Agency Use of Indirect Benefits to Justify Regulation
Executive agencies have long used indirect benefits—meaning benefits beyond the express purpose of a regulation—to justify their rulemakings. However, the statutes that provide agencies with regulatory authority rarely explicitly direct agencies to consider indirect benefits. Lower courts disagree over whether consideration of indirect benefits is permissible, and the Supreme Court has reserved the question for a future case. Courts and existing scholarship have largely asked whether particular statutory provisions authorize consideration of indirect benefits. This Note contends that, even without such statutory authorization, indirect benefits are presumptively permissible because they further three traditional administrative law values: rational decisionmaking, transparency, and accountability. It then argues that, even if statutory authorization was required, the Clean Air Act (CAA)—the statute at the center of the indirect benefits controversy—provides the Environmental Protection Agency (EPA) with unmistakable authority to consider indirect benefits. Finally, this Note recommends two lines of judicial review to prevent agency abuses of indirect benefits.
Introduction
Because presidents want regulations that benefit the public to the greatest degree possible, they direct agencies to consider all effects that a regulation will have. These effects include indirect benefits, also called co-benefits, that produce beneficial effects beyond the express purpose of a regulation.1 Off. of Mgmt. & Budget, Exec. Off. of the President, Circular No. A-4, at 39–40 (2003).
For example, a Clean Water Act regulation that targets water pollution from coal power plants might require those plants to limit their operations, which would indirectly benefit the climate more broadly. At the same time, Congress gives agencies authority to solve only particular problems. Administrative law, constrained by the separation of powers, prohibits agencies from considering regulatory factors irrelevant to the problems that Congress provided.2See Lopez v. Davis, 531 U.S. 230, 242 (2001); Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 (1983).
It is controversial whether indirect benefits are irrelevant. Although Congress does not explicitly authorize their use, Congress does want agencies to make reasonable and informed decisions. Thus, agencies are arguably caught between executive and legislative approaches.
Agencies have generally followed presidential direction to include co-benefits in their cost-benefit analyses and justifications for regulations.3See infra notes 28–35 and accompanying text.
This practice was uncontroversial until 2015, when the Supreme Court expressed skepticism about the practice but declined to address its legality.4Michigan v. EPA, 576 U.S. 743, 760 (2015).
Since then, the D.C. Circuit has embraced agency consideration of indirect benefits in certain circumstances,5U.S. Sugar Corp. v. EPA, 830 F.3d 579, 625 (D.C. Cir. 2016) (per curiam).
but one federal district court has struck down a regulation because the agency relied primarily on co-benefits to justify the regulation.6Wyoming v. U.S. Dep’t of the Interior, 493 F. Supp. 3d 1046, 1074 (D. Wyo. 2020).
And these judicial developments have come with a shift in executive philosophies as well: The Trump Administration has attempted to limit agency reliance on co-benefits,7See Increasing Consistency and Transparency in Considering Benefits and Costs in the Clean Air Act Rulemaking Process, 85 Fed. Reg. 84130, 84156 (Dec. 23, 2020) (to be codified at 40 C.F.R. pt. 83).
and the Biden Administration—despite protestations otherwise8See Lisa Friedman, Biden Administration to Repeal Trump Rule Aimed at Curbing E.P.A.’s Power, N.Y. Times (May 14, 2021), https://www.nytimes.com/2021/05/13/climate/EPA-cost-benefit-pollution.html [perma.cc/25LS-BQLG].
—has quietly done the same in some rulemakings.9 Off. of Pollution Prevention & Toxics, EPA, Economic Analysis of the TSCA Section 6 Proposed Rule for Asbestos Risk Management, Part 1, at ES-12 (2022), https://downloads.regulations.gov/EPA-HQ-OPPT-2021-0057-0008/content.pdf [perma.cc/387H-XPER] (quantifying climate co-benefits of the hazardous substance rule but declining to include co-benefits in net benefits calculations).
Given the current legal climate of increased suspicion toward agency action, the issue of indirect benefits seems sure to come to a head in the near future.
This Note shows that co-benefits further three traditional administrative law values: transparency, accountability, and rational decisionmaking.10Infra notes 67–81 and accompanying text.
Because Congress structured the system of administrative law to pursue these values, this Note argues that consideration of indirect benefits is presumptively permissible, even without statutory authorization.11Infra notes 82–84 and accompanying text.
Still, because some courts may apply a stricter standard to the use of co-benefits, this Note also finds implicit authorization to consider co-benefits under the Clean Air Act (CAA),12Infra Section II.B.
by far the most important and controversial statute in the struggle over co-benefits. The CAA is one of the most effective laws ever enacted, driving trillions—with a “t”—of dollars in health benefits to Americans every year by reducing air pollution.13 Off. of Air & Radiation, EPA, The Benefits and Costs of the Clean Air Act from 1990 to 2020, at 5-24 (2011), https://www.epa.gov/sites/default/files/2015-07/documents/fullreport_rev_a.pdf [perma.cc/NUE2-G9GC].
But the CAA gives rise to a significant portion of these benefits indirectly.14 Anne E. Smith, NERA Econ. Consulting, An Evaluation of the PM2.5 Health Benefits Estimates in Regulatory Impact Analyses for Recent Air Regulations 8 fig.1 (2011), https://www.nera.com/content/dam/nera/publications/archive2/PUB_RIA_Critique_Final_Report_1211.pdf [perma.cc/Y3KY-RL2B] (reviewing cost-benefit analyses to find that co-benefits accounted for the majority of benefits in most CAA rules aimed at nonparticulate matter pollution).
In some rules, more than 95 percent of quantified benefits are indirect.15 Innovative Strategies & Econ. Grp., EPA, EPA-452/R-04-002, Regulatory Impact Analysis for the Industrial Boilers and Process Heaters NESHAP 10-1 (2004), https://www.epa.gov/sites/default/files/2020-07/documents/ici-boilers_ria_final-neshap_2004-02.pdf [perma.cc/7Y42-5XGG]; Risk & Benefits Grp., EPA, MD-C439-02, Regulatory Impact Analysis: Standards of Performance for New Stationary Sources and Emission Guidelines for Existing Sources: Commercial and Industrial Solid Waste Incineration Units 5-9 (2011), https://www.epa.gov/sites/default/files/2020-07/documents/ciswi_ria_final-nsps-neshap_2011-02.pdf [perma.cc/7A9T-29SK].
The CAA includes at least a dozen programs, each aimed at different types of air pollutants and each with different statutory thresholds governing regulation.16 Richard K. Lattanzio, Cong. Rsch. Serv., RL30853, Clean Air Act: A Summary of the Act and Its Major Requirements 3-17 (2022), https://crsreports.congress.gov/product/pdf/RL/RL30853 [perma.cc/HR26-JDGK].
But, for example, because pollutants are often emitted together, regulation of a hazardous air pollutant like mercury (which is not a greenhouse gas) could create an indirect benefit: reducing co-emitted air pollutants that are greenhouse gases. Critically, as climate change worsens, co-benefits under the CAA account for billions of dollars of benefits to the climate.17E.g., Health & Env’t Impacts Div., EPA, EPA-452/R-11-011, Regulatory Impact Analysis for the Final Mercury and Air Toxics Standards 5-91 tbl.5-17 (2011), https://www.epa.gov/sites/default/files/2020-07/documents/utilities_ria_final-mats_2011-12.pdf [perma.cc/JM2G-8NAQ] (estimating that the hazardous air pollutant rule drives 0 million in climate co-benefits under a 3 percent discount rate); Off. of Air & Radiation, EPA, EPA-HQ-OAR-2009-0491, Regulatory Impact Analysis for the Federal Implementation Plans to Reduce Interstate Transport of Fine Particulate Matter and Ozone in 27 States; Correction of SIP Approvals for 22 States 181 tbl.5-16 (2011), https://www.epa.gov/sites/default/files/2020-07/documents/transport_ria_final-csapr_2011-06.pdf [perma.cc/T2TE-DAMY] (estimating that the particulate matter and ozone rule drives 0 million in climate co-benefits under a 3 percent discount rate).
These significant indirect impacts are likely why the majority of indirect benefits case law has centered on the CAA.18See U.S. Sugar Corp. v. EPA, 830 F.3d 579, 625 (D.C. Cir. 2016) (per curiam); Michigan v. EPA, 576 U.S. 743, 760 (2015); Sanne H. Knudsen, The Flip Side of Michigan v. EPA: Are Cumulative Impacts Centrally Relevant?, 2018 Utah L. Rev. 1, 42 (2018) (noting that the EPA frequently invokes co-benefits in justifying CAA regulations).
Part I of this Note lays out the history of co-benefits and the legal controversy surrounding them. Part II contends that consideration of co-benefits improves agency transparency, accountability, and rational decisionmaking. Because Congress designed the administrative apparatus to pursue these values, specific statutory authorization to consider co-benefits should not be required. However, even if a court were to require authorization, the CAA would pass with flying colors. Congress designed it to be implemented as a whole to protect the nation’s air resources. Co-benefits, then, are a critical tool for carrying out the CAA’s design. Finally, Part III recommends two standards of judicial review of co-benefits to prevent agency abuses. It argues courts must prohibit consideration of irrelevant co-benefits but defines “irrelevance” narrowly to include only co-benefits that contradict statutory text and purpose. It also suggests that co-benefits could, theoretically, indicate a pretextually promulgated regulation but recommends courts defer to agencies in most cases.
I. The History and Controversy of Co-Benefits
For decades, the executive branch has instructed agencies that “[t]he same standards of information and analysis quality that apply to direct benefits and costs should be applied to ancillary benefits and countervailing risks.”19 Off. of Mgmt. & Budget, supra note 1, at 26.
However, Congress rarely explicitly directs agencies to consider these “ancillary” benefits. Although courts have gestured at this tension between legislative and executive intent, they have not resolved it.20See Michigan, 576 U.S. at 760 (reserving the question of whether co-benefits are a permissible agency consideration).
Every president since Reagan has required that agencies assess costs and benefits to justify their major rules.21Exec. Order No. 12,291, 3 C.F.R. 127 (1982); Michael A. Livermore, Polluting the EPA’s Long Tradition of Economic Analysis, 70 Case W. Rsrv. L. Rev. 1063, 1067–68 (2020).
President Clinton established the modern cost-benefit infrastructure; his administration told agencies to assess “all costs and benefits of available regulatory alternatives,” thus including co-benefits.22Exec. Order No. 12,866, 3 C.F.R. 638, 638–39 (1994) (emphasis added).
Presidents Bush, Obama, Trump, and Biden all reaffirmed Clinton’s general approach and have required agencies to show the Office of Information and Regulatory Affairs (OIRA) that the benefits of their rules exceed their costs.23Exec. Order No. 13,422, 3 C.F.R. 191 (2008) (Bush); Exec. Order 13,563, 3 C.F.R. 215 (2012) (Obama); Exec. Order No. 13,771, 3 C.F.R. 284 (2018) (Trump); Modernizing Regulatory Review, 86 Fed. Reg. 7223 Jan. 20, 2021 (Biden).
However, not all presidents since Clinton have been consistent on whether this OIRA review includes co-benefits.24See infra notes 61–65 and accompanying text.
Under President Bush, the Office of Management and Budget issued its influential Circular A-4, which, since 2003, has directed agencies to “look beyond the direct benefits and direct costs . . . and consider any important ancillary benefits and countervailing risks.”25 Off. of Mgmt. & Budget, supra note 1, at 26.
This approach was designed to help each agency determine whether its action would cause more good than harm and to select the most effective alternative.26Id. at 2.
This directive also allowed the public and the courts to be more aware of the regulation’s effects. The Bush Administration released the Circular partially in response to concerns from economists that cost-benefit analyses (CBAs) focused too narrowly on direct costs and benefits, failing to assess the totality of a regulation’s effects.27Kimberly M. Castle & Richard L. Revesz, Environmental Standards, Thresholds, and the Next Battleground of Climate Change Regulations, 103 Minn. L. Rev. 1349, 1425–27 (2019) (arguing that Circular A-4 was issued in response to Risk vs. Risk: Tradeoffs in Protecting Health and the Environment (John D. Graham & Jonathan Baert Wiener eds., 1995)).
Even before President Bush formalized the practice, the EPA had consistently considered co-benefits under the CAA.28Livermore, supra note 21, at 1073–74.
For example, President Reagan’s EPA considered co-benefit reductions of criteria pollutants in a rule regulating hazardous air pollutants29Assessment of Municipal Waste Combustor Emissions Under the Clean Air Act, 52 Fed. Reg. 25399, 25406 (proposed July 7, 1987) (to be codified at 40 C.F.R. pt. 60).
and co-benefit reductions of ozone in a regulation targeting lead emissions from gasoline.30 Joel Schwartz et al., EPA, EPA-230-05-85-006, Costs and Benefits of Reducing Lead in Gasoline: Final Regulatory Impact Analysis, at VI-9 (1985), https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=9100YK16.TXT [perma.cc/NFF5-D2A6].
President George H.W. Bush’s EPA similarly justified emissions standards for landfill gases in part based on “the ancillary benefit of reducing global loadings of methane.”31Standards of Performance for New Stationary Sources and Guidelines for Control of Existing Sources: Municipal Solid Waste Landfills, 56 Fed. Reg. 24468, 24469 (proposed May 30, 1991) (to be codified at 40 C.F.R. pts. 51, 52, 60).
And President George W. Bush’s EPA supported an interstate air pollution rule by pointing to a co-benefit reduction in mercury emissions.32Rule To Reduce Interstate Transport of Fine Particulate Matter and Ozone (Clean Air Interstate Rule), 70 Fed. Reg. 25162, 25170 (May 12, 2005) (to be codified at 40 C.F.R. pts. 51, 72, 73, 74, 77, 78, 96).
Though this historical record under the CAA runs deep, it appears to be limited to co-benefits related to air quality.33I have not reviewed a CAA CBA that relies on non-air co-benefits except for CBAs that consider indirect effects on energy production, as required by the CAA. E.g., Off. of Air Quality Plan. & Standards, EPA, EPA-452/R-22-002, Regulatory Impact Analysis for the Proposed National Emission Standards for Hazardous Air Pollutants: Gasoline Distribution Technology Review and Standards of Performance for Bulk Gasoline Terminals Review 1-8 (2022), https://www.epa.gov/system/files/documents/2022-06/gasoline_distribution_ria_proposal_2022-06.pdf [perma.cc/D4MW-T7BP].
Outside the CAA context, other agencies have also occasionally considered indirect effects. Even President Trump’s Administration considered the “indirect benefit of . . . reduced distortions in the nationwide labor market” in a Supplemental Nutrition Assistance Program rulemaking.34 U.S. Dep’t of Agric., Supplemental Nutrition Assistance Program: Requirements for Able-Bodied Adults Without Dependents 7 n.3 (2019), https://www.regulations.gov/document/FNS-2018-0004-19016 [perma.cc/K2EC-FMCS].
This indirect benefit was used to justify a rulemaking that made it more difficult for states to waive work requirements under the program.35Id. at 2–3.
Despite this history, even a longstanding executive practice like consideration of co-benefits is subject to judicial review. Judicial review of agency action is governed by section 706 of the Administrative Procedure Act, which directs courts to set aside regulations that are “arbitrary [and] capricious.”365 U.S.C. § 706(2)(A).
The Supreme Court has interpreted this standard as being deferential to agencies. Though agencies are required to provide “a concise general statement of [a rule’s] basis and purpose,”375 U.S.C. § 553(c).
courts should set aside the rule only if:
[T]he 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.38Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 42–43 (1983).
Because unelected executive branch officials promulgate these regulations, the reasoned-explanation requirement “ensure[s] that agencies offer genuine justifications for important decisions, reasons that can be scrutinized by courts and the interested public.”39Dep’t of Com. v. New York, 139 S. Ct. 2551, 2575–76 (2019). Because courts require agencies to provide their true reasons as justifications for a rule, id., this Note uses agency consideration and justification interchangeably.
Co-benefits themselves are subject to judicial review because agencies often rely on them to explain why they promulgated a regulation.40Nat’l Ass’n of Home Builders v. EPA, 682 F.3d 1032, 1039–40 (D.C. Cir. 2012) (“[W]hen an agency decides to rely on a cost-benefit analysis as part of its rulemaking, a serious flaw undermining that analysis can render the rule unreasonable.”); Wyoming v. U.S. Dep’t of the Interior, 493 F. Supp. 3d 1046, 1068 (D. Wyo. 2020) (reviewing the co-benefits an agency relied on in promulgating a rule); U.S. Sugar Corp. v. EPA, 830 F.3d 579, 625 (D.C. Cir. 2016) (per curiam) (same).
Courts, up to this point, have not held that an agency must show that the benefits of a rule exceed its costs (though the agency may need to demonstrate that for OIRA review).41See supra note 23.
However, courts must ensure that an agency does not rely on factors irrelevant to their authorizing statute, potentially including co-benefits.42See State Farm, 463 U.S. at 42–43.
Additionally, a court can hold that co-benefits indicate the agency was pretextually attempting to control something other than the object of its regulation.
The only federal court to strike down a regulation for reliance on co-benefits apparently did so both because they were irrelevant and indicated pretextual reasoning.43Wyoming, 493 F. Supp. 3d 1046.
In the leadup to that case, Wyoming v. U.S. Department of the Interior, the U.S. Bureau of Land Management (BLM) regulated gas leaks from existing gas wells pursuant to statutory authority to promulgate rules “for the prevention of undue waste.”44Id. at 1063, 1067 (emphasis omitted).
The EPA had regulated greenhouse gases from new wells under the CAA’s New Source Performance Standards program but was slow to do so for existing wells.45Id. at 1066, 1068–69.
The Bureau stated that it promulgated its regulation in part because of “the length of this [CAA] process and the uncertainty regarding the final outcomes.”46Id. at 1066 n.18 (quoting Waste Prevention, Production Subject to Royalties, and Resource Conservation, 81 Fed. Reg. 83008, 83019 (Nov. 18, 2016) (to be codified at 43 C.F.R. pts. 3100, 3160, 3170)).
The court, however, believed the BLM promulgated its rule to circumvent the EPA’s process and, thus, “bootstrap[ped] itself into an area in which it has no jurisdiction.”47Id. at 1067 (quoting Adams Fruit Co. v. Barrett, 494 U.S. 638, 650 (1990)). The court acknowledged that Michigan v. EPA leaves the question of co-benefits unanswered but argued that Michigan v. EPA viewed reliance on co-benefits to push benefits above costs “with skepticism.” Id. at 1080. Another recent district court case, ruling on a later version of the BLM rule, rejected the argument that Michigan v. EPA foreclosed reliance on co-benefits. California v. Bernhardt, 472 F. Supp. 3d 573, 616 n.32 (N.D. Cal. 2020).
The Wyoming v. U.S. Department of the Interior court pointed to the agency’s CBA as evidence.48Wyoming, 493 F. Supp. 3d at 1069.
The benefits of the rule outweighed its costs only if the greenhouse gas co-benefits were factored in.49Id.
The court reiterated that the BLM could not bootstrap regulatory jurisdiction using co-benefits, finding that the “ancillary benefits of a rule cannot provide the primary justification for the rule, particularly where those ancillary benefits fall outside the scope of the agency’s statutory authority.”50Id. at 1074 (emphasis added).
The court worried that such use of co-benefits offered “no limit on agency authority and any colorable tie to an agency’s authority would permit the agency to act on a problem Congress never intended the agency to solve.”51Id.
Other courts have found co-benefits to be permissible, at least when they are consistent with statutory purpose. The D.C. Circuit held, in 2016, that the EPA was free to consider co-benefits from other hazardous air pollutants when crafting a hazardous air pollutant rule. The court reasoned that the CAA “does not foreclose the Agency from considering co-benefits and doing so is consistent with the CAA’s purpose—to reduce the health and environmental impacts of hazardous air pollutants.”52U.S. Sugar Corp. v. EPA, 830 F.3d 579, 625 (D.C. Cir. 2016).
The Supreme Court had a chance to resolve the controversy surrounding co-benefits in 2015 but declined to do so. Michigan v. EPA concerned the legality of the EPA’s 2012 Mercury and Toxic Substances (MATS) rule.53Michigan v. EPA, 576 U.S. 743 (2015).
In the CBA for the MATS rule, the EPA estimated that the regulation would create benefits of $37 to $90 billion per year and costs of only $9.6 billion.54Id. at 750.
However, over 99 percent of those quantified benefits were co-benefit reductions in particulate matter and sulfur dioxide air pollution.55Id.
Amici argued the EPA’s use of co-benefits was a “power grab” to regulate “outside the specific authority under which they are acting.”56Motion for Leave to File Amicus Curiae Brief and Brief of the Chamber of Commerce of the United States of America as Amicus Curiae in Support of Petitioners at 16–17, Michigan v. EPA, 576 U.S. 743 (2015) (Nos. 14-46, 14-47, 14-49) [hereinafter Chamber of Commerce Amicus Brief].
But the Court reserved the question for another day, holding that the EPA did not balance costs against benefits early enough in its regulatory process. The Court, therefore, declined to address if the EPA “could have considered ancillary benefits when deciding whether regulation is appropriate and necessary.”57Michigan, 576 U.S. at 760.
Although the full Court did not answer the question, at least two current justices have questioned the EPA’s approach. In the lower court, then-Judge Brett Kavanaugh described the EPA’s use of co-benefits to justify the MATS rule as “disputed.”58White Stallion Energy Ctr. v. EPA, 748 F.3d 1222, 1263 (D.C. Cir. 2014) (Kavanaugh, J., dissenting in part).
And at Supreme Court oral argument, Chief Justice Roberts expressed concern that the “disproportionate amount of benefit” might raise a “red flag” that the rule is an “illegitimate way of avoiding the . . . quite different limitations on EPA that apply in the criteria program [for particulate matter pollution].”59Transcript of Oral Argument at 62–64, Michigan, 576 U.S. 743 (Nos. 14-46, 14-47, 14-49).
In other words, Chief Justice Roberts suggested the co-benefits might indicate that the EPA was pretextually regulating particulate matter under a section of the CAA that provided lower barriers to regulation.60See infra notes 135–137 (describing the respective barriers to regulation in the criteria program and the Hazardous Air Pollutant program).
The EPA has recently reduced its reliance on co-benefits in CBAs. The Trump Administration was particularly skeptical of co-benefits, and, in 2020, the EPA released a rule requiring disaggregation and separate presentation of co-benefits and co-costs in CAA rulemakings.61Increasing Consistency and Transparency in Considering Benefits and Costs in the Clean Air Act Rulemaking Process, 85 Fed. Reg. 84130, 84151 (Dec. 23, 2020) (to be codified at 40 C.F.R. pt. 83).
The rule explains that “[t]he disaggregation of welfare effects will be important to ensure that the [CBA] may provide, to the maximum extent feasible, transparency in decisionmaking.”62Id.
The Trump Administration also attempted to rescind a 2016 BLM rule because it relied on climate co-benefits.63Waste Prevention, Production Subject to Royalties, and Resource Conservation; Rescission or Revision of Certain Requirements, 83 Fed. Reg. 49184 (Sept. 28, 2018) (to be codified at 43 C.F.R. pts. 3160, 3170).
The EPA revoked the Trump co-benefit rule in 2021 under the Biden Administration, expressing concerns that the rule had no basis in the CAA, unjustly called into question the significance of co-benefits, and promoted arbitrary decisionmaking.64Rescinding the Rule on Increasing Consistency and Transparency in Considering Benefits and Costs in the Clean Air Act Rulemaking Process, 86 Fed. Reg. 26406, 26413 (May 14, 2021) (to be codified at 40 C.F.R. pt. 83).
But even though President Biden’s EPA theoretically returned to the status quo, it may be quietly shying away from using co-benefits to justify regulations. For example, in a recent toxic substances rulemaking, the EPA quantified $10 to $99 million dollars a year in climate co-benefits but declined to include those benefits in its net CBA.65 Off. of Pollution Prevention & Toxics, EPA, Economic Analysis of the TSCA Section 6 Proposed Rule for Asbestos Risk Management, Part 1, at ES-12 (2022), https://www.regulations.gov/document/EPA-HQ-OPPT-2021-0057-0008 [perma.cc/387H-XPER].
This is exactly the approach the EPA criticized in its 2021 rule,66Rescinding the Rule on Increasing Consistency and Transparency in Considering Benefits and Costs in the Clean Air Act Rulemaking Process, 86 Fed. Reg. at 26413.
and it may show the Biden Administration’s concern with recent legal controversy surrounding co-benefits in Wyoming v. U.S. Department of the Interior and Michigan v. EPA. This Note argues that these concerns are misplaced because co-benefits are consistent with congressional design of the administrative apparatus.
II. Indirect Benefits Help Agencies Achieve Congressional Design
Co-benefits accomplish several regulatory goals. First, they allow agencies to rationally prioritize the most effective regulations and assess how these regulations will impact other sections of a statute. Furthermore, by allowing the public, political branches, and courts to understand the effects of and reasoning for a regulation, co-benefits serve both agency transparency and accountability. Because Congress wants agencies to pursue these values, statutory authorization is not required to consider co-benefits. And even if authorization were necessary, the CAA, at least, provides the EPA with unmistakable authorization.
A. Indirect Benefits Are Presumptively Permissible
Rational decisionmaking, transparency, and accountability are the foundational principles of administrative law.67See Benjamin Eidelson, Reasoned Explanation and Political Accountability in the Roberts Court, 130 Yale. L.J. 1748, 1751–52 (2021) (noting that judicial review of agency action is designed to ensure rational, transparent, and accountable decisions); Kristin E. Hickman & Richard J. Pierce, Jr., Administrative Law Treatise § 11.1, 11.5.2 (7th ed. 2023) (same, citing cases).
Consideration of co-benefits furthers these values. Co-benefits are presumptively permissible even without specific statutory authorization because Congress designed the administrative apparatus for agencies to make decisions in alignment with these values.
1. Rational Decisionmaking
Agencies do not have the resources to pursue every regulatory opportunity. Courts usually allow agencies to determine their own priorities because the “agency is far better equipped than the courts to deal with the many variables involved in the proper ordering of its priorities.”68Heckler v. Chaney, 470 U.S. 821, 831–32 (1985).
Co-benefits allow agencies to rationally prioritize the most effective regulations by considering the full effects of each option. Otherwise, an agency might prioritize inferior rules.
Co-benefits also allow agencies to quantify the interactions between different programs in the same statute, allowing agencies to rationally prioritize rules the way Congress designed. General principles of statutory construction support the conclusion that Congress wants agencies to implement statutes as a whole. Courts presume that statutes are cohesive texts,69 Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of Legal Texts 169 (2012); Law v. Siegel, 571 U.S. 415, 422 (2014) (applying the “normal rule of statutory construction that words repeated in different parts of the same statute generally have the same meaning” (internal quotation omitted)).
so each provision in a statute should be read in “harmony” with the others.70 Scalia & Garner, supra note 69, at 180.
Congress often adds statements of purpose that apply to the entire statute,71See, e.g., 42 U.S.C. § 7401 (CAA-wide statement of purpose); Safeco Ins. Co. of Am. v. Burr, 551 U.S. 47, 62 (2007) (citing Fair Credit Reporting Act statement of purpose in 15 U.S.C. § 1861(a) to interpret the statutory section).
implying that each section is intended to work towards that policy goal. A regulation may indirectly impact the purpose of the statute as a whole even if the effect is not regulated under that section. That effect would, therefore, still be relevant to Congress’s design of the whole statute. Co-benefits allow agencies to quantify those indirect effects.
Co-benefits can also allow agencies to promulgate rational regulations that accomplish statutory goals when they would otherwise falter under OIRA review. For instance, consider the EPA’s 2012 Oil Refineries New Source Performance Standard under the CAA.72 Off. of Air Quality Plan. & Standards, EPA, Regulatory Impact Analysis: Petroleum Refineries New Source Performance Standards Ja 1-5 (2012), https://www.epa.gov/sites/default/files/2020-07/documents/refineries_ria_final-nsps_2012-06.pdf [perma.cc/6ENS-RT9B]. To reach this number, I assume a 3 percent discount rate and average the upper and lower bounds to arrive at 0 million monetized benefits. Of those monetized benefits, 83 percent are from co-benefit reductions in SO2 and NOX as those are not pollutants under the New Source Performance Standards Program. Therefore, 83 percent of 0 million results in 9 million in co-benefits. The remaining 17 percent results in million in direct benefits. I assume a 3 percent discount rate and average upper and lower bounds.
The rule created $71 million in direct benefits, $349 million in particulate matter co-benefits, and $96 million in costs.73Id. at 1-2, 6-18 (noting that SO2 and NOX are not pollutants under the New Source Performance Standard program and are, therefore, co-benefits).
Without consideration of co-benefits, this rule would not have survived OIRA review because its costs exceed its direct benefits.74See supra text accompanying note 19.
Consideration of co-benefits, though, allowed the EPA to justify the rule with both its direct benefits and co-benefits, thus vindicating the rule and enhancing the quality of the nation’s air resources. Similarly, once the EPA decides to promulgate a regulation, co-benefits allow it to justify a regulatory option that is maximally protective of air quality even if that option would create additional costs.
2. Transparency
Next, an agency transparently relying on co-benefits allows the public and the political branches to understand both the effects of and the reasons for a regulation. Regulations will create indirect effects regardless of whether agencies are allowed to consider co-benefits. If co-benefits are prohibited, agencies may surreptitiously consider them without disclosing their reliance.75See infra Section III.B.
Prohibiting co-benefits would also deprive the public of understanding the full effects of a rule. The public might react differently to a costly rule if it learns the rule will indirectly protect it from carcinogenic particulate matter. Similarly, the political reaction to a regulation might differ if an agency presents one set of justifications versus another.76Eidelson, supra note 67, at 1759.
Congress might also want an agency to focus its attention on particular problem areas, as it did when it amended the CAA in 1990 to force the EPA to more aggressively limit hazardous air pollutants.77Clean Air Act Amendments of 1990, Pub. L. No. 101-549, 104 Stat. 2399 (codified as amended at 42 U.S.C. §§ 7401–7671).
Co-benefits can allow Congress to transparently see how well the agency is addressing these priorities.
3. Accountability
Transparency also gives rise to agency accountability. For example, voters can elect a new president if they disagree with the priorities of the administration, and the president can create political accountability through their removal power over agency administrators.78See Seila Law v. Consumer Fin. Prot. Bureau, 140 S. Ct. 2183 (2020) (holding that most agency heads must be removable at-will by the president to ensure political accountability).
The president can also direct OIRA to veto inappropriate agency uses of co-benefits through its review of CBAs. The legislative branch has considerable oversight authority as well. Congress can override a regulation via the Congressional Review Act (subject to a presidential veto),795 U.S.C. §§ 801–808.
amend an authorizing statute, or cut agency funding, among other options.
An agency’s open reliance on co-benefits also facilitates accountability through judicial review. For example, the Wyoming v. U.S. Department of the Interior court’s dual holding—that the BLM’s use of co-benefits relied on an irrelevant factor and that the co-benefits indicated a pretextual purpose—was possible only because the BLM included the co-benefits in its justification for the regulation.80See Wyoming v. U.S. Dep’t of the Interior, 493 F. Supp. 3d 1046, 1068–74 (D. Wyo. 2020).
Co-benefits, therefore, allow for more meaningful review of agency action by “offer[ing] genuine justifications for important decisions, [and] reasons that can be scrutinized by courts and the interested public.”81Dep’t of Com. v. New York, 139 S. Ct. 2551, 2575–76 (2019).
Because co-benefits serve these administrative values, courts should apply a presumption in favor of their consideration even if analysis of a particular statute does not reveal an implicit authorization. As the Supreme Court has recognized, Congress wants agencies to make informed, reasonable, and transparent decisions, for which courts and the political branches can hold agencies accountable.82See id.
This presumption runs so deep within agency structure and judicial review of agency action that Congress may not feel the need to say so explicitly in a particular statute.83See Jerry L. Mashaw & David L. Harfst, The Struggle For Auto Safety 215 (1990) (“This agency, any agency, should always read between the lines of its statute an implicit qualification of the form: ‘Don’t forget that this statute does not exhaust our vision of the good life or the good society. Remember that we have other goals and other purposes that will sometimes conflict with the goals and purposes of this statute. If we forgot to mention all those potential conflicting purposes in your instructions, take note of them anyway. For heaven’s sake, be reasonable.’ ”).
This presumption is presumably the same reason courts have allowed CBAs even when a statute does not call for them.84See, e.g., Mingo Logan Coal Co. v. EPA, 829 F.3d 710, 732–33 (D.C. Cir. 2016) (“[C]onsideration of costs is an essential component of reasoned decisionmaking under the Administrative Procedure Act.”); Entergy Corp. v. Riverkeeper, Inc., 556 U.S. 208, 223 (2009) (permitting a CBA under a best-technology standard); Michigan v. EPA, 576 U.S. 743, 752–53 (2015) (“Agencies have long treated cost as a centrally relevant factor when deciding whether to regulate.”).
It would subvert congressional design of the administrative apparatus to prevent agencies from using co-benefits to align their decisionmaking processes with these values.
B. The CAA Authorizes Indirect Benefits Under Any Standard of Review
Although this Note contends that agencies do not require statutory authorization to consider co-benefits, if a court was to require such authorization, the EPA would have a strong case that the CAA provides it. Consideration of co-benefits allows the EPA to meaningfully consider costs, as the CAA repeatedly requires, and it furthers the Act’s purpose: “to protect and enhance the quality of the Nation’s air resources.”8542 U.S.C. § 7401.
In fact, Congress wants the EPA to achieve this purpose by enacting the law as a whole, which makes sense; air regulations, by their nature, have substantial indirect effects on other air pollutants.86See supra notes 14–17.
Co-benefits are, therefore, a critical tool for the EPA to rationally implement the statute.
There are textual indications that Congress intends the EPA to consider co-benefits. Congress requires the EPA to consider costs in every major CAA program, save the National Ambient Air Quality Standards (NAAQS),8742 U.S.C. § 7411(a) (stating that standard of performance should be set “considering cost”); 42 U.S.C. §§ 7412(d)(2), (f)(2) (setting technology-based emissions standards “taking into consideration the cost” then, as a backstop, setting health-based emissions standards “taking into consideration costs”); Michigan, 576 U.S. 743 (2015) (requiring the EPA to consider cost to determine if regulation of power plants under the Hazardous Air Pollutant program is appropriate and necessary); 42 U.S.C. § 7479(3) (defining the term “best available control technology” as “taking into account . . . economic impacts and other costs”); 42 U.S.C. § 7521(a)(2) (stating that emissions standards should be set “giving appropriate consideration to the cost of compliance”). Furthermore, the EPA may consider costs of compliance in setting cross-state air pollution (or “transport”) rules. EPA v. EME Homer City Generation, 572 U.S. 489 (2014).
and courts consistently require consideration of costs to include co-costs.88E.g., Corrosion Proof Fittings v. EPA, 947 F.2d 1201, 1229–30 (5th Cir. 1991) (requiring the EPA to consider the potential negative safety impacts of non-asbestos car brakes in asbestos rulemaking); Competitive Enter. Inst. v. Nat’l Highway Traffic Safety Admin., 956 F.2d 321, 323–25 (D.C. Cir. 1992) (requiring an agency to consider the negative safety effects of smaller cars in a fuel efficiency rulemaking).
At the same time, the Supreme Court has implied that, to meaningfully consider costs, an agency must compare them against benefits.89Michigan, 576 U.S. at 752 (“One would not say that it is even rational, never mind ‘appropriate,’ to impose billions of dollars in economic costs in return for a few dollars in health or environmental benefits.”).
It would be anomalous for costs to include co-costs but for benefits to exclude co-benefits. This would put a thumb on the scale against regulation that is nowhere to be found in the statute.90Livermore, supra note 21, at 1077.
In addition to the antiregulatory anomaly created by considering co-costs but not co-benefits, it can be practically difficult to distinguish between the two because costs and benefits are merely useful labels attached to negative and positive effects of a regulation.91Christopher C. DeMuth & Douglas H. Ginsburg, Rationalism in Regulation, 108 Mich. L. Rev. 877, 887 (2010) (reviewing Richard L. Revesz & Michael A. Livermore, Retaking Rationality: How Cost-Benefit Analysis Can Better Protect the Environment and Our Health (2008)).
Consider the following hypothetical. Say that the EPA sets an emissions limit for a (non-greenhouse gas) hazardous air pollutant emitted by boat manufacturers and expects that, in response, boat manufacturers will add a particular pollution control technology to their plants. This control technology creates $30 million in co-benefits by capturing greenhouse gases from the smokestack, but it also creates $10 million in greenhouse gas co-costs by requiring additional electricity to power the controls. On net, the rule would therefore lead to $20 million worth of greenhouse gas reductions. Should the EPA ignore the $30 million in gross climate co-benefits and state in its CBA that the rule will harm the climate by $10 million a year? Or should the EPA ignore all climate impacts entirely because the net impact is a co-benefit and, therefore, impermissible to consider? Because costs and benefits are two sides of the same coin, there would not be a principled way for the agency to choose between these options. It would, therefore, be difficult to implement the CAA’s text directing the EPA to consider costs (and, therefore, co-costs) without considering co-benefits.
Beyond individual provisions, the CAA is structured to be enacted as a whole to achieve the goals of the statute, rather than in hermetically sealed sections as Chief Justice Roberts seemed to suggest.92Supra note 59 and accompanying text.
Congress enacted a CAA-wide section of findings and purpose rather than divvying up statements of findings and purpose section by section,9342 U.S.C. § 7401(a).
implying that it wants each section to build toward the same broader purposes. Congress also repeatedly instructed the EPA to consider the effects of its regulations under other sections of the CAA. To give a few examples: Congress required the EPA to study the effects of its other regulations before regulating hazardous air pollutants from power plants;9442 U.S.C. § 7412(n)(1)(A).
instructed the EPA to consider whether it was necessary to use its vehicle tailpipe standards to maintain the NAAQS;9542 U.S.C. § 7521(i)(2)(A).
and asked the EPA to maintain consistency between its source categories under the Hazardous Air Pollutant program and the New Source Performance Standards program.9642 U.S.C. § 7412(c)(1).
Congress has good reason to instruct the EPA to implement the CAA as a whole. The CAA is deeply interconnected along a number of axes. It is the rare CAA regulation that affects one pollutant alone. Many regulations reduce emissions of other pollutants even more than the regulated pollutant.97Supra notes 14–15.
Many facilities are regulated under multiple CAA programs,98E.g., N.C. Dep’t of Env’t Quality, Chemours Air Quality Permit No. 03735T48, at 3–5 (2020), https://www.deq.nc.gov/coastal-management/gis/data/airquailty-emissions-testing/final-permitno48-chemours-ves-sw-cas-1/download [perma.cc/AWQ5-AWSW] (regulating the source under both the Hazardous Air Pollutant and NAAQS programs).
so changing their obligations under one program may affect their ability to meet obligations under another. Similarly, several CAA programs regulate some pollutants, such as greenhouse gases.99 David R. Wooley & Elizabeth M. Morss, Clean Air Act Handbook § 10:10 (2023) (noting that greenhouse gases are regulated under the New Source Performance Standard program, Prevention of Significant Deterioration program, and Vehicle Tailpipe program).
Other pollutants, like volatile organic compounds, are regulated under one program when in one form but under another program when they transform into different pollutants once emitted.10042 U.S.C. § 7511b(b)(3) (regulating volatile organic compounds); Does EPA Regulate Volatile Organic Compounds (VOCs) in Household Products?, EPA, https://www.epa.gov/indoor-air-quality-iaq/does-epa-regulate-volatile-organic-compounds-vocs-household-products [perma.cc/L8FY-9T28] (last updated Mar. 5, 2024) (noting that volatile organic compounds photochemically react in the atmosphere to create ozone, which is regulated under the NAAQS program).
Thus, the EPA must keep a careful eye on how its regulations interact with one another to rationally implement the CAA.
Only co-benefits allow the EPA to quantify the effects a regulation has on other CAA pollutants. As the EPA has limited capacity to promulgate regulations, it would be useful to know that a greenhouse gas regulation will have substantial indirect effects on benzene (which is not a greenhouse gas), for example. The EPA might then prioritize regulating other hazardous air pollutants if it knows it has already indirectly, but effectively, addressed benzene.
The CAA’s legislative history is even more explicit that the EPA should consider the co-benefits of its regulations with respect to other CAA programs, at least in the context of hazardous air pollutant rulemakings. The Senate Committee to the 1990 CAA amendments expected that, when regulating hazardous air pollutants, the EPA would “consider the benefits which result from control of air pollutants that are not listed but the emissions of which are, nevertheless, reduced by control technologies or practices necessary to meet the prescribed limitation” because they “may produce substantial health and environment benefits.”101 S. Rep. No. 101-228, at 172 (1989).
By suggesting that the EPA should consider other types of pollutants in hazardous air pollutant rulemakings, the committee report goes beyond the holding of U.S. Sugar Corp., which upheld the EPA consideration of co-benefits from other hazardous air pollutants.102U.S. Sugar Corp. v. EPA, 830 F.3d 579, 584 (D.C. Cir. 2016).
Though this Part has thus far focused on only co-benefits that affect air quality, its argument also extends to some other types of co-benefits. The EPA rarely includes other co-benefits in its regulations,103E.g., Off. of Air Quality Plan. & Standards, supra note 33, at 1-8 (considering reduced energy prices in a CAA rulemaking and classifying this benefit as a negative cost, further illustrating the difficulty of distinguishing between costs and benefits).
but a future CAA regulation might indirectly reduce other environmental harms, such as water pollution. Congress created the EPA to be the primary agency tasked with environmental protection and pollution reduction.104See Reorganization Plan No. 3 of 1970, 84 Stat. 2086 (July 9, 1970).
It is, therefore, relevant to the EPA’s purpose—though not necessarily to the CAA’s purpose—to consider the indirect effects CAA regulations will have on water quality.
And, in the most extreme case, the EPA might consider a co-benefit that stems from an entirely nonenvironmental effect, like reduced consumer prices. This consideration should still be encouraged because it also promotes rational decisionmaking, transparency, and accountability. But, unlike environmental co-benefits, the EPA could point to little statutory basis to consider these co-benefits, except that Congress may not want the EPA to consider co-costs without considering co-benefits.
III. Judicial Review of Indirect Benefits
To prevent abuses, courts should review agency consideration of co-benefits in two ways. First, a court should ask if the co-benefits were an irrelevant factor for the agency to rely on when promulgating a regulation. Supreme Court case law on “irrelevance” suggests that most uses of co-benefits pass muster, and only reliance on co-benefits that conflict with statutory purposes should be considered irrelevant. Second, even if the co-benefits are relevant, a court could hold that co-benefits indicate the agency pretextually issued a regulation to evade political accountability or skirt statutory requirements. Although such pretextual reasoning would theoretically be unlawful, practical difficulties in determining agency intent militate in favor of generally granting deference to the agency. 105Co-benefits do not implicate the major questions doctrine. That doctrine is a recent development that strikes down agency actions if they involve major questions unless there is a clear statement of congressional intent to authorize the action. Daniel T. Deacon & Leah M. Litman, The New Major Questions Doctrine 109 Va. L. Rev. 1009, 1012 (2023). But the doctrine has been reserved for questions of an agency’s substantive authority, such as an agency’s authority to close down coal plants, West Virginia v. EPA, 142 S. Ct. 2587 (2022), or to enact an eviction moratorium during a pandemic, Ala. Ass’n of Realtors v. Dep’t of Health and Hum. Servs., 141 S. Ct. 2485 (2021). Nobody doubts the EPA’s authority to regulate air pollutants, though its method of doing so might raise questions, as in West Virginia v. EPA. Here, the question is the EPA’s internal method of analysis. When reviewing reliance on co-benefits, a court would ask whether the EPA was reasonable in promulgating a regulation rather than whether the regulation is substantively invalid. It would, therefore, be a significant expansion of the major questions doctrine to apply it in the co-benefit context.
A. Indirect Benefits as an Irrelevant Factor
It is a bedrock principle of administrative law that an agency may not promulgate a regulation by relying “on factors which Congress has not intended it to consider.”106Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 42–43 (1983). Petitioner in Michigan v. EPA seemed to argue this principle implied the EPA may never consider co-benefits under the Hazardous Air Pollutant program: “Section 7412(n)(1)(A) is about regulating something specific: hazardous air pollutants. So, the cost-benefit analysis must focus on the benefits of reducing those particular pollutants and the costs of the regulation that would create the reductions. The benefits of reducing non-hazardous air pollutants do not factor in.” Reply Brief for Petitioners State of Michigan, et al. at 21, Michigan v. EPA, 576 U.S. 743 (2015) (Nos. 14-46, 14-47, 14-49) (citation omitted).
Although Congress does not explicitly authorize the EPA to consider co-benefits, Congress also does not explicitly prohibit such consideration. The Supreme Court has generally held that silence is insufficient on its own to render a factor irrelevant under arbitrary and capricious review,107See Richard J. Pierce, Jr., What Factors Can an Agency Consider in Making a Decision?, 2009 Mich. St. L. Rev. 67, 72–73.
with one notable exception.108Massachusetts v. EPA, 549 U.S. 497, 532–33 (2007) (holding that the agency may consider only factors that “relate to” statutory text).
This Note argues that only co-benefits in conflict with the statute’s purpose are irrelevant. However, under stricter standards of review, agencies will be on firmer ground with co-benefits that are squarely related to the purpose of the statute or the agency. Finally, even when an agency considers irrelevant co-benefits, courts should reject the regulation only if they believe the agency would not have issued the regulation but for consideration of those co-benefits.
The Supreme Court has been reasonably consistent in determining that silence is insufficient to render a factor irrelevant.109See Pierce, supra note 107, at 72–73; Pension Benefit Guar. Co. v. LTV Corp., 496 U.S. 633, 648, 656 (1990) (stating that the statutory text “does not evince a clear congressional intent to deprive the [agency] of the ability to” consider additional factors and that the agency’s “failure to consider all potentially relevant areas of law did not render its restoration decision arbitrary and capricious”).
In Lopez v. Davis, the Court held that the agency need only show the extra-statutory factor it considered is “reasonable in light of the legislature’s revealed design.”110Lopez v. Davis, 531 U.S. 230, 242 (2001) (quoting NationsBank of N.C., N.A. v. Variable Annuity Life Ins. Co., 513 U.S. 251, 257 (1995)); see INS v. Yang, 519 U.S. 26 (1996) (holding that an agency could consider an immigrant’s reliance on fraud as means of entering the country when determining whether to waive his deportation, though the statute does not specify fraud as a factor for consideration).
In Entergy Corp. v. Riverkeeper, Inc., the Court gave the agency discretion to read cost-benefit considerations into a “best technology” standard. And in Whitman v. American Trucking Ass’ns, the Court interpreted silence to preclude consideration of a factor only when there was clear indication from the context of the statute that Congress intended that result.111Whitman v. Am. Trucking Ass’ns, 531 U.S. 457 (2001); Entergy Corp. v. Riverkeeper, Inc., 556 U.S. 208, 223 (2009) (describing Whitman as standing for “the rather unremarkable proposition that sometimes statutory silence, when viewed in context, is best interpreted as limiting agency discretion,” but declining to extend that holding). Whitman prohibited cost considerations under the CAA’s NAAQS program because cost considerations were deemed incompatible with the statutory requirement that emissions standards be set “requisite to protect the public health.” Whitman, 531 U.S. at 469. Following Whitman, the EPA still conducts a CBA when it sets a NAAQS rule but does not consider the results. Michael A. Livermore & Richard Revesz, Rethinking Health-Based Environmental Standards and Cost-Benefit Analysis, 46 Env’t. L. Inst. 10674, 10674 n.4, 10680 (2016). Therefore, unlike other CAA programs, the EPA does not consider co-benefits under the NAAQS, and the EPA could not to do so without future Supreme Court intervention. See Whitman, 531 U.S. at 457; Michael A. Livermore & Richard L. Revesz, Rethinking Health-Based Environmental Standards, 89 N.Y.U. L. Rev. 1184, 1264 (2014).
Because Congress rarely provides agencies with an exhaustive list of factors to consider, the Supreme Court should stick to this line of cases as a normative matter and as a matter of precedent. It also dovetails with this Note’s recommendation to consider co-benefits as presumptively permissible. Under this permissive standard, co-benefits would be considered irrelevant only when they contradict the statutory purposes Congress communicated to the agency. For example, if a statute directs the agency to consider only public health effects when setting an emissions standard, economic co-benefits to a regulated industry would be an irrelevant consideration because those co-benefits will generally be inversely correlated with the public health.112This would, in effect, be the practice Whitman foreclosed. Whitman, 531 U.S. 457. The Whitman court said the EPA could not consider costs to industry when setting the NAAQS. Id. at 486. Considering benefits to industry (via reduced costs) is the other side of the same coin.
However, more troublingly for agency consideration of co-benefits, the Supreme Court in Massachusetts v. EPA took a harsher stand on extra-statutory factors.113Massachusetts v. EPA, 549 U.S. 497 (2005).
In that case, the EPA refused a petitioner’s request that it determine whether greenhouse gases emitted from cars “may ‘reasonably be anticipated to endanger public health or welfare’ ” under section 202 of the CAA.114Id. at 514 (quoting 42 U.S.C. § 7521(a)(1)).
The EPA considered a number of factors in declining to make this “endangerment finding,” including the futility of combating climate change with a piecemeal approach and fears of weakening the president’s international negotiating position.115Id. at 513–14, 533.
The Court ruled that the EPA’s reasoning was arbitrary and capricious, in part, because the Court appeared to conflate the EPA’s denial of the petition with the endangerment finding itself. The Court held that the EPA’s consideration of these factors did not “relate to” the statutory question of whether greenhouse gasses emitted from cars endangered the public health.116Id. at 532–34.
Massachusetts v. EPA likely did not signal a doctrinal sea change away from the Court’s other irrelevant factor cases. It held that an agency may consider only factors “relate[d] to” the statutory text, which is admittedly a narrower standard than that in the other cases. But the Court did not cite any other cases interpreting the State Farm irrelevant-factor standard and did not seem to focus on the doctrinal implications of its decision.117Id.; Pierce, supra note 107, at 81–85 (arguing for a narrow reading of Massachusetts v. EPA).
The D.C. Circuit has apparently decided to read Massachusetts v. EPA narrowly and maintains that a factor is irrelevant only when there is clear congressional intent to preclude consideration.118See FTC v. Tarriff, 584 F.3d 1088 (D.C. Cir. 2009) (allowing an agency to consider a factor not included in a statutory list of nine factors); Catawba Cnty. v. EPA, 571 F.3d 20 (D.C. Cir. 2009) (refusing to interpret statutory silence as a bar on agency consideration of an extra-statutory factor).
And the Supreme Court itself has not yet construed Massachusetts v. EPA to have modified its precedents on the issue.119See Entergy Corp. v. Riverkeeper, Inc., 556 U.S. 208, 223 (2009) (not citing Massachusetts v. EPA).
Should the Court adopt the strict Massachusetts v. EPA standard in the future, co-benefits directly related to statutory text would still be considered relevant. In the CAA context, air co-benefits are relevant because they directly achieve the stated goal of the CAA to protect the nation’s air resources.12042 U.S.C. § 7401.
Furthermore, other environmental co-benefits would, at least, not be irrelevant under this standard. Although the EPA could not point to the CAA itself, it could rely on the EPA’s organic statute, which directs the EPA to protect the environment.121See Reorganization Plan No. 3 of 1970, 84 Stat. 2086 (July 9, 1970).
However, co-benefits beyond statutory text would be considered irrelevant. In the EPA’s case, other than the antiregulatory anomaly created by considering co-costs but not co-benefits, the EPA has little to demonstrate that such consideration is consistent with statutory text.
Even when an agency considers an irrelevant co-benefit, courts should not strike the regulation if the error was harmless. The Administrative Procedure Act instructs that “due account shall be taken of the rule of prejudicial error” on judicial review.1225 U.S.C. § 706.
This means courts should “uphold[] unsound agency decisions when they are confident that the agency would reach the same decision on remand.”123Nicholas Bagley, Remedial Restraint in Administrative Law, 117 Colum. L. Rev. 253, 302 (2017); Shinseki v. Sanders, 556 U.S. 396, 407 (2009) (applying this standard).
The harmless error doctrine is intended to preserve both judicial and administrative resources when a remand would ultimately lead to the same regulation.
The question is, therefore, whether the agency would have prioritized promulgating a different regulation had it not considered the irrelevant co-benefits.124Although the Supreme Court has not required the benefits of a rule to exceed its costs, the Court may be edging towards prohibiting a gross disparity between costs and benefits. Michigan v. EPA, 576 U.S. 743, 752 (2015) (“One would not say that it is even rational, never mind ‘appropriate,’ to impose billions of dollars in economic costs in return for a few dollars in health or environmental benefits.”); Michigan v. EPA, 213 F.3d 663, 679 (D.C. Cir. 2000) (“[R]eviewing courts will read statutes as authorizing regulations with benefits at least ‘roughly commensurate with their costs, unless there is a clear legislative statement to the contrary.’ ” (quoting Cass R. Sunstein, Interpreting Statutes in the Regulatory State, 103 Harv. L. Rev. 405, 487 (1989))). If a court disallows consideration of irrelevant co-benefits, it may strike the regulation if the remaining benefits are far lower than costs. That court would, in effect, override the harmless error inquiry by holding that the rule could not legally be promulgated on remand.
Co-benefits directly measure only the effect of the regulation and not the intent of the agency. When an agency relies on co-benefits, it points to them as one reason it pursued that regulation. But, when the regulation is overdetermined—that is, when the agency had many good reasons to pursue the regulation aside from the co-benefits—courts should not waste resources on a remand.
Harmless error review of agency reliance on irrelevant co-benefits would be fact intensive. The inquiry should include, among other factors:125Courts only rarely apply the harmless error standard, Bagley, supra note 123, at 259, and have apparently never applied it when remanding an agency action for consideration of an irrelevant factor.
the proportion of co-benefits to direct benefits, as a disproportionate number of co-benefits may indicate they were an important factor in the agency’s analysis; whether the agency was required to promulgate the regulation or a similar regulation under a statute or court order; whether the agency had other strong reasons to prioritize the regulation at issue; and what role the agency said the co-benefits played in their decision to promulgate the regulation.126For example, in Wyoming v. U.S. Department of the Interior, the BLM said it wanted to obtain the climate co-benefits because the EPA was too slow to regulate greenhouse gases. 493 F. Supp. 3d 1046, 1066 n.18 (D. Wyo. 2020) (quoting Waste Prevention, Production Subject to Royalties, and Resource Conservation, 81 Fed. Reg. 83008, 83019 (Nov. 21, 2016) (to be codified at 43 C.F.R. pts. 3100, 3160, 3170). The Court dinged the BLM for “hubristically justif[ying]” its rule on that basis. Id.
B. Indirect Benefits as an Indication of Pretext
Under this Note’s proposal, even if a court does not find a co-benefit to be an irrelevant factor, it could still hold that reliance on the co-benefit indicates the agency regulated under a pretextual purpose. The Supreme Court recently held that pretextual reasoning is a sufficient reason to strike a regulation.127Dep’t of Com. v. New York, 139 S. Ct. 2551, 2575–76 (2019).
Co-benefits are an awkward fit in the pretext context because disclosing reliance on co-benefits is, by nature, more transparent than not doing so. Still, significant or disproportionate co-benefits could indicate the agency was solely targeting those co-benefits under the pretext of another purpose.
An agency may pretextually promulgate a regulation for two primary reasons. First, it may do so to avoid the political blowback from promulgating an unpopular regulation that achieves its goals directly, instead attempting to promulgate a more popular regulation that achieves its goals indirectly, via co-benefits.128See, e.g., id. (striking a pretextual agency action).
Second, it might attempt to obtain a regulation’s co-benefits to avoid statutory limitations on gaining those benefits directly.129E.g., Wyoming, 493 F. Supp. 3d at 1068–74.
It would subvert statutory structure if an agency’s sole purpose is to obtain co-benefits. Nonetheless, courts should not strike down regulations on the basis of pretext when co-benefits are the only evidence of pretext. It would be nearly impossible for a court to determine why an agency issued a particular regulation using co-benefits alone, and there would be high costs to an incorrect judicial decision.
1. Motivations for an Agency to Engage in Pretextual Regulation
In Department of Commerce v. New York, the first Supreme Court case to hold that a regulation may be struck down for pretextual reasoning, the agency appeared to invent a justification for its action to conceal its true, but politically unpopular, motivation.130Dep’t of Com., 139 S. Ct. at 2575–76.
That case began when Secretary of Commerce Wilbur Ross attempted to add a census question related to citizenship.131Id. at 2562.
The agency maintained that Secretary Ross added the question pursuant to a request from the Department of Justice (DOJ) to help it enforce the Voting Rights Act.132Id.
However, litigation revealed that Secretary Ross had repeatedly pushed an unwilling DOJ to send him the request and had decided to add the citizenship question beforehand.133Id. at 2575. After oral argument, but before the Court’s decision, the New York Times released a trove of documents from a deceased Republican strategist, which showed the strategist pushed top Department of Commerce officials to add a citizenship question to aid Republican gerrymandering efforts. Michael Wines, Deceased G.O.P. Strategist’s Hard Drives Reveal New Details on the Census Citizenship Question, N.Y. Times (May 30, 2019), https://www.nytimes.com/2019/05/30/us/census-citizenship-question-hofeller.html [perma.cc/8RE2-AXNS].
Chief Justice Roberts, writing for the Court, remanded the agency decision because “[u]nlike a typical case in which an agency may have both stated and unstated reasons for a decision, here the [Voting Rights Act] enforcement rationale—the sole stated reason—seems to have been contrived.”134Dep’t of Com., 139 S. Ct. at 2575.
First, an agency might promulgate a pretextual regulation to avoid political blowback, like in Department of Commerce v. New York itself.135Eidelson, supra note 67, at 1761–63 (arguing the pretext in Dep’t of Com. v. New York was designed to avoid political costs associated with the true motivation).
That case was motivated, at least in part, by concerns about ensuring political accountability for agency action. Agencies must offer “reasoned explanation” for their regulatory decisions so that such “reasons . . . can be scrutinized by . . . the interested public.”136Dep’t of Com., 139 S. Ct. at 2575–76.
Permitting agencies to lie about the reasons behind their actions allows them to evade political blowback from the public and from Congress.137Eidelson, supra note 67, at 1785–93.
Political accountability is a critical check on agency action performed by unelected officials.138Id. at 1751; see supra Section II.A.
Second, an agency might, in theory, regulate indirectly to evade the statutory requirements associated with regulating directly. For example, imagine that the EPA perceives that the statutory threshold makes it too difficult to further regulate particulate matter under the CAA’s NAAQS program.139NAAQS must be set to the level “requisite to protect the public health” with an “adequate margin of safety.” 42 U.S.C. § 7409(b)(1).
Instead, it might pretextually issue a hazardous air pollutant regulation that is solely intended to indirectly reduce particulate matter. That regulation would likely have significant particulate matter co-benefits. During oral argument in Michigan v. EPA, Chief Justice Roberts alluded to this situation when he expressed concern that the “disproportionate amount of [particulate matter co-]benefit” might raise a ‘red flag’ that the rule was a pretextual attempt to avoid the strictures of the NAAQS program.140Transcript of Oral Argument, supra note 59, at 62–63.
Chief Justice Roberts is correct in implying that the structure of the CAA demonstrates that Congress intends for certain pollutants to be directly regulated under certain sections.141Three significant CAA programs for stationary sources (the NAAQS program, the Hazardous Air Pollutant program, and the New Source Performance Standard program) each regulate an entirely separate universe of pollutants. See Richard K. Lattanzio, Cong. Rsch. Serv., RL30853, Clean Air Act: A Summary of the Act and Its Major Requirements 3–17 (2022) (describing the substantial regulatory authority granted to the EPA under the NAAQS, Hazardous Air Pollutant, and New Source Performance Standard programs); 42 U.S.C. 7412(b)(2) (noting that the Hazardous Air Pollutant program may not regulate pollutants under the NAAQS program); 42 U.S.C. 7411(d) (noting that the New Source Performance Standard Program may not regulate pollutants under the NAAQS or Hazardous Air Pollutant program).
And Congress created different statutory guardrails for the EPA to meet when it regulates those pollutants. To regulate particulate matter under the NAAQS, the EPA must show the emissions standard is “requisite to protect the public health” with an “adequate margin of safety.”14242 U.S.C. § 7409(b)(1). Although this is a stringent standard, it is also uniform throughout the country and does not fully protect sensitive populations from all health effects of particulate matter. Hazardous air pollutant regulations can, therefore, create substantial co-benefits via reductions of particulate matter below the NAAQS.
This is a “health-based” standard. In contrast, hazardous air pollutant emissions standards are “technology-based”; the EPA must show they are commensurate with the “maximum degree of reduction” achievable with the application of the best available pollution control technology.14342 U.S.C. § 7412(d)(2).
In some situations, the EPA may, therefore, be able to more strictly regulate particulate matter indirectly through the Hazardous Air Pollutant program than it could directly through the NAAQS program.
As in the example above, even co-benefits related to statutory purpose could be an indication of pretext. The EPA is permitted to consider these co-benefits as one factor when prioritizing regulations; Chief Justice Roberts seemed to allow as much, noting that “it’s a good thing if your regulation also benefits in other ways.”144Transcript of Oral Argument, supra note 59, at 64. The Court has elsewhere said that it is typically unremarkable for an agency or agencies to have overlapping authorities over a single issue. Massachusetts v. EPA, 549 U.S. 497, 532 (2007).
But the EPA would go too far if it promulgated the regulation solely to obtain those co-benefits.145See Transcript of Oral Argument, supra note 59, at 64.
The same pretext concerns might apply if the co-benefits targeted effects that an agency is allowed to regulate only under an entirely different statute. For example, if a CAA regulation created substantial water quality co-benefits, a court could justifiably worry that the EPA was attempting to dodge the requirements of the Clean Water Act. Although this Note contends water co-benefits are a relevant factor under the CAA,146See supra note 104 and accompanying text.
the EPA may not subvert Clean Water Act statutory guardrails by promulgating a CAA regulation intended solely to protect water quality.
These concerns would be heightened if an agency considered co-benefits outside of its statutory purview, like the BLM considering climate co-benefits in Wyoming v. U.S. Department of the Interior.147See Wyoming v. U.S. Dep’t of the Interior, 493 F. Supp. 3d 1046, 1068 (D. Wyo. 2020).
The same might be true if the EPA considered lower consumer prices of a manufactured product, for example. The EPA might be more incentivized to regulate these effects indirectly because it does not have any authority to do so directly. However, the EPA rarely, if ever, relies on nonenvironmental co-benefits in its CAA regulations. And a court might not even reach the pretext issue if it finds these co-benefits to be irrelevant.
2. Issues with Judicial Review of Pretext
Even though an agency could hypothetically promulgate pretextual regulations to evade political accountability or statutory guardrails, it would be unwise for courts to strike down a regulation if co-benefits are the only evidence of pretext. Co-benefits are a blunt tool for determining agency intent. Furthermore, the costs of a false positive judicial decision are high, and the risk of agency malfeasance is low.
It would be very difficult to determine, based on co-benefits alone, whether the agency promulgated a regulation solely to obtain co-benefits or if the co-benefits were only one reason the agency prioritized that regulation.148Cf. Dep’t of Com. v. New York, 139 S. Ct. 2551, 2575–76 (2019) (“Unlike a typical case . . . the sole stated reason [here] seems to have been contrived.”).
This inquiry would be more difficult than the harmless-error inquiry described above.149See supra Section III.A.
There, the court would need only ask whether the co-benefits formed an important part of the agency’s erroneous justification for the regulation. Here, a court would need to determine that the EPA was solely targeting the co-benefits to avoid political accountability or statutory guardrails.
It is unclear how a court could find that an agency was trying only to achieve co-benefits if a regulation’s direct benefits otherwise meet the statutory requirements for regulation. However, if the agency gives some other indication of pretextual motivation—like the BLM’s admission in Wyoming v. U.S. Department of the Interior that it was pursuing the regulation because the EPA was slow to regulate greenhouse gases from gas wells—co-benefits can serve as additional evidence of pretext.150Wyoming, 492 F. Supp. 3d at 1066.
A false positive judgment against the agency on pretext grounds would be costly, particularly under the CAA. It is utterly ordinary for the EPA to issue regulations with a disproportionate amount of co-benefit.151Supra notes 14–15 and accompanying text.
Air pollution regulations create co-benefits by their nature. It would put the EPA in a nearly untenable position if each of these regulations raised suspicions that the EPA was acting pretextually. The EPA would need to blind itself to the full effects of its regulations and, therefore, prioritize inferior rules.
To think about the issue another way, striking down regulations with a disproportionate amount of co-benefit would, in effect, raise the statutory threshold for particularly effective regulations. Under such a rule, if a New Source Performance Standard regulation created no co-benefits, the EPA would need to show only that the pollutant endangered public health and that the standard reflected the best available technology.15242 U.S.C. § 7411(a)(1); § 7411 (b)(1)(A).
But, if the regulation created significant co-benefits, the EPA would additionally need to show that direct benefits are not disproportionately smaller than co-benefits. This would give a single CAA provision different effects under different circumstances. Nothing in the statute indicates that Congress intended such a result.
Finally, the risk of agency malfeasance is low, at least for the EPA. The historical consistency of co-benefit consideration somewhat insulates the practice from charges of malfeasance by any particular administration.153See Livermore, supra note 21, at 1069.
And, specifically, there is little indication the EPA has attempted to avoid political accountability in its CAA regulations. The Supreme Court recently struck down an Obama-era CAA rule regulating greenhouse gases from power plants—not because the EPA was trying to obscure its motives but because the EPA attempted to resolve a question of “political significance.”154See West Virginia v. EPA, 142 S. Ct. 2587 (2022).
The Trump Administration also came under political scrutiny for withdrawing that same regulation.155E.g., John Walke, Trump’s “Affordable Clean Energy” Rule: A Dirty Lie, Nat. Res. Def. Council (Jan. 28, 2019), https://www.nrdc.org/bio/john-walke/trumps-affordable-clean-energy-rule-dirty-lie [perma.cc/RK4M-GEWE].
Furthermore, the vast majority of CAA co-benefits stem from other air pollutants. Unlike in Department of Commerce v. New York, where the pretextual purpose was much more politically tolerable than was the true purpose, there is little reason to think that regulation of one air pollutant is substantially more popular than regulation of another.
The risk of EPA pretext to evade statutory guardrails is similarly low. If the EPA wants to regulate a particular air or water pollutant, it has the authority to regulate the pollutant directly. Though that authority comes with limitations, it might not be much easier for the EPA to regulate the pollutant indirectly. And, although the EPA may be more tempted to indirectly regulate effects outside of its statutory authority, it has rarely, if ever, relied on these effects to justify a rule.
Conclusion
Critics describe consideration of co-benefits as an unlawful and underhanded expansion of regulatory authority.156See, e.g., Chamber of Commerce Amicus Brief, supra note 56, at 16–17.
On the contrary, co-benefits do not expand regulatory authority, but they do improve its functioning. This Note contends that co-benefits further the traditional administrative law values of rational decisionmaking, transparency, and accountability. Since Congress wants agencies to pursue these values, courts should consider co-benefits presumptively permissible. And, even if statutory authorization were required, implicit authorization for co-benefits exists under the Clean Air Act, which was designed to be implemented in a rational and holistic manner to achieve its goals. The same could likely be said for many other statutes and agencies.
* J.D., May 2024, University of Michigan Law School. A special thank you to Professor Nina Mendelson, who sparked my interest in administrative law. Her expertise and guidance were essential to this Note. Thanks as well to Professor Rachel Rothschild, who planted the seed for this piece, and Professor Andy Buchsbaum, who helped shape its early versions. Thank you to the editors of Michigan Law Review who gave edits and support: Eddie Plaut, Hannah Cohen Smith, Kassie Fotiadis, Katie M. Osborn, Jordan Schuler, Ashley Munger, Hannah Juge, John Grosboll, and Ethan Haughie, and thanks as well as to First Lieutenant Joseph Westberry. All errors are my own. Finally, thanks most of all to my parents, Max and Eric.