On Monday, June 29, 2015, by a vote of 5-4 with Justice Kennedy in the majority, the U.S. Supreme Court (Court) invalidated a rule issued by the U.S. Environmental Protection Agency (EPA) limiting power plant emissions of mercury and air toxics (Mercury and Air Toxics Standards or MATS). Writing for the majority, Justice Scalia found unlawful EPA's failure to consider costs at the time EPA determined that regulating these emissions is appropriate and necessary—the first step required under a provision of the Clean Air Act (the Act) unique to hazardous air pollutants (HAPs) emitted by coal- and oil-fired power plants. Michigan v. EPA, No. 14-46, slip op. at 3 (June 29, 2015). While in prior decisions the Court has sometimes found EPA's injection of cost considerations to be unlawful, the decision does not necessarily create any new law. It does suggest, however, a new willingness to scrutinize cost-benefit analyses, which have been required by the last six presidents and enforced for the White House by OMB. 

The immediate impact of the decision will likely be limited. The Court did not vacate the rule; it remanded it for further proceedings consistent with its opinion. Under its precedents, the D.C. Circuit, on the other hand, could vacate the rule or stay it in whole or in part while remanding it to EPA to address costs as instructed. If some or all of the MATS rule survives, then its less costly obligations, such as those pertaining to operating and reporting, will still apply to the estimated 70% of the plants that either installed emissions control systems before the date the rule became effective, April 16, 2015, or "shuttered" (see testimony of Michael J. Bradley) Many of these plants did so because of other EPA rules, or state rules limiting mercury emissions, and/or decisions to convert to natural gas. Thus, the most significant costs to be imposed by any replacement rule—those associated with installing controls—will apply only to the proportionally smaller number of plants subject to the EPA compliance deadline of April 16, 2016 (MATS Compliance Extension Status Update). 

The longer-term impacts of the Court's ruling, however, may be more significant, and have more significance to a broader swath of regulated entities. Both the majority and minority opinions discussed at length the role that agency cost analyses should play in review of rules. This exchange could be of great consequence as EPA prepares to issue a new round of Clean Air Act rules to address climate risk. In the final analysis, Michigan's biggest impact may be not on power plants, but rather, on the process of executive branch consideration of draft Clean Air Act rules, as EPA attempts to fill statutory gaps with, or reconcile, its White House-driven cost-benefit analysis. 


In 2014, the Court struck down a portion of an EPA rule that identified which sources need permits to emit greenhouse gases (GHGs) on grounds that EPA's interpretation of Sections 165 and 501 (Title V) of the Act would have required an enormous and transformative expansion in its regulatory authority without clear congressional authorization (UARG v. EPA). In Michigan, while the Court found that Section 112(n)(1)(A) of the Act did not expressly refer to costs, the "capaciousness" of the phrase "appropriate and necessary" "[r]ead naturally in the present context," "requires at least some attention to cost." The Court further stated: "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." Indeed, the Court appeared to be particularly offended that EPA failed to consider costs at the early stage of rule-making, where the agency itself pegged the costs at $9.6 billion a year, but the benefits of HAP reductions at only $4-6 million per year. Michigan, slip op. at 4. 

The Court's reaction is noteworthy for two reasons. First, it is significant that while the Court chastised EPA for failing to consider costs in making an "appropriate and necessary" finding, and purported to leave EPA discretion to decide (within reason) how it would account for costs, it nevertheless attacked EPA for inappropriately balancing costs against the subset of benefits the Court chose to highlight. Id. at 13. Pursuant to the governing executive order, to obtain White House approval for issuing a "significant" rule, an agency must provide a full accounting of, and evaluate, a rule's costs and benefits, including those that cannot be quantified, as well as the "co-benefits" or ancillary side effects associated with a proposed action. The $4-6 million figure cited by the Court captures only the direct benefits from mercury reductions EPA could quantify—those associated with preventing IQ decrements in children exposed to freshwater fish caught recreationally. However, EPA's regulatory impact analysis (RIA) noted that the rule would produce some costs and important benefits EPA could not monetize, such as other mercury reduction benefits (for example, neurological effects), as well as benefits it could monetize from non-mercury HAP reductions. Indeed, EPA projected health and climate co-benefits to amount to a whopping $37-90 billion, largely because the same technology that reduces mercury emissions also reduces other pollutants, such as fine particulate matter (PM2.5) emissions, which cause asthma and heart attacks. RIA, at ES-3. The Court dismissed the co-benefits as "not covered by the hazardous air pollutants program." Michigan, slip op. at 4. However, under long-standing principles governing review of draft rules submitted to the White House for pre-promulgation review, OMB has concluded that "[t]he consideration of co-benefits, including the co-benefits associated with reduction of particulate matter, is consistent with standard accounting practices and has long been required under OMB Circular A-4." 2014 OMB Draft Report to Congress. Further, non-HAP reductions are relevant, as they serve as surrogates for reductions of some regulated HAPs (PM2.5 is a surrogate for non-mercury metals, RIA, p. 1-7). 

When all of these factors are considered, it is not clear that the rule "impos[es] costs far in excess of benefits," as the Court concluded. Nor is it possible to determine whether the Court would have found EPA's failure to consider costs "unreasonable," had the Court given credence to the more fulsome set of benefits set forth in the RIA. 

Second, the Court's reaction is noteworthy insofar as it dismissed EPA's election to consider costs in Step 2 of its rule-making procedure. Section 112(n)(1)(A) of the Clean Air Act required that EPA first make an "appropriate and necessary" determination to regulate, after studying whether regulations issued under other sections of the Act would mitigate the public health hazards associated with power plant HAP emissions. If EPA determined that it would regulate, Section 112(n)(1)(A) also required that EPA set emissions limits in consideration of multiple criteria, including the costs of achieving the limits. See Sections 112(n)(1)(A), 112(d)(2). The Court found it irrelevant that EPA considered costs later, when EPA decided how to regulate. It held that the phrase "appropriate and necessary" compelled EPA to consider cost "before deciding whetherregulation is appropriate and necessary" (emphasis added). 

What Do We Take Away from the Decision? 

The decision is not necessarily inconsistent with the Court's prior rulings, which limit an agency's discretion to decline to consider costs to those instances where Congress has expressly set forth another standard. InWhitman v. American Trucking, which opinion was again authored by Justice Scalia, the Court foundunreasonable EPA's choosing to consider costs in standard setting under Section 109 of the Act, which expressly directs EPA to regulate on the basis of a factor [standards must be those requisite to protect the public health, allowing an adequate margin of safety] that on its face does not include cost. Similarly, inAmerican Textile, the Court held that OSHA acted reasonably when it chose not to conduct a cost-benefit analysis, finding it "reasonably necessary and appropriate" to set a standard for cotton dust. There the Court held that "cost-benefit analysis by OSHA is not required by the statute because feasibility analysis is." 

Thus, while Michigan may not alter the legal landscape, it does highlight the Court's willingness to "interpret" a legally irrelevant cost-benefit analysis. We say "irrelevant," because the Court held that it could not uphold the rule based on an analysis EPA did not itself invoke when it took action. Slip op. at 12. Nevertheless, the Court's frequent remarks that the rule's nearly $10 billion price tag far exceeded its purportedly negligible health benefits suggest that as a practical matter, the Court's selective reading of EPA's cost-benefit analyses had some role in the Court's finding EPA's action unreasonable.