By Yuvraj Tuli, Competitive Enterprise Institutue
Debates over the nondelegation doctrine usually begin with constitutional structure. Article I vests “[a]ll legislative Powers” in Congress, so when Congress gives agencies sweeping discretion to make policy, the usual objection is that it disturbs the Constitution’s allocation of power. Critics of a revived nondelegation doctrine respond that this concern proves too much. The modern administrative state depends on broad delegations, the intelligible principle test has long tolerated them, and stricter enforcement could threaten ordinary governance.
That debate matters, but it is incomplete. The nondelegation doctrine is traditionally understood as a structural separation of powers principle. It can also be understood as serving due process values, especially fair notice. When Congress legislates in highly open-ended terms, it often leaves the real content of legal obligations to agencies, sometimes years later, through guidance rather than formal rulemaking, and in ways that may shift across administrations. The result is a legal order in which citizens and firms often cannot know with confidence what the law requires until the government announces, or enforces, its preferred interpretation.
Viewed that way, nondelegation is not only about institutional boundaries. It is also about whether the people receive clear and prospective notice of the rules they must follow.
Vagueness and nondelegation are usually treated as distinct doctrines. Vagueness is a due process problem because unclear laws fail to provide fair warning and invite arbitrary enforcement. Nondelegation is a separation of powers problem because Congress may not transfer its legislative authority to the executive without meaningful guidance. But the two doctrines respond to closely related constitutional dangers.
In both settings, the central problem is legal indeterminacy. A vague statute leaves people uncertain about what conduct is prohibited. A sweeping delegation leaves regulated parties uncertain about what obligations the law will eventually impose once an agency fills in the details. In one case, indeterminacy is troubling because it deprives the governed of fair warning. In the other, it is troubling because it transfers lawmaking discretion to executive actors. The doctrines remain distinct, but the practical harms often overlap. As Evan Criddle argues, unconstrained delegation is dangerous not merely because it relocates power, but because it enables arbitrary government.
Justice Gorsuch has gestured toward a similar insight. In his dissent in Gundy v. United States, he emphasized that legislation must be made by elected representatives through a public process. That point is often read as purely structural, but it also has a notice dimension. Rules announced clearly by Congress are more likely to be knowable in advance than rules created through broad delegations and later administrative elaboration. Likewise, in Sessions v. Dimaya, Justice Gorsuch explained that vague laws effectively hand lawmaking power to enforcers and judges. That logic arises in the vagueness context, but it closely tracks the concerns that animate nondelegation as well.
Modern regulation shows how broad delegation can undermine fair notice in three ways.
First, broad delegations create temporal uncertainty. Congress enacts a statute using open-ended language and leaves the specifics to an agency. The agency may not issue implementing rules for years. During that period, regulated parties know that legal obligations are coming, but they cannot know with confidence what those obligations will be. The public is thus governed under a framework in which the law’s practical content remains unsettled long after enactment.
Second, agencies often elaborate statutory obligations through guidance, interpretive rules, or enforcement positions rather than notice and comment rulemaking. That magnifies the notice problem because guidance can shift quickly and often lacks the procedural formality that ordinarily accompanies binding legal change. The Supreme Court confronted this concern in Christopher v. SmithKline Beecham Corp., where it declined to defer to the Department of Labor’s interpretation because doing so would have imposed “unfair surprise” on regulated parties. The significance of the case is not just that the Court refused deference. It is that the Court recognized that regulated parties are entitled to fair warning before agencies attach serious legal consequences to a new interpretation.
Third, broad delegations permit administrative oscillation. When Congress uses terms such as “public interest” or “unfair methods of competition,” successive administrations can pursue sharply different regulatory agendas under the same statutory text. That may be defended as a feature of democratic responsiveness, but from the standpoint of the regulated party it means legal obligation turns heavily on electoral turnover and agency preference. At some point, the problem is not simply breadth. It is instability.
To be sure, some discretion is inevitable in any complex regulatory system. Not every delegation is unconstitutional, and agencies cannot be stripped of all policymaking authority. But when Congress fails to specify the basic terms of legal obligation with sufficient clarity, the resulting regime creates the same sort of notice problem that due process doctrine has long treated with suspicion.
Criminal law offers a useful parallel. The rule of lenity requires ambiguity in criminal statutes to be resolved in the defendant’s favor. That principle rests on fair notice and the idea that legislatures, not prosecutors or courts, must make the core policy choices that expose people to punishment. Cass Sunstein has described lenity as a constitutionally inspired nondelegation canon because it prevents lawmaking by enforcement and interpretation where Congress has not spoken clearly.
The same logic has relevance beyond criminal law. If Congress wants to authorize agencies to impose binding obligations backed by substantial penalties, it should bear responsibility for specifying the basic contours of those obligations itself. Civil regulation is not identical to criminal punishment, but the distinction is not absolute. Civil penalties, compliance burdens, licensing consequences, and retroactive liability can be severe. The Court recognized as much in Christopher and in FCC v. Fox Television Stations, Inc., where it reaffirmed that due process requires fair warning when regulated parties face enforcement consequences.
This framing matters because it clarifies what is truly at stake. A purely structural account can sound abstract. The fair notice framing instead highlights the doctrine’s effect on the people subject to regulation. It asks a practical constitutional question: can a person reading the statute, at the time she must act, determine with reasonable confidence what the law requires?
That question better reflects the lived reality of broad delegation. The problem is not only that Congress may have ceded too much authority. It is that the governed may be left without clear, knowable, and prospective legal rules. The nondelegation doctrine can be understood as helping guard against that danger. Both vagueness and nondelegation, in different ways, resist government through indeterminacy. And both reflect a common constitutional intuition: people should know the rules before they are punished.
