Federalist Society members tend to applaud the Supreme Court's Chevron doctrine,1 because it seeks to restrict the lawmaking powers of unelected federal courts. When Congress directs an agency to administer a federal statute, Chevron says that reviewing courts, as agents of Congress, must enforce any unambiguous statutory directives. But, Chevron continues, we presume in these circumstances that Congress intended the agency, not the reviewing courts, to resolve ambiguities in the statute, provided they do so reasonably. Moreover, according to Chevron, the primary reason for giving agencies this ambiguity-resolving authority is not so much because of their expertise. Rather, it is because agencies must answer to the President, who is in turn elected by all the people. Thus, the Chevron doctrine rests on a fundamental commitment to confining lawmaking power as much as possible to the democratic branches of government — the Congress and the executive branch agencies— as opposed to the unelected federal courts.

Unfortunately, Chevron's restrictions on federal courts make it more difficult to enforce a second and equally fundamental constitutional principle: that an agency's power to make law is strictly limited to the authority delegated to it by Congress. The problem is sometimes discussed in terms of whether courts must give Chevron deference to agency interpretations of ambiguities regarding the agency's own jurisdiction.2 Suppose Congress in delegating authority to an agency under a particular statute does not describe the agency's jurisdiction in air-tight terms, but rather uses vague language or language that contains significant gaps. If reviewing courts must defer to agency interpretations of such ambiguities, then what is to stop the agency from exploiting these uncertainties to try to grab more power than Congress intended to give it?

In other words, Chevron places lawmaking by agencies on the same plane as lawmaking by Congress, and on grounds of democratic theory prefers both to lawmaking by unelected courts. But the Constitution rather clearly puts lawmaking by Congress on a higher plane than lawmaking by agencies. If we are to preserve this second principle— the basic principle of the nondelegation doctrine— then we may need to give the courts greater leeway than they have under the Chevron doctrine to police agency interpretations of the statutes they are charged with administering.

A classic illustration of this dilemma is provided by the Fourth Circuit's recent decision striking down the Clinton Administration's ambitious program to put the Food and Drug Administration (FDA) in charge of tobacco advertising and marketing.3 The Supreme Court has agreed to hear the case, with argument likely occurring in the fall. The Clinton Administration's anti-tobacco initiative is second only to health care reform as an example of its on-again, off-again commitment to big government solutions to social problems. Conceived by Dr. David Kessler, former Commissioner of the FDA, these regulations go far toward federalizing the marketing of cigarettes. For example, print advertising for cigarettes is restricted to black-on-white tombstone advertising in any publication with more than 15 percent youth readership; tobacco vending machines are banned; tobacco billboards are prohibited within one thousand feet of schools and playgrounds; tobacco brand names cannot be placed on merchandise; promotional gifts for purchasing tobacco products are banned; and brand-name sponsorship of concerts and sporting events is prohibited.

Not surprisingly, the tobacco companies went to court to challenge FDA's legal authority to issue these far-reaching regulations. FDA argued that it had the power to issue the regulations because tobacco falls within the definitions of "drug" and "device" as set forth in the Food, Drug, and Cosmetic Act of 1938.4 Those definitions broadly define drugs and devices as "articles (other than food) intended to affect the structure or function of the body."5 FDA maintained that tobacco products affect the structure or function of the body by causing addiction, and that the tobacco companies "intended" this result because they knew or should have known that tobacco is addictive.

The tobacco companies responded that FDA was entering new territory it has never previously sought to regulate: the marketing of a consumer product that is not sold as a medicinal drug, a cosmetic, or a food product. Indeed, the companies argued that until Commissioner Kessler came along, no one in Congress, the FDA, or the courts ever suggested that Congress had given FDA authority to regulate the marketing of tobacco.

The legal battle over FDA's authority in the Fourth Circuit was framed by the Chevron decision. More specifically, it was fought out in terms of whether a court should identify statutory "ambiguities" subject to Chevron deference by looking to the objective meaning of the enacted text, or to the subjective intentions of Congress. The FDA argued for objective meaning: tobacco products fall within the literal definition of drugs and devices, or at most there is an ambiguity about the meaning of the word "intended," and FDA is entitled to Chevron deference in interpreting this term. The tobacco industry argued for subjective intentions: the legislative history of the 1938 Act and history of post-enactment interpretations by FDA, the courts, and other statutes all show that however broad the language of the Act, Congress never intended to give FDA authority to regulate tobacco.

The Fourth Circuit, by a 2-1 vote, adopted the subjective intentions construction of Chevron, concluding that Congress did not intend to confer authority over tobacco on FDA, and accordingly struck down the regulations.6 The majority made clear that it was impelled to adopt the subjective intentions construction of Chevron because only this approach, at least in this particular case, allowed the court to engage in the independent reading of the statute necessary to confine the agency to the scope of power delegated to it by Congress.7

I doubt that there is any necessary connection between the subjective intentions conception of the Chevron doctrine and the ability of courts to keep agency's within their established jurisdiction. In some cases the subjective intentions approach will provide a greater check on agency aggrandizement; in other cases the objective meaning approach would serve this end better; in still others, it will make no difference. However, what the tobacco case reveals, I think, is that the Chevron doctrine, by its very nature, impairs the ability of courts to police efforts by agencies to stray beyond the sphere assigned to them by Congress. Once agencies are given primary authority to fill gaps and ambiguities — however those are defined — and this includes gaps and ambiguities in the outer perimeter of the agency's authority, then the courts will be at a relative disadvantage in interpreting the law so as to confine agency's to their designated jurisdiction. Chevron deference in these circumstances is at war with the nondelegation doctrine.

Finding a solution to this problem is not an easy task. The root of the difficulty is that "agency jurisdiction" is a difficult if not impossible concept to define in the abstract. As Justice Scalia has observed, "there is no discernible line between an agency's exceeding its authority and an agency's exceeding authorized application of its authority. To exceed authorized application is to exceed authority."8 Thus, where the outer perimeter of an agency's authority is not clearly delineated by statute, it is doubtful that the line can be deduced very often from what Congress has said.

Nevertheless, while it may be impossible to deduce boundary lines from first principles, it may still be possible to identify boundaries that exist based on historical practice and understanding. The tobacco case presents a controversy in which virtually everyone recognizes that, given the history of FDA enforcement since 1938, the agency is proposing to expand its jurisdiction. Perhaps such cases can only be identified by undertaking a particularized analysis of post-enactment history. Such post-enactment history is of course disfavored as a means of identifying the "intent" of the enacting Congress. But it may be the best we can do in identifying the "settled equilibriums" of the administrative state, and hence in highlighting those agency proposals that most directly threaten the values associated with the nondelegation doctrine.

If we can identify, based on post-enactment history, existing boundary lines that mark the outer limit of agency authority, what should we do with such an understanding? I would not endorse a rigid rule to the effect that settled boundaries may never be breached by agencies without an affirmative act of Congress. That would represent too much of an iron grip for the status quo. But it might be appropriate to recognize an exception to Chevron in such circumstances. Thus, where history has given us a settled understanding of the outer limits of an agency's authority, and an agency proposes to breach that understanding by expanding the sphere of its regulatory activity further, courts should review the issue of the agency's statutory authority to undertake this additional step with less deference to the agency than required under the Chevron doctrine. Perhaps Skidmore deference9 would be appropriate in these circumstances.

Now that the Supreme Court has granted the government's petition for certiorari in Brown & Williamson, it has a prime opportunity to reconsider the balance between Chevron and the nondelegation doctrine. Whether it will use this opportunity to good effect remains to be seen. The Court's recent decisions on Chevron have not been especially inspired.

More fundamentally, the notion that tobacco policy, like other basic policy choices, should be made by the legislative branch of government seems like an increasingly quaint ideal. Even if the Supreme Court stops the FDA from legislating without the approval of Congress, it is very doubtful that anyone will block the transfer of state legislative authority to state court judges, who will soon start administering tobacco advertising restrictions contained in consent decrees negotiated between the tobacco companies, state attorneys general, and a bunch of rich plaintiffs lawyers. If legislators want to duck an issue badly enough, it is very difficult to force them to stand up and be counted. But the relationship between the tobacco settlement and the nondelegation doctrine is, as they say, a story for another day.


   1. Chevron U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984).
   2. The question whether there should be a "jurisdictional" exception to the Chevron doctrine was debated between Justice Scalia and Justice Brennan in separate opinions in Mississippi Power & Light Co v. Mississippi ex rel. Moore, 487 U.S. 354 (1988). See id. at 381-82 (Scalia, J. concurring in judgment); id. at 386-87 (Brennan, J. dissenting). The full Court has never expressly addressed the issue.
   3. Brown & Williamson Tobacco v. Food & Drug Administration, 153 F.3d 155 (4th Cir. 1998), cert. filed, 67 U.S.L.W. 3490 (Jan. 19, 1999).
   4. FDA's authority under that Act extends to food, drugs, devices, and cosmetics.
   5. 21 U.S.C. §321(g)(1)(C) ("drug"); id. § 321(h)(3) ("device").
   6. The Fourth Circuit majority tried to dress up its reliance on legislative history and post-enactment history by making what it called a textual argument. The court noted that if tobacco products are drugs and devices, then it would seem to follow that FDA would have no choice under various provisions of the Act but to ban all sales of tobacco products, because cigarettes cannot be shown to be "safe and effective" for their intended use. FDA was not proposing to ban all sales of tobacco products, however, because this would obviously be politically infeasible. Hence FDA's premise, that tobacco is a drug and a device, must be flawed. This argument, however, in the end boils down to more evidence of Congress's subjective intentions. The Fourth Circuit majority could not say that calling tobacco a drug or device (and hence committing potentially to banning its use) was absurd on its face. Its point rather was that the FDA's premise might lead to results that Congress would find politically unacceptable, hence providing further inferential support that Congress did not intend this premise, notwithstanding the words it choose to use to define drugs and devices.
   7. See 153 F.3d at 161-62. ("Applying the principles set forth by the Supreme Court in Chevron, we examine whether Congress intended to give the FDA jurisdiction over tobacco products....[We] note that ascertaining congressional intent is of particular importance where, as here, an agency is attempting to expand the scope of its jurisdiction.").
   8. Mississippi Power & Light, 487 U.S. at 354 (Scalia, J. concurring in judgment).
   9. See Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944) (stating that the weight to be given to an agency interpretation of a statute rendered in an advisory capacity "will depend upon the thoroughness evident in its consideration, the validity of its reasoning, its consistency with earlier and later pronouncements, and all those factors which give it power to persuade, if lacking power to control").

Thomas W. Merrill is the John Paul Stevens Professor of Law at Northwestern University. Professor Merrill is chairman of the Judicial Review Subcommittee of the Administrative Law Practice Group.