Montana Environmental Information Center v. Debra H. Thomas, in her capacity as Acting Regional Administrator, United States…

FOR PUBLICATION
UNITED STATES COURT OF APPEALS
FOR THE NINTH CIRCUIT
MONTANA ENVIRONMENTAL
INFORMATION CENTER,
Petitioner,
v.
DEBRA H. THOMAS, in her capacity
as Acting Regional Administrator,
United States Environmental
Protection Agency, Region 8; U.S.
ENVIRONMENTAL PROTECTION
AGENCY,
Respondents,
MONTANA DEPARTMENT OF
ENVIRONMENTAL QUALITY; TALEN
MONTANA, LLC,
Respondents-Intervenors.
No. 16-71933
OPINION
On Petition for Review of an Order of the
Environmental Protection Agency
Argued and Submitted February 14, 2018
San Francisco, California
Filed August 30, 2018

-- 1 of 17 --

MONTANA ENVTL. INFO CTR. V. THOMAS 2
Before: Mary M. Schroeder and Johnnie B. Rawlinson,
Circuit Judges, and William K. Sessions III,* District
Judge.
Opinion by Judge Rawlinson
SUMMARY**
Environmental Law
The panel denied a petition for review challenging an
action of the United States Environmental Protection Agency
approving a 1994 revision to Montana’s State Implementation
Plan.
Petitioner alleged that the EPA’s approval was arbitrary
and capricious because the Montana Department of
Environmental Quality (“DEQ”) interpreted “actual
emissions” less stringently than the Clean Air Act would
allow. The DEQ’s interpretation was advanced in unrelated
litigation (the “Talen case”). Petitioner further alleged that
Montana’s 1994 Revised State Implementation plan was
deficient, and the EPA should not have approved the 2008-
2015 revisions until the state definition of “actual emissions”
complied with federal standards. Petitioner submitted a
comment during the EPA’s notice and comment period
* The Honorable William K. Sessions III, United States District Judge
for the District of Vermont, sitting by designation.
** This summary constitutes no part of the opinion of the court. It has
been prepared by court staff for the convenience of the reader.

-- 2 of 17 --

MONTANA ENVTL. INFO CTR. V. THOMAS 3
following Montana’s submission on December 17, 2015 of its
final implementation plan.
The panel held that the language of the Implementation
Plan concerning the “two year period” in determining a
source’s actual emissions was ambiguous where the DEQ and
the EPA reasonably interpreted the phrase to mean two
different things. The panel further held that it was
appropriate to give deference to the EPA’s reasonable
interpretation, consistent with the deference given under
Chevron to the EPA’s rulemaking authority. The panel
agreed with the EPA that petitioner’s comment raised a
question of implementation of a program rather than approval
of a plan, and as such, DEQ’s statements in the Talen case
need not be resolved at the approval phase of the state plan.
The panel held that because the EPA’s interpretation of
ambiguous text in the 1994 Revised Implementation Plan was
a permissible one, and because the EPA’s interpretation
controlled, its approval of the succeeding 2015
Implementation Plan was not arbitrary or capricious. The
2015 Implementation Plan was otherwise in conformance
with the EPA’s Prevention of Significant Deterioration
program under the Clean Air Act.

-- 3 of 17 --

MONTANA ENVTL. INFO CTR. V. THOMAS 4
COUNSEL
George E. Hays (argued), San Francisco, California; Derf
Johnson, Montana Environmental Information Center,
Helena, Montana; for Petitioner.
Sheila Baynes (argued) and Jeffrey H. Wood, Acting
Assistant Attorney General; Environmental Defense Section,
Environment & Natural Resources Division, United States
Department of Justice, Washington, D.C.; Kristi M. Smith,
Brian Doster, Zach Pilchen, and Melina Williams, Office of
General Counsel, Environmental Protection Agency,
Washington, D.C.; for Respondents.
Norman J. Mullen (argued), Montana Department of
Environmental Quality, Helena, Montana, for Respondent-
Intervenor Montana Department of Environmental Quality.
Joshua Frank (argued), Baker Botts LLP, Washington, D.C.,
for Respondent-Intervenor Talen Montana LLC.

-- 4 of 17 --

MONTANA ENVTL. INFO CTR. V. THOMAS 5
OPINION
RAWLINSON, Circuit Judge:
Petitioner, Montana Environmental Information Center
(Information Center), challenges an action of the United
States Environmental Protection Agency (Agency) approving
a 1994 revision to Montana’s State Implementation Plan
(Implementation Plan). Information Center asserts that the
Agency’s approval was arbitrary and capricious because
Montana interprets one of its provisions less stringently than
the Clean Air Act would allow. We have jurisdiction to
review the EPA’s action under 42 U.S.C. § 7607(b)(1), and
deny the petition for review.
I. Statutory and Regulatory Background
A. Clean Air Act
Congress passed the Clean Air Act to protect and enhance
the quality of the nation’s air. See 42 U.S.C. § 7401(b)(1).
To achieve this, “the States and the Federal Government
partner[ed] in the struggle against air pollution.” General
Motors Corp. v. United States, 496 U.S. 530, 532 (1990). We
have deemed this partnership “a uniquely important system
of cooperative federalism in the quest for clean air.”
Committee for a Better Arvin v. EPA, 786 F.3d 1169, 1173
(9th Cir. 2015) (citation omitted).
B. Federal-State Partnership
One of the Agency’s primary responsibilities under the
Clean Air Act is to identify air pollutants that endanger the
public health and welfare. See 42 U.S.C. § 7408(a). Once

-- 5 of 17 --

MONTANA ENVTL. INFO CTR. V. THOMAS 6
identified, the Agency must then set National Ambient Air
Quality Standards (Air Quality Standards), which specify the
maximum allowable concentration of those pollutants in the
atmosphere. See id. § 7409. The Air Quality Standards are
subject to periodic review and revision. See id. In sum, the
federal government’s role in the federal-state partnership is to
combat air pollution by identifying pollutants and then setting
(and updating) Air Quality Standards.
But what of the states? “The [Clean Air Act] requires the
states to submit State Implementation Plans, or ‘SIPs,’
showing how the states will attain [Air Quality Standards] . . .
” El Comite Para el Bienestar de Earlimart v. EPA, 786 F.3d
688, 692 (9th Cir. 2015) (citing 42 U.S.C. § 7410(a)(1)); see
also Whitman v. Am. Trucking Assn’s, Inc., 531 U.S. 457, 470
(2001) (“It is to the States that the [Clean Air Act] assigns
initial and primary responsibility for deciding what emissions
reductions will be required from which sources. . . .”)
(citations omitted). These Implementation Plans must “make
demonstrations (of how attainment, maintenance, and
progress will be achieved) and [] provide a control strategy
that will achieve the necessary reductions and otherwise meet
the requirements of the Act.” Hall v. EPA, 273 F.3d 1146,
1153 (9th Cir. 2001), as amended (citation omitted). Each
plan must “specify the manner in which [Air Quality
Standards] will be achieved and maintained within each air
quality control region.” 42 U.S.C. § 7407(a).
C. State Implementation Plans
If the Agency determines that a proposed Implementation
Plan meets the applicable requirements, then that
Implementation Plan “bec[o]me[s] federal law, not state
law.” Safe Air For Everyone v. EPA, 488 F.3d 1088, 1097

-- 6 of 17 --

MONTANA ENVTL. INFO CTR. V. THOMAS 7
(9th Cir. 2007), as amended (emphasis in the original). That
is true in part because Implementation Plans cannot be
amended without the EPA’s further approval. See id.
When the Agency updates its Air Quality Standards,
states have three years to revise their Implementation Plans
to comply with the new standards. See 42 U.S.C.
§ 7410(a)(1). “These revisions need not be wholesale
recastings of [Implementation Plans]; instead, the [Clean Air
Act] allows the states to submit, and [the] Agency to review,
piecemeal amendments dealing with discrete [Implementation
Plan] provisions, leaving most of the plan untouched.” Safe
Air, 488 F.3d at 1092 (citation omitted).
In addition, Implementation Plans must comply with the
Clean Air Act’s Prevention of Significant Deterioration
program (PSD program). See 40 C.F.R. § 51.166(a)(1)
(“[E]ach applicable State Implementation Plan . . . shall
contain emission limitations and such other measures as may
be necessary to prevent significant deterioration of air
quality.”). The purpose of the PSD program is to ensure that
major sources of air pollution do not degrade areas that either
meet Air Quality Standards (also known as “attainment”
areas) or cannot be classified as meeting—or not meeting—
Air Quality Standards (also known as “unclassifiable” areas).
42 U.S.C. § 7407(d)(1). The PSD program combats
degradation of these areas by requiring developers to acquire
permits before constructing new sources of emissions or
modifying existing ones. See 40 C.F.R. § 51.166(a)(7). The
permit requirement applies to existing sources that are
planned to undergo a “major modification,” id., resulting in
a “significant emissions increase” and a “significant net
emissions increase.” Id. § 51.166(a)(1)(7)(iv)(a).
“Significant emissions increases” are calculated by taking the

-- 7 of 17 --

MONTANA ENVTL. INFO CTR. V. THOMAS 8
“actual emissions” of a source, which establishes a
representative baseline level of emissions, id.
§ 51.166(b)(47), and comparing that baseline to the projected
emissions, post-modification. See id. § 51.166 (a)(7)(iv)(c).
This formula evokes the question of how to determine a
source’s “actual emissions.” The answer has evolved over
time. In 1980, the Agency stated that actual emissions
shall equal the average rate, in tons per year,
at which the unit actually emitted the pollutant
during a two-year period which precedes the
particular date and which is representative of
normal source operation. The reviewing
authority may allow the use of a different time
period upon a determination that it is more
representative of normal source operation.
40 C.F.R. § 51.24(b)(21)(1980).
In 1992, the Agency clarified that “a 2-year period which
precedes the particular date” has “historically [meant] the
2 years immediately preceding the proposed change to
establish the baseline.” 57 Fed. Reg. 32,314, 32,323 (July 21,
1992) (citations omitted). Going forward, however, “the
[Agency] would presume that any 2 consecutive years within
the 5 years prior to the proposed change is representative.”
Id.
Ten years later, the Agency formally updated its
definition of “baseline actual emissions.” This update
involved creating separate definitions for steam power plants
and all other stationary sources. As of 2002, the baseline
actual emissions of steam power plants equaled the average

-- 8 of 17 --

MONTANA ENVTL. INFO CTR. V. THOMAS 9
rate of a pollutant actually emitted “during any consecutive
24-month period selected by the owner or operator within the
5-year period immediately preceding [the] . . . actual
construction of the project. The reviewing authority shall
allow the use of a different time period upon a determination
that it is more representative of normal source operation.”
40 C.F.R. § 51.166(b)(47)(i). The baseline actual emissions
of non-steam sources equaled omissions occurring “during
any consecutive 24-month period . . . within the 10-year
period immediately preceding” construction of a project. Id.
§ 51.166(b)(47)(ii).
Although the Clean Air Act requires each Implementation
Plan to contain permitting processes that comply with the
PSD program, the Act does not require verbatim adoption of
the PSD program. Rather, a state is free to deviate so long as
it “specifically demonstrates” that those deviations “are more
stringent than or at least as stringent in all respects as the
corresponding provisions [of the Clean Air Act.]” 40 C.F.R.
51.166(a)(7)(iv).
II. Factual and Procedural Background
Montana, “like every other state, was first required to
submit [an Implementation Plan] to the [Agency] within
thirteen months of the Act’s . . . passage.” Safe Air, 488 F.3d
at 1093 (citing Train v. NRDC, 421 U.S. 60, 65 (1975). In
addition, Montana is required to revise its Implementation
Plan concomitantly with EPA updates to federal standards.
See 42 U.S.C. § 7410. Relevant to this case is a revision
Montana submitted on March 30, 1994 (1994 Revised
Implementation Plan). The revision contained the following
definition of “actual emissions:”

-- 9 of 17 --

MONTANA ENVTL. INFO CTR. V. THOMAS 10
Actual emissions as of a particular date shall
equal the average rate, in tons per year, at
which the unit actually emitted the pollutant
during a two-year period which precedes the
particular date and which is representative of
normal source operation. The department may
determine that a different time period is more
representative of normal source operation. . . .
This definition, in large part, mirrors the Agency’s 1980
definition of actual emissions, see 40 C.F.R. § 51.24(b)(21)
(1980), and was approved by the Agency in 1995. See
60 Fed. Reg. 36,715, 36,719 (July 18, 1995).
Between 2008 and 2012, the Agency updated its Air
Quality Standards. See 81 Fed. Reg. 4225 (Jan. 26, 2016).
These updates triggered Montana’s obligation to revise its
Implementation Plan within three years. See 42 U.S.C.
§ 7410. On December 17, 2015, Montana submitted its final
Implementation Plan in response to the 2008–2012 updates.
This Implementation Plan contained the same emissions
definition that was approved by the Agency in 1995. During
the Agency’s subsequent notice and comment period,
Information Center submitted a comment to the effect that
Montana interpreted the definition of “actual emissions” from
its 1994 Revised Implementation Plan less stringently than
the PSD program requires.
The basis for Information Center’s comment was the
interpretation of “actual emissions” advanced by the Montana
Department of Environmental Quality (DEQ) in unrelated
litigation. See Sierra Club and MEIC v. Talen Montana,
LLC, CV13-32-BLG-DLC-JCL, 2015 WL 13714343 (D.
Mont. Dec. 31, 2015). In Talen, the DEQ acknowledged that

-- 10 of 17 --

MONTANA ENVTL. INFO CTR. V. THOMAS 11
the Information Center and the Agency read the definition of
“actual emissions” to mean “‘the’ two-year period
immediately preceding” a modification. DEQ argued that
“no deference [should be given] to [the Agency’s
interpretation] because . . . the interpretation that [it is] ‘the’
two-year period immediately preceding [a modification] is
inconsistent within the rule language which says . . . ‘a’ two-
year period.” (emphases added).
The Agency responded that it “appreciates and takes
seriously [Information Center’s comment] that Montana has
adopted ‘policy interpretations’ outside the context of the
[Implementation Plan] that may undermine the State’s
implementation of the [Implementation Plan] as approved by
the [Agency].” Still, the Agency did not find it necessary to
“evaluat[e] the merits of these assertions concerning
implementation of the [Implementation Plan] in the context
of this action” because “this action involves a review of the
[Implementation Plan] itself.” Accordingly, the Agency
expressed its “inten[t] to evaluate the merits of these
assertions, separate from this action, at a future time.” The
Agency determined that the previously approved 1994
Revised Implementation Plan and the 2015 Implementation
Plan “[met] the relevant structural requirements,” and
approved the 2015 Implementation Plan. Information Center
timely petitioned this court to review the Agency’s action.
III. Standard of Review
We review the approval of an Implementation Plan “by
considering whether the [Agency’s] decision was arbitrary,
capricious, an abuse of discretion, or contrary to law.”
Committee for a Better Arvin, 786 F.3d at 1174–75 (citations
omitted); see also Hall, 273 F.3d at 1155. Whether the

-- 11 of 17 --

MONTANA ENVTL. INFO CTR. V. THOMAS 12
Agency acted arbitrarily and capriciously “rests on whether
it articulated a rational connection between the facts found
and the choice made.” Friends of Yosemite Valley v. Norton,
348 F.3d 789, 793 (9th Cir. 2003) (citation and internal
quotation marks omitted), opinion clarified, 366 F.3d 731
(9th Cir. 2004). In conducting this review, we may neither
“rubber-stamp administrative decisions” nor “substitute our
judgment for that of the agency.” Id. (citation and alterations
omitted).
“With respect to the [Clean Air Act], Congress has given
[the Agency] general rulemaking authority, 42 U.S.C.
§ 7601(a)(1), which, when exercised, requires our deference
in accordance with Chevron. . . .” Sierra Club v. EPA,
671 F.3d 955, 962 (9th Cir. 2012) (citation and internal
quotation marks omitted).
IV. Analysis
A. Montana Department of Environmental Quality
Statements
The entirety of Information Center’s appeal rests on the
DEQ’s statements in the Talen litigation. Information Center
argues that Montana’s 1994 Revised Implementation Plan did
not comply with the Clean Air Act, because the DEQ
interprets “actual emissions” less stringently than federal
standards would allow. The DEQ’s interpretation,
Information Center contends, carries the force of law.
Information Center maintains that Montana’s 1994 Revised
Implementation Plan was therefore deficient, and that the
Agency should not have approved the 2008–2015 revisions
until the state definition of “actual emissions” complied with
federal standards.

-- 12 of 17 --

MONTANA ENVTL. INFO CTR. V. THOMAS 13
Information Center relies on Go v. Holder, 744 F.3d 604,
611 (9th Cir. 2014), to support its argument that the DEQ’s
statements have the force of law. But reliance on Go is
misplaced in the circumstances of this case. It is well settled
that once the Agency approves either an Implementation Plan
or a Revised Implementation Plan, that plan becomes federal
law. See Committee for a Better Arvin, 786 F.3d at 1174. As
such, “a state may not unilaterally alter the legal
commitments of its [Implementation Plan] once [the Agency]
approves the plan.” Safe Air, 488 F.3d at 1097 (citation
omitted). Thus, the DEQ’s interpretation of “actual
emissions” could not invalidate Montana’s 1994 Revised
Implementation Plan. And, where the Agency has officially
interpreted a vague regulatory term, the Agency’s
interpretation prevails. See Auer v. Robbins, 519 U.S. 452,
461 (1997).
B. Implementation of Program As Opposed to Approval
of Plan
The Clean Air Act “permits the [Agency] to issue ‘partial
approvals’ [of Implementation Plans]. Hall, 273 F.3d at
1159 (citation and alteration omitted). In light of this
circumstance, the question becomes to what extent the
Agency must consider the DEQ’s statements as part of the
approval process for an Implementation Plan.
“In interpreting [an Implementation Plan], we begin with
a look toward the plain meaning of the plan and stop there if
the language is clear. . . .” Safe Air, 488 F.3d at 1095. The
language at issue is:
(1) “Actual emissions as of a particular date
shall equal . . . a two-year period which

-- 13 of 17 --

MONTANA ENVTL. INFO CTR. V. THOMAS 14
precedes the particular date [of construction,]”
and
(2) “The department may determine that a
different time period is more representative of
normal source operation.”
Under Information Center’s theory, the DEQ and the Agency
reasonably interpreted “a two-year period” to mean two
different things. Therefore, the language at issue is
ambiguous. See NRDC v. Cnty. of Los Angeles, 725 F.3d
1194, 1205 (9th Cir. 2013) (explaining that language is
ambiguous “if reasonable people could find its terms
susceptible to more than one interpretation”) (citation
omitted). And while Safe Air, 488 F.3d at1095–96, provides
guidance for instances where Implementation Plan language
is clear, it does not do so for Implementation Plan language
that is unclear. We do so here: Where the plain meaning of
an Implementation Plan cannot be readily discerned from the
text, we think it appropriate to give deference to the Agency’s
reasonable interpretation. This approach is consistent with
the deference given under Chevron to the Agency’s
rulemaking authority, and with how we have reviewed the
Agency’s interpretation of Implementation Plan-related
documents. See, e.g., El Comite Para el Bienestar de
Earlimart, 786 F.3d at 696 (“Because the plain language of
the relevant documents is ambiguous, we defer to the EPA’s
interpretation if it is reasonable, i.e., if it ‘sensibly conforms
to the purpose and wording of the regulations.’”) (citation
omitted).
The Agency’s interpretation of the two-year period
immediately preceding the commencement of construction is
consistent with the tenor of the regulation to fix a finite

-- 14 of 17 --

MONTANA ENVTL. INFO CTR. V. THOMAS 15
period for the measurement of baseline emissions. See
40 C.F.R. § 51.24(b)(2). In any event, we agree with the
Agency that Information Center’s comment raises a question
of implementation of a program rather than approval of a
plan. As such, DEQ’s statements in the Talen case need not
be resolved at the approval phase of the state plan. As
explained above, DEQ’s policy interpretations do not carry
the force of law, contrary to Information Center’s contention.
See Committee for a Better Arvin, 786 F.3d at 1174 (“Once
approved by [the Agency, an Implementation Plan] becomes
federal law, and cannot be changed unless and until [the
Agency] approves any change. . . .”) (citation, alterations and
internal quotation marks omitted). At this point, we are not
even sure if Montana will adhere to its interpretation of the
plan language at issue should the occasion arise to implement
that language. As matters currently stand, Montana still has
the option to change course and adopt the Agency’s
interpretation that “a 2-year period which precedes the
particular date” means “the 2 years immediately preceding
the proposed change.” 57 Fed. Reg. at 32,323. As a result,
we cannot say at this juncture that the Agency acted
arbitrarily in approving the 1994 Revised Implementation
Plan.
At bottom, we conclude that the Agency’s interpretation
of the regulation was a reasonable one. See Auer, 519 U.S. at
461. The EPA interpreted “a 2-year period which precedes
the particular date” to mean “the 2 years immediately
preceding” the particular date. 57 Fed. Reg. at 32,323. In
Talen, Montana argued that this interpretation is
irreconcilable with the text because the text employs the
indefinite article “a,” not the definite article “the.” Although
the Agency’s interpretation does incorporate a definite article,
the fact that the language is ambiguous, as discussed, gave the

-- 15 of 17 --

MONTANA ENVTL. INFO CTR. V. THOMAS 16
Agency leeway to pose a reasonable interpretation of the
language. See Comite Para el Bienstar de Earlimart,
786 F.3d at 696.
Because the Agency’s interpretation of ambiguous text in
the 1994 Revised Implementation Plan was a permissible one,
and because the Agency’s interpretation controlled, its
approval of the succeeding 2015 Implementation Plan was
not arbitrary or capricious. See Committee for a Better Arvin,
786 F.3d at 1174–75. The 2015 Implementation Plan was
otherwise in conformance with the Agency’s PSD program
under the Clean Air Act. See 40 C.F.R. 51.166(a)(7)(iv)
(requiring Implementation Plan compliance with the
Agency’s PSD program).
V. Conclusion
The Agency’s interpretation of “a 2-year period which
precedes the particular date” was a permissible one. 57 Fed.
Reg. at 32,323. DEQ’s contrary interpretation had no effect
on the Agency’s approval process. Accordingly, the
Agency’s approval of Montana’s 2015 Implementation Plan
was neither arbitrary nor capricious, and Information Center’s
comment regarding Montana’s interpretation of the language
in question raised a question of implementation, better
addressed at a different time. As the Agency mentioned in its
response to Information Center’s comment, “there are
multiple statutory tools that the [Agency] can use to rectify
problems with state implementation of its [Implementation
Plan]. For example, the [Clean Air Act] provides the
[Agency] the authority to issue [an Implementation Plan] call,
42 U.S.C. § 7410(k)(5); make a finding of failure to
implement, id. §§ 7410(m), 7509(a)(4); and take measures to

-- 16 of 17 --

MONTANA ENVTL. INFO CTR. V. THOMAS 17
address specific permits pursuant to the [Agency’s] case-by-
case permitting oversight. See, e.g., id. § 7661d(b).”
PETITION FOR REVIEW DENIED.

-- 17 of 17 --