Legal Analysis

Clean Air Power Sector

MATS, Cost-Benefit Analysis, and the Appropriate and Necessary Finding


Shortly after Christmas, EPA is expected to issue a proposal reviewing the Mercury and Air Toxics Standards, which were signed by EPA Administrator Lisa Jackson 7 years ago and which have been operating since 2015 to achieve significant reductions in hazardous air pollutant emissions from power plants at lower-than-projected costs.

The proposal is expected to rescind or revise a 2016 “supplemental finding” that, after considering the cost of regulation, it is “appropriate and necessary” to regulate power plant hazardous air pollutants. Congress required the finding in section 112(n)(1)(A) of the Clean Air Act because of the possibility that other provisions of the Clean Air Act might operate to reduce power plant hazardous air pollutant emissions. If so, then EPA would have no need to use section 112 to achieve those reductions.[i] If the other Clean Air Act provisions requiring reductions in power plant emissions did not also result in reductions in hazardous air pollutants from power plants, then EPA was to use section 112 to achieve the hazardous air pollutant reductions from power plants. Congress clearly did not contemplate that neither other provisions in the Clean Air Act nor section 112 would be used to achieve hazardous air pollutant reductions, or that hazardous air pollutants emitted by power plants would remain uncontrolled. Yet, EPA is about to propose an outcome that either directly by rescinding the MATS standard, or indirectly by inviting further litigation challenging the MATS standards, would do just that.

The pending proposal would turn the Clean Air Act on its head: by rescinding the appropriate and necessary finding or revising it in a way to make it more vulnerable to legal challenge, the EPA would be ensuring that none of the tools of the Clean Air Act would be available for reducing power plant hazardous air pollutants.

In addition, the proposal is likely to rely on the application of cost-benefit analysis to reach this outcome and do so in a way that will add to the EPA’s legal challenges.

Cost-Benefit Analysis may be the Goal

Earlier this week, we published a background memo detailing many of the issues EPA could face if it proposes to change or withdraw the MATS appropriate and necessary finding, the supplemental finding and/or the pollution standards themselves.

We looked at actions the EPA has taken since the beginning of the Trump administration, issues raised in litigation on the supplemental finding, and Acting Administrator Wheeler’s reported claims that EPA will not change the MATS pollution control standards, and concluded that altering the methods and analysis underlying the supplemental appropriate and necessary finding may be the EPA’s Trojan Horse objective in issuing the pending proposal.

EPA may be seeking to accomplish two things: 1) narrow the scope of benefits evaluated in cost-benefit analysis (which we discussed in our earlier memorandum) and 2) elevate the use of cost-benefit analysis in setting, as opposed to simply evaluating, pollution control standards. In so doing, EPA may be going beyond its legal authority and be departing from its own decades-long approach to considering cost under the Clean Air Act.

As a result, EPA will also have to contend with a threshold legal question: is it permissible for EPA to rely on cost-benefit analhttps://eelp.law.harvard.edu/rolling-back-the-mercury-and-air-toxics-standards-proposed-withdrawal-of-appropriate-and-necessary/ysis in making the appropriate and necessary finding? The answer may be “no”.

Cost-Benefit Analysis may be Arbitrary and Capricious if Applied to the Appropriate and Necessary Finding

If EPA’s justification for rescinding the finding relies too heavily on cost-benefit analysis, that is if EPA proposes to rescind the finding simply because it calculates that the costs exceed the benefits, then it risks making a decision that is arbitrary and capricious. After the proposal is issued, commenters will be able to fill the record with expert analysis and a long list of past examples demonstrating 1) that the monetary value attributed to the public health and societal benefits of reducing air pollution understates those benefits and 2) that the projected costs of reducing air pollution exceeds the actual costs.

The seven years since EPA issued MATS provide examples of both points.

EPA’s January 2017 brief to the D.C. Circuit defending the supplemental appropriate and necessary finding in Murray Energy v. EPA reported: “Indeed, EPA was unable to quantify many important benefits of the Standards, including the majority of the public health benefits associated with reductions in hazardous air pollutants, which are the focus of the Standards. EPA was able to quantify only a small subset of health benefits related to reducing mercury (the $4-6 million value on which Petitioners seek to focus the benefit-cost analysis, see Pet. Br. 56), and could not quantify other health and environmental benefits related to reducing mercury or other hazardous air pollutants. See id. at 75,040/2.”

Since then scientists have reported:

“(1) Recent research demonstrates that quantified societal benefits associated with declines in mercury deposition attributable to implementation of MATS are much larger than the amount estimated by EPA in 2011. (2) As-yet-unquantified benefits to human health and wildlife from reductions in EGU mercury emissions are substantial. (3) Contributions of EGUs to locally deposited mercury have been underestimated by EPA’s regulatory assessment.”

When it issued MATS in 2012, EPA projected that the annual costs of complying with MATS would be $9.8 billion as of 2015. This past summer the utility industry reported that the total cost of complying with the 7-year-old program has been $18 billion.

An approach that over-relied on cost-benefit analysis in decision-making – in this instance to rescind the appropriate and necessary finding – would be inherently flawed since cost-benefit analysis does not encompass the full societal benefits of reducing hazardous air pollutants nor is it reliably predictive of the costs of achieving air pollution reductions. If EPA were to center its proposed decision to rescind the appropriate and necessary finding on a cost-benefit analysis it would risk being arbitrary and capricious in its decision-making.

The Clean Air Act may not Authorize the use of Cost-Benefit Analysis in the Appropriate and Necessary Finding

It is not clear where EPA would find the legal authority to adopt cost-benefit analysis as the method by which it made its decision to rescind an appropriate and necessary determination. When the Supreme Court directed EPA to consider cost in making the appropriate and necessary finding in Michigan v EPA, the Court made a point of not addressing how EPA should consider cost.

When EPA issued the supplemental appropriate and necessary finding in 2016, its preferred approach to considering cost was not to perform a cost-benefit analysis (although it did include, as an alternative, a cost-benefit analysis). Instead, EPA sought to ground its approach in its understanding of the Clean Air Act’s mandates with respect to hazardous air pollutants. Referring in the final supplemental finding preamble to the proposed supplemental finding, EPA stated:

“EPA explained that it preferred this approach to a formal benefit-cost analysis given the statutory objectives of CAA section 112, in particular Congress’ determination that HAP emissions are inherently harmful, and the instruction from Congress to protect the most sensitive populations from those harms. See Legal Memorandum at 6–20. EPA found that CAA section 112(n)(1)(A)’s emphasis on the required studies supported its interpretation that while cost is an important factor that it must consider in making the appropriate and necessary finding, it is one of several factors that must be considered and the statutory text does not support a conclusion that cost should be the predominant or overriding factor. See id. at 11–15. EPA’s preferred approach to considering cost allows the Administrator to weigh the full range of factors relevant to making a determination under CAA section 112(n)(1)(A) of whether it is appropriate and necessary to regulate HAP emissions from EGUs. Moreover, because the Supreme Court’s holding did not disturb the scientific assessments and conclusions made in the original appropriate and necessary finding, many of which were challenged and upheld by the D.C. Circuit in White Stallion, the Administrator concluded that the task on remand was to determine whether a consideration of cost caused her to alter her prior conclusion that it was appropriate to regulate HAP emissions from EGUs under CAA section 112. See 80 FR 75038; Legal Memorandum at 20.”

Instead of using cost-benefit analysis as the foundation of its appropriate and necessary determination EPA offered a different approach as its preferred one. It assessed whether the costs of complying with the mandates of section 112 using the MATS pollution control standards were reasonable. EPA compared annual industry-wide compliance costs with annual revenues, projected the impact of MATS compliance on consumers’ electricity bills, and took account of similar factors showing that MATS compliance was affordable. EPA concluded that the costs of complying with MATS were reasonable and on that basis confirmed the original appropriate and necessary finding after considering costs, as the Supreme Court required.

In its 2016 supplemental finding, EPA’s preferred approach was true to the Clean Air Act itself by avoiding reading the Supreme Court’s Michigan as a mandate to second-guess section 112 and Congress. As EPA stated, section 112 reflected Congress’ determination that hazardous air pollutants were inherently threatening to air quality and public health. Section 112 mandated that EPA follow a prescribed methodology in setting pollution control standards for 189 hazardous air pollutants listed in section 112 itself for every major industrial source of hazardous air pollutants, chief among which is the power sector.

Neither the section 112(n)(1)(A) appropriate and necessary finding requirement nor the Michigan majority’s “consider cost” ruling was a mandate that EPA revisit the fundamental decision Congress made when it enacted section 112 (which required EPA to set standards for hazardous air pollutants). Instead, section 112(n) reflected Congress’ view that other provisions of the Clean Air Act might reduce power plant hazardous air pollutant emissions that would remove the need for EPA to use section 112 to achieve those reductions.[ii] Again, Congress clearly did not contemplate that neither other provisions in the Clean Air Act nor section 112 would be used to achieve hazardous air pollutant reductions, and that hazardous air pollutants emitted by power plants would remain uncontrolled.

Even though EPA also presented the results of its 2016 cost-benefit analysis as an alternative basis for the supplemental finding, it did so without changing its cost-reasonableness framing of the Supreme Court’s “consider cost” directive.

If EPA’s proposal now reverses that approach, and presents cost-benefit analysis as the primary analysis and justification for rescinding the appropriate and necessary finding, then that cost-benefit analysis will have hijacked the authority of Congress to require reductions in power plant hazardous air pollutants under section 112 if other Clean Air Act provisions have proven ineffective in doing so.[iii]


[i] At the time of enactment the Clean Air Act Amendments of 1990 included provisions requiring significant reductions in sulfur dioxide emissions as a precursor of acid rain as well as other provisions expected to require power plants to install and operate pollution control equipment. Congress viewed as an open question whether the pollution controls used in meeting those provisions would result in significant reductions in hazardous air pollutants as well, rendering separate regulations of power plant hazardous air pollutants under section 112 duplicative and unnecessary. Hence, the inclusion of section 112(n)(1)(A)’s requirement of an appropriate and necessary finding before EPA issued power plant regulations under section 112.

[ii] Congress easily could have directed EPA to apply cost-benefit analysis as the decisive tool in making the appropriate and necessary finding, but did not. If EPA acts on this proposal, it is doing so without precedent under the Clean Air Act. At the time of enactment the Clean Air Act Amendments of 1990 included provisions requiring significant reductions in sulfur dioxide emissions as a precursor of acid rain as well as other provisions expected to require power plants to install and operate pollution control equipment. Congress viewed as an open question whether the pollution controls used in meeting those provisions would result in significant reductions in hazardous air pollutants as well, rendering separate regulations of power plant hazardous air pollutants under section 112 duplicative and unnecessary. Hence, the inclusion of section 112(n)(1)(A)’s requirement of an appropriate and necessary finding before EPA issued power plant regulations under section 112.

[iii] Nearly every provision of the Clean Air Act – from those that require EPA to set health-based ambient air quality standards to those that authorize EPA to set technology-based standards for automobile tailpipe emissions to those that mandate smokestack pollution control standards for power plants and factories (including section 112) – conspicuously avoids requiring EPA to compare costs and benefits in setting those standards – and section 112(n)(1)(A), which makes no reference to cost-benefit analysis, is no exception. Instead, these provisions list a wide range of factors – including but not limited to cost – that EPA must take into account in carrying out the pollution control and air quality improvement purposes of Congress. That Congress did not direct EPA to apply cost-benefit analysis as the decisive tool in making the appropriate and necessary finding is a strong indication that EPA’s obligation under section 112(n)(1)(A) is to take a range of factors into consideration, not least of which is Congress’ clear directive that whether through the provisions of section 112 or by operation of other provisions in the Clean Air Act, power plant hazardous air pollutants be significantly reduced.