what direction the
other Skidmore factors point.
207
Because courts often do not make clear
the precise basis for their approach to Skidmore, this subset of applications
merits further study.
Twelve of 79, or 15%, of the sliding-scale Skidmore applications fell
into this subset by applying only the validity factor. We do not believe
that the heavy reliance of some courts on the validity factor undermines
our overall conclusion regarding the predominance of the sliding-scale
approach to Skidmore. As discussed above, the tone of a reviewing court’s
rhetoric speaks volumes about whether a court is acting independently or
deferentially in evaluating the merits of the agency’s analysis.
208
Moreo-
ver, even excluding these applications, the sliding-scale conception of
Skidmore is still over three times more common than the independent
judgment conception (67 sliding-scale applications versus 20 indepen-
dent judgment applications).
209
Finally, as an interesting side note, we found the Supreme Court de-
cision a reviewing court cites in support of Skidmore deference to be sug-
gestive of whether the court will engage in independent judgment or slid-
Chevron: Step Two Reconsidered, 72 Chi.-Kent L. Rev. 1253, 1292–94 (1997) (disagreeing
with Lawson’s framework “because ‘outcome’ issues and ‘process’ issues should not be
analyzed in isolation from each other”); Mark Seidenfeld, A Syncopated Chevron:
Emphasizing Reasoned Decisionmaking in Reviewing Agency Interpretations of Statutes,
73 Tex. L. Rev. 83, 125–32 (1994) (arguing that “[s]ubstituting something akin to hard
look review for the deferential reasonableness standard that courts have used in Chevron’s
step two” would help Chevron review better serve interests of democracy and political
accountability of agencies).
206. For example, although the Ninth Circuit in G.V.M. Trust invoked terms of
deference, such as “reasonable,” the opinion’s reasoning gives one the impression that the
court agreed with the agency’s conclusion independently. See 366 F.3d at 989.
207. See, e.g., George Harms, 371 F.3d at 162 (rejecting Secretary of Labor’s
interpretation because it lacked any valid basis); Barber Foods, 360 F.3d at 280–81 (rejecting
Secretary of Labor’s construction that would lead to “absurd result” and “threaten[ ] to
undermine Congress’s purpose” for statute).
208. See supra notes 188–190 and accompanying text.
R
209. One might argue that these applications, in which the court considers only the
validity factor and so assesses only the merits of the agency’s interpretation, actually reflect
the independent judgment model rather than the sliding-scale model. This would result
in a 2:1 ratio of 67 sliding-scale applications to 32 independent judgment applications, a
ratio that still heavily favors the sliding-scale model. However, we perceive the courts in
these cases generally to be affording weight to the agency view (and so falling in the
sliding-scale model) because the agency view falls within the realm of reasonableness
rather than being the best or right answer. In contrast, we perceive that courts in the
independent judgment mode generally accept agency interpretations for the very different
and nondeferential reason of aligning with the courts’ own views. See supra notes 188–190
R
and accompanying text.
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 41 11-OCT-07 15:18
2007] MODERN SKIDMORE 1275
ing-scale review.
210
Recall that the Court’s opinion in Christensen reflects
the independent judgment model of Skidmore deference, while Mead
more clearly contemplates the sliding-scale model. Of the 106 Skidmore
applications we identified, 13 cited only Christensen as the applicable stan-
dard of review, and 6 of those 13, or 46%, proceeded to apply indepen-
dent judgment of the statute.
211
In contrast, of the 9 Skidmore applica-
tions that cited solely Mead for the standard of review only 1, or 11%,
applied independent judgment of the statute.
212
And of the 84 Skidmore
applications that cited Skidmore, whether alone or along with Mead or
Christensen, only 13, or 15%, applied independent judgment to the stat-
ute.
213
These findings support our contention that Christensen and Mead
do, in fact, represent competing models of Skidmore and that courts tend
to follow the model of whichever case they cite.
b. Documenting Skidmore’s Deferential Character. — Our second major
finding was that Skidmore is relatively deferential as applied by the federal
courts of appeals. In 64 of 106, or 60.4%, of the Skidmore applications
studied, the courts sided with the agency.
T
ABLE
2: O
UTCOME OF
S
KIDMORE
A
PPLICATIONS
Number of Percent of
Skidmore Application Outcomes Applications Applications
Court accepts agency’s interpretation 64 60.4%
Court rejects agency’s interpretation 42 39.6%
106 100.0%
210. By this we mean the case that a court cited for its description of what the Skidmore
standard entails, not for another proposition, such as the fact that Skidmore was the
appropriate standard of review.
211. See Padilla-Caldera v. Gonzales, 426 F.3d 1294, 1300–01 (10th Cir. 2005);
Trowell v. Beeler, 135 F. App’x 590, 595–96 (4th Cir. 2005); Broad. Music, Inc. v. Roger
Miller Music, Inc., 396 F.3d 762, 778–80 (6th Cir. 2005); Howard v. Surface Transp. Bd.,
389 F.3d 259, 270 & n.15 (1st Cir. 2004); Forest Park II v. Hadley, 336 F.3d 724, 732 n.6
(8th Cir. 2003); Mack v. Otis Elevator Co., 326 F.3d 116, 127 (2d Cir. 2003).
212. See James v. Von Zemenszky, 284 F.3d 1310, 1318–19 (Fed. Cir. 2002).
213. Eight of 40 applications (20%) citing Skidmore alone applied independent
judgment. See Zhang v. Gonzales, 426 F.3d 540, 543–49 (2d Cir. 2005); Dabit v. Merrill
Lynch, Pierce, Fenner & Smith, Inc., 395 F.3d 25, 34–36 & n.6, 39–44 & n.9 (2d Cir. 2005)
(two applications); White v. Burlington N. & Santa Fe Ry. Co., 364 F.3d 789, 798–800 (6th
Cir. 2004); Town of Stratford v. FAA, 292 F.3d 251, 253 (D.C. Cir. 2002); Old Ben Coal Co.
v. Dir., Office of Workers’ Comp. Programs, 292 F.3d 533, 542 n.8 (7th Cir. 2002); Franklin
v. United States, 289 F.3d 753, 757–61 (Fed. Cir. 2002); Am. Fed’n of Gov’t Employees v.
Veneman, 284 F.3d 125, 129–30 (D.C. Cir. 2002). Five of 26 applications (19%) citing
Skidmore and Mead applied independent judgment. See New Orleans Stevedores v. Ibos,
317 F.3d 480, 483–88 (5th Cir. 2003); Herman v. Fabri-Centers of Am., Inc., 308 F.3d 580,
592–93 (6th Cir. 2002); Landmark Legal Found. v. IRS, 267 F.3d 1132, 1135–37 (D.C. Cir.
2001); Matz v. Household Int’l Tax Reduction Inv. Plan, 265 F.3d 572, 574–76 (7th Cir.
2001); Morris v. Bus. Concepts, Inc., 259 F.3d 65, 70–71 & n.3 (2d Cir. 2001). Surprisingly,
however, none of the 13 applications citing Skidmore and Christensen applied independent
judgment.
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 42 11-OCT-07 15:18
1276 COLUMBIA LAW REVIEW [Vol. 107:1235
Not surprisingly, when courts engage in independent judgment, they
are less likely to accept the agency’s interpretation. Courts accepted the
agency’s view in 10 of 20, or 50%, of the applications in which the court
followed the independent judgment approach to Skidmore.
T
ABLE
2
A
: O
UTCOME IN
I
NDEPENDENT
J
UDGMENT
A
PPLICATIONS
Number of Percent of
Skidmore Application Outcomes Applications Applications
Court accepts agency’s interpretation 10 50.0%
Court rejects agency’s interpretation 10 50.0%
20 100.0%
By contrast, courts accepted the agency’s view in 47 of 79, or 59.5%, of
the applications in which courts applied the sliding-scale model of
Skidmore and evaluated whether deference was merited. All 7 of the
Skidmore applications classified as indeterminate resulted in the court ac-
cepting the agency’s position.
T
ABLE
2
B
: O
UTCOME IN
S
LIDING
-S
CALE
A
PPLICATIONS
Number of Percent of
Skidmore Application Outcomes Applications Applications
Court accepts agency’s interpretation 47 59.5%
Court rejects agency’s interpretation 32 40.5%
79 100.0%
These findings suggest that Skidmore is in fact deferential to agency
statutory interpretations, but as compared to what? Studies by other
scholars support the unsurprising conclusion that Skidmore is measurably
less deferential than Chevron, regardless of the Skidmore model employed.
In particular, a study by Orin Kerr of federal appellate decisions from
1995 and 1996 demonstrated that such courts accepted the agency’s view
in 89% of Chevron applications decided at step two.
214
Kerr’s study pre-
dates Mead but is nevertheless consistent with other studies in showing a
high government success rate in Chevron cases.
215
Yet, another substan-
214. See Kerr, supra note 90, at 31. Because the 89% deference rate documented by
R
Kerr reflects only those cases resolved by a reviewing court at Chevron step two, this finding
offers the best comparison to our study. Kerr separately found an overall deference rate
among identified “Chevron applications” of 73%. See id. at 30. However, that latter finding
includes both Chevron step one and Chevron step two outcomes. Id.
215. See, e.g., Eskridge & Baer, supra note 7 (manuscript at 13 tbl.1) (showing agency
R
win rates of 76.2% in Chevron cases and 73.5% in Skidmore cases decided by Supreme Court
between Chevron decision in 1984 and end of 2005 term, as compared to 66% agency win
rate when Court applied no deference standard); Thomas J. Miles & Cass R. Sunstein, Do
Judges Make Regulatory Policy? An Empirical Investigation of Chevron, 73 U. Chi. L. Rev.
823, 849 (2006) (finding overall government success rate of 64% in EPA and NLRB cases
citing Chevron before federal courts of appeals between 1990 and 2004). These studies did
not separately track Chevron step one and step two outcomes and thus are more
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 43 11-OCT-07 15:18
2007] MODERN SKIDMORE 1277
tial body of literature claims judicial deference to government agencies as
litigants generally.
216
If so, one must ask whether Skidmore’s 60% success
rate in the courts of appeals is any more deferential than the norm. Two
studies of federal administrative agency success rates covering roughly the
same period offer seemingly contradictory results.
Martha Anne Humphries and Donald Songer documented a 58%
success rate for federal administrative agencies generally in the federal
courts of appeals from 1969 through 1988.
217
However, Humphries and
Songer did not detail comparative success rates among different subcat-
egories of administrative law cases. Further, Humphries and Songer lim-
ited their study to those cases in which a federal government agency was a
party,
218
while our study included cases in which the federal government
participated only as an amicus or not at all.
219
Given the significantly
higher success rate for Chevron cases identified by Kerr and others and
the roughly comparable success rate for Skidmore cases found by our
study, one could reasonably hypothesize that federal government agen-
cies must enjoy appreciably lower success rates in other types of adminis-
trative law cases to yield an overall 58% success rate.
In another study of samples of federal appellate cases drawn from
periods in 1965, 1974–75, 1984–85, and 1988, however, Peter Schuck and
comparable to Kerr’s finding of a 73% overall deference rate among all Chevron
applications. See supra note 214. Also, the Miles & Sunstein study evaluated the influence
R
of judges’ political views in Chevron decisionmaking, and their lower 64% rate of
government success may be influenced by their evaluation of only two agencies, the EPA
and the NLRB, chosen explicitly for the highly political nature of their actions. See Miles
& Sunstein, supra, at 848 (outlining rationale for limiting data set to challenges to EPA and
NLRB interpretations); see also Schuck & Elliott, supra note 35, at 1054–55 (questioning
R
whether studies that focus on particular agencies are reliable indicators of patterns across
all administrative agencies and the courts).
216. See, e.g., Linda R. Cohen & Matthew L. Spitzer, The Government Litigant
Advantage: Implications for the Law, 28 Fla. St. U. L. Rev. 391, 395 (2000) (arguing that
strategic behavior by government litigators yields pro-government advantage in Supreme
Court); Martha Anne Humphries & Donald R. Songer, Law and Politics in Judicial
Oversight of Federal Administrative Agencies, 61 J. Pol. 207, 208–10 (1999) (surveying
numerous studies demonstrating high overall success rate among agencies in Supreme
Court litigation); Reginald S. Sheehan, Federal Agencies and the Supreme Court: An
Analysis of Litigation Outcomes, 1953–1988, 20 Am. Pol. Q. 478, 483–95 (1992)
(considering impact of economic versus social policy orientation and executive branch
versus independent status on success rates in Supreme Court litigation).
217. See Humphries & Songer, supra note 216, at 215.
R
218. See id.
219. Nevertheless, our review found that agency interpretations tended to fare only
slightly worse under Skidmore analysis if the agency was not a party to the case. For
example, our study included 21 applications in which the federal government was neither
a party nor submitted an amicus brief. The government’s interpretation was accepted in
12, or 57%, of those 21 applications. Similarly, our study identified 14 applications in
which the federal government participated as amicus, and the government’s interpretation
was accepted in 8, or 57%, of those cases. By comparison, the government’s interpretation
was accepted in 44, or 62%, of the 71 applications in which the government was a party.
See infra Appendix.
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 44 11-OCT-07 15:18
1278 COLUMBIA LAW REVIEW [Vol. 107:1235
Donald Elliott found increasing and ultimately substantially higher
agency success rates in administrative cases, ranging from 55.1% in 1965
up to 76.7% in 1984–85 and 75.5% in 1988.
220
If correct, these findings
may suggest that Skidmore is in fact less deferential than the norm. Yet
despite the overlap in periods covered by the two studies, Schuck and
Elliott studied a substantially different group of cases than Humphries
and Songer.
221
We simply lack the data to reconcile the findings of these
studies or to compare them meaningfully as a baseline against our own
study of Skidmore applications.
None of these studies take into account how different anticipations
regarding potential outcomes under different administrative law doc-
trines may influence the behavior of agencies and regulated parties in
deciding whether and when to pursue litigation.
222
Without a better un-
derstanding of how the pools of Skidmore, Chevron, and administrative law
cases generally differ from one another, we cannot make strong infer-
ences about just how deferential Skidmore review really is. Nevertheless, a
comparison of our findings with the Kerr study offers support for the
widely shared belief that Skidmore is less deferential than Chevron, and
with further research may demonstrate that Skidmore represents a thumb
on the scale in favor of administrative interpretations.
Finally, our finding that the appellate courts accepted the agencies’
views under Skidmore in 60% of the applications studied reveals that
Skidmore is substantially more agency-friendly than other scholars con-
ducting post-Mead analysis have supposed. Eric Womack and Amy Wil-
dermuth each conducted smaller-scale examinations of Skidmore in the
federal courts and reported significantly lower acceptance rates.
223
Womack looked at decisions from all federal courts, both trial and appel-
late, that cited Mead in the six months following Mead.
224
Womack found
that a mere 9 of 29, or 31%, of cases reviewed accepted the agency’s pre-
ferred interpretation under Skidmore review.
225
However, the percentage
of acceptances was slightly higher for the court of appeals decisions he
220. See Schuck & Elliott, supra note 35, at 1003, 1038–39.
R
221. Humphries and Songer limited their study to 734 cases between 1969 and 1988
in which a federal administrative party was either an appellant or a respondent. See
Humphries & Songer, supra note 216, at 215. Schuck and Elliott employed a series of
R
electronic searches focused principally on Westlaw topical categories to identify
administrative law cases decided in six-month periods in 1965, 1974–75, 1984, and 1985,
and a two-month period in 1988. See Schuck & Elliott, supra note 35, at 992.
R
Notwithstanding their limited sample periods, however, Schuck and Elliott analyzed more
than three times as many cases as did Humphries and Songer. See id. at 1003 tbl.1
(charting results from study that included total of 2,325 cases).
222. See Eskridge & Baer, supra note 7 (manuscript at 34) (noting potential role of
R
selection biases in post-Chevron agency win rates).
223. See Wildermuth, supra note 7, at 1897–99; Womack, supra note 7, at 327–28.
R
224. See Womack, supra note 7, at 323 & n.189.
R
225. See id. at 327 & n.199.
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 45 11-OCT-07 15:18
2007] MODERN SKIDMORE 1279
studied: 6 of 15, or 40%, of cases.
226
Also, Womack included two cases
that applied Skidmore to an agency’s interpretation of a regulation,
227
which we excluded since, as explained above, we view such applications as
a class apart from agency statutory interpretations.
228
Additionally,
Womack included a case that we contend does not contain a Skidmore
application,
229
and two more cases in which the court rejected the agency
view as contravening the statute’s plain meaning.
230
Without these cases,
Womack would have found an acceptance rate of 5 of 10, or 50%—a rate
that more closely resembles the rate we found courts to have maintained
in the years following his study.
Wildermuth’s 2006 article purports to have examined “federal ap-
peals court cases citing Skidmore since Womack’s work” and reports that
only 39% of cases affirmed the agency’s view.
231
Yet, Wildermuth’s study
included only 23 cases, far fewer than the 104 that we identified.
232
Wildermuth provides no explanation for how she selected the cases she
studied, which makes her finding difficult to assess. Moreover, review of
Wildermuth’s article shows that she included two cases that we excluded
as dismissing the agency’s view as contrary to the statute’s plain mean-
ing,
233
and one case that we excluded as interpreting a regulation rather
than a statute.
234
Excluding these three cases from Wildermuth’s set of
23 cases raises courts’ acceptance rate of administrative interpretations to
9 of 20, or 45%. This is closer to but still significantly lower than the rate
of acceptance we found. Without knowing how Wildermuth chose her
cases, whether by sampling or some other criteria, we cannot assess her
conclusion. Regardless, our research shows that agencies fare reasonably
well under Skidmore analysis, winning about 60% of their battles, thus re-
futing Womack’s and Wildermuth’s conclusions that Skidmore is not at all
deferential.
226. See id. at 325 n.194 (listing 15 appellate decisions among others); id. at 327
n.199 (listing 6 appellate decisions).
227. See id. at 325 n.194 (citing Tate v. Farmland Indus., Inc., 268 F.3d 989 (10th Cir.
2001); Navarro v. Pfizer Corp., 261 F.3d 90 (1st Cir. 2001)).
228. See supra notes 151–155 and accompanying text.
R
229. See Nat’l Org. of Veterans’ Advocates, Inc. v. Sec’y of Veterans Affairs, 260 F.3d
1365, 1379 (Fed. Cir. 2001) (“[W]e need not decide whether Skidmore deference would be
sufficient to support the government’s interpretation, for 38 C.F.R. § 3.22, as revised, must
be remanded to the agency for another reason. Without further explanation of § 3.22, it is
inconsistent with the agency’s interpretation of another virtually identical statute.”).
230. See Catskill Mountains Chapter of Trout Unlimited, Inc. v. City of New York, 273
F.3d 481, 489 (2d Cir. 2001); Michigan v. EPA, 268 F.3d 1075, 1082 (D.C. Cir. 2001). We
excluded such cases because when a statute’s meaning is clear, courts lack discretion to
defer to an agency’s interpretation such that there can be no meaningful Skidmore review.
231. See Wildermuth, supra note 7, at 1898–99.
R
232. See id. at 1899 nn.176–177.
233. See Orlando Food Corp. v. United States, 423 F.3d 1318, 1325 (Fed. Cir. 2005);
Tax & Accounting Software Corp. v. United States, 301 F.3d 1254, 1261–62, 1267 (10th Cir.
2002).
234. See Moore v. Hannon Food Serv., Inc., 317 F.3d 489, 492 (5th Cir. 2003).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 46 11-OCT-07 15:18
1280 COLUMBIA LAW REVIEW [Vol. 107:1235
c. One Final Note: Skidmore “Step One.” — As noted above,
235
this
study defined a Skidmore application to exclude cases in which the court
found the statute clear or unambiguous. However, this criterion retained
cases in which the court did not reach an explicit conclusion that the
statute was clear; after all, the Court has not said that Skidmore necessarily
includes a “step one” inquiry along the lines of Chevron step one. Al-
though we did not track data specifically with respect to this conclusion,
we nevertheless discovered that, in practice, Skidmore generally does in-
clude a “step one.” In many Skidmore applications, the court first re-
viewed the statute for a plain meaning, determined that the statute was
ambiguous, and then proceeded to apply Skidmore.
236
This approach
should be uncontroversial, given that the administrative interpretation
can have little effect on the way the court reads the statute if Congress’s
intent is certain. Yet, this finding is important for scholars who have
urged that Skidmore should include a step one inquiry.
237
The fact of the
matter is that, in many cases, Skidmore already does.
4. Summary. — In summary, careful analysis of 106 identified
Skidmore applications in the federal courts of appeals demonstrates that,
in a strong majority of cases, the Skidmore doctrine represents a bona fide
standard of review, rather than merely an excuse for reviewing courts to
follow their own interpretive preferences. Additionally, the evidence
shows that Skidmore review is highly deferential—less so than Chevron, but
still weighted heavily in favor of government agencies over their challeng-
235. See supra notes 165–176 and accompanying text.
R
236. See, e.g., McGraw v. Barnhart, 450 F.3d 493, 499–500 (10th Cir. 2006) (stating
that relevant statute “reasonably can be read either narrowly or broadly” and that “neither
of [the statute’s] purposes directly addresses the question before us”); United States v.
W.R. Grace & Co., 429 F.3d 1224, 1237 (9th Cir. 2005) (“[T]he statutory definition of
‘removal’ is vague and, consequently, the EPA’s construction of this statutory term
warrants our deference.”); La. Envtl. Action Network v. EPA, 382 F.3d 575, 583 (5th Cir.
2004) (finding ambiguity in Clean Air Act’s requirement that state antipollution plans
include measures “to take effect” in certain instances, and then applying Skidmore); Fed.
Nat’l Mortgage Ass’n v. United States, 379 F.3d 1303, 1307–08 (Fed. Cir. 2004) (applying
Skidmore only after stating that the “language at issue . . . is equally subject to both
proffered interpretations”); Ammex, Inc. v. United States, 367 F.3d 530, 535 (6th Cir.
2004) (noting that Internal Revenue Code did not define “export” in context of excise
taxes, then deferring to agency’s interpretation); Wells Fargo Bank, N.A. v. FDIC, 310 F.3d
202, 205–09 (D.C. Cir. 2003) (finding Congress did not clearly speak in Federal Deposit
Insurance Act to issue at hand, then assessing FDIC’s interpretation under Skidmore’s
standard); see also Reimels v. Comm’r, 436 F.3d 344, 346–47 (2d Cir. 2006); Rabin v.
Wilson-Coker, 362 F.3d 190, 196–99 (2d Cir. 2004); Bolen v. Dengel (In re Dengel), 340
F.3d 300, 308–10 (5th Cir. 2003).
237. See, e.g., Murphy, Counter-Marbury, supra note 7, at 43–44 (“[T]he Skidmore
R
framework should include a step one at which a court uses the ‘traditional tools of
statutory construction’ to check for clear meaning.” (quoting Chevron U.S.A. Inc. v.
Natural Res. Def. Council, Inc., 467 U.S. 837, 843 n.9 (1984))); Rossi, supra note 7, at 1139
R
(“[E]ven when courts review the types of agency statements for which Skidmore would
normally apply, they presumably engage in the Chevron step-one inquiry.”).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 47 11-OCT-07 15:18
2007] MODERN SKIDMORE 1281
ers. Finally, as with Chevron, the courts generally apply Skidmore defer-
ence only in the face of statutory ambiguity.
C. Skidmore’s Sliding Scale in Practice
Once one accepts that the courts of appeals overwhelmingly ap-
proach Skidmore in the mode of a sliding scale, evaluating the functional-
ity of that model takes on added significance. In fact, the overarching
impression that one receives from the Skidmore cases is a lack of uniform-
ity in how courts apply the sliding-scale conception of Skidmore.
At one level, the ad hoc quality of the Skidmore applications is com-
pelled by Skidmore’s instruction that courts consider “all those factors
which give [the agency’s interpretation] power to persuade.”
238
Hence
Justice Scalia’s mocking description in his Mead dissent of Skidmore review
as “th’ol’ ‘totality of the circumstances’ test.”
239
However, the varying
ways in which courts apply Skidmore’s factors arguably run deeper than
the variety to be found in the typical totality of the circumstances test.
The cases reveal disparate approaches to which factors should be applied
first, how the factors relate to each other, and what each factor means.
To describe in more detail how courts apply the sliding-scale conception
of Skidmore, the following subsections address the individual factors, not-
ing particularly where relevant (1) the courts’ varying understandings of
the factor, (2) the factor’s relationship to other factors, and (3) the
weight courts appear to assign to the factor.
1. Thoroughness of Consideration. — Among the Skidmore factors iden-
tified in Part II, the “thoroughness” of the agency’s consideration of an
issue was one of the most cited in the sliding-scale applications of the
Skidmore standard.
240
That said, the courts’ opinions reflected two differ-
ent conceptions of what this factor entails. Many courts conceived of the
thoroughness factor as testing the extensiveness of the agency’s explana-
tion of its interpretation.
241
Others viewed thoroughness as measuring
238. Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944).
239. United States v. Mead Corp., 533 U.S. 218, 241 (2001) (Scalia, J., dissenting).
240. Identifying discussions of thoroughness was occasionally difficult. For example,
in OfficeMax, Inc. v. United States, the court referred dismissively to the agency’s “one-page
analysis” of the relevant issue as “not contain[ing] the traditional hallmarks for receiving
deference.” 428 F.3d 583, 594 (6th Cir. 2005). While the significance of such a limited
statement could be debated, more than anything else, the statement seems to convey the
court’s sense that the agency’s consideration of the issue was not especially thorough. For
purposes of determining which factor was being applied, we classified an assessment of the
extensiveness of an agency’s explanation as “thoroughness” and an assessment of the
agency’s procedures as “formality.” See infra notes 241–242. By our highly subjective
R
count, 40 of the 79 sliding-scale Skidmore applications offered some discussion related to
the thoroughness of the agency’s consideration. See infra Appendix.
241. See, e.g., Coke v. Long Island Care at Home, Ltd., 376 F.3d 118, 134–35 (2d Cir.
2004) (finding that Department of Labor failed to “exhibit thoroughness in its
consideration” because it “offered virtually no explanation for the direct inconsistency” of
its regulations), rev’d on other grounds, 127 S. Ct. 2339 (2007); Heinz v. Cent. Laborers’
Pension Fund, 303 F.3d 802, 812 n.17 (7th Cir. 2002) (declining to defer to IRS manual
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 48 11-OCT-07 15:18
1282 COLUMBIA LAW REVIEW [Vol. 107:1235
the formality of the agency’s procedures in adopting the interpreta-
tion,
242
notwithstanding Mead’s identification of formality as a separate
factor.
243
Both inquiries permit a generalist court to evaluate whether
the agency has brought its expertise to bear on the interpretive question
and whether the agency has respected regulated parties’ interests in regu-
latory process. However, these two views of thoroughness cover distinct
aspects of agencies’ actions.
Consider, for example, De La Mota v. United States Department of Educa-
tion, in which the government sought deference for an interpretation ad-
vanced by midlevel agency officials through two handbooks and an
email.
244
The De La Mota court weighed the thoroughness factor
against the agency based on the interpretation’s informality:
245
“[T]horoughness is impossible for an agency staff member to demon-
strate when the staff member does not report to the Secretary, bears no
lawmaking authority, and is unconstrained by political accountability.
Thorough consideration requires a macro perspective that a staff mem-
ber, acting alone, lacks.”
246
A reviewing court that defined thoroughness
by reference to the extensiveness of the legal analysis contained in the
email might have resolved the factor in the agency’s favor.
Some evidence suggests that the thoroughness factor carries the po-
tential to negate an agency’s poor performance on the consistency factor.
In five applications, courts held that an agency’s thoroughly considered
interpretation can merit deference even if it is inconsistent with previous
agency views.
247
For instance, in Horn v. Thoratec Corp., the Third Circuit
accepted the FDA’s view that a medical device approval process pre-
empted state law—even though the FDA had previously concluded other-
wise—because the FDA supported its position with well-reasoned analy-
sis.
248
Five applications do not constitute a large set, but the pattern of
because “[t]he single statement in the manual does not tell us anything about the
thoroughness of the agency’s analysis”).
242. See, e.g., Pension Benefit Guar. Corp. v. Wilson N. Jones Mem’l Hosp., 374 F.3d
362, 370 (5th Cir. 2004) (deeming PBGC’s view to have been thoroughly considered
because PBGC order was promulgated “after a thorough review” by agency official); Cline
v. Hawke, 51 F. App’x 392, 397 (4th Cir. 2002) (“Because the OCC implemented a formal
notice-and-comment procedure and consulted [regulated parties] in reaching its decision,
we find that the OCC’s consideration was thorough.”); Heartland By-Products, Inc. v.
United States, 264 F.3d 1126, 1135 (Fed. Cir. 2001) (concluding that Customs gave
“thorough consideration” to its revocation ruling because it “was issued pursuant to a
notice and comment process”).
243. See Mead, 533 U.S. at 228.
244. 412 F.3d 71, 78 (2d Cir. 2005).
245. See id. at 78–80.
246. Id. at 80.
247. See Warner-Lambert Co. v. United States, 425 F.3d 1381, 1385–86 (Fed. Cir.
2005); Horn v. Thoratec Corp., 376 F.3d 163, 178–79 (3d Cir. 2004); Rubie’s Costume Co.
v. United States, 337 F.3d 1350, 1358 (Fed. Cir. 2003); Springfield, Inc. v. Buckles, 292 F.3d
813, 819–20 (D.C. Cir. 2002); Heartland, 264 F.3d at 1136.
248. See 376 F.3d at 179.
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 49 11-OCT-07 15:18
2007] MODERN SKIDMORE 1283
these applications is reinforced by the fact that no courts withheld defer-
ence on the basis of an agency’s inconsistent position when the agency
thoroughly explained its new interpretation. Moreover, this relationship
surfaces in courts’ review of the agencies’ processes, and several courts in
this study supported deferring to an inconsistent but well-explained
agency position by citing Motor Vehicles Manufacturers Ass’n v. State Farm
Mutual Automobile Insurance Co.,
249
the leading process-review case.
250
2. Formality of the Agency’s Procedure and Interpretation. — Courts as-
sessed the formality of the administrative interpretation’s procedural ped-
igree and format with somewhat less frequency than other factors.
251
The lower frequency may stem from the fact that Skidmore did not include
this factor in its list—Mead added it to the list of relevant factors.
252
An-
other explanation may be that, because Mead makes this factor relevant
to the determination of whether to apply Chevron or Skidmore,
253
courts
do not wish to repeat the inquiry in the actual application of Skidmore.
That is, a court might find that an interpretation’s informality disqualifies
it from Chevron deference and then move on to apply other Skidmore fac-
tors besides formality.
254
Accordingly, the relative infrequency with
which the courts specifically discuss the formality factor within the
Skidmore analysis is not necessarily indicative of the courts’ assessment of
its importance.
As noted above, some courts conducted this inquiry under the label
of thoroughness rather than formality.
255
Whatever the label, strong rea-
249. 463 U.S. 29, 43, 57 (1983).
250. See Horn, 376 F.3d at 179 (citing State Farm, 463 U.S. at 42); Springfield, Inc. , 292
F.3d at 819–20 (citing State Farm, 463 U.S. at 57). In the context of discussing Chevron
deference, the Supreme Court has also reiterated that “if the agency adequately explains
the reasons for a reversal of policy, ‘change is not invalidating.’” Nat’l Cable &
Telecomms. Ass’n v. Brand X Internet Servs., 545 U.S. 967, 981 (2005) (quoting Smiley v.
Citibank (S.D.), N.A., 517 U.S. 735, 742 (1996)); cf. Long Island Care at Home Ltd. v.
Coke, 127 S. Ct. 2339, 2349 (2007) (deferring under Auer v. Robbins, 519 U.S. 452 (1997),
and stating, “as long as interpretive changes create no unfair surprise . . . the change in
interpretation alone presents no separate ground for disregarding the Department’s
present interpretation”). At least one court in this study pointed to Brand X to support
prioritizing thoroughness over consistency. See Warner-Lambert Co., 425 F.3d at 1385–86.
251. By our count, 16 of the 79 sliding-scale Skidmore applications discussed the
formality factor. See infra Appendix. See supra note 240 for an explanation of how we
R
defined “formality.”
252. Compare Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944), with United States v.
Mead Corp., 533 U.S. 218, 228 (2001).
253. See Mead, 533 U.S. at 229–30 (noting that “a very good indicator of delegation
meriting Chevron treatment” exists when agencies possess authority to engage in “relatively
formal administrative procedure tending to foster the fairness and deliberation that should
underlie” a regulation with legal force).
254. See, e.g., George Harms Constr. Co. v. Chao, 371 F.3d 156, 161–63 (3d Cir.
2004) (refusing to apply Chevron to Secretary of Labor’s litigation position because it was
“informal interpretation” and rejecting Skidmore deference without reference to
informality of interpretation).
255. See, e.g., Shikles v. Sprint/United Mgmt. Co., 426 F.3d 1304, 1316 (10th Cir.
2005) (“The EEOC’s brief provides no indication of whether the agency has been
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 50 11-OCT-07 15:18
1284 COLUMBIA LAW REVIEW [Vol. 107:1235
sons exist for courts to include the formality factor in their Skidmore analy-
sis. Even more than the thoroughness factor, under which many courts
focus on the extensiveness of an agency’s explanation of its interpreta-
tion, the formality factor protects the interests of regulated parties by re-
warding agencies for engaging in procedures that allow for public in-
put.
256
For example, Mead suggests that interpretations adopted
pursuant to notice-and-comment rulemaking typically, though not always,
deserve Chevron deference.
257
When such interpretations do not merit
Chevron deference, courts are right to find that notice-and-comment
rulemaking bolsters the agency’s claim for strong Skidmore deference.
258
The procedure provides the public notice and opportunity to participate
in rulemaking that ameliorates concerns about agency overreaching. By
contrast, courts found the formality factor cut against deferring to the
agency in cases where the agency’s interpretation appeared only in for-
mats lacking public scrutiny: an internal agency manual;
259
an agency
amicus brief;
260
an email from a midlevel agency official;
261
a Customs
tariff schedule classification;
262
and a permit decision that neglected the
required National Environmental Policy Act analysis, for example.
263
In
all of these cases, the reviewing courts valued deliberative agency proce-
dures and interpretations.
It is possible to take this factor too far, however. For example, in
Structural Industries, Inc. v. United States, the court indicated that a
Customs classification ruling of the sort at issue in Mead was not entitled
even to Skidmore deference in part because it was not promulgated
thorough in its consideration of the issue, and it appears that the agency’s position has not
been subjected to any sort of public scrutiny.”); see also supra note 240 (explaining how we
R
defined thoroughness versus formality); supra notes 244–246 and accompanying text
R
(discussing De La Mota v. U.S. Dep’t of Educ., 412 F.3d 71, 76–80 (2d Cir. 2005)).
256. See, e.g., Research Triangle Reg’l Pub. Transp. Auth. v. United States, 83 F.
App’x 505, 510 (4th Cir. 2003) (noting favorably that agency officials met directly with
regulated party regarding issue at hand).
257. Mead, 533 U.S. at 229–30; see also Long Island Care at Home, Ltd. v. Coke, 127
S. Ct. 2339, 2350–51 (2007) (strengthening connection by finding regulation to be legally
binding and thus entitled to Chevron deference partly because it was promulgated through
notice-and-comment rulemaking).
258. See, e.g., Rubie’s Costume Co. v. United States, 337 F.3d 1350, 1356 (Fed. Cir.
2003) (finding Customs classification warranted Skidmore deference in part because it was
adopted “pursuant to a deliberative notice-and-comment rulemaking process”). But cf.
Coke v. Long Island Care at Home, Ltd., 376 F.3d 118, 131–33 (2d Cir. 2004) (noting that
agency interpretation was adopted pursuant to notice-and-comment rulemaking but
nonetheless finding interpretation impermissible), rev’d on other grounds, 127 S. Ct. 2339
(2007).
259. See Ebbert v. DaimlerChrysler Corp., 319 F.3d 103, 115 (3d Cir. 2003) (analyzing
deference to EEOC Compliance Manual).
260. See Shikles, 426 F.3d at 1315–16.
261. See De La Mota v. U.S. Dep’t of Educ., 412 F.3d 71, 80 (2d Cir. 2005).
262. See Structural Indus., Inc. v. United States, 356 F.3d 1366, 1371 (Fed. Cir. 2004).
263. See High Sierra Hikers Ass’n v. Blackwell, 390 F.3d 630, 648 (9th Cir. 2004).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 51 11-OCT-07 15:18
2007] MODERN SKIDMORE 1285
through notice-and-comment rulemaking.
264
This heightened formality
requirement contradicts Christensen and Mead and also risks collapsing
Mead’s test for whether Skidmore or Chevron applies into the inquiry of
whether Skidmore deference is warranted. Also, since the vast majority of
Skidmore applications will involve interpretations that do not issue from
notice-and-comment rulemaking, it makes little sense to view the absence
of such rulemaking as meaningful.
3. Validity of the Agency’s Reasoning. — Courts evaluated the validity of
an interpretation’s reasoning more often than any other Skidmore fac-
tor.
265
The high frequency with which courts consider the validity factor
may be explained partly by our defining “validity” to include discussion of
the reasonableness and plausibility of the interpretation itself.
266
It is no
surprise that most courts consider the substantive merits of the agency’s
interpretation in determining whether to defer to it; even those courts
that do not explicitly apply this factor but accept the agency view un-
doubtedly make an implicit determination that the agency’s view is valid.
Validity is a unique factor within the Skidmore framework because it
alone tests the merits of the agency’s interpretation. The other factors
assess an interpretation’s context, but not its merits. Most courts pay no
attention to this distinction and evaluate the interpretation’s reasonable-
ness in the midst of the other contextual factors.
267
However, the nature
of the validity inquiry suggests—and some courts accordingly imple-
ment—a logical progression of the Skidmore factors. First, the courts eval-
uate the contextual factors—thoroughness, formality, consistency, and
expertise—to gauge the level of deference an interpretation deserves.
Then, having determined how much leeway the agency has earned, the
court applies the validity factor to decide whether the interpretation falls
within that interval.
268
Heartland By-Products, Inc. v. United States contains
a prototypical example of this approach to Skidmore’s sliding scale.
269
Re-
viewing a challenge to a Customs classification ruling, the Federal Circuit
264. Structural Indus., Inc., 356 F.3d at 1370.
265. By our count, 63 of the 79 sliding-scale Skidmore applications included a
discussion of this factor. See infra Appendix.
266. See supra notes 188–190 and accompanying text.
R
267. See, e.g., Pension Benefit Guar. Corp. v. Wilson N. Jones Mem’l Hosp., 374 F.3d
362, 370 (5th Cir. 2004) (describing interpretation as “logical” in midst of noting its
consistency with previous interpretations, its compatibility with other regulations, and the
fact that it resulted from thorough analysis by expert agency); Cal. State Legislative Bd. v.
Mineta, 328 F.3d 605, 607–08 (9th Cir. 2003) (remarking upon agency’s thorough
consideration, consistent stance, and its “not unreasonable” views before concluding that
deference was due and accepting interpretation).
268. See, e.g., Cathedral Candle Co. v. U.S. Int’l Trade Comm’n, 400 F.3d 1352,
1366–67 (Fed. Cir. 2005) (assessing various Skidmore factors besides validity and then,
having determined deference is due, deciding interpretation is reasonable); Heartland By-
Products, Inc. v. United States, 264 F.3d 1126, 1135–37 (Fed. Cir. 2001) (same).
269. See 264 F.3d at 1135–37; see also Cathedral Candle Co., 400 F.3d at 1366–67
(noting, among other things, agency’s consistent adherence to interpretation, its
explanation of interpretation, and its expertise).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 52 11-OCT-07 15:18
1286 COLUMBIA LAW REVIEW [Vol. 107:1235
first noted that the ruling resulted from notice-and-comment rulemaking,
was supported by a thorough explanation, and hailed from an agency
with expertise in the subject matter, but was inconsistent with prior
Customs pronouncements.
270
Taking each of these factors into account,
the court determined that a measure of deference was due, and then the
court evaluated the interpretation, finding it neither illogical nor
unreasonable.
271
4. Consistency of the Agency’s Interpretation. — As with thoroughness,
courts discussed the consistency of the agency’s interpretation in many of
the cases we evaluated.
272
However, despite its numerous appearances in
judicial opinions, “consistency” seems less dispositive than other Skidmore
factors. In 18 applications, the court found the agency’s interpretation to
be inconsistent with former or following positions; yet in 7 of those appli-
cations, the court accepted the agency’s position.
273
By contrast, an ex-
plicitly negative finding with respect to any of the other factors was virtu-
ally always associated with a decision against the agency.
274
This suggests
that courts are willing to accept changes in agencies’ policies so long as
the agency accompanies those shifts with procedures and reasoning that
alleviate concerns about arbitrariness and unfairness to regulated
parties.
275
Consistency’s decline in Skidmore analysis may stem in part from
Chevron’s deemphasis of that factor.
276
Indeed, the opinions reviewed in
this study indicate that some courts are uncertain why exactly they should
value an agency’s consistency and that courts rarely explain the impor-
tance of this Skidmore factor.
277
Generally, courts value consistency be-
270. See Heartland, 264 F.3d at 1135–36.
271. See id. at 1136.
272. By our count, the courts discussed the consistency factor in 36 of the 79 sliding-
scale Skidmore applications. See infra Appendix.
273. See Warner-Lambert Co. v. United States, 425 F.3d 1381, 1386 (Fed. Cir. 2005);
Horn v. Thoratec Corp., 376 F.3d 163, 179–80 (3d Cir. 2004); In re New Times Sec. Servs.,
Inc., 371 F.3d 68, 87 (2d Cir. 2004); Rubie’s Costume Co. v. United States, 337 F.3d 1350,
1356 (Fed. Cir. 2003); Springfield, Inc. v. Buckles, 292 F.3d 813, 819–20 (D.C. Cir. 2002);
Chao v. Russell P. Le Frois Builder, Inc., 291 F.3d 219, 228 (2d Cir. 2002); Heartland, 264
F.3d at 1136–37.
274. See, e.g., De La Mota v. U.S. Dep’t of Educ., 412 F.3d 71, 80 (2d Cir. 2005)
(rejecting Department of Education’s interpretation after concluding it did not reflect
agency’s congressionally delegated “authority or expertise”); Tum v. Barber Foods, Inc.,
360 F.3d 274, 280 (1st Cir. 2004) (rejecting Secretary of Labor’s interpretation after
reaching negative finding of its validity, noting that it would lead to “absurd” results);
Smith v. City of Jackson, 351 F.3d 183, 189 n.5 (5th Cir. 2003) (declining deference for
EEOC interpretive guidelines because they lack “significant analysis”).
275. This comports with the discussion above, see supra notes 247–250 and
R
accompanying text, regarding cases in which courts explicitly explained that a thoroughly
considered and explained policy change need not be invalidated solely due to its
inconsistency. See, e.g., Warner-Lambert Co., 425 F.3d at 1385–86.
276. See supra note 40 and accompanying text.
R
277. See, e.g., Butterbaugh v. Dep’t of Justice, 336 F.3d 1332, 1341–42 (Fed. Cir.
2003) (speculating as to relevance of interpretation’s longstanding, consistent nature in
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 53 11-OCT-07 15:18
2007] MODERN SKIDMORE 1287
cause it protects parties’ reliance interests, promotes the rule of law by
ensuring similarly situated parties are treated similarly, and guards
against capricious or ill-intentioned agency action.
278
Especially since
Chevron’s ascendance, however, courts have been more attuned to admin-
istrative agencies’ need for flexibility in administering statutes, which
sometimes requires changing positions on a given policy.
279
Thus, there
exists a need for courts to spell out why the consistency—or inconsis-
tency—of a given interpretation affects the availability of judicial
deference.
The Tenth Circuit’s opinion in Southern Utah Wilderness Alliance v.
Bureau of Land Management reflects this concern.
280
The Bureau of Land
Management (BLM) altered an interpretation in a manner that
threatened to destabilize vested property rights but claimed its new inter-
pretation deserved deference on the basis of the other Skidmore factors.
281
The court rejected BLM’s argument and accorded the new interpretation
little deference principally because BLM had changed its interpretation
of the statute three times in thirty years, upsetting settled expectations of
rights holders at each turn.
282
Adding to the confusion regarding the meaning of consistency, some
judges merge their evaluations of consistency with their inquiries into
congressional intent.
283
As discussed further below, other courts style dis-
cussions of congressional intent as implicating the “longstanding or con-
temporaneous” factor.
284
Yet these two conceptions of consistency reflect
Chevron context, but also applying Skidmore); cf. Landmark Legal Found. v. IRS, 267 F.3d
1132, 1137 (D.C. Cir. 2001) (noting in citations that “consistency” could refer to temporal
or logical consistency but not explaining which definition it was using to evaluate agency’s
interpretation).
278. See Yoav Dotan, Making Consistency Consistent, 57 Admin. L. Rev. 995, 1000–01
(2005).
279. See Nat’l Cable & Telecomms. Ass’n v. Brand X Internet Servs., 545 U.S. 967, 981
(2005) (explaining Chevron’s leeway for agencies to alter their interpretations); see also
Dotan, supra note 278, at 1018–29 (discussing Chevron’s approach to consistency and its
R
impact on administrative law).
280. See 425 F.3d 735, 760–61 (10th Cir. 2005).
281. See id at 759–60.
282. See id at 760–61.
283. See OfficeMax, Inc. v. United States, 428 F.3d 583, 602 (6th Cir. 2005) (Rogers,
J., dissenting) (“The logical basis for taking consistency into account is perforce that
Congress must have acquiesced in the agency’s interpretation where it has been
consistently applied.”); Butterbaugh v. Dep’t of Justice, 336 F.3d 1332, 1350–51 (Fed. Cir.
2003) (Bryson, J., dissenting) (arguing that agency’s consistent interpretation merits
deference because Congress was aware of interpretation and reenacted statute several
times).
284. See infra Part II.C.6. We distinguished between applications of the “consistency”
factor and the “longstanding or contemporaneous” factor by asking if the court sought to
determine whether the interpretation gave insight into congressional intent. If so, we
considered the application to be an example of longstanding or contemporaneous factor
evaluation. If instead the court discussed the interpretation’s consistency in the abstract or
in relation to reliance interests or fairness, we thought of the application as evaluating the
consistency factor.
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 54 11-OCT-07 15:18
1288 COLUMBIA LAW REVIEW [Vol. 107:1235
very different reasons for deferring to an agency. Consistency counsels
deference to protect regulated parties and reward well-crafted interpreta-
tions; under this factor, the court focuses on society and the agency.
285
A
contemporaneous interpretation may trigger deference because the
agency’s proximity to the statute’s enactment suggests that the interpreta-
tion benefited from special insight into Congress’s wishes. Similarly,
longstanding interpretations may trigger deference because, in theory,
Congress has acquiesced, especially where it has reenacted the statutory
provision after the agency’s interpretation was made public.
286
Concomi-
tantly, the reasons for not deferring under these inquiries will be very
different, and courts would improve the transparency and coherency of
their decisions by keeping these inquiries separate.
287
5. Agency Expertise. — While the Skidmore Court did not include
agency expertise in its oft-cited list of factors,
288
the “expertise” factor
nevertheless played a prominent role in that opinion
289
and appeared
explicitly in Mead’s compilation of factors.
290
In our view it undergirds
the Skidmore doctrine.
291
Therefore, it is unsurprising that a sizable num-
ber of the sliding-scale Skidmore applications included discussion of the
agency’s expertise.
292
Yet the expertise factor generally lacks teeth, as
285. See OfficeMax, 428 F.3d at 598 (evaluating IRS’s consistency in relationship to
IRS’s treatment of taxpayers filing claims for refunds).
286. Under a strong version of the reenactment doctrine, an agency interpretation
may become legally binding solely because Congress has reenacted the relevant statutory
provision without explicitly rejecting the agency’s view. See, e.g., Nat’l Lead Co. v. United
States, 252 U.S. 140, 146 (1920) (describing reenactment as “an implied legislative
recognition and approval of the executive construction of the statute”); United States v.
Cerecedo Hermanos y Compa˜
nia, 209 U.S. 337, 339 (1908) (“[T]he re¨
enactment by
Congress, without change, of a statute, which had previously received long continued
executive construction, is an adoption by Congress of such construction.” (citing United
States v. Falk, 204 U.S. 143, 152 (1907))). Although the reenactment doctrine is arguably
inconsistent with modern deference doctrine’s emphasis on interpretive flexibility, the
Court still occasionally emphasizes legislative reenactment as supporting deference. See,
e.g., Barnhart v. Walton, 535 U.S. 212, 220 (2002) (citing reenactments as evidence that
“Congress intended the Agency’s interpretation, or at least understood the interpretation
as statutorily permissible”); United States v. Cleveland Indians Baseball Co., 532 U.S. 200,
219–20 (2001) (“ ‘Treasury regulations and interpretations long continued without
substantial change, applying to unamended or substantially reenacted statutes, are deemed
to have received congressional approval and have the effect of law.’” (quoting Cottage Sav.
Ass’n v. Comm’r, 499 U.S. 554, 561 (1991))).
287. For a positive example of keeping these factors separate and a good discussion of
the dangers inherent in finding congressional approval of longstanding agency views, see
Butterbaugh, 336 F.3d at 1342–43.
288. See Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944) (listing particularly
thoroughness, validity, and consistency).
289. Id. at 137–39 (emphasizing agency expertise).
290. See United States v. Mead Corp., 533 U.S. 218, 228 (2001) (listing “relative
expertness” as relevant to gauging deference).
291. See infra Part III.
292. By our count, 29 of the 79 sliding-scale Skidmore applications included discussion
of the agency’s expertise. See infra Appendix.
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 55 11-OCT-07 15:18
2007] MODERN SKIDMORE 1289
courts only counted this factor against agency deference in three of the
cases we evaluated.
293
Indeed, many invocations of expertise seem not to
be careful evaluations of agencies’ interpretive competence vis-`
a-vis the
court, but are instead mere throwaway lines tacked onto independent de-
cisions to defer to the agency.
294
Nevertheless, one of the applications that counted the expertise fac-
tor against deference suggests an approach that could give the factor
more heft. In Hall v. EPA, the Ninth Circuit reviewed a challenge to the
Environmental Protection Agency’s interpretation of the Clean Air Act’s
requirement that state pollution reduction plans must not “interfere”
with other regulatory goals.
295
In rejecting the EPA’s bid for Skidmore
deference, the court noted that the “EPA has given us no basis to con-
clude that the EPA has drawn on any special expertise in advocating this
interpretation.”
296
The Hall court did not focus on whether the EPA pos-
sessed expertise in administering the Clean Air Act generally—surely, the
EPA does. Instead, the Hall court demanded evidence that the EPA actu-
ally applied that expertise in forming the particular interpretation.
297
Conceptualizing the expertise factor in this way usefully hones in on the
interpretation at hand, helping the court focus on whether it should de-
fer to the interpretation rather than to the agency generally. Moreover,
this conception creates incentives for agencies to form their views more
deliberately, for example, by forcing an agency to explain the relation-
ship among a particular interpretation, the overall statutory scheme, and
the agency’s policy choices.
6. Longstanding or Contemporaneous Interpretations. — As noted above,
neither Skidmore nor Mead mentioned the factor of an interpretation’s
longevity or contemporaneity with the statute’s adoption. Yet pre-Chevron
courts often included this factor in their Skidmore analysis,
298
and the
293. See De La Mota v. U.S. Dep’t of Educ., 412 F.3d 71, 80 (2d Cir. 2005) (stating
that interpretation adopted by midlevel agency official did not reflect exercise of agency
expertise); Wilderness Soc’y v. U.S. Fish & Wildlife Serv., 353 F.3d 1051, 1069 (9th Cir.
2003) (declining to defer to agency’s decision because it did not “reflect the product of
specialized agency expertise” (citing Mead, 533 U.S. at 228, 235)); Hall v. EPA, 273 F.3d
1146, 1156 (9th Cir. 2001) (declining to defer to EPA’s approval of state implementation
plan under Clean Air Act).
294. See, e.g., Santiago v. GMAC Mortgage Group, Inc., 417 F.3d 384, 389 n.4 (3d Cir.
2005) (stating that HUD’s “expertise in the area of federally-related home mortgages” adds
weight to its interpretation); Student Loan Fund of Idaho, Inc. v. U.S. Dep’t of Educ., 272
F.3d 1155, 1164 (9th Cir. 2001) (concluding without analysis that Secretary of Education’s
“expertise and long experience with the [student loan program]” favor giving deference to
his interpretation).
295. 273 F.3d at 1155–56.
296. Id. at 1156.
297. See id.; see also Wilderness Soc’y, 353 F.3d at 1069 & n.17 (focusing on agency’s
decisionmaking process).
298. See, e.g., Steadman v. SEC, 450 U.S. 91, 103 (1981) (noting SEC’s longstanding
practice, in place since 1938, as evidence of congressional intent); Adamo Wrecking Co. v.
United States, 434 U.S. 275, 287 n.5 (1978) (rejecting agency’s “contemporaneous
construction” of statute because other Skidmore factors weighed against it); Gen. Elec. Co.
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 56 11-OCT-07 15:18
1290 COLUMBIA LAW REVIEW [Vol. 107:1235
Court has mentioned longevity and contemporaneity in more recent ju-
risprudence.
299
Consequently, we wondered whether this factor had, in
practice, become grafted onto modern Skidmore applications.
We found that, for the most part, courts have kept longevity and con-
temporaneity distinct from the other Skidmore factors. Only a small num-
ber of modern Skidmore applications, 11 by our count, involved discussion
of whether the agency’s longstanding or contemporaneous interpretation
revealed congressional intent on the issue at hand.
300
Two of those 11
cases considered only this factor in determining whether to defer under
Skidmore, which suggests that they comprehended the interpretation’s
longevity or contemporaneity as a distinct inquiry from the other Skidmore
factors, even as they purported to apply Skidmore’s standard.
301
Hence,
while courts do not uniformly separate their evaluation of contemporane-
ity or longevity from Skidmore analysis, the great majority do.
As mentioned in the discussion of consistency above, it makes sense
to keep inquiry into longevity and contemporaneity separate from the
Skidmore test because this factor measures congressional approval of the
agency view.
302
The implications of this factor are quite unlike those of
other Skidmore factors. If the court applies this factor to find that
Congress intends the agency’s conclusion, no other Skidmore factors
should matter: Congress’s intent should trump the court’s preferred in-
terpretation. Yet modern courts are wary of reading too much certainty
into Congress’s failure to address even a longstanding agency interpreta-
tion.
303
The Skidmore factors are geared toward situations in which
v. Gilbert, 429 U.S. 125, 141–42 (1976) (noting in midst of Skidmore analysis that EEOC’s
interpretation was not “contemporaneous”); see also 2 Davis, supra note 2, § 7.14, at 64–69
R
(discussing role of contemporaneity, longevity, and reenactment in pre-Chevron judicial
review of interpretative rules).
299. See Alaska Dep’t of Envtl. Conservation v. EPA, 540 U.S. 461, 487 (2004) (“We
‘normally accord particular deference to an agency interpretation of “longstanding”
duration.’” (quoting Barnhart v. Walton, 535 U.S. 212, 220 (2002))); United States v.
Cleveland Indians Baseball Co., 532 U.S. 200, 219 (2001) (“‘A regulation may have
particular force if it is a substantially contemporaneous construction of the statute by those
presumed to have been aware of congressional intent.’” (quoting Nat’l Muffler Dealers
Ass’n, Inc. v. United States, 440 U.S. 472, 477 (1979))).
300. By our count, the courts discussed longevity and/or contemporaneity in only 11
of the 79 sliding-scale Skidmore applications. See infra Appendix. See also supra note 284
R
(describing our definition of this factor for purposes of interpreting applications).
301. See Reimels v. Comm’r, 436 F.3d 344, 347–48 (2d Cir. 2006) (accepting agency’s
“longstanding revenue ruling” in part to preserve agency’s “unbroken pattern of narrow
statutory interpretation”); O’Shaughnessy v. Comm’r, 332 F.3d 1125, 1130–31 (8th Cir.
2003) (finding district court’s decision not to grant deference was justifiable in part
because, as compared to other agency interpretations that had earned deference, the
revenue ruling at issue “d[id] not reflect a similarly longstanding or consistent
interpretation”).
302. See supra Part II.C.4 (discussing consistency factor).
303. See, e.g., OfficeMax, Inc. v. United States, 428 F.3d 583, 595–96 (6th Cir. 2005)
(expressing doubts about viability of reenactment doctrine and construing its applicability
narrowly); Butterbaugh v. Dep’t of Justice, 336 F.3d 1332, 1340–42 (Fed. Cir. 2003)
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 57 11-OCT-07 15:18
2007] MODERN SKIDMORE 1291
Congress’s intent is uncertain, and the court must determine whether the
agency’s actions, not Congress’s approval, compel deference.
7. Summary. — The appellate courts seem to believe that Skidmore
review represents something more than mere totality of the circum-
stances evaluation. Yet the courts are uncertain as to precisely what that
something is. In the end, the overlapping definitions of the various fac-
tors may make little difference to the outcome in most cases. Put another
way, one might argue, for example, that it matters little whether the
courts label their evaluation of the process by which an agency adopted
its interpretation as thoroughness or formality, so long as they are consid-
ering the implications of that process for deference purposes. Yet the
confusion over the inquiry each factor represents in turn feeds the uncer-
tainty over how the Skidmore standard should function. What the courts
seem to be searching for, and what seems to be lacking in many cases, is
an underlying guiding principle that links the various factors and ex-
plains why one informal, nonbinding agency action is superior to
another.
III. R
ECONCEPTUALIZING
S
KIDMORE
R
EVIEW
It is apparent that the courts of appeals lack a coherent conception
of how Skidmore’s sliding scale should function. By now, it is equally plain
that the Court is unlikely to follow Justice Scalia’s suggestion of eliminat-
ing Skidmore deference altogether.
304
The Supreme Court continues to
assert a role for Skidmore deference; and the courts of appeals clearly take
seriously Skidmore’s admonition that they weigh the various contextual
factors to determine the extent to which administrative interpretations
should be accorded deference. Yet the courts’ approach to the factors is
rather ad hoc and lacking in consistency, from the Supreme Court on
down.
Given the disarray that characterizes the courts’ application of the
Skidmore standard, it is unsurprising that other scholars have taken up the
task of reconceptualizing Skidmore to alleviate some of the difficulty in its
application. Based upon his perceptive analysis of Christensen, Jim Rossi
draws from Justice Breyer’s dissenting opinion to suggest viewing Skidmore
as an intensified version of reasonableness review under Chevron—in es-
sence advocating Chevron step two with a dash of hard look review.
305
Although Rossi’s analysis predates Mead and uses slightly different phrase-
(questioning relevance of longevity absent other pro-agency Skidmore factors and then
conducting separate analysis of reenactment doctrine and noting “congressional inaction
is perhaps the weakest of all tools for ascertaining legislative intent, and courts are loath to
presume congressional endorsement unless the issue plainly has been the subject of
congressional attention”); O’Shaughnessy, 332 F.3d at 1130–31 (declining to infer
congressional acquiescence to longstanding IRS revenue ruling).
304. See United States v. Mead Corp., 533 U.S. 218, 241 (2001) (Scalia, J., dissenting).
305. See Rossi, supra note 7, at 1137–46; see also supra note 79 (explaining hard look
R
review).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 58 11-OCT-07 15:18
1292 COLUMBIA LAW REVIEW [Vol. 107:1235
ology, his approach is consistent with the notion of treating Mead as a
Chevron step one-and-a-half.
306
Also, appropriately in our view, Rossi rec-
ognizes the relevance to Skidmore review of the Chevron step one search
for a statute’s plain meaning, and he acknowledges Skidmore’s potential
for allowing a reviewing court to uphold an agency’s interpretation even
if the court would prefer an alternate reading.
307
Nevertheless, Rossi
merely folds Skidmore review into the Chevron regime rather than treating
Skidmore and Chevron as independent standards of review.
308
In so doing,
Rossi in effect diminishes the significance of the distinct justifications for
the Chevron and Skidmore standards. With the benefit of post-Mead hind-
sight, the courts of appeals, if not the Supreme Court, are seeking a more
robust articulation of Skidmore as an independent standard of review.
Amy Wildermuth endeavors to achieve precisely that goal.
Wildermuth acknowledges that courts understand Skidmore’s approach as
a sliding scale of judicial deference in which “‘agency interpretations re-
ceive various degrees of deference, ranging from none, to slight, to great,
depending on the court’s assessment of the strength of the agency inter-
pretation under consideration.’”
309
Yet, Wildermuth doubts that “it is
possible to sort the varying degrees of deference Skidmore imagines.”
310
Accordingly, Wildermuth proposes viewing Skidmore as representing a sin-
gle, fixed point of “intermediate” deference in which the reviewing court
evaluates the agency’s interpretation much as it would that of any other
litigant, but with a thumb on the scale in the agency’s favor ostensibly
represented by the contextual factors.
311
Wildermuth’s revision of the Skidmore standard promises simplicity,
but to a fault. As we read it, Wildermuth’s approach seems largely analo-
gous to Colin Diver’s suggestion twenty years ago that “[c]ourts generally
use ‘deference’ in an intermediate sense, between ‘courteous regard’ and
‘submission,’” with an administrative interpretation’s pedigree as a
nondispositive plus factor.
312
Her approach thus resembles only a
slightly more deferential version of the independent judgment model of
Skidmore that our study found most courts of appeals declining to follow.
Such a one-size-fits-all approach effectively dismisses the contextual signif-
icance that Skidmore aims to promote. Wildermuth nods toward
306. Cf. supra notes 65–67 and accompanying text (discussing Mead’s inquiry as step
R
zero or step one-and-a-half in deference framework created by Mead, Chevron, and
Skidmore).
307. See Rossi, supra note 7, at 1137–46.
R
308. See id.
309. See Wildermuth, supra note 7, at 1896 (quoting Merrill & Hickman, supra note
R
42, at 855).
R
310. See id.
311. See id. at 1905.
312. See Diver, supra note 80, at 565–66. The principal difference between
R
Wildermuth and Diver seems to be that, while Diver attributes the courts’ ultimately self-
justificatory utilization of Skidmore deference to judicial obfuscation, Wildermuth blames
such behavior on excessive doctrinal complexity.
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 59 11-OCT-07 15:18
2007] MODERN SKIDMORE 1293
Skidmore’s factors, suggesting agency expertise and the interpretation’s
longevity as guides in applying them.
313
Yet those elements are all funda-
mentally variable across interpretations; and there simply is no room for
such variability in the sort of on/off switch model that Wildermuth pro-
poses. In short, Wildermuth forecloses the possibility that a court could
or should feel more or less compelled to defer depending upon how the
Skidmore factors apply in a given case.
While we disagree with Rossi’s and Wildermuth’s approaches toward
Skidmore, we agree that the courts’ haphazard application of Skidmore’s
contextual factors suggests the need for reconceptualization. Further,
based on review of Skidmore’s day-to-day application, we share
Wildermuth’s concern that courts are unwilling and likely unable to dis-
tinguish finely between numerous degrees of deference. In our view,
however, the problem seems to arise from courts thinking of Skidmore
review solely in terms of its factors rather than the goals those factors
serve. With this view in mind, we offer two suggestions to clarify and yet
simplify Skidmore’s standard.
First, the solution to Skidmore’s inconsistent application lies in the
two competing concepts that are at the heart of Skidmore review: compar-
ative agency expertise and the potential for arbitrariness in the exercise
of that expertise. As a doctrine of judicial prudence, Skidmore is premised
principally on the courts’ recognition that sometimes agencies are simply
better situated to resolve certain issues of statutory construction.
314
This
institutional superiority arises from several sources. As compared to the
generalist courts, agencies focus more narrowly in their endeavors upon a
particular statute or group of statutes.
315
Agencies are staffed with attor-
neys and other experts who specialize in the areas they regulate.
316
Agen-
cies dedicate their resources to investigating and evaluating the implica-
tions of alternative interpretations of the statutes they administer and
enforce.
317
All of these elements culminate in expertise being not merely
an isolated factor that the courts consider in employing Skidmore analysis.
313. See Wildermuth, supra note 7, at 1909–11.
R
314. See Skidmore v. Swift & Co., 323 U.S. 134, 139–40 (1944) (emphasizing agency’s
superior expertise regarding issue at bar). Whether an agency possesses relatively superior
expertise in a given case will often depend partly on the subject matter of the question at
hand. Along these lines, the forthcoming study by Eskridge and Baer found the Supreme
Court more likely to side with the agency in cases involving technical and comparatively
apolitical subject matters (e.g., tax, intellectual property, energy) and less likely to side with
the agency regarding less technical or more highly charged subject matters (e.g., civil
rights, labor, criminal law). See Eskridge & Baer, supra note 7 (manuscript at 57–59).
R
315. See Skidmore, 323 U.S. at 137–38 (observing also that statute gave Administrator
both duties and powers over its administration).
316. See United States v. Mead Corp., 533 U.S. 218, 235 (2001) (recognizing agency’s
“specialized experience” with case’s “subtle questions”); Skidmore, 323 U.S. at 138–39
(conceding Administrator’s superior expertise); see also Gonzales v. Oregon, 546 U.S. 243,
269 (2006) (noting “Attorney General’s lack of expertise in this area”).
317. See Skidmore, 323 U.S. at 137–39 (acknowledging agency’s investigatory efforts
and consequent accumulated knowledge).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 60 11-OCT-07 15:18
1294 COLUMBIA LAW REVIEW [Vol. 107:1235
More importantly, agency expertise is the principle that guides Skidmore
as a doctrine and to which all the other factors relate.
Yet the modern, post-Mead version of Skidmore, like Chevron, is also
guided by the need to balance judicial respect for agency policy discre-
tion against concerns over agency arbitrariness. Statutory ambiguity sig-
nals the opportunity for policy choice.
318
Agencies are more democrati-
cally accountable and thus better suited than courts to decide between
policy alternatives,
319
but agencies are not as politically accountable as
the elected branches. The informal interpretive formats for which
Skidmore provides the appropriate standard often lack the procedural
safeguards that the Administrative Procedure Act imposes on notice-and-
comment rulemaking and formal adjudication.
320
The Administrative
Procedure Act calls upon the courts to serve as a bulwark against arbitrary
or capricious agency action.
321
Skidmore’s factors should thus be understood as ferreting out two
things: first, the extent to which agencies have deliberately employed
their superior expertise and resources in evaluating the statutory ambigu-
ity at hand; and second, the potential for arbitrariness in agency action.
Considering the factors in these terms, rather than as isolated inquiries,
should help clarify their meaning and relative weight.
322
For example, the fact of an agency’s inconsistency, standing alone,
tells the court little. Inconsistency may signal arbitrariness, but inconsis-
tency is not troubling if a new interpretation was prompted by new infor-
mation or changed circumstances, or accompanied by a well-reasoned
justification, all of which would alleviate concern that the agency’s waver-
318. See Chevron U.S.A. Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837, 864–66
(1984) (discussing policy implications of ambiguous statutory provisions).
319. See id. (recognizing significance of agencies’ superior political accountability).
320. See, e.g., Eskridge & Baer, supra note 7 (manuscript at 93) (emphasizing
R
relevance of public participation and/or feedback from elected officials as a criteria for
deference); Nina A. Mendelson, Regulatory Beneficiaries and Informal Agency
Policymaking, 92 Cornell L. Rev. 397, 425–33 (2007) (noting reduced opportunities for
public participation in informal guidance documents and negative implications for agency
accountability); Patricia M. Wald, Judicial Review in Midpassage: The Uneasy Partnership
Between Courts and Agencies Plays On, 32 Tulsa L.J. 221, 223 (1996) (recognizing APA-era
view of procedure as “the best mechanism for controlling agency discretion”).
321. See 5 U.S.C. §§ 702, 703, 706 (2000) (supporting and providing standards for
judicial review of administrative action); see also, e.g., Wald, supra note 320, at 223
R
(describing the role of judicial review under the APA as a secondary “retrospective check
on the legality and rationality of administrative action”).
322. Eskridge and Baer suggest similarly that courts applying Skidmore analysis should
focus on core inquiries into agency expertise; “rule-of-law variable[s],” such as reliance
issues; and the level of public participation or feedback in the interpretation’s formation.
See Eskridge & Baer, supra note 7 (manuscript at 92–94). Their formulation seems
R
designed to require agencies to earn courts’ respect, while our emphasis on arbitrariness
suggests that courts should presume such respect while checking for reasons to disregard
that presumption. Hence, our conception of Skidmore would permit strong judicial
deference more often than that of Eskridge and Baer. See infra notes 326–329 and
R
accompanying text (discussing strong form of Skidmore deference under our conception).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 61 11-OCT-07 15:18
2007] MODERN SKIDMORE 1295
ing represents arbitrariness.
323
Similarly, the informality of an interpreta-
tion is important in Skidmore analysis only to the extent it shows that the
agency has not brought its institutional superiority to bear in forming its
interpretation. Thus, an agency handbook that senior agency officials
have thoroughly reviewed and approved for use throughout the agency
may reflect the agency’s expertise despite its informality.
324
Carefully de-
liberated and explained interpretations that implicate an agency’s core
area of competency and draw upon the agency’s experiences in that area,
particularly when adopted through consultative processes, reflect the
agency’s expertise and eschew arbitrariness. Interpretations that do not
possess these qualities—that stray to the boundaries of the agency’s spe-
cialized knowledge, that are adopted ad hoc without collaborative and
transparent deliberation, or that are insufficiently explained, for exam-
ple—suggest that the agency has not truly brought its expertise to bear
and should trigger a court’s concern that the agency has behaved
capriciously.
By emphasizing the juxtaposition of expertise and arbitrariness as
the guiding principles underlying Skidmore’s multifactor analysis, we do
not mean to suggest that the courts should superimpose yet another two-
part test on top of the current, already too complicated deference frame-
work. Rather, courts should simply use these principles to focus their
inquiry, ensuring they do not miss the forest for the trees in applying the
factors of Skidmore’s sliding scale.
Second, in evaluating the relationship between expertise and arbi-
trariness with respect to a particular interpretation, we suggest that courts
view Skidmore’s sliding scale as a choice among three identifiable moods
or attitudes that the courts may adopt toward agencies’ legal interpreta-
tions. Conceptualizing Skidmore’s sliding scale in this manner should ease
the burden of its application while also fulfilling its promise of tailoring
deference to the unique interpretation at hand. The three moods of
Skidmore share some important characteristics. Each turns on an evalua-
tion of the contextual factors, guided by the overarching principles of
expertise and arbitrariness. Also, a substantively invalid or unreasonable
interpretation would not pass muster under any of the three, any more
than it would under Chevron analysis. Rather, the three attitudinal zones
on the Skidmore sliding scale merely reflect the attitudes courts may adopt,
based upon the contextual factors, in deciding whether their own pre-
ferred, reasonable interpretation or the agency’s should prevail.
The first of these three moods is highly deferential. This strong def-
erence is appropriate when the court finds that Skidmore’s factors more or
323. For example, in McGraw v. Barnhart, the court accepted the agency’s change of
course as responding to several other courts rejecting its former, more contemporaneous
interpretation and in light of the agency’s subsequent consistency and valid reasoning in
support of its present view. See 450 F.3d 493, 500–01 (10th Cir. 2006).
324. See, e.g., Lapine v. Town of Wellesley, 304 F.3d 90, 106 (1st Cir. 2002) (giving
considerable deference to Department of Labor handbook).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 62 11-OCT-07 15:18
1296 COLUMBIA LAW REVIEW [Vol. 107:1235
less uniformly fall in the agency’s favor, suggesting applied expertise
rather than arbitrariness as the basis for the agency’s action. In such a
case, a court should defer to a reasonable agency interpretation even if
the court might prefer an alternative view. Thus, if the reviewing court
finds that the agency’s interpretation is the product of extensive agency
evaluation, backed by careful reasoning that demonstrates the considera-
tion of alternatives, consistently applied, and within the range of reasona-
ble interpretations, then the court should simply defer to the agency.
Skidmore deference at this level is distinctly Chevron-like in effect, in that
courts should only reject agency interpretations that are clearly statutorily
impermissible. In fact, for a court to assert its contrary interpretational
preferences under these circumstances would smack of judicial capri-
ciousness and intrusion into the agency’s policy sphere, rather than pro-
tection against arbitrary agency behavior.
Rocknel Fastener, Inc. v. United States offers a good example of this par-
ticular Skidmore mood.
325
Rocknel Fastener involved an importer’s chal-
lenge to a Customs Service tariff classification.
326
The Federal Circuit
noted that Customs had adopted the classification not only in multiple
Headquarters Ruling Letters, but also in a Customs Service publication.
The classification had been supported with “thorough analysis,” consist-
ently maintained for “more than 16 years,” and harmonized with widely
accepted standards of the industry.
327
Finally, the court noted Customs’s
expertise with the “highly detailed” “regulatory scheme,” and so deter-
mined that deference was due to Customs’s interpretation.
328
Satisfied
that Customs had applied its expertise in forming the interpretation, and
finding no hint of arbitrariness, the court deferred to Customs’s view.
329
At the opposite end of the sliding scale is an attitude of little or no
deference toward the agency’s interpretation. When the contextual fac-
tors fall more or less uniformly against the agency, suggesting that an
interpretation is an ad hoc, arbitrary conclusion rather than the result of
325. 267 F.3d 1354 (Fed. Cir. 2001). For other examples of courts properly giving
strong deference under Skidmore’s sliding scale, see Santiago v. GMAC Mortgage Group,
Inc., 417 F.3d 384, 389 n.4 (3d Cir. 2005) (stating, as alternative basis for holding, that
HUD policy statement deserved deference because it was reasonable position and
demonstrated agency’s ongoing, thorough consideration of issue within its expertise);
Cathedral Candle Co. v. U.S. Int’l Trade Comm’n, 400 F.3d 1352, 1365–67 (Fed. Cir. 2005)
(deferring to International Trade Commission’s interpretation as “product of the
Commission’s ‘specialized expertise,’” informed by thorough investigation and
consideration, and having no hint of arbitrariness, since it had been consistently held and
cohered with agency’s other positions (citing United States v. Mead Corp., 533 U.S. 218,
234 (2001); Nat’l Org. of Veterans’ Advocates, Inc. v. Sec’y of Veterans Affairs, 260 F.3d
1365, 1379 (Fed. Cir. 2001))).
326. See Rocknel Fastener, 267 F.3d at 1356 (challenging classification of metal
fasteners).
327. Id. at 1356–58 (noting classification’s agreement with American National
Standards Institute and its long standing position in Customs documents).
328. See id. at 1358.
329. See id.
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 63 11-OCT-07 15:18
2007] MODERN SKIDMORE 1297
the agency’s applied expertise, then a court should without hesitation im-
pose its own interpretation of the statute. For example, a court finding
that an agency has interpreted the statute inconsistently in comparable
circumstances with little or no explanation for the deviation would rightly
suspect that the agency’s view is a product of arbitrary and ad hoc reac-
tion rather than reasoned judgment and applied expertise.
330
This mode
of analysis differs from pure independent judgment in that a reviewing
court is constrained to justify its adoption of its own interpretation versus
that of the agency through analysis of Skidmore’s factors. Nevertheless, in
practical terms, the end result is little different from independent judg-
ment because the court rejects the agency’s interpretation as inferior to
its own, even if the agency’s interpretation otherwise falls within the
realm of reasonableness.
Rosales-Garcia v. Holland illustrates this lower end of the Skidmore def-
erence spectrum.
331
Rosales-Garcia involved an interpretation of the
Illegal Immigration Reform and Immigrant Responsibility Act of 1996 ad-
vanced by the then-Immigration and Naturalization Service solely
through briefs in the course of litigation.
332
While the Sixth Circuit cited
Skidmore as the appropriate evaluative standard, the court noted both the
litigation context in which the agency asserted the interpretation and also
the agency’s inconsistency in having asserted a different interpretation in
other, similar litigation.
333
Hence the court declared the agency’s inter-
pretation “unpersuasive” and instead adopted its own interpretation of
the statute.
334
Finally, true intermediate Skidmore deference occupies the middle
ground, where a court may legitimately assert its own preferences but
should be wary of doing so for fear of intruding too deeply upon agency
policy prerogatives. Intermediate deference is appropriate when assess-
ment of the factors leaves the court with some question of whether the
interpretation resulted from the agency’s applied expertise or from arbi-
trary action. For example, an agency may adopt a permissible interpreta-
tion through a thorough vetting process that includes consideration by
the agency’s highest officials, demonstrating applied expertise; but the
330. See, e.g., Rodriguez v. Whiting Farms, Inc., 360 F.3d 1180, 1189 (10th Cir. 2004)
(declining to defer to Department of Labor opinion letter that was contradicted by
another opinion letter, especially where neither letter included persuasive analysis).
331. 322 F.3d 386 (6th Cir. 2003). For other examples of courts properly giving little
to no deference under Skidmore’s sliding scale, see Shickles v. Sprint/United Mgmt. Co.,
426 F.3d 1304, 1315–16 (10th Cir. 2005) (denying deference to EEOC’s position that was
first asserted in amicus brief, gave no sign of thorough, agency-wide deliberation or public
scrutiny, and contradicted various judicial precedents—all of which suggest arbitrariness);
Hall v. EPA, 273 F.3d 1146, 1156 & n.6 (9th Cir. 2001) (denying deference to EPA’s final
ruling because it contradicted other rulings and lacked any supporting explanation, thus
giving court “no basis to conclude that the EPA has drawn on any special expertise in
advocating this interpretation”).
332. See Rosales-Garcia, 322 F.3d at 403 n.22.
333. See id.
334. See id. at 403–08 & n.22.
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 64 11-OCT-07 15:18
1298 COLUMBIA LAW REVIEW [Vol. 107:1235
same interpretation may represent a change in agency policy or may ap-
pear in a wholly internal, informal format that gave regulated parties little
notice or opportunity for input.
335
Or, an agency may have never articu-
lated a clear basis for its interpretation nor adopted it in a format that
suggests thorough consideration, leaving the court with no evidence that
the agency employed its expertise deliberately; yet the same interpreta-
tion may reflect a long-established agency practice, perhaps dating back
to the statute’s enactment, to which regulated parties generally have con-
formed their primary behavior.
336
In such circumstances, even if it might
prefer another interpretation, the prudent court should think twice
before asserting its own views over those of the agency and upending set-
tled regulatory understanding if, in litigation, the agency can demon-
strate the reasonableness of its interpretation.
337
Heartland By-Products, Inc. v. United States nicely exemplifies this inter-
mediate attitude of Skidmore deference.
338
Like Mead, the Heartland case
involved a Customs classification ruling. Unlike the Mead ruling, how-
ever, the Heartland ruling was adopted after a notice-and-comment pro-
cess, which in turn represented an effort by the Customs Service to recon-
sider and ultimately revoke an earlier, contrary ruling letter.
339
The
Federal Circuit recognized the statute as ambiguous and susceptible of
more than one interpretation, but did not express its own interpreta-
335. For example, in In re New Times Securities Services, Inc., the court noted that the
SEC’s interpretation related to a statutory scheme, the Securities Investor Protection Act,
which lies outside of the SEC’s core competency. See 371 F.3d 68, 83 (2d Cir. 2004).
While having carefully considered the issue, the SEC’s interpretation had first been
advanced in its amicus brief to the case even though the SEC could have intervened with its
view in any number of previous cases. Id. at 80–82. This mixed picture of expertise and
arbitrariness led the court to afford “some degree of deference,” but not the “considerable
deference” the SEC requested. Id. at 83.
336. For example, in New Hampshire v. Ramsey, the court gave some respect to a
Department of Transportation memorandum of guidance which was relatively informal
and laid out its views without much justification, but which had been adhered to since
1983. 366 F.3d 1, 25–26 (1st Cir. 2004); see also Rabin v. Wilson-Coker, 362 F.3d 190, 198
(2d Cir. 2004) (acknowledging agency’s general expertise regarding issue but finding no
evidence that agency either applied its expertise through deliberative process or
considered prominent contrary court ruling).
337. Admittedly, intermediate deference in our framework may become the default
mode that is applied to a wide variety and the majority of interpretations. That result,
however, is entirely appropriate given the types of informal interpretations which are
generally Skidmore eligible.
338. See 264 F.3d 1126 (Fed. Cir. 2001). For other examples of courts properly giving
intermediate deference under Skidmore’s sliding scale, see Cmty. Bank of Ariz. v. G.V.M.
Trust, 366 F.3d 982, 987–89 (9th Cir. 2004) (concluding that Office of the Comptroller of
Currency’s view was “entitled to respect” because OCC’s expertise extended to statute at
issue, and OCC’s interpretation was sensible, but was only implicitly adopted in series of
opinion letters); Cunningham v. Scibana, 259 F.3d 303, 307–09 (4th Cir. 2001) (accepting
Bureau of Prison (BOP) classification of offense as “crime of violence” because decision
was “sound,” based on BOP’s expertise, and consistently held, although classification
appeared in program statement that provided no reasoning to support classification).
339. See Heartland, 264 F.3d at 1129–32.
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 65 11-OCT-07 15:18
2007] MODERN SKIDMORE 1299
tion.
340
Instead, analyzing the ruling under Skidmore, the Federal Circuit
noted several factors supporting deference, including Customs’s thor-
oughness in collecting and considering public comments, the “logical
and well-reasoned explanation” that Customs published, and Customs’s
“‘specialized experience’ in classifying goods.”
341
The court weighed the
lack of consistency represented by the earlier ruling against deference,
but the court also noted that the statute expressly permitted revocation of
earlier rulings, that the earlier ruling was not adopted with the benefit of
notice-and-comment rulemaking, and that the earlier ruling did not ad-
dress the issue raised in the revocation process.
342
Given that analysis
and the reasonableness of the agency’s interpretation, the Heartland
court deferred to the agency’s view.
343
In sum, we are confident that, as courts center their analysis of
Skidmore’s factors around the dual goals of assessing relative expertise and
checking arbitrary decisionmaking, Skidmore’s sliding scale can assume a
more certain form. Also in the interests of simplification and trans-
parency, courts need not bother trying to ascertain a precise degree of
deference that the Skidmore factors prescribe in isolation from the facts of
the given case. Instead, it suffices to keep in mind just three distinct
moods that are available—strong deference, no deference, or intermedi-
ate deference.
IV.
S
KIDMORE
’
S
D
OMAIN
In conducting our study, we observed that questions remain regard-
ing the range of Skidmore’s applicability. Our primary goals with this Arti-
cle were to evaluate the modern Skidmore standard and offer a vision for
how it should be applied. Full consideration of Skidmore’s reach is thus
beyond this Article’s scope. Nevertheless, emerging questions over when
Skidmore applies reinforce the importance of settling how Skidmore should
operate. Accordingly, a brief synopsis of our observations and prelimi-
nary thoughts on this issue is warranted.
Christensen and Mead make clear that Skidmore stands as an important
standard of review alongside of Chevron, but those cases do not limit judi-
cial review of agency interpretations solely to Chevron or Skidmore. In Part
III we suggested that Skidmore’s sliding scale, applied under the right cir-
cumstances, can result in little or no deference at all.
344
Yet recall from
Part II our observation that deferential review—even where it leads to a
conclusion that little or no deference is warranted—nevertheless differs
from independent judgment at the outset.
345
Straight de novo review,
340. See id. at 1134–35, 1136–37.
341. Id. at 1135–36 (quoting United States v. Mead Corp., 533 U.S. 218, 234–35
(2001)).
342. See id. at 1136.
343. See id. at 1137.
344. See supra notes 330–334 and accompanying text.
R
345. See supra Part II.A.1.
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 66 11-OCT-07 15:18
1300 COLUMBIA LAW REVIEW [Vol. 107:1235
with a reviewing court exercising its independent judgment, is conceiva-
bly appropriate under certain circumstances as well. In short, there must
be some outer limits on Skidmore’s domain. The question is where those
outer limits lie.
Courts often apply Skidmore as the evaluative standard for administra-
tive interpretations that fail to qualify for Chevron deference on the basis
of Mead step one because the agency lacks the authority to bind regulated
parties with the force of law.
346
Many if not most of these cases involve
agencies that nevertheless possess some degree of congressionally dele-
gated authority over the statute in question. Both Christensen and
Skidmore involved informal interpretations of the FLSA by the
Administrator of the Wage and Hour Division of the Department of
Labor, an officer with many congressionally imposed duties in the area of
labor and employment but not Chevron-requisite authority.
347
Likewise,
the courts regularly apply the Skidmore standard in reviewing interpretive
guidelines, compliance manuals, and amicus briefs from the EEOC inter-
preting Title VII of the Civil Rights Act of 1964, though that statute gives
the EEOC only limited enforcement authority.
348
Courts also regularly apply Skidmore review when agencies that pos-
sess binding authority nevertheless interpret the statutes they administer
through less formal, nonbinding formats.
349
In other words, such inter-
346. See, e.g., S. Utah Wilderness Alliance v. Bureau of Land Mgmt., 425 F.3d 735,
759–60 (10th Cir. 2005) (applying Skidmore to proposed regulations of Bureau of Land
Management that could not, per operation of statute, take effect without later statutory
authorization); Butterbaugh v. Dep’t of Justice, 336 F.3d 1332, 1339–40 (Fed. Cir. 2003)
(applying Skidmore partly because Office of Personnel Management lacked explicit
rulemaking authority regarding question at hand); Cal. State Legislative Bd. v. Mineta, 328
F.3d 605, 607 (9th Cir. 2003) (applying Skidmore to Federal Railroad Administration’s
interpretive rule because agency “does not have rulemaking power with respect to the
Act”).
347. See Christensen v. Harris County, 529 U.S. 576, 587–88 (2000); Skidmore v. Swift
& Co., 323 U.S. 134, 137–38 (1944) (discussing Administrator’s authority); see also
Herman v. Fabri-Centers of Am., Inc., 308 F.3d 580, 592 (6th Cir. 2002) (noting that
Skidmore and Mead may support deference to such opinion letters); Arriaga v. Fla. Pac.
Farms, LLC, 305 F.3d 1228, 1238 (11th Cir. 2002) (applying Skidmore deference to
Department of Labor opinion letter).
348. See Noviello v. City of Boston, 398 F.3d 76, 90 & n.3 (1st Cir. 2005) (deferring to
EEOC Compliance Manual interpreting Title VII under Skidmore); United States v. City of
New York, 359 F.3d 83, 93 (2d Cir. 2004) (citing Christensen for deference to EEOC
Compliance Manual interpreting Title VII); White v. Burlington N. & Santa Fe Ry. Co., 364
F.3d 789, 798 (6th Cir. 2004) (applying Skidmore in evaluating EEOC amicus brief and
guidelines interpreting Title VII); Smith v. City of Jackson, 351 F.3d 183, 189 n.5 (5th Cir.
2003) (considering court “bound to treat [EEOC guidelines] as having persuasive force, to
the extent that they are thoughtfully considered” (citing Christensen, 529 U.S. at 587)).
349. See, e.g., Reimels v. Comm’r, 436 F.3d 344, 347 n.2 (2d Cir. 2006) (granting
Skidmore deference to IRS revenue ruling interpreting Internal Revenue Code); La. Envtl.
Action Network v. EPA, 382 F.3d 575, 583 (5th Cir. 2004) (applying Skidmore review to
informal EPA interpretation of Clean Air Act); Horn v. Thoratec Corp., 376 F.3d 163,
178–80 & n.25 (3d Cir. 2004) (deferring to FDA amicus brief interpreting Food, Drug &
Cosmetic Act).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 67 11-OCT-07 15:18
2007] MODERN SKIDMORE 1301
pretations satisfy Mead step one but not Mead step two, and thus are ineli-
gible for Chevron deference solely on that basis.
350
Mead itself was just
such a case, as the Customs Service possesses the authority to bind with
the force of law but the interpretation in question was advanced through
a ruling letter with much more limited application.
351
Hence, in speak-
ing of deference generally, the Mead Court explained, “The fair measure
of deference to an agency administering its own statute has been under-
stood to vary with circumstances . . . .”
352
As the Mead Court observed,
the Customs Service issues 10,000 to 15,000 such rulings each year.
353
Unsurprisingly, many are challenged, and the courts apply Skidmore defer-
ence in evaluating such interpretations.
354
Other common examples are
rulings and other informal interpretations issued by the Internal Revenue
Service concerning the Internal Revenue Code
355
or by the
Environmental Protection Agency interpreting the Clean Air Act and the
Clean Water Act.
356
Skidmore deference may vary within this category
based in part upon the relative formality of the agency’s interpretation:
Published rulings issued by an agency head may be more persuasive than
mere post-hoc litigating positions.
357
In these common scenarios, where an administering agency either
possesses expertise but not the power to bind or enjoys Chevron-requisite
authority but chooses to act more informally, Mead’s two prongs apply
neatly to deny Chevron deference. Such cases thus fall in the heartland of
Skidmore’s domain and represent the majority of Skidmore applications.
Under such circumstances, the courts are generally in agreement that
Skidmore provides the appropriate standard of review. Applying Skidmore
in these contexts makes sense, as deferential review clearly serves
Skidmore’s appreciation for agency expertise and concern for uniformity
350. See United States v. Mead Corp., 533 U.S. 218, 226–27 (2001) (establishing this
two-part test); see also supra notes 64–67 and accompanying text (discussing Mead test).
R
351. Mead, 533 U.S. at 222–23.
352. Id. at 228 (emphasis added).
353. See id. at 233.
354. See, e.g., Warner-Lambert Co. v. United States, 407 F.3d 1207, 1209 (Fed. Cir.
2005); Structural Indus., Inc. v. United States, 356 F.3d 1366, 1370 (Fed. Cir. 2004); Int’l
Trading Co. v. United States, 281 F.3d 1268, 1274 n.2 (Fed. Cir. 2002); Mead Corp. v.
United States, 283 F.3d 1342, 1346 (Fed. Cir. 2002); Rocknel Fastener, Inc. v. United
States, 267 F.3d 1354, 1357 (Fed. Cir. 2001).
355. See, e.g., Appoloni v. United States, 450 F.3d 185, 193–94 (6th Cir. 2006);
Reimels v. Comm’r, 436 F.3d 344, 347 n.2 (2d Cir. 2006); OfficeMax, Inc. v. United States,
428 F.3d 583, 594–95 (6th Cir. 2005); Ammex, Inc. v. United States, 367 F.3d 530, 535 (6th
Cir. 2004); O’Shaughnessy v. Comm’r, 332 F.3d 1125, 1130 (8th Cir. 2003); Omohundro v.
United States, 300 F.3d 1065, 1067–68 (9th Cir. 2002).
356. See, e.g., La. Envtl. Action Network v. EPA, 382 F.3d 575, 583–84 (5th Cir. 2004);
Hall v. EPA, 273 F.3d 1146, 1155–60 (9th Cir. 2001).
357. Compare, e.g., Omohundro, 300 F.3d at 1068–69 (deferring under Skidmore to
published revenue ruling consistent with subsequently adopted regulation), with U.S.
Freightways Corp. v. Comm’r, 270 F.3d 1137, 1142–47 (7th Cir. 2001) (finding
interpretation advanced solely through application and litigation unpersuasive under
Skidmore analysis).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 68 11-OCT-07 15:18
1302 COLUMBIA LAW REVIEW [Vol. 107:1235
of interpretation. Occasionally, however, cases present circumstances
where the courts of appeals are less clear as to whether Chevron, Skidmore,
or no deference is appropriate for the administrative interpretations
before them.
As between the standards of Chevron and Skidmore, Mead’s two-part
test ultimately ought to determine which standard a court applies. How-
ever, it appears that courts may have extended the scope of Skidmore’s
applicability arguably beyond what the Mead Court envisioned simply to
avoid the complexity of the Mead test.
358
Whether congressional delega-
tion exists (Mead step one) and whether an interpretation holds legal
force (Mead step two) are often unclear. Thus, where courts agree with
the agency’s arguably but not obviously Chevron-eligible interpretation,
they may be inclined to dodge the Mead inquiry entirely by finding that
the interpretation withstands Skidmore’s less deferential scrutiny.
359
In
this way, some interpretations that arguably merit Chevron deference re-
ceive Skidmore review.
Likewise, on the opposite end of the Skidmore spectrum, where
Chevron is obviously inappropriate and courts agree with the agency’s in-
terpretation, they may assume without further inquiry the applicability of
Skidmore deference.
360
Consequently, interpretations that arguably are
not entitled to Skidmore deference seem to receive it. The result in such
cases is the same—the agency’s view stands.
361
After all, the true test of
the scope of a judicial deference doctrine only comes when the reviewing
court prefers a different interpretational alternative from that advanced
by the agency. Yet assumptions made in the course of judicial agreement
with agency interpretations may be taken as precedents in later cases
358. See Lisa Schultz Bressman, How Mead Has Muddled Judicial Review of Agency
Action, 58 Vand. L. Rev. 1443, 1464–69 (2005) (observing same and discussing problems).
359. See, e.g., United States v. W.R. Grace & Co., 429 F.3d 1224, 1241 (9th Cir. 2005)
(“Whichever of [Chevron or Skidmore] applies, we reach the same result . . . .”); Santiago v.
GMAC Mortgage Group, Inc., 417 F.3d 384, 389 n.4 (3d Cir. 2005) (“[B]ecause we would
find HUD’s interpretation to be persuasive under Skidmore, we would not need to reach
whether Chevron deference is warranted.”); Southco, Inc. v. Kanebridge Corp., 390 F.3d
276, 286 n.5 (3d Cir. 2004) (declining to “decide what degree of deference is warranted
under the circumstances” and citing Skidmore to justify deferring to Copyright Office’s
interpretation); Pension Benefit Guar. Corp. v. Wilson N. Jones Mem’l Hosp., 374 F.3d
362, 369 (5th Cir. 2004) (“We do not need to decide whether the PBGC’s
interpretation . . . warrants Chevron deference because it is clear that the PBGC’s order may
be upheld . . . under the less deferential standard set forth in [Mead’s affirmation of
Skidmore].”); Stephen Bronte, Advisors, LLC v. Commodities Futures Trading Comm’n, 90
F. App’x 251, 252 (9th Cir. 2004) (“We need not determine whether to afford this
interpretation Chevron or Skidmore deference, since even under Skidmore deference we
would sustain the CFTC’s interpretation of the Act as a reasonable one.”).
360. See, e.g., Dabit v. Merrill Lynch, Pierce, Fenner & Smith, Inc., 395 F.3d 25, 34 n.6
(2d Cir. 2005) (assuming but declining to apply Skidmore review because agreed with
agency).
361. See Herz, Judicial Review, supra note 7, at 143–44 (recognizing that courts can
R
resolve many cases without deciding which deference doctrine applies); Sunstein, Step
Zero, supra note 59, at 229–30 (same).
R
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 69 11-OCT-07 15:18
2007] MODERN SKIDMORE 1303
when judges and agencies disagree. The Court only took up Christensen
and Mead after several years of the lower courts’ grappling with questions
over Chevron’s applicability under various circumstances—difficulties
prompted at least in part by mixed messages in the Court’s own
jurisprudence.
362
The Court has not enunciated a rule comparable to Mead for deter-
mining which interpretations should be the subject of independent judg-
ment rather than Skidmore review. Broad-based application of Skidmore
deference in the post-Mead era really has only just begun. Cases raising
true scope-of-applicability issues at the lower end of the scale are the ex-
ceptions rather than the rule of Skidmore applications. Consequently, ap-
pellate court jurisprudence to date is too limited and fragmented to yield
many clear patterns. Yet a few such circumstances have emerged requir-
ing courts to address the line between Skidmore deference and no defer-
ence at all, and we suspect that others will follow.
For example, interpretations advanced for the first time in the
course of litigation represent an agency interpretation that arguably does
not deserve Skidmore deference. In Skidmore, the Court deferred to the
agency’s interpretation as expressed in an amicus brief, buttressed by an
interpretive bulletin and informal rulings addressing analogous though
not precisely identical circumstances.
363
Hence the courts seem gener-
ally inclined to include amicus briefs filed by agencies among the
Skidmore-eligible formats.
364
Where the agency is a party to litigation in
the course of enforcement, however, the courts are less clearly unani-
mous in their support for Skidmore’s applicability. In Chao v. Russell P.
Le Frois Builder, Inc., for example, the Second Circuit considered an inter-
pretation of the Occupational Safety and Health Act advanced by the
Secretary of Labor for the first time in the course of litigation challenging
an enforcement decision by the Occupational Safety and Health Review
Commission, and concluded that Skidmore provided the appropriate stan-
dard of review.
365
Other courts, such as the Federal Circuit in Caribbean
362. For example, the circuits disagreed over whether Chevron deference was available
for interpretative rules and policy statements. Compare Elizabeth Blackwell Health Ctr.
for Women v. Knoll, 61 F.3d 170, 181–82 (3d Cir. 1995) (applying Chevron in reviewing
interpretative rule), with Massachusetts v. FDIC, 102 F.3d 615, 621 (1st Cir. 1996) (denying
Chevron’s applicability for “[l]ess formal interpretations” including policy statements and
guidelines). See generally Merrill & Hickman, supra note 42, at 848–52 (detailing lower
R
court confusion over Chevron’s applicability pre-Mead).
363. See Skidmore v. Swift & Co., 323 U.S. 134, 139–40 (1944).
364. See, e.g., Conn. Office of Prot. & Advocacy for Persons with Disabilities v.
Hartford Bd. of Educ., 464 F.3d 229, 239–40 (2d Cir. 2006); M. Fortunoff of Westbury
Corp. v. Peerless Ins. Co., 432 F.3d 127, 139 (2d Cir. 2005); Cmty. Bank of Ariz. v. G.V.M.
Trust, 366 F.3d 982, 987 (9th Cir. 2004); Matz v. Household Int’l Tax Reduction Inv. Plan,
265 F.3d 572, 574–75 (7th Cir. 2001).
365. See 291 F.3d 219, 227–28 (2d Cir. 2002); see also Shikles v. Sprint/United Mgmt.
Co., 426 F.3d 1304, 1315–16 & n.7 (10th Cir. 2005) (applying Skidmore factors but
declining to defer to interpretation advanced in EEOC amicus brief); Old Ben Coal Co. v.
Dir., Office of Workers’ Comp. Programs, 292 F.3d 533, 542 n.8 (7th Cir. 2002)
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 70 11-OCT-07 15:18
1304 COLUMBIA LAW REVIEW [Vol. 107:1235
Ispat Ltd. v. United States, have employed Skidmore review in evaluating
agency litigating positions, only to apply Skidmore’s contextual factors to
conclude that such interpretations merited little or no deference.
366
In
still other cases, such as Connecticut Office of Protection & Advocacy for
Persons with Disabilities v. Hartford Board of Education, courts have suggested
that agency litigation positions are mere “‘post hoc rationalizatio[ns]’” to
which any form of deference is inappropriate.
367
Another issue of Skidmore’s reach pertains to the relationship be-
tween Skidmore and stare decisis. Prior to Mead, jurisprudence regarding
the primacy of Chevron or judicial precedent was extraordinarily con-
flicted.
368
In its 2005 decision in National Cable & Telecommunications
Ass’n v. Brand X Internet Services, the Supreme Court resolved that contro-
versy in favor of deference.
369
“Only a judicial precedent holding that
the statute unambiguously forecloses the agency’s interpretation, and
therefore contains no gap for the agency to fill, displaces a conflicting
agency construction.”
370
Yet Brand X was a Chevron case, involving a le-
gally binding Federal Communications Commission interpretation of the
Communications Act of 1934 promulgated through public notice and
comment.
371
Skidmore is neither discussed nor even cited in any of the
opinions issued in Brand X, even though Skidmore deference shares the
(suggesting that Skidmore deference may be appropriate for some agency litigation
positions).
366. See 450 F.3d 1336, 1340–41 (Fed. Cir. 2006) (applying Skidmore but declining
deference based on inconsistency and unpersuasiveness); see also Padash v. INS, 358 F.3d
1161, 1168 n.6 (9th Cir. 2004) (employing Skidmore analysis in dicta and rejecting agency
view); Natural Res. Def. Council v. Abraham, 355 F.3d 179, 201 (2d Cir. 2004) (“‘[T]o
carry much weight, an agency’s interpretation must be publicly articulated at some time
prior to the embroilment of the agency in litigation over the disputed provision.’”
(quoting Natural Res. Def. Council v. Herrington, 768 F.2d 1355, 1428 (D.C. Cir. 1985))).
367. See 464 F.3d at 239–40 (contrasting amicus brief in case at bar with “ ‘post hoc
rationalizatio[ns]’” and suggesting latter may not be entitled to Skidmore deference
(quoting Auer v. Robbins, 519 U.S. 452, 462 (1997))); see also Old Ben Coal Co., 292 F.3d at
542 n.8 (distinguishing between agency litigating position “trying to defend an action it
already had taken” from another with “a less-vested interest in the outcome”); Defenders of
Wildlife v. Norton, 258 F.3d 1136, 1145 n.11 (9th Cir. 2001) (“[W]e ordinarily will not
defer ‘to agency litigating positions that are wholly unsupported by regulations, rulings, or
administrative practice.’” (quoting Bowen v. Georgetown Univ. Hosp., 488 U.S. 204, 212
(1988))).
368. In a series of cases in the 1990s, the Supreme Court seemed to suggest that its
own precedents trumped Chevron deference. See Neal v. United States, 516 U.S. 284, 290,
294–95 (1996); Lechmere, Inc. v. NLRB, 502 U.S. 527, 536–37 (1992); Maislin Indus., U.S.,
Inc. v. Primary Steel, Inc., 497 U.S. 116, 126–31 (1990). The courts of appeals were divided
on the question with respect to their own precedents. Compare, e.g., Satellite Broad. &
Commc’ns Ass’n of Am. v. Oman, 17 F.3d 344, 348 (11th Cir. 1994) (favoring Chevron over
circuit precedent), and Chem. Waste Mgmt., Inc. v. EPA, 873 F.2d 1477, 1480–82 (D.C.
Cir. 1989) (same), with Bankers Trust N.Y. Corp. v. United States, 225 F.3d 1368, 1376
(Fed. Cir. 2000) (finding no conflict between applying Chevron and stare decisis).
369. See 545 U.S. 967, 981 (2005).
370. Id. at 982–83.
371. See id. at 980–81.
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 71 11-OCT-07 15:18
2007] MODERN SKIDMORE 1305
same tension with stare decisis as Chevron previously did.
372
Furthermore,
the dilemma of Skidmore and stare decisis falls squarely into the long-
standing problem of agencies declining to follow adverse decisions from
the courts of appeals,
373
as such nonacquiescence is rarely expressed
through Chevron-eligible formats.
374
And it is far from clear that the out-
come should be the same and that Skidmore should trump judicial
precedent.
Mead expressly premised Chevron deference upon a presumption
that Congress vested primary interpretive authority in an agency, such
that the agency’s interpretation enjoys a superior pedigree to that of the
court.
375
Brand X not only post-dated Mead but reiterated and relied
upon the same delegation premise in prioritizing Chevron over stare deci-
sis.
376
Under such circumstances, the Brand X Court had no problem
sacrificing its expressed interpretative preferences to the results of the
FCC’s notice-and-comment process.
377
Under both Skidmore and Mead,
however, where Congress has not vested the agency with primary inter-
pretive authority, expertise and other factors may call for judicial defer-
ence, but the courts rather than the agencies hold ultimate interpretive
authority.
378
Moreover, whether or not the agency possesses the author-
ity to bind with the force of law, it is more difficult to imagine courts
throwing over their own precedents for agency interpretations expressed
through amicus briefs or opinion letters, no matter how thoroughly and
expertly reasoned.
372. See, e.g., Kenneth A. Bamberger, Provisional Precedent: Protecting Flexibility in
Administrative Policymaking, 77 N.Y.U. L. Rev. 1272, 1300–01 (2002) (recognizing tension
between Skidmore and stare decisis); Bressman, supra note 358, at 1466–69 (same);
R
Murphy, Counter-Marbury, supra note 7, at 38–39 (same).
R
373. See Samuel Estreicher & Richard L. Revesz, Nonacquiescence by Federal
Administrative Agencies, 98 Yale L.J. 677, 687 (1989) (recognizing three types of agency
nonacquiescence); see also, e.g., Ross E. Davies, Remedial Nonacquiescence, 89 Iowa L.
Rev. 65, 76 (2003) (noting that agencies engaged in nonacquiescence “might be seen as
participating in the complex separation-of-powers dance in which Congress and the
executive branch play roles equal to the role of the Supreme Court, or at least equal to the
lower federal courts that, are after all, creations of Congress and the executive” (footnote
omitted)).
374. Agencies typically signal such nonacquiescence through policy statements or
informal rulings. See Gary W. Carter, The Commissioner’s Nonacquiescence: A Case for a
National Court of Tax Appeals, 59 Temp. L.Q. 879, 885 (1986) (describing IRS
nonacquiescence through notices called Actions on Decisions); Estreicher & Revesz, supra
note 373, at 694–99 (describing Social Security Administration’s approach to
R
nonacquiescence through agency handbook, interim circular, and informal rulings);
Samuel Figler, Executive Agency Nonacquiescence to Judicial Opinions, 61 Geo. Wash. L.
Rev. 1664, 1680–82 (1993) (detailing NLRB nonacquiescence through informal agency
policy).
375. See United States v. Mead Corp., 533 U.S. 218, 226–27 (2001).
376. See Brand X, 545 U.S. at 982.
377. See id. at 980–81 (noting legally binding effect of FCC interpretation at bar).
378. See Mead, 533 U.S. at 226–27; Skidmore v. Swift & Co., 323 U.S. 134, 137 (1944).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 72 11-OCT-07 15:18
1306 COLUMBIA LAW REVIEW [Vol. 107:1235
Our conceptualization of the Skidmore standard is not a comprehen-
sive solution for resolving all the questions regarding Skidmore’s bounda-
ries. Nevertheless, we believe our approach will offer assistance in deal-
ing with many such issues. Returning to the above examples, and
assuming that an agency is operating within the sphere of its core compe-
tency, both litigation positions and interpretations that contradict ex-
isting judicial precedent should fall within the realm of Skidmore defer-
ence. Reviewing courts may properly apply Skidmore’s factors to give little
or no deference to such interpretations, particularly when one views
Skidmore’s sliding scale as having three potential attitudinal zones.
379
Yet
one can envision a litigation position that is thoroughly reasoned, sub-
stantively valid, and consistently applied over time in like cases as worthy
of at least an intermediate level of Skidmore deference. Likewise, one can
imagine that an agency might attempt to resolve a disagreement in the
courts of appeals over a particular interpretation question by issuing an
official, thoroughly considered but nonbinding notice of its intent to
adopt one of the competing, reasonable interpretations. Perhaps the
courts would be wise in deferring under Skidmore to the agency’s applica-
tion of its expertise in such situations rather than insisting rigidly upon
adherence to judicial precedent absent notice-and-comment rulemaking.
Nevertheless, the courts may properly resolve other potential issues
of Skidmore’s domain in favor of independent judgment. For example,
the Supreme Court in the past has given conflicting signals as to how it
intends to address agency interpretations that are arguably unconstitu-
tional. In the Chevron context, the Court has taken two slightly different
approaches to such questions.
380
On some occasions, the Court has exer-
cised independent judgment in lieu of Chevron deference to evaluate in-
terpretations that presented mere constitutional questions (not necessa-
rily constitutional violations) and thereby avoided the constitutional
inquiry.
381
In Rust v. Sullivan, however, the Court was not so quick to
discard Chevron deference when faced with an arguably unconstitutional
interpretation.
382
Instead, the Court suggested that independent judg-
ment was required only where the agency interpretation would in fact
render the statute unconstitutional; but because the interpretation
379. See supra Part III (conceptualizing Skidmore deference in three zones).
380. See Merrill & Hickman, supra note 42, at 914–15 (discussing relationship
R
between deference doctrines and constitutional avoidance).
381. See Solid Waste Agency v. U.S. Army Corps of Eng’rs, 531 U.S. 159, 172–74
(2001) (declining Chevron deference to interpretation “invok[ing] the outer limits of
Congress’ power” absent “a clear indication that Congress intended that result”); Edward J.
DeBartolo Corp. v. Fla. Gulf Coast Bldg. & Constr. Trades Council, 485 U.S. 568, 574–77
(1988) (“Even if [the agency] construction of the Act were thought to be a permissible
one, . . . we must independently inquire whether there is another interpretation, not
raising these serious constitutional concerns, that may fairly be ascribed to [the Act].”).
382. 500 U.S. 173, 190–91 (1991) (deferring to agency regulations that “do not raise
the sort of ‘grave and doubtful constitutional questions’ that would lead us to assume
Congress did not intend to authorize their issuance” (internal citation omitted)).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 73 11-OCT-07 15:18
2007] MODERN SKIDMORE 1307
passed constitutional muster, the Court afforded it Chevron deference.
383
In either approach, independent judgment plays a significant role in
trumping Chevron deference when the Constitution is at issue. From the
Skidmore perspective, constitutional interpretation is beyond the range of
any agency’s core competency, particularly vis-`
a-vis the judiciary. Thus,
Skidmore review should be applied sparingly, if at all, to such questions.
Finally, Skidmore’s domain is also being extended to another category
of agency interpretations that arguably fall under a different standard en-
tirely—interpretations of the agency’s own regulations. The Supreme
Court has frequently applied a rule, often traced to Seminole Rock & Sand
Co.,
384
that an agency’s interpretation of its regulations receives strong
deference and is “controlling unless ‘plainly erroneous or inconsistent
with the regulation.’”
385
However, the Court has not clearly established
the bounds of Seminole Rock deference—also known as Auer deference—
and has hinted that Skidmore may instead provide the appropriate stan-
dard in certain circumstances.
386
Most notably, in Martin v. Occupational
Safety & Health Review Commission, the Court suggested in dicta that
Skidmore rather than Seminole Rock was the appropriate standard for “less
formal means of interpreting regulations,” embodied in such forms as
“interpretive rules” and “enforcement guidelines,” because they did not
“derive from the exercise of the Secretary’s delegated lawmaking
powers.”
387
The Court’s later reliance in Christensen and Mead on similar ideas of
delegation and interpretive format to draw the line between Chevron and
Skidmore has led some to contend that the same principles also cabin
Seminole Rock deference.
388
The Court did not confirm this contention,
383. See id. at 184–91.
384. 325 U.S. 410 (1945).
385. Auer v. Robbins, 519 U.S. 452, 461 (1997) (quoting Robertson v. Methow Valley
Citizens Council, 490 U.S. 332, 359 (1989) (quoting Seminole Rock, 325 U.S. at 414)); see
also Nat’l Ass’n of Home Builders v. Defenders of Wildlife, 127 S. Ct. 2518, 2537–38 (2007)
(applying same standard in deferring to agency interpretation of its own regulations);
Long Island Care at Home, Ltd. v. Coke, 127 S. Ct. 2339, 2349 (2007) (same); Wash. State
Dep’t of Soc. & Health Servs. v. Guardianship Estate of Keffeler, 537 U.S. 371, 387–88
(2003) (same); Methow Valley Citizens Council, 490 U.S. at 358–59 (same); cf. Gonzales v.
Oregon, 546 U.S. 243, 255–58 (2006) (declining to apply Auer deference to interpretation
of regulation that parroted statute); Christensen v. Harris County, 529 U.S. 576, 588 (2000)
(declining to apply Auer deference because regulation at issue was unambiguous).
386. See Auer, 519 U.S. at 461; see also supra note 385.
R
387. Martin v. Occupational Safety & Health Review Comm’n, 499 U.S. 144, 157
(1991) (citing Skidmore and other cases in explaining that “informal interpretations” of
regulations that might not deserve Seminole Rock deference “are still entitled to some
weight on judicial review”); cf. Gonzales, 546 U.S. at 256–69 (applying Skidmore after
declining Auer and Chevron deference; Court perceived interpretation as construing
statute, not regulation, because regulation merely parroted statute).
388. See, e.g., Krent, supra note 155, at 156–58 (discussing merits of this argument);
R
cf. Keys v. Barnhart, 347 F.3d 990, 993 (7th Cir. 2003) (suggesting that after Christensen and
Mead, agency interpretations promulgated in legal briefs could not receive strong Auer
deference).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 74 11-OCT-07 15:18
1308 COLUMBIA LAW REVIEW [Vol. 107:1235
however, in Long Island Care at Home, Ltd. v. Coke, its most recent applica-
tion of Seminole Rock.
389
There the Court cited Seminole Rock in showing
strong deference to an agency interpretation appearing in an internal
advisory memorandum written in response to litigation, a format that
surely would not receive Chevron deference were it interpreting a
statute.
390
Yet it is clear that Christensen and Mead have led some federal courts
of appeals to invoke Skidmore rather than Seminole Rock to review informal
agency interpretations of their regulations. Some courts cite Christensen
as dictating this result, neglecting Seminole Rock completely and stating
simply that Skidmore applies to agency interpretations lacking legal
force.
391
In this mold, courts have invoked Skidmore when reviewing an
agency’s view of its regulations advanced in opinion letters,
392
legal
briefs,
393
or through the course of administering its regulations.
394
This
may be explained partly by the fact that, as a matter of practice, agency
interpretations of regulations often appear in the very same formats that
Christensen and Mead identified as meriting only Skidmore’s standard when
the interpretation construes a statute.
395
Other courts, though, have read
389. See Long Island Care at Home, 127 S. Ct. at 2349.
390. See id. at 2342–43 (applying Seminole Rock deference despite interpretation’s
format in part because “agency’s course of action indicates that its interpretation . . .
reflects its considered views”); see also Auer, 519 U.S. at 462–63 (stating that fact that “the
Secretary’s interpretation comes to us in the form of a legal brief . . . does not, in the
circumstances of this case, make it unworthy of deference”); cf. Geier v. Am. Honda Motor
Co., 529 U.S. 861, 883–84 (2000) (citing Auer and affording “some weight” to Department
of Transportation’s interpretation of its regulations, as contained in amicus brief).
391. See, e.g., Mercy Catholic Med. Ctr. v. Thompson, 380 F.3d 142, 155 (3d Cir.
2004).
392. See Arriaga v. Fla. Pac. Farms, L.L.C., 305 F.3d 1228, 1237–39 (11th Cir. 2002).
This case is especially noteworthy because the court had no doubt that the letters reflected
the agency’s considered judgment. See id. at 1239 (acknowledging that “the
Administrator’s opinion letters since 1969 consistently have taken the [same] position”).
393. See Union Pac. R.R. v. Cal. Pub. Utils. Comm’n, 346 F.3d 851, 858, 866 (9th Cir.
2003) (concluding that agency’s amicus brief, which argued that its regulations did not
preempt state law, was entitled to at least Skidmore deference).
394. See U.S. Freightways Corp. v. Comm’r, 270 F.3d 1137, 1141–42 (7th Cir. 2001)
(applying Skidmore to interpretations that “ha[ve] emerged inferentially in the way the IRS
has applied the rules to different cases” and “appeared through the litigating positions the
Service has taken”).
395. See United States v. Mead Corp., 533 U.S. 218, 234 (2001) (concluding that
classification rulings, like “‘interpretations contained in policy statements, agency manuals,
and enforcement guidelines,’” lie beyond Chevron’s pale (quoting Christensen v. Harris
County, 529 U.S. 576, 587 (2000))). Illustrating the overlap between Seminole Rock and
Skidmore, in Reutter ex rel. Reutter v. Barnhart, the Eighth Circuit considered the Social
Security Administration’s program manual interpreting its regulations. See 372 F.3d 946,
951 (8th Cir. 2004). The court stated both that the interpretation deserved strong
deference because it was an interpretation of a regulation, and also that it deserved respect
under Skidmore because it was a program manual. See id.
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 75 11-OCT-07 15:18
2007] MODERN SKIDMORE 1309
Christensen and Mead as not precluding Seminole Rock’s application to in-
terpretations lacking legal force.
396
For many, federal appellate courts’ extension of Skidmore into
Seminole Rock’s domain is a welcome development.
397
Critics of Seminole
Rock have argued that the doctrine may tempt agencies to issue vague
regulations through the relatively burdensome notice-and-comment pro-
cess and then clarify the regulations later through less formal—and less
costly—interpretations.
398
Given this and other critiques, Skidmore may
appear as an attractive alternative. Its factors direct courts to consider a
key justification for Seminole Rock—the agency’s expertise and familiarity
with the regulations it has crafted—but also to scrutinize the agency inter-
pretation for unfairness or arbitrariness. If calibrated in the manner we
suggest in this Article, Skidmore’s standard should still result in strong def-
erence to agency interpretations in many cases.
C
ONCLUSION
Without question, Skidmore’s standard of review for administrative in-
terpretations holds a place of great importance in the wake of Christensen
and Mead. In this modern Skidmore era, courts use Skidmore frequently to
evaluate a wide variety of interpretive formats spanning many regulatory
areas. This study answers a number of questions and raises several new
ones. While Skidmore is indeed less deferential than Chevron, it leads
courts to uphold more agency interpretations than was previously as-
sumed. This study demonstrates that most courts understand Skidmore to
prescribe deference that is tailored in accordance to contextual factors.
Comparatively few cases cite Skidmore to justify courts independently fix-
ing the statute’s meaning.
396. See Taylor v. Vt. Dep’t of Educ., 313 F.3d 768, 780 n.7 (2d Cir. 2002) (rejecting
plaintiff’s argument that Christensen precluded Auer deference to agency’s informal policy
letter); Chalenor v. Univ. of N.D., 291 F.3d 1042, 1046–47 (8th Cir. 2002) (same); cf. Air
Brake Sys., Inc. v. Mineta, 357 F.3d 632, 644 (6th Cir. 2004) (contending that “Seminole
Rock deference appears to have survived Mead” but that it nevertheless was inapplicable
because opinion letters of agency’s chief counsel did not represent agency’s authoritative
view).
397. See Robert A. Anthony, The Supreme Court and the APA: Sometimes They Just
Don’t Get It, 10 Admin. L.J. Am. U. 1, 4–11 (1996) (arguing that complete deference to
agencies’ interpretations “allows agencies to take self-serving actions, and to be the judges
in their own causes”); Krent, supra note 155, at 154–55 (addressing criticism of Seminole
R
Rock deference); Manning, supra note 22, at 686–90 (discussing benefits of using Skidmore
R
as doctrinal check on Seminole Rock). But see Angstreich, supra note 66, at 112–28
R
(reasoning that while clarity may suffer under Seminole Rock, agencies are better able to
achieve desired substantive effects when given deference on regulatory interpretation).
398. See Manning, supra note 22, at 655–60 (arguing that Seminole Rock creates
R
heightened risk of regulatory vagueness); see also Thomas Jefferson Univ. v. Shalala, 512
U.S. 504, 519 (1994) (Thomas, J., dissenting) (claiming that Seminole Rock deference in that
case “permit[ted] the Secretary to transform by ‘interpretation’ what self-evidently are
mere generalized expressions of intent into substantive rules of reimbursability”).
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 76 11-OCT-07 15:18
1310 COLUMBIA LAW REVIEW [Vol. 107:1235
Yet within the realm of cases applying the sliding-scale conception of
Skidmore, consistency and coherence is lacking. Courts blur distinctions
between factors and often appear uncertain of the rationale underlying
the various factors. Amidst the confusion, however, certain cases provide
promising examples of more coherent ways to apply Skidmore’s factors.
Recognizing agency expertise and the avoidance of arbitrariness as the
guiding principles behind Skidmore’s multifactor review should help in
stabilizing the courts’ approach to the different factors. And reconceptu-
alizing Skidmore as prescribing a choice of three distinct zones of defer-
ence should simplify its application.
Meanwhile, to some extent, we will simply have to take a “wait and
see” approach to questions of Skidmore’s domain. Nevertheless, with fur-
ther study and discussion, we have high hopes that the modern Skidmore
test will evolve into a useful, stable standard for assessing administrative
interpretations.
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 77 11-OCT-07 15:18
2007] MODERN SKIDMORE 1311
A
PPENDIX
A
PPLICATIONS OF THE
S
KIDMORE
S
TANDARD IN
U.S. C
OURTS OF
A
PPEALS FROM
J
UNE
17, 2001
TO
J
UNE
17, 2006
Model of Skidmore Accept/Reject Case(s) Cited Role of United
Application Applied* Agency View for Standard Factors Discussed** States in Case***
Forrester v. Am. Dieselelectric, Inc., 255 F.3d 1205 (9th Cir. I Accept Christensen None
2001)
Morris v. Bus. Concepts, Inc., 259 F.3d 65 (2d Cir. 2001) IJ Accept Skidmore None
Mead
Cunningham v. Scibana, 259 F.3d 303 (4th Cir. 2001) SS Accept Skidmore Thoroughness Party
Christensen Validity
Consistency
Expertise
Am. Fed’n of Gov’t Employees v. Rumsfeld, 262 F.3d 649 SS Accept Skidmore Validity Party
(7th Cir. 2001) Christensen Consistency
Mead Longstanding
Heartland By-Products, Inc. v. United States, 264 F.3d 1126 SS Accept Skidmore Thoroughness Party
(Fed. Cir. 2001) Mead Formality
Validity
Consistency
Expertise
Matz v. Household Int’l Tax Reduction Inv. Plan, 265 F.3d IJ Reject Skidmore None
572 (7th Cir. 2001) Mead
Landmark Legal Found. v. IRS, 267 F.3d 1132 (D.C. Cir. IJ Accept Skidmore Party
2001) Mead
Rocknel Fastener, Inc. v. United States, 267 F.3d 1354 (Fed. SS Accept Skidmore Thoroughness Party
Cir. 2001) Mead Validity
Consistency
Expertise
Student Loan Fund of Idaho, Inc. v. U.S. Dep’t of Educ., SS Accept Skidmore Validity Party
272 F.3d 1155 (9th Cir. 2001) Mead Expertise
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 78 11-OCT-07 15:18
1312 COLUMBIA LAW REVIEW [Vol. 107:1235
Model of Skidmore Accept/Reject Case(s) Cited Role of United
Application Applied* Agency View for Standard Factors Discussed** States in Case***
Hall v. EPA, 273 F.3d 1146 (9th Cir. 2001) SS Reject Skidmore Thoroughness Party
Mead Consistency
Expertise
Int’l Trading Co. v. United States, 281 F.3d 1268 (Fed. Cir. SS Reject Mead Thoroughness Party
2002) Consistency
Glover v. Standard Fed. Bank, 283 F.3d 953 (8th Cir. 2002) SS Accept Skidmore Validity None
Expertise
Mead Corp. v. United States, 283 F.3d 1342 (Fed. Cir. 2002) SS Reject Skidmore Validity Party
Mead Expertise
Am. Fed’n of Gov’t Employees v. Veneman, 284 F.3d 125 IJ Accept Skidmore Party
(D.C. Cir. 2002)
James v. Von Zemenszky, 284 F.3d 1310 (Fed. Cir. 2002) IJ Reject Mead Party
Franklin v. United States, 289 F.3d 753 (Fed. Cir. 2002) IJ Reject Skidmore Party
Chao v. Russell P. Le Frois Builder, Inc., 291 F.3d 219 (2d SS Accept Skidmore Consistency Party
Cir. 2002) Expertise
Pronsolino v. Nastri, 291 F.3d 1123 (9th Cir. 2002) SS Accept Skidmore Validity Party
Consistency
Expertise
Old Ben Coal Co. v. Dir., Office of Workers’ Comp. IJ Accept Skidmore Party
Programs, 292 F.3d 533 (7th Cir. 2002)
Springfield, Inc. v. Buckles, 292 F.3d 813 (D.C. Cir. 2002) SS Accept Skidmore Thoroughness Party
Validity
Consistency
Expertise
Town of Stratford v. FAA, 292 F.3d 251 (D.C. Cir. 2002) IJ Accept Skidmore Party
Jewelpak Corp. v. United States, 297 F.3d 1326 (Fed. Cir. I Accept Skidmore Party
2002)
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 79 11-OCT-07 15:18
2007] MODERN SKIDMORE 1313
Model of Skidmore Accept/Reject Case(s) Cited Role of United
Application Applied* Agency View for Standard Factors Discussed** States in Case***
Omohundro v. United States, 300 F.3d 1065 (9th Cir. 2002) SS Accept Skidmore Validity Party
Mead Consistency
Heinz v. Cent. Laborers’ Pension Fund, 303 F.3d 802 (7th SS Reject Mead Thoroughness None
Cir. 2002) Validity
Consistency
Lapine v. Town of Wellesley, 304 F.3d 90 (1st Cir. 2002) SS Accept Skidmore Validity None
Consistency
Longstanding
Herman v. Fabri-Centers of Am., Inc., 308 F.3d 580 (6th Cir. IJ Accept Skidmore Party
2002) Mead
Bryson v. Shumway, 308 F.3d 79 (1st Cir. 2002) SS Accept Skidmore Validity None
Mead Expertise
Bank of Am. v. City of San Francisco, 309 F.3d 551 (9th Cir. SS Accept Christensen Formality Amicus
2002) Validity
Cmty. Health Ctr. v. Wilson-Coker, 311 F.3d 132 (2d Cir. SS Accept Mead Validity None
2002) Consistency
Expertise
Cline v. Hawke, 51 F. App’x 392 (4th Cir. 2002) SS Accept Skidmore Thoroughness Party
Formality
Validity
Wells Fargo Bank, N.A. v. Fed. Deposit Ins. Corp., 310 F.3d SS Accept Skidmore Validity Party
202 (D.C. Cir. 2002) Consistency
Expertise
New Orleans Stevedores v. Ibos, 317 F.3d 480 (5th Cir. IJ Accept Skidmore Party
2003) Mead
Ebbert v. DaimlerChrysler Corp., 319 F.3d 103 (3d Cir. SS Reject Skidmore Thoroughness Amicus
2003) Formality
Rosales-Garcia v. Holland, 322 F.3d 386 (6th Cir. 2003) SS Reject Skidmore Validity Party
Consistency
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 80 11-OCT-07 15:18
1314 COLUMBIA LAW REVIEW [Vol. 107:1235
Model of Skidmore Accept/Reject Case(s) Cited Role of United
Application Applied* Agency View for Standard Factors Discussed** States in Case***
Mack v. Otis Elevator Co., 326 F.3d 116 (2d Cir. 2003) IJ Accept Christensen None
Brown v. United States, 327 F.3d 1198 (D.C. Cir. 2003) SS Accept Skidmore Thoroughness Party
Formality
Validity
Expertise
Vernazza v. SEC, 327 F.3d 851 (9th Cir. 2003) SS Accept Skidmore Validity Party
Mead Expertise
Cal. State Legislative Bd. v. Mineta, 328 F.3d 605 (9th Cir. SS Accept Skidmore Thoroughness Party
2003) Mead Validity
Consistency
Longstanding
O’Shaughnessy v. Comm’r, 332 F.3d 1125 (8th Cir. 2003) SS Reject Skidmore Longstanding Party
Mead
Public Citizen, Inc. v. Dep’t of Health & Human Servs., 332 SS Reject Skidmore Thoroughness Party
F.3d 654 (D.C. Cir. 2003) Christensen Validity
Mead
Butterbaugh v. Dep’t of Justice, 336 F.3d 1332 (Fed. Cir. SS Reject Skidmore Thoroughness Party
2003) Mead Formality
Validity
Consistency
Forest Park II v. Hadley, 336 F.3d 724 (8th Cir. 2003) IJ Reject Christensen None
Colorado v. Sunoco, Inc., 337 F.3d 1233 (10th Cir. 2003) SS Accept Skidmore Validity Amicus
Consistency
Expertise
Rubie’s Costume Co. v. United States, 337 F.3d 1350 (Fed. SS Accept Skidmore Thoroughness Party
Cir. 2003) Mead Formality
Validity
Consistency
Expertise
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 81 11-OCT-07 15:18
2007] MODERN SKIDMORE 1315
Model of Skidmore Accept/Reject Case(s) Cited Role of United
Application Applied* Agency View for Standard Factors Discussed** States in Case***
Bolen v. Dengel (In re Dengel), 340 F.3d 300 (5th Cir. SS Accept Skidmore Validity Party
2003) Christensen
IA 80 Group, Inc. v. United States, 347 F.3d 1067 (8th Cir. SS Accept Christensen Validity Party
2003)
O’Brien v. Town of Agawam, 350 F.3d 279 (1st Cir. 2003) SS Accept Skidmore Validity None
Smith v. City of Jackson, 351 F.3d 183 (5th Cir. 2003) SS Reject Christensen Thoroughness None
Collins v. Nat’l Transp. Safety Bd., 351 F.3d 1246 (D.C. Cir. SS Accept Skidmore Validity Party
2003) Expertise
Wilderness Soc’y v. U.S. Fish & Wildlife Serv., 353 F.3d 1051 SS Reject Skidmore Thoroughness Party
(9th Cir. 2003) Mead Validity
Expertise
Malacara v. Garber, 353 F.3d 393 (5th Cir. 2003) SS Reject Skidmore Validity None
Christensen
Research Triangle Reg’l Pub. Transp. Auth. v. United States, SS Accept Skidmore Thoroughness Party
83 F. App’x 505 (4th Cir. 2003) Formality
Validity
Hecht v. Barnhart, 68 F. App’x 244 (2d Cir. 2003) I Accept Skidmore Party
Christensen
Structural Indus., Inc. v. United States, 356 F.3d 1366 (Fed. SS Reject Skidmore Formality Party
Cir. 2004) Mead Consistency
Padash v. INS, 358 F.3d 1161 (9th Cir. 2004) SS Reject Skidmore Thoroughness Party
Christensen Formality
Mead Validity
United States v. City of New York, 359 F.3d 83 (2d Cir. SS Accept Christensen Validity Party
2004)
Tum v. Barber Foods, Inc., 360 F.3d 274 (1st Cir. 2004) SS Reject Skidmore Validity Amicus
Rodriguez v. Whiting Farms, Inc., 360 F.3d 1180 (10th Cir. SS Reject Skidmore Thoroughness None
2004) Christensen Consistency
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 82 11-OCT-07 15:18
1316 COLUMBIA LAW REVIEW [Vol. 107:1235
Model of Skidmore Accept/Reject Case(s) Cited Role of United
Application Applied* Agency View for Standard Factors Discussed** States in Case***
Rabin v. Wilson-Coker, 362 F.3d 190 (2d Cir. 2004) SS Reject Mead Thoroughness None
Formality
Validity
Consistency
Expertise
White v. Burlington N. & Santa Fe Ry. Co., 364 F.3d 789 IJ Reject Skidmore Amicus
(6th Cir. 2004)
Coal. for Gov’t Procurement v. Fed. Prison Indus., Inc., 365 SS Accept Skidmore Validity Party
F.3d 435 (6th Cir. 2004) Christensen
Mead
Cmty. Bank of Ariz. v. G.V.M. Trust, 366 F.3d 982 (9th Cir. SS Accept Skidmore Validity None
2004) Christensen
New Hampshire v. Ramsey, 366 F.3d 1 (1st Cir. 2004) SS Accept Skidmore Validity Party
Mead Longstanding
Ammex, Inc. v. United States, 367 F.3d 530 (6th Cir. 2004) SS Accept Skidmore Formality Party
Mead Validity
Consistency
Expertise
George Harms Constr. Co. v. Chao, 371 F.3d 156 (3d Cir. SS Reject Skidmore Validity Party
2004)
In re New Times Sec. Servs., Inc., 371 F.3d 68 (2d Cir. SS Accept Skidmore Validity Amicus
2004) Mead Consistency
Expertise
Pension Benefit Guar. Corp. v. Wilson N. Jones Mem’l SS Accept Skidmore Thoroughness Party
Hosp., 374 F.3d 362 (5th Cir. 2004) Mead Validity
Consistency
Expertise
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 83 11-OCT-07 15:18
2007] MODERN SKIDMORE 1317
Model of Skidmore Accept/Reject Case(s) Cited Role of United
Application Applied* Agency View for Standard Factors Discussed** States in Case***
Coke v. Long Island Care at Home, Ltd., 376 F.3d 118 (2d SS Reject Skidmore Thoroughness Amicus
Cir. 2004) Validity
Consistency
Longstanding
Horn v. Thoratec Corp., 376 F.3d 163 (3d Cir. 2004) SS Accept Mead Thoroughness Amicus
Consistency
Expertise
Goswami v. Am. Collections Enter., Inc., 377 F.3d 488 (5th I Accept Christensen None
Cir. 2004)
Fed. Nat’l Mortgage Ass’n v. United States, 379 F.3d 1303 SS Reject Skidmore Thoroughness Party
(Fed. Cir. 2004) Longstanding
Perez-Gonzalez v. Ashcroft, 379 F.3d 783 (9th Cir. 2004) SS Reject Christensen Thoroughness Party
Validity
Intercontinental Marble Corp. v. United States, 381 F.3d SS Reject Skidmore Consistency Party
1169 (Fed. Cir. 2004)
La. Envtl. Action Network v. EPA, 382 F.3d 575, 582–84 (5th SS Accept Skidmore Thoroughness Party
Cir. 2004) Christensen Validity
La. Envtl. Action Network v. EPA, 382 F.3d 575, 584–86 (5th SS Reject Skidmore Validity Party
Cir. 2004) Christensen
Howard v. Surface Transp. Bd., 389 F.3d 259 (1st Cir. 2004) IJ Accept Christensen Party
High Sierra Hikers Ass’n v. Blackwell, 390 F.3d 630 (9th Cir. SS Reject Skidmore Thoroughness Party
2004) Mead Formality
Validity
Stephen Bronte, Advisors, LLC v. Commodities Futures SS Accept Skidmore Validity Party
Trading Comm’n, 90 F. App’x 251 (9th Cir. 2004) Christensen Consistency
Beck v. City of Cleveland, 390 F.3d 912 (6th Cir. 2004) SS Accept Skidmore Validity Amicus
Mead Consistency
Kort v. Diversified Collection Servs., Inc., 394 F.3d 530 (7th SS Accept Mead Validity None
Cir. 2005) Expertise
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 84 11-OCT-07 15:18
1318 COLUMBIA LAW REVIEW [Vol. 107:1235
Model of Skidmore Accept/Reject Case(s) Cited Role of United
Application Applied* Agency View for Standard Factors Discussed** States in Case***
Dabit v. Merrill Lynch, Pierce, Fenner & Smith, Inc., 395 IJ Accept Skidmore Amicus
F.3d 25, 34–36 & n.6 (2d Cir. 2005)
Dabit v. Merrill Lynch, Pierce, Fenner & Smith, Inc., 395 IJ Reject Skidmore Amicus
F.3d 25, 39–44 & n.9 (2d Cir. 2005)
Broad. Music, Inc. v. Roger Miller Music, Inc., 396 F.3d 762 IJ Reject Christensen None
(6th Cir. 2005)
Noviello v. City of Boston, 398 F.3d 76 (1st Cir. 2005) I Accept Skidmore None
Christensen
Cathedral Candle Co. v. U.S. Int’l Trade Comm’n, 400 F.3d SS Accept Skidmore Thoroughness Party
1352 (Fed. Cir. 2005) Validity
Consistency
Expertise
Longstanding
Carnival Cruise Lines, Inc. v. United States, 404 F.3d 1312 SS Accept Mead Thoroughness Party
(Fed. Cir. 2005) Validity
Warner-Lambert Co. v. United States, 407 F.3d 1207 (Fed. SS Reject Skidmore Thoroughness Party
Cir. 2005) Validity
De La Mota v. U.S. Dep’t of Educ., 412 F.3d 71 (2d Cir. SS Reject Skidmore Thoroughness Party
2005) Formality
Validity
Expertise
Lin v. U.S. Dep’t of Justice, 416 F.3d 184 (2d Cir. 2005) SS Reject Skidmore Thoroughness Party
St. Mary’s Hosp. v. Leavitt, 416 F.3d 906 (8th Cir. 2005) SS Accept Skidmore Validity Party
Christensen
Santiago v. GMAC Mortgage Group, Inc., 417 F.3d 384 (3d SS Accept Skidmore Thoroughness Amicus
Cir. 2005) Validity
Expertise
Packard v. Pittsburgh Transp. Co., 418 F.3d 246 (3d Cir. SS Reject Skidmore Thoroughness None
2005)
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 85 11-OCT-07 15:18
2007] MODERN SKIDMORE 1319
Model of Skidmore Accept/Reject Case(s) Cited Role of United
Application Applied* Agency View for Standard Factors Discussed** States in Case***
Fujitsu Am., Inc. v. United States, 422 F.3d 1364 (Fed. Cir. I Accept Mead Party
2005)
Warner-Labmert Co. v. United States, 425 F.3d 1381 (Fed. SS Accept Skidmore Thoroughness Party
Cir. 2005) Validity
Consistency
S. Utah Wilderness Alliance v. Bureau of Land Mgmt., 425 SS Reject Skidmore Consistency Party
F.3d 735 (10th Cir. 2005) Christensen
Zhang v. Gonzales, 426 F.3d 540 (2d Cir. 2005) IJ Reject Skidmore Party
Padilla-Caldera v. Gonzales, 426 F.3d 1294 (10th Cir. 2005) IJ Reject Christensen Party
Shikles v. Sprint/United Mgmt. Co., 426 F.3d 1304 (10th SS Reject Skidmore Thoroughness Amicus
Cir. 2005) Formality
Validity
Longstanding
OfficeMax, Inc. v. United States, 428 F.3d 583 (6th Cir. SS Reject Skidmore Thoroughness Party
2005) Christensen Validity
Mead Consistency
Longstanding
United States v. W.R. Grace & Co., 429 F.3d 1224 (9th Cir. SS Accept Skidmore Thoroughness Party
2005) Validity
Expertise
M. Fortunoff of Westbury Corp. v. Peerless Ins. Co., 432 SS Accept Skidmore Validity Amicus
F.3d 127 (2d Cir. 2005) Christensen
Trowell v. Beeler, 135 F. App’x 590 (4th Cir. 2005) IJ Reject Christensen Party
Reimels v. Comm’r, 436 F.3d 344 (2d Cir. 2006) SS Accept Skidmore Longstanding Party
McGraw v. Barnhart, 450 F.3d 493 (10th Cir. 2006) SS Accept Skidmore Thoroughness Party
Validity
Consistency
Appoloni v. United States, 450 F.3d 185 (6th Cir. 2006) I Accept Skidmore Party
\\server05\productn\C\COL\107-6\COL601.txt unknown Seq: 86 11-OCT-07 15:18
1320 COLUMBIA LAW REVIEW [Vol. 107:1235
Model of Skidmore Accept/Reject Case(s) Cited Role of United
Application Applied* Agency View for Standard Factors Discussed** States in Case***
Caribbean Ispat Ltd. v. United States, 450 F.3d 1336 (Fed. SS Reject Skidmore Thoroughness Party
Cir. 2006) Mead Formality
Consistency
* Model of Skidmore Applied:
IJ — Independent judgment model of the Skidmore standard.
SS — Sliding-scale model of the Skidmore standard.
I — Indeterminate: application of Skidmore fit neither the independent judgment model nor the sliding-scale model.
** Because courts employ varying notions of what Skidmore’s factors mean, our assessment of whether a court was applying a given factor entailed a measure of
subjectivity. Accordingly, this list of what factors a court applied should not be taken as absolute or definitive. For discussion of the varying ideas of what the
factors mean, see Part II.C.
*** Role of United States in Case:
Party — The agency was a party to the litigation.
Amicus — The agency submitted an amicus brief but was not a party.
None — The agency did not participate in the litigation.
