EPA is rolling back protections with methodology no respectable economist would endorse

In some ways, federal agencies are like people: When reality does not serve our purposes, we make something up and hope nobody notices. Seeking to boost museum revenue, P.T. Barnum once displayed the remains of a “mermaid,” constructed by sewing together a monkey and a fish. Similarly, when agencies cannot find justifications for their decisions, they sometimes invent them. The results can be both bizarre and dangerous.

Case in point: The Environmental Protection Agency recently proposed that regulating hazardous emissions from power plants is not “appropriate and necessary,” reversing Obama-era findings and exposing current emissions standards to legal challenges. To arrive at this conclusion, the agency has invented a new, unnatural way of evaluating regulatory cost that finds no support in the economics literature or in the regulatory practices of prior administrations of either political party. If the agency succeeds using its novel methodology, it will likely deploy the same strange method in other rulemakings, leading to large numbers of premature deaths and other significant adverse health consequences for the American people. 

{mosads}The specific regulation under threat is the Mercury and Air Toxics Standards, or “MATS,” an Obama-era regulation that limits the emissions of hazardous pollutants like neurotoxic mercury, carcinogenic arsenic, and other toxins from power plants running on fossil fuels. The Clean Air Act’s section 112 gives EPA the authority to regulate hazardous emissions from these plants, so long as the agency determines that such regulation is “appropriate and necessary.” The Supreme Court held in 2015 that, in making this determination, EPA must consider the costs associated with regulation. The court left it to EPA, however, to decide how to consider cost.

Fortunately, EPA and other federal agencies are accustomed to taking cost into account, and the Obama-era EPA turned to two established methodologies for doing so. Under the first “reasonableness” method, EPA examined the expected costs to regulated firms to ensure that these costs were reasonable. The EPA found that the expected annual cost of the regulation represented only 2 to 4 percent of the power sector’s annual revenues from electricity sales, that the cost of equipment necessary for compliance amounted to only 3 percent of what industry usually spent in capital expenditures, and that regulation would only minimally affect the retail price of electricity. So the agency concluded that the costs of the mercury emissions limits were reasonable. 

In addition, the Obama-era EPA determined that the regulation of power plant emissions was “appropriate and necessary” using another methodology: cost-benefit analysis, a monetized comparison of a regulation’s pros and cons. This was an understandable choice. Presidents of both parties since Ronald Reagan have made agencies conduct cost-benefit analyses to justify significant federal regulations. The White House’s Office of Management and Budget has published a special guide, “Circular A-4,” that lays out the rules agencies must follow when conducting such analyses.

Again, the regulation passed with flying colors. EPA predicted that its emission limits on power plants would save up to 11,000 lives per year, yielding $37 billion to $90 billion in annual benefits that dwarfed an annual $9.6 billion in costs. In light of its two forms of cost consideration, EPA deemed regulating power plants “appropriate and necessary.” Since then, power plants have complied with the mercury limits at costs even lower than those predicted. 

But EPA now argues that the Clean Air Act and Supreme Court prohibit using established forms of cost consideration to evaluate whether it is “appropriate and necessary” to regulate mercury emissions. Instead, the agency claims that the law calls for a totally new methodology. EPA then stitches together the parts of cost-benefit analysis it likes and saws off the pieces it does not. The result is a weird, Frankenstein methodology that no respectable economist would endorse.

As an initial matter, the agency’s claim that Congress and the Supreme Court banned established forms of cost consideration is inventive, but flatly wrong. The statute is silent about what makes regulation “appropriate and necessary,” and the court said it would be “up to the agency…to decide how to account for cost,” so long as the methodology was reasonable.

EPA seems aware that it has created something odd, so before revealing its new methodology, the agency begins with the familiar. Invoking the language of cost-benefit analysis, EPA says it intends to “directly compare the cost [of regulating power plants] with the benefits.” So far, so good. But after referencing cost-benefit analysis, the agency makes clear its methodology is a totally different beast.

Specifically, the agency takes a warped approach to benefits. By reducing over-all emissions from power plants, the mercury standards have produced benefits of two kinds. First, the regulation reduces emissions of hazardous air pollutants expressly targeted by the governing statutory provision, creating so-called “direct benefits,” which are significant. The regulation eliminates 90 percent of industry’s mercury emissions and also reduces acid gases and hazardous metals like antimony, arsenic, beryllium, cadmium, chromium, cobalt, manganese, nickel, lead and selenium.

EPA eviscerates the regulation’s direct benefits by focusing only on a small subset of mercury reduction benefits associated with subsistence fishing communities. In focusing on this small class, the agency ignores many direct benefits that the agency knows to exist, but cannot exactly quantify because of data limitations. EPA acknowledges that these unquantified benefits from limiting hazardous air pollutants are “substantial and important,” but, without any explanation, goes on to say that they are insufficient to make up for the regulation’s cost.

In addition to creating direct benefits by curbing hazardous air pollutants, the steps plant owners take to reduce mercury emissions — installing new technologies and switching to cleaner fuels — necessarily reduce other forms of pollution as well, creating so-called “co-benefits,” or salutary indirect effects. For example, the regulation cuts back emissions of particulate matter that can cause death, heart problems and respiratory risks. 

{mossecondads}Including co-benefits is an established feature of cost-benefit analysis. Circular A-4 says conclusively that agencies should “look beyond the direct benefits and direct costs” of a regulation and include both co-benefits and indirect costs when evaluating a regulation 

But now EPA claims that the regulation’s co-benefits, which include saving thousands of lives annually, has no bearing on whether regulation is appropriate and necessary. Direct benefits must be counted, says EPA, but co-benefits must be excluded. This approach makes as much sense as counting benefits that accrue on Monday and Tuesday, but ignoring benefits that accrue any other day of the week.

It gets weirder. EPA’s refusal to consider indirect benefits is especially problematic because courts have made clear that agencies must consider indirect costs. This approach would systematically distort cost analyses and stand in the way of desirable regulation.

Although EPA invokes the language of the applicable Clean Air Act provision to justify its new creation, the agency’s actions in other proceedings strongly suggest that EPA hopes this creation will develop a life of its own, spawning sequels and spin-offs beyond this particular statutory section. Fortunately, this seems unlikely. When courts review agency decisions, they strike down those made in an “arbitrary and capricious” fashion. It is difficult to imagine a more arbitrary and capricious methodology than a rule under which EPA must take into account the indirect consequences of regulation if they are negative but must ignore them if they are positive. Such an irrational methodology should not be long for this world.

Natalie Jacewicz is a legal fellow at the Institute for Policy Integrity, who has authored scholarship and public comments on environmental and regulatory issues.

Richard L. Revesz is the Lawrence King Professor of Law and Dean Emeritus at NYU Law School, where he directs the Institute for Policy Integrity.