Bahr v. EPA, No. 14-72327 (9th Cir. 2016)

Annotate this Case
Justia Opinion Summary

Petitioners seek review of the EPA's rule approving Arizona’s Five Percent Plan for airborne particulate matter around Maricopa County. The court upheld the EPA’s determination that the control measures in Arizona’s Five Percent Plan did not need to be updated, and that the 135 exceedances were exceptional events that are excluded from consideration under the EPA’s regulation and guidance documents. The court concluded, however, that it will not defer to the EPA’s interpretation of the contingency measures requirement because, under the plain language of 42 U.S.C. 7502(c)(9), contingency measures are measures that will be taken in the future, not measures that have already been implemented. Accordingly, the court granted in part and denied in part the petition for review.

Court Description: Environmental Law. The panel granted in part and denied in part a petition for review of an order of the United States Environmental Protection Agency approving Arizona’s Five Percent Plan for airborne particulate matter around Maricopa County, promulgated under the Clean Air Act. Arizona submitted a new State Implementation Plan revision on May 25, 2012 – the Five Percent Plan – to achieve a five percent annual reduction in PM-10, a harmful air pollutant. Petitioners alleged that the EPA acted contrary to law by failing to require that Arizona include an updated analysis of best available control measures and most stringent measures in the Five Percent Plan, excluding 135 exceedances from the monitoring data as “exceptional events,” and allowing BAHR V. U.S. EPA 3 Arizona to satisfy the “contingency measures” requirement with previously implemented control measures. The panel held that it would apply Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984), deference to the EPA’s interpretation of the Clean Air Act issued in connection with a State Implementation Plan approval. The panel upheld the EPA’s determination that the control measures in Arizona’s Five Percent Plan did not need to be updated, and that the 135 exceedances were exceptional events that were excluded from consideration under the EPA’s regulation and guidance documents. The panel did not defer to the EPA’s interpretation of the contingency measures requirement, however, because under the plain language of 42 U.S.C. § 7502(c)(9) contingency measures are measures that will be taken in the future, not measures that have already been implemented. The panel remanded to the EPA for further consideration of this portion of the State Implementation Plan, but otherwise denied the petition. Judge Clifton concurred in sections I-IV of the majority opinion, and dissented from the majority’s conclusion in section V that EPA’s approval of the contingency measures in Arizona’s State Implementation Plan was contrary to the clear language of the Clean Air Act. In his view, the scope of the Clean Air Act’s contingency measures requirement was ambiguous and EPA’s reasonable interpretation of that requirement was entitled to deference. 4 BAHR V. U.S. EPA

Download PDF
FOR PUBLICATION UNITED STATES COURT OF APPEALS FOR THE NINTH CIRCUIT SANDRA L. BAHR; DAVID MATUSOW, No. 14-72327 Petitioners, OPINION v. U.S. ENVIRONMENTAL PROTECTION AGENCY; GINA MCCARTHY, Administrator, United States Environmental Protection Agency; JARED BLUMENFELD, Regional Administrator, EPA Region IX, Respondents, STATE OF ARIZONA, Respondent-Intervenor. On Petition for Review of an Order of the Environmental Protection Agency Argued and Submitted June 17, 2016 San Francisco, California Filed September 12, 2016 2 BAHR V. U.S. EPA Before: Richard R. Clifton and Sandra S. Ikuta, Circuit Judges, and William Q. Hayes,* District Judge. Opinion by Judge Ikuta; Partial Concurrence and Partial Dissent by Judge Clifton SUMMARY** Environmental Law The panel granted in part and denied in part a petition for review of an order of the United States Environmental Protection Agency approving Arizona’s Five Percent Plan for airborne particulate matter around Maricopa County, promulgated under the Clean Air Act. Arizona submitted a new State Implementation Plan revision on May 25, 2012 – the Five Percent Plan – to achieve a five percent annual reduction in PM-10, a harmful air pollutant. Petitioners alleged that the EPA acted contrary to law by failing to require that Arizona include an updated analysis of best available control measures and most stringent measures in the Five Percent Plan, excluding 135 exceedances from the monitoring data as “exceptional events,” and allowing * The Honorable William Q. Hayes, United States District Judge for the Southern District of California, 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. BAHR V. U.S. EPA 3 Arizona to satisfy the “contingency measures” requirement with previously implemented control measures. The panel held that it would apply Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984), deference to the EPA’s interpretation of the Clean Air Act issued in connection with a State Implementation Plan approval. The panel upheld the EPA’s determination that the control measures in Arizona’s Five Percent Plan did not need to be updated, and that the 135 exceedances were exceptional events that were excluded from consideration under the EPA’s regulation and guidance documents. The panel did not defer to the EPA’s interpretation of the contingency measures requirement, however, because under the plain language of 42 U.S.C. § 7502(c)(9) contingency measures are measures that will be taken in the future, not measures that have already been implemented. The panel remanded to the EPA for further consideration of this portion of the State Implementation Plan, but otherwise denied the petition. Judge Clifton concurred in sections I-IV of the majority opinion, and dissented from the majority’s conclusion in section V that EPA’s approval of the contingency measures in Arizona’s State Implementation Plan was contrary to the clear language of the Clean Air Act. In his view, the scope of the Clean Air Act’s contingency measures requirement was ambiguous and EPA’s reasonable interpretation of that requirement was entitled to deference. 4 BAHR V. U.S. EPA COUNSEL Joy E. Herr-Cardillo (argued) and Timothy M. Hogan, Arizona Center for Law in the Public Interest, Tucson, Arizona, for Petitioners. Alan D. Greenberg (argued), Attorney; Sam Hirsch, Acting Assistant Attorney General; Environment & Natural Resources Division, United States Department of Justice, Denver, Colorado; Geoffrey Wilcox, Office of General Counsel; Kara Christenson, Office of Regional Counsel, Region 9; United States Environmental Protection Agency, San Francisco, California; for Respondents. Monique Coady, Assistant Attorney General, Office of the Attorney General, Phoenix, Arizona, for RespondentIntervenor. OPINION IKUTA, Circuit Judge: Sandra Bahr and David Matusow petition for review of a final rule issued by the Environmental Protection Agency (EPA) approving Arizona’s Five Percent Plan for airborne particulate matter around Maricopa County. They argue that the EPA erred in approving this plan because it did not include best available control measures (BACM) and most stringent control measures (MSM) as of 2012. The petitioners also argue that the EPA failed to follow its own published guidance in approving Arizona’s claim that 135 exceedances of the air emission standard could be excluded from consideration. See 42 U.S.C. § 7619(b)(1)(A). Finally, BAHR V. U.S. EPA 5 the petitioners argue that the EPA’s approval of the contingency measures included in Arizona’s Five Percent Plan was contrary to 42 U.S.C. § 7502(c)(9) because the measures had already been implemented. We uphold the EPA’s determination that the control measures in Arizona’s Five Percent Plan did not need to be updated, and that the 135 exceedances were exceptional events that are excluded from consideration under the EPA’s regulation and guidance documents. We do not defer to the EPA’s interpretation of the contingency measures requirement, however, because under the plain language of § 7502(c)(9) contingency measures are measures that will be taken in the future, not measures that have already been implemented. I We begin by briefly describing the statutory framework. The Clean Air Act (CAA), 42 U.S.C. § 7401 et seq., establishes “cooperative Federal, State, regional, and local programs to prevent and control air pollution,” id. § 7401(a)(4). Under the CAA, the EPA is required to “publish . . . a list which includes each air pollutant . . . emissions of which, in [the EPA’s] judgment, cause or contribute to air pollution which may reasonably be anticipated to endanger public health or welfare.” Id. § 7408(a)(1). The EPA is then required to “prescrib[e] a national primary ambient air quality standard” (NAAQS) for that pollutant. Id. § 7409(a). One such harmful air pollutant is “PM-10,” defined as “particulate matter with an aerodynamic diameter less than or 6 BAHR V. U.S. EPA equal to a nominal ten micrometers.” Id. § 7602(t). According to the EPA, “PM-10 causes adverse health effects by penetrating deep into the lungs, aggravating the cardiopulmonary system.” Approval and Promulgation of Implementation Plans—Maricopa County PM-10 Nonattainment Area, 79 Fed. Reg. 7118, 7118 (Feb. 6, 2014). The EPA established a NAAQS for PM-10 of 150 micrograms per cubic meter, averaged over a 24-hour period. 40 C.F.R. § 50.6(a). This standard, which is sometimes referred to as the “24-hour PM-10 standard,” is “attained when the expected number of days per calendar year with a 24-hour average concentration above 150 ìg/m³ . . . is equal to or less than one.” Id. The CAA provides that “[e]ach State shall have the primary responsibility for assuring air quality” within the state “by submitting an implementation plan” explaining how the state will meet and maintain the NAAQS and other standards. 42 U.S.C. § 7407(a). An area within a state that does not meet a NAAQS is designated as a “nonattainment” area, id. § 7407(d). Each state’s implementation plan (called a State Implementation Plan or SIP) must provide for the “implementation, maintenance, and enforcement” of the NAAQS. Id. § 7410(a)(1). The CAA requires each SIP for a nonattainment area to contain specified information, including a requirement for reasonable further progress, id. § 7502(c)(2), an emissions inventory, id. § 7502(c)(3), and a list of “contingency measures” to “be undertaken if the area fails to make reasonable further progress, or to attain the BAHR V. U.S. EPA 7 national primary ambient air quality standard by the attainment date applicable under this part,” id. § 7502(c)(9).1 The CAA sets out a series of deadlines for states to meet the NAAQS for PM-10, with increasingly stringent requirements if a state misses a deadline. Id. §§ 7513–7513b. The sequence is as follows: A nonattainment area is initially designated as a “moderate” area. Id. § 7513(a). A SIP for a “moderate” PM10 nonattainment area must explain how that area will meet the PM-10 NAAQS by the “attainment date,” which for nonattainment areas designated by Congress was no later than December 31, 1994. Id. § 7513(c)(1). The SIP must “assure that reasonably available control measures for the control of PM-10” are implemented. Id. § 7513a(a). If a moderate nonattainment area fails to meet the PM-10 NAAQS by the attainment date, the EPA must reclassify it as a “Serious PM-10 nonattainment area.” Id. § 7513(b). After redesignation, the state must submit a SIP that demonstrates how the area will meet the PM-10 NAAQS within 10 years of 1 42 U.S.C. § 7502(c)(9) states: (9) Contingency measures Such plan shall provide for the implementation of specific measures to be undertaken if the area fails to make reasonable further progress, or to attain the national primary ambient air quality standard by the attainment date applicable under this part. Such measures shall be included in the plan revision as contingency measures to take effect in any such case without further action by the State or the Administrator. 8 BAHR V. U.S. EPA the original nonattainment designation, or, for areas originally designated as nonattainment by Congress, no later than December 31, 2001. Id. § 7513(c)(2). A SIP for a serious nonattainment area must also “assure that the best available control measures [BACM] for the control of PM-10 shall be implemented.” Id. § 7513a(b).2 If a state fails to meet the deadline for bringing a Serious Area into compliance, the EPA may grant the state a single five-year extension of the deadline to meet the NAAQS for PM-10, but only if the state submits a SIP that “includes the most stringent measures [MSM] that are included in the implementation plan of any State or are achieved in practice in any State, and can feasibly be implemented in the area.” Id. § 7513(e).3 2 42 U.S.C. § 7513a(b) provides: (b) Serious Areas (1) Plan provisions. In addition to the provisions submitted to meet the requirements of paragraph [1] (a)(1) (relating to Moderate Areas), each State in which all or part of a Serious Area is located shall submit an implementation plan for such area that includes each of the following: . . . (B) Provisions to assure that the best available control measures for the control of PM-10 shall be implemented no later than 4 years after the date the area is classified (or reclassified) as a Serious Area. 3 42 U.S.C. § 7513(e) provides in pertinent part: (e) Extension of attainment date for Serious Areas Upon application by any State, the Administrator may BAHR V. U.S. EPA 9 If a Serious Area fails to achieve compliance by the attainment date after receiving the one-time five-year extension under § 7513(e), the CAA requires the state to “submit within 12 months after the applicable attainment date, plan revisions which provide for attainment of the PM10 air quality standard.” Id. § 7513a(d).4 The SIP revisions extend the attainment date for a Serious Area beyond the date specified under subsection (c) of this section, if attainment by the date established under subsection (c) of this section would be impracticable, the State has complied with all requirements and commitments pertaining to that area in the implementation plan, and the State demonstrates to the satisfaction of the Administrator that the plan for that area includes the most stringent measures that are included in the implementation plan of any State or are achieved in practice in any State, and can feasibly be implemented in the area. . . . The Administrator may not approve an extension until the State submits an attainment demonstration for the area. The Administrator may grant at most one such extension for an area, of no more than 5 years. 4 42 U.S.C. § 7513a(d) states: (d) Failure to attain In the case of a Serious PM-10 nonattainment area in which the PM-10 standard is not attained by the applicable attainment date, the State in which such area is located shall, after notice and opportunity for public comment, submit within 12 months after the applicable attainment date, plan revisions which provide for attainment of the PM-10 air quality standard and, from the date of such submission until attainment, for an annual reduction in PM-10 or PM-10 precursor emissions within the area of not less than 5 percent of 10 BAHR V. U.S. EPA must provide for an annual five percent reduction in PM-10 within the Serious Area from the date the SIP revision was submitted to the EPA until the state attains the NAAQS in that area. Id. States are required to conduct ambient air quality monitoring to determine whether a geographical region or area in the state is meeting the NAAQS for PM-10. Id. § 7410(a)(2)(B)(i). State air quality monitoring systems must use the criteria and methodology established by the EPA. Id. § 7619(a). Congress recognized that air quality monitoring data could be affected by exceptional events that could not reasonably be controlled by the states, and directed the EPA to promulgate regulations “governing the review and handling of air quality monitoring data influenced by exceptional events.” Id. § 7619(b)(2)(A). The statute defines an “exceptional event” as an event that “(i) affects air quality; (ii) is not reasonably controllable or preventable; (iii) is an event caused by human activity that is unlikely to recur at a particular location or a natural event; and (iv) is determined by the Administrator through the process established in the regulations promulgated under paragraph (2) to be an exceptional event.” Id. § 7619(b)(1)(A). Pursuant to this direction, the EPA promulgated the “Exceptional Events Rule,” 40 C.F.R. § 50.14. The rule repeats the statute’s definition of “exceptional event,” id. the amount of such emissions as reported in the most recent inventory prepared for such area. BAHR V. U.S. EPA 11 § 50.1(j),5 and allows a state to “request EPA to exclude data showing exceedances or violations of the national ambient air quality standard that are directly due to an exceptional event,” id. § 50.14(a)(1). In order to obtain EPA approval to exclude exceptional event data, a state must provide evidence that “[t]he event satisfies the criteria set forth in 40 C.F.R 50.1(j)” and meets other criteria. Id. § 50.14(c)(3)(iv). If a state makes the required showing, the “EPA shall exclude [the exceptional event] data from use in determinations of exceedances and NAAQS violations.” Id. § 50.14(b)(1). The EPA has recognized that PM-10 levels can be affected by natural events such as dust storms, see id. § 50.1(j)–(k), and has therefore developed guidance for applying the Exceptional Events Rule to high wind events. See Treatment of Data Influenced by Exceptional Events, 72 Fed. Reg. 13560 (Mar. 22, 2007) (Treatment of Data Guidance). The Treatment of Data Guidance states that increased particulate matter concentrations “raised by unusually high winds will be treated as due to uncontrollable natural events where (1) the dust originated from nonanthropogenic sources, or (2) the dust originated from 5 40 C.F.R. § 50.1(j) provides: (j) Exceptional event means an event that affects air quality, is not reasonably controllable or preventable, is an event caused by human activity that is unlikely to recur at a particular location or a natural event, and is determined by the Administrator in accordance with 40 CFR 50.14 to be an exceptional event. It does not include stagnation of air masses or meteorological inversions, a meteorological event involving high temperatures or lack of precipitation, or air pollution relating to source noncompliance. 12 BAHR V. U.S. EPA anthropogenic sources within the State, that are determined to have been reasonably well-controlled at the time that the event occurred, or from anthropogenic sources outside the State.” Id. at 13576 (emphasis added). In May 2013, the EPA published additional “guidance and interpretation” explaining how the Exceptional Events Rule and the Treatment of Data Guidance applies to high wind events. See EPA, Interim Guidance on the Preparation of Demonstrations in Support of Requests to Exclude Ambient Air Quality Data Affected by High Winds Under the Exceptional Events Rule (May 2013) (Interim Guidance).6 The Interim Guidance addresses when an anthropogenic source within the State is “reasonably well-controlled at the time that the event occurred,” 72 Fed. Reg. at 13576, as required by the Treatment of Data Guidance. II We now provide the background of this case. Congress designated Maricopa County, Arizona, as a “moderate” PM10 nonattainment area in 1990. 42 U.S.C. § 7407(d)(4)(B); PM10 Group I and Group II Areas, 52 Fed. Reg. 29383, 29384 (Aug. 7, 1987). The designated nonattainment area, termed the “Maricopa County PM-10 Nonattainment Area” (Maricopa Area), covers the eastern portion of Maricopa County, including the cities of Phoenix, Mesa, Scottsdale, 6 The Interim Guidance is available at: https://www.epa.gov/ sites/production/files/2015-09/documents/exceptevents_highwinds_ guide_130510.pdf; see also Draft Guidance To Implement Requirements for the Treatment of Air Quality Monitoring Data Influenced by Exceptional Events, 77 Fed. Reg. 39959, 39960 (July 6, 2012) (announcing the availability of a draft version of the Interim Guidance on the EPA’s website). BAHR V. U.S. EPA 13 Tempe, Chandler, and Glendale. 79 Fed. Reg. at 7118. It also covers unincorporated parts of Maricopa County and portions of Pinal County. Id. Because the Maricopa Area was designated as a nonattainment area by Congress, Arizona’s first deadline for meeting the NAAQS for PM-10 was December 31, 1994. 42 U.S.C. § 7513(c). The Maricopa Area was not in attainment by 1994, so the EPA reclassified the Maricopa Area as a Serious PM-10 nonattainment area. 42 U.S.C. § 7513(b)(2); Clean Air Act Reclassification; ArizonaPhoenix Nonattainment Area; PM-10, 61 Fed. Reg. 21372, 21373 (May 10, 1996). This required Arizona to submit a SIP that would demonstrate how the Maricopa Area would meet the NAAQS for PM-10 by December 31, 2001, see 42 U.S.C. § 7513(c)(2), and to explain how it would implement the best available control measures for PM-10, id. § 7513a(b). The Maricopa Area did not meet the NAAQS for PM-10 by the end of 2001. Rather, in 2000 Arizona preemptively applied for a five-year extension (until December 2006) under § 7513(e) and at the same time submitted a SIP for the Maricopa Area (the 2000 SIP). The 2000 SIP proposed to implement over 70 best available control measures for major dust sources, Approval and Promulgation of Implementation Plans; Arizona—Maricopa County PM-10 Nonattainment Area; Serious Area Plan for Attainment of the Annual PM-10 Standard, 65 Fed. Reg. 19964, 19972–83 (Apr. 13, 2000), and stated that its measures were “the most stringent measures that are included in the implementation plan of any State, or are achieved in practice in any State,” id. at 19984. In 2002, the EPA issued a final rule approving the 2000 SIP and granting the requested extension. Approval and Promulgation 14 BAHR V. U.S. EPA of Implementation Plans; Arizona–Maricopa County PM-10 Nonattainment Area; Serious Area Plan for Attainment of the PM-10 Standards, 67 Fed. Reg. 48718, 48718–19 (July 25, 2002) (the 2002 Final Rule). The 2002 Final Rule stated that the control measures in the 2000 SIP met the BACM and MSM standards. Id.7 By December 2006, the Maricopa Area had still failed to meet the NAAQS for PM-10. See Findings of Failure To Attain; State of Arizona, Phoenix Nonattainment Area; State of California, Owens Valley Nonattainment Area; Particulate Matter of 10 Microns or Less, 72 Fed. Reg. 31183, 31184–85 (June 6, 2007). At that point, the CAA gave Arizona 12 months to submit revisions to the SIP that would achieve attainment of the NAAQS for PM-10, provide for an annual five percent reduction in PM-10 in the Maricopa Area, 42 U.S.C. § 7513a(d), and contain appropriate contingency measures, id. § 7502(c)(9). Arizona submitted revisions to the Maricopa Area SIP in December 2007 (the 2007 SIP). In addition to proposing 53 control measures, the 2007 SIP proposed revising a previously approved agricultural control measure, namely an agricultural general permit specifying best management practices for reducing PM-10 from agricultural activities. See Approval and Promulgation of Implementation Plans— Maricopa County (Phoenix) PM-10 Nonattainment Area; 7 We considered a challenge to EPA’s 2002 Final Rule approving the 2000 SIP and held that the EPA’s approval of Arizona’s rejection of a measure requiring the use of a reformulated diesel fuel as BACM was arbitrary and capricious. Vigil v. Leavitt, 381 F.3d 826, 841–46 (9th Cir. 2004). We upheld the rest of the EPA’s approval, including its approval of Arizona’s general permit rule for agricultural emissions of PM-10 as BACM. Id. at 836–38, 847. BAHR V. U.S. EPA 15 Serious Area Plan for Attainment of the 24-Hour PM-10 Standard; Clean Air Act Section 189(d), 75 Fed. Reg. 54806, 54810, 54812–13 (Sept. 9, 2010) (the 2010 Proposed Rule). The 2010 Proposed Rule stated that the EPA would disapprove this revision to the agricultural general permit rule on the ground that other states and local agencies had “acquired additional expertise about how to control emissions from these sources,” and as a result the EPA no longer believed that the requirements in the agricultural general permit rule in the 2000 SIP were best available control measures. Id. To avoid a partial disapproval, Arizona withdrew the plan in 2011. 79 Fed. Reg. at 7119. As a result of this withdrawal of the 2007 SIP, the EPA found that Arizona had failed to make a required SIP submittal. Finding of Failure To Submit State Implementation Plan Revisions for Particulate Matter, PM-10, Maricopa County (Phoenix) PM10 Nonattainment Area, AZ, 76 Fed. Reg. 8300, 8300–01 (Feb. 14, 2011). This finding required Arizona to submit another SIP by 2013. Id. at 8301; 42 U.S.C. § 7509(a). Arizona submitted a new SIP revision on May 25, 2012. See 79 Fed. Reg. at 7119. Because this SIP was prepared pursuant to § 7513a(d), which requires a state to achieve a five percent annual reduction in PM-10, we adopt the EPA’s term and refer to it as the “Five Percent Plan.” The Five Percent Plan proposed to achieve the five percent annual reduction required by § 7513a(d) by implementing many of the 53 control measures previously proposed in the 2007 SIP, as well as adopting a new emissions control measure, the “Dust Action General Permit.” Unlike the 2007 SIP, the Five Percent Plan did not propose any changes to the agricultural general permit. 16 BAHR V. U.S. EPA As required by § 7502(c)(9), the Five Percent Plan proposed a number of contingency measures. 79 Fed. Reg. at 7124. Four of the five measures were permanent changes to infrastructure that had been completed in the years 2008 through 2011, namely, paving existing dirt roads and alleys, paving and stabilizing unpaved shoulders, repaving or overlaying paved roads with rubberized asphalt, and lowering speed limits on dirt roads and alleys. Id. The fifth contingency measure required the purchase of PM-10 certified sweepers (which had already been accomplished by the end of 2009), and ongoing sweeping of ramps, freeways, and frontage roads. Id. Arizona also acknowledged that there had been a number of exceedances of the 24-hour PM-10 standard during 2011 and 2012, but claimed they should be deemed “exceptional events” and excluded from a determination of whether the Maricopa Area met the NAAQS for PM-10. In support of this claim, Arizona submitted documentation to the EPA to demonstrate that 137 exceedances of the NAAQS for PM-10 on 27 days during the period from 2011 to 2012 were the result of “exceptional events,” namely high wind dust events. 79 Fed. Reg. at 7122.8 The EPA wrote separate reports on each of Arizona’s submissions. The reports analyzed the data and concluded that the flagged exceedances met “the definition of an exceptional event: the exceedances affected air quality, were not reasonably controllable or preventable, and meet the definition of a natural event.” In February 2014, the EPA published its proposed decision regarding Arizona’s Five Percent Plan. 79 Fed. Reg. 8 Though the EPA’s proposed rulemaking mentioned 133 exceedances, the parties agree that Arizona requested approval of 137. BAHR V. U.S. EPA 17 at 7118. The EPA noted that the monitoring data for the Maricopa Area showed 133 exceedances of the 24-hour PM10 NAAQS during the 2011–2012 time period, but stated that 131 of those exceedances “were caused by high wind exceptional events,” and “should not be used for regulatory purposes,” including for evaluation of the Five Percent Plan. Id. at 7122. Excluding these exceedances, the EPA proposed to determine that the Maricopa Area had attained the NAAQS for 24-hour PM-10 by December 31, 2012. Id. at 7125. The EPA also proposed to approve Arizona’s contingency measures. Id. at 7124. In doing so, the EPA explained that it had previously “interpreted [42 U.S.C. § 7502(c)(9)] to allow states to implement contingency measures before they are triggered by a failure of . . . attainment as long as those measures are intended to achieve emission reductions over and beyond those relied on in the attainment and [reasonable further progress] demonstrations.” 79 Fed. Reg. at 7124. The EPA issued a final rule on June 10, 2014, approving the Five Percent Plan “as meeting all relevant statutory and regulatory requirements.” Approval and Promulgation of Implementation Plans—Maricopa County PM-10 Nonattainment Area; Five Percent Plan for Attainment of the 24-Hour PM-10 Standard, 79 Fed. Reg. 33107, 33107 (June 10, 2014) (2014 Final Rule). The 2014 Final Rule stated that the EPA was excluding 135 exceedances9 from the monitoring data as exceptional events. Id. at 33111. For each of the events that EPA determined was exceptional, “EPA found that the event was not reasonably controllable or preventable,” and that “reasonable controls” were in place for 9 The EPA’s final rule clarified that the number of approved exceedances was 135, which occurred on 25 days over the period of 2010–12. 79 Fed. Reg. at 33110–11. 18 BAHR V. U.S. EPA anthropogenic sources of dust. Id. The EPA reached its conclusion that “reasonable controls” were in place by relying on its 2002 Final Rule approving Arizona’s 2000 SIP as including best available control measures and most stringent control measures. Id. at 33112. The EPA also determined that Arizona sufficiently demonstrated that dust sources outside the Maricopa Area were reasonably controlled. Id. at 33113. The 2014 Final Rule included the EPA’s response to comments made by petitioners Sandra Bahr and David Matusow, two residents of Phoenix. First, the EPA addressed petitioners’ argument that the EPA should have required Arizona to update its control measures to ensure that they were BACM and MSM, rather than letting Arizona rely on the EPA’s approval of the 2000 SIP. 79 Fed. Reg. at 33108–10. The EPA rejected this claim, explaining that “the requirement for BACM is triggered by a specific event: The reclassification of a moderate PM-10 nonattainment area to serious.” Id. at 33108. Likewise, the EPA explained that § 7513(e) provides “that the requirement for MSM is triggered by a particular event: EPA’s granting of a state’s request for an extension of the attainment deadline for a serious nonattainment area.” Id. The EPA then stated that § 7513(d) (the section requiring the submission of a five percent plan) “does not contain a specific requirement that the state update the previously approved requirements for BACM and MSM as a consequence of failing to reach attainment by the applicable deadline for serious PM-10 nonattainment areas.” Id. at 33109. Because there was no statutory trigger requiring Arizona to update its demonstration that its control measures were best available and most stringent, the EPA explained, the Five Percent Plan could rely on the EPA’s approval of BACM and MSM in the 2000 SIP. Id. The EPA BAHR V. U.S. EPA 19 also disagreed with petitioner’s comment that the Five Percent Plan was inadequate because the agricultural control measures were not BACM, as indicated by the EPA’s 2010 Proposed Rule disapproving of Arizona’s proposed revision to its agricultural general permit. In response to this comment, the EPA stated that the Five Percent Plan satisfied the requirements in § 7513a(d) “without relying on additional emissions reductions from agricultural sources.” Id. at 33109. Second, the EPA addressed petitioners’ argument that its determination that the 135 exceedances constituted exceptional events was contrary to the Interim Guidance. Petitioners interpreted the Interim Guidance as preventing the EPA from concurring that best available control measures were in place unless the EPA had determined control measures for windblown dust to be BACM within the past three years. 79 Fed. Reg. at 33111–12. The EPA disagreed with this interpretation and concluded that Arizona’s controls were reasonable. Id. at 33112. The EPA also rejected petitioners’ argument that Arizona had failed to provide an adequate description of upwind sources and control measures as required by the Interim Guidance. Id. at 33113–14. Finally, the EPA dismissed petitioners’ comment that the EPA erred in accepting Arizona’s “contingency measures” in its Five Percent Plan because the measures had already been implemented. Id. at 33114–15. The petitioners argued that because § 7502(c)(9) requires contingency measures that “are automatically and immediately implemented if a milestone for reasonable further progress or attainment is not met,” the previously implemented measures in the Five Percent Plan could not qualify. Id. at 33114. In rejecting this comment, the EPA stated it interpreted § 7502(c)(9) as requiring only 20 BAHR V. U.S. EPA that “[c]ontingency measures must provide for additional emission reductions” that were not otherwise included “in the attainment demonstration,” and that “[n]othing in the statute precludes a state from implementing such measures before they are triggered,” relying on the Fifth Circuit decision in Louisiana Environmental Action Network v. EPA, 382 F.3d 575 (5th Cir. 2004). 79 Fed. Reg. at 33114. The plaintiffs filed a petition for review of the 2014 Final Rule on July 29, 2014. III Under 42 U.S.C. § 7607(b)(1), we have jurisdiction over “[a] petition for review of the Administrator’s action in approving or promulgating any implementation plan.” In reviewing a challenge to the EPA’s approval of a SIP under § 7607(b)(1), we apply “the general standard of review for agency actions set forth in the Administrative Procedure Act (APA).” Latino Issues Forum v. EPA, 558 F.3d 936, 941 (9th Cir. 2009); see also Vigil v. Leavitt, 381 F.3d 826, 833 (9th Cir. 2004). Under the APA, we must uphold an agency action unless it is “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. § 706(2)(A). This standard is “highly deferential, presuming the agency action to be valid and affirming the agency action if a reasonable basis exists for its decision.” Ranchers Cattlemen Action Legal Fund United Stockgrowers of Am. v. U.S. Dep’t of Agric., 499 F.3d 1108, 1115 (9th Cir. 2007). Generally, “[a]n agency decision will be upheld as long as there is a rational connection between the facts found and the conclusions made.” Barnes v. U.S. Dep’t of Transp., 655 F.3d 1124, 1132 (9th Cir. 2011). We will deem an BAHR V. U.S. EPA 21 agency action to be arbitrary and capricious only “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.” Motor Vehicle Mfrs. Ass’n of the U.S. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 (1983). Where the question presented for review is a factual dispute which implicates “a high level of technical expertise” we defer to “the informed discretion of the responsible federal agencies.” Kleppe v. Sierra Club, 427 U.S. 390, 412 (1976). “Even when an agency explains its decision with ‘less than ideal clarity,’ a reviewing court will not upset the decision on that account ‘if the agency's path may reasonably be discerned.’” Alaska Dep’t of Envtl. Conservation v. EPA, 540 U.S. 461, 497 (2004) (quoting Bowman Transp., Inc. v. Ark.-Best Freight Sys., Inc., 419 U.S. 281, 286 (1974)). Where the petitioner challenges the agency’s action as inconsistent with the agency’s own policies, we examine whether the agency has actually departed from its policy and, if so, whether the agency has offered a reasoned explanation for such departure. See Encino Motorcars, LLC v. Navarro, 136 S. Ct. 2117, 2125–26 (2016). Generally, “[a]gencies are free to change their existing policies as long as they provide a reasoned explanation for the change.” Id. at 2125. In contrast, where the agency is not offering a policy explanation but is instead interpreting a binding regulation, the agency’s interpretation is “controlling” unless “plainly erroneous or inconsistent with the regulation.” Auer v. Robbins, 519 U.S. 452, 461 (1997). 22 BAHR V. U.S. EPA Where the agency’s action is an interpretation of a statute that the agency administers, “we follow the two-step approach set out in Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984).” Latino Issues Forum, 558 F.3d at 941. First, “if Congress has ‘directly spoken to the precise question at issue,’ then the matter is capable of but one interpretation by which the court and the agency must abide.” Vigil, 381 F.3d at 834 (quoting Chevron, 467 U.S. at 842). At the second step of Chevron, if we determine that Congress was silent on the issue, or the statute is subject to multiple interpretations, we must determine the degree of deference to give the agency’s interpretation of a statute. United States v. Mead Corp., 533 U.S. 218, 227–29 (2001). “Not all agency statutory interpretations are entitled to Chevron deference.” Sierra Club v. EPA, 671 F.3d 955, 962 (9th Cir. 2012). “Rather, Chevron deference is appropriate where ‘the agency can demonstrate that it has the general power to make rules carrying the force of law and that the challenged action was taken in the exercise of that authority.’” Id. (quoting Wilderness Soc’y v. U.S. Fish & Wildlife Serv., 353 F.3d 1051, 1067 (9th Cir. 2003) (en banc)). We generally deem Congress to have delegated such authority when it authorizes the agency to engage in noticeand-comment rulemaking. Mead, 533 U.S. at 229–30. In such circumstances, a court should accept the agency’s interpretation “if Congress has not previously spoken to the point at issue and the agency’s interpretation is reasonable.” Id. at 229. Though we have previously applied Chevron deference to the EPA’s interpretation of the CAA issued in connection with a SIP approval, see Association of Irritated Residents v. BAHR V. U.S. EPA 23 EPA, 686 F.3d 668, 679–81 (9th Cir. 2011), we have never expressly held that such deference is appropriate, but rather have held the question open, see Vigil, 381 F.3d at 835 (declining to address “whether the EPA’s interpretation of the Act in the course of approving Arizona’s SIP is entitled to Chevron deference”). But because “[a] very good indicator of delegation meriting Chevron treatment [is] express congressional authorizations to engage in the rulemaking or adjudication process that produces the regulations or rulings for which deference is claimed,” Mead Corp., 533 U.S. at 229, and because the EPA engages in such a rulemaking process in approving a SIP, we agree with the Fifth Circuit that “EPA’s final rules approving the [Arizona] SIP, to the extent they involve the reasonable resolution of ambiguities in the CAA, will be afforded Chevron deference,” BCCA Appeal Grp. v. EPA, 355 F.3d 817, 825 (5th Cir. 2003). IV On appeal, petitioners argue that the EPA acted contrary to law by failing to require that Arizona include an updated analysis of best available control measures and most stringent measures in the Five Percent Plan. They also argue that the EPA abused its discretion by excluding 135 exceedances from the monitoring data as “exceptional events.” Finally, they argue that the EPA violated the CAA by allowing Arizona to satisfy the “contingency measures” requirement with previously implemented control measures. We consider each argument in turn. A We first consider petitioners’ argument that the EPA’s approval of the Five Percent Plan constituted an abuse of 24 BAHR V. U.S. EPA discretion because the EPA did not require Arizona to demonstrate that the plan included best available control measures (BACM) or most stringent measures (MSM) for the control of PM-10. According to petitioners, the Five Percent Plan must demonstrate that Arizona’s control measures meet the BACM and MSM standards because Arizona has a continuing obligation to comply with § 7513a(b)(1)(B) and § 7513(e), which set the control requirements for serious nonattainment areas. Petitioners argue that had a BACM demonstration been required, the Five Percent Plan would have failed because the agricultural control measures in Arizona’s 2000 SIP are no longer BACM, as stated in EPA’s 2010 Proposed Rule. Further, petitioners argue, the EPA’s statement in the 2010 Proposed Rule regarding Arizona’s failure to meet the BACM standard is evidence that the EPA generally requires an updated BACM demonstration in each SIP, meaning that the EPA acted inconsistently by not requiring such a demonstration in the Five Percent Plan. We disagree with these arguments. The EPA’s decision not to require an updated demonstration of BACM and MSM in the Five Percent Plan was not an abuse of discretion because it was not contrary to any language in the CAA. Section 7513a(d), which governs five percent plans such as the one before us here, provides only that if a state fails to achieve attainment after receiving the five-year extension, it must submit plan revisions providing for “attainment of the PM-10 air quality standard,” and an annual five percent reduction in PM-10 within the Serious Area. It does not mention BACM or MSM. The CAA sections that do require BACM and MSM demonstrations do not expressly apply to a five percent plan. Section § 7513a(b) provides that when the EPA reclassifies a Moderate Area as a Serious Area, the state must then submit a SIP that “assure[s] that the best BAHR V. U.S. EPA 25 available control measures for the control of PM-10 shall be implemented.” It does not require the state to update that assurance when submitting a five percent plan, nor does it require the EPA to review its previous BACM determination. Similarly, § 7513(e) provides that when the EPA grants a state a five-year extension of the deadline to meet the NAAQS for PM-10, the state must submit a SIP that “includes the most stringent measures” that are included in any SIP or achieved in any state. Again, this language does not require the state to include an updated demonstration of MSM when submitting a five percent plan. We also disagree with petitioners’ argument that the EPA acted inconsistently in failing to require an updated demonstration of BACM and MSM in the Five Percent Plan. Even though the EPA had previously reviewed and proposed to disapprove the agricultural control measure in Arizona’s 2007 SIP as not meeting the BACM standard, the EPA provided a reasonable interpretation of its approach in the 2014 Final Rule. According to the EPA, the CAA lays out a series of escalating control measures that are triggered by a finding of noncompliance with a series of statutory requirements. The EPA will assess compliance with the control measures that were triggered at each step, but will not reassess compliance with those measures at subsequent steps unless a state proposes changes to control measures that were previously approved. 79 Fed. Reg. at 33108–09. This approach is consistent with the CAA, which does not require the EPA to reassess a state’s controls in each SIP submission. Here, the EPA’s 2010 Proposed Rule reassessed Arizona’s agricultural control measures under the BACM standard because Arizona proposed to revise agricultural control measures that were previously approved as BACM in the 2002 Final Rule. Arizona did not propose any revisions to its 26 BAHR V. U.S. EPA agricultural controls in the Five Percent Plan, however, so EPA did not have any occasion to reevaluate those measures. Given the EPA’s reasonable explanation for its approach, and the lack of any contrary statutory command in the CAA, we conclude that the EPA did not abuse its discretion or act contrary to law by declining to require an updated demonstration of BACM or MSM in the Five Percent Plan. Petitioners also argue that the EPA acted in an arbitrary and capricious manner in reviewing the Five Percent Plan because it evaluated Arizona’s compliance with CAA requirements regarding emission inventories, reasonable further progress, and contingency measures, but ignored BACM and MSM. We also reject this argument. The EPA reasonably explained that those particular CAA requirements are procedural or otherwise applicable to all SIP submissions, see 42 U.S.C. §§ 7410(a), 7502(c), 7506(c), 7513a(c)(1), and the EPA reviews such measures whenever it reviews a proposed SIP. By contrast, the BACM and MSM requirements are not applicable to all SIP submissions, and so the EPA reviews them only when the state is required to demonstrate compliance with these requirements. B We next turn to the petitioners’ argument that the EPA acted contrary to law by excluding 135 exceedances in the Maricopa Area from Arizona’s air quality monitoring data. Petitioners raise several arguments as to why the EPA erred in concluding that the dust sources causing the 135 exceedances were from anthropogenic sources that were “reasonably well-controlled,” and therefore were excludable as exceptional events. See 40 C.F.R. § 50.14(b)(1). The BAHR V. U.S. EPA 27 petitioners begin by pointing to the statement in the Interim Guidance that “[g]enerally, the EPA will consider windblown dust BACM to constitute reasonable controls if these measures have been reviewed and approved in the context of a SIP revision for the emission source area within the past three years.” Interim Guidance at 15. The EPA’s decision was inconsistent with this guidance, petitioners argue, for two reasons. First, Arizona’s dust control measures for the Maricopa Area had not been approved since 2002, well over three years before the 2014 Final Rule. Second, agricultural emissions are among the sources of windblown dust in the Maricopa Area, and the EPA’s 2010 Proposed Rule had proposed to disapprove of Arizona’s agricultural control measures because they were not BACM. Moreover, petitioners argue, the EPA failed to offer a reasonable explanation for its departure from its guidance. We disagree; the EPA’s 2014 Final Rule did not conflict with the Interim Guidance. First, nothing in the Interim Guidance indicates that EPA must find that control measures for windblown dust have been reviewed and approved as BACM within the past three years in order for the dust to be deemed reasonably well-controlled. Rather, the Interim Guidance gives the EPA flexibility to consider a wide range of issues, and emphasizes that a prior BACM determination “may be a reference point, but not the sole means, by which the EPA assess the reasonableness of controls.” Interim Guidance at 15. The EPA’s interpretation is therefore consistent with the Interim Guidance and is a reasonable interpretation of the Exceptional Events Rule, to which we owe deference. Auer, 519 U.S. at 461. Second, the EPA provided a reasonable explanation as to why the Maricopa Area had reasonable controls for 28 BAHR V. U.S. EPA windblown dust even though the 2010 Proposed Rule had proposed to disapprove of Arizona’s agricultural control measures. The EPA’s 2002 Final Rule determined that the 2000 SIP contained the best available control measures for the highest emitters of PM-10 (including unpaved roads and alleys, construction, paved road dust, and non-agricultural windblown dust). 79 Fed. Reg. at 33111–13. The 2014 Final Rule then explained that it was still appropriate to rely on that determination because neither the highest emitters of PM-10 nor the techniques for controlling fugitive dust had changed significantly since 2002. Id. at 33112. Moreover, Arizona made its dust control rules even more stringent in the years following its 2000 SIP, further bolstering the conclusion that the controls remained reasonable. Id. Although the EPA had proposed to disapprove of the agricultural controls as not meeting the BACM standard in 2010, data showed that agricultural sources were only a minimal contributor to the Maricopa Area’s overall level of PM-10 emissions. Id. Accordingly, the EPA’s judgment was that the control measures in the 2000 SIP ensured that anthropogenic windblown dust was reasonably controlled for purposes of the Exceptional Events Rule. “[W]e generally must be at [our] most deferential when reviewing scientific judgments and technical analyses within the agency’s expertise.” Lands Council v. McNair, 629 F.3d 1070, 1074 (9th Cir. 2010) (second alteration in original) (internal quotation marks omitted). Here, the EPA considered the relevant factors and articulated a rational connection between the facts found and the choice made. As a general rule, a determination that particular control measures are reasonable relies on technical considerations that are BAHR V. U.S. EPA 29 “properly left to the informed discretion of” the EPA. Kleppe, 427 U.S. at 412. We defer to its conclusion here.10 C Petitioners also claim that the EPA’s approval of Arizona’s 135 exceedances as exceptional events violated the Interim Guidance because the EPA failed to adequately address the controls in upwind areas outside the Maricopa Area. To qualify as an exceptional event under the Exceptional Events Rule, an exceedance must be “caused by human activity that is unlikely to recur at a particular location or a natural event.” 40 C.F.R. §§ 50.1(j), 50.14. The Treatment of Data Guidance states that high wind events will be considered “natural” when “(1) the dust originated from nonanthropogenic sources, or (2) the dust originated from anthropogenic sources within the State, that are determined to have been reasonably well-controlled at the time that the event occurred, or from anthropogenic sources outside the State.” 72 Fed. Reg. at 13576. To assist the EPA in determining compliance with this requirement, the Interim Guidance requires a state to provide “a brief description” of “all contributing emission sources in upwind areas and provide evidence that those sources were reasonably controlled, whether anthropogenic or natural.” Interim Guidance at 42. Petitioners argue that Arizona’s submissions were inadequate because Arizona did not identify all contributing emission sources outside the Maricopa Area, failed to 10 Because we conclude that the EPA did not depart from its Interim Guidance, we do not reach the petitioners’ argument that the EPA failed to provide a reasoned explanation for departing from its guidance. 30 BAHR V. U.S. EPA distinguish between natural and anthropogenic sources, and failed to submit evidence that Pinal County had reasonable controls in place. We again disagree. Under the Treatment of Data Guidance, a high wind event may meet the criteria of the Exceptional Events Rule when the dust originated from nonanthropogenic sources, or from anthropogenic sources that are reasonably well-controlled at the time that the event occurred. 72 Fed. Reg. at 13576–77. Arizona explained that the high wind events causing the 135 exceedances stemmed from monsoonal dust storms. Because “outflow from thunderstorms can carry dust over vast distances encompassing many source areas,” Arizona could not clearly distinguish between nonanthropogenic and anthropogenic sources of dust. Nevertheless, Arizona adequately provided a “brief description” of contributing dust sources outside the Maricopa Area and demonstrated that reasonable controls were in place for any anthropogenic sources of dust. For instance, in describing exceedances that occurred during the August 11, 2012, event, Arizona provided a “conceptual model” identifying a nonanthropogenic source, the “undeveloped lands south of Maricopa County,” as being the primary contributing source areas in Pima and Pinal Counties. Arizona also provided evidence that any anthropogenic sources of dust in those areas were reasonably controlled, pointing to two Pinal County rules applicable to fugitive dust and construction sites. Arizona compiled similar submissions for each of its exceptional event submittals. Arizona’s submissions provided enough detail for the EPA to reasonably conclude that the dust originated either from “nonanthropogenic sources” or “anthropogenic sources” that were “reasonably well-controlled.” Interim Guidance at 42. Accordingly, the EPA’s conclusion that Arizona’s description BAHR V. U.S. EPA 31 of the upwind sources was adequate was not an abuse of discretion. Petitioners also argue that the EPA ignored the Interim Guidance and thus abused its discretion in concluding that the anthropogenic dust sources in the areas of Pinal County outside of the Maricopa Area were reasonably controlled. The Interim Guidance states that “[f]or the anthropogenic sources to be considered to be reasonably controlled, the EPA anticipates that it is reasonable for an air agency to have the controls required for an area’s attainment status.” Interim Guidance at 15. “[T]he EPA does not expect areas classified as attainment, unclassifiable, or maintenance for a NAAQS to have the same level of controls as areas that are nonattainment for the same NAAQS.” Id. In other words, an area that is in attainment should have the control measures appropriate for an attainment area, while an area that has been designated a serious nonattainment area should have the control measures appropriate for that level of classification. Id. Where “an area has been recently designated to nonattainment but has not yet been required to implement controls, the EPA will expect the level of controls that is appropriate for the planning stage.” Id. Under the Interim Guidance, the EPA did not abuse its discretion in concluding that the anthropogenic dust sources in the areas of Pinal County outside of the Maricopa Area were reasonably well-controlled. The EPA provided a reasoned explanation as to why it deemed the Pinal County area to have the controls required for that area’s attainment status. The 135 exceedances approved by the EPA all occurred between 2010 and 2012. 79 Fed. Reg. at 33111. From January 2011 to June 2012, Pinal County (excluding the portion within the Maricopa Area) was an attainment area, 32 BAHR V. U.S. EPA and had adopted two rules addressing fugitive dust emissions that were appropriate for this status. Although Pinal County was redesignated a nonattainment area in July 2012, it was not required to submit a SIP until 18 months after redesignation, leaving it in the planning stage for the remainder of 2012. See Designation of Areas for Air Quality Planning Purposes; State of Arizona; Pinal County; PM10, 77 Fed. Reg. 32024, 32030 (May 31, 2012). Because Pinal County had not yet been required to implement controls, the EPA reasonably concluded that Pinal County’s fugitive dust rules were appropriate for the planning stage. V Finally, we turn to petitioners’ argument that the EPA erred in approving the contingency measures in the Five Percent Plan because those measures had already been implemented. As noted above, four of five contingency measures in the Five Percent Plan were permanent changes to infrastructure that had been completed by 2012 (paving and stabilizing existing public dirt roads and alleys, paving and stabilizing unpaved shoulders, repaving or overlaying paved roads with rubberized asphalt, and lowering speed limits on dirt roads and alleys). The fifth contingency measure involved the purchase of PM-10 certified sweepers and sweeping of freeways, though the purchase occurred in 2009 and Arizona had procured contracts for sweeping services by 2010. The CAA provides that a nonattainment SIP: [S]hall provide for the implementation of specific measures to be undertaken if the area fails to make reasonable further progress, or to attain the national primary ambient air quality standard by the attainment date BAHR V. U.S. EPA 33 applicable under this part. Such measures shall be included in the plan revision as contingency measures to take effect in any such case without further action by the State or the Administrator. 42 U.S.C. § 7502(c)(9). In its 2014 Final Rule, the EPA explained its interpretation of this requirement. While “[c]ontingency measures must provide for additional emission reductions that are not relied on for RFP [reasonable further progress] or attainment and that are not included in the attainment demonstration,” the EPA concluded that “[n]othing in the statute precludes a state from implementing such measures before they are triggered.” 79 Fed. Reg. at 33114. Although we defer to the EPA’s interpretation of the CAA contained in a final rule approving a SIP if that interpretation involves the reasonable interpretation of ambiguous statutory terms, see supra at 22, we cannot defer to the EPA’s interpretation of § 7502(c)(9) here. Where Congress has “directly spoken to the precise question at issue . . . 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.” Chevron, 467 U.S. at 842–43. The statutory language in § 7502(c)(9) is clear: it requires the SIP to provide for the implementation of measures “to be undertaken” in the future, triggered by the state’s failure “to make reasonable further progress” or to attain the NAAQS. These measures are included in the SIP as “contingency measures” and are “to take effect” automatically in the future. Although the statute does not define the word “contingency,” the meaning of the term is not ambiguous. According to the dictionary definition, it means “a possible future event or 34 BAHR V. U.S. EPA condition or an unforeseen occurrence that may necessitate special measures.” Webster’s Third New International Dictionary (2002). Because Congress was clear that “contingency measures” are control measures that will be implemented in the future, and the statutory language is not susceptible to multiple interpretations, we must give effect to its plain meaning. Chevron, 467 U.S. at 842–43. In arguing that its interpretation of § 7502(c) is entitled to deference despite the clear language of the statute, the EPA relies on the Fifth Circuit’s decision in Louisiana Envtl. Action Network v. EPA, 382 F.3d 575, 580 (5th Cir. 2004). In Louisiana Envtl. Action Network, the petitioners challenged the EPA’s 2002 approval of Louisiana’s SIP because the contingency measure in the SIP (a compressor station’s permanent reduction of its emissions) had been implemented in 1998, and therefore was not a measure “to be undertaken” or “to take effect” in the future, as § 7502(c)(9) requires. Id. at 582. The Fifth Circuit first acknowledged that “a plain reading of the terms ‘to take effect’ and ‘to be undertaken’ imply a prospective, forward looking orientation” that would “preclude the use of past reductions which have already failed to achieve attainment.” Id. at 583. Nevertheless, the Fifth Circuit held that § 7502(c)(9) was ambiguous because it “neither affirms nor prohibits continuing emissions reductions—measures which originate prior to the SIP failing, but whose effects continue to manifest an effect after the plan fails—from being utilized as a contingency measure.” Id. (emphasis omitted). Having found this “ambiguity,” it deferred to the EPA’s interpretation that “contingency measures” could include measures that had already been implemented by the state. BAHR V. U.S. EPA 35 We cannot agree with the Fifth Circuit’s interpretative approach. Having determined that the “plain reading of the terms” indicates a forward looking approach, the Fifth Circuit was bound by Chevron to give effect to the plain meaning of the statute. We disagree that the lack of any discussion in § 7502(c)(9) regarding treatment of continuing emissions reductions makes the statute ambiguous. Rather, unless such continuing emissions reductions are “to be undertaken” in the event of a contingency, they do not fit the definition of “contingency measures” provided in § 7502(c)(9). We also disagree with the dissent’s contention that previously implemented control measures that provide continuing emissions reductions “take effect” and are “undertaken” both “at the time they are first implemented but also thereafter.” Dissent at 38. This is a misreading of the statute, which defines contingency measures as measures “to be undertaken” or “to take effect” if a future event occurs, namely “if the area fails to make reasonable further progress, or to attain the [NAAQS].” 42 U.S.C. § 7502(c)(9) (emphases added). Control measures that have already been implemented are not measures “to be undertaken” or “to take effect” in the future, and the statute cannot reasonably be so interpreted. The EPA argues that its interpretation is consistent with the CAA’s policy goals, because permitting early implementation of contingency measures is consistent with the overall policy of the CAA to reduce particulate emissions and protect public health. La. Envtl. Action Network, 382 F.3d at 583. The Fifth Circuit likewise relied on these policy considerations, stating that allowing states to implement measures before the contingency occurs was consistent with the CAA’s purpose of creating incentives for states to reach NAAQS compliance earlier and more efficiently. Id. at 583–84. The dissent agrees, adopting the 36 BAHR V. U.S. EPA Fifth Circuit’s policy analysis.11 Dissent at 39–40. Even if we agreed that the EPA’s policy considerations are compelling, such considerations cannot override the plain language of the statute. We therefore cannot give them controlling weight here. Because the “contingency measures” in Arizona’s SIP were not “specific measures to be undertaken if the area fails to make reasonable further progress, or to attain the national primary ambient air quality standard by the attainment date applicable under this part,” the EPA’s approval of this part of the Five Percent Plan was contrary to the CAA. Accordingly, we remand to the EPA for further consideration of this portion of the SIP but otherwise deny the petition.12 PETITION GRANTED IN PART AND DENIED IN PART. 11 In addition to relying on the Fifth Circuit’s policy arguments, the dissent argues that precluding the use of previously implemented controls as contingency measures “imposes an additional and unnecessary burden upon states where the failure to attain the NAAQS also triggers a bump up of an area’s classification under the Act” because states will not be able to “focus their efforts on implementing the newly imposed requirements.” Dissent at 40. This is incorrect: because contingency measures automatically take effect when the contingency occurs, “without further action by the State,” 42 U.S.C. § 7502(c)(9), the implementation of contingency measures cannot distract a state from meeting the other CAA requirements. 12 Each party is to bear its own costs on appeal. BAHR V. U.S. EPA 37 CLIFTON, Circuit Judge, concurring in part and dissenting in part: I fully concur in sections I–IV of the majority opinion. I disagree, however, with the majority’s conclusion in section V that EPA’s approval of the contingency measures in Arizona’s SIP is contrary to the clear language of the CAA. In my view, the scope of the CAA’s contingency measures requirement is ambiguous and EPA’s reasonable interpretation of that requirement is entitled to deference. Like the majority, I begin by analyzing the text of the relevant statutory provision, 42 U.S.C. § 7502(c)(9). That section of the CAA requires state nonattainment plans to include contingency measures “to be undertaken” and “to take effect” in the event that an area “fails to make reasonable further progress, or to attain” the NAAQS by the applicable attainment date. Id. Although I agree with the majority that this language most often refers to measures that are to be implemented in the future, in the event that the other measures included in a state’s SIP are not sufficient to meet CAA requirements, I am not persuaded that the provision’s text forecloses the interpretation advanced by EPA and applied in this case. The language of the statute prohibits states from labeling as “contingency measures” the same proposed reductions relied upon to achieve NAAQS compliance. See La. Envtl. Action Network v. EPA, 382 F.3d 575, 583 (5th Cir. 2004) (“Such a prospective reading of the text would seemingly preclude the use of past reductions which have already failed to achieve attainment.”). It requires states to identify additional measures that must be put into effect without further action by the state or EPA if reasonable progress is 38 BAHR V. U.S. EPA not made or the air quality standard is not met by the attainment date. Those additional measures are what the statute describes as “contingency measures.” Arizona’s SIP identified additional measures that were not relied upon to obtain the anticipated compliance. The practical issue before us in this case is whether Arizona was prohibited from putting those additional measures into effect in advance. The majority opinion concludes that it was, that the state’s contingency measures must be left undone, sitting on the sidelines in reserve. I do not believe that the language or intent of the statute requires that conclusion. The early implementation of infrastructure improvements that are expected to result in additional and continuing emissions reductions is consistent with the language of § 7502(c)(9). These early-implemented contingency measures result in a net reduction in emissions following their implementation. They “take effect” and are “undertaken” not only at the time they are first implemented but also thereafter, including at the time they might formally be required due to nonattainment. So long as these reductions are not relied upon to meet other CAA requirements, they function as a backup plan, reducing the likelihood that the state will fail to attain the NAAQS even in the “contingency” that the measures explicitly included in the SIP for that purpose are not enough. The majority responds by asserting that the statute “defines contingency measures as measures ‘to be undertaken’ or ‘to take effect’ if a future event occurs, namely ‘if the area fails to make reasonable further progress, or to attain the [NAAQS].’” Majority opinion at 35 (emphasis in original). But this language fits just as well with EPA’s interpretation as it does with the view of the majority. In both BAHR V. U.S. EPA 39 scenarios, the contingency measures must be in effect at the time an area fails to achieve the goals outlined in the SIP. What is at issue here is whether states are prohibited from also implementing the measures before that “future event occurs.” The language quoted by the majority contains no such prohibition and it is not our role to read one into the statute. EPA’s interpretation also comports well with the purpose of the CAA. “In determining whether Congress has specifically addressed the question at issue, a reviewing court should not confine itself to examining a particular statutory provision in isolation. . . . It is a ‘fundamental canon of statutory construction that the words of a statute must be read in their context and with a view to their place in the overall statutory scheme.’” FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 132–33 (2000) (quoting Davis v. Mich. Dept. of Treasury, 489 U.S. 803, 809 (1989)). Allowing states to implement contingency measures before they are triggered makes sense in light of that same provision’s requirement that such measures “take effect . . . without further action by the State or the Administrator.” That requirement, read in context with the Act’s mandate that states implement emission-control measures “as expeditiously as practicable,” id. § 7502(c)(1), evinces a clear congressional preference that contingency measures operate to reduce the emission of harmful pollutants in as efficient and timely a manner as possible. Taking this statutory purpose into account, “it seems illogical to penalize nonattainment areas that are taking extra steps, such as implementing contingency measures prior to a deadline” as a cushion to ensure NAAQS compliance and prevent the need for the contingency measures requirement to be triggered in the first place. La. 40 BAHR V. U.S. EPA Envtl. Action Network, 382 F.3d at 584. It does not benefit ordinary citizens to read the CAA to incentivize states to hold off from purchasing new street sweepers or repaving their roads until the contingency measures requirement is triggered. The majority’s interpretation also imposes an additional and unnecessary burden upon states in circumstances where the failure to attain the NAAQS also triggers a bump up of an area’s classification under the Act. For example, the failure of a moderate nonattainment area to achieve NAAQS compliance by the applicable deadline triggers both § 7502(c)(9)’s contingency measures requirement and the additional requirements imposed upon serious nonattainment areas. 42 U.S.C. § 7513(b)(2). The early implementation of contingency measures allows states in this situation to focus their efforts on implementing the newly imposed requirements while the contingency measures continue to operate in the interim. See La. Envtl. Action Network, 382 F.3d at 583.1 1 The majority argues that because contingency measures must take effect “without further action by the State” or EPA, 42 U.S.C. § 7502(c)(9), they “cannot distract a state from meeting the other CAA requirements.” Majority opinion at 36 n. 11. However, EPA has interpreted this language to require only “that no further rulemaking activities by the State or EPA would be needed to implement the contingency measures.” Greenbaum v. EPA, 370 F.3d 527, 541 (6th Cir. 2004) (quoting State Implementation Plans; General Preamble for the Implementation of Title I of the Clean Air Act Amendments of 1990, 57 Fed. Reg. 13498, 13512 (Apr. 16, 1992)). Thus, under the majority’s interpretation, states would still be required to perform the actual business of implementing the identified contingency measures at the same time they address other requirements imposed by a failure to attain the NAAQS. BAHR V. U.S. EPA 41 I recognize that what is lost by EPA’s interpretation of the statute is the advance identification of “still more” measures aimed at improving air quality that can be newly and additionally implemented in the event of nonattainment. If Arizona’s SIP does not reach its goal, then reliance upon contingency measures that have already been put into effect will not further reduce airborne particulate matter, and more will have to be done at that point. But EPA applied its expertise and exercised its judgment in concluding that the goal was likely to be reached by the measures proposed in the SIP, without regard to the additional contingency measures. By letting Arizona identify actions that have already been implemented as contingency measures, EPA has obtained a cushion. It is not an unreasonable judgment for EPA to conclude that implementing the cushion right away is more valuable than advance identification of what else might be done, if necessary. For these reasons, I would give Chevron deference to the EPA’s interpretation of § 7502(c)(9). I respectfully dissent from the majority’s conclusion to the contrary.

Some case metadata and case summaries were written with the help of AI, which can produce inaccuracies. You should read the full case before relying on it for legal research purposes.

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.