ACCOUNTABILITY FOR NONENFORCEMENT.

• Author: Velikonja, Urska
• Position: Symposium: Administrative Lawmaking in the Twenty-First Century

In recent decades, almost every new presidential administration has come into office after having made campaign promises to deregulate some area of social or economic activity. Democratic candidates promise to lessen the burden on those living in poverty, including limiting enforcement against law-abiding undocumented immigrants. Republican candidates promise to lighten the regulatory burden on business: to pare back laws perceived as excessively costly, to reduce environmental and workplace safety standards, or to deregulate regulated markets. The current administration is no different.

Even large changes in enforcement can generally be defended as reasoned policy shifts, and protected from judicial review by Heckler v. Chaney, (1) a 1985 Supreme Court decision that affords agencies considerable enforcement discretion, so long as the change is not a prospective categorical program of nonenforcement, such as President Obama's immigration policies (Deferred Action for Childhood Arrivals (DACA) and Deferred Action for Parents of Americans (DAPA)). This Article suggests that the choice between discretionary nonenforcement, which courts cannot touch, and categorical nonenforcement, which they can, is not binary. Enforcement priorities can result in enforcement declines that are substantial, but not down to zero, even in the absence of a public declaration of nonenforcement. If the availability of judicial review hinges on a public declaration of nonenforcement, the doctrine has a built-in bias in favor of well-heeled, well-connected classes of defendants. When the universe of those affected by the nonenforcement policy is small, an agency can communicate such a shift quietly, without a public declaration, and thus effectively immunize itself from judicial review under Heckler. An agency cannot do so when the universe of those affected includes several million undocumented immigrants or a large market of marijuana growers, sellers, and users.

But many significant shifts in enforcement are not categorical in the sense that enforcement does not decline to zero. Instead, enforcement declines by a substantial amount, observable as a pattern of low-priority enforcement. Like a categorical program, large shifts in enforcement over a short period of time are, in effect, similar to rule changes. Unlike categorical shifts, large changes in enforcement are immune from judicial review, and should be immune. Their adoption in the dark, however, is not consistent with the ideas that underpin the rule of law, including transparency, predictability, and accountability.

Changes in enforcement can move in more than one direction: enforcement can increase significantly as the Securities and Exchange Commission saw in the aftermath of the accounting scandals or the Madoff Ponzi scheme, and decrease precipitously, as evidenced at the Consumer Financial Protection Bureau under Acting Director Mick Mulvaney. There is no reason in constitutional or administrative law to treat changes in enforcement policy differently depending on whether enforcement increases or decreases. (2) Policy choices raise similar questions about reviewability and accountability, regardless of whether they increase or decrease enforcement. They also raise symmetrical questions about fair notice and due process and about the separation of powers. We demand that agencies give reasons for changes in rules; reason giving seems appropriate for significant shifts in enforcement, in order to match given reasons with observed enforcement practices, and to subject those reasons to political scrutiny through media coverage and congressional attention, even when judicial review is not available or appropriate.

1. ALTERNATIVES FOR DEREGULATION

   1. Three Options for Presidents Who Deregulate

      Presidents can deregulate in a variety of ways. First, a President with solid support in Congress can implement a statutory agenda. During his first (and only) term in office, during which Democrats controlled both chambers of Congress, President Carter passed the Airline Deregulation Act of 1978, (3) on the hope that competition would reduce ticket prices. Second, a President can direct agencies to deregulate by adopting new rules that revoke existing rules and regulations deemed costly or unfair. Additionally, a President can, by executive order, impose roadblocks that make new regulations less likely. Finally, a President can deregulate by appointing department and agency heads who will reduce enforcement of disfavored laws and regulations by directing their enforcement hands to do less, and thereby reduce the impact of existing laws on regulated parties.

      The three alternatives are substitutes, if imperfect ones. New statutes are difficult to adopt. They usually require the President to enjoy majority support in both houses of Congress--a feat that ordinarily occurs only in the first two years of an administration but not after that. Even with full control of Congress new statutes are difficult to adopt. This is because statutes, including deregulatory ones, face many layers of review, both inside Congress, as well as in the media, and often in courts. But the advantage of statutory (de)regulation is that statutes are difficult to undo once adopted. President Obama's marquee achievement, the Affordable Care Act, (4) is a case in point: President Trump and the Republican Congress tried to repeal the statute on several different occasions, only to fail each time. (5)

      Deregulatory rulemaking faces considerably fewer political challenges, and less media scrutiny, but is usually no easier to implement than a statute because of significant procedural requirements. Rules and regulations must go through the notice-and-comment process required by the Administrative Procedure Act (APA). (6) The process requires agencies to justify legal rules, including showing that the social welfare benefits of a proposed rule outweigh its costs. Moreover, the interested public has the right to participate in the process and offer comments, to which the agency must respond. in addition, economically significant rules proposed by departments and federal agencies (except for independent agencies that include most financial regulators) must be reviewed by the Office of Information and Regulatory Affairs (OIRA), an agency that is closely associated with the White House and the President of the United States. (7) OIRA often requires substantial amendments to proposed rules, and delays decisions on rules it dislikes, effectively vetoing them. (8) Even if a rule survives the many procedural steps, adopted rules are subject to judicial review, where a court may vacate the rule for failure to respond adequately to public comments or for failure to consider the costs and benefits of adopted rules. (9) There is no safe harbor for deregulatory rules or actions to rescind a properly adopted rule; they, too, must jump through all procedural hoops. (10) A rule can be repealed by statute, but statutes are difficult to adopt, as explained above. As a result, repealing rules is just as difficult as adopting new rules, if not more so. A statute can pass Congress even if Congress is uninformed and even if the costs of the amendment outweigh its benefits. The same is not true for agency rules. Once a rule has survived the notice-and-comment process, it is very difficult for an agency to reverse course absent compelling evidence showing that the original rule significantly underestimated the costs and/or overestimated the benefits. (11)

      Not surprisingly, the vast majority of successful deregulatory regulation by the Trump administration in its first year has been pursuant to the Congressional Review Act. (12) The Act allows a new Congress to repeal rules and regulations published late in the term of an outgoing administration through an expedited legislative process. (13)

      There is a third way to deregulate. In recent decades, Presidents have often resorted to nonenforcement as the preferred method to deregulate in lieu of legislative or regulatory processes. (14) The President, and by extension, agency heads, can achieve deregulatory objectives through nonenforcement without significant delays and without any real threat of judicial review, contrary to the very real roadblocks present in legislating and rulemaking. This is because the Supreme Court afforded agencies considerable enforcement discretion for piecemeal enforcement decisions. As long as a change in enforcement policy is not prospective and categorical, it is immune from judicial review. (15) Similar to nonenforcement are soft agency policy positions, such as guidance documents that can be reversed very quickly. (16)

      Deregulation through nonenforcement may not outlast the administration but has very real consequences nonetheless. In the first year of the Trump administration, agency enforcement has declined across the board. (17) Staffing changes at the Environmental Protection Agency (EPA), the Department of Education, and the Food and Drug Administration (FDA) have resulted in laxer enforcement against polluters, for-profit colleges, and medical device manufacturers. (18) Declines in enforcement have also permeated into the financial regulatory agencies that historically have not faced such cutbacks. This development raises several interesting questions about appropriate accountability mechanisms. The following Sections explore, in turn, the legal requirements for enforcement and the history of deregulation through nonenforcement.

   2. The Duty to Enforce the Law and Enforcement Discretion

      The Take Care Clause of the Constitution commands the President "to put the laws into effect, or at least to see that they are put into effect, 'without failure' and 'exactly.'" (19) The words "faithfully executed" do imply some degree of discretion, but most commentators agree that the Clause establishes a presumption that Presidents will dutifully follow existing law, in...