Can agencies rewrite statutes? In general, the obvious answer is no; this post, however, defends EPA’s effort to do just that in regulating greenhouse gas emissions under the Clean Air Act.
The issue on which the Supreme Court granted cert. in Utility Air Regulatory Group v. EPA (UARG) (consolidated with five other cases, all arising out of the same decision by the D.C. Circuit) is quite narrow. Acting pursuant to Clean Air Act authority, the Environmental Protection Agency recently issued new corporate average fuel economy standards, or CAFE standards, to regulate greenhouse gas emissions from new cars and light trucks. The agency maintains that the issuance of this rule automatically triggered provisions of the Act that require control of greenhouse gas emissions from major stationary sources like coal-fired power plants. The industry petitioners, on the other hand, argue that Congress did not intend the Act’s stationary sources provisions to extend to greenhouse gases like carbon dioxide, which are emitted from a much wider array of sources (including from me, as I write this) in much larger quantities than other familiar air pollutants like carbon monoxide and lead. The D.C. Circuit sided with the agency; the industry sought cert.
Lurking behind this arcane issue of statutory triggers, though, is a broader question about agency authority: how much deference should agencies receive as they remold complex, technical, and aging statutory and regulatory regimes to fit complex, technical, and new policy challenges? In my view, deference doctrines only gain in importance with statutory age. Congress drafts legislation at a particular point in time, with specific problems in mind. Inevitably, therefore, statutes grow increasingly ill-fitting and unwieldy as tools to address new and newly recognized problems. Sometimes, the ill-fitting statute can simply be ignored, but in other cases – such as this one – the statutory strictures remain in place, either inhibiting necessary action or requiring regulators or regulated entities to do something senseless. Ideally, of course, Congress would solve these kinds of problems by amending the original statute. But when no legislative solution is forthcoming, the expert agency to which the statute delegates regulatory authority remains the entity best suited to massaging the existing statute and its implementing regulations into functional shape.
To understand the agency authority question lurking in the greenhouse gas cases, you need some – simplified! – background on the history and intricacies of the Act. Congress passed the Clean Air Act in 1970. As amended several times since then, the Act gives both EPA and the states a role in reducing air pollution. Broadly speaking, EPA sets air quality standards, and the states determine how best to achieve and maintain those standards. Superimposed on that broad structure are a few federally administered programs to control emissions from specific categories of polluters. As relevant here: the Act directs EPA to regulate the automobile tailpipe emissions of any “air pollution agent or combination of such agents” that “may reasonably be anticipated to endanger public health or welfare”; and the Act separately charges EPA with administering permitting and pollution control programs for large new stationary sources that will emit specific amounts of “any air pollutant” already subject to regulation under the Act, and for large existing sources that are undertaking major modifications that will increase their emissions of “any air pollutant” subject to regulation under the Act.
When Congress devised this complex scheme, climate science was in its infancy. The science is still evolving, but at this point, virtually all respected scientists consider the impacts of greenhouse gas emissions on global temperatures, precipitation patterns, sea levels, and storm intensity as among the most dire public health and welfare (not to mention environmental) threats facing this country and the world.
Fortunately, the Clean Air Act’s drafters recognized the need to grant EPA regulatory flexibility to respond to new and newly recognized risks to ourselves and our way of life. As the Supreme Court has twice observed, first in Massachusetts v. EPA and then in Connecticut v. American Electric Power, that regulatory flexibility readily extends to the increasingly certain and serious risks of climate change. For a Court as committed to plain language and precedent as this Court professes to be, therefore, there can be no real question that the Clean Air Act grants EPA ample authority to regulate greenhouse gas emissions.
How, then, does the agency find itself in court (again) to defend an exercise of such authority? The problem in the greenhouse gases cases arose because the Act has aged, and its language and regulatory structures don’t comfortably fit the modern problem of greenhouse gas emissions and climate change. Under the Act, EPA’s (statutorily required) regulation of greenhouse gas emissions from tailpipes automatically triggered permitting and pollution control requirements for all “major” stationary sources of the same pollutant. But what is a “major” source? According to the express language of the Act, a “major” stationary source is any source that emits more than 100 (or, for some sources, 250) tons per year of the regulated pollutant. This statutory triggering threshold works well for the trace pollutants that Congress had in mind when it drafted the Act – only the most significant stationary sources, like large coal-fired power plants, emit more than the threshold amount of lead or carbon monoxide and thus find themselves subject to permitting and pollution control requirements. But the most common greenhouse gas, carbon dioxide, is different in kind. It is not a trace pollutant but instead the inevitable end product of almost all combustion. As a result, applying the statutory threshold to stationary sources of greenhouse gases would bring millions of additional stationary sources under EPA’s regulatory umbrella (including millions of universities, hospitals, and other mid-size businesses), impose billions of additional dollars in permitting costs, and result in years or decades of permitting delay.
As a temporary solution to this problem, EPA promulgated something it dubiously dubbed a “Tailoring” Rule – a rule that purports to elevate by several orders of magnitude the triggering threshold for stationary sources of greenhouse gases, in order to continue to target only the most egregious sources. The agency has emphasized that this is not a permanent solution but a temporary fix, a “phase-in” of the permitting requirements. Indeed, EPA has pledged to conduct future rulemakings to address greenhouse gas emissions from sources whose emissions fall between the statutory triggering threshold and the Tailoring Rule’s elevated threshold.
In the interim, though, the Tailoring Rule gave industry petitioners an oversized target at which to shoot. The Act spells out the number “one hundred tons” in black and white. The intent of Congress is clear, petitioners have argued, EPA has no room to deviate from that clear intent.
In fact, though, the intent of Congress is not as clear as the industry petitioners would have it. Or rather, four different and irreconcilable congressional intents are evident on the face of the Act: (1) in the words of Massachusetts v. EPA, Congress intentionally defined the term “air pollutant” quite broadly, to “confer” on EPA the “regulatory flexibility” to respond to “changing circumstances and scientific developments” that would otherwise render the Clean Air Act “obsolete”; (2) relatedly, Congress plainly intended the Act’s mandate to extend to pollutants that affect “weather . . . and climate”; (3) Congress then required EPA to regulate tailpipe emissions of any compound that meets the (broad) air pollution definition and also affects weather and climate; and, finally, (4) Congress expressly provided for the automatic triggering of permitting and pollution control requirements for stationary sources that emit more than 100 (or 250) tons per year of any pollutant already regulated under the tailpipe emissions program (or any other Clean Air Act program).
These four congressional intents are clear, but that cannot be, in the words of Chevron v. NRDC, “the end of the matter.” In the context of greenhouse gases, giving simultaneous effect to all four intents would lead to a prohibitively expensive bureaucratic nightmare. Thus, someone – Congress, the agency, or the courts – must do the difficult work of reconciling the disparate intents of this aging statute with the new and unexpected challenges posed by greenhouse gas emissions. Congress shows no sign of acting (on this or any other pressing policy problem), so take that potential solution off the table for the moment. The other two proposed solutions are either (1) to read the automatic triggering provisions out of the Act, as the industry petitioners ask the Supreme Court to do, or (2) to elevate the numeric triggering threshold to a more sensible level, as EPA attempted to do in the challenged Tailoring Rule.
Neither is a perfect solution; both would do some measure of violence to the statute and to congressional intent (though if EPA does conduct future rulemakings to lower the triggering threshold back to the statutory level, then its phase-in will ultimately conform to congressional intent). In my view, the only way to decide between these imperfect solutions is to focus not on the action but on the actor. The first solution would require the Court to rewrite the statute to give the best possible effect to conflicting congressional intents; the second would rely on the expert agency to do the same. Given the intricacies of the Clean Air Act and the complexities of climate science, my money is with EPA.