How the Supreme Court Created a Constitutional Case Against the Administrative State

One of the Supreme Court’s first cases in October threatens to unsettle the power of federal agencies to implement laws through regulation.

In the first week of the Supreme Court's upcoming term, the justices will hear argument in a case asserting that a federal statute is unconstitutional because it gives an official in the executive branch insufficiently precise instructions about how to implement the law. Although the case has attracted little attention from progressive groups so far – not one amicus brief was filed in support of the federal law – there is reason to believe that conservative justices may view the case as an opportunity to revitalize a long-moribund constitutional doctrine and, in the process, rework the balance of power in modern government.

Watch Lisa Heinzerling's pre-argument analysis of Gundy v. United States

The Supreme Court has long interpreted the Constitution to require Congress to provide intelligible instructions when it delegates authority to the executive branch to implement legislation. This principle, known as the doctrine of "nondelegation," aims to ensure that Congress does not cede its legislative power to the executive. Only twice in history, however, has the Supreme Court invalidated a federal statute on nondelegation grounds – both times in 1935, at the height of the Court's resistance to President Roosevelt's New Deal legislation.

Since then, the Court has not rejected a single federal statute because the instructions given by Congress to the executive were insufficiently precise. As Justice Scalia once wrote, given the "multifarious" and "highly political" nature of the relevant inquiry, the Court has "almost never felt qualified to second-guess Congress regarding the permissible degree of policy judgment that can be left to those executing or applying that law." A large worry is that unelected judges applying a revitalized nondelegation doctrine would be left to follow their own, personal impulses in drawing the line separating acceptable from unacceptable delegations.

Nevertheless, four of the Court's current justices have in recent years expressed constitutional anxieties about Congress's broad delegations of power to administrative agencies. Justices Alito and Thomas have complained that the Court has gone too far in condoning broad delegations of power to administrative agencies. Chief Justice Roberts has openly lamented the power of administrative agencies and has questioned the practice of deferring to agencies' interpretive judgments. Likewise, as a lower-court judge, Justice Gorsuch also expressed dissatisfaction with the Court's policy of "Chevron" deference, citing the constitutional constraint on delegation as a reason for his uneasiness.

It seems fair to surmise that the Court's conservative justices have been itching for a fight on nondelegation. They've now found one, and it may prove to be just what they've been looking for.

In Gundy v. United States, set to be argued the first week of October, a convicted sex offender argues that the Sex Offender Registration and Notification Act (SORNA) violates the nondelegation doctrine by giving the Attorney General the authority to decide whether the law's registration requirements apply to offenders convicted before the statute was enacted. Failure to register as required under SORNA is a criminal offense, carrying a penalty of up to ten years in prison.

A sticking point during legislative debates was how the law should deal with pre-Act offenders. Congress defined a "sex offender" as "an individual who was convicted of a sex offense," thus apparently including within the Act's requirements individuals who were convicted either before or after the law was passed. Yet Congress also provided that "[t]he Attorney General shall have the authority to specify the applicability of the requirements" of the statute to pre-Act offenders.

These provisions together raised a question: do the Act's registration requirements apply to pre-Act offenders by the force of the Act itself, or do they apply only once the Attorney General specifies that they apply? Under the first interpretation, the statute would present no nondelegation problem, since Congress itself would have made the choice to create criminal liability for pre-Act offenders. Under the second interpretation, however, the Act would appear to give to the Attorney General the power to decide whether to impose registration requirements, backed by criminal penalties, on the 500,000 sex offenders whose convictions pre-dated the Act.

In Reynolds v. United States, decided in 2013, the Court chose the second interpretation. Writing for a lopsided majority, Justice Breyer found that the Act's conferral of authority to determine the Act's "applicability" to pre-Act offenders was "more naturally read as conferring the authority to apply the Act, not the authority to make exceptions" – meaning that the Act's registration requirements did not apply to pre-Act offenders until the Attorney General said they did. Chief Justice Roberts, who at oral argument had asked counsel for the government whether delegating this degree of authority would create a nondelegation problem, joined Justice Breyer's opinion without comment. As did Justices Alito and Thomas.

Justice Scalia dissented, joined by Justice Ginsburg. Justice Scalia saw no reason to interpret the statute in the way the majority did, but he did see a good reason not to: "[I]t is not entirely clear to me that Congress can constitutionally leave it to the Attorney General to decide – with no statutory standard whatever governing his discretion – whether a criminal statute will or will not apply to certain individuals. That seems to me sailing close to the wind with regard to the principle that legislative powers are nondelegable…" Justice Scalia observed that it was thelis Court's "settled policy to avoid an interpretation of a federal statute that engenders constitutional issues if a reasonable alternative interpretation poses no constitutional question." Indeed, the Court has in the past chosen a narrow interpretation over a broad one precisely because it thought the broad interpretation might be inconsistent with the nondelegation principle.

The basic problem in Gundy is that SORNA, as interpreted by the Supreme Court in Reynolds, provides not just imprecise guidance, but no guidance to the Attorney General in deciding whether to apply the law to pre-Act offenders. The government made no bones about this limitless discretion in Reynolds itself; indeed, it used the unconstrained nature of the authority given to the Attorney General to argue against making the application of the statute to pre-Act offenders turn on the Attorney General's action. Under the provision on the Attorney General's authority, the government said, the Attorney General could apply the Act to "every" pre-Act offender or to "some but not all" or could "do nothing at all," and could "change his mind at any given time." Despite these cautions, the Court ruled against the government in Reynolds.

In defending SORNA against the constitutional attack in Gundy, the government now attempts to create a limiting principle for the Attorney General's authority. It argues that "the text and context" of SORNA convey an instruction to the Attorney General to "specify the applicability of SORNA's registration requirements to pre-Act offenders 'to the maximum extent he finds to be feasible.'" In Gundy, the government has thus discovered a limiting principle that it could not locate when it litigated Reynolds. The trouble is that the limiting principle – "to the maximum extent he finds to be feasible" – does not appear in the statute.

By interpreting SORNA to require action from the Attorney General before the law's registration requirements may apply to pre-Act offenders, the Court in Reynolds created the constitutional conflict that Justices Scalia and Ginsburg sought to avoid. In this way, Reynolds gave birth to the constitutional controversy in Gundy.

Unless the Court is willing to overturn its decision in Reynolds – which is unlikely, especially given the recentness of the decision – the Court in Gundy will not be able to wriggle out of the constitutional question posed there. Given the four conservative justices' apparent desire to police nondelegation problems with a firmer hand, one cannot help but wonder whether the legal box they've put themselves in suits them just fine. Add Judge Kavanaugh to the mix, and one may have a solid majority for reworking present understandings of the separation of powers.

If the Court invalidates SORNA as an unconstitutional delegation of legislative authority, the consequences could be gigantic. A reawakened nondelegation doctrine could run like a scythe through the scores of statutes that grant broad authority to administrative agencies. Operating that destructive instrument would be the least politically accountable of all the branches of government.

The Court could try to limit a decision invalidating SORNA, citing such factors as the criminal context; the worrisome power of the Attorney General as both prosecutor and quasi-legislator; and the backward-looking nature of the application of registration requirements to pre-Act offenders. But given that this is the very fight that the conservative justices appear to have been spoiling for, and given that they could have easily avoided it in Reynolds, it seems fair to say that no matter how Gundy comes out, this will not be the Court' last word on nondelegation.

I am grateful to Jonathan Adler for helpful comments on an earlier draft.

Regulation and the Administrative State