Yesterday, the Supreme Court in Michigan v. EPA threw out EPA’s regulations protecting the American public from mercury and other hazardous emissions of power plants.
In another instance of judicial activism by the Roberts court, the majority refused to defer to EPA’s decision to ignore costs in deciding whether to regulate power plant emissions.
The decision turned on the meaning of the word “appropriate” in a section of the Clean Air Act that addressed hazardous air pollutant (HAP) emissions from fossil fuel-fired power plants. Before EPA subjected HAPs emissions from power plants to stringent technology-based regulations, it had to decide whether regulating those emissions was “appropriate and necessary,” given the other controls that the statute imposed on power plants to reduce acid rain.
If EPA made the “appropriate and necessary” finding, the statute required EPA to subdivide power plants into various categories and promulgate emissions limitations for new and existing plants in those categories based on the “maximum achievable control technology” (MACT). The statute defined MACT to mean the very best operating plant in the category in the case of new power plants above a certain size and the top 12 percent of plants in the category for existing power plants.
Fifteen years ago, EPA determined that regulation of HAPs emissions from power plants was appropriate and necessary. In making that finding, it considered the adverse effects of HAPs on human health and the environment (which were considerable in the case of mercury) and the availability of controls. But it did not consider the cost of complying with the standards that it would be promulgating as a result of the finding. Among other things, at the time it made the finding, it had no way to quantify the cost of compliance with a regulation that it had not yet begun to think about.
EPA pointed out at the time that costs would play a role in categorizing the industry and in determining MACT for the various categories, and cost considerations (both by EPA in categorizing the industry and Congress in determining the stringency of MACT) did play a dominant and repeated role in setting the standards.
The Supreme Court, in an opinion written by Justice Antonin Scalia, held that EPA’s interpretation of the word “appropriate” was “unreasonable.” The majority held that even in making the threshold decision whether to regulate HAPs emissions from power plants, EPA had to consider the costs imposed by the regulations that it would eventually promulgate.
To Justice Scalia, the word “appropriate” was “the classic broad and all-encompassing term that naturally and traditionally includes consideration of all the relevant factors.” He continued: “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.” Riding a conservative warhorse called the perversity thesis, he observed that an environmental regulation might cause more harm to health and the environment than it prevented and concluded that “[n]o regulation is ‘appropriate’ if it does significantly more harm than good.”
The problem with Justice Scalia’s analysis is that it focused on the final product of the regulatory process -- the requirements that the regulations imposed on the industry -- and not the decision whether to initiate the process of determining those requirements.
EPA fully conceded that it should (and Congress did) consider costs in deciding how to regulate HAPs emissions from power plants, and both EPA and Congress clearly did consider costs at the stage of the decision-making process where the agency specified the MACT-based standards. EPA simply took the position that costs were not a relevant consideration whether to initiate the regulatory process in the first place. After all, the purpose of the Clean Air Act, under which EPA was contemplating regulation, is to protect the air Americans breathe from pollution. So it’s perfectly reasonable for EPA to consider first whether some regulation is “appropriate” to protect the nation’s air-breathers, and then consider later the question of how much and what type of regulation is practical.
Under the well-known Chevron doctrine, reviewing courts are supposed to defer to an agency interpretation of ambiguous language in its own statute if the agency’s interpretation is “reasonable.” In doing so, the courts are to consider the ambiguous language in the context of the entire statute and the congressional intent in enacting the statute, as the Court did last week in rejecting the most recent legal challenge to the Affordable Care Act.
The majority opinion in Michigan v. EPA did not defer at all to the agency’s perfectly reasonable interpretation of the word “appropriate” in the context of the threshold decision whether to regulate HAPs emissions from power plants. As Justice Kagan’s dissent noted, interpreting words like “appropriate” to provide a low threshold for regulating dangerous pollutants is quite consistent with the approach that Congress has prescribed for promulgating national ambient air quality standards, new source performance standards, and hazardous emissions standards for every emitter of HAPs other than power plants. In deciding whether to invoke the regulatory process, Congress did not want for EPA to go to a great deal of trouble gathering data and analyzing the costs and benefits of possible outcomes of that process.
Surely it was not unreasonable for EPA to take the same approach to HAPs emissions from power plants.
More to the point, one might say that it is not even rational to require an agency to speculate about the costs that a final regulation to be written more than a decade after a threshold finding would impose on an industry at the time that the agency initiated a rulemaking process that could result in a wide variety of different regulatory requirements.
Justice Scalia analogized EPA’s threshold determination to a decision whether it is “appropriate” to buy a Ferrari without thinking about cost because one plans to think about cost later when deciding whether to upgrade the sound system. But that is a misleading analogy. The appropriate analogy is whether it would be “appropriate” for a billionaire to disregard cost considerations in deciding whether to walk through the door of a Ferrari dealership to kick some tires and gather information, as part of a much larger car search. Walking through the door commits the potential car-buyer to nothing, and it’s hardly the end of the process. In this case, EPA decided that technologies existed for reducing mercury emissions in making its threshold determination — technologies that the exceedingly profitable electric power industry is able to afford. So while the relative costs of various cars will undoubtedly be a consideration in deciding whether to purchase a particular model, that decision comes later. That’s exactly what EPA did. It decided regulation was in order, and then subsequently decided how much of it was appropriate, considering costs as part of that process. And, not for nothing, the monetized benefits of the eventual regulation are projected to vastly exceed the costs, saving thousands of American lives a year along the way.
The majority opinion also contains language suggesting that the agency would have to undertake some assessment of the benefits of various regulatory outcomes as well. Justice Scalia was willing to concede for the moment that the threshold analysis need not require the agency “to conduct a formal cost-benefit analysis in which each advantage and disadvantage is assigned a monetary value.” But we can expect industry lawyers to argue that the word “appropriate” is synonymous with cost-benefit analysis in this and other statutory contexts in the future. Worse, industry lawyers will cite Michigan v. EPA for the proposition that an agency interpretation of any statutory language that does not explicitly exclude cost considerations is unreasonable and therefore not worthy of Chevron deference.
In the wake of Michigan v. EPA, the agency will have to go back to the drawing board and make a fresh determination whether it is “appropriate” to regulate mercury emissions from power plants after considering the costs of the regulations.
Fortunately, in the process of writing the HAPs regulations for power plants, EPA’s economists prepared a regulatory impact analysis (RIA) detailing costs and benefits of the alternatives that it identified later in the regulatory process. The RIA concluded that the benefits ($37 to $90 billion) far exceeded the costs ($9.6 billion). This assessment included the “co-benefits” in terms of lives saved and environmental loss prevented because of the reductions in fine particulate matter and other pollutants that would also be removed by the technologies that removed HAPs from power plant emissions.
Justice Scalia (and Chief Justice Roberts at oral argument) emphasized the “direct” benefits yielded by the removal of only the HAPs themselves ($4 to $6 million because mercury does not kill people, it only causes neurological damage in fetuses, infants and children). The opinion did not say that EPA’s analysis on remand was limited to direct benefits. To do so would be wholly inconsistent with cost-benefit analysis as it is practiced by federal agencies (and most economists). If the agency is limited to direct benefits, then it is unlikely that it will promulgate any rule on remand, and it will have wasted 25 years’ worth of effort, while the Supreme Court moved the regulatory goal posts.
Now for some good news.
The Supreme Court’s decision should not have a disastrous impact in the real world. EPA’s efforts to regulate mercury emissions from power plants have been in the works for so long that nearly all coal-fired power plants built in the past 10 years and many older power plants have installed technology to address mercury emissions. The Department of Energy’s Energy Information Institute estimates that as much as 70 percent of existing coal-fired power plants currently comply with EPA’s regulations. And many older plants with the poorest controls will close down if EPA’s forthcoming regulations limiting greenhouse gas emissions from power plants survive judicial review (not necessarily a given with this court).
In addition, several states have already established stringent emissions limitations for mercury under their own laws, some of which are more stringent than those imposed by EPA’s regulations.
But in the 21 states that challenged the regulation, including states like Texas and Ohio that are home to some of the largest mercury emitters in the country, the Supreme Court decision will probably result in the termination of ongoing mercury control projects and may well inspire companies to turn off mercury controls that have already been installed.
In short, the Supreme Court may have just given some of the country’s oldest and dirtiest power plants a decidedly inappropriate reprieve.