Does the Centers for Disease Control and Prevention have the authority to impose a mask mandate on people who travel on planes, trains, and buses? In April a federal district court in Florida offered a clear answer: Absolutely not. The court gave an exceedingly narrow reading to the CDC’s powers under laws enacted by Congress. In the process, it sent an unmistakable signal: some conservative judges will not allow federal agencies to protect public safety and health unless Congress has unambiguously given them the authority to do so.

That signal is ominous. In a period of congressional deadlock, federal agencies often have to take the lead in responding to urgent social problems. During the Covid-19 pandemic, many of the nation’s most important decisions about vaccinations, air travel, masks, social distancing, and more have been made by White House officials, the CDC, the Food and Drug Administration, and the Occupational Safety and Health Administration (OSHA). Policy responses to climate change have also primarily come from the White House, the Environmental Protection Agency (EPA), and other agencies.

Because public policy is often made by administrative agencies, it can shift dramatically from one administration to another. Whether we are speaking about public health, civil rights, clean air, health care, food safety, tobacco, or immigration, fundamental policy judgments might well depend less on Congress than on who wins the presidency.

The administrative state has been with us since the founding. But much of modern government can be traced to the 1930s, when in response to the Great Depression Franklin Roosevelt created a host of new agencies, including the Securities and Exchange Commission, the Federal Communications Commission, and the Social Security Administration. These agencies exercise a great deal of discretion, and they affect the lives of millions of Americans every day. (They also have international influence.) They were born in a period of enthusiasm for technical expertise: Roosevelt and his New Dealers believed in the rule of law, but they did not believe in the rule of courts; they wanted to give authority to specialists.

Many conservatives were skeptical about New Deal agencies. They insisted that the expanded administrative state was a constitutional outrage, because it would give rise to an unelected fourth branch of government operating essentially on its own, without congressional control. In 1941 the legal scholar Roscoe Pound warned of the rise of “absolutism” (i.e., tyranny) and the “Marxian idea of the disappearance of law.” The debate reached a fever pitch in the 1930s and early 1940s, but it quieted down in 1946 with the enactment of the Administrative Procedure Act, a compromise between the New Dealers and their critics. Among other constraints, the act created “notice-and-comment rulemaking,” which allows members of the public to weigh in on proposed regulations.

The debate over the administrative state is no longer quiet. Contemporary critics on the right object vigorously to the power and even the existence of the nation’s regulatory agencies—the EPA, the Department of Education, the Department of Health and Human Services, and more. They ask: Shouldn’t Congress be making the law? Why should our lives be controlled by unelected bureaucrats?

It is true and important that federal agencies cannot act unless Congress has authorized them to do so. Frequently, however, Congress gives them broad discretion. OSHA is allowed to issue safety standards that are “reasonably necessary or appropriate to provide safe or healthful employment and places of employment”; the EPA is allowed to issue national air quality standards that are “requisite to protect the public health,” with “an adequate margin of safety.”

Even when authorizations by Congress are not so broad and open-ended, they often include ambiguous language. The EPA may regulate “air pollutants,” defined to include “any air pollution agent…which is emitted into or otherwise enters the ambient air.” Do greenhouse gas emissions count as air pollutants? The Endangered Species Act says that no person may “take” an endangered or threatened species, and it defines “take” to mean “harass, harm, pursue, hunt, shoot, wound, kill, trap, capture, or collect.” Does modification of a habitat, so that endangered species are injured or killed, count as a “harm”? And by the way, who should answer these questions? Federal judges or the agencies?

These may seem to be technical matters, but they have fundamental importance. If agencies get to interpret such words, their power will be significantly increased. You might be alarmed by that. If independent courts get to interpret Congress’s words, you might feel more comfortable. After all, interpretation of legal terms is quintessential judicial business. But if judges have biases of their own (as they do), and if it turns out that their interpretations would be, in effect, policy judgments, you might think that agencies, with their technical expertise and accountability to the president, should be making those decisions. And if governments need to make decisions quickly and to change their approach as new problems arise—perhaps to deal with emergencies—you might think agencies have a major advantage over the courts.


In 1984 the Supreme Court ruled, in Chevron v. Natural Resources Defense Council, that when the language of statutes enacted by Congress is ambiguous, federal agencies are entitled to interpret it as they see fit, as long as their interpretations are not unreasonable. The Chevron doctrine, as it is often called, has been a terrific gift to Democratic presidents, including Bill Clinton and Barack Obama, who took advantage of it to move federal policy in their preferred directions. But the same has been true of Republican presidents, including George W. Bush and Donald Trump.

The Chevron case pitted President Ronald Reagan’s EPA against one of the nation’s most respected environmental groups. The legal issue arose when the EPA rejected, on economic grounds, an interpretation of the Clean Air Act (1970) that had been made by the same agency under President Jimmy Carter. The contested word in the Clean Air Act was “source.” Carter’s EPA had concluded that each individual smokestack within a plant should count as a pollution source subject to environmental regulation. In a move that was expected to reduce regulatory costs, Reagan’s EPA adopted a different interpretation, ruling that the entire plant, and not an individual smokestack within it, was the source. This meant that if polluters modified smokestacks so that they emitted more pollution, they could do so without worrying about getting permission from federal regulators.

In an opinion written by Justice John Paul Stevens, the Supreme Court upheld the EPA’s decision. It determined that the word “source” is ambiguous and that federal agencies can interpret ambiguous words as they like, unless their interpretation is unreasonable. Here’s the most important passage in Justice Stevens’s opinion:

When a court reviews an agency’s construction of the statute which it administers, it is confronted with two questions. First, always, is the question whether Congress has directly spoken to the precise question at issue. If the intent of Congress is clear, that is the end of the matter; for the court, as well as the agency, must give effect to the unambiguously expressed intent of Congress. If, however, the court determines Congress has not directly addressed the precise question at issue, the court does not simply impose its own construction on the statute, as would be necessary in the absence of an administrative interpretation. Rather, if the statute is silent or ambiguous with respect to the specific issue, the question for the court is whether the agency’s answer is based on a permissible construction of the statute.

This passage has been quoted innumerable times by lower federal courts, and for decades it was understood to give new power to the administrative state and to federal regulators. In defending the decision, Justice Stevens made two points. First, “judges are not experts in the field”; they have far less technical expertise than agencies do. Second, “while agencies are not directly accountable to the people, the Chief Executive is, and it is entirely appropriate for this political branch of the Government to make such policy choices.” Think about Covid-19 and greenhouse gases—or about tobacco regulation, food safety, sex discrimination, and immigration. Technical expertise or political accountability, or both, might suggest that agencies, rather than courts, ought to be given the task of sorting out ambiguities in Congress’s words.

For at least two decades, the Chevron doctrine was widely celebrated on the political right. Deeply mistrustful of left-leaning federal courts, conservatives argued that interpretation of ambiguous terms involves political choices—and that such interpretations should be made by agencies subject to the president, not by unelected judges. Conservatives added that in sorting out ambiguities in complex statutes, agencies often know what they are doing (as courts do not). Justice Antonin Scalia was a particularly enthusiastic supporter of the Chevron doctrine. He wrote in 1989 that “broad delegation to the Executive is the hallmark of the modern administrative state,” and “agency rulemaking powers are the rule rather than, as they once were, the exception.” A major advantage of Chevron is that it permits “needed flexibility, and appropriate political participation, in the administrative process.”

By contrast, many progressives saw Chevron as a disaster and an affront to the very idea of the rule of law. In their view, administrative agencies were subject to the influence of powerful private interests, including wealthy corporations. Independent courts were a necessary safeguard against “capture” of the administrative state. Emphasizing the importance of clean air, workplace safety, and protection against discrimination on the basis of race and sex, progressives argued that the Chevron doctrine was a grave threat to crucial regulatory safeguards.


In the last decade or so, everything has been turned upside down. Many conservatives now disavow Chevron, while progressives enthusiastically defend it. Conservatives are increasingly drawn to the view, going back to the 1930s, that much of the modern administrative state is unconstitutional and that there is something seriously wrong with a situation in which major policy decisions are made by agencies instead of by Congress. Some of these critics would like important provisions of the Occupational Safety and Health Act and the Clean Air Act to be struck down on constitutional grounds.

If the modern administrative state is a constitutional disgrace, the Chevron doctrine, which increases agency power, starts to look impossible to defend. Among conservatives on the Supreme Court today, there’s a lot of anti-Chevron sentiment. Justice Clarence Thomas, who once applied Chevron with zest, now believes that it is inconsistent with the Constitution. In his view, the decision “wrests from Courts the ultimate interpretive authority to ‘say what the law is,’ and hands it over to the Executive.” Justice Neil Gorsuch similarly objects that the Chevron doctrine transfers “the job of saying what the law is from the judiciary to the executive.” Justice Brett Kavanaugh describes the doctrine as “nothing more than a judicially orchestrated shift of power from Congress to the Executive Branch.” Questioning Chevron, Chief Justice John Roberts insists that “the Framers could hardly have envisioned today’s ‘vast and varied federal bureaucracy’ and the authority administrative agencies now hold over our economic, social, and political activities.”

By contrast, Justices Stephen Breyer, Sonia Sotomayor, and Elena Kagan are generally comfortable with Chevron, and they show no interest in wanting to abandon it. When she joins the Court, the newest justice, Ketanji Brown Jackson, may or may not agree with them. But it’s clear that before long, the Supreme Court might well overrule Chevron and insist that when a congressional enactment is ambiguous, the courts, not the agency, get to decide what it means.

In his wise and illuminating new book, The Chevron Doctrine: Its Rise and Fall, and the Future of the Administrative State, Thomas Merrill has two goals. The first is to give an account of Chevron’s history, and in the process he uncovers an assortment of major shifts and turns in its uses and scope. The second is to offer his own preferred resolution of the underlying questions, a kind of Chevron 2.0 that would be, in his view, better suited to the powers of courts and agencies in the modern era.

When Justice Stevens wrote his opinion in Chevron, he meant to solve a knotty problem, but he did not mean to produce a major ruling, or even to make any change in the law. Justice Harry Blackmun’s private papers, which are now public, show that members of the Court found the case to be highly technical and difficult to decide. But Stevens patiently demonstrated that the Reagan administration’s EPA had not violated any provision of the Clean Air Act.

Revealingly, Chevron had hardly any influence on the Supreme Court in its first years. Everything changed after Justice Scalia joined the Court in 1986 and became Chevron’s champion, urging that it inaugurated a new approach for courts to apply in reviewing the interpretations of administrative agencies. Justice Stevens repeatedly disagreed with him; he insisted that Chevron did not make any big change in the law, and that questions of law were for courts, not agencies. By the early 1990s, Scalia had prevailed: whenever an agency’s interpretation of a congressional enactment was at issue, Chevron was widely understood to give the administrative state a lot of room to maneuver. If you worked at a federal agency at the time, Chevron was your best friend.

To be sure, there remained plenty of open questions. Sometimes agencies adopt interpretations that resolve major social and economic debates—for example, by declaring that tobacco is indeed a drug and subject to regulation as such, or that Covid-19 poses a grave danger in the workplace and is subject to emergency regulation. Should courts use the Chevron doctrine in such cases, or should they say that ambiguity in some old law is not enough to justify a new interpretation, and that Congress must explicitly authorize agencies to adopt those new interpretations? Suppose that under President Clinton the Department of Transportation had shifted from a long-standing interpretation under Reagan and George H.W. Bush—concluding, say, that fuel economy standards should be issued with little regard for costs. Should courts be skeptical about that shift just because it was a shift, and perhaps because it disrupted so-called settled expectations, for example the expectations of automobile companies?

In accordance with the Administrative Procedure Act, agencies are ordinarily required to give advance notice of their interpretations and invite public comment. Sometimes they simply announce their interpretations, informing people through what is called an “interpretive rule.” Sometimes they act immediately because of an emergency (such as a pandemic). Does it matter whether agencies have gone through the notice-and-comment process? If they do not, do they lose the benefit of Chevron deference?

Merrill shows that, in answering these and other questions over the decades, courts have transformed the seemingly straightforward Chevron doctrine into something much messier. The Supreme Court has been increasingly supportive of the idea that Chevron should not apply to “major questions,” understood to include situations in which an agency claims large-scale power that it has not exercised before. (The Biden administration’s Covid-19 regulations have been vulnerable for that reason.) In general, courts give less deference to agencies when they reject long-standing interpretations and when they disrupt settled expectations. And if agencies have not used the notice-and-comment process, courts might refuse to apply the Chevron standard—and agencies might well lose in court.

To evaluate these shifts, and the Chevron doctrine in general, Merrill draws attention to four values: the rule of law, including protection of settled expectations; constitutional commitments, including individual rights, separation of powers, and federalism; accountability to the public; and better agency decisions, simply as a matter of good government. We might think, for example, that with some version of Chevron, public policy would be better, because agencies know what they are doing; but we might also fear that if agencies shift too quickly, they will make it hard for people to adapt. With these four values in mind, Merrill argues that Chevron should be reaffirmed, but also rethought.

His Chevron 2.0 would have three steps. In step 1, courts would make their own independent decisions on the general boundaries of an agency’s authority. If the EPA is seeking to treat greenhouse gas emissions as “air pollutants” under the Clean Air Act, judges would have to decide whether greenhouse gases are, in fact, pollutants. In Merrill’s view, judges should be more skeptical if an agency has violated settled expectations (by, for example, suddenly telling companies to make energy-efficient refrigerators). In step 2, courts would decide whether the agency’s interpretation violates any specific provision of a law. If OSHA is telling employers to require their employees to be vaccinated against Covid-19, judges would have to ask whether Congress has forbidden OSHA to do that. Here Merrill would give the agency the benefit of the doubt. In step 3, courts would ask whether the agency has gone through a notice-and-comment process. If it has, Merrill insists that courts should accept the agency’s interpretation, subject to step 1 and step 2. If the agency has not used the notice-and-comment process, courts should be more willing to strike down the administrative interpretation.

Merrill’s reformulation is helpful and reasonable enough, but let’s step back a bit. The sheer intensity of the current debate over Chevron and the ferocity of the conservative opposition seem quite baffling. Everyone agrees that if Congress has been clear, agencies must comply with its instructions. Everyone agrees that even if Congress has given agencies discretion, they must act reasonably. (No matter who the president is, the EPA could not deny that climate change is occurring, and OSHA could not say that Covid-19 is fake news.) The narrow question, then, is whether we want agencies or courts to interpret terms that are genuinely ambiguous, such as “source,” “drug,” and “harm.”

Merrill can be read to lean toward the anti-Chevron camp, but he insists that everything depends on the kind of legal question being asked. If it concerns the meaning of an open-ended term, like “feasible,” he thinks it is fair to say that Congress has granted some interpretive authority to the agency. He agrees that courts should be reluctant to overturn a reasonable agency interpretation of a more specific term, like “source.” That’s sensible, but what’s all the shouting about? Merrill doesn’t really tell us.

Here’s a clue. When Chevron was decided in 1984, during Reagan’s presidency, the federal courts contained numerous Democratic appointees who had been pushing agencies hard on environmental and other issues, and thus to act in ways favored by political progressives. When conservatives defended the Chevron doctrine under Reagan and his successor, George H.W. Bush, and when progressives challenged it, the ideological background was unmistakable: Chevron was terrific for Republican presidents and for their various efforts to mold the administrative state as they saw fit. The conservative opposition to Chevron did not emerge until between 2008 and 2016, during the presidency of Barack Obama.

Can that possibly be a coincidence? Merrill does not mention a revealing and even astonishing fact, which is that according to Justice Samuel Alito, Scalia was rethinking Chevron (and was very possibly poised to repudiate it) not long before his death in 2016. Conservative enthusiasm for Chevron wanes when Democrats control the administrative state—and in such a period, progressive skepticism about Chevron softens.

Since the 1980s some conservatives have been arguing, with increasing confidence, that the constitutional assault on the New Deal and the modern administrative state, most visible in the 1930s, had a lot of force, and that courts should say so. They speak of “the Constitution in Exile,” and they see it as raising fundamental doubts about the powers of modern regulatory agencies. To say the least, that’s radical stuff.

The most important idea here is “the nondelegation doctrine,” last used in 1935, which is said to forbid Congress to grant too much policymaking discretion to administrative agencies. If the nondelegation doctrine were reinvigorated, it would raise serious questions about a host of administrative powers, including those now exercised by the EPA, the Department of Transportation, OSHA, and the Federal Communications Commission.1 If you believe that the modern administrative state is unconstitutional, you might well think that the Chevron doctrine is terrible, because it gives agencies the authority not only to implement the law but also to interpret it—in other words, to decide on the scope of their own authority.

Merrill’s treatment of the rise of Chevron, and its various twists and turns over the decades, is keenly insightful. For those in search of a path forward, his attempted solution is certainly reasonable. The problem is that we cannot understand the Chevron debate, or current disputes about Covid-19, climate change, and much more, without understanding that for the first time in eighty years, the modern administrative state is under an ominous constitutional cloud.

We are already starting to see the implications. Without even referring to Chevron, the Supreme Court ruled in January that OSHA lacked the authority to impose a vaccine mandate on large employers.2 In striking down the CDC’s national mask mandate for travelers on trains and airplanes in April, the Florida district court quickly brushed Chevron aside.3 And in West Virginia v. EPA, the Supreme Court is now considering whether and to what extent the EPA has the authority to regulate greenhouse gas emissions from power plants; a decision is expected in June. There is a good chance that the Court will severely limit the EPA’s authority to combat climate change.

The long-term consequences are clear. It is not exactly news that because of partisan divisions, Congress is often unable to respond to serious national challenges, which means that the burden of problem-solving inevitably falls on federal agencies. Sometimes those agencies face genuine emergencies, which require immediate action; sometimes they face unanticipated problems, which need to be addressed. In either case, Chevron has long given agencies the authority they need. But if Chevron is not quite dead, it is dying. Whether the US government will continue to have the capacity to address pressing national problems will depend on whether the Supreme Court transforms the dark cloud that now looms over the administrative state into some kind of hurricane.

—April 28, 2022