My name is Andrew Grossman. I am an Adjunct Scholar at the Cato Institute and a partner in the Washington, D.C., office of Baker & Hostetler LLP. The views I express in this testimony are my own and should not be construed as representing those of the Cato Institute, my law firm, or its clients.

This Task Force’s work is both important and urgent. It is important because the actions of the Obama Administration, especially in President Barack Obama’s second term, threaten to permanently upend our constitutional separation of powers by arrogating to the Executive Branch powers that had previously been reserved to the Legislative Branch and subject to check by the Judicial Branch. Although apologists for this Administration can rightly claim that its aggressive use of executive power has continued trends in executive unilateralism that began in previous administrations, unwilling to engage a Congress wary of its policy priorities, it has pushed more forcefully to expand the limits of executive action beyond those recognized by its predecessors. And that, in turn, is why the issue of executive overreaching is so urgent. We simply do not know whether this President’s actions have established a new status quo, one that will be impossible for future presidents—no matter their political affiliation—to abandon in favor of the old understandings and arrangements. Indeed, we do not know whether future presidents will consider the Obama Administration’s executive actions to be a baseline for further, even more aggressive actions that depart still further from our constitutional separation of powers.

Interest in the issue of executive overreaching is not confined to the legal profession and the policy community. The use, abuse, and limits of executive power have been overriding issues of public concern in the current and previous administration. Many members of the public, as well as members of this body, question the legitimacy of numerous actions taken by the current administration, from circumventing Congress to “enact” immigration reform, to circumventing Congress to regulate greenhouse gas emissions and ban new coal-fired power plants, to circumventing Congress to repeal the work requirements that were the centerpiece of 1996’s welfare reform, to circumventing Congress to “rewrite” problematic provisions of the Patient Protection and Affordable Care Act.

Inevitably, reaction to individual examples of overreaching is colored by politics, with the President’s supporters often willing to suspend their wariness of executive unilateralism when it is deployed to achieve policy goals that they favor and his detractors sometimes quick to pounce on perceived abuses that do not always pan out. But that understandable division should not obscure the fact that the Obama Administration has launched us into a new era of executive administration by seizing the prerogative to make the kind of decisions of enormous economic, social, and practical significance that had heretofore been made by Congress—particularly in the area of domestic policy. The current President’s supporters should understand, no less than those who disagree with his policy agenda, that that placing so much power in the executive carries great risks, risks that ultimately outweigh the policy results that they now celebrate. How many of those who approve the unilateral actions of the Obama Administration would be content to see those same powers, or even greater ones premised on Obama-era precedents, exercised by an administration helmed by a President Donald Trump, Ted Cruz, Hillary Clinton, or Bernard Sanders?

What’s really at stake here are not fleeting political victories, but the rights and liberties of all Americans. The Constitution provides for separation of powers to protect individual liberty1 and for checks and balances to confine each branch of government to its proper place and thereby enforce the separation of powers. Departing from the Constitution’s structure because doing so may be convenient in some mundane political dispute jeopardizes Americans’ political freedoms and individual liberties over the long term.

Congress is not powerless to reverse this trend, and it can and should work to assert and reclaim its place in the constitutional order. That will not be easy. It will require the fortitude to wield the power of the purse against executive prerogative and to conduct intensive and forceful oversight. It will also require thoughtful legislation to address the pathologies of our current arrangements that the current Administration has exploited so skillfully. And it will require rethinking the ways that Congress has—sometimes intentionally, sometimes inadvertently—facilitated executive overreaching through delegations of broad discretion.

Although there is little prospect that any substantial regulatory reforms will become law in this Congress—why would the President sign a bill abolishing techniques that have proven so useful to his Administration?—now is the time to lay the intellectual and political groundwork for an aggressive firstone-hundred-days regulatory reform agenda for the next administration. It is in that spirit that this testimony analyzes several instances of executive overreaching by the current Administration, identifies the means by which that overreaching was achieved, and then presents several modest proposals for reform to prevent future abuses and shore up the separation of powers.

I. Waiving Welfare Reform’s Work Requirements

The centerpiece of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 was Section 407, which is entitled “Mandatory Work Requirements” and sets out an absolute requirement that state welfare programs achieve specific work-participation rates or forfeit federal funding.

In a 2012 “Information Memorandum” to states, the Obama Administration Department of Health and Human Services encouraged states to submit proposals for state welfare programs that do not comply with Section 407’s work requirements, asserting that HHS has the authority to waive compliance with that provision. This episode reflects several themes in executive overreaching: circumvention of Congress to achieve policy goals, aggressive statutory interpretation to increase executive discretion, abuse of waiver authority, and reliance on lack of standing to avoid potential legal challenges.

A. The 1996 Act Established “Mandatory Work Requirements”

The 1996 Act replaced the failed Aid for Families with Dependent Children program, which perversely encouraged dependency on government by offering states additional federal funding as their welfare rolls grew. The new program, Targeted Aid for Needy Families (TANF), offered block grants to states with programs that met certain conditions. Foremost among these conditions were that states require able-bodied welfare recipients to engage in “work activities” and that each state achieve specified work-participation rates for welfare recipients.

Section 407 lays out these requirements in clear, imperative language. The statute contains two tables specifying minimum work-participation rates, one for all families receiving assistance and one for two-parent families receiving assistance.2 A state receiving TANF funding “shall achieve the minimum participation rate” specified in each table for each applicable year.3 For 2002 and thereafter, the applicable participation rates are 50 percent for all families and 90 percent for two-parent families.4 To prevent gaming, the statute even contains a provision specifying the precise method of calculating participation rates.5

The work requirements for welfare recipients are equally clear and equally mandatory. The statute provides that, “if an individual in a family receiving assistance…refuses to engage in work…, the State shall” either “reduce the amount of assistance” to that family on at least a pro rata basis or simply “terminate such assistance.”6 States can decline to impose a penalty for violations only in three circumstances: for “good cause,” for exceptions established by the state and approved by HHS, and for a single parent where childcare is otherwise completely unavailable.7 Such exceptions are not counted, however, in calculating states’ work-participation rates.8

It is apparent on the face of Section 407 that Congress was concerned that the Department of Health and Human Services (HHS), which administers TANF, or states would attempt to evade the law’s strict work requirements. To prevent backsliding, it legislated in great detail, defining terms with specificity and setting hard caps on exemptions. For example, rather than leave the matter to administrative discretion, Section 407 enumerates 12 “work activities”—including subsidized and unsubsidized employment, onthe-job training, and vocational training—that satisfy the state and individual work requirements.9 It specified the number of hours per week that family members would be required to work to be considered “participating in work activities.”10 It put a hard cap of 30 percent on the proportion of a state’s welfare recipients who could participate in educational activities and still be counted as engaged in work.11 Finally, the law requires HHS to oversee and verify states’ compliance with all work requirements.12

In addition to the penalties for individuals refusing to work, the 1996 Act established penalties for states that did not comply with Section 407. States that failed to cut off or reduce assistance to such individuals would lose between 1 and 5 percent of their TANF funding in the subsequent year, amounting to millions of dollars.13 And states that failed to meet the minimum workparticipation rates specified in Section 407 would lose 5 percent of their federal funding in the subsequent year, increased by 2 percentage points for each year of non-compliance, up to 21 percent.14 In this way, Congress gave the work requirements real teeth.

B. The Obama Administration Asserts Authority To Waive Work Requirements

On July 12, 2012, HHS issued an “Information Memorandum” to state welfare plan administrators regarding “waiver and expenditure authority un-der Section 1115.”15 Despite the prosaic title, the memorandum signaled a major shift in policy for HHS regarding the mandatory nature of the work requirements contained in Section 407.

HHS, the memorandum explained, “is encouraging states to consider new, more effective ways to meet the goals of TANF, particularly helping parents successfully to prepare for, find, and retain employment.”16 To achieve these goals, the memorandum announced that HHS would accept applications for waivers from TANF requirements “to allow states to test alternative and innovative strategies, policies, and procedures that are designed to improve employment outcomes for needy families.” Specifically, “to improve employment outcomes,” HHS would exercise its Section 1115 waiver authority to “waive compliance” with Section 407 and authorize states to adopt different “definitions of work activities and engagement, specified limitations, verification procedures, and the calculation of participation rates.”17 The memorandum contained a single paragraph of legal analysis supporting HHS’s novel contention that it could waive any aspect of Section 407:

Section 1115 authorizes waivers concerning section 402.… While the TANF work participation requirements are contained in section 407, section 402(a)(1)(A)(iii) requires that the state plan “[e]nsure that parents and caretakers receiving assistance under the program engage in work activities in accordance with section 407.” Thus, HHS has authority to waive compliance with this 402 requirement and authorize a state to test approaches and methods other than those set forth in section 407, including definitions of work activities and engagement, specified limitations, verification procedures, and the calculation of participation rates.18

That same day, Representative Dave Camp and Senator Orrin Hatch sent a letter to HHS Secretary Kathleen Sebelius requesting that she provide “a detailed explanation of your Department’s legal reasoning” underlying its assertion of authority to waive Section 407’s requirements.19

The Secretary responded a week later with a three-page letter explaining that “Republican and Democratic Governors have requested more flexibility in welfare reform” and, in particular, that governors of both parties had supported legislation in 2005 to broaden waiver authority.20

Accompanying Secretary Sebelius’s letter was a one-page attachment setting forth the Administration’s “Legal Basis for Utilizing Waiver Authority in TANF.” This document recapitulates the legal basis offered in HHS’s earlier Information Memorandum—i.e., that because Section 1115 authorizes waiver of requirements in Section 402, and because Section 402 mentions Section 407, Section 1115 authorizes HHS to waive Section 407.21

HHS, the Secretary’s letter further explains, “has long interpreted its authority to wave state plan requirements under Section 1115 to extend to requirements set forth in other statutory provisions that are referenced in the provisions governing state plans.” As an example, it mentions Wisconsin’s “Work Not Welfare” program, which included a waiver of rules related to the distribution of child support contained in Section 454, despite that Section 1115 only references the child support state plan provisions of Section 457 (which, in turn, references Section 454). Even if there were doubt as to this authority, the document continues, Congress has ratified HHS’s more expansive interpretation by declining to amend the statute.22

Finally, the document dismisses the argument that a separate provision, Section 415, precludes HHS from waiving Section 407’s work requirements, on the basis that this limitation applied only to the “former AFDC program” and “does nothing to restrict the Secretary’s waiver authority with respect to the current TANF program.”23

C. Legal Analysis: An Overreach

By its own terms, Section 407 establishes a set of obligations on states accepting TANF funding from the federal government. It expressly conditions their entitlement to funds on satisfying specified “work requirements.” It contains no exception to its reach and no provision giving the Secretary of HHS authority to relax or waive its requirements. There can be no question but that, by default, it applies to all states accepting TANF funding. HHS does not dispute this point, nor could it.

The questions that HHS’s actions raise, however, are (1) whether the Secretary possesses authority, from some other statutory source, to excuse states accepting TANF funding from full compliance with Section 407’s requirements and (2) if so, whether that authority is limited by any other provision. As to the first question, HHS points to Section 1115’s waiver authority, but as is discussed below, that provision cannot be read to reach Section 407. As to the second, even if Section 1115 could be interpreted, standing alone, to authorize the waiver of Section 407’s requirements, that interpretation would ultimately have to be rejected in light of the more specific language of Section 415, which precludes the waiver of work requirements and confirms Congress’s intention that Section 407’s work requirements not be subject to waiver.

1. Section 1115 Waiver Authority

HHS argues that Section 1115 authorizes it to waive Section 407’s work requirements. It does not.

Section 1115 provides, in relevant part:

In the case of any experimental, pilot, or demonstration project which, in the judgment of the Secretary, is likely to assist in promoting the objectives of [various human welfare programs], in a State or States—(1) the Secretary may waive compliance with any of the requirements of section 302, 602, 654, 1202, 1352, 1382, or 1396a of this title, as the case may be, to the extent and for the period he finds necessary to enable such State or States to carry out such project[.]24

(Because it refers to U.S. Code provisions, rather than the organic statute, its reference to Section 602 corresponds to Section 402 of the Social Security Act.)

Section 402, in turn, defines what it means to be an “eligible state,” i.e., one that is eligible to receive a TANF block grant.25 In particular, it requires a state to “submit[] to the Secretary a plan,” in the form of a “written document that outlines how the State intends to” carry out various requirements for federal funding.26 Among other things, a state must outline how it intends to “[e]nsure that parents and caretakers receiving assistance under the program engage in work activities in accordance with section [407].”27

This provision, HHS argues, allows it to waive Section 407’s work requirements. But that contention must be rejected on three grounds.

The first, and simplest, is the negative-implication canon, or expressio unius est exclusio alterius (the expression of one thing implies the exclusion of others). Section 1115 lists seven provisions the requirements of which the Secretary may waive. Section 407 is not among them. Ergo, the Secretary has no authority to waive its requirements. The enumeration of statutory provisions subject to waiver manifests congressional intent to limit the Secretary’s discretion, not to allow her free reign over the entirety of Title 42.28

Second is the precise language and structure of Sections 402 and 407. Section 407 establishes freestanding requirements for state programs receiving TANF funding and does not depend on Section 402 for its effectiveness. Its text contains commands for states participating in TANF: they “shall achieve the minimum participation rate” and “shall reduce the amount of assistance otherwise payable” to a family whose members refuse to work.29 These provisions establish independent obligations on states participating in TANF and are effective irrespective of any requirement of Section 402. In other words, even had Section 402 omitted any reference to Section 407, they would still continue in force; a state would merely be relieved from “outlin[ing]” in a “written document … how the State intends to satisfy” any portion of Section 402.

This interpretation is confirmed by Section 402’s limited reference to Section 407’s requirements. As described above, Section 407 imposes two separate types of requirements for states: (1) that they attain certain “minimum participation rate[s]” and (2) that they impose penalties on any recipient of assistance (with certain exceptions) who “refuses to engage in work.”30 But Section 402 only refers to the latter requirement; it does not so much as mention the minimum participation requirements. Accordingly, those requirements cannot possibly be among the “requirement[s] of section [402]” that Section 1115 authorizes the Secretary to waive. And there is no basis in the text of Section 407 to distinguish between the two types of work requirements; both are specified in the same imperative language, as freestanding commands on participating states.31

Third is the distinction between Section 402, which is concerned with states’ discretion in carrying out their TANF programs, and other provisions (including Section 407) intended to deprive them of any discretion. Section 402 lays out the minimum contents for a state plan that is “eligible” for funding, requires that the plan be submitted in a “written document,” and requires the state to certify that it will carry out the provisions of the written plan.32 This mechanism allows the states discretion as to how they structure and operate their TANF programs, within the parameters allowed by the statute. That discretion may be broadened by Section 1115 waivers that relax Section 402 requirements.

But the statutory structure reflects that Congress did not intend to give the states such broad discretion with respect to all aspects of their programs and, in particular, with respect to work requirements. This is why minimum work-participation requirements are nowhere mentioned in Section 402; Section 407 affords states zero discretion as to whether they will meet these requirements, such that there is no reason for the states to “outline” their preferred policy choices. They have no choice, other than declining to seek TANF funds.33 Conversely, because states have some discretion as to how they intend to implement the individual work requirements for welfare recipients, they are required to outline how they intend to exercise that discretion.34 There is no basis in Section 402 to conclude, however, that their failure to do so—for example, if the outlining requirement is waived—somehow absolves them from carrying out the individual work requirements altogether.

To the contrary, Congress carefully and deliberately distinguished between areas where the states would have some discretion (and where waivers might be appropriate) and those where they would not (and waivers would be unavailable). This is apparent in comparing the broad and discretionconferring language of Section 402 with the absolute commands of Section 408, which specifies non-waivable “prohibitions” and “requirements,” and of Section 409, which specifies in comprehensive fashion penalties for states’ violation of TANF requirements. Section 408 contains a number of bedrock requirements for all state TANF programs, such as prohibiting assistance to families without minor children.35 Although containing three separate penalties for violations of Section 407’s work requirements, Section 409 does not impose penalties for any “requirement” of Section 402.36 Instead, it establishes a number of additional requirements for state TANF programs. As a result, states are not penalized for legitimate exercise of their discretion under Section 402, but they are for violations of the requirements of Sections 407, 408, and 409.

The history of Section 402 also shows that Congress intended this distinction. Prior to the 1996 Act, Section 402 contained all requirements for state welfare programs, while providing the states substantially less flexibility in the structure and operation of their programs. It opened with the command that “[a] State plan for aid and services to needy families with children must …” and proceeded through the subsequent nine pages of the official U.S. Code to enumerate in excruciating detail every requirement for state programs, all of them mandatory.37 Accordingly, Section 1115 (which did then, as now, apply to Section 402) permitted the Secretary to waive any requirement whatsoever respecting states’ welfare programs.

TANF, however, scrapped the prior approach, replacing the specific strictures of Section 402 with general requirements that afforded states substantial flexibility in the design of their programs, over which the Secretary retained waiver authority to provide still-further flexibility.38 But where Congress sought to preclude state flexibility, as with work requirements, it used mandatory language and placed those requirements in separate provisions not subject to Section 1115.

This also cuts against the Administration’s argument that Congress ratified HHS’s broad assertions of waiver authority; to the contrary, Congress acted to pick and choose to which requirements it would apply. The Administration’s interpretation, however, would render the distinctions drawn by Congress in the text and structure of the 1996 Act entirely ineffective, as if it had merely amended Section 402 and left it at that. Of course, Congress did no such thing, and a court would not so casually deprive amendments made by Congress of any meaning.39 Indeed, when a previous Secretary raised a similar argument concerning his Section 1115 authority, the court rejected it in favor of “[t]he plain language of the statute.”40 There is little doubt that a court would view things the same way in this instance, perhaps even without wading into the intricacies of the TANF program.

2. Section 415’s Limitation on the Secretary’s Waiver Authority

That the Secretary lacks authority to waive Section 407’s work requirements is confirmed by another provision of the 1996 Act, Section 415, which provides additional limitations on the Secretary’s waiver power with respect to work requirements.

Section 415(a) sets out rules governing the treatment of waivers in place at the time the 1996 Act came into effect and those “granted subsequently.” For waivers already in effect, states may continue to receive funding without complying with the Act’s new requirements, although only until the expiration of the waiver, without regard to any extensions.41 Similarly, for waivers submitted and approved between the date of the Act’s passage (August 22, 1996) and its effectiveness (July 1, 1997), states may continue to receive funding without complying with the Act’s new requirements, so long as the waiver does not increase federal costs.42 But a third provision states that, notwithstanding the exception for plans submitted and approved during the interim period, “a waiver granted under section [1115] or otherwise which relates to the provision of assistance under a State program funded under this part (as in effect on September 30, 1996) shall not affect the applicability of section [407] to the State.”43

The Administration argues that this third provision “has no application” to present-day waivers “because it is a transitional provision applicable only to waivers under the former AFDC program.…”44 But there is some ambiguity in the language of the statute. It can be ready broadly to preclude any “waiver granted under section [1115]” from waving Section 407’s work requirements. Or it can be read more narrowly, to apply only to “a waiver granted under section [1115]…which relates to the provision of assistance under a State program funded under this part (as in effect on September 30, 1996).” While the Administration may have the better argument on this point—in light of the placement of this provision in a subsection regarding “continuation of waivers” and its parallel placement with the interim-waiver provision—its interpretation is not inevitable. (Also, for what little it may be worth, the legislative history says nothing on this point, one way or the other.)

But the Administration’s interpretation, even if correct, is fatal to its position regarding Section 1115 authority. It concedes, as it must, that Congress allowed states obtaining interim-period waivers to ignore every single new requirement of the 1996 Act except for the work requirements contained in Section 407, which they were required to implement immediately upon their becoming effective. This confirms the absurdity of the Administration’s central claim that those same states could, under subsequent waivers granted after the 1996 Act went into effect, abandon those same work requirements that Congress specifically required they implement even under interim-period waiver plans. It makes no sense to suggest that Congress was so concerned about ensuring that the work requirements were not waived that it inserted a stop-gap provision to prevent waiver during the interim period following the passage but then authorized HHS to waive those requirements at will at any time thereafter.

The absurdity of this argument demonstrates its fallacy: if the Administration’s interpretation of Section 415 is correct, then its interpretation of Section 1115 to allow it to waive work requirements is surely wrong.

D. The Aftermath

HHS’s 2012 guidance attracted considerable attention and criticism, particularly from Members of Congress. It may be that the resulting controversy doomed this particular attempt to circumvent the law: given the popularity with the public of the 1996 Act’s work requirements, perhaps state officials were unwilling to bear the political cost—including intensive oversight by their own legislatures and by Congress—of seeking to waive them. So far as I am aware, no states ultimately took advantage of HHS’s invitation to dispense entirely, or even substantially, with Section 407 compliance. But the episode is, nonetheless, instructive for several reasons.

First, the guidance was a blatant attempt to circumvent Congress to achieve the policy goals of the Administration and its allies, who have long opposed the 1996 Act’s work requirements. Shortly before the guidance was announced, the President stated, “We’re going to look every single day to figure out what we can do without Congress.”45 In this instance, the Administration never even attempted to work with Congress to change the law, perhaps recognizing that any such effort would most likely be futile or even backfire by drawing greater attention to an attempt to strike or water down a popular measure. So the Administration determined to go it alone.

Second, to do that, it relied on one of the Administration’s most useful tools, aggressive interpretation of statutory authority, typically to increase executive discretion. As described above, the Administration appears not to have devoted all that much attention to the legal basis of its 2012 guidance, instead contriving a superficially plausible rationale and going no further. While in other instances the Administration has employed interpretative gymnastics to reach preferred policy results, here it settled for much less.

Third, and related, the 2012 guidance took advantage of a limited waiver provision. The Administration has used waiver provisions—often intended for special circumstances, emergencies, or (in the case of “cooperative federalism” programs) facilitating state experimentation in particular areas—to expand its policy discretion across the board. Waiver authority in programmatic areas can rarely be challenged in court and often—even if intended for narrow circumstances—can be employed to override the most detailed statutory provisions that Congress put care and thought into crafting. The implicit understanding that waiver authority might be intended only for unusual or emergency circumstances, with general provisions to otherwise govern, is not one that is necessarily enforceable and not one that this Administration has felt compelled to observe. It has also recognized that waiver authority can be used to effectively “enact” new programs, by waiving onerous statutory requirements for parties who agree to carry out policies not mentioned in the statute but favored by officials. In this way, even the most narrowly conceived waiver provisions can become open-ends grants of executive authority unless they are subject to clear statutory limitations.

Fourth, the Administration felt no need to drill down on its statutory interpretation, or to take seriously limitations on its waiver authority, because it did not expect to have to justify its legal position in court. The Supreme Court has interpreted Article III’s “case” or “controversy” requirement to require a plaintiff to demonstrate “standing”—in short, that it (1) has suffered, or will imminently suffer, a concrete injury (2) caused by the conduct challenged (3) that can be redressed by a favorable decision.46 The 2012 guidance was probably viewed by the Administration as immune from judicial review, because it was just guidance (which is not reviewable under the Administrative Procedure Act47) and because any action under it would not result in conferring standing on any plaintiff. Under the guidance, a state would approach HHS with a proposed program, and HHS in turn would waive statutory re-quirements in approving that program. No particular party48 would be injured, and so no one could bring suit. Unsurprisingly, a number of the Obama Administration’s most controversial assertions of executive authority have been in circumstances where the Administration anticipated that judicial review would be unavailable due to standing limitations.

In sum, the Administration’s 2012 attempt to waive the 1996 Act’s work requirements may be most notable as an exemplar of how executive overreach works: stretching statutory authority, and evading judicial review, so as to avoid having to win congressional approval. That it was ultimately unsuccessful reflects the fact that the President, acting unilaterally, is still constrained by reality—much as even the Administration’s most heroic overreaching has been insufficient to drive up enrollment in PPACA insurance plans.

II. EPA’s “Tailoring Rule”

In 2014, the Supreme Court struck down, in part, an Obama Administration EPA regulation purporting to “tailor” numerical thresholds in the Clean Air Act by reading them to be entirely different numbers, so that the agency could impose regulation on a larger set of facilities than it otherwise would be able to.49 The episode is a clear example of attempting to rewrite a statute to achieve particular policy goals, in reliance on both judicial deference canons and standing-based defenses to evade judicial review—a strategy that, in the end, failed for the Administration.

In 2009 and 2010, the Obama Administration EPA released the federal government’s first round of regulations addressing greenhouse gas emissions under the Clean Air Act. The “Endangerment Finding” announced EPA’s determination that greenhouse-gas emissions by motor vehicles endanger public health and welfare by fostering climate change.50 The “Triggering Rule” announced that the effectiveness of standards for motor-vehicle greenhousegas emissions would trigger permitting requirements for stationary sources (e.g., factories, power plants, etc.), as well, under certain Clean Air Act programs.51 One of those programs, Prevention of Significant Deterioration (“PSD”), makes it unlawful to construct a “major emitting facility” without first obtaining a permit, which in turn requires that the facility employ the “best available control technology” for emissions of regulated pollutants.52 A “major emitting facility” is defined by the statute to be 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).53

EPA’s “Tailoring Rule” addressed those statutory thresholds. EPA explained that the PSD program was meant to regulate only “a relatively small number of large industrial sources,” but that applying the statutory thresholds would require thousands of sources to obtain permits, due to the large amounts (much larger than other regulated emissions) in which greenhouse gases are typically emitted. That, the agency said, would make PSD both unadministrable and “unrecognizable to the Congress that designed” it. Rather than read the statute not to regulate greenhouse-gas emissions as regulated “pollutants”—a capacious term arguably subject to some agency interpretative discretion—the agency acted to “tailor” the PSD threshold, requiring sources emitting 100,000 tons per year of carbon-dioxide (or its equivalent) to obtain a permit, while leaving open the possibility that it might reduce the threshold in the future.54

In other words, EPA rewrote the statutory language “two hundred and fifty tons per year or more of any air pollutant” as “10,000 tons per year of carbon-dioxide.” It then argued that, because this rewrite alleviated burdens on sources that would otherwise be regulated, no one was injured by it such that a party would have standing to challenge it.55 The D.C. Circuit bought that argument.56

Not a single justice on the Supreme Court did. Instead, the Court addressed the merits and held that the statute could not be read to trigger PSD requirements based greenhouse-gas emissions. The majority, in an opinion by Justice Antonin Scalia, rejected EPA’s call for deference to its statutory interpretation under the standard set forth in Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc.57 The Court agreed with EPA that “that requiring permits for sources based solely on their emission of greenhouse gases at the 100– and 250–tons–per–year levels set forth in the statute would be ‘incompatible’ with ‘the substance of Congress’ regulatory scheme.’”58 But, unlike EPA, it concluded that “[a]n agency has no power to ‘tailor’ legislation to bureaucratic policy goals by rewriting unambiguous statutory terms.”59 While an agency may have interpretative discretion in areas of statutory ambiguity, “[i]t is hard to imagine a statutory term less ambiguous than the precise numerical thresholds at which the Act requires PSD…permitting.”60 Agencies, it concluded, “are not free to adopt…unreasonable interpretations of statutory provisions and then edit other statutory provisions to mitigate the unreasonableness.”61 Allowing them to do so “would deal a severe blow to the Constitution’s separation of powers.”62

EPA’s power-grab, the Court recognized, was basically unprecedented:

In the Tailoring Rule, EPA asserts newfound authority to regulate millions of small sources—including retail stores, offices, apartment buildings, shopping centers, schools, and churches— and to decide, on an ongoing basis and without regard for the thresholds prescribed by Congress, how many of those sources to regulate. We are not willing to stand on the dock and wave goodbye as EPA embarks on this multiyear voyage of discovery. We reaffirm the core administrative-law principle that an agency may not rewrite clear statutory terms to suit its own sense of how the statute should operate.63

But, as will be seen below, neither the agency nor the Administration were chastened.

This episode again reflects an attempt to circumvent Congress in achieving policy priorities—here, imposing limitations on new sources of greenhouse gases that could remain in operation for years. EPA adopted an interpretation of broad language in the Clean Air Act that was at odds with more specific provisions of the statutory text, and deployed a number of openended legal theories—including the need to ensure its own “administrative convenience”—as grounds to expand its discretion. It also relied heavily on judicial deference canons, as well as standing doctrine to evade review entirely of its most suspect statutory interpretations. These are all recurring motifs in the annals of executive overreach.

III. EPA’s “Utility MACT” RuleZ

EPA’s “Utility MACT” Rule, also known as the “Mercury and Air Toxic Standards” (“MATS”) Rule, was an instance where the Administration was able to shield a legally vulnerable regulation from effective challenge, allowing it to impose tens of billions of dollars in costs and force the premature retirement of a number of power plants despite that its rule was ultimately held unlawful.

At issue was the application of the Clean Air Act’s hazardous air pollutants program, contained in Section 112 of the Act, to power plants. The Section 112 program targets stationary-source emissions of a number of listed hazardous air pollutants.64 The program’s focus is on categories of sources (e.g., petroleum refineries, industrial process cooling towers, etc.) that emit those pollutants. EPA is required to “list” all categories of sources that emit hazardous air pollutants and then issue emissions standards for each listed category.65 Unlike other pollution-control programs, Section 112 provides little discretion in setting minimum standards for “major sources”—those emitting or with the potential to emit more than 10 tons of a single pollutant or more than 25 tons of a combination of pollutants per year.66 In general, under the “maximum achievable control technology” standard, major sources are subject to a “floor” based on “the average emission limitation achieved by the best performing 12 percent of the existing sources.”67 EPA then may in some circumstances go “beyond the floor”—that is, make them even more stringent—based on cost considerations and other factors.68 But the general idea of Section 112 and MACT is that every major source—no matter its age or unique characteristics—is required to minimize emissions of hazardous air pollutants to the same extent as the very best performing sources in the same category.69

When Congress created the current Section 112 program in the 1990 Clean Air Act Amendments, it required EPA to identify, list, and regulate nearly all categories of sources emitting hazardous air pollutants but made an exception for fossil fuel-fired power plants. Recognizing that other provisions of the Amendments would directly lead to significant reductions in power plants’ emissions of hazardous air pollutants through market-based measures and could therefore render Section 112 regulation unnecessary, it directed EPA to study power-plant emissions and review “alternative control strategies.”70 It then directed EPA to regulate power plants under Section 112 only if it “finds such regulation is appropriate and necessary after considering the results of the study.”71

Pursuant to a consent decree, EPA proposed MACT standards for power plants in May 2011 and published a final rule in February 2012.72 The final rule’s preamble features a dense 54-page discussion of the basis for regulation, ultimately “affirm[ing]” that application of Section 112 to power plants remained “appropriate and necessary.”73 Although greatly expanded, the 2012 analysis relies on the same interpretation of the statutory trigger as an earlier 2000 finding.74 Under that earlier finding (as well as the 2012 one), regulation is appropriate, in the agency’s view, if power plants emit a listed hazardous air pollutant that poses risks to public health or the environment and if controls are available to reduce those emissions. Regulation is necessary if other Clean Air Act programs do not eliminate those risks. Despite the statute’s special treatment for power plants, EPA viewed the costs of regulation—in this instance, the application of the Clean Air Act’s most stringent program to the nation’s largest category of industrial sources—as irrelevant.

The agency did respond to comments that it was required to consider costs in assessing the “appropriate[ness]” of regulation. According to EPA, it was reasonable to make the decision listing power plants without consideration of the costs of regulation because it is forbidden from considering costs when making listing decisions under Section 112 for other source categories.75 It also claimed discretion to adopt an interpretation of “appropriate” turning only on the ability of Section 112 regulation to address power plants’ emissions of hazardous air pollutants. “Cost,” it concluded, “does not have to be read into the definition of ‘appropriate.’”76 In this, the agency appeared to argue that it had discretion to consider costs, but was not obligated to do so. And so it decided not to, on the view that Section 112 was geared to reducing hazards to human health and the environment.77

Although EPA did not take into account costs when determining whether to regulate, it did produce a “Regulatory Impact Analysis” tabulating the expected costs and benefits of the standards. The regulation would force power plants to bear costs of $9.6 billion per year78—making the rule one of the most expensive in the history of the federal government.79 It projected monetized direct benefits—that is, benefits flowing directly from reduced emissions of hazardous air pollutants, particularly mercury, that could be quantified—of $4 to $6 million per year, chiefly from “avoided IQ loss” resulting from reduced mercury exposure.80 It also projected ancillary benefits attributable to reductions in emissions of particulate matter (and to a much lesser extent, carbon dioxide) amounting to $37 to $90 billion per year, while acknowledging that these particulate matter “co-benefits” are subject to “uncertainty” based on limitations in its research linking particulate-matter levels with health outcomes.81

The rule was challenged on numerous grounds but ultimately upheld by the D.C. Circuit, over the dissent of Judge Kavanaugh, who argued that it was “entirely unreasonable for EPA to exclude consideration of costs in determining whether it is ‘appropriate’ to regulate electric utilities under the MACT program.”82 In Judge Kavanaugh’s view, the result was the same “whether one calls it an impermissible interpretation of the term ‘appropriate’ at Chevron step one, or an unreasonable interpretation or application of the term ‘appropriate’ at Chevron step two, or an unreasonable exercise of agency discretion under State Farm.”83 The Supreme Court granted three petitions raising that point and directed the parties to address a single question that it had formulated: “Whether the Environmental Protection Agency unreasona-bly refused to consider cost in determining whether it is appropriate to regulate hazardous air pollutants emitted by electric utilities.”84

The Court’s opinion in Michigan v. EPA, authored by Justice Scalia, forcefully declares, “‘Not only must an agency’s decreed result be within the scope of its lawful authority’”—that is, within its statutory authority—“‘but the process by which it reaches that result must be logical and rational.’”85 And that process, it continues, must rest “‘on a consideration of the relevant factors.’”86

The opinion reasons that there is a presumption that agencies will consider the costs of their actions. “Agencies,” it says, “have long treated cost as a centrally relevant factor when deciding whether to regulate. Consideration of cost reflects the understanding that reasonable regulation ordinarily requires paying attention to the advantages and the disadvantages of agency decisions.”87 And it is not “even rational, never mind ‘appropriate,’ to impose billions of dollars in economic costs in return for a few dollars in health or environmental benefits.”88 Accordingly, to overcome the presumption that costs will be taken into account, EPA’s burden was to identify “an invitation to ignore cost.”89

The “appropriate and necessary” language, the Court concluded, is not anything of the sort. While recognizing that the word “appropriate” is “capacious[],” which would ordinarily provide an agency a wide scope of interpretative discretion, the majority explains that a reasonable statutory interpretation may not, like any agency action, “‘entirely fai[l] to consider an important aspect of the problem’”—which “naturally” includes costs.90 After all, “[n]o regulation is ‘appropriate’ if it does significantly more harm than good.”91 For example, an agency could not reasonably deem something like emissions limitations “appropriate” if “the technologies needed to eliminate these emissions do even more damage to human health.”92

Michigan thus held that EPA acted unlawfully when it issued the Utility MACT Rule, but the Court’s decision has not had any impact on the Rule itself. One day after the decision dropped, EPA Air Administrator Janet McCabe took to the agency’s weblog to boast that the Court’s decision wouldn’t make any difference to the agency’s plans. Because the Rule hadn’t been stayed, “the majority of power plants are already in compliance or well on their way to compliance.”93 In other words, plants had spent (by EPA’s own estimation) tens of billions of dollars to comply with a Rule that, at the end of the day, was unlawful. On that basis, the D.C. Circuit declined to vacate the Rule, instead remanding it to EPA for further consideration. Challengers are currently seeking to interest the Supreme Court in reviewing that decision.94

As discussed below, EPA attempted to take advantage of this precedent in its next major action, the Clean Power Plan.

IV. EPA’s “Clean Power Plan”

EPA’s “Clean Power Plan” represents the agency’s attempt to arrogate to itself authority over the nation’s generation fleet and electric grid that has always been exercised by the states and, in certain aspects, by the Federal Energy Regulatory Commission. The Plan relies on aggressive statutory interpretation, as bolstered by Chevron deference, to effect a transformation of the nation’s energy economy. Keenly aware of the Utility MACT precedent, and no doubt concerned over the legal infirmities of the Clean Power Plan, the Supreme Court acted to stay the rule on February 9, 2016.95

The Plan relies on Section 111(d) of the Clean Air Act. That provision charges states to establish and apply “standards of performance” for certain existing stationary sources of air pollutants. A “standard of performance” is “a standard for emissions of air pollutants which reflects the degree of emission limitation achievable through the application of the best system of emission reduction.” Under Section 111(d), EPA “establish[es] a procedure” for states to submit plans establishing such standards and providing for their implementation and enforcement. EPA’s procedure must allow states “to take into consideration, among other factors, the remaining useful life” of a source. Only if a state fails to submit a compliant plan may EPA step in and promulgate a federal plan to regulate sources within a state directly.96

EPA promulgated the Plan at the same time that states and utilities were making final decisions whether to upgrade or retire coal-fired facilities in response to Utility MACT—which EPA projected would result in the retirement of 4,700 megawatts of coal-fired generating capacity and require tens of billions of dollars in investments for the remaining facilities to achieve compliance by the April 16, 2016 deadline. The Plan aims to reduce carbondioxide emissions from the power sector by 32 percent by 2030, relative to 2005 levels. These emissions reductions are premised on states’ actions to overhaul their electric sectors, shifting from coal generation to natural gas and from fossil fuels to renewable sources like wind and solar.

It specifies numerical emissions rate- and mass-based CO2 goals for each state, based on its existing coal-fired and gas-fired generation fleet. These goals are based on projected emissions reductions that EPA believes can be achieved through the combination of three “building blocks” that it says represent a baseline “best system of emission reduction”: (1) require power plants to make changes to increase their efficiency in converting fuel into energy, (2) replace coal-fired generation with increased use of natural gas, and (3) replace fossil-fuel-fired generation with generation from new, zero-carbon-emitting renewable energy sources, such as wind and solar. 80 Fed. Reg. at 64,667/1. In other words, the EPA Power Plan requires states to transition away from coal-fired generation and take all steps that are necessary to integrate other generating sources and to maintain electric service. EPA, however, itself lacks the authority to carry out all but the first of these building blocks, as well as supporting actions necessary to reorganize the production, regulation, and distribution of electricity.

Yet EPA recognizes that such “generation-shifting” will be required for states to comply with the Plan. EPA acknowledges that source-specific efficiency improvements are insufficient to achieve anywhere near that required magnitude of reductions. Accordingly, whether a state adopts a state plan to meet these targets or EPA promulgates a federal plan, the Plan forces the the state to undertake and facilitate generation-shifting, as well as substantial legislative, regulatory, planning, and other activities to accommodate the changes required by the Plan and to maintain electric service throughout the state.

Serious federalism issues aside, the Plan contravenes two separate statutory limitations on EPA authority. First is the bar on regulation of source categories already subject to Section 112—as power plants are following EPA’s Utility MACT rule. Section 111(d) states that EPA may not require states to issue “standards of performance for any existing source for any air pollutant…emitted from a source category which is regulated under section [112].”97 The Supreme Court recognized the plain meaning of the Section 112 exclusion in AEP v. Connecticut, finding that “EPA may not employ § 7411(d) if existing stationary sources of the pollutant in question are regulated under the national ambient air quality standard program, §§ 7408–7410, or the ‘hazardous air pollutants’ program, § 7412. See § 7411(d)(1).”98 EPA promulgated Section 112 regulations for electric utility generating units—that is, power plants—in 2012. EPA therefore lacks authority to require Section 111(d) emissions standards for power plants—full stop.

In fact, EPA likewise has recognized for years that “a literal reading” of the language codified at 42 U.S.C. § 7411(d)(1) mandates “that a standard of performance under section 111(d) cannot be established for any air pollutant—HAP and non-HAP—emitted from a source category regulated under section 112.”99

Of course, where the “literal reading” of the text is clear, “that is the end of the matter; for the court, as well as the agency, must give effect to the unambiguously expressed intent of Congress.”100 And that should be the end of the matter here: the Clean Air Act unambiguously withholds authority from EPA to require states to establish Section 111(d) performance standards for a source category, like power plants, that is regulated under Section 112.

EPA’s primary defense to the plain language of the Act is to assert that there is an ambiguity in the Statutes at Large concerning Section 111(d), based on two portions of the 1990 Clean Air Act Amendments that EPA claims conflict.101 The first is a substantive amendment to Section 111(d) (the “House Amendment”). Before 1990, the Section 112 exclusion prohibited EPA from requiring States to regulate under Section 111(d) any air pollutant “included on a list published under…112(b)(1)(A).”102 This meant that if EPA had listed a pollutant under Section 112, the agency could not regulate that pollutant under Section 111(d). In order “to change the focus of section 111(d) by seeking to preclude regulation of those pollutants that are emitted from a particular source category that is actually regulated under section 112,”103 the House Amendment provides:

strik[e] “or 112(b)(1)(A)” and insert[] “or emitted from a source category which is regulated under section 112.”104

The second amendment (the “Senate Amendment”) appears in a list of “Conforming Amendments” that make clerical changes to the Act. Conforming amendments are “amendment[s] of a provision of law that [are] necessitated by the substantive amendments or provisions of the bill.”105 Consistent with this description, the Senate Amendment merely updated the crossreference in the Section 112 exclusion. It states:

strik[e] “112(b)(1)(A)” and insert[] in lieu thereof “112(b)”.106

This clerical update was necessitated by the fact that substantive amendments expanding the Section 112 regime—broadening the definition of “hazardous air pollutant” and changing the program’s focus to source categories—had renumbered and restructured Section 112(b).

As an initial matter, there is no true conflict between the amendments. Amendments are executed in the order of their appearance,107 and the House Amendment appears first in the 1990 Act, striking the reference to “112(b)(1)(A).” Accordingly, the Senate Amendment fails to have any effect, because it is no longer necessary to “strik[e] ‘112(b)(1)(A)’” to conform the Section 112 exclusion to the revised Section 112.108 The U.S. Code provision, in other words, fully enacts both amendments.

In any case, the U.S. Code provision is also consistent with Congress’s intent in enacting both amendments, which address different aspects of the scope of EPA’s authority. The House Amendment added a limitation to the scope of Section 111(d): where a category of sources is regulated under Section 112, Section 111(d) cannot be used to impose additional performance standards on that source category. The purpose was to ensure that existing source categories regulated under Section 112—which the 1990 Act substantially revised to focus on source categories rather than pollutants—would not face additional costly regulation under Section 111.109

The Senate Amendment had a different focus, seeking to maintain the pre-1990 prohibition on using Section 111(d) to regulate emissions of hazardous air pollutants from existing sources regulated under Section 112. Failure to retain that limitation would have allowed EPA to undo Congress’s considered decision to regulate only certain sources of hazardous air pollutants: the 1990 Act requires EPA to regulate all major sources of hazardous air pollutants, but only those area sources representing 90 percent of area source emissions, thereby sparing many smaller sources from the stringent Section 112 regime.110 42 U.S.C. § 7412(c)(3). In other words, the Senate Amendment restrains EPA from circumventing this limitation by simultaneously regulating the same emissions under both Section 112 and 111(d) and thereby burdening all sources, even the ones Congress exempted from regulation.

Thus, by blocking both double regulation and circumvention of the Section 112(c)(3) area-source limitation, the U.S. Code provision achieves Congress’s intent underlying both amendments and constitutes a statutory limitation on EPA’s authority. But even if there were a conflict, an agency or court “must read [allegedly conflicting] statutes to give effect to each if [it] can do so while preserving their sense and purpose.”111 Thus, even assuming arguendo that there is a potential conflict, EPA’s interpretation must be rejected because it deprives the House Amendment of any effect. EPA cannot attempt to manufacture ambiguity to expand its interpretative license and ability to pursue its policy goals.

The second statutory defect in EPA’s overall approach is more fundamental: the statute does not permit EPA’s generation-shifting approach, only allowing it to specify standards of performance applicable to, and achievable by, particular sources.112 In EPA’s view, “anything that reduces the emissions of affected sources may be considered a ‘system of emission reduction’” for purposes of Section 111. 79 Fed. Reg. at 34,886/1.That includes requiring a source owner or operator to shift generation to another source. But while the first “building block”—reducing emissions by improving sources’ efficiency— may be lawful to the extent that it is “achievable,” measures that involve reducing the utilization of coal-fired power plants in favor of other generation sources are not permissible components of the “best system of emission reduction” that underlies a Section 111 standard.

This is plain on the face of the statute. Section 111(d) requires states to “establish[] standards of performance for any existing source” that is already subject to a new source performance standard.113 Likewise, Section 111(d) requires EPA to establish “standards of performance for new sources” within listed categories.114 These provisions simply do not authorize obligations regarding other sources—for example, that application of a performance standard to a coal-fired plant would require increased utilization of some other facility that is not subject to the standard. Confirming as much, Section 111(e) enforces new source performance standards by providing that it is “unlawful for any owner or operator” of a regulated source to violate any such applicable standard.115

Indeed, a “best system of emission reduction,” which is used to determine an emission standard, must be both “achievable” and “adequately demonstrated,” but those requirements would be nullified if generationshifting (which is always an achievable and adequately demonstrated means of reducing emissions) in favor of other sources or reduced output were a permissible basis for a performance standard.116 Achievability, the D.C. Circuit has long held, must therefore be demonstrated with respect to the regulated source category itself.117

Moreover, Section 111 expressly regulates sources’ emissions “performance,” which concerns the rate of emissions at a particular level of production, and not the level of production. In other words, mandating that a highemissions facility shifting production to another facility may reduce emissions, but it has nothing to do with that facility’s emissions performance. In-deed, in its Section 111 regulations, EPA determines “performance” by measuring “pollutant emission rates” with respect to particular levels of production.118 Similarly, its regulations do not regard “[a]n increase in production rate of an existing facility” as a modification triggering application of new source performance standards.119

In light of these and many other statutory features, the courts have had no difficultly in recognizing that “best system of emission reduction” refers to measures applicable to a particular facility. The Supreme Court, viewing this language, recognized that it refers to “technologically feasible emission controls”—that is, emission-reduction technologies implemented at the source.120

EPA’s own regulations reflect the same understanding. Its regulations establishing procedures for state plans pursuant to Section 111(d) define compliance in terms of the purchase and construction of “emission control systems” and “emission control equipment,” as well as other “on-site” activities.121 They require EPA to publish guidelines “containing information pertinent to control of the designated pollutant form [sic] designated facilities,” which in turn refers to “any existing facility which emits a designated pollutant.”122 Likewise, EPA’s guidelines must reflect “the application of the best system of emission reduction (considering the cost of such reduction) that has been adequately demonstrated for designated facilities.”123 These citations are just the tip of the iceberg. A complete recitation of all the EPA regulatory actions that treat “best system of emission reduction” as referring to on-site measures would go on for pages. A recent example is the agency’s proposed performance standards for new power plants—released less than two weeks after the Plan—which reaffirms that Section 111 standards of performance “apply to sources” and must be “based on the BSER achievable at that source.”124

The Plan was, as noted above, stayed by the Supreme Court, and its lawfulness is currently being litigated before the D.C. Circuit, with a decision ex-pected before the end of the year. It is not premature, however, to identify the Plan as an example of executive overreaching: it seeks to effect a transformation of the U.S. energy economy by contorting an obscure statutory provision into license for EPA to regulate the central aspects of electricity generation and distribution across the nation. Whether or not ultimately upheld by the Judicial Branch, this kind of action involves the sort of major questions that are properly decided by Congress in exercise of its legislative power.

And it certainly bears all the hallmarks of overreaching: circumvention of Congress to achieve a major policy priority; a rush to change the facts on the ground prior to the completion of judicial review; aggressive statutory interpretation; the “discovery” of broad authority in long-dormant statutory provisions; and extensive reliance on interpretative deference canons. It also adds one not previously discussed in this testimony: trenching on the vertical separation of powers, which (as my colleague David Rivkin has observed) often results when the Executive Branch acts to breach the horizontal separation of powers.

V. Opportunities for Reform

The Legislative Branch can and should act to restrain executive overreaching and thereby assert and defend its own interests, those of the citizens its members represent, and the liberties of the American people. The section proposes several different areas of action.

  • Rethink Judicial Deference to Agency Interpretations of Statutes and Regulations. Judicial review is one of the most important checks on executive action. But it is also crucial for safeguarding the interests of the Legislative Branch, because it is the judiciary that measures the execution of the law against what Congress has actually legislated. It is therefore appropriate that this body should consider the effectiveness of judicial review and opportunities for improvement and reform. In particular, it should consider whether and how to address the deference that courts afford to agencies’ interpretations of their own regulations (often referred to as “Auer deference” or “Seminole Rock deference”) and of statutes they administer (“Chevron deference”).

    Giving agencies the authority to interpret their rules is not a constitutional command, but a matter of congressional delegation or authorization. The Court “presume[s] that the power authoritatively to interpret its own regulations is a component of the agency’s delegated lawmaking powers.”125 Thus, when considering which of several competing actors should be entitled to such deference, the Court has asked “to which…did Congress dele-gate this ‘interpretive’ lawmaking power.”126 There is no reason to believe that the presumption of delegated or conferred authority is inviolable; any power that Congress may confer on an agency, it can also rescind. Nor is there any reason to believe that the power to interpret regulations—to say what the law is, without deferring—is one that the Constitution forbids assigning to the courts, consistent with the requirements of Article III.127 Indeed, the courts routinely exercise that power today, in cases where agencies have not addressed a particular interpretative question or have been denied deference.128 Accordingly, through legislation, Congress could abrogate Auer deference, leaving courts to interpret agency rules de novo or according to their “power to persuade.”

    There are good reasons to do so. As Professor John Manning has written, according “the agency lawmaker…effective control of the exposition of the legal text that it has created,” Auer deference, unlike Chevron, “leaves in place no independent interpretive check on lawmaking by an administrative agency.”129 This is problematic for the reason identified by Montesquieu and embraced by the Framers: “[w]hen legislative power is united with executive power in a single person or in a single body of the magistracy, there is no liberty, because one can fear that the same monarch or senate that makes tyrannical laws will execute them tyrannically.”130

    As Manning explains, allocating legislative and executive power to the same entity has serious consequences for individual liberty. First, it encourages an agency to issue imprecise or vague regulations, “secure in the knowledge that it can insist upon an unobvious interpretation, so long as its choice is not ‘plainly erroneous.’”131 Second, it undermines accountability, by removing an independent check on the application of law that is illconsidered or unwise. Third, it “reduces the efficacy of notice-andcomment rulemaking” by permitting the agency “to promulgate imprecise or vague rules and to settle upon or reveal their actual meaning only when the agency implements its rule through adjudication.”132 Fourth, “[Auer] deference disserves the due process objectives of giving notice of the law to those who must comply with it and of constraining those who enforce it.”133 Finally, Auer may distort the political constraints on agency action by making it “more vulnerable to the influence of narrow interest groups” who are able “to use “ambiguous or vague language to conceal regulatory outcomes that benefit [themselves] at the expense of the public at large.”134 In recent opinions, members of the Supreme Court have expressed similar views.135

    Overruling Auer—whether by judgment or by legislation—would hardly be an avulsive change in the law. And it would have the benefits of fortifying the constitutional separation of powers, improving notice of the law, and ultimately advancing individual liberty. It is a reform worthy of serious consideration.

    Congress may also wish to consider courts’ continued application of Chevron deference. One aspect of Justice Scalia’s Perez concurrence that has attracted considerable attention is his suggestion that fixing the pathologies of administrative law may require reconsideration of Chevron deference.136 His remark speaks to a broader dissatisfaction—on the Court, among regulated parties and the public, and in the academy—with the current state of administrative authority. Where agencies once were viewed as delegates of Congress, simply “fill[ing] up the details” of congressional enactments,137 the Executive Branch has become a primary, if not the primary, mover in making federal law, supplanting Congress.138 Scalia’s criticism is notable because he is often seen as the leading expo-nent of judicial deference to agencies, in general, and of Chevron, in particular.

    Chevron’s impact cannot be overstated—at least, its impact on the Executive Branch. It has fundamentally changed the way that agencies go about their business of interpreting governing statutes. The search for meaning in Congress’s commands has been replaced with a hunt for ambiguities that might allow the agency to escape its statutory confines.139 In other words, whatever its effect in court cases—which is hotly disputed—Chevron has transformed the way that the Executive Branch pursues its policy objectives. No matter Chevron’s specifics in judicial proceedings, executive agencies have come to see it as a license for improvisation and lawmaking, so long as an escape-hatch of ambiguity can be found—and it always can.140 Whether or not Chevron has reduced judicial discretion, it has unleashed the Executive Branch and upset the balance of power between it and Congress. This is the “mood” of Chevron deference.141

    And yet Chevron has arguably failed at its primary purposes of cabining judicial discretion and increasing deference to agencies’ policy determinations. Empirical studies “show that immediately after the Chevron decision, the rate of affirmance of agency interpretations rose substantially, especially at the court of appeals level, but then in subsequent years it has settled back to a rate that is very close to where it was before Chevron.”142 One “study found that approval of an agency interpretation is less likely in cases in which Chevron is cited.”143 And another found that Chevron has been unsuccessful in “eliminat[ing] the role of policy judgments in judicial review of agency interpretations of law.”144 Despite Chevron’s conceptual merits, its actual application in the courts leaves much to be desired.

    As with Auer, Congress may supplant Chevron. Congress could, for example, specify that agency interpretations would be subject only to Skidmore deference—that is, according to their power to persuade—just as it has done with review of certain agency action under the Dodd–Frank Wall Street Reform and Consumer Protection Act.145 Or it could specify, as the Supreme Court actually once held post-Chevron, that “a pure question of statutory construction [is] for the courts to decide.”146 In fact, Congress already has specified that, in the Administrative Procedure Act.147 So it will apparently have to be more emphatic if it intends to overrule or limit Chevron.

  • Reconsider Congress’s Role in Rulemaking. The Congressional Review Act provides a (cumbersome and generally ineffective) procedure for Congress to disapprove certain agency action, subject to the President’s veto power. But it bears considering whether the CRA inverts the proper order of lawmaking under the Constitution, which makes Congress—not any agency—the primary actor in setting generally applicable laws. Proposals like the REINS Act would reassert Congress’s constitutional authority by requiring that at least major rules be subject to congressional approval before they take effect. Professor Jonathan Adler persuasively argues that the REINS Act would “enhance regulatory accountability and popular input on major regulatory proposals,” without compromising the government’s ability to undertake needed regulatory initiatives.148 It would also go a long way to deterring, or as necessary blocking, regulations that contravene congressional understanding of the Executive’s statutory authority and discretion.
  • Ensure the Availability and Effectiveness of Judicial Review. Recent actions by EPA suggest that the Executive Branch may be seeking to take improper advantage of its agility, relative to the other branches, by forcing compliance with legally questionable rules like Utility MACT and the Clean Power Plan. A party seeking to enjoin a lawfully promulgated rule bears an unusually heavy burden, such that many rules that ultimately fail judicial review are nonetheless allowed to go into force, and remain in effect, during the pendency of litigation challenging their lawfulness. As a result, rules that are ultimately held unlawful may nonetheless compel parties to invest heavily in compliance, achieving their supporters’ policy prerogatives through sheer force of edict. In this way, the executive can act to deny regulated parties their rights under law.

    Congress can act to curb this kind of abuse. The most straightforward approach, in general, would be to provide for an automatic stay of rules that impose costs above a certain threshold and are subject to judicial challenge, while also providing that, when a stay is imposed, judicial proceedings be expedited. In this way, serious questions regarding agency authority could be decided in relatively short order, before parties are required to undertake burdensome compliance measures that may ultimately prove unwarranted. Such relief could be limited, as appropriate, to rules promulgated under particular statutes (e.g., the Clean Air Act and Communications Act) or by certain agencies, with narrow exceptions for rules that the President certifies are vital to national security. The point here is not so much to identify every possible exception that may be worthy of consideration, but only to note that a law providing for automatic stays need not be absolute and can be flexibly crafted so as to address possible concerns. In other words, Congress can strike a better balance between the benefits of timely federal action and adherence to the rule of law.

  • Reconsider Broad Delegations and Discretionary Authority in Domestic Affairs. In many instances of executive overreaching, at least some blame should be placed at Congress’s feet. In the episodes described in this paper, as well as many others, the Administration has sought to take advantage of broad, vague, or otherwise uncertain statutory delegations of authority. At one time, Congress could reasonably expect that the Executive Branch would not seek to take advantage of unclear or ambiguous statutory language as a basis for launching broad policy initiatives—those kinds of issues, it was understood, would be left to Congress. But that time has past, and many statutes on the books are invitations to mischief by agencies that wish to achieve their policy priorities without going through the legislative process. At the least, Congress should take care in the laws that it enacts so as to avoid providing greater discretion than it intends. Congress should also revisit existing laws that have proven problematic or that appear open to abuse. In particular, many statutes contain waiver provisions that, while intended to authorize exceptions from rules that are otherwise generally applicable, are sufficiently broad as to provide at leastan arguable basis for general application themselves. These are ripe for reform.
  • Limit Collusive Litigation (AKA “Sue and Settle”). “Sue and settle” raises serious concerns about the conduct and resolution of litigation that seeks to set agency regulatory priorities and (in some instances) actually influences the content of those regulations. Since the House Judiciary Committee first directed its attention to the problem of collusive settlements in 2012,149 there have been a myriad of hearings and reports focusing on this problem, as well as the introduction of legislation to constructively address it. Recent examples show that the problem is real, it is serious, and it is, if anything, getting worse.

    Congress can and should adopt certain common-sense policies that provide for transparency and accountability in settlements and consent decrees that compel future government action. Legal experts have given considerable thought on how to alter the incentives and the legal environment that facilitate collusive settlements. Over the past three years, Members of the House and Senate have developed several bills that seek to carry out the principles identified in my 2012 testimony on abuses of settlements and consent decrees. The most comprehensive of those bills, the Sunshine for Regulatory Decrees and Settlements Act, passed the House in the previous Congress, and (as reintroduced this Congress) has drawn strong support in the Senate.

    That bill’s approach represents a leap forward in transparency, requiring agencies to publish proposed settlements before they are filed with a court and to accept and respond to comments on proposed settlements. It also requires agencies to submit annual reports to Congress identifying any settlements that they have entered into. The bill loosens the standard for intervention, so that parties opposed to a “failure to act” lawsuit may intervene in the litigation and participate in any settlement negotiations. Most substantially, it requires the court, before approving a proposed consent decree or settlement, to find that any deadlines contained in it allow for the agency to carry out standard rulemaking procedures. In this way, the federal government could continue to benefit from the appropriate use of settlements and consent decrees to avoid unnecessary litigation, while en-suring that the public interest in transparency and sound rulemaking is not compromised.

    Other proposed legislation focuses on settlements under specific statutory regimes. For example, the Endangered Species Act (ESA) Settlement Reform Act150 would amend the ESA to provide, in cases seeking to compel the Fish and Wildlife Service to make listing determinations regarding particular species, many of the procedural reforms contained in the Sunshine for Regulatory Decrees and Settlements Act, such as broadening intervention rights to include affected parties and allowing them to participate in settlement discussions. In addition, as particularly relevant in this kind of litigation, the bill would require that notice of any settlement be given to each state and county in which a species subject to the settlement is believed to exist and gives those jurisdictions a say in the approval of the settlement. In effect, this proposal would return discretion for the sequencing and pace of listing determinations under the ESA to the Fish and Wildlife Service, which would once again be accountable to Congress for its performance under the ESA. Similarly, the Reducing Excessive Deadline Obligations Act of 2013,151 which was introduced in the last Congress and passed the House, would have amended the Resource Conservation and Recovery Act to remove a nondiscretionary duty that EPA review and, if necessary, revise all current regulations every three years and the Comprehensive Environmental Response Compensation and Liability Act to remove a 1983 listing deadline that has never been fully satisfied.152 The effect of these amendments would have been to reduce the opportunity for citizen suits seeking to set agency priorities under these obsolete provisions.

  • Be Realistic in Setting Mandatory Duties. The “sue and settle” phenomenon suggests that there may be a broader issue at play than just collusion in litigation. Congress may wish to consider a more comprehensive approach that limits the ability of third parties to compel Executive Branch action. Suing to compel an agency to act on a permit application or the like is different in kind from seeking to compel it to issue generally applicable regulations or take action against third parties. As Justice Anthony Kennedy has observed, “Difficult and fundamental questions are raised” by citizen-suit provisions that give private litigants control over actions and decisions (including the setting of agency priorities) “committed to the Executive by Article II of the Constitution of the United States.”153 Constitutional concerns aside, at the very least, the ability to compel agency action through litigation and settlements gives rise to the policy concerns identified above, suborning the public interest to special interests and sacrificing accountability.

    The sue-and-settle phenomenon is facilitated by the combination of broad citizen-suit provisions with unrealistic statutory deadlines that private parties may seek enforced through citizen suits. According to William Yeatman of the Competitive Enterprise Institute, “98 percent of EPA regulations (196 out of 200) pursuant to [Clean Air Act] programs were promulgated late, by an average of 2,072 days after their respective statutorily defined deadlines.”154 Furthermore, “65 percent of the EPA’s statutorily defined responsibilities (212 of 322 possible) are past due by an average of 2,147 days.”155 With so many agency responsibilities past due, citizen-suit authority allows special-interest groups (whether or not in collusion or philosophical agreement with the agency) to use the courts to set agency priorities. Not everything can be a priority, and by assigning so many actions unrealistic and unachievable nondiscretionary deadlines, Congress has inserted the courts into the process of setting agency priorities, but without providing them any standard or guidance on how to do so. It should be little surprise, then, that the most active repeat players in the regulatory process—the agency and environmentalist groups—have learned how to manipulate this situation to advance their own agendas and to avoid, as much as possible, accountability for the consequences of so doing.

    Two potential solutions suggest themselves. First, a deadline that Congress does not expect an agency to meet is one that ought not to be on the books. If Congress wants to set priorities, it should do so credibly and hold agencies to those duties through oversight, appropriations, and its other powers. In areas where Congress has no clear preference as to timing, it should leave the matter to the agencies and then hold them accountable for their decisions and performance. What Congress should not do is empower private parties and agencies to manipulate the litigation process to set priorities that may not reflect the public interest while avoiding the po-litical consequences of those actions. To that end, Congress should seriously consider abolishing all mandatory deadlines that are obsolete and all recurring deadlines that agencies regularly fail to observe.156 Second, Congress should consider narrowing citizen-suit provisions to exclude “failure to act” claims that seek to compel the agency to consider generally applicable regulations or to take actions against third parties. As a matter of principle, these kinds of decisions regarding agency priorities should be set by government actors who are accountable for their actions, and subject to congressional oversight, not by litigants and not through abusive litigation.

V. Conclusion

Executive overreach is a serious problem and the Task Force should be commended for its efforts to identify the scope of the problem and potential solutions. I thank the subcommittee for the opportunity to testify on these important issues and look forward to your questions.

Notes

1See Bond v. United States, 131 S. Ct. 2355, 2365 (2011).
2 42 U.S.C. § 607(a).
3 42 U.S.C. §§ 607(a)(1), (a)(2).
4Id.
5 5 42 U.S.C. § 607(b).
6 42 U.S.C. § 607(e)(1).
7 42 U.S.C. §§ 607(e)(1), (e)(2).
8 42 U.S.C. § 607(b)(1)(B).
9 42 U.S.C. § 607(d).
10 42 U.S.C. §§ 607(c)(1)(A), (1)(B).
11 42 U.S.C. § 607(c)(1)(D).
12 42 U.S.C. § 607(i). See also 42 U.S.C. § 609(a)(15) (imposing penalties for states’ failure tocomply with work participation verification procedures).
13 42 U.S.C. § 609(a)(14).
14 42 U.S.C. § 609(a)(3).
15 Memorandum from Earl Johnson, Director, Office of Family Assistance, to States administeringthe TANF Program and other interested parties (July 12, 2012), at 1, available at http://​www​.acf​.hhs​.gov/​p​r​o​g​r​a​m​s​/​o​f​a​/​p​o​l​i​c​y​/​i​m​-​o​f​a​/​2​0​1​2​/​i​m​2​0​1​2​0​3​/​i​m​2​0​1​2​0​3​.html.
16Id.
17Id. at 2.
18Id.
19 Letter from Dave Camp, Chairman, House Ways and Means Committee, and OrrinHatch, Ranking Member, Senate Finance Committee, to Kathleen Sebelius, Secretary of Health and Human Services (July 12, 2012), available at http://​waysand​means​.house​.gov/​U​p​l​o​a​d​e​d​F​i​l​e​s​/​7​.​1​2​.​1​2​_​T​A​N​F​_​w​o​r​k​_​r​e​q​u​i​r​e​m​e​n​t​s​_​l​etter.
20 Letter from Kathleen Sebelius, Secretary of Health and Human Services, to Orrin Hatch,Ranking Member, Senate Finance Committee (July 18, 2012), available at http://www.washingtonpost.com/blogs/ezra-klein/files/2012/07/Sen-Hatch-TANF‑7–18-.pdf..
21Id. at 4.
22Id.
23Id.
24 42 U.S.C. § 1315(a)(1).
25 42 U.S.C. §§ 602(a), 603(a)(1)(A).
26 42 U.S.C. § 602(a)(1)(A).
27 42 U.S.C. § 602(a)(1)(A)(iii).
28See, e.g., United States v. Giordano, 416 U.S. 505, 514 (1974).
29 42 U.S.C. §§ 607(a)(1), (a)(2), (e)(1).
30Id.
31 Compare 42 U.S.C. §§ 607(a)(1), (a)(2) with 42 U.S.C. § 607(e)(1)
32 42 U.S.C. § 602(a)(4).
33See 42 U.S.C. § 607(a)(1) (work requirements apply only to “a State to which a grant ismade).
34 4 In particular, 42 U.S.C. 607(e)(1) grants states some discretion to establish “exceptions” to the recipient work requirement, although they remain subject to the minimum work participation rate requirements.
35 42 U.S.C. § 608(a)(1).
36 42 U.S.C. §§ 609(a)(3)(A), (a)(14)(A), (a)(15)(A).
37 42 U.S.C. § 602(a) (1994)
38 42 U.S.C. § 601(a).
39United States v. Wells, 519 U.S. 482, 495–96 (1997) (rejecting ratification argument whereCongress reenacted statute).
40Portland Adventist Medical Center v. Thompson, 399 F.3d 1091, 1099 & n.9 (9th Cir. 2005).
41 42 U.S.C. § 615(a)(1)(A).
42 42 U.S.C. § 615(a)(2)(A)
43 42 U.S.C. § 615(a)(2)(B).
44 Sebelius letter, at 4.
45 Remarks by the President on College Affordability, Oct. 26, 2011,http://​www​.white​house​.gov/​t​h​e​-​p​r​e​s​s​-​o​f​f​i​c​e​/​2​0​1​1​/​1​0​/​2​6​/​r​e​m​a​r​k​s​-​p​r​e​s​i​d​e​n​t​c​o​l​l​e​g​e​a​f​f​o​r​d​a​b​ility.
46See Lujan v. Defenders of Wildlife, 504 U.S. 555, 560–61 (1992).
47See 5 U.S.C. § 704.
48 Other than taxpayers, but taxpayer status is generally insufficient to confer standing. SeeFrothingham v. Mellon, 262 U.S. 447 (1923).
49Utility Air Regulatory Group v. EPA, 134 S. Ct. 2427 (2014).
50 74 Fed. Reg. 66496 (2009).
51 75 Fed. Reg. 17004 (2010).
52 42 U.S.C. § 7475.
53 42 U.S.C. § 7479(1).
54 This account is somewhat simplified for concision. The rule actually involves severalthresholds and phases, as well as the “Title V” program, none of which alters this legal analysis.
55See Coalition for Responsible Regulation v. EPA, 684 F.3d 102, 146 (D.C. Cir. 2012).
56Id.
57 467 U.S. 837, 842–843 (1984).
58 8 134 S. Ct. at 2443 (quoting FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 156(2000)).
59Id. at 2445.
60Id.
61Id. at 2446 (quotation marks omitted).
62Id.
63Id. at 2446.
64 4 42 U.S.C. § 7412.
65Id. § 7412(c)(1)–(2).
66Id. § 7412(a)(1).
67Id. § 7412(d)(3)(A).
68Michigan, 135 S. Ct. at 2705.
69See generally, Nat’l Lime Ass’n v. EPA, 233 F.3d 625, 633–34 (D.C. Cir. 2000); S. Rep. No.101–228, at 131–33 (1989), reprinted in 1990 U.S.C.C.A.N. 3385, 3516–18.
70 42 U.S.C. § 7412(n)(1)(A). See also 136 Cong. Rec. 3493 (Mar. 6, 1990) (statement of Sen.Symms); 136 Cong. Rec. H12911, 12934 (daily ed. Oct. 26, 1990) (statement of Rep. Oxley).
71 42 U.S.C. § 7412(n)(1)(A) (emphasis adde
72Id. For an account of the events leading up to entry of the consent decree, see Andrew M.Grossman, Regulation Through Sham Litigation: The Sue and Settle Phenomenon 5–7(2014), available at http://​www​.her​itage​.org/​r​e​s​e​a​r​c​h​/​r​e​p​o​r​t​s​/​2​0​1​4​/​0​2​/​r​e​g​u​l​a​t​i​o​n​-​t​h​r​o​u​g​h​s​h​a​m​-​l​i​t​i​g​a​t​i​o​n​-​t​h​e​-​s​u​e​-​a​n​d​-​s​e​t​t​l​e​-​p​h​e​n​o​menon.
73 77 Fed. Reg. 9,304, 9,310–64 (Feb. 16, 2012).
74 See id. at 9,311.
75Id. at 9,327
76Id. (emphasis added).
77 See id.
78Id. at 9,305–06.
79 White Stallion Energy Ctr., LLC v. EPA, 748 F.3d 1222, 1263 (D.C. Cir. 2014), (Kavanaugh, J., concurring in part and dissenting in part) (citing James E. McCarthy, Congressional Research Service, R42144, EPA’s Utility MACT: Will the Lights Go Out? 1 (2012)).
80 See 77 Fed. Reg. at 9,306, Table 2, 9,427–28. For a description of EPA’s convoluted approach to estimating and monetizing these benefits, see Brief for the Cato Institute as Amicus Curiae, Michigan v. EPA, 135 S. Ct. 2699 (2015), available at https://​object​.cato​.org/​s​i​t​e​s​/​c​a​t​o​.​o​r​g​/​f​i​l​e​s​/​p​u​b​s​/​p​d​f​/​m​i​c​h​i​g​a​n​-​v​-​e​p​a​-​m​e​r​i​t​s.pdf (“Cato Michigan Brief”).
81 77 Fed. Reg. at 9,306 & Table 2.
82White Stallion, 748 F.3d at 1261 (Kavanaugh, J., concurring in part and dissenting in part).
83Id. at 1261 (Kavanaugh, J., concurring in part and dissenting in part).
84 135 S. Ct. 702 (2014).
85 135 S. Ct. at 2706 (quoting Allentown Mack Sales & Serv., Inc. v. NLRB, 522 U.S. 359, 374(1998)).
86Id. (quoting State Farm, 463 U.S. at 43).
87Id. at 2707.
88Id.
89Id. at 2708.
90Id. at 2707 (quoting State Farm, 463 U.S. at 43) (alteration in original).
91Id.
92Id.
93 Janet McCabe, In Perspective: the Supreme Court’s Mercury and Air Toxics Rule Decision,June 30, 2015, https://​blog​.epa​.gov/​b​l​o​g​/​2​0​1​5​/​0​6​/​i​n​-​p​e​r​s​p​e​c​t​i​v​e​-​t​h​e​-​s​u​p​r​e​m​e​-​c​o​u​r​t​s​m​e​r​c​u​r​y​-​a​n​d​-​a​i​r​-​t​o​x​i​c​s​-​r​u​l​e​-​d​e​c​i​sion/.
94 See Petition for Certiorari, Michigan v. EPA, No. 15–1152 (filed Mar. 14, 2016).
95 Order, Chamber of Commerce v. EPA, No. 15A787, et al. (filed Feb. 9, 2016).
96 42 U.S.C. § 7411.
97 42 U.S.C. § 7411(d)(1)(A)(i).
98 131 S. Ct. 2527, 2537 n.7 (2011).
99 70 Fed. Reg. 15,994, 16,031 (Mar. 29, 2005). Accord EPA, Air Emissions From MunicipalSolid Waste Landfills — Background Information For Final Standards And Guidelines 1–6(1995) (explaining that the Section 112 exclusion applies “if the designated air pollutantis…emitted from a source category regulated under section 112”); Final Brief of Respondentat 105, New Jersey v. EPA, 517 F.3d 574 (D.C. Cir. 2008) (No. 05–1097) (“[A] literal readingof this provision could bar section 111 standards for any pollutant, hazardous or not, emittedfrom a source category that is regulated under section 112.”); 69 Fed. Reg. 4,652, 4,685 (Jan.30, 2004) (“A literal reading…is that a standard of performance under CAA section 111(d)cannot be established for any air pollutant that is emitted from a source category regulatedunder section 112.”); EPA, Legal Memorandum for Proposed Carbon Pollution EmissionGuidelines for Existing Electric Utility Generating Units 26 (2014) (“EPA Legal Memorandum”)(“[A] literal reading of that language would mean that the EPA could not regulate anyair pollutant from a source category regulated under section 112.”).
100Chevron, U.S.A., Inc. v. NRDC, 467 U.S. 837, 842–43 (1984).
101E.g., EPA Legal Memorandum 22–23.
102 42 U.S.C. § 7411(d) (1989).
103 70 Fed. Reg. at 16,031.
104 Pub. L. No. 101–549, § 108(g), 104 Stat. 2399, 2467 (1990).
105 Legislative Drafting Manual, Office of the Legislative Counsel, United States Senate 28(1997) (“Senate Manual”).
106 Pub. L. No. 101–549, § 302(a), 104 Stat. 2399, 2574 (1990).
107 Manual on Drafting Style, House Legislative Counsel 42 (1995); Senate Manual 33. Seealso Donald Hirsch, Drafting Federal Law § 2.2.3, p.13 (U.S. House Office of LegislativeCounsel, 2d ed. 1989); Lawrence E. Filson & Sandra L. Strokoff, The Legislative Drafter’sDesk Reference § 14.4, p.191 (CQ Press, 2d ed. 2008). The Supreme Court recognizes thesetreatises as authoritative on legislative drafting. See Koons Buick Pontiac GMC, Inc. v. Nigh, 543U.S. 50, 60–61 & n.4 (2004); id. at 71 (Scalia, J., dissenting).
108 See Revisor’s Note, 42 U.S.C. § 7411 (Senate Amendment “could not be executed, becauseof the prior [House] amendment”). The failure of a subsequent amendment to have any effect,due to changes made by an earlier amendment in the same legislation, is not at all unusual.Oklahoma is aware of more than 30 other instances—including dozens in Title 42alone—in which an amendment to the U.S. Code failed to have any effect due to an earlier amendment. Petitioner’s Opening Brief at 31–32 n.9, In re Murray Energy Corp., 788 F.3d 330(D.C. Cir. filed Mar. 9, 2015).
109 See 70 Fed. Reg. at 16,031 (discussing legislative history and concluding that the HouseAmendment sought to avoid “duplicative or overlapping regulation”).
110 “Major” sources emit or have the potential to emit above a statutorily prescribed thresholdof hazardous air pollutants; “area” sources are those that fall below this threshold. 42 U.S.C.§ 7412(a)(1)–(2).
111Watt v. Alaska, 451 U.S. 259, 267 (1981).
112 42 U.S.C. § 7411(a)(1).
113 42 U.S.C. § 7411(d)(1)(A) (emphasis added).
114 42 U.S.C. § 7411(b)(1)(B).
115 42 U.S.C. § 7411(e).
116 42 U.S.C. § 7411(a)(1).
117Essex Chem. Corp. v. Ruckelshaus, 486 F.2d 427, 433 (D.C. Cir. 1973).
118 40 C.F.R. § 60.8(e).
119 40 C.F.R. § 60.14(e).
120Hancock v. Train, 426 U.S. 167, 193 (1976). See also Bethlehem Steel Corp. v. EPA, 651 F.2d861, 869 (3d Cir. 1981) (“system” is something that a source can “install”); PPG Indus., Inc. v.Harrison, 660 F.2d 628, 636 (5th Cir. 1981) (holding that, prior to an amendment authorizingoperational standards, EPA could not “require a use of a certain type of fuel” that would reduceemissions).
121 40 C.F.R. § 60.21(h).
122 §§ 60.22(a), 60.21(b) (cross-reference omitted).
123 § 60.22(b)(5) (emphasis added).
124 79 Fed. Reg. 36,880, 36,885 (June 30, 2014).
125Martin v. Occupational Safety & Health Review Comm’n, 499 U.S. 144, 151 (1991). See alsoFord Motor Credit Co. v. Milhollin, 444 U.S. 555, 566 (1980).
126Martin, 499 U.S. at 151.
127 Congress has, in fact, codified Skidmore-style deference in certain cases. See 12 U.S.C.§ 25b(b)(5)(A).
128E.g., Christopher 132 S. Ct. 2156, 2168–69 (quoting Skidmore v. Swift & Co., 323 U.S. 134,140 (1944)).
129 John Manning, Constitutional Structure and Judicial Deference to Agency Interpretationsof Agency Rules, 96 Colum. L. Rev. 612, 639 (1996).
130Id. at 645 (quoting Montesquieu, The Spirit of the Laws bk. XI, ch. 6, at 157 (Anne Cohleret al. eds. & trans., 1989) (1768)).
131Id. at 657.
132Id. at 662.
133Id. at 669.
134Id. at 676. 135See, e.g., Talk America, Inc. v. Michigan Bell Telephone Co., 131 S. Ct. 2254, 3366 (2011) (Scalia,J., concurring); Christopher v. SmithKline Beecham Corp., 132 S. Ct. 2156, 2168 & n.17(2012); Decker v. Northwest Environmental Defense Center, 133 S. Ct. 1326, 1338 (2013) (Roberts,C.J., concurring); id. at 1339 (Scalia, J., concurring); Perez v. Mortgage Bankers Association, 135S. Ct. 1199, 1211–13 (2015) (Scalia, J., concurring in the judgment); id. at 1220 (Thomas, J.,concurring in the judgment).
136Perez, 135 S. Ct. at 1213 (Scalia, J., concurring in the judgment). But should it really havebeen such a surprise? See Decker, 133 S. Ct. at 1340 (Scalia, J., concurring in part and dissentingin part) (“…Chevron (take it or leave it)…”); Mead, 533 U.S. at 241–42 (“There is somequestion whether Chevron was faithful to the text of the Administrative Procedure Act (APA),which it did not even bother to cite.”).
137See Wayman v. Southard, 23 U.S. (10 Wheat.) 1, 43 (1825); United States v. Grimaud,220 U.S. 506, 517 (1911).
138See, e.g., Tim Devaney, Obama’s ‘pen and phone’ barrage, The Hill, Dec. 28, 2014, http://​the​hill​.com/​r​e​g​u​l​a​t​i​o​n​/​p​e​n​d​i​n​g​-​r​e​g​s​/​2​2​8​0​9​3​-​o​b​a​m​a​s​-​p​e​n​-​a​n​d​-​p​h​o​n​e​-​b​a​rrage.
139See, e.g., Jonathan Adler and Michael Cannon, Taxation Without Representation: TheIllegal IRS Rule To Expand Tax Credits Under the PPACA, 23 Health Matrix 119, 195(2013) (describing the “frantic, last-ditch search for ambiguity by supporters who belatedlyrecognize the PPACA threatens health insurance markets with collapse, which in turn threatensthe PPACA”).
140 Cf. Timothy Noah, Bill Clinton and the Meaning of “Is,” Slate, Sep. 13, 1998,http://​www​.slate​.com/​a​r​t​i​c​l​e​s​/​n​e​w​s​_​a​n​d​_​p​o​l​i​t​i​c​s​/​c​h​a​t​t​e​r​b​o​x​/​1​9​9​8​/​0​9​/​b​i​l​l​_​c​l​i​n​t​o​n​_​a​n​d​_​t​h​e​_​m​e​a​n​i​n​g​_​o​f​_​i​s​.html.
141 See Kent Barnett, Codifying Chevmore, 90 N.Y.U. L. Rev. 1, 54 n.275 (2015).
142 Jack M. Beerman, End the Failed Chevron Experiment Now: How Chevron Has Failedand Why It Can and Should be Overruled, 42 Conn. L. Rev. 779, 829 (2010) (citing studies).
143Id. (emphasis added).
144 Thomas J. Miles and Cass R. Sunstein, Do Judges Make Regulatory Policy? An EmpiricalInvestigation of Chevron, 73 U. Chi. L. Rev. 823, 825–27 (2006).
145 See 12 U.S.C. § 25b(b)(5)(A).
146I.N.S. v. Cardoza-Fonseca, 480 U.S. 421, 446 (1987).
147 5 U.S.C. § 706 (“To the extent necessary to decision and when presented, the reviewingcourt shall decide all relevant questions of law, interpret constitutional and statutory provisions,and determine the meaning or applicability of the terms of an agency action.”); § 706(“The reviewing court shall… hold unlawful and set aside agency action, findings, and conclusionsfound to be… n excess of statutory jurisdiction, authority, or limitations, or short ofstatutory right.”). One strike against Chevron (and Seminole Rock and Auer) is that it’s flatlyinconsistent with the APA. Historical evidence suggests that Congress meant what it said in1946, but that courts ultimately adopted the more deferential views expressed in the AttorneyGeneral’s Manual on the Administrative Procedure Act. See Beerman, supra, at 789–90.
148 See Jonathan H. Adler, Would the REINS Act Rein In Federal Regulation?, Regulation,Summer 2011, at 22, available athttps://object.cato.org/sites/cato.org/files/serials/files/regulation/2011/7/regv34n2‑2.pdf.
149See generally The Use and Abuse of Consent Decrees in Federal Rulemaking: Hearing beforethe Subcommittee on the Courts, Commercial and Administrative Law, Committee onthe Judiciary, United States House of Representatives, 112th Congress (Feb. 3, 2012), availableat http://judiciary.house.gov/_files/hearings/Hearings%202012/Grossman%2002032012.pdf (written testimony of Andrew M. Grossman, Visiting Legal Fellow, The Heritage Foundation)[hereinafter “2012 Testimony”].
150 H.R. 585; S. 293.
151 H.R. 2279 (113th Cong.).
152See generally Reducing Excessive Deadline Obligations Act of 2013, House Report 113–179(113th Cong.).
153Friends of the Earth, Inc. v. Laidlaw Envt’l Servs. (TOC), Inc., 528 U.S. 167, 197 (2000) (Kennedy,J., concurring).
154 William Yeatman, EPA’s Woeful Deadline Performance Raises Questions about AgencyCompetence, Climate Change Regulations, “Sue and Settle,” July 10, 2013,http://cei.org/sites/default/files/William%20Yeatman%20-%20EPA%27s%20Woeful%20Deadline%20Performance%20Raises%20Questions%20About%20Agency%20Competence.pdf.
155Id.
156 One commentator endorses allowing agencies to set their own non-binding deadlines, subjectto congressional oversight. Alden F. Abbott, The Case Against Federal Statutory andJudicial Deadlines: A Cost-Benefit Appraisal, 39 Admin. L. Rev. 171, 200–02 (1987).