HOW AGENCIES CHOOSE WHETHER TO ENFORCE THE LAW: A PRELIMINARY INVESTIGATION.

• Author: Nielson, Aaron L.
• Position: Symposium: Administrative Lawmaking in the Twenty-First Century

INTRODUCTION

The U.S. Supreme Court recently did a remarkable thing. In 2016, the Court granted certiorari to review whether the "Deferred Action for Parents of Americans and Lawful Permanent Residents" executive action--DAPA for short--should be enjoined as unlawful. Through this program, the Obama administration purported to use enforcement "discretion to permit an alien to be 'lawfully present,'" so long as certain conditions were satisfied. (1) The Supreme Court's decision to hear the case was not remarkable; once the Fifth Circuit upheld an injunction of DAPA, (2) nearly everyone expected the Court to hear the case.

What was remarkable, however, was the Court's sua sponte instruction that the parties also brief "[w]hether the Guidance violates the Take Care Clause of the Constitution, Art. II, [section] 3." (3) That command turned what was already a contentious dispute about immigration in particular and administrative law in general into an all-out constitutional battle. Anticlimactically, however, the Court later announced that because the Justices were split four to four, the Fifth Circuit's decision would be affirmed without opinion by an equally divided court. (4) The Court therefore did not resolve the thorny question of whether there are meaningful limits on the executive branch's discretion to not enforce the law.

The fight over DAPA (5) is just one example of what has become an increasingly controversial aspect of administrative law: nonenforcement. (6) Similar fights have played out over federal narcotics policy, (7) the Affordable Care Act, (8) various welfare programs, (9) and how to regulate "in the wake of the financial crisis." (10) All the while, scholars have also considered whether there are bounds to the nonenforcement power (11) and, if so, where the line should be drawn and how such a line could be enforced. (12) Indeed, some have even suggested that if a future administration does not like certain types of federal taxes, perhaps it can simply decline to collect them. (13)

This recent focus on nonenforcement is noteworthy because an agency's decision to let violations occur is not the typical emphasis of administrative law scholarship. After all, when one thinks of regulatory decisions, what often comes to mind are affirmative actions to ensure that the law is obeyed, which often includes ensuring that legal prohibitions are respected and, if appropriate, that violations of such prohibitions are sanctioned. Yet nonenforcement demonstrates that is not all that agencies do. Instead, agencies sometimes allow violations of the law to go without enjoinment or sanction.

Some nonenforcement, of course, is often beneficial and, in any event, inevitable. Agencies have finite resources; it is impossible for them to investigate and punish every potential violation of the law. (14) And sometimes there are good reasons to not even try. After all, a situation may be so distant from what Congress or the agency had in mind when a general prohibition was put on the books that enforcement makes no sense. "Indeed, a central normative reason for separating legislative and executive functions, as articulated by Montesquieu, the Federalist Papers, and other foundational sources, is to create a safety valve that protects citizens from overzealous enforcement of general prohibitions." (15) As Judge Stephanos Bibas has observed, no one should bring a knife to a school. Yet there is a world of difference between an adult brandishing a switchblade and a child carrying a kitchen knife to cut a birthday cake. (16) Just as prosecutorial discretion makes sense when it comes to crimes by school children, nonenforcement has a place in administrative law. (17)

At the same time, nonenforcement can raise troubling questions. Even apart from separation of powers concerns, nonenforcement implicates basic notions of fairness and administrative regularity. As with other forms of administrative discretion, (18) discretionary authority to determine when the law should and should not be enforced can be put to good ends but is also subject to abuse. (19) In fact, "government by waiver," if taken too far, is antithetical to liberty. As Richard Epstein has explained, agencies may be tempted to overregulate, knowing that they can fall back on nonenforcement. (20) The result may be that regulated parties find themselves at the mercy of the government, and "when currying the favor of capricious government officials is required for a person's well-being or a firm's very existence, government abuse becomes nearly impossible to oppose." (21) This danger comes to the forefront when agencies deliberately engage in nonenforcement for self-aggrandizing reasons, especially to obtain ends that are outside of the agency's statutory authority. (22) But the risk of abuse can also arise in more mundane situations, for instance, where agency officials, perhaps unconsciously, play favorites. (23)

Although nonenforcement has received some theoretical consideration in the academy, it has not received much empirical examination. This Article attempts to begin filling that void. Drawing on survey data from nine agencies and interviews with officials from the Consumer Financial Protection Bureau (CFPB), Federal Aviation Administration (FAA), Mine Safety and Health Administration (MSHA), and the Alcohol and Tobacco Tax and Trade Bureau (TTB), this Article explores nonenforcement (specifically waivers, exemptions, and prosecutorial discretion) at these agencies. This empirical investigation, moreover, serves as a springboard for a hopefully more fulsome conceptual understanding of when nonenforcement is dangerous and how to safeguard against abuse. Indeed, this Article makes three contributions to the nonenforcement literature.

First, Part I describes how certain agencies evaluate potential nonenforcement. For instance, what are the steps and who is involved? What procedures do they use? How regularly do agencies engage in nonenforcement? Do they reject requests for nonenforcement and, if so, how often? Empirically, it turns out that the answers to these questions vary widely across agencies. Some agencies have robust nonenforcement practices; the FAA, for instance, makes hundreds and sometimes even thousands of formal nonenforcement decisions per year. Because of the volume of nonenforcement decisions, the Agency has developed regularized procedures to evaluate requests, which often include a public process. By contrast, the CFPB has developed two formal nonenforcement programs, one of which has never been used, and the other of which has only been used once. And the TTB makes dozens of formal nonenforcement decisions each year but, due to privacy concerns, essentially never publishes them. All the while, agency inspectors informally may be making countless nonenforcement decisions each day, particularly in certain industries. There is at least one characteristic, however, that cuts across agencies: very few of these decisions are ever challenged in court.

Second, reflecting this heterogeneity, this Article begins to sketch a taxonomy of nonenforcement. Not all nonenforcement is the same and some varieties of it are especially dangerous. For example, nonenforcement may be particularly susceptible to abuse when the agency acts sua sponte in a scheme with opaque substantive standards and no notice-and-comment procedures. By contrast, nonenforcement may be relatively less problematic when the agency, acting upon a formal request, applies neutral standards following a public comment process. And these factors do not exhaust the full range of potential considerations. Until nonenforcement is disaggregated, evaluation of it is incomplete.

And third, this Article explores how to prevent nonenforcement from being abused. Most importantly, nonenforcement and retrospective review should not be understood as discrete aspects of administrative law. Instead, successful requests for nonenforcement, especially when there are many similar ones, should be understood as a presumptive signal that the agency should engage in retrospective review. If a prohibition no longer makes sense, it is understandable that agencies may employ their nonenforcement authority. But rather than rely too much on nonenforcement, the better path may be simply to change the law for everyone.

1. BRIEF BACKGROUND

   Nonenforcement is an important part of administrative law. The issue has taken on particular significance in recent years as presidents have begun aggressively using nonenforcement for ideological ends.

   Administrative law has long recognized the benefits of nonenforcement. This recognition is well illustrated by Heckler v. Chaney, (24) "one of the modern landmarks of administrative law." (25) In Heckler, the Court confronted alleged nonenforcement by the Food and Drug Administration (FDA). A group of death row inmates challenged the Agency's refusal to regulate certain drugs that states were using for executions. The Agency disagreed with the prisoners' view of the law, but also declared that even if the prisoners were right, the Agency would not enforce federal law in these circumstances because "enforcement proceedings in this area are initiated only when there is a serious danger to the public health or a blatant scheme to defraud." (26) The D.C. Circuit held that the Agency's decision was reviewable. (27)

   The Supreme Court disagreed, concluding that at least when it comes to whether to bring an enforcement action, agency nonenforcement should "be presumed unreviewable under the 'committed to agency discretion' exception to the general rule of reviewability under the APA." (28) The Court explained, among other points, that nonenforcement requires "a complicated balancing" that considers not only whether the law has been violated, but also "whether agency resources are best spent on this violation or another, whether the agency is likely to succeed if it acts, whether...