Full opinion text
BETTY B. FLETCHER, Circuit Judge: Eleven states, the District of Columbia, the City of New York, and four public interest organizations petition for review of a rule issued by the National Highway Traffic Safety Administration (NHTSA) entitled “Average Fuel Economy Standards for Light Trucks, Model Years 2008-2011,” 71 Fed.Reg. 17,566 (Apr. 6, 2006) (“Final Rule”) (codified at 49 C.F.R. pt. 533). Pursuant to the Energy Policy and Conservation Act of 1975 (EPCA), 49 U.S.C. §§ 32901-32919 (2007), the Final Rule sets corporate average fuel economy (CAFE) standards for light trucks, defined by NHTSA to include many Sport Utility Vehicles (SUVs), minivans, and pickup trucks, for Model Years (MYs) 2008-2011. For MYs 2008-2010, the Final Rule sets new CAFE standards using its traditional method, fleet-wide average (Unreformed CAFE). For MY 2011 and beyond, the Final Rule creates a new CAFE structure that sets varying fuel economy targets depending on vehicle size and requires manufacturers to meet different fuel economy levels depending on their vehicle fleet mix (Reformed CAFE). Petitioners challenge the Final Rule under the EPCA and the National Environmental Policy Act of 1969 (NEPA), 42 U.S.C. §§ 4321-4347 (2007). First, they argue that the Final Rule is arbitrary, capricious, and contrary to the EPCA because (a) the agency’s cost-benefit analysis does not set the CAFE standard at the “maximum feasible” level and fails to give due consideration to the need of the nation to conserve energy; (b) its calculation of the costs and benefits of alternative fuel economy standards assigns zero value to the benefit of carbon dioxide (C02) emissions reduction; (c) its calculation of costs and benefits of alternative fuel economy standards fails to evaluate properly the benefit of vehicle weight reduction; (d) Reformed CAFE standards will depend on manufacturer fleet mix and not guarantee a minimum average fuel economy or “backstop”; (e) the transition period during which manufacturers may choose to comply with either Unreformed or Reformed CAFE is contrary to the “maximum feasible” requirement and unnecessary; (f) it perpetuates the “SUV loophole,” which allows SUVs, minivans, and pickup trucks to satisfy a lower fuel economy standard than cars; and (g) it excludes most vehicles rated between 8,500 and 10,000 pounds gross vehicle weight (comprised mostly of large pickup trucks) from any fuel economy regulation, even though these vehicles satisfy the statutory criteria for regulation. Second, Petitioners argue that NHTSA’s Environmental Assessment is inadequate under NEPA because it fails to take a “hard look” at the greenhouse gas implications of its rulemaking and fails to analyze a reasonable range of alternatives or examine the rule’s cumulative impact. Petitioners also argue that NEPA requires NHTSA to prepare an Environmental Impact Statement. NHTSA argues that the Final Rule is not arbitrary and capricious or contrary to the EPCA, the Environmental Assessment’s evaluation of the environmental consequences of its action is adequate, and an Environmental Impact Statement is not required. We have jurisdiction under 49 U.S.C. § 32909(a) to review the Final Rule issued by NHTSA. We hold that the Final Rule is arbitrary and capricious, contrary to the EPCA in its failure to monetize the value of carbon emissions, failure to set a backstop, failure to close the SUV loophole, and failure to set fuel economy standards for all vehicles in the 8,500 to 10,000 gross vehicle weight rating (“GVWR”) class. We also hold that the Environmental Assessment was inadequate and that Petitioners have raised a substantial question as to whether the Final Rule may have a significant impact on the environment. Therefore, we remand to NHTSA to promulgate new standards as expeditiously as possible and to prepare a full Environmental Impact Statement. I. FACTUAL AND PROCEDURAL BACKGROUND A. CAFE Regulation Under the Energy Policy and Conservation Act In the aftermath of the energy crisis created by the 1973 Mideast oil embargo, Congress enacted the Energy Policy and Conservation Act of 1975, Pub.L. No. 94-163, 89 Stat. 871, 901-16. See H.R.Rep. No. 94-340 at 1-3 (1975), as reprinted in 1975 U.S.C.C.A.N. 1762, 1763-65. Congress observed that “[t]he fundamental reality is that this nation has entered a new era in which energy resources previously abundant, will remain in short supply, retarding our economic growth and necessitating an alteration in our life’s habits and expectations.” Id. at 1763. The goals of the EPCA are to “decrease dependence on foreign imports, enhance national security, achieve the efficient utilization of scarce resources, and guarantee the availability of domestic energy supplies at prices consumers can afford.” S.Rep. No. 94-516 (1975) (Conf.Rep.), as reprinted in 1975 U.S.C.C.A.N. 1956, 1957. These goals are more pressing today than they were thirty years ago: since 1975, American consumption of oil has risen from 16.3 million barrels per day to over 20 million barrels per day, and the percentage of U.S. oil that is imported has risen from 35.8 to 56 percent. NRDC Cmt. at 11; see also 71 Fed.Reg. at 17,644. In furtherance of the goal of energy conservation, Title V of the EPCA establishes automobile fuel economy standards. An “average fuel economy standard” (often referred to as a CAFE standard) is “a performance standard specifying a minimum level of average fuel economy applicable to a manufacturer in a model year.” 49 U.S.C. § 32901(a)(6) (2007). Only “automobiles” are subject to fuel economy regulation, and passenger automobiles must meet a statutory standard of 27.5 mpg, 49 U.S.C. § 32902(b), whereas non-passenger automobiles must meet standards set by the Secretary of Transportation, id. § 32902(a). Congress directs the Secretary to set fuel economy standards at “the maximum feasible average fuel economy level that the Secretary decides the manufacturers can achieve in that model year.” Id. § 32902(a). Under this subsection, the Secretary is authorized to “prescribe separate standards for different classes of automobiles.” Id. Congress also provides that “[wjhen deciding maximum feasible average fuel economy under this section, the Secretary of Transportation shall consider technological feasibility, economic practicability, the effect of other motor vehicle standards of the Government on fuel economy, and the need of the United States to conserve energy.” Id. § 32902(f). Under the EPCA’s definitional scheme, vehicles not manufactured primarily for highway use and vehicles rated at 10,000 lbs. gross vehicle weight or more are excluded from fuel economy regulation altogether because they are not “automobiles.” An “automobile” is defined as: a 4-wheeled vehicle that is propelled by fuel, or by alternative fuel, manufactured primarily for use on public streets, roads, and highways ..., and rated at— (A) not more than 6,000 pounds gross vehicle weight; or (B) more than 6,000, but less than 10,-000, pounds gross vehicle weight, if the Secretary decides by regulation that— (i) an average fuel economy standard under this chapter for the vehicle is feasible; and (ii) an average fuel economy standard under this chapter for the vehicle will result in significant energy conservation or the vehicle is substantially used for the same purposes as a vehicle rated at not more than 6,000 pounds gross vehicle weight. 49 U.S.C. § 32901(a)(3). Although NHTSA has the authority to regulate the fuel economy of vehicles up to 10,000 lbs. GVWR, see id. § 32901(a)(3)(B), the agency has excluded vehicles exceeding 8,500 lbs. (other than medium-duty passenger vehicles manufactured during MY 2011 or thereafter) from its definition of “automobile,” see 49 C.F.R. § 523.3(b). The CAFE standards NHTSA sets for non-passenger automobiles or “light trucks,” as referred to by the agency in its regulations, are lower than the standards for passenger automobiles. Compare 49 C.F.R. § 533.5(a) (2007) with 49 C.F.R. § 531.5(a) (2007). A “passenger automobile” is defined as: an automobile that the Secretary decides by regulation is manufactured primarily for transporting not more than 10 individuals, but does not include an automobile capable of off-highway operation that the Secretary decides by regulation— (A) has a significant feature (except dr-wheel drive) designed for off-highway operation; and (B) is a 4-wheel drive automobile or is rated at more than 6,000 pounds gross vehicle weight. 49 U.S.C. § 32901(a)(16). The Final Rule sets CAFE standards for “light trucks,” defined by NHTSA to include many SUVs, vans, and pickup trucks, for MYs 2008-2011. See 71 Fed. Reg. at 17,568; 49 C.F.R. § 533.5(a), (g), (h). A “light truck” is: an automobile other than a passenger automobile which is either designed for off-highway operation, as described in paragraph (b) of this section, or An automobile capable of off-highway operation is an automobile— (l)(i) That has 4-wheel drive; or designed to perform at least one of the following functions: (1) Transport more than 10 persons; (2) Provide temporary living quarters; (3) Transport property on an open bed; (4) Provide greater cargo-carrying than passenger-carrying volume; or (5) Permit expanded use of the automobile for cargo-purposes or other non-passenger-carrying purposes through [removable or foldable, stowable seats to create a flat floor], 49 C.F.R. § 523.5(a) (2007). For MYs 1996 to 2004, Congress froze the light truck CAFE standard at 20.7 mpg. See 71 Fed.Reg. at 17,568. After the legislative restrictions were lifted, NHTSA set new light truck CAFE standards in April 2003: 21.0 mpg for MY 2005, 21.6 mpg for MY 2006, and 22.2 mpg for MY 2007. Light Truck Average Fuel Economy Standards Model Years 2005-2007, 68 Fed.Reg. 16,868, 16,871 (Apr. 7, 2003) (codified at 49 C.F.R. pt. 533). In response to a request from Congress, the National Academy of Sciences (NAS) published in 2002 a report entitled “Effectiveness and Impact of Corporate Average Fuel Economy (CAFE) Standards.” The NAS committee made several findings and recommendations. It found that from 1970 to 1982, CAFE standards helped contribute to a 50 percent increase in fuel economy for new light trucks. Id. at 14. In the subsequent decades, however, light trucks became more popular since domestic manufacturers faced less competition in the light truck category and could generate greater profits. Id. at 18-19. The “less stringent CAFE standards for trucks ... provide[d] incentives for manufacturers to invest in minivans and SUVs and to promote them to consumers in place of large cars and station wagons.” Id. at 18. When the CAFE regulations were originally promulgated in the 1970s, “light truck sales accounted for about 20 percent of the new vehicle market,” but now they account for about half. Id. at 88. This shift has had a “pronounced” effect on overall fuel economy. Id. at 19. As the market share of light trucks has increased, the overall average fuel economy of the new light duty vehicle fleet (light trucks and passenger automobiles) has declined “from a peak of 25.9 MPG in 1987 to 24.0 MPG in 2000.” Id. Vehicle miles traveled (VMT) by light trucks has also been growing more rapidly than passenger automobile travel. Id. The NAS committee found that the CAFE program has increased fuel economy, but that certain aspects of the program “have not functioned as intended,” including “[t]he distinction between a car for personal use and a truck for work use/cargo transport,” which “has been stretched well beyond the original purpose.” Id. at 3. The committee also found that technologies exist to “significantly reduce fuel consumption,” for cars and light trucks and that raising CAFE standards would reduce fuel consumption. Id. at 3-4. Significantly, the committee found that of the many reasons for improving fuel economy, “[t]he most important ... is concern about the accumulation in the atmosphere of so-called greenhouse gases, principally carbon dioxide. Continued increases in carbon dioxide emissions are likely to further global warming.” Id. at 2, In addition, the committee found “exter-nalities of about $0.30/gal of gasoline associated with the combined impacts of fuel consumption on greenhouse gas emissions and on world oil market conditions” that “are not necessarily taken into account when consumers purchase new vehicles.” Id. at 4. B. National Environmental Policy Act NEPA requires a federal agency “to the fullest extent possible,” to prepare “a detailed statement on ... the environmental impact” of “major Federal actions significantly affecting the quality of the human environment.” 42 U.S.C. § 4332(2)(C)(i) (2007); see also 40 C.F.R. § 1500.2 (2007). The purpose of NEPA is twofold: “ ‘ensure! ] that the agency ... will have available, and will carefully consider, detailed information concerning significant environmental impacts!, and] guarantee! ] that the relevant information will be made available to the larger [public] audience.’ ” Idaho Sporting Cong. v. Thomas, 137 F.3d 1146, 1149 (9th Cir.1998) (quoting Robertson v. Methow Valley Citizens Council, 490 U.S. 332, 349, 109 S.Ct. 1835, 104 L.Ed.2d 351 (1989)); see also 40 C.F.R. § 1500.1(b) (stating that environmental information must be provided “before decisions are made and before actions are taken.”). “NEPA expresses a Congressional determination that procrastination on environmental concerns is no longer acceptable.” Found. for N. Am. Wild Sheep v. U.S. Dep’t of Agric., 681 F.2d 1172, 1181 (9th Cir.1982). NEPA “is our basic national charter for protection of the environment.” 40 C.F.R. § 1500.1(a). If there is a substantial question whether an action “may have a significant effect” on the environment, then the agency must prepare an Environmental Impact Statement (EIS). See, e.g., Blue Mountains Biodiversity Project v. Blackwood, 161 F.3d 1208, 1212 (9th Cir.1998) (internal quotation marks omitted). An EIS should contain a discussion of significant environmental impacts and alternatives to the proposed action. See 40 C.F.R. §§ 1502.1, 1502.14, 1508.7. As a preliminary step, an agency may prepare an Environmental Assessment (EA) in order to determine whether a proposed action may “significantly affect[ ]” the environment and thereby trigger the requirement to prepare an EIS. See 40 C.F.R. § 1508.9(a)(1) (2007). An EA is “a concise public document” that “[b]riefly provide[s] sufficient evidence and analysis for determining whether to prepare an environmental impact statement or a finding of no significant impact.” Id. An EA “[s]hall include brief discussions of the need for the proposal, of alternatives as required by sec. 102(2)(E), of the environmental impacts of the proposed action and alternatives, and a listing of agencies and persons consulted.” Id. § 1508.9(b). Whether an action may “significantly affect” the environment requires consideration of “context” and “intensity.” Id. § 1508.27; see also Nat’l Parks & Conservation Ass’n v. Babbitt, 241 F.3d 722, 731 (9th Cir.2001). “Context ... delimits the scope of the agency’s action, including the interests affected.” Nat’l. Parks & Conservation Ass’n, 241 F.3d at 731. Intensity refers to the “severity of impact,” which includes both beneficial and adverse impacts, “[t]he degree to which the proposed action affects public health or safety,” “[t]he degree to which the effects on the quality of the human environment are likely to be highly controversial,” “[t]he degree to which the possible effects on the human environment are highly uncertain or involve unique or unknown risks,” and “[wjhether the action is related to other actions with individually insignificant but cumulatively significant impacts.” 40 C.F.R. § 1508.27(b)(2), (4), (5), (7). C. NHTSA’s Proposed Rulemaking and Draft Environmental Assessment On December 29, 2003, NHTSA published an advance notice of proposed rulemak-ing (ANPRM) that solicited comments on several proposed regulatory changes intended to increase fuel economy, including a proposal to modernize the light truek/car distinction and a proposal to increase the GVWR limit on vehicles subject to CAFE standards. 68 Fed.Reg. 74,908 (Dec. 29, 2003). NHTSA acknowledged that its regulations define passenger and non-passenger vehicles “by the type of use to which they were generally put in the mid-1970s,” id. at 74,909, and that “[tjhe markets for, and designs of, cars and light trucks have changed substantially,” with “some light trucks ... used primarily to transport passengers,” id. at 74,913. NHTSA noted that in its original NPRM promulgated in December 1976, it concluded that “Congress intended that passenger automobiles be defined as those used primarily for the transport of individuals and that all other vehicles would fall within the category of non-passenger automobiles.” Id. at 74,926. NHTSA did not present any specific proposals for reforming the CAFE program, but it presented two options for including vehicles under 10,000 lbs. GVWR in the program: (1) regulating medium-duty passenger vehicles (MDPVs), which are vehicles between 8,500 and 10,000 lbs. GVWR designed primarily for the transportation of persons, and (2) regulating all vehicles between 8,500 and 10,000 lbs. GVWR. Id. at 74,930. On August 30, 2005, NHTSA issued proposed CAFE standards for light trucks MYs 2008-2011 of 22.5 mpg for MY 2008, 23.1 mpg for MY 2009, and 23.5 mpg for MY 2010. 70 Fed.Reg. 51,414, 51,424 (Aug. 30, 2005). NHTSA determined that these were the “maximum feasible” standards using a marginal cost-benefit analysis. See id. For MY 2011 and beyond, NHTSA proposed to adopt a “Reformed CAFE” system, which would set different CAFE standards for vehicles based on size, measured by the vehicle’s footprint (the product of multiplying wheelbase by track width). Id. at 51,414, 51,429-41. NHTSA proposed six footprint categories (a step function), id. at 51,430, and it proposed a transition period (MY 2008-2010) to Reformed CAFE, during which manufacturers could choose to comply with either Reformed or Unreformed CAFE. NHTSA also proposed not to change the criteria by which vehicles are classified as passenger automobiles or light trucks, id. at 51,422, and it proposed to regulate only MDPVs within the 8,500 to 10,000 lb. vehicle class as light trucks, id. at 51,455-56. NHTSA issued a Draft Environmental Assessment in August 2005. The Draft EA integrated much of the text from the Final EA that accompanied NHTSA’s light truck rulemaking for MYs 2005-2007 released in April 2003. See Draft Environmental Assessment, NHTSA Proposed Corporate Average Fuel Economy (CAFE) Standards 9 (Aug.2005) (Draft EA). The Draft EA analyzed three alternatives to the proposed rule. Alternative A (“No Action”) would extend the MY 2007 standard of 22.2 mpg through MY 2011. Alternative B would be Unreformed CAFE in MY 2008-2010 and Reformed CAFE in MY 2011. Alternative C would be Reformed CAFE set at equalized cost with Unreformed CAFE in MY 2008-2010 and Reformed CAFE in MY 2011. Id. The Draft EA noted that “C02 ... has started to be viewed as an issue of concern for its global climate change potential.” Id. at 18. With regard to biological resources, the Draft EA stated, “emissions of criteria pollutants and greenhouse gases could result in ozone layer depletion and promote climate change that could affect species and ecosystems.” Id. at 19. The projected lifetime fuel savings for MY 2008-2011 light trucks under Alternatives B and C would “rang[e] from 1.3% to 1.7% of their fuel compared to the baseline, corresponding to 4.7-6.0 billion gallons.” Id. at 25. The estimated lifetime emissions of C02 ranged from 1,341.4 million metric tons (mmt) under baseline to 1,306.4 and 1,304.0 mint under Alternatives B and C, respectively. Id. at 29. The Draft EA concluded that the proposed standards would “result in reduced emissions of C02, the predominant greenhouse gas emitted by motor vehicles,” “reductions in contamination of water resources,” and “minor reductions in impacts to biological resources.” Id. at 30-31. In addition, “the cumulative effects estimated to result from both the 2005-2007 and 2008-2011 light truck rulemakings over the lifetimes of the vehicles they would affect are projected to be very small.” Id. at 34. NHTSA received over 45,000 comments on the NPRM and Draft EA from states, consumer and environmental organizations, automobile manufacturers and associations, members of Congress, and private individuals. See 71 Fed.Reg. at 17,577. Manufacturers argued that reliance on a cost-benefit analysis might not “adequately account for the capabilities of the industry.” Id. They also generally opposed subjecting vehicles greater than 8,500 lbs. GVWR to CAFE regulation, arguing that those vehicles are used in a different manner than lighter vehicles and that their regulation would not result in significant fuel savings. Id. at 17,577-78. The states and environmental and consumer organizations generally argued that: • The need of the nation to conserve energy and national security require more stringent standards, and such standards are feasible and practicable. E.g., NRDC Cmt. at 4-5; Environmental Defense. Cmt. at 1-7; Public Citizen Cmt. at 1-2. For example, the Alliance to Save Energy' — American Council for an Energy-Efficient Economy (ACEEE) argued that “[t]he 10 billion gallons of fuel that NHTSA claims will be saved through the new standards over the three-decade life of model year 2008-2011 vehicles amount to less than one month’s supply of gasoline for U.S. light-duty vehicles. These savings are also insufficient to offset the expected growth in gasoline usage for even the four-year period covered by the rule.” ACEEE Cmt. at l. • NHTSA’s use of marginal cost-benefit analysis unlawfully overemphasizes cost at the expense of technological feasibility and energy conservation and is not “technology-forcing,” as EPCA intended. E.g., NRDC Cmt. at 14-16; Environmental Defense Cmt. at 4-5; Public Citizen Cmt. at 1-2. • Even if NHTSA’s cost-benefit analysis is permissible, the “maximum feasible” standard cannot be determined properly without taking environmental impacts into account, and the failure to monetize certain benefits such as greenhouse gas (GHG) emissions underestimates benefits of stricter standards. E.g., CBD Cmt. at 1-4; NRDC Cmt. at 8 (suggesting specific figures and sources for the value per ton of C02 emissions avoided, from $8/ton to $26.50/ton); Environmental Defense Cmt. at 5-6; Environmental Defense Cmt. Re: Carbon Costs at 1-3 (citing new studies from the United Kingdom that value carbon at $96-174/ton carbon). • Reformed CAFE “rewards fuel economy laggards while penalizing industry leaders,” Sierra Club Cmt. at 4, and, like Unreformed CAFE, promotes the manufacture of larger, less fuel-efficient vehicles. E.g., App. G to NRDC Cmt. at 3-4; States Cmt. at 5; Environmental Defense Cmt. at 12-13. • NHTSA’s analysis of the adverse safety effects of vehicle weight reduction is flawed and confounds size and weight. E.g., Sierra Club Cmt. at 8-10; App. C to Environmental Defense Cmt. at 1-4; Public Citizen Cmt. at 12-19; App. B to Public Citizen Cmt. at 1-16. • Since the Reformed CAFE standard for a particular manufacturer depends on its fleet mix, NHTSA should include a “backstop” or guarantee that average fuel economy levels will not fall below the “maximum feasible” level. E.g., NRDC Cmt. at 24-25; ACEEE Cmt. at 5; see also App. E to NRDC Cmt. at 6-12 (analysis of gaming scenarios and upsizing trends); Environmental Defense Cmt. at 13-14 (same); App. G. to Environmental Defense Cmt. at 1-14 (same). • The transition period (MY 2008-2010) to Reformed CAFE is unnecessary and undercuts fuel economy savings. E.g., NRDC Cmt. at 27-28; ACEEE Cmt. at 2; Environmental Defense Cmt. at 8-9; UCS Cmt. at 9. • All Class 2b trucks between 8,500-10,000 lbs. GVWR should be regulated because fuel economy standards for them are feasible, would result in significant energy conservation, and they are used for substantially the same purposes as vehicles 6,000 lbs. or less. Environmental Defense Cmt. at 9-11; App. F to Environmental Defense Cmt. at 1-2; and Polk Study. • Higher fuel economy standards would help domestic automakers remain competitive in the long term and protect U.S. jobs. App. D to NRDC Cmt. at 22. The California Energy Resources Conservation and Development Commission commented that “[ujpgrading CAFE requirements could enhance jobs in the United States, especially in the automobile manufacturing sector.... Increasing lighttruck CAFE to 26.9 mpg in 2010 and 31 mpg in 2015 (with corresponding changes for passenger cars) would increase net jobs up to 346,000.” California Energy Commission Cmt. at 9-10. • NHTSA’s draft EA is inadequate and fails to consider the proposed rule’s impact on climate change. States Cmt. at 1-11; CBD Cmt. at 5-12. See also 71 Fed.Reg. at 17,578-79 (summarizing comments). Commenters also submitted to NHTSA numerous scientific reports and studies regarding the relationship between climate change and greenhouse gas emissions and the expected impacts on the environment. Emissions from light trucks make up about eight percent of annual U.S. greenhouse gas emissions. Final EA at 22 (citing EPA, EPA-430-R-05-003, Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2004 (Draft 2006)). The transportation sectors account for about 31 percent of human-generated C02 emissions in the U.S. economy. NAS Report at 14. “Overall, U.S. light-duty vehicles [passenger cars and light trucks] produce about 5 percent of the entire world’s greenhouse gases.” Id. at 20. The NAS committee concluded, “Since the United States produces about 25 percent of the world’s greenhouse gases, fuel economy improvements could have a significant impact on the rate of C02 accumulation in the atmosphere.” Id. at 14. The Intergovernmental Panel on Climate Change (IPCC)’s “Third Assessment Report,” published in 2001, presented the consensus view of hundreds of scientists on key issues relating to climate change. The IPCC concluded that “C02 concentrations increasing over [the] 21st century[are] virtually certain to be mainly due to fossil-fuel emissions,” and that “[stabilization of atmospheric C02 concentrations at 450, 650, or 1,000 ppm would require global anthropogenic C02 emissions to drop below year 1990 levels, within a few decades, about a century, or about 2 centuries, respectively, and continue to decrease steadily thereafter to a small fraction of current emissions.” Id. The average earth surface temperature has increased by about 0.6 degree Celsius since the late 19th century, see Technical Summary of IPCC Working Group I Report at 26; snow and ice cover have decreased about 10 percent since the late 1960s, id. at 30; and global average sea level has risen between 10 to 20 cm during the 20th century, id. at 31. The IPCC also developed a range of emissions scenarios as its basis for predicting the environmental effect of increased emissions. Id. at 62-63. More recent evidence shows that there have already been severe impacts in the Arctic due to warming, including sea ice decline. See Attachments J, L, & Q to CBD Cmt. Global warming has already affected plants, animals, and ecosystems around the world. See, e.g., Attachment M to CBD Cmt. at 15-16. Some scientists predict that “on the basis of mid-range climate-warming scenarios for 2050, that 15-37% of species in our sample of regions and taxa will be ‘committed to extinction.’ ” Attachment O to CBD Cmt (Thomas, Ax-tinction Risk from Climate Change, 427 Nature at 145). In addition, there will be serious consequences for human health, including the spread of infectious and respiratory diseases, if worldwide emissions continue on current trajectories.- See, e.g., Attachment F to CBD Cmt. at 32-64. Sea level rise and increased ocean temperatures are also associated with increasing weather variability and heightened intensity of storms such as hurricanes. Id. at 21-24. Past projections have underestimated sea level rise. See id. at 20. Several studies also show that climate change may be non-linear, meaning that there are positive feedback mechanisms that may push global warming past a dangerous threshold (the “tipping point”). Id. at 26-27; see also Technical Summary of IPCC Working Group I Report at 46-53; Attachment F to CBD Cmt. at 26-27; IPCC Report at 14-16; States Cmt. 9. D. The Final Rule: CAFE Standards for Light Trucks MYs 2008-2011 NHTSA issued the Final Rule on April 6, 2006. 71 Fed.Reg. at 17,566. NHTSA set the CAFE standards for MY 2008-2010 (Unreformed CAFE) at the same levels as proposed in the NPRM. Unreformed CAFE sets a fleet-wide average fuel economy standard “with particular regard to the ‘least capable manufacturer with a significant share of the market.’ ” 71 Fed.Reg. at 17,580. NHTSA has reformed the structure of the CAFE program for light trucks, effective MY 2011 (Reformed CAFE). Under Reformed CAFE, fuel economy standards are based on a truck’s footprint, with larger footprint trucks subject to a lower standard and smaller footprint trucks subject to higher standards. 71 Fed.Reg. at 17,566. Instead of six footprint categories (a step function) as proposed in the NPRM, Reformed CAFE would be based on a continuous function, meaning a separate fuel economy target for each vehicle of a different footprint. See id. at 17,595-96. “A particular manufacturer’s compliance obligation for a model year will be calculated as the harmonic average of the fuel economy targets for the manufacturer’s vehicles, weighted by the distribution of manufacturer’s production volumes among the footprint increments.” Id. at 17,566. A manufacturer’s CAFE compliance obligation will vary with its fleet mix. A manufacturer that produces more large footprint light trucks will have a lower required CAFE standard than one that produces more small footprint light trucks. During MYs 2008-2010, manufacturers may choose to comply with Unreformed CAFE or Reformed CAFE. See id. at 17,593-94. NHTSA used the manufacturers’ preexisting product plans as the baseline for its analyses of technical and economic feasibility under both Unreformed and Reformed CAFE. Id. at 17,579. NHTSA made adjustments to the product plans by applying additional technologies in a “cost-minimizing fashion,” id. at 17,582, and stopping at the point where marginal costs equaled marginal benefits, id. at 17,597. NHTSA considered the cost of new technologies and the benefits of fuel savings over the lifetime of the vehicle as the costs and benefits of higher fuel economy standards. Id. at 17,585-87, 17,622-28. NHTSA monetized some externalities such as emission of criteria pollutants during gasoline refining and distribution and crash and noise costs associated with driving. See Final Regulatory Impact Analysis, Corporate Average Fuel Economy and CAFE Reform for MY 2008-2011 Light Trucks at VIII-60, VIII-74-80 (March 2006) (FRIA). However, NHTSA did not monetize the benefit of reducing carbon dioxide emissions, which it recognized was the “the main greenhouse gas emitted as a result of refining, distribution, and use of transportation fuels.” FRIA at VIII-61 to 62. NHTSA acknowledged the estimates suggested in the scientific literature, see 71 Fed.Reg. at 17,638; FRIA at VIII-63, but concluded: [T]he value of reducing emissions of C02 and other greenhouse gases [is] too uncertain to support their explicit valuation and inclusion among the savings in environmental externalities from reducing gasoline production and use. There is extremely wide variation in published estimates of damage costs from greenhouse gas emissions, costs for controlling or avoiding their emissions, and costs of sequestering emissions that do occur, the three major sources for developing estimates of economic benefits from reducing emissions of greenhouse gases. 71 Fed.Reg. at 17,638; see also FRIA at VIII-64 to 65. In its cost-benefit analysis, NHTSA also excluded weight reduction for vehicles between 4,000 and 5,000 lbs. curb weight as a potential measure that manufacturers could use to increase fuel economy. 71 Fed.Reg. at 17,627. NHTSA accepted the possibility of weight reduction for vehicles over 5,000 lbs. curb weight as a cost-effective technology that would not reduce overall safety. Id. NHTSA relied on a study by Dr. Charles Kahane for this 5,000 lb. figure: [T]he net safety effect of removing 100 pounds from a light truck is zero for light trucks with a curb weight greater than 3,900 lbs. However, given the significant statistical uncertainty around that figure, we assumed a confidence bound of approximately 1,000 lbs. and used 5,000 lbs. as the threshold for considering weight reduction. Id. (footnotes omitted). By “net safety effect,” NHTSA means that 3,900 lbs. is the breakeven point: “the point where the total effect of reducing all vehicles heavier than the breakeven weight by an equal amount is zero.” Id. at 17,628. In the FRIA, NHTSA explained that it chose the approximately 1,000 lb. confidence bound based on additional empirical work found in Kahane’s study: Kahane estimated a crossover weight of 5,085 lbs. if manufacturers changed both weight and footprint, and the interval estimated ranged from 4,224 lbs. to 6,121 lbs[.], i.e., an interval +/-1000 lbs[J around the point estimate. Although the crossover weight differs from the point of zero net impact, they would both tend to have similar sampling errors. We applied this interval to the 3,900 lbs. point of zero net impact (which is based on the assumption that footprint is held constant); therefore, the agency felt it would be prudent to limit weight reductions to those vehicles above 5,000 lbs. curb weight. FRIA at V-15 (internal citation omitted). NHTSA rejected the idea of a “backstop” under Reformed CAFE. 71 Fed.Reg. at 17,592; id. at 17,617. NHTSA stated that a backstop, or a required fuel economy level applicable to a manufacturer if its required level under Reformed CAFE fell below a certain minimum, “would essentially be the same as an Unreformed CAFE standard.” Id. at 17,592. NHTSA argued that “EPCA permits the agency to consider consumer demand and the resulting market shifts in setting fuel economy standards,” id. at 17,593, and that a backstop “would essentially limit the ability of manufacturers to respond to market shifts arising from changes in consumer demand. If consumer demand shifted towards larger vehicles, a manufacturer potentially could be faced with a situation in which it must choose between limiting its production of the demanded vehicles, and failing to comply with the CAFE light truck standard.” Id. Finally, NHTSA declined to change the regulatory definition of cars and light trucks to close the SUV loophole and refused to regulate vehicles between 8,500 and 10,000 lbs. GWVR, other than MDPVs. See id. at 17,574. II. STANDARD OF REVIEW The Administrative Procedure Act (APA), 5 U.S.C. §§ 701-706 (2007), provides that agency action must be set aside by the reviewing court if it is “ ‘arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.’ ” Competitive Enter. Inst. v. NHTSA (CEI III), 45 F.3d 481, 484 (D.C.Cir.1995) (quoting 5 U.S.C. § 706(2)(A)) (applying the APA to review a rulemaking under the EPCA). The scope of review is narrow, but “the agency must examine the relevant data and articulate a satisfactory explanation for its action including a ‘rational connection between the facts found and the choice made.’ ” Motor Vehicle Mfrs. Ass’n v. State Farm, Mut. Auto. Ins. Co., 463 U.S. 29, 43, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983) (citation omitted). An agency rule would normally be arbitrary and capricious if: the agency has relied on factors which Congress has not intended it to consider, entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise. Id. The reviewing court “ ‘may not supply a reasoned basis for the agency’s action that the agency itself has not given.’ ” Id. (quoting SEC v. Chenery Corp., 332 U.S. 194, 196, 67 S.Ct. 1575, 91 L.Ed. 1995 (1947)). If Congress has spoken directly to the “precise question at issue,” then we must give effect to Congress’s “unambiguously expressed intent.” Chevron U.S.A., Inc. v. NRDC, 467 U.S. 837, 842-43, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984). However, “if the statute is silent or ambiguous with respect to the specific issue, the question for the court is whether the agency’s answer is based on a permissible construction of the statute.” Id. at 843, 104 S.Ct. 2778. We “must reject administrative constructions which are contrary to clear congressional intent.” Id. at 843 n. 9, 104 S.Ct. 2778. NHTSA’s compliance with NEPA is reviewed under an arbitrary and capricious standard pursuant to the APA. See, e.g., Nat’l Parks & Conservation Ass’n, 241 F.3d at 730. With respect to NEPA documents, the agency must take a “hard look” at the impacts of its action by providing “ ‘a reasonably thorough discussion of the significant aspects of the probable environmental consequences.’ ” Thomas, 137 F.3d at 1149 (quoting Or. Nat. Res. Council v. Lowe, 109 F.3d 521, 526 (9th Cir.1997)). We must determine whether the EA “ ‘foster[s] both informed decision-making and informed public participation.’ ” Native Ecosystems Council v. U.S. Forest Serv., 418 F.3d 958, 960 (9th Cir.2005) (quoting California v. Block, 690 F.2d 753, 761 (9th Cir.1982)). III. DISCUSSION A. Energy Policy and Conservation Act Issues 1. NHTSA’s use of marginal cost-benefit analysis to determine “maximum feasible average fuel economy level” With respect to non-passenger automobiles (i.e., light trucks), the fuel economy standard “shall be the maximum feasible average fuel economy level that the Secretary decides the manufacturers can achieve in that model year.” 49 U.S.C. § 32902(a). “Maximum feasible” is not defined in the EPCA. However, the EPCA provides that “[w]hen deciding maximum feasible average fuel economy under this section, the Secretary of Transportation shall consider technological feasibility, economic practicability, the effect of other motor vehicle standards of the Government on fuel economy, and the need of the United States to conserve energy.” Id. § 32902(f). Petitioners argue that the meaning of “maximum feasible” is plain, and that NHTSA’s decision to maximize economic benefits is contrary to the plain language of the EPCA because “feasible” means “ ‘capable of being done,’ ” not economically optimal. But even if “feasible” means ‘“capable of being done,’” technological feasibility, economic practicability, the effect of other motor vehicle standards, and the need of the nation to conserve energy must be considered in determining the “maximum feasible” standard. American Textile Manufacturers Institute v. Donovan does not support Petitioners’ interpretation of “feasible.” 452 U.S. 490, 101 S.Ct. 2478, 69 L.Ed.2d 185 (1981). In that case, no other language in the statute modified the phrase at issue: “to the extent feasible.” Id., 452 U.S. at 508-11, 101 S.Ct. 2478. Here, “maximum feasible” standards are to be determined in light of technological feasibility, economic practicability, the effect of other motor vehicle standards, and the need of the nation to conserve energy. The EPCA clearly requires the agency to consider these four factors, but it gives NHTSA discretion to decide how to balance the statutory factors — as long as NHTSA’s balancing does not undermine the fundamental purpose of the EPCA: energy conservation. In Center for Auto Safety v. NHTSA the D.C. Circuit considered whether NHTSA gave “impermissible weight to shifts in consumer demand” in setting the MY 1985 and 1986 standards for light trucks. 793 F.2d 1322, 1338 (D.C.Cir.1986). Petitioners in that case challenged NHTSA’s rule that revised the standards downward. Id. at 1323-24. The court held that since Congress had not directly spoken to the issue of consumer demand, the court must determine whether the agency’s interpretation represented a “ ‘reasonable accommodation of conflicting policies that were committed to the agency’s care by the statute.’ ” Id. at 1338 (quoting Chevron, 467 U.S. at 845, 104 S.Ct. 2778). The court reasoned that: Congress intended energy conservation to be a long term effort that would continue through temporary improvements in energy availability. Thus, it would clearly be impermissible for NHTSA to rely on consumer demand to such an extent that it ignored the overarching goal of fuel conservation. At the other extreme, a standard with harsh economic consequences for the auto industry also would represent an unreasonable balancing of EPCA’s policies. Id. at 1340 (footnote omitted). The court concluded that NHTSA’s consideration of consumer demand was permissible because Congress did not speak to the precise issue, and “it specifically delegated the process of setting light truck fuel economy standards with broad guidelines concerning the factors that the agency must consider. NHTSA has remained within the reasonable range permitted by those factors.” Id. at 1341; see also Pub. Citizen v. NHTSA, 848 F.2d 256, 265 (D.C.Cir.1988) (R. Ginsburg, J.). In Public Citizen, the petitioners challenged the NHTSA’s lowering of the fuel economy standard for passenger cars for MY 1986. 848 F.2d at 259. They argued that NHTSA’s determination that the statutory 27.5 mpg standard was not economically practicable improperly elevated consumer demand and market forces, subordinated the statute’s technology-forcing design, and ignored the need of the nation to conserve energy. Id. at 264. The court held that NHTSA’s “consideration of the likelihood of economic hardship within its assessment of ‘economic practicability[ ]’ must be accorded due weight.” Id. at 264-65. Based on economic analyses supplied by other governmental agencies, “NHTSA concluded that the industry-wide economic effects of the higher CAFE standard would be severe,” id. at 265, “including sales losses well into the hundreds of thousands, and job losses well into the tens of thousands,” id. at 264; see also 49 Fed.Reg. 41,250, 41,252 (Oct. 22, 1984). Petitioners cite Public Citizen for the proposition that consideration of “economic practicability” allows lowering fuel economy standards only if a higher standard would cause substantial economic hardship to a manufacturer with a substantial share of the market. But that is not precisely what Public Citizen held. Rather, that court concluded that given the extensive evidence in the record showing that severe economic hardship would result from a higher standard, NHTSA’s decision to lower the standard under those circumstances was not devoid of rational support. Pub. Citizen, 848 F.2d at 265. The Public Citizen court held that NHTSA’s balancing of the statutory factors in 49 U.S.C. § 32902(f) was reasonable given that the possible energy savings from the higher standard did not outweigh the severe economic costs, id. at 265, since “the maximum potential increase in annual fuel consumption attributable to th[e] rule would amount to less than 0.1 percent of current consumption,” id. at 268. The court observed, “NHTSA found the maximum yearly impact of the lower (26.0 mpg) standard on U.S. gasoline consumption to be 210 million gallons, 0.3% of annual U.S. gasoline consumption and 0.09% of annual U.S. petroleum consumption. That savings, NHTSA stated, was not commensurate with ‘potential sales losses to the industry in the hundreds of thousands, job losses in the tens of thousands, or the unreasonable restriction of consumer choices.’ ” Pub. Citizen, 848 F.2d at 260-61 (citation omitted). In sum, Congress did not “offer[] a more precise balancing formula for the agency to apply to the four ... factors,” and “[i]n the absence of a sharper congressional delineation,” the court could not conclude that, under the circumstances presented there, NHTSA’s decision was not a reasonable accommodation of conflicting statutory policies. Id. at 265. In this rulemaking, NHTSA does not set forth its interpretation of the four factors in 49 U.S.C. § 32902(f). It simply states that in determining the “maximum feasible” fuel economy level, NHTSA “assesses what is technologically feasible for manufacturers to achieve without leading to adverse economic consequences, such as a significant loss of jobs or the unreasonable elimination of consumer choice.” 70 Fed.Reg. at 51,425; 71 Fed.Reg. at 17,585 (citing Pub. Citizen, 848 F.2d at 264). NHTSA “balance[s]” the four factors in § 32902(f), “along with other factors such as safety,” in determining the CAFE standards. 71 Fed.Reg. at 17,588, 17,655. In earlier rulemakings, NHTSA interpreted “technological feasibility” to mean “whether particular methods of improving fuel economy will be available for commercial application in the model year for which a standard is being established,” “economic practicability” to mean “whether the implementation of projected fuel economy improvements is within the economic capability of the industry,” “effect of other Federal motor vehicle standards on fuel economy” to mean “an analysis of the unavoidable adverse effects on fuel economy of compliance with emission, safety, noise, or damageability standards,” and “the need of the Nation to conserve energy” to mean “the consumer cost, national balance of payments, environmental, and foreign policy implications of our need for large quantities of petroleum, especially imported petroleum.” 42 Fed.Reg. 63,184, 63,188 (Dec. 15, 1977) (emphasis added); see also Ctr. for Auto Safety, 793 F.2d at 1325 n. 12. NHTSA “recognize[s] that [it] in the past has expressed its belief that the statutory consideration of economic practicability differs from, but does not preclude consideration of, cost/ benefit analysis.” 70 Fed.Reg. at 51,435. In its final rule establishing passenger automobile CAFE standards for MYs 1981-1984, NHTSA stated, “not equating cost-benefit considerations with economic practicability is consistent with the goal of achieving maximum feasible fuel economy by allowing economically and technologically possible standards which will improve fuel economy but which an analysis, subject to many practical limitations, might indicate are not cost-beneficial.” See 42 Fed.Reg. 33,534, 33,536 (1977). The agency further opined, “A cost-benefit analysis would be useful in considering [economic practicability], but sole reliance on such an analysis would be contrary to the mandate of the Act.” Id. at 33,537. In this rulemaking, however, NHTSA states that “the cost/benefit anal-yses conducted today ... are substantially more robust than those conducted in decades past and provide a more substantial basis for consideration of economic practicability.” 70 Fed.Reg. at 51,435. We agree with NHTSA that “EPCA neither requires nor prohibits the setting of standards at the level at which net benefits are maximized.” Id. at 51,435. The statute is silent on the precise question of whether a marginal cost-benefit analysis may be used. See Chevron, 467 U.S. at 843, 104 S.Ct. 2778. Public Citizen and Center for Auto Safety persuade us that NHTSA has discretion to balance the oft-conflicting factors in 49 U.S.C. § 32902(f) when determining “maximum feasible” CAFE standards under 49 U.S.C. § 32902(a). To be clear, we reject only Petitioners’ contention that EPCA prohibits NHTSA’s use of marginal cost-benefit analysis to set CAFE standards. Whatever method it uses, NHTSA cannot set fuel economy standards that are contrary to Congress’s purpose in enacting the EPCA — energy conservation. We must still review whether NHTSA’s balancing of the statutory factors is arbitrary and capricious. Additionally, the persuasiveness of the analysis in Public Citizen and Center for Auto Safety is limited by the fact that they were decided two decades ago, when scientific knowledge of climate change and its causes were not as advanced as they are today. The need of the nation to conserve energy is even more pressing today than it was at the time of EPCA’s enactment. See, e.g., NRDC Cmt. at 4, 11 (“When fuel economy legislation was first enacted, America consumed 16.3 million barrels of oil per day and 35.8 percent of U.S. oil came from imports. In the nearly 30 years since then, oil consumption has risen to over 20 million barrels per day and 56 percent of U.S. oil is imported. If fuel economy standards are not strengthened, these trends are only expected to get worse, with transportation oil use driving 80 percent of U.S. oil demand growth through 2025 and imports rising to 68 percent of U.S. oil demand. The light duty vehicle fleet currently consumes 8.3 million barrels per day, and in the absence of stronger standards, that is projected to grow to 12.45 million barrels by 2025.”); NAS Report at 13-14, 20. What was a reasonable balancing of competing statutory priorities twenty years ago may not be a reasonable balancing of those priorities today. 2. Failure to monetize benefits of greenhouse gas emissions reduction Even if NHTSA may use a cost-benefit analysis to determine the “maximum feasible” fuel economy standard, it cannot put a thumb on the scale by undervaluing the benefits and overvaluing the costs of more stringent standards. NHTSA fails to include in its analysis the benefit of carbon emissions reduction in either quantitative or qualitative form. It did, however, include an analysis of the employment and sales impacts of more stringent standards on manufacturers. See 71 Fed.Reg. at 17,590-91. To determine the “maximum feasible” CAFE standards, NHTSA began with the fuel economy baselines for each of the seven largest manufacturers — that is, “the fuel economy levels that manufacturers were planning to achieve in those years.” Id. at 17,581. NHTSA then “add[ed] fuel saving technologies to each manufacturer’s fleet until the incremental cost of improving its fuel economy further just equalled] the incremental value of fuel savings and other benefits from doing so.” Id. at 17,-596. The standard is further adjusted “until industry-wide net benefits are maximized. Maximization occurs when the incremental change in industry-wide compliance costs from adjusting it further would be exactly offset by the resulting incremental change in benefits.” Id. NHTSA claims that this “cost-benefit analysis carefully considers and weighs all of the benefits of improved fuel savings,” and that “there is no compelling evidence that the unmonetized benefits would alter our assessment of the level of the standard for MY2011.” Mat 17,592. Under this methodology, the values that NHTSA assigns to benefits are critical. Yet, NHTSA assigned no value to the most significant benefit of more stringent CAFE standards: reduction in carbon emissions. Petitioners strongly urged NHTSA to include this value in its analysis, and they cited peer-reviewed scientific literature in support. NRDC cited figures for the benefit of carbon emissions reduction ranging from $8 to $26.50 per ton C02, based on values assigned by the California Public Utilities Commission, the Idaho Power Company, and the European Union (EU) carbon trading program. NRDC Cmt. at 8. NRDC also cited a study published by the National Commission on Energy Policy, which “found that measures mitigating climate change emissions have estimated benefits of $3-19 per ton of carbon dioxide equivalent. The Commission recommends a price of $7 per ton beginning in 2010 and then rising 5 percent each year.” Id. at 23 (footnote omitted). Environmental Defense and the Union of Concerned Scientists recommended a minimum value of $50 per ton carbon (or $13.60 per ton C02), which reflects a mean marginal damage cost developed in 28 peer-reviewed studies. Environmental Defense Cmt. at 6, A-4; UCS Cmt. at 16. Valuing carbon emissions at $50 per ton carbon translates into approximately $0.15 per gallon of gasoline saved. UCS Cmt. at 16. The NAS committee, on which NHTSA relies for other aspects of its analysis, also valued the benefit of carbon emissions reduction at $50 per ton carbon. NAS Report at 85. NHTSA acknowledged that “Mon-serving energy, especially reducing the nation’s dependence on petroleum, benefits the U.S. in several ways. [It] has benefits for economic growth and the environment, as well as other benefits, such as reducing pollution and improving security of energy supply.” 71 Fed.Reg. at 17,644. NHTSA also acknowledged the comments it received that recommended values for the benefit of carbon emissions reduction; however, the agency refused to place a value on this benefit. See id. at 17,638. NHTSA stated: The agency continues to view the value of reducing emissions of C02 and other greenhouse gases as too uncertain to support their explicit valuation and inclusion among the savings in environmental externalities from reducing gasoline production and use. There is extremely wide variation in published estimates of damage costs from greenhouse gas emissions, costs for controlling or avoiding their emissions, and costs of sequestering emissions that do occur, the three major sources for developing estimates of economic benefits from reducing emissions of greenhouse gases. Moreover, ... commenters did not reliably demonstrate that the un-monetized benefits, which include C02, and costs, taken together, would alter the agency’s assessment of the level of the standard for MY 2011. Thus, the agency determined the stringency of that standard on the basis of monetized net benefits. Id.; see also FRIA, at VIII-64 to 65. NHTSA’s reasoning is arbitrary and capricious for several reasons. First, while the record shows that there is a range of values, the value of carbon emissions reduction is certainly not zero. NHTSA conceded as much during oral argument when, in response to questioning, counsel for NHTSA admitted that the range of values begins at $3 per ton carbon. NHTSA insisted at argument that it placed no value on carbon emissions reduction rather than zero value. We fail to see the difference. The value of carbon emissions reduction is nowhere accounted for in the agency’s analysis, whether quantitatively or qualitatively. This position also contradicts NHTSA’s own explanation in the Final Rule that “the agency determined the stringency of [the MY 2011] standard on the basis of monetized net benefits.” 71 Fed.Reg. at 17,638 (emphasis added). By presenting a scientifically-supported range of values that does not begin at zero, Petitioners have shown that it is possible to monetize the benefit of carbon emissions reduction. Second, NHTSA gave no reasons why it believed the range of values presented to it was “extremely wide”; in fact, several commenters and the NAS committee recommended the same value: $50 per ton carbon. The NAS committee selected the value of $50 per ton carbon although it acknowledged the wide range of values in the literature and the potential controversy in selecting a particular value. NAS Report at 85. NHTSA argues that the problem was not simply “the ultimate value to be assigned, but the wide variation in published estimates of the three major underlying costs of carbon dioxide emissions — the cost of damages caused by such emissions, the costs of avoiding or controlling such emissions, and the costs of sequestering resulting emissions.” NHTSA Br. at 49. But NHTSA fails to explain why those three “underlying costs” are relevant to the question of how carbon emissions should be valued. We are convinced by Petitioners’ response: To monetize the benefits of reducing C02 emissions from automobiles, NHTSA did not need to calculate the “costs of sequestering emissions.” Carbon capture and sequestration, though a feasible means of reducing emissions from large stationary sources such as coal-fired power plants, was not within the range of actions at issue in this automobile fuel economy rulemaking. Nor were “costs for controlling or avoiding [C02] emissions” a genuine methodological barrier here: NHTSA already performed an elaborate analysis of the costs of mandating increases in fuel economy. For purposes of this rule-making, that was the relevant category of control costs. EPCA Reply Br. at 10-11. In sum, there is no evidence to support NHTSA’s conclusion that the appropriate course was not to monetize or quantify the value of carbon emissions reduction at all. Citizens for Clean Air v. EPA, 959 F.2d 839 (9th Cir.1992), which NHTSA cites to support its contention that agencies may decline to adopt a “particular monetary value” when the “costs and benefits are too uncertain,” NHTSA Br. at 48, is inappo-site. In Citizens for Clean Air, petitioners filed for administrative review of a state agency’s grant of a permit for construction of a solid waste incinerator. Citizens for Clean Air, 959 F.2d at 841. EPA denied the petitions, and this court held that the decision of the EPA not to consider recycling as a possible “best available control technology” under the Clean Air Act was not arbitrary or capricious. Id. at 841-42. The EPA noted in its proposed rule that “it was ‘unable to reliably quantify the emission reductions attributable to materials separation when a[ ] [waste incinerator] is equipped with highly efficient at-the-stack air pollution control devices.’ ” Id. at 844 (citation omitted). Petitioners submitted “no hard evidence” that recycling would reduce air pollution when the waste incinerators are already equipped with “state-of-the-art pollution control equipment” (e.g., scrubbers). Id. at 847-48. In addition, the Clean Air Act required “that the proposed technology [i.e., recycling] be the best available control technology, and in the absence of anything specific or quantifiable in support ... we conclude that EPA’s decision not to consider recycling in permitting Spokane’s incinerator was not arbitrary or capricious.” Id. at 848. The petitioners in Citizens for Clean Air had to satisfy such a high statutory threshold (“best available control technology”), and they could not satisfy that threshold without hard evidence. By contrast, Petitioners here provided substantial evidence of the value of carbon emissions reduction, and they do not have to satisfy a high statutory threshold. Third, NHTSA’s reasoning is arbitrary and capricious because it has monetized other uncertain benefits, such as the reduction of criteria pollutants, crash, noise, and congestion costs, see FRIA at VIII-73 to 80, and “the value of increased energy security,” 71 Fed.Reg. at 17,592. Dr. Michael Wang of the Center for Transportation Research at Argonne National Laboratory stated in his peer review of the CAFE compliance and effect model used by NHTSA in its rule-making that the wide range of dollar values per ton of C02 “is not a good reason that CO[2] dollar values are not included .... The same can be said [of] dollar values for criteria pollutants. Yet, monetary values for criteria pollutant emissions are included in the model.” Wang Cmt. at 6. Fourth, NHTSA’s conclusion that com-menters did not “reliably demonstrate” that monetizing the value of carbon reduction would have affected the stringency of the CAFE standard “ ‘runs counter to the evidence’ ” before it. NRDC v. U.S. Forest Serv., 421 F.3d 797, 806 (9th Cir.2005) (citation omitted). The Union of Concerned Scientists concluded that “including [a $50/tC value] in the determination of cost-efficient fuel economy could increase the 2011 targets by an average of 0.4-1.1 mpg.” UCS Cmt. at 16. Given that the CAFE standards set by NHTSA increase only 1.5 mpg from MY 2008 to 2011, an additional 0.4 to 1.1 mpg increase b