abstract. This Article investigates the normative and constitutional case for a particular form of congressional delegation that is of increasing practical importance: delegations that give agencies the power to deprive statutory provisions of legal force and effect, a power this Article calls “administrative forbearance authority.” Although legal scholars have recently noted the rise of administrative forbearance authority, they have largely ignored how exactly such a power might operate in the hands of the agency and the various governance functions it performs. Without such knowledge, the case for administrative forbearance authority is necessarily incomplete.

This Article thus makes two principal contributions to the literature. First, it describes the variety of functions that administrative forbearance authority serves at the agency level, drawing on the previously unexplored histories of various agencies’ experience with such authority. Second, it uses the descriptive account both to develop a fuller normative and constitutional case for administrative forbearance authority and to illuminate the various circumstances in which forbearance can be beneficially employed as a policy tool.

To defenders of delegation generally, this Article posits that there is no special reason to be wary of administrative forbearance authority and that forbearance can be used as a governance device in previously underappreciated ways. To critics who urge a stronger nondelegation doctrine than the one we have today, I argue that there may be reasons to actually support administrative forbearance in a world where delegations of the traditional type are unlikely to go anywhere anytime soon.

author. Climenko Fellow and Lecturer on Law, Harvard Law School. Many thanks to Nick Bagley, Susan Crawford, Dan Epps, Barry Friedman, Jake Gersen, Jacob Goldin, Don Herzog, Sam Issacharoff, David Kamin, Jo Langille, Maggie Lemos, Leah Litman, Todd Rakoff, Ricky Revesz, Matthew Stephenson, Cass Sunstein, and participants in the Climenko Fellows works-in-progress workshop, the NYU School of Law Furman Scholars Seminar, and the Harvard Law School “Becoming a Law Professor” Seminar for valuable guidance and comments. I previously represented telecommunications carriers in some of the proceedings discussed below. All opinions, errors, etc. are my own.

Introduction

In 2015, the Federal Communications Commission (FCC) reclassified broadband Internet service providers (ISPs)—companies like Comcast and Verizon—as common carriers under the Communications Act.1 That decision, heralded by many,2 automatically subjected such providers to a range of statutory obligations. At the same time, however, the Commission announced it will “forbear from”—render inapplicable—many of these requirements. The Commission can do so because of a provision in the Communications Act, little known outside of communications-law circles, allowing the FCC to formally deprive portions of the Act of their legal force.3 In other words, the statute expressly allows the Commission to render statutory requirements no longer legally binding.

Delegations to agencies of the power to deprive statutory provisions of legal force and effect—what this Article calls “administrative forbearance authority”4—raise a set of questions distinct from those associated with traditional delegations of authority to agencies to fill in the details of a regulatory scheme. What roles does such an authority serve in the hands of an agency? Do the traditional justifications for delegating authority to make law also apply to delegations that allow an agency to relieve regulated parties of their statutory obligations? How does such a power compare to other forms of agency action, such as nonenforcement? Should we have more reason to fear such delegations than we do delegations of the normal sort?

Although the literature has begun to investigate these and other questions,5 it has yet to offer a full descriptive and normative evaluation of administrative forbearance authority. This Article thus makes two principal contributions to the literature. First, it describes the variety of functions that administrative forbearance serves at the agency level, drawing on the ways the FCC and other agencies have used their forbearance authority. Second, the Article uses this descriptive account to mount a normative case for forbearance as a particular form of delegation and to illuminate the range of circumstances in which it might be used. It thus provides a robust defense of administrative forbearance authority that is firmly grounded in both the realities of administration and administrative-law theory.

The time is ripe for a fuller evaluation of forbearance authority. As many have written,6 the current age is characterized by legislative gridlock in which agencies face increasing pressure to “tailor” statutes that are out of date, overbroad, or simply unworkable.7 An expressly delegated power to ease or eliminate statutory requirements, such as that held by the FCC, presents the promise of much-needed regulatory flexibility.

Forbearance authority is also potentially more legitimate than other tools. Judges and commentators have recently expressed concerns about different kinds of agency actions—especially discretionary decisions not to enforce certain applications of federal statutes—that border on unilateral executive lawmaking.8 An express forbearance authority avoids many of the criticisms leveled against these other forms of executive lawmaking: forbearance involves the exercise of a power that Congress explicitly gave to an agency, and an agency’s exercise of its forbearance authority triggers procedures that result in more opportunities for judicial review and public comment than decisions not to enforce a statute.

That said, forbearance does raise its own set of normative issues, which have been thrust to the fore by forbearance-like provisions contained in high-profile programs such as the No Child Left Behind Act,9 as well as by Mitt Romney’s presidential campaign pledge to dismantle the Affordable Care Act through executive action.10 For one, the traditional policy justifications for delegating authority to an administrative agency—most prominently, agencies’ greater expertise and flexibility vis-à-vis Congress—apply less clearly to administrative forbearance authority. With “normal” delegations, the typical narrative is that Congress wants to do this or that but isn’t quite sure how and thus delegates power to an agency. Among other reasons, we tolerate delegation because the agency is, under the received view, often in a better position than Congress to know the precise course to chart and can more easily change direction if necessary. But with negative delegations, Congress has already set the requirements and defined to whom (or what) they apply, at least using broad strokes.11 For that reason, the traditional story we tell for “normal” delegations does not seem, at least at first glance, to apply with full force to forbearance-like delegations.

Scholars have also recently raised questions about the propriety of forbearance-type delegations. Law professor and historian Philip Hamburger has called an express agency power to nullify statutory requirements “astonishing even by administrative standards.”12 Another scholar has written that conferring negative-lawmaking authority on an agency “amounts to an abdication of Congress’s core legislative functions.”13 And writing at the Volokh Conspiracy blog, Professor David Post recently penned that it is “[h]ard for [him] to believe that” such provisions “can pass constitutional muster; it’s like a repeal process, but one not involving Congressional action.”14 For some, agency action formally dispensing with legal requirements simply seems more “legislative” than other types of agency action. Indeed, a majority of the Supreme Court came close to endorsing this position in Clinton v. City of New York, which invalidated the Line Item Veto Act.15 Under this view, forbearance authority is constitutionally suspect either because it impermissibly delegates power to agencies or because it violates Article I, Section 7’s bicameralism and presentment requirements.16

Even apart from these formal legal concerns, some have voiced alarm that forbearance-type delegations are more likely to be used by agencies in nefarious ways—perhaps to override the will of Congress or to grant regulatory breaks to powerful groups.17 In an essay noting the rise of what he called “government by waiver,” for example, Richard Epstein writes that, in part because of the potential for favoritism created by a broad power in the Executive to nullify statutory requirements, such authority represents a “particularly dangerous form of government power.”18

This Article addresses these issues in the following ways. First, it develops an understanding of the roles played by forbearance authority, and, based on this descriptive account, reveals that the primary justifications for delegations to an expert agency apply with full force to “negative” or forbearance-type delegations. Second, the descriptive account reveals that forbearance, like the more familiar authority to fill in the details of statutes, is properly viewed as implementing (and not overriding) the statute Congress has written, which includes the delegation itself. Thus, an agency exercising forbearance authority no more exercises a purely legislative power to “repeal” the law than an agency filling in the gaps exercises a legislative power to enact law. Finally, in practice, agencies’ histories with administrative forbearance power do not reveal them to be behaving in particularly problematic ways. Instead, agencies often use forbearance authority to address perennial governance problems that Congress anticipated when it included the delegation in a statute. Of course, the potential for abuse remains, just as it does for any type of government power. But these histories indicate that forbearance need not be viewed as a particularly troubling form of delegation.

This Article does more than respond defensively to various critiques of forbearance, however. It also highlights several underappreciated benefits of forbearance-like delegations, providing a roadmap for policymakers who are considering when and in what circumstances to include forbearance provisions. The descriptive account shows that forbearance can be used as a policy tool in seemingly counterintuitive ways. For example, Congress has used forbearance as an anticapture device, allowing it to set an initial proregulatory baseline while permitting the agency some flexibility to depart downward from that baseline. Similarly, even though forbearance might appear to be a form of deregulation, forbearance actually can be used in a range of circumstances to facilitate more socially beneficial regulation than might otherwise be possible. Forbearance-type delegations also have underappreciated ancillary benefits for the administrative process. Forbearance may substitute for other types of agency action, such as agency nonenforcement decisions, that pose graver normative concerns. In addition, a number of the standard critiques of delegation apply with less force to forbearance-like authority. Thus, to the extent forbearance-type delegations can stand in for garden-variety positive delegations, even critics of delegation generally may have reason to support them.

The argument proceeds in four parts. Part I sets up the problem. In the current age of congressional gridlock, scholars and politicians have increasingly turned their attention to the executive branch for potential solutions. Administrative forbearance is thus a particularly attractive tool at this point in time. Yet it still is only partially understood.

Part II turns to the task of describing the functions that administrative forbearance serves in the hands of an agency. A forbearance option can be used to address familiar problems, such as when a statutory requirement has become unnecessary or counterproductive (the problem of statutory obsolescence), or when the application of a requirement in a particular setting does not seem justified in light of the reasons for the statute (the problem of overinclusiveness or “fit”). But there are also a number of less obvious ways an agency can use forbearance to pursue beneficial regulatory ends. Somewhat paradoxically, one such use is to enable regulation: forbearance can be deployed to selectively regulate in contexts where applying a set of statutory requirements across the board would be unwise.

Part III pivots from the descriptive to the normative. Section III.A applies the primary policy-based justifications for administrative delegation to the case of forbearance-type delegations and argues that those justifications apply with equal force to forbearance authority. Section III.B then shows that agency action pursuant to forbearance authority is superior to other forms of agency action, such as statutory nonenforcement, for which it may substitute. Because enforcement discretion often operates “underground,” shielded from public view, policymaking through enforcement is vulnerable to criticism on process grounds: nonenforcement often is not transparent and can be used to evade the Administrative Procedure Act’s (APA) notice-and-comment requirements, escape judicial “arbitrariness” review, and limit opportunities for cost-benefit analysis. By contrast, forbearance decisions typically are made openly, are immediately appealable to the courts, and are announced in a form amenable to calculating the relevant costs and benefits.

Section III.C then addresses various objections that might be lodged against administrative forbearance. An important theme of this Section is that, in many cases, the objections to traditional positive delegations apply with less force to negative delegations. After all, when it delegates forbearance authority, Congress is the one responsible for setting the regulatory default. As I explain, agency deviations from the default often are easier to monitor and address than agency actions setting regulatory requirements in the first place.

Finally, Part IV considers the promises and limits of forbearance as a policymaking tool. Section IV.A describes the conditions under which Congress should include forbearance authority in its statutes and under which we might be skeptical of forbearance as a normative or policy matter. Section IV.B then invokes recent regulatory controversies involving the Voting Rights Act (VRA) and the Clean Air Act to describe how a forbearance authority might operate in areas in which it currently does not exist.

I. background

A. Old Statutes, “Hyper”-Statutes, and Executive Policymaking

According to many scholars, agencies—and the executive branch more generally19—are in a tough spot. The current political climate, marked by congressional gridlock and extreme partisanship, has hampered legislative policymaking.20 For any number of reasons, Congress increasingly lets “old” statutes that are out of step with current realities languish on the books.21 In the rare political moments where Congress produces legislation, the legislation tends to be sprawling and, at least according to some, ill-conceived or even “incoherent”—a trend toward what one recent article calls “hyper-legislation.”22 In this climate, agencies are increasingly responsible for making sense of that legislation. The result is what two scholars have recently described as a significant uptick in executive “policymaking in the absence of Congress.”23 In particular, agencies face growing pressure to “tailor” statutes—using enforcement discretion or otherwise—to address real governance concerns.24

Such agency action seems almost inevitable, especially in the current environment.25 If Congress’s statutes pose serious governance problems that Congress may not be prepared or able to address, the task of working out those problems often will fall, for good or ill, to the agencies charged with administering the statutes. However, these agency-centered responses to congressional dysfunction pose potential legitimacy issues: most fundamentally, are these exercises of power lawful in the absence of express congressional approval? Questions like this have vexed lawyers and scholars in the debate over President Obama’s immigration policies, for instance.26 And if recent history is an indication, they will continue to plague administrative-law practice for years to come.

B. A Way Out? Agency Forbearance Authority

Because of the governance problems posed by congressional gridlock, administrative forbearance authority—by which Congress grants agencies the express power to deprive the laws it passes of legal force and effect27—may be particularly appealing to modern policymakers. Of course, forbearance does not itself solve the problem of gridlock. Because forbearance is an express power, Congress itself must pass legislation to grant it. Nonetheless, there is reason to think that Congress may be more able to pass legislation involving broad forbearance provisions than legislation lacking such provisions: as Judge (then Professor) David Barron and Professor Todd Rakoff have noted, forbearance-like delegations may help facilitate compromise on legislation, especially in the current political environment.28

Administrative forbearance is distinct from other kinds of agency power that it resembles. It is different from enforcement discretion in that it operates to formally nullify statutory requirements on a prospective basis, at least if not reinstated by the agency during a subsequent rulemaking.29 It is also different from, though related to, longstanding concepts such as administrative waivers and variances. Traditionally, waiver has referred to agencies’ ability to waive their own regulations, not the requirements contained in a statute.30 Some statutes, however, do contain provisions allowing agencies to waive statutory requirements in exceptional circumstances and in particular cases—often, though not always, when national security is at issue.31 Similarly, statutes may allow agencies to grant variances from statutory or regulatory requirements; these variances allow the agency to make such requirements more or less strict for a particular regulated party.32 Forbearance is different. While waivers and variances tend to involve one-off exceptions based on facts particular to the regulated entity involved, forbearance operates on the wholesale level, resembling the exercise of prospective policymaking. With forbearance, in other words, an agency decides based on shared facts that a statutory requirement should be eliminated across the board or with respect to an entire category of regulated entity.33

An expressly granted authority to void statutory requirements solves the particular legitimacy problem posed by unilateral executive action. When an agency acts pursuant to a congressionally delegated power to deprive statutory provisions of legal force or effect, the agency is not going it alone but rather exercising authority granted by Congress. In these cases, the executive branch’s power is traditionally at its apex.34

Although an expressly granted authority to deprive statutory provisions of legal force and effect is a solution to one problem, it also raises another: that of delegation. Simply put, does the Constitution permit Congress to grant agencies such a power? And normatively, as opposed to strictly legally, should we allow Congress to provide agencies with the express authority to override statutes that Congress itself has passed?

The answers to these questions are unresolved, and the literature on delegation thus far has not fully addressed them. As a purely doctrinal matter, it is indeed unclear whether current law allows for such delegations. It is true, of course, that Congress has nearly unfettered ability to grant agencies the power to promulgate rules with binding legal force. Such “positive” lawmaking delegations—delegations of the authority to make law—are valid as long as the delegation supplies an “intelligible principle” to guide the agency’s lawmaking discretion.35 And the Court has upheld, over the years, several laws containing intelligible principles as vague as the “public interest.”36

In theory, the “intelligible-principle” test should apply no differently to negative delegations than to the more traditional positive variety. The intelligible-principle test focuses on the scope of discretion granted to an agency; it is satisfied as long as the agency’s discretion does not amount to a limitless, truly “legislative”-type power. In terms of the agency’s discretion, a congressional direction that an agency may regulate greenhouse-gas emissions if doing so would be in the public interest is the same as a direction that an agency shall regulate greenhouse-gas emissions unless doing so would be contrary to the public interest.37 The only difference between these cases is the default state (regulation or no regulation), an issue about which the intelligible-principle test is theoretically agnostic.38

The rub comes with the Supreme Court’s decision in Clinton v. City of New York.39 That decision invalidated the Line Item Veto Act, which allowed the President to “cancel” certain kinds of spending items and tax benefits after they were formally signed into law.40 After President Clinton exercised that power, the Supreme Court struck down the Act, stating that “[i]n both legal and practical effect, the President has amended two Acts of Congress by repealing a portion of each.”41

Scholars have debated the meaning of Clinton, which was famously cryptic. They have disagreed, for instance, about whether the result rested on Presentment Clause grounds, as the Court claimed, or whether the case was really a disguised application of the nondelegation doctrine instead.42 And regarding the scope of the decision, some scholars have argued that Clinton is best read as essentially limited to its peculiar facts. In particular, Barron and Rakoff have argued that the Court’s real concern lay in the five-day limitation on the President’s exercise of his cancellation authority.43 In their view, the five-day limitation meant that the delegation in question could not have been based on the traditional justifications for delegating authority to the Executive (such as expertise), and was “thus proof of abdication, pure and simple.”44 Outside of such “extreme cases,” however, Barron and Rakoff argue that Clinton leaves plenty of room for negative-type delegations.45

While Barron and Rakoff make a persuasive case for a narrow reading of Clinton, a broader reading may also be plausible and is supported by several other aspects of the Court’s opinion.46 For example, Clinton repeatedly noted, without reference to the five-day limitation or any other feature unique to the Line Item Veto Act, that a cancellation effectively “amend[ed] or repeal[ed]” legislation by depriving a statutory provision of legal force or effect, something the Court appeared to believe violated the Presentment Clause ipso facto.47 Moreover, in distinguishing prior cases upholding forbearance-like delegations, the Court stressed two points. First, the Court explained that its decisions had only upheld negative delegations in the context of international trade, an area in which the Executive is afforded both greater deference and greater discretion.48 Second, the Court emphasized the extremely limited nature of the Executive’s discretion under prior negative delegations, which the Court characterized as more akin to factfinding than policymaking.49 Clinton’s discussion on these points can thus be interpreted to suggest that something more than a lax form of the “intelligible-principle” test applies to negative delegations. This broader reading of Clinton cuts against negative delegations more generally, perhaps even prohibiting them entirely.

Some scholars have pushed this point even further. Most prominently, Philip Hamburger suggests that an executive power to deprive legal provisions of legal force is fundamentally at odds with the separation of powers embodied by the Constitution.50 For him, a power to “waive” (or “dispense with”) statutory requirements for individual parties is a kind of “extralegal” power that is not legislative, judicial, or executive in nature and cannot be legally authorized, even by the legislature.51 Hamburger believes that a more broad-based power to “suspend” legal requirements rests with the legislature, but that power is nondelegable unless a constitution specifically authorizes its delegation.52 It is unclear in which category Hamburger would place the various forbearance provisions discussed in this Article. In any event, his critiques share a common premise—namely, that when an agency lifts statutory requirements, it is doing something categorically different than the executive task of implementing legislation as written.53

In addition, R. Craig Kitchen, drawing on Clinton, argues that Article I, Section 7’s bicameralism and presentment requirements mean “[t]hat [statutory] text should not have its legal force or effect undone through the exercise of unilateral executive discretion.”54 In particular, Kitchen worries that the minority-protective nature of Article I, Section 7 will be undermined if “specific compromises in the negated statutory text” are undone by the Executive through the exercise of forbearance authority.55 As he elaborates:

The Article I, Section 7 test of Clinton, properly understood, represents fidelity to a highly specific constitutional bargain regarding the procedure for lawmaking. . . . Negating or altering the legal force or effect of that text outside of bicameralism and presentment risks undermining whatever bargain was made for that specific text, and it does so without the input of the parties who made the bargain and without providing them any opportunity to demand compromise in other areas in exchange for their agreement to the change.56

Scholars are thus divided over whether negative delegations are constitutional. As the foregoing discussion suggests, much of the scholarship on negative delegations has focused on the proper interpretation of the Court’s opinion in Clinton. In doing so, however, the literature has yet to fully explore the constitutional case for negative delegations.

C. The Need for a Functional Analysis

What has been missing from the small but growing literature on forbearance-type delegations is, in part, a fuller understanding of how such delegations operate within the overall system of government. Even defenders of negative delegations treat them as a new phenomenon, and a potentially irregular one at that.57 But, though agency forbearance authority has operated mostly in the shadows until recently, it is not so new.

This lack of knowledge contributes to some of the uneasiness about negative delegations. The propriety of positive delegations has rested in part on a set of functional considerations that depend on a particular understanding of the roles played by Congress and agencies in the policymaking process.58 As Rafael Pardo and Kathryn Watts recently put it, “Administrative law teaches that broad delegations of policymaking power to agencies may well be desirable—and, hence, will generally be tolerated as a constitutional matter—because of a variety of functional considerations relating to agencies’ institutional structures and capacities.”59 The prodelegation position has (often implicitly) assumed that Congress and agencies perform particular, complementary roles in the lawmaking process. Both critics and defenders of broad congressional delegations assume that Congress does not engage in detailed legislative drafting for various reasons—for example, the costs associated with drafting detailed legislation, a desire to harness agency expertise in fleshing out policy details, or, more nefariously, a desire to shift blame for hard choices onto others, such as the President.60 According to one metaphor, a decision to leave a statute “incomplete”—and allow an agency (or court) to fill in the relevant details—is similar to the decision made by contracting parties to write an incomplete contract.61 Whatever Congress’s reasons for writing vaguely worded laws, however, the agency’s role in such situations is clear. The agency makes rules governing situations where Congress has not spoken. It is here, in filling in the policy details, where agencies’ greater expertise, flexibility, and the like are usually brought to bear.62

But the functional considerations supporting delegations of negative-lawmaking authority to agencies are less readily grasped. With negative delegations, Congress specifies certain requirements and gives an agency broad power to dispense with those requirements. To return to the contract metaphor, this type of legislation resembles a contract in which the contracting parties specify many details of their transaction while inserting a broad force majeure clause allowing a court to dispense with the contract’s terms under certain circumstances. Yet little is known about what agencies are to do, and what they in fact do, when given such an authority.

Existing defenses of forbearance authority are thus necessarily incomplete. To date, the most thorough defense of forbearance-like authority has been supplied by Barron and Rakoff.63 Apart from drawing attention to the phenomenon of forbearance and providing their own interpretation of the Court’s decision in Clinton, Barron and Rakoff helpfully catalogue the reasons, sounding in considerations of political economy, that Congress increasingly delegates forbearance authority to administrative agencies.64 In their telling, forbearance is a salutary development largely because it facilitates lawmaking by allowing Congress to “establish new regulatory frameworks” knowing that those frameworks can be revised by the executive branch.65

Yet Barron and Rakoff’s account focuses almost entirely on Congress. Although Barron and Rakoff make a compelling case that forbearance authority is a useful tool for Congress, their analysis, as a normative defense of forbearance, tells only part of the story.66 That is because they do not focus on the functions that forbearance might serve in the hands of the agency itself.

Understanding those functions is useful for at least two reasons. First, as the above analysis suggests, without understanding the functions a statutory forbearance authority might serve, it is harder to know whether such delegations capture the benefits of agencies’ greater expertise and information relative to Congress, or whether Congress could perform a similar role just as effectively. And if Congress could perform those roles as well or better than an agency, thereby depriving such delegations of a public-interested purpose, we should be more worried that Congress is engaged in pure buck-passing. Second, an examination of the functions forbearance serves in the hands of the agency can reveal the potentially counterintuitive roles that forbearance may serve as part of an overall regulatory system, thus providing further information for policymakers to consider when deciding whether to include forbearance-like provisions in statutes of various kinds. Moreover, as argued below, forbearance may be able to substitute for other, more problematic forms of agency action, thus buttressing the normative case for forbearance-like delegations.

The analysis also responds in part to broad constitutional critiques recently leveled against administrative forbearance authority.67 Negative delegations involve Congress using its legislative power to specify the means by which real governance concerns are addressed—namely, through the exercise of agency discretion. Negative delegations, like their positive cousins, thus implement choices made by the statutes that Congress has passed through the constitutionally prescribed Article I, Section 7 process.Moreover, the history of agencies’ use of forbearance authority shows that agencies exercising such authority are not ordinarily “undermining” whatever bargain the statute represents. Instead, agencies are implementing the statute as a whole, which includes the negative delegation itself. And because the statute as a whole has satisfied bicameralism and presentment, the minority-protective functions of those procedures should be satisfied.68

II. the uses of forbearance

This Part describes how agencies have used forbearance authority in practice, creating a taxonomy of the various functions that such authority performs. As will become apparent, forbearance comes in many shapes and sizes and has rarely involved nullifying a requirement altogether. For example, agencies have used forbearance to exempt certain categories of regulated entities from a requirement’s scope, to subject entities to regulation under a statute while relieving them of the obligation to follow certain requirements, or to defer a decision on classifying entities one way or another while forbearing from any obligations that might apply in the interim.

Because the following Sections frequently draw on the FCC’s experience, a brief introduction to the statutory source of the FCC’s power will be useful. The FCC has two different sources of forbearance authority. The first is § 332(c) of the Communications Act, which was added in 1993.69 That section subjects commercial mobile services—i.e., cellular telephony—to common carrier status, which automatically imposes a host of statutory obligations on mobile carriers.70 But § 332 also allows the Commission to render most of those obligations “inapplicable” to mobile carriers “by regulation.”71 Second, a more expansive grant of forbearance authority, added as part of the 1996 overhaul of the communications laws, is contained in § 160.72 That section provides that if certain conditions are met, the Commission “shall forbear” from applying any requirements contained either in its regulations or in the Communications Act to “a telecommunications carrier or telecommunications service, or class of telecommunications carriers or telecommunications services, in any or some of its or their geographic markets.”73 Although both sources of the Commission’s forbearance authority require it to make certain fact-sounding determinations, they are worded broadly enough that the Commission has substantial policymaking discretion to determine whether forbearance is justified.74

The following Sections explain several problems forbearance authority could be used to address and then draw on examples to show how forbearance authority has served various purposes in the hands of agencies.

A. Addressing Changed Circumstances

Perhaps the most obvious use of a statutory forbearance authority is in implementing statutory provisions that have become obsolete or counterproductive due to changed circumstances.75 Two kinds of changes are particularly relevant here. First, the factual conditions that justified a statutory provision might change.76 The FCC’s decades-long efforts to alter the Communications Act’s requirements in the face of newly emerging competition—culminating in the Commission’s defeat in MCI Telecommunications Corp. v. AT&T77—is what initially prompted Congress to enact the 1996 Telecommunications Act’s forbearance provision.78 Indeed, the legislative history of the 1996 amendments suggests that Congress acted specifically in order to overturn the decision in MCI,79 which held that the FCC’s authority to “modify” the Act’s tariffing requirements did not create a forbearance-like power to dispense with those requirements completely.80

Second, our assumptions about facts might change even when the facts on the ground do not. In other words, what we know or understand about facts might change even while the actual facts remain exactly the same. A vivid illustration is provided by the “famous” Delaney Clause,81 which banned food additives containing carcinogens at a time, the 1950s, “when carcinogenic substances were difficult to detect and all detectable carcinogens were extremely dangerous.”82 As time went on, however, our knowledge of carcinogens changed, and it became clear that a literal application of the Delaney Clause required the FDA to ban substances that posed only a minuscule risk of cancer and that may even have been safer than the available alternatives.83

How have agencies used forbearance to deal with such issues? To return to the main case study, the FCC has used its forbearance authority to relieve regulated entities of obligations in circumstances where the facts that once supported the imposition of those obligations have changed. One example comes from the 1996 Telecommunications Act’s “local competition” provisions,84 which were designed to spur competition in local (as distinct from long-distance) telephone markets that had long been dominated by monopolist firms.85 Among the Act’s reforms, Congress controversially required incumbent carriers to lease portions of their network to competitors at cost-based rates.86

By the mid-2000s, however, the Commission had already seen fit to exercise its forbearance authority to relieve incumbent carriers of their forced-leasing obligations in two geographic markets where competition had developed with unexpected rapidity.87 What had changed? The answer was something that Congress could barely have anticipated in 1996: the use of cable television facilities to provide point-to-point telephony services via Voice-over-Internet-Protocol (VoIP) technology.88 The entry into local telephone markets by facilities-based competitors unsettled a regulatory paradigm that assumed that, absent forced leasing, entry by new market players would be very unlikely.89 If that premise does not hold, forced leasing imposes costs related both to implementation and market incentives that likely would not be offset by corresponding benefits.90 Seizing on these facts, in 2004 Qwest (a former Bell company) formally asked the FCC to exercise its forbearance authority and, among other things, relieve Qwest of its unbundling obligations in Omaha, Nebraska.91 The FCC agreed.92 As the Commission explained:

While the costs of such regulatory intervention may be warranted in order to foster competitive entry into the local exchange and exchange access markets where such competition would not otherwise be generated, we find that these costs are unwarranted and do not serve the public interest once local exchange and exchange access markets are sufficiently competitive . . . . In addition to furthering the congressional goal of creating competitive local exchange markets, our decision today also furthers another of Congress’s primary aims in the 1996 Act—to deregulate telecommunications markets to the extent possible. We act today in accord with Congress’s clear intent in section 10 to sunset in a narrowly tailored fashion any regulatory requirements that are no longer necessary in the public interest so long as consumer interests and competition are protected.93

Another agency that has used a forbearance-like authority in order to address changed circumstances is the SEC. Under the statutes it administers, the SEC has several sources of what the securities laws refer to as “exemptive” authority—essentially, a forbearance power to relieve parties of statutory requirements similar to that held by the FCC.94 In 2005, the SEC exercised that authority to significantly alter many restrictions on how companies may offer securities stemming from the 1933 Act, essentially “forbearing” from those restrictions.95 As part of the reforms, the Commission relieved “well-known seasoned issuers”96 from compliance with the anti-gun jumping prohibition contained in section 5(c) of the Act, which makes it unlawful to communicate regarding any security offering prior to filing a registration statement (including a prospectus) with the SEC.97 As the Commission explained, that statutory provision was “enacted at a time when the means of communications were limited and restricting communications (without regard to accuracy) to the statutory prospectus appropriately balanced available communications and investor protection.”98 However, with the advent of modern communications technology and changes in investor behavior, the pendulum had swung. The Commission believed that “the gun-jumping provisions of the Securities Act,” as applied to seasoned issuers, had “impose[d] substantial and increasingly unworkable restrictions on many communications that would be beneficial to investors and markets and would be consistent with investor protection.”99 The Commission thus removed the restriction for well-seasoned issuers.

B. Statutory Fit

The problem of changed circumstances emerges when portions of a statute become unnecessary or counterproductive over time. Some statutes, however, are born ill fitted. Most relevant here, a statute may be overinclusive—that is, its “language, read without sufficient regard to context or its intended field of application, will reach situations that it could not reasonably cover.”100 For example, a statutory prohibition might on its face prohibit certain conduct that the reasons for the prohibition do not seem to reach. (Think, for example, of H.L.A. Hart’s famous “vehicles in the park” hypothetical.)101 Likewise, a statutory requirement may obligate regulated entities to do something that appears unjustified in light of the reasons for the requirement.

A traditional way to deal with overinclusiveness is through purposive statutory interpretation, wherein the purposes of a statute are invoked in order to justify creating carveouts from the statutory text.102 But purposivism cannot cure all overinclusiveness, especially in an age where textualism enjoys substantial support.103 Forbearance authority in the hands of an agency, on the other hand, can help cure overinclusiveness by allowing an agency to tailor the statute to the surrounding circumstances. This would ultimately create a regulatory regime that makes more sense than the “provisional” statute enacted by Congress.

The FCC has performed such a tailoring function in the mobile wireless context. Prior to 1993, the Commission regulated mobile-wireless providers—then emerging on a widespread basis—under an ad hoc regulatory regime that relied on a hodgepodge of statutory sources of authority.104 In its 1993 amendments to the Communications Act, Congress brought mobile-wireless providers within the Act’s scope by subjecting “commercial mobile radio services”—the kind of mass-market voice services now provided by carriers like AT&T and Verizon—to common-carrier status.105 As an immediate result of that decision, many of the various “Title II” common-carriage obligations, ranging from tariffing requirements to licensing rules obligating providers to gain permission from the FCC before entering or exiting the market, were to apply to mobile-wireless providers.106 At the same time, however, Congress understood that mobile-telephone markets exhibited much greater levels of competition than traditional wired telephony. So Congress allowed the Commission to “specify by regulation” that some statutory obligations would be “inapplicable” to mobile carriers, provided that the Commission had first determined that such regulations were in the public interest.107

The Commission accordingly responded by forbearing from applying many of Title II’s provisions, including all statutory requirements dealing with tariffing and entry and exit from the market, to mobile carriers.108 Regarding the entry and exit certifications required by section 214 of the Communications Act, the Commission stated:

[I]n a competitive market, application of Section 214 could harm firms lacking market power since certification procedures can actually deter entry of innovative and useful services, or can be used by competitors to delay or block the introduction of such innovations. The presence of Section 214 barriers to exit may also deter potential entrants from entering the marketplace . . . . [T]he time involved in the decertification process can impose additional losses on a carrier after competitive circumstances have made a particular service uneconomic and, if adequate substitute services are abundantly available, the discontinuance application is unnecessary to protect consumers.109

Other agencies have also used forbearance authority to perform a tailoring role. One example stems from Congress’s 1990 amendments to the Clean Air Act. Section 112 of the 1970 amendments required the EPA to create emission standards governing hazardous air pollutants from stationary sources; the standards had to “provide[] an ample margin of safety to protect the public health.”110 The EPA was also tasked with creating a list of hazardous air pollutants from stationary sources within ninety days of the passage of the 1970 amendments.111 Between 1970 and 1990, however, the EPA had regulated only seven air toxins (out of potentially hundreds) under section 112.112 As the Senate report accompanying the amendments noted, “[T]he law ha[d] worked poorly.”113

The 1990 amendments took the task of designating hazardous air pollutants out of the EPA’s hands by specifying an “initial list” of nearly two hundred such pollutants.114 At the same time, however, Congress recognized that its list may have been overinclusive—that is, certain chemical substances included on the initial list actually may not have posed a threat to public health. Congress thus empowered the EPA to “delete” any substance from the list of pollutants upon a showing that the substance “may not reasonably be anticipated to cause any adverse effects to the human health or adverse environmental effects.”115 “[A]ny person” may petition the EPA to delete (or add) a substance to the list, and the EPA is required to “either grant or deny the petition by publishing a written explanation” of the reasons for its decision within eighteen months.116 The EPA has occasionally exercised its authority to delete substances from the list.117 When it does so, such substances are no longer regulated under section 112.

C. Enabling Regulation

In each of the examples described above, an agency deregulates in some way: Congress set a proregulatory baseline that the Executive can then (to a greater or lesser extent) dismantle. That is not surprising, of course, because forbearance usually empowers agencies to eliminate or limit the scope of a statutory requirement.

In a somewhat paradoxical way, however, forbearance authority may also be used to enable regulation. Imagine the following scenario: an agency is tasked with deciding whether something or someone falls within the meaning of a statutory term. The statute is ambiguous in the Chevron sense, meaning that the agency therefore has policymaking discretion (within bounds) to decide the classification question either way.118 In many settings, the decision, if made in the affirmative, will automatically trigger a number of statutory requirements (let’s call them Requirements A, B, and C).119 Assume that the social benefits of applying Requirement A to the entity in question are net positive—yielding, say, one billion dollars in net present value. But if the net costs of applying Requirements B and C are sufficiently high, they may swamp those benefits. The rub is this: in most situations, the agency’s choice is all-or-nothing—that is, the agency can decide the initial classification one way or another, but once it does, it cannot pick and choose among the regulatory consequences of that decision. Thus, a rational agency faced with a situation in which Requirement A is socially beneficial but the costs of Requirements B and C outweigh those benefits will choose not to regulate at all.

Statutory forbearance authority may solve this dilemma. An agency armed with such authority can simultaneously make the classification decision in question—essentially flipping the “on” switch—while at the same time avoiding the undesirable regulatory consequences of doing so. For example, the agency in the above hypothetical could choose to apply A while at the same time negating B and C.

This is essentially what the FCC has done in the continuing controversy over so-called “net-neutrality” rules.120 Proponents of net neutrality seek to regulate the relationship between ISPs, like Comcast, and Internet content providers, such as Google or Netflix.121 By reclassifying broadband ISPs as Title II “telecommunications carriers,” the Commission placed net neutrality on sounder legal footing.122 But the decision to reclassify broadband ISPs also automatically subjected those providers to the full suite of “common-carrier” obligations contained in Title II of the Communications Act, including retail rate tariffing requirements.123 Many believe that a number of the common carrier obligations, such as tariffing requirements, should not apply to ISPs.124 Thus, proposals to reclassify broadband ISPs as common carriers have almost uniformly demanded forbearance from certain Title II rules, and in particular ex ante rate regulation, as a key component.125 And a large part of the Commission’s February 2015 Title II reclassification order was devoted to forbearing from such obligations.126 As the Commission explained, the order forbore “from 30 statutory provisions and render[ed] over 700 codified rules inapplicable” in order “to establish a light-touch regulatory framework tailored to preserving those provisions that advance our goals of more, better, and open broadband.”127

D. Reducing Uncertainty

Another use for a negative-lawmaking delegation in the hands of an agency is to reduce regulatory uncertainty while the agency makes (or defers) a decision on an issue with broad consequences. Recent scholarship has highlighted how agencies often wait to make decisions with important regulatory consequences.128 For example, prior to Massachusetts v. EPA,129 the EPA had studiously avoided deciding whether greenhouse gases were air pollutants requiring regulation under the Clean Air Act.130

One obvious cost arising from agency decisions not to decide, however, is regulatory uncertainty. Take the hypothetical described above, in which an agency is tasked with deciding whether something or someone meets a statutory definition, with an affirmative answer triggering three distinct regulatory requirements (A, B, and C). Now imagine that the agency declares that, at least for the time being, it is agnostic on the classification question—that is, it has chosen not to decide. The costs of choosing not to decide may not be great as long as the agency is the only relevant enforcer and it is clear that the agency will not enforce requirements A, B, and C until it has decided the underlying question. However, that often will not be the case. Frequently, whether regulatory requirements A, B, or C apply will be determined in the course of a legal dispute between private parties. And in those situations, courts—including district courts—may ultimately have to decide the classification question itself, subject to possible later override by the agency.131 Not only can this dynamic cause significant ex ante regulatory uncertainty, but it also may undermine one of the oft-cited rationales for the Chevron doctrine—namely, the value of uniformity in federal law.132

Now imagine that the agency could exercise forbearance authority and specify that certain requirements do not apply while also deferring decision on the underlying classification question. In other words, the agency in the above hypothetical could declare that Requirements A, B, and C (or just A, or B and C, etc.) lack legal force and effect while remaining mum on the broader classification issue. In this way, the agency could defer a large decision and engage in “administrative minimalism” while still maintaining national uniformity and reducing uncertainty, at least with respect to the application of requirements A, B, and C.133

The FCC has just this power under its forbearance authority, although that power had to be thrust upon the agency by litigants and the D.C. Circuit. The FCC had for years avoided deciding whether certain Internet Protocol-based services, including VoIP services providing two-way voice communication, are “telecommunications services” under the Communications Act; this determination carries with it an array of important regulatory consequences, including various common-carrier obligations under Title II of the Communications Act.134

It was in this context that in 2004, SBC Communications (now AT&T) petitioned the Commission to forbear from placing Title II requirements on “IP platform services”—including VoIP—“to the extent that such regulation might otherwise be found to apply.”135 The Commission rejected the petition on the procedural ground that the FCC’s forbearance authority did not allow it to forbear from obligations that could only hypothetically apply in the future.136 In other words, because it had yet to decide the initial question regarding whether IP services were telecommunications services, the Commission concluded that it could not forbear from specific regulatory requirements that would flow from an affirmative answer to that initial question.

The D.C. Circuit disagreed.137 It rebutted the Commission’s claim that conditional forbearance was never in the “public interest.”138 The court stated that “[p]arties petitioning for conditional forbearance seek elimination of regulatory uncertainty, and even the Commission recognizes that ‘regulatory uncertainty . . . in itself may discourage investment and innovation’ regarding the very technologies Congress intended the Act to promote.”139 In essence, the court determined that forbearance can be used to reduce uncertainty regarding the specific obligations of regulated parties while the agency decides (or defers) an issue of broader importance.

III. forbearance as delegation

Part II described the functions that a forbearance authority can serve. This Part asks a different question: is delegation of those functions to an agency normatively desirable?

Section III.A applies the traditional justifications for delegation to the special case of negative-lawmaking delegations. Those justifications apply rather straightforwardly in this different context. In particular, agencies’ greater expertise and flexibility likely make forbearance by agencies superior to action by Congress through legislative repeals or “sunset” clauses specifying an expiration date for statutory requirements.140

While Section III.A asks whether agency forbearance is superior to action by the other branches, Section III.B analyzes whether forbearance authority is superior to other regulatory tools that agencies possess, such as the power to selectively enforce statutes. An express forbearance authority may serve as a substitute for agency nonenforcement in a range of circumstances, and a variety of process considerations—including concerns about “underground” policymaking outside the normal accountability channels—make the exercise of forbearance authority normatively preferable as a policymaking device. Perhaps ironically, even opponents of delegation may prefer express forbearance authority to one frequently used alternative—detailed legislation coupled with de facto delegation to agencies of the power to tailor legislation through enforcement discretion.141

Section III.C addresses a number of objections that may be raised against agency forbearance authority. One important theme is that the charges leveled against negative delegations are simply variants on concerns raised with respect to more traditional positive delegations. These charges do not apply with any more force to negative delegations; in fact, several of those charges are considerably weaker with respect to negative delegations. Thus, once again, even opponents of delegation generally may have reason to tolerate forbearance authority as an alternative to more open-ended positive delegations.

A. Traditional Justifications for Delegation to Agencies

This Section applies the traditional justifications for delegation to agency forbearance authority. Because of the greater expertise and flexibility of agencies vis-à-vis Congress, agencies might be better at wielding the power to deprive statutory provisions of their legal force and effect.

1. Expertise-Information

Perhaps the classic justification for delegation is to harness agencies’ superior expertise and information vis-à-vis Congress.142 Agencies are assumed to know more about the likely effects of a given policy choice than either Congress or the courts.143 Thus, in the traditional positive-delegation model, agencies are tasked with a goal and allowed broad latitude to select the policy means to achieve that goal.

It is not difficult to see how agencies’ greater expertise and information might also justify delegations of the “negative” sort. Take the issue of statutory obsolescence. When Congress has made some effort to specify the details of a statute, and even when Congress has made a large one-time investment to acquire the information necessary to do so, the requirements Congress has fashioned may become unnecessary or counterproductive over time. Congress might assume responsibility for determining when statutes become outdated: it might trust a future Congress to repeal obsolete requirements, or it might specify ex ante that certain provisions will expire on a certain date. A better option, however, may be to delegate that obsolescence determination to an agency. Because agencies often have large staffs of experts devoted to studying a particular area,144 they are more likely to be able to make informed decisions about when statutory requirements have become obsolete. They likely are also better at assessing the often complex effects of lifting a statutory requirement.145

Agencies’ expertise and informational advantages also may make them better at tailoring statutes to cure overinclusiveness. As noted above, sometimes Congress realizes that its “initial” legislation is likely to require tailoring, and it delegates that task to an expert agency.146 Although Congress may have enough information to write a “first draft” of legislation147 and may prefer for any number of reasons that its draft be the default until the agency acts, Congress may also reasonably predict that an agency will have greater expertise regarding how the statute should be narrowed going forward.

With regard to the two other functions of negative delegations, enabling regulation and reducing uncertainty, it is less clear that Congress is ever in a particularly good position to perform these tasks in the first place. These two functions address situations in which old regulatory structures are applied in new or unanticipated ways. For example, Congress barely anticipated the rise of broadband Internet when it wrote the communications laws. As a consequence, both the definitions and requirements contained in the Communications Act apply rather crudely to broadband. In theory, Congress could intervene to create a more rational regulatory structure in such situations. But as will be discussed below, Congress’s crowded docket and lethargic pace make such an intervention unlikely.148

One might object at this point that the Executive’s decision to forbear often may not be motivated by expert judgment. Rather, we might predict that political opposition to a given statutory requirement is often what will fuel an agency’s exercise of forbearance authority.

The general point—that agency decision making might be influenced by political as well as expert judgment—cannot be denied.149 Denying Congress the ability to include forbearance provisions in its statutes is unlikely to be the right response, however, and may actually make the problem worse.150 The problem of politics in agency decision making—to the extent it is a problem—is one endemic to the administrative state, and not one specific or unique in any way to agency forbearance authority. Moreover, as argued below, including forbearance provisions in statutes may actually provide Congress a way to reduce the amount of slack between the enacting Congress’s wishes and the actions of the agency, at least as compared to traditional delegations.151 Finally, the increased opportunity that expressly delegated forbearance authority provides for judicial review and for monitoring by Congress—because, for example, it is subject to notice-and-comment requirements—reduces the risk of unchecked political influence compared to executive branch nonenforcement.152

2. Flexibility

Another common justification for delegation to agencies is their superior flexibility. As Pardo and Watts have noted, “[A]gencies are better able to adapt rules to respond to new information, different facts, or changed circumstances than Congress (which is constrained by political roadblocks and institutional barriers) or the courts (which are constrained by stare decisis).”153 Indeed, under prevailing administrative-law principles, an agency is provided broad discretion both to fashion the details of policy and to change or repeal those details.154

The flexibility rationale for delegation applies particularly well to negative delegations. First, the same factors that make agencies more likely to revise their own regulations in response to new information make them more likely to be able to revise congressionally specified requirements through forbearance. One might think Congress should respond vigorously to changed circumstances, but a number of factors may prevent it from doing so.155 Second, the fact that an agency can reverse or revise its decisions more easily than Congress means that an agency is more likely to fix an erroneous decision. This flexibility is particularly valuable when statutory requirements are eliminated in response to changed circumstances. After all, there may be uncertainty regarding whether the statute itself is responsible for the changed circumstances and a risk that the problems addressed by the statute will reoccur after the statutory requirements in question are rendered inapplicable.

Third, and relatedly, agencies’ greater flexibility can allow valuable experimentation. Recent work has shown that when laws can be more easily reversed, the optimal approach may be to experiment with policies in order to gain valuable information that can help policymakers formulate general rules.156 A forbearance authority allows an agency to experiment not with regulation but with deregulation. Of course, deregulatory experimentation carries its own risks. Agencies might, for example, experiment in ways that undo requirements that in fact remain necessary, resulting in backsliding.157 But while the risks associated with experimentation through forbearance must be taken seriously, the added flexibility afforded by forbearance nonetheless makes it a potentially valuable tool in the hands of agencies.

B. Advantages over Policymaking Through Enforcement

Section III.A focused on institutional choice, asking whether agencies, Congress, or courts were better at updating outdated statutes. This Section asks a slightly different set of questions. How does an express forbearance authority compare to other policymaking tools that an agency possesses and that serve similar ends? And should we prefer agencies exercising power pursuant to an express forbearance authority to them shaping statutes through more informal means, especially nonenforcement policies?

Recent scholarship has highlighted the extent to which executive nonenforcement of federal statutes can substitute for more explicit forms of policymaking.158 Take the area of immigration. The immigration laws resemble the tax code in the number of congressionally specified details they contain.159 However, as Adam Cox and Cristina Rodríguez have explained, because the Executive cannot (due to resource constraints) possibly remove all “formally deportable” noncitizens, the system provides the Executive broad discretion to shape the law through enforcement policies.160 Indeed, Cox and Rodríguez maintain that this ability to shape immigration policy through enforcement practices amounts to a “de facto delegation of power” from Congress to decide which particular persons (or classes of persons) stay in the country and which do not.161

This Section examines the process reasons we might prefer agencies to act pursuant to an express forbearance authority rather than pursuant to nonenforcement.162 One assumption this Section makes is that action pursuant to an express forbearance authority can reasonably substitute for nonenforcement practices, such that the delegation of express forbearance authority to agencies will, at least in some cases, lead to less reliance on nonenforcement as a policymaking tool. Although agencies debating whether to act through one or the other means may still have incentives to rely on nonenforcement, perhaps particularly when they want to shield their actions from public scrutiny or judicial review, agencies also have strong incentives pointing in the opposite direction. For one, action taken pursuant to an express delegation is more durable and more likely to carry over into the next administration. Thus, agencies looking to “lock in” policies may prefer to act through delegated rulemaking authority. In addition, as Justice Scalia recently explained, the power to deprive statutory requirements of their legal force is stronger than simple nonenforcement because it can stop private parties who otherwise hold private rights of action, and not just the agency, from enforcing statutory requirements.163 The following discussion will therefore assume that agencies sometimes will act, if given the opportunity, through an express negative delegation when they otherwise would have pursued similar ends through nonenforcement.

1. The Take Care Clause

One advantage of agency action pursuant to an express forbearance authority is its greater consonance with the Constitution’s Take Care Clause, which provides that the President “shall take Care that the Laws be faithfully executed.”164 Scholars have argued that when the Executive uses its enforcement discretion in a manner that is too categorical or policy-based, it violates the Take Care Clause.165 On this view, while the Executive has broad discretion to decide whether to enforce statutory requirements in individual cases, decisions not to enforce a statute prospectively for entire categories of violations go too far, effectively amending the laws that Congress has passed.

Whatever one thinks of these claims, the important point is that an express congressional delegation of authority to deprive statutory requirements of legal force and effect for certain categories of cases solves any dilemma posed by the text of the Take Care Clause.166 That is because the negative delegation is endogenous to the statute itself. Executive action pursuant to such a delegation therefore is not “unfaithful” to the law Congress has passed.167 Instead, an agency exercising forbearance authority pursuant to a congressional statute is faithfully executing that statute as written.

2. Visibility

A second reason that we might prefer agency action pursuant to an express forbearance authority is that it provides greater opportunities to monitor agency activity. One persistent criticism of the administrative state, and particularly of broad delegations, is that policy is frequently formulated through allegedly less transparent means.168 Less visible policymaking may lead to less accountable government.169 Visibility also may be linked to other good government values invoked by opponents of delegation, such as the rule of law.170

But agencies themselves can act in more or less visible (and thus accountable) ways.171 And one function of administrative law is to encourage agencies to announce policies in ways that the people and their elected representatives can more easily monitor.172

Agency action pursuant to an express forbearance authority is much more likely to be visible. Several of the delegations discussed above specifically require the agency to invoke forbearance authority by rule or regulation, which in turn requires the agency to publish its actions in the Federal Register.173 And in every one of the above examples of agencies using their forbearance authority, the agency has done so following a notice-and-comment process.174 Such processes are likely mandated by the APA because negative delegations prospectively alter substantive legal obligations.175

A notice-and-comment process resulting in a formal regulation is, for obvious reasons, a much more visible form of policymaking than is statutory nonenforcement. Following the FCC’s May 2014 Notice of Proposed Rulemaking on net neutrality, the agency received almost four million comments.176 Although that proceeding is surely something of an outlier, it shows the potential of notice and comment to bring regulatory issues into the public discourse.177 Less dramatically but probably more importantly, notice-and-comment processes provide information to political insiders, including elected politicians. Influential work by Mathew McCubbins, Roger Noll, and Barry Weingast, for example, has argued that such processes contribute to a “fire alarm” system of oversight in which interested parties can more easily monitor agency behavior and alert Congress (or the President) to potential problems.178 This contrasts with more costly forms of oversight such as congressional committee investigations, in which politicians must acquire information on their own.179

3. Arbitrariness

Another objection lodged against the modern administrative state is that agency policymaking encourages “arbitrary” government decision making.180 The core of the concern is that delegation allows policy to be formulated outside the deliberative structures set up by the Constitution. These deliberative mechanisms were once thought to ensure that the government acts with public purposes in mind.181 Administrative arbitrariness, defined by the government’s failure to act rationally and according to public purposes, is thus one threat traditionally associated with delegation.182

The primary way American law deals with administrative arbitrariness is through various subconstitutional rules derived from the APA and enforced through judicial review. Most importantly, agencies proceeding through notice-and-comment rulemaking generally must (a) provide reasons for their decisions, including a response to any significant objections;183 and (b) engage in “reasoned-decisionmaking” such that the policies they adopt are at least minimally rational.184 They also must base their decisions on the factors made relevant by Congress.185 These requirements seek to discipline agency decision making and promote public-regarding (i.e., nonarbitrary) action by agencies.186

Agency nonenforcement decisions generally are not subject to judicial review and thus do not have to satisfy the requirements of reason giving and rationality that apply to other categories of agency action.187 In Heckler v. Chaney, the Supreme Court held that agency nonenforcement decisions are presumptively unreviewable under the APA.188 That presumption has limits,189 but those limits are unlikely to play a meaningful role in cabining agency action.190 In addition, even apart from the APA, constitutional standing requirements often prevent members of the public from challenging agency nonenforcement decisions.191

By contrast, and as explained above, agency action taken pursuant to a forbearance authority will result in a “rule” under the APA. Judicial review of such a rule is presumptively available.192 The FCC’s forbearance decisions, for example, have been routinely subject to judicial review under the normal “arbitrary and capricious” standard.193

4. Cost-Benefit Analysis

Judicial review is not the only way to discipline agency decision making. Since the Reagan Administration, OIRA, which is part of the Office of Management and Budget (OMB), has subjected all significant rules promulgated by executive branch agencies to cost-benefit review.194 The original impetus for centralized cost-benefit review was distrust of agency power and, in particular, “a fear that, if left unchecked, regulatory agencies would consistently regulate ‘too much.’”195 But recent work also has highlighted how cost-benefit review might usefully be applied to agency deregulation or inaction.196 More generally, review of agency rules by officials closer to the President can be seen as a way of monitoring agency behavior and mitigating any number of pathologies that could lead agencies to act in undesirable ways.197

Cost-benefit review is much more easily applied to agency forbearance decisions than to agency deregulation through nonenforcement.198 The former will usually result in a discrete rule, the effects of which can be isolated and studied just as with any other rule. Agency enforcement practices, by contrast, are not so easily policed. Even concerted efforts to reduce enforcement of a particular requirement will often take the form of a series of discrete (though related) decisions not to prosecute.199 In these kinds of nonenforcement actions, it is unclear how costs and benefits could be measured in the first place. The relative lack of visibility associated with nonenforcement practices compounds this difficulty, making it easier for agencies to hide policymaking from potential reviewers. Thus, by shifting agency policymaking away from nonenforcement, statutorily authorized forbearance may make it easier for the executive branch to engage in cost-benefit review of administrative deregulation.

C. Addressing Concerns

This Section addresses some possible critiques of forbearance authority. An overarching theme of the Section is that the objections lodged against negative delegations are just variants on traditional concerns associated with more familiar types of delegation. In many cases, these objections in fact apply with less force to negative delegations than to positive ones. Thus, those already comfortable with delegation to agencies generally should have no particular reason to fear the growth of forbearance authority. More interestingly, opponents of broad delegations to agencies may actually prefer forbearance-type delegation in a world—such as our own—where positive delegations are “here to stay.”200 Opponents of delegation generally should support negative delegation as a second-best solution because banning negative delegations may simply increase the law’s already heavy reliance on positive ones.201

1. Agency Power

One objection to an agency forbearance authority is that it simply places too much power—or perhaps power of a too dramatic sort—in the hands of an agency. As noted above, critics of forbearance have called such negative delegations “astonishing” and have questioned whether allowing agencies such a power comports with the constitutional separation of powers.202 This criticism echoes Clinton, where the Court described the ability to deprive statutory requirements of legal force and effect as a power “to enact, to amend, or to repeal statutes” that the Constitution does not allow the Executive to wield.203

But if one steps back and considers more carefully what is involved in forbearance, the power is not at all “astonishing.” In many ways it is a tamer version of the traditional, positive-type delegation that is firmly established as a matter of constitutional law.

First, as a formal matter, to say that forbearance authority allows a legislative “repeal” (or “modification” or “unmaking”) of the statute Congress has passed fundamentally mischaracterizes forbearance. Since the negative-lawmaking delegation is part of the statute, action pursuant to such a delegation is an implementation of the statute, not a repeal or modification of it. Here is a slightly different way to approach the point: when Congress simultaneously enacts a requirement and makes it defeasible at the agency’s discretion, the characteristic of defeasibility already has been built into the requirement as passed. It is thus incorrect to posit that when an agency uses its authority, it is engaged in a “repeal” of the statute that requires the constitutional processes of bicameralism and presentment.204

It still could be argued that, at least as a functional matter, an express forbearance authority is simply “too big” an authority to vest in an agency. But is that right? Certainly it is correct that, from the standpoint of agency power, a statute containing a negative-lawmaking delegation allows the agency more power than a relatively detailed statute that does not contain such a delegation. However, that is often the wrong comparison, for Congress will always have the option of instead enacting a garden-variety positive-type delegation that specifies few, if any, details in advance. In terms of agency power, negative-lawmaking delegations in fact occupy a middle ground between completely specified statutes and more open-ended positive delegations. Recall, for instance, the 1993 mobile-wireless amendments to the Communications Act.205 Congress could have written a statute providing that the FCC “shall have the authority to regulate commercial mobile services to promote competition and protect the public interest.” What Congress did instead was to presumptively apply a number of requirements—such as entry and exit certifications and tariffing obligations—to commercial mobile services and then allow the agency to forbear from applying some of those requirements by regulation. As a result, the agency had at its disposal a narrower set of options than if Congress had allowed the Commission to regulate on a blank slate.

2. “Unmaking” Deals

Another objection to negative delegations is that such delegations allow an agency to unmake deals that were necessary to strike the legislative compromise that resulted in the requirement at issue.206 This objection similarly crumbles on closer inspection.

First, as a purely formal matter, to say that agency action pursuant to an expressly delegated authority undermines the deal struck by Congress makes little sense. After all, the “deal” includes the negative delegation. The risk that the agency would use the delegation as authorized was built into the “deal.” The objection appears to assume that the compromise is reflected in the specific language of the requirement and that the delegation is somehow external to that compromise.207 But the delegation is just as much a part of the overall statute as any other statutory language, and one cannot assume that prohibiting Congress from making the delegation of forbearance authority would have resulted in the same statute otherwise. Quite the contrary. In the absence of agency authority to forbear from a particular requirement, the resulting statute might have been less favorable toward the interests that lobbied in favor of that requirement. At the extreme, the statute might not have even passed.

Perhaps, however, the objection reflects a more nuanced concern along the following lines. Every delegation involves the risk of what political scientists refer to as bureaucratic drift in which the agency pursues outcomes different from those preferred by the principal (here, Congress).208 This risk may be heightened for negative-lawmaking delegations, as agents can “cancel” the instructions they receive from Congress.

But the intuition that negative delegations create greater principal-agent problems than positive ones is empirically untested and, for reasons similar to those discussed in Section III.C.1,209 unlikely to be true. Forbearance authority is in important ways a more modest power than that more typically given to agencies. Most important here, when Congress delegates forbearance authority, it still sets the default regulatory regime, including the presumptive requirements that apply to regulated entities. The agency thus has the burden of inertia to depart from that regulatory regime, which, as argued above, must be done through notice-and-comment processes. Moreover, just as in the traditional case, Congress has the option to write a more or less stringent intelligible principle to govern the exercise of agency discretion, and may place the burden of proof on the agency (or petitioners) to satisfy that principle.210 Congress can thus monitor and assess any departures from its baseline regime and respond accordingly. All of this is reason to think that, as compared to the positive type, negative delegations may actually mitigate rather than exacerbate the principal-agent problems arising from delegation.

3. Congressional Shirking and Accountability

A third possible objection to negative-lawmaking delegations is that such delegations allow Congress to shirk its responsibility and hide potentially unpopular policymaking from public view. One form of this argument has appeared in the popular press and takes the following general form: negative delegations allow Congress to write laws that appear tough (and thus, by hypothesis, please the public).211 However, those laws are a mirage because members of Congress (and, perhaps, powerful regulated entities) know that the law will be implemented only partially. The agency tasked with implementing the statute will, by exercising its delegated authority, grant exceptions and lift statutory requirements. Because such agency action is—again by hypothesis—less visible, lines of accountability will be severed and democracy will suffer.

This argument is really just a reframing of the conventional critique of delegation—associated with David Schoenbrod212—that delegation allows Congress to engage in “happy talk” about the laws it passes while at the same time handing agencies the obligation to make hard choices.213 This argument is also open to the same responses. Most persuasively, under the assumptions favored by critics of delegation, voters could still hold legislators responsible for the decision to delegate in the first place.214 A vote for legislation containing a delegation (including a “negative” one) signals Congress’s judgment that the delegation is in the public interest. If that turns out to be wrong, voters can punish those who supported the law. As Eric Posner and Adrian Vermeule have argued with respect to traditional positive delegations:

If the agency performs its function poorly, citizens will hold Congress responsible for the poor design of the agency, or for giving it too much power or not enough, or for giving it too much money or not enough, or for confirming bad appointments, or for creating the agency in the first place.215

And if voters are inattentive to the details of legislation, there is no reason to think they are more easily duped by happy talk concerning delegatory legislation than any other kind of legislation.216

Negative delegations also may fare better than positive ones when assessed in terms of the potential for congressional shirking. When Congress legislates with specificity while allowing an agency to forbear from implementing some of those detailed requirements, at least Congress is the one responsible for setting the default policy.217 Often this is not true with traditional delegation, where Congress sets only a goal or principle and leaves agencies to fill in the details.218 Thus, it is quite plausible that even “provisional” legislation that establishes a default policy can supply voters with more information about Congress’s performance than more open-ended legislation.

One might still object that the act of delegating forbearance authority inherently involves the transfer of power from Congress to agencies. To the extent Congress is seen as the more accountable actor, the critique has force and echoes a common concern with delegation generally.219 However, the point remains that there is no reason to fear greater accountability losses with respect to forbearance delegations than any other kind of delegation. In fact, for the reasons just given, forbearance delegations involve more of a commitment by Congress than the ordinary case.220

One further point bears mentioning. If Congress legislates with specificity and does not delegate forbearance authority to an agency, it increases the likelihood that courts might help themselves to such power. Guido Calabresi argues that judges should rather explicitly “overrule” statutory provisions that have become obsolete, much as a common-law judge would overrule a precedent.221 Less jarringly, William Eskridge suggests that judges should “update” statutes to reflect changed circumstances through a process of “dynamic statutory interpretation.”222 Of course, agencies have an undoubted accountability advantage over courts.223 Thus, when deciding whether power should be lodged in an agency or, as a practical matter, fall instead to the courts, accountability concerns should generally favor the agency approach.

4. Capture

Another objection to agency forbearance authority is that it facilitates agency capture, which can be broadly defined as agency action for the benefit—and at the behest of—a special interest, often industry.224 This has sometimes been expressed as a problem of “favoritism.”225 Another proregulatory variant on the objection is that forbearance allows the Executive to achieve a kind of backdoor deregulation, in which statutory requirements designed to benefit the public are eroded over time.226

Putting aside whether there is a problem of agency capture generally,227 the capture critique of negative delegations is unpersuasive on several counts. First, to the extent that it treats deregulation as synonymous with industry interest, the critique is mistaken. Indeed, modern capture theory in its original form stressed how industries used regulation itself—not the lack of it—to advance their interests.228 Thus, the prescription for solving agency capture was often deregulation. On this view, to remedy capture, agencies should abolish regulations that powerful parties could use to suppress competition and innovation, such as entry requirements and tariffing obligations.229 These are exactly the sorts of regulations that the FCC, one of the few remaining industry-specific competition agencies, has targeted with its forbearance authority.

More recent analyses have accepted that capture can lead to agency inaction or deregulation as well as to regulation.230 Thus, negative delegations might seem to grant agencies a potent tool for engaging in action motivated by this “corrosive” form of capture, eroding not only the agency’s own administrative regulations, but also statutory requirements.231

This argument has a ring of plausibility, but whether negative delegations facilitate corrosive capture is uncertain. Negative delegations may actually combat some of the ills associated with such capture. Imagine a relatively public-spirited legislature that wants to protect the public from air pollution but is unsure of exactly how to do so.232 Such uncertainty is a classic reason the legislature might delegate policymaking authority to an expert agency. However, the legislature may also fear that the agency will underregulate, perhaps because of capture. One option in such circumstances is to write stringent but crude requirements and allow the agency to depart downwards—that is, to grant the agency forbearance authority. That option allows the legislature to set relatively heavy-handed default rules and place the burdens of inertia, explanation, and proof on the party seeking relief from those requirements; at the same time, the legislature can preserve some of the benefits associated with agency expertise.233 From the perspective of agency capture theory, such an option appears to be superior to delegations of the positive variety, which allow agencies to achieve underregulation through mere inaction.234

Such a dynamic actually appears to have motivated Congress in its efforts to address hazardous air pollutant (HAP) regulation.235 Earlier environmental-regulation regimes allowed the EPA to determine which pollutants would be regulated as HAPs, but the agency had been slow to identify pollutants as HAPs.236 In response, Congress switched the default rule, subjecting a laundry list of pollutants to regulation while allowing the EPA to remove any of those pollutants from the list.237 The resulting regime has not worked perfectly.238 However, it does show the potential for forbearance to facilitate tougher, rather than weaker, legislation and to combat the potential for corrosive agency capture.239

There is always a concern that an agency may use its delegated authority to advance special interests, just as there is a concern that Congress will use its legislative authority to do so. But negative delegations to agencies do not necessarily present a special challenge in this regard. Indeed, in the right circumstances, negative delegations may be able to reduce the effects of capture on the administrative state.

IV. applications

The prior Parts have sought to establish the general propriety of forbearance-like delegations. They have argued that, as a normative and legal matter, there is nothing particularly troubling about forbearance as opposed to the more familiar agency authority to fill in the details of a statutory scheme. This Part asks a slightly different question: as a policy matter, when should Congress include an administrative forbearance authority? The guidance provided is tentative and incomplete; after all, the Article suggests throughout that Congress generally should be able to decide when to include negative delegations in statutes, just as it does with “normal,” positive delegations. Nevertheless, as with any type of delegation, there are good and bad reasons that Congress might give agencies forbearance authority. Section IV.A briefly considers some of those reasons. Section IV.B then fleshes out the analysis by showing how forbearance authority might have helped resolve two recent regulatory controversies.

A. The Promise (and Limits) of Forbearance

1. When Should Congress Give Agencies Forbearance Authority?

The most obvious context in which forbearance might be beneficial is when Congress writes statutory requirements, expects that the need for those requirements may recede over time, and predicts that an agency will have better information about—and be more likely to adjust—the statutory framework as needed. Because Congress may be uncertain about when or even whether statutory obligations will become obsolete, other statutory tools—such as sunset provisions—might be too crude. In the 1996 Telecommunications Act, for example, Congress had hoped that competition in local telephone markets might someday obviate the need for heavy-handed, prescriptive regulation, but it did not know when that day would come, or if it ever would. The FCC seemed better positioned to make that determination. This example highlights that in substantive areas of law characterized by rapid technological change, statutes might be particularly likely to become outdated.240 However, it would be a mistake to conclude that statutory obsolescence is a challenge unique to the regulation of high-tech markets. Indeed, it may be an enduring feature of the administrative state.241

Congress also might beneficially use forbearance when it believes that the statutes it writes are currently overbroad but wants—for one reason or another—to set a proregulatory baseline instead of letting an agency build from the ground up. The legislative response to HAPs is an instructive example of this approach.242 Congress tried letting the EPA craft the list of pollutants to be regulated as hazardous but, because of capture or otherwise,243 the agency did a bad job.244 Congress then flipped the baseline by specifying which pollutants were to be regulated, and it allowed the agency to subtract from the list if the substances included did not actually pose health risks.245

Congress might also anticipate that future developments or information might subject new entities to regulation under existing statutory requirements that were not designed to address these novel entities. As discussed above, granting an agency a forbearance option can potentially enable sound regulation in such circumstances and can also allow the agency to deal with uncertainty that arises while it decides what to do.246 The Congresses that wrote the Communications Act only dimly foresaw the rise and importance of broadband Internet, and so it is not surprising that the Act as written applies rather crudely to it.247 The history of greenhouse-gas regulation under the Clean Air Act, discussed below,248 provides another example. Giving forbearance power to an agency preserves some amount of flexibility to adapt old regulatory structures to changing environments.

All of the above scenarios may be more likely to occur in statutes where a threshold definitional question—say, whether a product is a “drug,” a substance is a “pollutant,” or a service is “telecommunications”—triggers broad statutory consequences. Because so much turns on a single determination, such situations likely pose a threat of overbreadth that forbearance may mitigate. It is thus not surprising that many of the existing examples of forbearance are contained within such statutes.

2. The Limits of Forbearance

Finally, a word about when we might be wary of Congress including a forbearance-like delegation. With any kind of delegation, Congress sometimes may not appear to be harnessing the advantages associated with agency decision making; instead, it may be engaged in political buck-passing “in an attempt to shift responsibility for the negative impacts of law to other governmental branches.”249 That buck-passing concern raises a question: in what situations might we be more skeptical of Congress including a forbearance provision in a statute? At least two dimensions to the answer warrant discussion here.

First, as a matter of constitutional law, this Article’s analysis suggests that judges should rarely strike down forbearance provisions. Indeed, judges should invalidate negative delegations no more frequently than they invalidate traditional positive delegations, which are almost always upheld because the strong form of the nondelegation doctrine is essentially moribund.250 The dilution of the nondelegation doctrine has occurred because of judicial administrability concerns that disrupt any clean attempt to separate delegations passed for good reasons from those passed for bad251 or to judge whether a particular delegation is simply too sweeping for purposes of the intelligible-principle doctrine.252 Thus, outside of “extreme cases” where there appears to be no good reason to delegate,253 courts should not nullify forbearance delegations as long as they contain an intelligible principle, which can be quite broad.254

Second, however, to say that courts should not strike down forbearance provisions is not to say that forbearance provisions do not raise normative concerns in some circumstances. For one, the judicial administrability concerns just mentioned may leave certain norms underenforced.255 Underenforcement is especially concerning where, based on the breadth of the forbearance provision itself and the surrounding circumstances, there does not appear to be a nexus between the reasons for the forbearance decision and the underlying purposes of the statutory requirements from which the agency will forbear. Suppose, for example, that in a future Republican administration, a Republican Congress were to pass the “Obamacare Forbearance Act,” providing that the President (or an executive branch agency, such as the Department of Health and Human Services) could permanently forbear from any provision of the Affordable Care Act, provided that forbearance was in the public interest.256 Imagine further that circumstances demonstrated that support for the forbearance provision was fueled by political opposition to the law itself, and not by a belief that the law’s purposes would be effectuated best by allowing an agency to tailor the Act’s requirements. In such circumstances, we might be concerned that the forbearance provision in question was a kind of backdoor repeal that had been designed merely to gain some relative political advantage.257

This disjunction between the constitutionally permissible and the normatively desirable suggests a role for subconstitutional rules of interpretation. Indeed, scholars have long invoked certain “nondelegation canons” to deal with normative issues raised by broad delegations to agencies.258 Two rules of interpretation may be particularly useful for forbearance-like delegations.

The first rule of interpretation underlies the Supreme Court’s opinion in MCI Telecommunications Corp. v. AT&T: an agency may not exercise forbearance authority unless Congress has granted it that authority using relatively clear language.259 That is, a forbearance power may not be lightly implied, at least where forbearance would result in a “fundamental revision” of the regulatory scheme enacted by Congress.260 Although MCI was murky about the basis for this rule, this Article provides some arguments in its favor. When Congress expressly includes a forbearance provision in the statute, the “unmaking deals” and principal-agent objections discussed above are considerably weaker, as Congress’s express inclusion of the provision makes clear that it intends the forbearance power to be part of the legislative package. Requiring Congress to be explicit about the power it gives to the agency also undercuts the “backdoor-repeal” objection by making clear to outsiders that Congress may in fact be authorizing the dismantling of a particular law.

The second rule of interpretation relates to the nexus between the reasons to forbear and the underlying purposes of the statute. If possible, when reviewing agency forbearance decisions, courts should presume that Congress intends the agency to consider the underlying purposes of the statute when deciding whether to forbear.261 Of course, when Congress has required the forbearing agency to consider a list of specific factors that relate to the purposes of the statute in question, this likely will not be difficult. For example, the FCC’s forbearance provision requires the Commission to consider factors related to competition and consumers, the twin aims of most of the Communication Act’s protections.262 But if Congress speaks in broader language—say, allowing an agency to forbear when it is in the “public interest”—courts should still presume that the agency’s decision should be guided (though perhaps not exclusively) by the purposes underlying the statute. Such a presumption would require agencies to articulate more clearly how their decisions advance the goals of the statute as a whole. It would minimize the dangers associated with a runaway Executive negating the will of Congress. And it would reduce the risk that Congress and the Executive could act in tandem to achieve a “backdoor repeal” of a duly enacted statute.

B. Examples

This Section imagines the role a forbearance authority might have played—had one been held by an agency—in two recent controversies: first, the debate over the continued validity of the VRA’s preclearance provisions, and second, the debate over the EPA’s regulations regarding greenhouse-gas emissions. A statutory forbearance power could have led to better regulatory outcomes in both controversies and might have saved the programs in question from (partial) judicial invalidation.

1. The Voting Rights Act

Passed in 1965 as a response to decades of black disenfranchisement, the VRA has several key provisions.263 The Act categorically bars voting practices intended to deny voting rights on the basis of race.264 It also suspends certain kinds of “test[s] or device[s]”265 in “covered jurisdictions,”which were determined according to a formula contained in section 4(b).266 As relevant here, section 5 of the Act subjected those covered jurisdictions to a “preclearance” regime.267 Under that regime, covered jurisdictions were required to seek permission for changes to their voting laws from the Department of Justice or a three-judge panel of the U.S. District Court for the District of Columbia; permission would be granted if the covered jurisdiction demonstrated that the change had neither the purpose nor effect of abridging the right to vote based on race.268

Section 4(b) of the original 1965 Act determined which jurisdictions were “covered” based on whether they had been using a forbidden test or device and had less than fifty percent voter registration or turnout in the November 1964 election.269 As originally enacted, the preclearance provisions of the VRA were set to expire in five years,270 reflecting the fact that, in the Supreme Court’s words, they were from the beginning “intended to be temporary.”271 In addition, and as fleshed out by subsequent amendments, otherwise covered jurisdictions could “bail out” of section 5’s requirements under certain conditions.272

Congress reenacted the VRA in 1970, 1975, 1982, and 2006, extending the preclearance provisions long past their original five-year expiration date.273 The 1970 and 1975 reauthorizations updated the Act’s coverage formula to include election data from the 1968 and 1972 elections, but they otherwise left the formula unchanged.274 Subsequent reauthorizations did not alter the coverage formula either.275 By the late 2000s, covered jurisdictions—which included much of the South—were still determined based on their characteristics some forty years prior.

Scholars have found much to criticize about the Supreme Court’s decision in Shelby County v. Holder, which concluded that the preclearance scheme—or, more accurately, the statute’s coverage formula—had become unconstitutional.276 What cannot be doubted, however, is that America’s racial politics looked much different in 2013 than in 1965.277 This fact mattered to Shelby County’s five-member majority. As pointed out by Chief Justice Roberts, in many jurisdictions still covered by section 5, black and white voting rates were near parity.278 And some evidence suggested that at least as of 2004, voting disparities were greatest in certain northern states that were not required to seek preclearance.279 For the majority, the fact that Congress had, in the face of these facts, acted “as if nothing had changed” sealed the Act’s fate.280 As the Court explained, “Our country has changed, and . . . Congress must ensure that the legislation it passes to remedy that problem speaks to current conditions.”281

The difficult counterfactual is whether a more agency-focused scheme—one that involved, at least in part, a congressional delegation of the power to remove jurisdictions from section 5’s scope—would have done a better job addressing the concerns that ultimately led the Court to invalidate the VRA’s coverage formula.282 The germ of such a solution is already found in the Act’s “bailout” mechanism.283 But that mechanism, besides lodging bailout authority in the courts and not in an agency, proved too rigid to provide meaningful relief in most cases.284 There are reasons to think that a more flexible delegation, in place of the bailout mechanism and the continued cycle of congressional reauthorization, might have blunted some of the main criticisms of the Act. The main impediment to changing the coverage formula itself was political—namely, a fear that “a debate over the coverage formula . . . would have led to the complete unraveling of the bill.”285 Instead of reauthorizing the Act for a finite period of time and including a bailout mechanism, however, Congress instead could have reenacted the Act and given an agency authority to designate a covered jurisdiction as no longer subject to the preclearance process. This amendment might have allowed Congress to sidestep the thorniest debates about the VRA while also creating a mechanism to address the coverage formula’s purported overinclusiveness in light of changed circumstances.286

By granting an agency the power to alter the coverage formula, Congress could have potentially captured some of the benefits associated with agency decision making surveyed above.287 An agency tasked with forbearance authority could have naturally developed expertise bearing on whether the preclearance regime had become unnecessary or counterproductive in a particular locality. An agency-focused solution also would have been more flexible than either the sunset-and-reauthorization regime, which depended on Congress, or the Supreme Court’s ultimate nullification of the Act’s entire coverage formula. With forbearance authority, the agency could have more quickly reversed course and reimposed preclearance requirements if backsliding occurred.

2. The EPA and Climate Change

Forbearance authority also might have played a critical role in the controversy over the EPA’s regulation of greenhouse-gas emissions. That saga began with the Supreme Court’s decision in Massachusetts v. EPA, which held (among other things) that greenhouse gases count as “air pollutants” for purposes of the Clean Air Act’s mobile-source (i.e., motor-vehicle) provisions.288 On remand several years later, the EPA ruled that greenhouse-gas emissions from mobile sources “may reasonably be anticipated to endanger both public health and welfare,”289 a determination many saw as the inevitable result of the Court’s holding in Massachusetts.290 As soon as the EPA made an endangerment finding, as it did on remand, emissions of greenhouse gases automatically became subject to a comprehensive set of rules governing emissions of such air pollutants from automobiles.291

The EPA’s regulation of mobile-source emissions also triggered the regulation of greenhouse gases under several other parts of the Clean Air Act, two of which proved especially relevant.292 First, under the Prevention of Significant Deterioration (PSD) program, stationary sources emitting large amounts of pollutants regulated elsewhere under the Act are subject to onerous preconstruction permitting requirements.293 Such sources include “any stationary source with the potential to emit 250 tons per year of ‘any air pollutant’ (or 100 tons per year for certain types of sources).”294 Second, Title V of the Act requires “major sources”—defined as sources having the potential to emit one hundred tons or more per year of “any air pollutant”—to obtain comprehensive operating permits295 that “facilitate compliance and enforcement by consolidating into a single document all of a facility’s obligations under the Act.”296

The EPA’s determination that greenhouse gases were regulated pollutants for purposes of mobile-source regulation thus required the agency to regulate greenhouse gases under the PSD and Title V provisions as well, or so the agency ruled.297 But the EPA faced a problem. Because greenhouse gases are emitted at much higher levels than are “classic” pollutants, the statutory triggers for the PSD and Title V programs would pull in an enormous number of new stationary sources—estimated at over six million, including many smaller, nonindustrial sources—resulting in “undue costs for sources and impossible administrative burdens for permitting authorities.”298 The EPA thus decided to “tailor” those triggers, subjecting only those otherwise unregulated sources emitting at least one hundred thousand tons of greenhouse gases to eventual regulation under the PSD program and Title V, with the possibility of lowering that threshold at a later date.299 That determination exempted many sources facially subject to regulation under the Clean Air Act from the Act’s scope.

The appeal from the EPA’s greenhouse-gas regulations eventually reached the Supreme Court in Utility Air Regulatory Group (UARG) v. EPA.300 In an opinion authored by Justice Scalia, who had dissented in Massachusetts, the Court vacated the EPA’s treatment of stationary sources that would have been unregulated under the Act but for their greenhouse-gas emissions.301 The Court reasoned that the EPA was correct that regulating greenhouse gases from sources emitting amounts exceeding the statutory thresholds would have created intolerable and indeed absurd results.302 However, the Court said, that did not mean that the EPA had authority to “tailor” those thresholds as it saw fit. Indeed, Justice Scalia described this kind of tailoring authority as a power “to alter [the Act’s] requirements and to establish with the force of law that otherwise-prohibited conduct will not violate the Act.”303 The Court then suggested that the intolerable and absurd results stemming from EPA’s regulatory efforts meant that the EPA’s interpretation of air pollutant as encompassing greenhouse gases in the context of the PSD and Title V programs was unreasonable.304 The Court thus invalidated that interpretation, finding that the EPA went beyond its statutorily delegated authority in reading the phrase air pollutant—at least as used in the Clean Air Act’s PSD and Title V provisions—to include greenhouse gases.305

UARG was a curious case. In Massachusetts, the Court heldthat its interpretation of air pollutant was mandated in the context of mobile-source regulation, but that interpretation was affirmatively forbidden in other statutory contexts.306 Nevertheless, the EPA’s chosen solution—effectively rewriting the statute to exempt certain sources—suffered from its own infirmities, most notably the lack of any clear source of statutory authority to do so. As Justice Scalia pointed out, an agency’s general powers do not “include a power to revise clear statutory terms that turn out not to work in practice.”307

A forbearance-like statutory authority to remove the Clean Air Act’s requirements for certain categories of sources, had one been available, might have provided a straightforward solution to the conundrum faced by both the EPA and the courts following Massachusetts. Indeed, the Court’s opinion in UARG can be read to endorse just such a power.308 By wielding it, the EPA could have simultaneously recognized greenhouse gases to be “pollutants” wherever that term appeared in the statute while blunting the extreme results that rendered this interpretation unreasonable in the Court’s eyes. The agency thus could have created a regulatory regime that was more rational, coherent, and protective of the environment than the one left in the wake of UARG’s holding.

Conclusion

This Article elaborates the normative and constitutional case for administrative forbearance authority, which empowers agencies to deprive statutory provisions of their legal force and effect. It makes two principal contributions to the growing literature on such authority. First, the Article describes how forbearance might be used and the various functions it might perform, drawing on examples from agencies that already have a forbearance power. Second, from that descriptive account, the Article sketches a normative and constitutional defense of agency forbearance authority. That analysis is necessarily comparative. It asks whether the traditional justifications for delegation of positive policymaking authority apply to negative delegations, as well. As this Article suggests, the answer is yes. Forbearance also might substitute for other forms of agency action, such as nonenforcement practices, that are for various reasons more troubling. Finally, the Article argues that the criticisms that may be lodged against forbearance authority are actually just generic objections to agency delegations and, in many cases, apply with less force to delegations of the negative type. Thus, where the alternative is delegation of the more open-ended, positive variety, even critics of delegation generally should find some comfort in administrative forbearance.