How Justice Scalia’s Last Canon Is Unhinging Statutory Interpretation
Justice Antonin Scalia was, as much as anything else, known for insisting that the text of a statute alone – not its purposes, not its legislative history – should serve as the basis for the courts’ interpretation of the statute. Justice Scalia promoted canons of statutory construction – or at least what he deemed the valid ones – as a way of limiting the power of judges by setting rules for their interpretation of statutes. Yet he also warned, in a 1997 book, against “presumptions and rules of construction that load the dice for or against a particular result.” He worried that such “dice-loading” rules might effect “a sheer judicial power-grab.”
It is striking, therefore, that in one of his last majority opinions for the Supreme Court, Justice Scalia went out of his way to create such an interpretive rule. Writing for a 5-4 majority in Michigan v. Environmental Protection Agency (EPA), he found that EPA had erred in declining to consider costs in determining that regulation of hazardous air pollutants – such as mercury – from power plants was “appropriate and necessary” under section 112 of the Clean Air Act.
Justice Scalia’s reasoning went beyond the statutory provision and agency regulation at hand and suggested that agencies’ purportedly general practice of considering costs in deciding whether to regulate had made the interpretive default one in which agencies must consider cost in order to engage in “reasonable regulation.” In Michigan v. EPA, in other words, Justice Scalia created a brand-new, dice-loading, anti-regulatory canon of statutory construction.
The lower courts have begun to apply this canon with gusto, and in cases far removed from section 112 of the Clean Air Act. In the biggest of these cases so far, MetLife v. Financial Stability Oversight Council (FSOC), Judge Rosemary Collyer of the federal district court in Washington, D.C. relied heavily on Michigan v. EPA in finding that the FSOC had erred in determining that the insurance giant MetLife was a systemically important financial institution – or “too big to fail” – because it had not considered the costs of this designation to MetLife.
In Independent Pilots Association v. Federal Aviation Administration (FAA), the D.C. Circuit relied on Michigan v. EPA in holding, in two quick paragraphs in an unpublished opinion, that the FAA was within its statutory authority in declining to regulate the hours of cargo plane pilots based on its finding that the costs of such action outweighed the benefits.
And in Mingo Logan Mining v. EPA, one of two D.C. Circuit judges present at oral argument cited Michigan v. EPA in suggesting that EPA may have erred in vetoing a permit for a large mountaintop removal mining operation because the agency did not, in considering whether the mine posed “unacceptable” risks, consider the costs to the mine of vetoing the permit.
This is a stunning and problematic turn of events, for two large reasons. The first is that the interpretive principle created in Michigan v. EPA is a made-up canon, with no basis in prior law. The second is that the lower courts have stretched even the broad principle of Michigan too far.
The question presented in Michigan v. EPA was whether EPA “unreasonably refused to consider costs in determining whether it is appropriate to regulate hazardous air pollutants emitted by electric utilities.” In addressing this question, the Supreme Court was working within a kind of mini-jurisprudence – a long series of cases in which the Court had addressed whether agencies were permitted to consider costs in developing health, safety, and environmental regulation. Several of these cases – including a major one in which Justice Scalia held, for a unanimous Court, that EPA must set the Clean Air Act’s National Ambient Air Quality Standards without considering economic costs or technological feasibility – precluded the relevant agency from considering costs. Others permitted but did not require an agency to consider costs. Michigan v. EPA broke with this line of cases by requiring, for the first time, an agency charged with protecting health, safety, or the environment to consider costs in deciding whether to regulate.
There was a good reason for the line of cases from which Michigan v. EPA departed. It is often difficult to determine with any precision the benefits of environmental regulation, and even more difficult to avoid the temptation, when comparing benefits to costs, of being unduly daunted by the costs alone. In fact, the history of section 112 of the Clean Air Act, the statutory provision at issue in Michigan v. EPA, demonstrates how little environmental protection can happen when an agency balances costs against benefits.
In the 20 years after the Clean Air Act required regulation of hazardous air pollutants, EPA managed to regulate only seven such pollutants. A prime reason for this inaction was the agency’s difficulty weighing the costs of such regulation against the benefits. As I have discussed elsewhere, Congress in 1990 overhauled section 112 in the hope of creating a regulatory program for hazardous air pollutants that would actually regulate some hazardous air pollutants. In Michigan v. EPA, the Supreme Court found it unnecessary to delve into the troubled history of cost-sensitive regulation under section 112, sufficing instead with fixation on the bare word “appropriate” and on its brand-new interpretive canon.
The Court could not cite any meaningful legal precedent for its new canon. Justice Scalia conceded that not all statutory references to judgments of appropriateness entail consideration of costs, but, he believed, statutory provisions triggering regulation are distinctive:
Section 7412(n)(1)(A) directs EPA to determine whether “regulation is appropriate and necessary.” (Emphasis added.) Agencies have long treated cost as a centrally relevant factor when deciding whether to regulate. Consideration of cost reflects the understanding that reasonable regulation ordinarily requires paying attention to the advantages and the disadvantages of agency decisions. It also reflects the reality that “too much wasteful expenditure devoted to one problem may well mean considerably fewer resources available to deal effectively with other (perhaps more serious) problems.” Entergy Corp. v. Riverkeeper, Inc., 556 U.S. 208, 233 (2009) (Breyer, J., concurring in part and dissenting in part). Against the backdrop of this established administrative practice, it is unreasonable to read an instruction to an administrative agency to determine whether “regulation is appropriate and necessary” as an invitation to ignore cost.
This is an argument from policy; it is not a legal argument. The only legal material Justice Scalia cites in this key passage is Justice Breyer’s solo opinion in Entergy Corp. v. Riverkeeper. There, Justice Breyer concurred in part with and dissented in part from Justice Scalia’s majority opinion for the Court, holding that EPA reasonably interpreted a provision of the Clean Water Act to allow the agency to weigh costs against benefits in regulating the cooling water intake structures of power plants.
This is a paltry precedent for the idea that an agency must consider costs. It is not strengthened by Justice Scalia’s unsourced, and factually incorrect, claim that agencies “have long treated costs as a centrally relevant factor when deciding whether to regulate.” Sometimes they have done so, sometimes they have not; it depends on the statute under which they are operating. In any event, past agency practice does not help us understand the permissible range of meanings of section 112 of the Clean Air Act.
The broad interpretive principle created in Michigan v. EPA thus has no basis in precedent. What is more, the lower courts that have begun applying the canon have stretched it beyond its limits. Here, I focus on the district court’s decision rejecting the Financial Stability Oversight Council’s designation of MetLife as a systemically important financial institution (SIFI), or one that is “too big to fail.”
In the aftermath of the financial crisis of 2008, Congress passed the Dodd-Frank Act. A central feature of this law is the delegation to the FSOC of authority to designate firms that, if they experienced “material financial distress,” would “pose a threat to the financial stability of the United States.” In MetLife v. FSOC, Judge Collyer rejected the designation of MetLife as a SIFI in part because the FSOC had not considered the costs to MetLife of this designation. She hypothesized that the extra regulatory requirements placed on SIFIs by the Dodd-Frank law might contribute to the firms’ failure. Leaning heavily on Michigan v. EPA, she required the FSOC to consider regulatory costs in the SIFI designation.
The structure of the SIFI designation bears a striking resemblance to the health, safety, and environmental decisions that pose difficulties for cost-sensitive analysis. Both kinds of decisions require judgments in the face of what can be extreme uncertainty. Both kinds of decisions require judgments about events that may have a low probability of occurrence but that may be catastrophic in magnitude. In these decision-making frameworks, weighing the benefits of regulation against the costs will often skew decisions against regulatory intervention. Judge Collyer should have thought twice before doubling down on the anti-regulatory bias of Michigan v. EPA.
The MetLife context, moreover, is legally distinguishable from Michigan. The Dodd-Frank law explicitly instructs the FSOC to consider ten factors in making a SIFI designation. The provision setting out these factors ends with a catch-all instruction to the agency to “consider … any other such risk-related factors that it deems appropriate.” The magic word “appropriate” also appeared, of course, in section 112 of the Clean Air Act, at issue in Michigan. But in Dodd-Frank, the broad term “appropriate” applies only to factors that are “risk-related.”
In holding that the cost to MetLife was “risk-related,” Judge Collyer called upon not the Dodd-Frank statute itself, but the FSOC’s guidance on implementing Dodd-Frank. In brushing aside the FSOC’s understanding of both Dodd-Frank and of the FSOC’s own guidance, Judge Collyer managed to disrespect not just one but two varieties of judicial deference to agency interpretations, one (the famous Chevron doctrine) having to do with interpretations of statutes and the other having to do with interpretations of rules.
The interpretive principle embraced in Michigan v. EPA departed from relevant judicial precedent, unhinged statutory interpretation from careful analysis of statutory context and history, and sprang from the Supreme Court’s policy-laden preferences for inert regulatory agencies. It is bad enough that the Court used the principle in Michigan v. EPA itself. It is more ominous still that the lower courts seem to be celebrating it.
Heinzerling is the Justice William J. Brennan, Jr., Professor of Law at Georgetown University Law Center. This post draws on her new paper, “The Power Canons,” to be published in 2017 in the William & Mary Law Review.