Administrative law has a remedy problem. Careful attention to procedural safeguards and standards of review in administrative cases often leaves remedial options undertheorized both in court opinions and in scholarly commentary.1 Recently, in Standing Rock Sioux Tribe v. U.S. Army Corps of Engineers,2 the D.C. Circuit upheld the vacatur of an easement to construct a major pipeline system known as the Dakota Access Pipeline (DAPL).3 In conducting its vacatur analysis, the Standing Rock court drew a key distinction: the vacatur inquiry looks to whether the agency could justify its procedural actions on remand, not whether it could justify its final decision. This approach is controversial4 but should become the new canon in vacatur analyses. By preserving the remand without vacatur remedy, but cabining its applicability, the Standing Rock rule helps align judicial practice with the twin mandates of the Administrative Procedure Act5 (APA), that a reviewing court “shall . . . set aside” flawed agency action while taking “due account” of “prejudicial error.”6 The Standing Rock decision also comes at an opportune time. With the Supreme Court yet to articulate its view on remand without vacatur, Standing Rock both fills the space left by the Court’s silence and offers a compelling analytic model for the Court should it seek to clarify the doctrine in this important area of administrative law.
In June 2014, Dakota Access Pipeline, LLC (Dakota Access) applied to the U.S. Army Corps of Engineers for an easement to construct an oil pipeline across the federally regulated waters of Lake Oahe in the Dakotas.7 The long, narrow lake — formed by an Army Corps dam in the 1950s — provides crucial water resources to the Standing Rock Sioux Tribe and other members of the Great Sioux Nation that have lands in the region.8 Dakota Access sought to transport oil from North Dakota to Illinois and selected a route that required the pipeline to pass underneath the lake on its way south.9 Crossing the lake required an Army Corps easement. To issue the easement, the Corps had to comply with the National Environmental Policy Act of 196910 (NEPA), which requires preparation of an Environmental Impact Statement (EIS) for any federal action likely to have a significant adverse effect on the environment.11 Instead of a full EIS, the Corps issued an Environmental Assessment (EA) and a mitigated Finding of No Significant Impact (FONSI).12 No EIS was necessary, the Corps argued, because the underground pipeline crossing of Lake Oahe would not “significantly affect the quality of the human environment.”13 Objecting to the Corps’s NEPA analysis, the Standing Rock and Cheyenne River Sioux Tribes sought a preliminary injunction against the Corps in federal court.14 Although the court denied the request for an injunction, the Departments of Justice, the Interior, and the Army issued a joint statement indicating that the Corps would reconsider its decision not to conduct an EIS.15 On January 18, 2017, the Assistant Secretary of the Army for Civil Works published a notice of intent to prepare an EIS in the Federal Register.16
Then came a change in administrative regime. The new Trump Administration sought expeditious approval of the DAPL project, and the Corps determined that in fact no EIS or further review would be necessary.17 On February 8, 2017, the Corps granted the easement.18 On motions for summary judgment, the District Court for the District of Columbia remanded the easement decision to the agency on the grounds that, inter alia, the Corps had violated NEPA by failing to consider sufficiently the extent that the project’s effects were likely to be “highly controversial.”19 After the Corps completed its analysis on remand, the Tribes again moved for summary judgment and the district court again concluded that the Corps had not satisfied its NEPA burden and must prepare an EIS.20 In a separate decision, the district court vacated the easement and ordered Dakota Access to “shut down the pipeline and empty it of oil.”21 Both the Corps and Dakota Access appealed.22
The D.C. Circuit affirmed the district court’s decision to remand and vacate the easement but reversed the order requiring Dakota Access to shut down the pipeline. Writing for the panel, Judge Tatel23 first held that unresolved disputes around DAPL’s leak-detection systems,24 operator-safety record,25 winter-conditions resilience,26 and worst-case discharge modeling27 rendered the decision “highly controversial” for NEPA purposes.28 Second, the court determined that vacatur was appropriate under circuit precedent.29 Third, the court held that the decision to shut down the pipeline must be analyzed separately from vacatur under the “traditional four-factor test” for an injunction.30 The court thus affirmed in part and reversed in part.
The D.C. Circuit began its analysis with the proper standard of review for NEPA violations. Citing its opinion in National Parks Conservation Ass’n v. Semonite,31 the court articulated that a decision is “highly controversial” if a “substantial dispute exists as to the size, nature, or effect of the major federal action.”32 In addition, “[t]he question is not whether the Corps attempted to resolve the controversy, but whether it succeeded.”33 Failure to satisfactorily resolve a sustained, credible set of concerns from “highly specialized governmental agencies and organizations” counseled in favor of preparing an EIS.34
The court then turned to the question of remedy and upheld the district court’s decision to vacate the DAPL easement. Unlawful agency action is presumptively subject to vacatur.35 Yet, under D.C. Circuit precedent, a court has discretion to remand a flawed agency decision without vacatur under certain circumstances.36 This analysis depends on a two-factor test laid out in Allied-Signal v. U.S. Nuclear Regulatory Commission37: (1) the “seriousness of the order’s deficiencies,” and (2) the “disruptive consequences” of vacating a decision that can be subsequently corrected.38 The first prong of the Allied-Signal test hinges on whether, on remand, there is a “significant possibility that the [agency] may find an adequate explanation for its actions.”39 More specifically, Judge Tatel emphasized that the likelihood of agency justification on remand applies to the relevant procedural requirement, not the ultimate agency decision.40 In other words, the court is to assess whether the agency could substantiate its decision not to issue an EIS, rather than whether it could substantiate its easement decision. The D.C. Circuit agreed with the district court that the flaws in the agency’s decisionmaking were such that the Corps would be unlikely to justify its decision to avoid an EIS.41 With respect to the second Allied-Signal factor, disruptive consequences, the court found that the district court had sufficiently wrestled with the severe economic effects of vacating the easement, and vacatur thus was not an abuse of discretion.42
Nevertheless, the D.C. Circuit reversed the pipeline shutdown order.43 In its analysis the court focused on the Supreme Court’s decision in Monsanto Co. v. Geertson Seed Farms,44 a NEPA case which held that “[a]n injunction should issue only if the traditional four-factor test is satisfied.”45 The D.C. Circuit emphasized that easement approval and pipeline operation were two distinct matters and that vacatur of the easement could not necessarily be said to require cessation of operation, as the pipeline would be an encroachment regardless of whether it was empty or full.46 The court thus found that the district court erred in not applying Monsanto, and reversed the lower court’s shutdown order.47
The Standing Rock court’s vacatur analysis clarified longstanding ambiguities in the canonical Allied-Signal test, and its focus on an agency’s discrete procedural steps ought to become the standard inquiry in remand without vacatur decisions. Flawed agency action should therefore face vacatur unless (1) the agency is likely to reach the same decision via the same procedures on remand or (2) the disruptive consequences of vacatur are prohibitively severe. With the second prong largely reserved for extreme circumstances, the first prong will decide the mine-run of cases. Under its terms, an agency will receive the benefit of a remand without vacatur only if its chosen procedural path merely requires further substantiation on remand. New procedures will merit vacatur. This interpretation of the D.C. Circuit’s Standing Rock rule threads the remand without vacatur needle by preventing agency abuse of judicial leniency while leaving space for remedial restraint at the margins. Should the Supreme Court take up the issue, the Court should endorse Standing Rock’s nuanced approach and clarify a significant doctrine that has percolated in the lower courts for over three decades.48
Substantial debate surrounds the question of whether courts have the discretion to grant remands without vacatur under the APA. The APA’s “shall . . . set aside”49 language arguably prohibits such remands, and Judge Randolph of the D.C. Circuit memorably argued that the remedy “rests on thin air.”50 Professor Nicholas Bagley disagrees, noting the APA provides that “due account shall be taken of prejudicial error,”51 and that “[t]here’s nothing to the argument that the APA, by its terms, strips courts of the authority to leave procedurally defective agency rules intact.”52 Under current D.C. Circuit precedent, Bagley’s view has won the day. Allied-Signal formalized that view, and the court regularly issues a handful of remand without vacatur orders each year.53 This remedial flexibility has substantial advantages, such as allowing a court to leave a partially protective (but flawed) rule in place while the agency corrects the defects on remand.54 However, it has significant disadvantages as well, including encouraging agency delay,55 reducing political accountability,56 and potentially weakening administrative legitimacy by allowing invalidated rules to remain on the books.
The Supreme Court has never formally sanctioned remand without vacatur, but the remedy has proliferated in the lower courts.57 Recent developments at the Court, however, indicate that the Justices may harbor reservations about the practice. In Department of Homeland Security v. Regents of the University of California,58 the Court held that reliance on a cabinet-level memo issued after the agency action in question was impermissible post hoc rationalization.59 Following the Court’s logic, remand without vacatur seems in peril: if an agency cannot rely on a supplemental memo from a cabinet secretary, it seemingly cannot rely on a series of post hoc rationalizations generated on remand.60 As Professor Benjamin Eidelson notes, the Roberts Court appeared in Regents to endorse the argument that “[t]he interim changes that [remand without vacatur] avoids are sometimes essential to ensuring meaningful political accountability for the agency’s revised reasoning.”61 Vacatur requires a new — and potentially politically costly — agency decision. Remand without vacatur proceeds largely in the dark.
The Standing Rock rule would help alleviate these concerns. By limiting most remands without vacatur to cases in which the agency convincingly asserts that it will proceed via the same procedural steps on remand, the rule would prevent agencies from circumventing key procedures and correcting their errors without political consequence. While the rule would have limited effect on minimum-procedure informal adjudications, such as the immigration policy at issue in Regents, its core contribution would be to extend the logic of Regents to more complex and procedure-laden agency actions where a party can allege discrete procedural errors. Remand without vacatur would likely become less prevalent under such an approach. But it would ensure that the remedy remains confined to the most appropriate circumstances of major disruption or truly harmless error.
An additional benefit of reducing the frequency of remand without vacatur arises in the context of cross-administration remands. In the distinct class of agency decisions that are both politically salient and that occur between administrations of different political parties, remand without vacatur can lead to the odd result of entrenching a decision rejected by both the judicial and executive branches. In Standing Rock, this situation was avoided by the Biden Administration’s decision to allow the DAPL EIS process to continue, but the Administration faced significant pressure to withdraw support for the pipeline.62 Had the Administration ultimately decided the other way — as well it could have — a remand without vacatur would have left in place a decision that the courts had deemed unlawful and the Executive deemed unwise.
For further evidence of the strange results that can ensue when a court issues a cross-administration remand without vacatur, consider the D.C. Circuit’s opinion in Vecinos para el Bienestar de la Comunidad Costera v. FERC,63 decided just seven months after Standing Rock. The court again reviewed a deficient NEPA analysis for a pipeline approval, again from a previous administration, but this time granted remand without vacatur.64 The agency — in this instance the Federal Energy Regulatory Commission (FERC) — had a highly unusual reaction. Its new Chairman, appointed by President Biden, celebrated the remand of his own agency’s decision, issuing a press release that cited his dissents from the original approvals and declared his intention to “expeditiously update” FERC’s policy on gas infrastructure.65 The remand without vacatur remedy — grounded as it is in antipathy to disruption — does not appear built for the shifting tectonics of regime change.
An intriguing inconsistency also arises between Standing Rock and Vecinos with respect to the discrete procedural-step analysis. The Vecinos court held that certain NEPA-implementing regulations required FERC to at least consider whether it would need to apply a “generally accepted” scientific protocol to compute certain climate change effects.66 The court insinuated that FERC might indeed have to adopt such a new computational protocol, as “[t]he regulation appear[ed] applicable on its face.”67 Yet the court remanded without vacatur because it found the agency could likely reach the same end result — even if via different procedures — on remand.68 The Vecinos court thus seemed to blur the lines that the Standing Rock court drew so sharply between discrete procedural requirements and a final approval decision. Should the Supreme Court adopt the rule from Standing Rock, the inquiry would focus on whether the agency must undergo a new and significant procedural step on remand, in which case vacatur would follow.
The Standing Rock approach to remand without vacatur accommodates both the Roberts Court’s emphasis on administrative accountability and a pragmatic model of how agencies actually behave. It prevents agencies from grappling with serious procedural defects under cover of a remand without vacatur order and rightfully refocuses judicial attention on the specifics of agency procedure. As Judge Tatel argued in another context, administrative procedures are not mere roadblocks; they “keep agencies tethered to Congress and to our representative system of government.”69 The test articulated in Standing Rock retains the strength of these tethers and offers a workable, accountability-forcing mechanism for deciding questions of vacatur. For regulated parties and regulatory beneficiaries, the sword will cut two ways: both proregulatory and deregulatory action will be subject to scrupulous procedural requirements. The substantive consequences of a strong reading of Standing Rock are thus difficult to gauge. But the conceptual and practical justifications for taming the remand without vacatur remedy are sound, and the Supreme Court would do well to formalize a version of Standing Rock’s elaboration of this key remedial doctrine.