By David Williams, Editor, Virginia Environmental Law Journal.
In the wake of Massachusetts v. EPA, the EPA fashioned new regulations to cover greenhouse gasses. As part of the new suite of regulations, the agency promulgated a “Tailoring Rule” that departed from the plain text of the Clean Air Act (“CAA”). The EPA justified this rule with reference to two canons of interpretation: absurd results and administrative necessity. The EPA describes the canon of administrative necessity as a three part test:
When an agency has identified what it believes may be insurmountable burdens in administering a statutory requirement, the first step the agency must take is to evaluate how it could streamline administration as much as possible, while remaining within the confines of the statutory requirements. The second step is that the agency must determine whether it can justifiably conclude that . . . the remaining administrative tasks are impossible for the agency because they are beyond its resources, e.g., beyond the capacities of its personnel and funding . . . .Then the agency may take the third step, which is to phase in or otherwise adjust the requirements so that they are administrable.
The way the agency describes and applies the administrative necessity doctrine suggests that it is a well-established, clearly defined doctrine that has been used often to justify agency departures from statutory requirements. I argue to the contrary. The doctrine of administrative necessity is actually a recently assembled collection of disparate statements from a small handful of D.C. Circuit cases. Never has a regulatory scheme that departs from statutory requirements been justified by administrative necessity. Such a piecemeal rule is inadequate to justify the EPA’s regulatory departure from the Clean Air Act.
I. Origins of the Doctrine of Administrative Necessity
The three-step test as articulated by the EPA was only assembled coherently in the tailoring rule itself. Administrative necessity applies when Congress gives an agency more non-discretionary duties than it can actually fulfill given constraints on budget, time, and manpower. The agency should thus be able to limit its regulatory reach in concert with its resource constraints.
The idea underlying administrative necessity was first posited in dicta inAlabama Power Co. v. Costel, which came in the wake of the 1977 CAA Amendments. The EPA interpreted the “potential to emit” provision of the CAA to exempt major emitting facilities from regulation if their actual emissions were below 50 tons per year (“tpy”), even though it would leave some facilities unregulated. The EPA tried to justify this regulation in litigation through cost benefit analysis, arguing that the huge burden imposed by regulating the major emitters in question would yield minimal benefits.
The Alabama Power Court decided the case on grounds that rendered the 50 tpy regulation moot, but still elected to address “the principles pertinent to an agency’s authority to adopt general exemptions to statutory requirements.” The Court began the discussion by stating that even though there is an implicit ability to create exemptions not explicitly authorized by the statute inherent in the administrative process, cost benefit analysis is not an adequate foundation for those exemptions. Rather, “the existence of an impossibility,” such as restraints caused by lack of resources or manpower, is required for an exemption. To show that an impossibility exists, a court must look to the statute to see if the legislature has built in flexibility that the agency could use that does not run directly counter to the statute.
II. Early Application of Pseudo Administrative Necessity
The doctrine continued to develop in Environmental Defense Fund v. EPA, where the agency tried to justify its interpretation of the Toxic Substances Control Act (“TSCA”). The EPA included exemptions for products with low concentrations of polychlorinated biphenyls (“PCBs”). The EPA tried to justify the exemptions by arguing that it did not have the capacity to enforce the statute completely. The Court found that the agency’s interpretation was not justified by administrative necessity because it failed to rely on exemption authority provided elsewhere in the statute. The agency should turn first to statutory exemptions before relying on self-created exemptions. Although the EPA’s assertion of administrative necessity was unsuccessful, the contours of the doctrine began to take shape.
In Sierra Club v. EPA, the EPA was required to promulgate regulations dealing with the problem of stationary sources raising stack heights to circumvent ground-level pollutant thresholds. Rather than confront the underlying issue, the EPA’s final rule only regulated three specific techniques used to avoid these thresholds. The Sierra ClubCourt held that the administrative necessity doctrine did not justify this regulation because the EPA had not carried the heavy burden of proving that full enforcement was impossible given its current resources. It had “offered mere predictions, rather than conclusions drawn from good faith efforts at enforcement,” and it had “caved in” without having “adequately explored . . . regulatory alternatives” like regulating plumes based on their engineering purposes.
Center for Biological Diversity v. EPA provides the most recent discussion of administrative necessity. For three years The EPA defered its decision whether to regulate sources of biogenic carbon dioxide the same way that they regulated other stationary sources. The court held the agency’s action was arbitrary and capricious despite an administrative necessity claim. The opinion emphasized that impossibility is the standard the agency must meet, not mere inconvenience, and added that agencies faced with impossibility must adopt the narrowest feasible exemption to the statutory mandate. The exemption did not meet the impossibility standard because it did not satisfy the Sierra Club v. EPArequirement that all feasible alternatives to the exemption be explored. The EPA had previously rejected without investigation a “middle-ground” approach where biogenic carbon dioxide sources could obtain permits but only if they fail to make “any effort to take into account net carbon cycle impacts.” The agency should have responded to this regulatory alternative with adequate exploration in order to establish impossibility.
The concurring opinion in Center for Biological Diversity also rejected the administrative necessity doctrine, but for a different reason. That opinion argued that the EPA had no statutory basis for distinguishing between biogenic carbon dioxide and other sources of carbon dioxide. The fact that the agency tried to distinguish between these sources concerned the concurring judge because it appeared that the decision to distinguish may have come from cost benefit analysis, which was impermissible underAlabama Power.
Even outside the context of environmental law, the administrative necessity doctrine is rarely applied and never successful. Public Citizen v. Shalala concerned a restaurant menu exemption from the nutrition and health labeling requirements of the Nutrition Labeling and Education Act (“NLEA”). The district court gave significant consideration to the FDA’s argument that administrative necessity justified the exception before ultimately deciding that the doctrine could not justify the exception. After acknowledging that the doctrine could theoretically justify this type of exception, the court held that the “FDA has not borne its ‘especially heavy’ burden of establishing the administrative impossibility of applying the . . . provisions of the NLEA to restaurant menus.”  The agency had twice stated that a lack of resources prevented it from enforcing its regulations in restaurants, but those statements had been “proffered in support of the agency’s decision to hold restaurants to a lower standard for substantiating claims of nutrition content and health, not of its decision to exempt menus altogether.”The proof that the FDA had given of its lack of resources pertained only to another context, so it had not met the heavy burden of proving impossibility.
In Public Citizen v. FTC, an FTC interpretation of the Smokeless Tobacco Act exempted utilitarian items from required health warnings on advertisements for smokeless tobacco. The court decided the case on the grounds that the statute was unambiguous, and therefore the regulatory departure failed under the first step of Chevron analysis. But in dicta, the Court hinted that it might be possible to uphold similar exceptions under the administrative necessity doctrine. The exemptions at issue, however, failed because they were based on cost-benefit analysis, which according to Alabama Power is not a permissible basis for exemptions.
It is important to note that none of the cases upon which the EPA relied for its formulation of the administrative necessity doctrine actually allowed the exemptions in question to stand. The same holds true generally. I found no cases in which administrative necessity justified an exception contrary to statutory language. The case law is notably sparse, and when the language which the EPA’s Tailoring Rule relies on to establish the three-part test is cited, it is usually used to establish propositions other than administrative necessity.
III. Defenses and Criticisms of Administrative Necessity
The main argument in support of the use of the administrative necessity doctrine is that it would help avoid the enormous costs that would come from directly applying the unambiguous language of certain statutes. Since “current administrative law doctrines do not adequately accommodate an agency’s inability to fully carry out an excess of delegated nondiscretionary power, and Congress has proven unwilling to adjust delegated authority in light of strained agency budgets,” the agency should have some recourse. When agencies are asked to do more than they are capable of, costs accrue to the agency, the taxpayers, and those the statute was meant to protect. Deadlines get pushed back indefinitely, agency resources get stretched so thin that otherwise remediable violations get left unenforced, and other agency responsibilities fall by the wayside.
To illustrate this problem, consider the costs that would come from immediate full enforcement of the CAA without the Tailoring Rule:
Over six million sources would be newly classified as major stationary sources for PSD purposes. EPA estimates that over 80,000 sources would be required to apply for PSD permits each year for new construction and modifications. Each of the over six million newly major stationary sources would also be required to apply for a Title V permit. The estimated cost of the PSD and Title V permitting programs would increase from $74 million annually to $22.5 billion annually. The annual number of work hours needed to run the permitting programs would increase from close to 1.5 million hours to nearly 480 million hours, which would require an additional 200,000 employees to be hired, trained, and managed.
This would mean huge administrative burdens and exorbitant costs, impeding the issuance of permits to the newly included sources and to the thousands of sources that Congress expected to be covered.
Another reason to favor the administrative necessity doctrine is that agencies have already been using undesirable non-transparent means to deal with this type of situation. Whether through under-enforcement or simple neglect, resource-strapped agencies have found ways to limit the non-discretionary power assigned to them. Application of the administrative necessity doctrine would bring those decisions out into the open, allowing agencies to publicize their inability to carry out their responsibilities and pressure Congress to address the issue. The public would receive valuable information on what statutes are going under-enforced and the risks associated with the regulatory limitations. Through notice-and-comment rulemaking or litigation, administrative necessity would bring transparency to the administrative state.
Although there are some potential benefits to use of the doctrine, the problems it creates are difficult to overcome. Primary among these is the fact that, by definition, exceptions justified by administrative necessity run counter to the statutory mandates given by Congress. Allowing agencies the discretion to deviate from the plain text of a statute creates concerns analogous to those motivating the non-delegation doctrine, which “prevents ‘agency lawmaking on the cheap’ by requiring the legislative power to be exercised through the Article I, Section 7 requirement of bicameralism and presentment.” By allowing agencies to cure the implementation issues that aspirational statutes create, the administrative necessity rationale arguably allows Congress to “avoid making the hard choice of where to direct scarce administrative resources.”  Allowing an agency to substitute its own policy choices for Congress’s policy choices in this manner would undermine core separation of powers principles.
A more pragmatic concern is that agencies may use the doctrine as a pretext to shirk their duties. This could be manifested in two ways. First, the “impossibility” standard announced in Alabama Power could feasibly be diluted over time to become “inconvenience,” allowing agencies to defy statutory responsibility whenever executing the statute would be costly or difficult. Second, agencies may claim they lack resources to enforce policies whenever they disagree with Congressional judgment. Agencies would then have a convenient end-run for any situation in which they would have legislated differently than Congress did.
One commentator downplays this concern by claiming that agencies who don’t want to enforce a statute use non-transparent means like under-enforcement anyway. This does not mean, however, that a robust administrative necessity doctrine would not expand this kind of agency quasi-legislation. Bringing administrative necessity into mainstream administrative law would grant it an air of legitimacy, which may make agency end-runs around statutory mandates pervasive.
IV. Whether Past Applications of Administrative Necessity Justify EPA’s GHG Tailoring Rule
The administrative necessity doctrine is not as well-defined as the EPA makes it appear. The novelty of the three-part test which the EPA presents is somewhat problematic, but every doctrine must have its genesis somewhere. The real issue is that no agency action has actually been upheld based on the administrative necessity doctrine. This means that there is no standard for what would actually have to be established to meet the requirements of the test.
Furthermore, the case law is so thin that it is not clear what any given element really means. The element of impossibility is a good example of this. The text of the tailoring rule makes the costs of directly applying the CAA sound so prohibitive that it is functionally impossible. But there is good reason to think that the impossibility is partially of the agency’s own making:
The need for PSD and Title V permits for greenhouse gas emissions is a result of prior regulatory choices made by EPA, described in Part I, each of which is under legal challenge . . . .Furthermore, EPA could have revised the language in its current regulations to avoid triggering a requirement to regulate greenhouse gases.
If the impossibility really is created by the agency’s own regulatory choices, perhaps a presumption that the statute is not actually impossible to administer should be established. But issues like these have not been addressed in the courts, so it really is not clear how to apply any given element of the doctrine.
Another reason why the administrative necessity doctrine cannot uphold the tailoring rule comes from the EPA’s other proffered defense of the rule: absurd results. Assuming that the absurd results doctrine actually applies to the tailoring rule, an issue I do not address here, the interplay between the two doctrines renders them incompatible. In their true form, they assume different things about congressional intent. The absurd results canon assumes that since the outcome of literal enforcement of the statutory text is so ridiculous, there is no way that Congress could have actually intended that outcome. So in this case, Congress could not actually have intended for the EPA to regulate greenhouse gasses like air pollutants. Administrative necessity, however, assumes that Congress did actually want the agency to enforce the statute in that way, but the agency does not have sufficient resources to do it. Thus, the EPA is simultaneously asserting two irreconcilable stances regarding Congressional intent. If the absurd results doctrine does apply here, then administrative necessity cannot apply as well.
In sum, administrative necessity is a patchwork doctrine that has never been applied by the courts to uphold agency decisions to limit enforcement. Its origins and existence are somewhat dubious, and its wisdom is questionable. The unstable foundation upon which the administrative necessity doctrine rests makes it seem unlikely to be a valid justification for the EPA’s Tailoring Rule.
 549 U.S. 497 (2007) (holding that the EPA has authority to regulate greenhouse gasses under the Clean Air Act).
 Prevention of Significant Deterioration and Title V Greenhouse Gas Tailoring Rule, 74 Fed. Reg. 55,292 (Oct. 27, 2009) (to be codified at 40 C.F.R. Parts 51, 52, 70, 71).
 Rather than regulate at the 100/250 tpy threshold required by the Clean Air Act, the EPA decided to regulate at thresholds of 75,000/100,000 tpy. Id. at 55,292.
 Id. at 55,295 (“The judicial doctrine of “absurd results” authorizes departure from a literal application of statutory provisions if it would produce a result that is inconsistent with other statutory provisions or congressional intent, and particularly one that would undermine congressional purposes.”).
 Id. at 55,313 (citing Alabama Power Co. v. Costel, 636 F.2d 323, 400 (D.C. Cir. 1979)) (“EPA does have discretion, in administering the statute’s ‘modification’ provision, to exempt from PSD review some emission increases on grounds of de minimis or administrative necessity.”).
 74 Fed. Reg. 55,315 (Oct 27, 2009).
 For a more in-depth discussion of the D.C. Circuit precedent leading to the use of the administrative necessity doctrine in the tailoring rule than I have space for here, see Kirti Datla, Note, The Tailoring Rule: Mending the Conflict Between Plain Text and Agency Resource Constraints, 86 N.Y.U. L. Rev. 1989 (2011); Travis Garrison, Comment,The EPA’s Greenhouse Gas Regulation Tailoring Rule: Administrative Necessity Avoiding or Pursuing Absurd Results? 56 Loy. L. Rev. 685 (2010).
 636 F.2d 323 (D.C. Cir. 1979).
 42 U.S.C. 7479(1) (2012).
 Alabama Power Co. v. Costel, 636 F.2d 323, 355–56 (D.C. Cir. 1979).
 Id. at 356–57.
 Id. at 357.
 Id. at 357–58.
 Id. at 359.
 Id. at 359–60.
 636 F.2d 1267 (D.C. Cir. 1980).
 Id. at 1272–73.
 Id. at 1283–84.
 719 F.2d 436 (D.C. Cir. 1983).
 Id. at 442–43.
 Id. at 462–63.
 Id. at 463.
 Id. at 464.
 722 F.3d 401 (D.C. Cir. 2013).
 Id. at 404.
 Id. at 410–11.
 Id. at 411.
 Id. at 411.
 Id. at 412 (Kavanaugh, J., concurring).
 Id. at 414 (Kavanaugh, J., concurring).
 932 F. Supp. 13 (D.D.C. 1996).
 Id. at 17.
 Id. (citing Environmental Defense Fund v. EPA, 636 F.2d 1267, 1283 (D.C. Cir. 1980)).
 Id. at 17 (emphasis in original).
 869 F.2d 1541 (D.C. Cir. 1989).
 Public Citizen v. FTC, 869 F.2d at 1555–56.
 Id. at 1557.
 See New York v. EPA, 443 F.3d 880, 888 (D.C. Cir. 2006) (citingAlabama Power in support of the general proposition that agencies can create exceptions to statutory language only under rare circumstances); United States v. Am. Elec. Power Serv. Corp., 218 F. Supp. 2d 931, 945 (S.D. Ohio 2002) (same); see also Potomac Elec. Power Co. v. EPA, 650 F.2d 509, 516–18 (4th Cir. 1981) (citing Alabama Power in support of its substantive finding about “common sense industrial groupings” under the Clean Air Act); Texas v. EPA, 726 F.3d 180, 191 (D.C. Cir. 2013) (same); Kentucky Waterways Alliance v. Johnson, 540 F.3d 466, 483 (Sixth Cir. 2008) (same); United States v. Ohio Edison Power Co., 276 F. Supp. 2d 829, 888 (S.D. Ohio 2003) (same); see also Public Citizen v. Mineta, 427 F. Supp. 2d 7, 12–13 (D.D.C. 2006) (citing administrative necessity to support the proposition that the agency has discretion to choose between categorical rulemaking and case by case adjudication); Pharm Research & Mfrs. of Am. v. FTC, 44 F. Supp. 3d 95 (D.D.C. 2014) (same).
 Datla, supra note 7, at 2023.
 Id. at 2001 (internal citations omitted).
 Id. at 2026 (internal citations omitted).
 Note that this concern may be more theoretical than practical since non-delegation has been rarely enforced since Mistretta v. United States, 488 U.S. 361 (1989).
 Datla, supra note 7, at 2024–25 .
 Id. at n. 88.