“The curious task of economics,” Hayek once wrote in The Fatal Conceit, “is to demonstrate to men how little they really know about what they imagine they can design” (76).That line could just as easily serve as an epigraph for administrative law. Agencies are routinely tasked with designing solutions to problems—market failures, externalities, the consequences of new technology—under conditions of genuine uncertainty, yet much of rulemaking doctrine still proceeds as though a sufficiently diligent agency could, in principle, get the rule right the first time.

In a forthcoming article, we argue that this assumption should be retired. We develop epistemic humility—the candid acknowledgment that regulators’ knowledge is necessarily partial, fallible, and provisional—into a design principle for rulemaking: one that should shape not whether agencies act, but how they structure the rules they adopt.

We trace epistemic humility’s deep intellectual roots from Socrates and Aquinas through Hume, Burke, and Mill. But the version most relevant to rulemaking comes from Hayek, who identified not one but two distinct knowledge problems facing any regulator.

The first is dispersion. As Hayek explained in “The Use of Knowledge in Society,” “the knowledge of the circumstances of which we must make use never exists in concentrated or integrated form, but solely as the dispersed bits of incomplete and frequently contradictory knowledge which all the separate individuals possess” (522). Notice-and-comment rulemaking under APA § 553 was designed, in part, to address exactly this problem by inviting affected parties to supply the information any single agency could never assemble on its own (one of us has written about this at greater length in connection with a separate Notice & Comment post).

But § 553 leaves Hayek’s second, less-appreciated problem untouched: Even an agency that successfully aggregates all the dispersed information available at the time of rulemaking still cannot deliberately engineer an optimal rule because some of the knowledge needed to evaluate that rule—how the market will adapt, what new technologies will emerge, how regulated parties will respond strategically—simply does not exist yet. Hayek called the belief that planners could overcome this problem “constructivist rationalism.” He regarded it as a category error: treating complex, evolving social orders as engineering problems that admit of a single correct design.

Acknowledging this second problem is not a counsel of paralysis. Following Aquinas’ distinction between true and false humility, we stress that epistemic humility does not mean regulatory timidity, minimalism, or an agency that waits for certainty before acting—certainty that, on Hayek’s account, will never arrive. Instead, an epistemically humble agency acts but structures its rules to remain provisional: built for learning, correction, and revision rather than permanence. The article surveys the existing toolkit for doing this—retrospective review, sunset provisions, the real-options approach to rule adoption, regulatory sandboxes and pilot programs, contingent-effectiveness clauses, costly opt-outs, and stock-market-based rulemaking—and organizes them into three functional categories: ex post evaluation mechanisms, experimental approaches, and flexibility-preserving design features.

We then make a claim that may be the article’s most distinctive contribution: epistemic humility is not merely a hedge against error, but a source of democratic legitimacy. Agency rulemaking is, after all, conducted by unelected officials, often without the full apparatus of legislative deliberation. Drawing on David Estlund’s epistemic proceduralism—legitimacy requires a process that is “epistemically the best among those that are better than random”—and Elizabeth Anderson’s Deweyan account of democracy as an experimental mechanism for “pooling widely distributed information,” we argue that adaptive design features do double duty. For example, the same sunset clause or pilot program that protects against a regulator’s overconfidence also institutionalizes the kind of ongoing responsiveness to evidence and public input that these accounts treat as the hallmark of legitimate governance. Legitimacy, on this view, is not a property an agency earns by getting the rule right on the first try; it rests on demonstrated openness to being wrong.

We conclude with American federalism, which we treat as a kind of epistemic humility operating at the systemic level: each state’s public utility commission is, in effect, running its own experiment, and the federal system as a whole can learn from the comparison. We illustrate this point through electric vehicle charging infrastructure regulation in California, Florida, and Texas—three large EV markets that have adopted strikingly different answers to the question of whether utilities or private firms should own and operate the charging networks. California has settled into a hybrid “make-ready” model that splits the infrastructure and operating roles between utilities and private operators; Florida initially let utilities build ratepayer-funded charging networks and then reversed course by statute in 2024 once that approach raised competitive concerns; Texas has gone furthest toward private ordering, generally barring utilities from owning public chargers at all. None of these states assumed ex ante that it had the right answer, and the resulting variation is itself a source of knowledge—for these states and for others watching from the sidelines.

Ultimately, the article calls on regulators, commenters, and courts to give up on a particular fantasy: that good rulemaking means identifying the single correct rule and locking it in. We think the more honest—and more legitimate—picture treats regulation as an iterative, fallible, and improvable enterprise, and we try to show what that looks like in practice, from sunset clauses to state public utility commissions.

L. Lynne Kiesling is Director of the Institute for Regulatory Law & Economics at Northwestern University and External Faculty at the Santa Fe Institute. Yoon-Ho Alex Lee is the Howard Friedman ’64 JD Professor of Law and Director of the Center on Law, Business, and Economics at Northwestern Pritzker School of Law. This post—prepared with assistance from Claude and revised by the authors—draws on our forthcoming article, “Epistemic Humility as a Principle of Adaptive Regulation.”