Last month, the Trump Administration announced that it would “reconsider” the Obama Administration’s conclusion that it is “appropriate and necessary” to regulate mercury emissions from power plants. Doing so would allow the Trump Administration to undo the Mercury and Air Toxics Standards (MATS), which regulates power-plant emissions of mercury and other “hazardous air pollutants.”
As others have pointed out, undoing the mercury rules would be both pointless and dangerous. But—and stop me if you’ve heard this one before—there’s another reason the Trump Administration shouldn’t take its preferred course of action: Doing so would be unlawful.
Start with “pointless and dangerous.” Revisiting the MATS rule is pointless because it has already been implemented—and power plants have already spent billions to install pollution-control equipment. (For that reason, many utilities actually want the regulation to remain in place). Somewhat paradoxically, though, undoing the rule would also be dangerous. As Jack Lienke points out, installing emissions technology isn’t the same as using it, and, in the absence of regulation, many power plants may well shut off their expensive new equipment to save some money. That could kill. EPA estimates that the MATS rule saves up to 11,000 lives annually. Even if only some power plants stop using their equipment, lives could be lost.
So, undoing the MATS rule is plainly bad policy. But to understand why revisiting the rule’s “appropriate and necessary” underpinning is also unlawful, we need to take a tour through the lengthy, complicated history surrounding the mercury rule and its statutory background.
In 1990, as part of a sweeping revision to the Clean Air Act, Congress required EPA to regulate emissions of mercury and other air toxins from various industrial sources. Bowing to industry pressure, however, Congress treated power plants differently. Congress instructed EPA to regulate toxic emissions from power plants only if EPA (1) conducted a study, and (2) decided it was “appropriate and necessary” to regulate. Eight years later, EPA completed its study, which showed that mercury emissions from power plants can cause neurological harm to children and fetuses. And in 2000, during the waning days of the Clinton Administration, EPA formally concluded that it was “appropriate and necessary” to regulate mercury emissions from power plants.
But then George W. Bush took office. Rather than issue power-plant mercury regulations (as it was required to do), the Bush Administration attempted to walk back the Clinton EPA’s “appropriate and necessary” finding. It tried to do so by issuing a finding that it was not “appropriate and necessary” to regulate power-plant mercury emissions. Having made that finding, the Bush EPA claimed, it could “delist” power plants as a pollution source that required regulation under the hazardous pollutant program.
Still with me? Here’s the crucial part. In a case captioned New Jersey v. EPA, the D.C. Circuit Court of Appeals unanimously struck down the Bush EPA’s attempt to undo the Clinton Administration’s “appropriate and necessary” finding. Once an “appropriate and necessary” finding has been made, the court concluded, EPA cannot undo it. Instead, if EPA wants to stop regulating mercury emissions from power plants, it has to go through a statutorily defined process for “delisting” pollution sources. And that process is onerous. To “delist” power plants, EPA must make a finding that not a single power plant nationwide emits pollutants that threaten “public health” or cause “adverse environmental effect.” That, it goes without saying, is a reallytough finding to make for power plants.
Then came the Obama Administration. Though the Clinton-era “appropriate and necessary” finding remained in effect, the Obama EPA opted to issue an “appropriate and necessary” finding that “confirmed” the 2000 finding. (It did so, presumably, to defend against legal challenge: there were potential procedural deficiencies in the Clinton EPA’s finding). And with that “confirmatory” finding in place, EPA finally—22 years after the Clean Air Act amendments—issued the MATS rule regulating mercury emissions.
Oh, but the saga still wasn’t over. In 2015, the Supreme Court invalidated the Obama EPA’s “appropriate and necessary” finding, ruling that EPA had failed to take account of the costs of regulation when it made the finding. The rule was thus “remanded” to EPA, which promptly issued a supplemental finding that—even when costs are taken into account—it is “appropriate and necessary” to regulate power plants’ mercury emissions. And that’s where we stand. That finding is final, and remains in effect, as does the substantive MATS rule.
So: what gives? In New Jersey v. EPA, the D.C. Circuit said, very clearly, that EPA can’t walk back an “appropriate and necessary” finding. Once it’s been made, power plants are subject to regulation—unless EPA goes through the statutory “de-listing” process. There are now three “appropriate and necessary” findings on the books (the Clinton finding, the “confirmatory” Obama finding, and the supplemental finding after remand from the Supreme Court). Given that, how can the Trump Administration possibly purport to “reconsider” the appropriate-and-necessary finding?
It’s not like the D.C. Circuit’s decision in New Jersey v. EPA was on shaky footing. The decision was unanimous, without concurrence or dissent, and was joined by Judge Janice Rogers Brown—a conservative judge who was never shy about expressing skepticism about perceived regulatory overreach. What’s more, the court’s holding in New Jersey seems plainly correct. In 1990, Congress effectively told EPA: “You make the call. Go do a study and figure out whether it’s ‘appropriate and necessary’ to regulate power plants under this program.” That seems very much like a one-time command. EPA’s decision whether to regulate, after all, was supposed to be based on a one-time study. Congress, moreover, did not provide any mechanism for periodically conducting new studies, or for periodically re-examining the “appropriate and necessary” finding. (Notably, such requirements are included in many other sections of the Clean Air Act).
And once power plants are subject to regulation under the hazardous pollutant program, Congress provided a very specific process EPA must follow to remove them. “Reconsidering” the appropriateness of mercury regulation simply can’t cut it. To be sure, agencies typically have the discretion to change their minds. But not when doing so would countermand what Congress instructed. As the D.C. Circuit noted in New Jersey, “Congress. . . undoubtedly can limit an agency's discretion to reverse itself.”
Thus, if EPA actually attempts to “reconsider” the appropriate-and-necessary finding, it will be doing so against both a statutory structure and a court of appeals decision that expressly prohibit it from doing so. Perhaps EPA has some reason to believe that New Jersey can be distinguished. Perhaps its statement that it will “reconsider” the appropriate-and-necessary finding is imprecise legalistic bluster. But right now, it sure looks like EPA is preparing to take an action that is not just dangerous, but is knowingly unlawful.
 The Bush Administration did, for what it is worth, try to put into place an alternative regulatory structure for mercury emissions—a cap-and-trade system. That didn’t make much sense. As I’ve written elsewhere, “[a]lthough such programs can work for globally dispersed pollutants like greenhouse gases—where an emissions reduction in Knoxville is as good as one in Kalamazoo—they do not work well for pollutants like mercury, the effects of which are highly localized in areas surrounding the emissions source.”
 The supplemental finding was also challenged in court, but (at the request of the Trump Administration) that litigation has been on hold since last year, as the Administration
 EPA could, for example, presumably stop defending the supplemental “appropriate and necessary” finding in court.