In Michigan v. EPA, the Supreme Court reviewed the Environmental Protection Agency's decision to regulate power plants under section 112 of the Clean Air Act. Section 112 is the provision regulating toxic air pollutants, such as mercury. The question before the Court was whether EPA reasonably interpreted the Clean Air Act to allow EPA to decline to consider costs in deciding whether to regulate power plants under section 112. The Court held that it was not reasonable to interpret the Act in this way. Thus, from the Court's decision, we know that EPA must consider costs in deciding whether to regulate power plants under section 112. There are, however, important questions that remain:
1. We do not yet know what happens to EPA's rule while EPA does the analytical work the Court has required of it. The Supreme Court reversed the judgment of the D.C. Circuit and remanded the case for further proceedings consistent with its opinion. The case will go back to the D.C. Circuit for it to figure out how to address the Supreme Court's ruling. Certainly the case will eventually have to return to EPA; the D.C. Circuit itself will not attempt to undertake the consideration of costs the Supreme Court has ordered. But what happens between the time the case goes back to EPA and the time EPA makes a decision in light of the Supreme Court's ruling? That depends on the D.C. Circuit. The court will need to decide whether to remand or to vacate and remand; that is, whether to simply send the matter back to EPA while leaving the rule in place, or undo the rule in the interim. The D.C. Circuit has lately remanded quite a few agency rules, especially environmental rules, without vacating them. Given the amount of discretion left to the agency by the Supreme Court's decision (see below), and the fact that EPA has previously stated that the rule is justified even in light of its costs, I believe there is a strong case for remand without vacatur.
2. We do not yet know how EPA will or should take costs into account in revisiting the issue of whether to regulate power plants under section 112. The Court left this matter to EPA, with the qualification that the agency's treatment of costs must be, "[a]s always, within the limits of reasonable interpretation." The Court emphasized that it was not holding that the agency must conduct "a formal cost-benefit analysis in which each advantage and disadvantage is assigned a monetary value." Beyond that, the Court gave little hint of the kind of analysis it would approve. At times, it seemed to be looking for a judgment about whether costs were disproportionate to benefits; at other points, it seemed to highlight cost-effectiveness analysis. These are different inquiries, and it will now be up to EPA – at least in the first instance – to decide which of several cost-sensitive frameworks to use. My sense from the Court's opinion is that as long as EPA considers costs in some fashion, whether through formal cost-benefit analysis or something far more qualitative, it will be sufficient.
3. We do not yet know which benefits EPA may consider for purposes of the additional analysis the Court has required. Although EPA conducted a formal cost-benefit analysis of the rule for purposes of White House regulatory review, and that analysis included billions of dollars in "ancillary" benefits due to reductions in air pollutants not covered by the air toxics program, the Court did not decide whether these ancillary benefits may be included in the analysis yet to come. That is, the Court said, "a point we need not address." Several justices seemed skeptical of these benefits at oral argument, and there is some (small but discernible) textual basis for excluding them. A number of commentators have argued that EPA must be able to consider such benefits because they are included in standard economic practice and because an OMB circular on cost-benefit analysis, dating from the George W. Bush administration, admits them in the cost-benefit framework. I don't think economic practice and an OMB circular will decide this question, but I do think it would be quite aggressive for a court to tell EPA which regulatory benefits count.
4. We do not yet know whether, if (when?) EPA again finds regulation of power plants appropriate and necessary under section 112, it must regulate them "on the same terms" as other sources. The Court assumed without deciding that EPA was correct on this point – meaning the question remains open. The question is important because EPA found, following D.C. Circuit case law, that once it had determined that regulating mercury from power plants was appropriate and necessary and thus listed power plants as regulated sources under section 112, it was obligated to regulate all toxic all pollutants from power plants. EPA's regulation of the toxic air pollutants other then mercury is what made EPA's rule have a powerfully positive cost-benefit profile. I think it is unlikely the D.C. Circuit will back away from its line of cases requiring regulation of all toxic air pollutants from sources that are listed and regulated under section 112. I do not know what the Supreme Court would do if faced with this question.
5. What we do know: all nine justices now agree with the proposition that costs matter to regulatory decisions. As Justice Kagan put it in her dissent, unless Congress indicates otherwise, "an agency must take costs into account in some manner before imposing significant regulatory burdens." This idea may sound reasonable, even self-evident, but one might ask: where in law does one find this principle? It doesn't come from the relevant statutes, since it is a default principle that operates across statutes; indeed, I think the principle that has just been embraced by all of the Supreme Court justices is close to the "cost-benefit default principles" championed by Professor Cass Sunstein some years ago. I had thought this default position had been rejected in Justice Scalia's unanimous opinion for the Court in Whitman v. American Trucking Associations – but in Michigan v. EPA, Justice Scalia limited the apparent reach of his own previous opinion.
There is a year and a half left in this administration. From today's decision, we know that EPA has enough leeway under the Clean Air Act to take either a very narrow position on regulation under section 112 or a more protective position on it, thus making any change in administrations highly relevant to the agency's stance. The timeline for EPA's response to the remand includes the time it takes for the D.C. Circuit to give remand instructions to the agency and the time it takes EPA to render a decision under the conditions the Court has stated. There is time for EPA to do the analytical work the Supreme Court has just required, but it is short.