Introduction

Imagine that a federal agency promulgates a new regulation that establishes a uniform process for adjudicating certain federal benefits. The U.S. Court of Appeals for the Eleventh Circuit, which has jurisdiction over three U.S. states, invalidates a benefits decision under that regulation, holding that the agency must use a different process. Now, the agency faces some choices: Should it avoid using its new adjudication process altogether, even in the other 47 states? Should it continue to use its new process in other states and defend that process in different courts? Can it apply that process to other benefits claimants in the Eleventh Circuit who may not have been a party to the original case?

Some of the agency's possible choices in this situation would involve a practice known as agency nonacquiescence. This report discusses the various forms of agency nonacquiescence, examines the elements of the relationship between federal courts and agencies that make nonacquiescence possible, and provides an overview of the arguments for and against the practice.

When Congress delegates the power to regulate to a federal agency, it often also provides for judicial review of that agency's actions.¹ Reviewing courts may "set aside" agency action that is arbitrary and capricious or contrary to law.² Agency nonacquiescence may arise after a court sets aside agency action because the court disagrees with the agency's interpretation of law. The agency must determine whether it will conform its future actions both internally and before the federal courts to the court's interpretation of law—acquiescence—or whether it will continue to apply its preferred interpretation in future actions—nonacquiescence. Nonacquiescence is possible primarily, although not exclusively, in situations when a reviewing court overturns a decision made by an agency through adjudication. Although the term nonacquiescence can apply to situations in which an agency disregards a court order as it applies to a specific party before the agency, this report focuses on situations in which an agency declines to adopt a court's reasoning as it applies to future actions and different parties before the agency.

From the agency's perspective, nonacquiescence raises two related issues: First, what other courts or agency adjudicators are bound to adopt the court's legal determination in a future case or a future proceeding before the agency (i.e., the precedential effect of the decision)? Second, what entities are bound by the court's order in the initial case challenging the agency action (i.e., the scope of relief)? The precedential effect of a court's decision and the scope of the relief it grants are governed by a number of features of the federal system that make nonacquiescence an available option for an agency. While there are significant constitutional questions about the legality of agency nonacquiescence (at least in some circumstances), nonacquiescence is made possible by a combination of features that either limit the precedential effect of a court's ruling or limit the scope of the court's relief. These features include the federal government's right to relitigate legal issues decided against it,³ the geographic limitations of the federal courts of appeals' jurisdiction,⁴ and the kind of action taken by the agency.⁵

A number of federal agencies have engaged in nonacquiescence of some kind since at least the 1940s. These agencies include the Social Security Administration,⁶ the Environmental Protection Agency,⁷ the National Labor Relations Board,⁸ the Internal Revenue Service,⁹ the Food and Drug Administration,¹⁰ and the Securities and Exchange Commission.¹¹ Some federal agencies, such as the Social Security Administration, have at times adopted broad nonacquiescence policies essentially declaring that they would not necessarily conform future actions to the decisions of the federal courts.¹² Others, however, have adopted more selective nonacquiescence policies aimed at addressing a single or small set of court decisions.¹³

While nearly all agree that an agency is bound by a court's interpretation of law as it applies to the parties to the case, there is significant debate about whether an agency must conform its future actions to the legal interpretations of a reviewing court's interpretation of law.¹⁴ The debate centers on the effect of a decision from a federal court of appeals.¹⁵ Some argue that Article III of the Constitution vests the entire federal judiciary (not just the Supreme Court) with the power to render interpretations of federal law that are binding on all entities within a specific court's geographic jurisdiction.¹⁶ Others contend that agencies are part of a coordinate branch of government vested with interpretive authority by Congress and accordingly, under certain circumstances, are not bound to adopt the legal interpretations of a reviewing court in its future actions.¹⁷ Rather, they have a duty to administer nationally uniform regulatory programs until the court system has settled on a nationally uniform interpretation.¹⁸ Congress has yet to address nonacquiescence in statute, but it is likely that Congress could limit or bar the practice were it to find it desirable.

Defining Agency Nonacquiescence

In its most general form, agency nonacquiescence refers to when a federal agency declines to follow a decision of a federal court interpreting a statute that the agency administers. There are, however, several variations of nonacquiescence, some more controversial than others:

• Intercircuit nonacquiescence refers to the practice of an agency refusing to follow the case law of one court of appeals in actions it takes that will be reviewed by a different court of appeals.
• Intracircuit nonacquiescence refers to the practice of an agency refusing to follow the case law of a court of appeals that will review the agency's decision.
• Venue choice nonacquiescence refers to a situation in which judicial review may be had in either a court that has rejected the agency's position or a court that has not.¹⁹

No matter the form, nonacquiescence raises fundamental questions about the separation of powers: first between the judiciary and the executive branch and second between Congress and the judiciary. Nonacquiescence raises the specter of the executive branch disregarding the legal pronouncements of the federal courts and prompts questions about the judiciary's role to announce generally applicable legal rules both within and across the geographic boundaries of the federal circuit courts of appeals.

Under certain circumstances, nonacquiescence can be a legally available but controversial option for an agency. Nonacquiescence can be legally available when (1) the federal government is free to relitigate a legal issue it lost in a prior case,²⁰ (2) the region in which the agency is taking action is not subject to the jurisdiction of the court that ruled against the agency,²¹ (3) the relief granted by the federal courts to parties challenging the agency's action is limited to the parties' case,²² (4) the challenged agency action applies to a limited number of parties before the agency,²³ and (5) there is more than one judicial venue in which to challenge the agency action.²⁴ None of these features taken individually or together authorizes any particular instance of agency nonacquiescence, but taken together they create the structural underpinnings making nonacquiescence possible. This report addresses each of these elements in turn.

The Federal Government's Right to Relitigate

Collateral estoppel, also known as issue preclusion, is a judicially developed doctrine that bars a party from relitigating in a subsequent case (in any court) any issue that was actually decided and material to the outcome of a prior case.²⁵ Collateral estoppel creates finality in court decisions in civil cases, ensuring that parties are not required to relitigate an issue that was already decided in a prior case.²⁶ It also serves to conserve judicial resources from being consumed by parties litigating the same issue more than once and prevents inconsistent decisions across multiple courts.²⁷ Collateral estoppel can apply not only when both parties are the same but also when only the defendant is the same but the plaintiff is a different party than in the first case.²⁸ This latter type of collateral estoppel is known as nonmutual offensive collateral estoppel.²⁹ In these types of cases, although the plaintiff was not a party to the original case, the defendant was.³⁰ As such, a plaintiff may assert that the defendant is bound by the decision of a prior case on an issue that arises in a subsequent case.³¹ Under this form of collateral estoppel, defendants are precluded from relitigating issues they lost in prior cases, even in situations where the plaintiff suing the defendant is different.³²

The federal government is not bound by nonmutual offensive collateral estoppel, allowing the federal government to relitigate an issue it lost in a prior case.³³ In United States v. Mendoza, the Supreme Court held that nonmutual offensive collateral estoppel does not apply to the federal government when the government is a defendant in a case and wishes to relitigate an issue that it lost in a prior case.³⁴ Mendoza, a Filipino national whose petition for naturalization was denied, argued that the federal government violated his Fifth Amendment Due Process rights by denying his petition.³⁵ Some years earlier, a group of sixty-eight Filipino nationals sued the federal government for denying their petitions for naturalization, asserting the same constitutional challenges.³⁶ The district court ruled in favor of the plaintiffs, finding that their Fifth Amendment rights were violated.³⁷ The federal government did not appeal the case or seek to have the judgment modified or set aside.³⁸ When Mendoza brought suit against the federal government for denying his naturalization petition on the same grounds as those that were brought by the sixty-eight Filipino nationals in the earlier case, Mendoza argued that the United States could not relitigate whether denying his naturalization petition violated the Fifth Amendment.³⁹ The district court and the court of appeals agreed and held that the United States was bound by the earlier decision.⁴⁰ The Supreme Court reversed, holding that the United States is not subject to collateral estoppel.⁴¹

The Court explained that the United States is unique among civil defendants due to the number of cases brought against it. For example, some cases addressing the legality of government conduct can be brought only against the federal government.⁴² The United States, the Court conveyed, is therefore more likely than any other civil defendant to be engaged in litigation against different parties that nonetheless raise the same issues.⁴³ Were the United States bound by collateral estoppel from relitigating those issues, the Court continued, this would "substantially thwart" the development of important legal issues at the first decision.⁴⁴ That would deprive the Supreme Court of the benefit of the dialogue among multiple federal courts exploring the most difficult legal questions.⁴⁵ That process is sometimes called percolation.⁴⁶ The Court has confirmed on multiple occasions its view that this dialogue among multiple courts helps develop and sharpen the most important issues, which in turn can signal to the Supreme Court that an issue is of such importance that it should hear an appeal.⁴⁷ Ultimately, without the ability to relitigate issues the government lost in prior cases, agency nonacquiescence would be all but impossible. Because collateral estoppel does not bar the federal government from relitigating an issue it lost in a prior case, an agency that lost on an issue before the federal courts can relitigate that same issue against a different plaintiff in a future case.

Percolation of Legal Issues Across the Federal Courts

The process of percolation points to other features of the federal court system that permit agency nonacquiescence: the geographic division of the lower federal courts of appeals and the absence of a requirement that courts in one circuit follow precedents from other circuits.⁴⁸ In other words, the federal court system does not require intercircuit stare decisis. Stare decisis refers to "[t]he doctrine of precedent, under which a court must follow earlier judicial decisions when the same points arise again in litigation."⁴⁹ As an initial matter, most agree that a decision of the Supreme Court interpreting federal law is binding precedent for not just the lower courts but also federal agencies.⁵⁰ Two elements of the Court define its role as the final arbiter of federal law. First, the Supreme Court has national jurisdiction, meaning its decisions are binding everywhere in the nation.⁵¹ Second, through a combination of two Supreme Court cases, the Court has determined that it is the final arbiter of federal law. In the seminal 1803 case Marbury v. Madison, the Court explained that "it is emphatically the province and duty of the judicial department to say what the law is."⁵² Read narrowly, however, Marbury establishes that a court's interpretation of federal law is binding as applied to the parties to the case, leaving room for other branches to continue to interpret federal law in future actions.⁵³ It was not until the 1958 case of Cooper v. Aaron that the Supreme Court, relying on Marbury, explicitly held that it was the supreme and final arbiter of questions of federal law.⁵⁴ The federal courts of appeals, however, do not enjoy the same position as the Supreme Court. The geographic limitations of the lower appeals courts' jurisdiction and their intermediate position underneath the Supreme Court limits (although does not extinguish) their ability to make nationally binding determinations of federal law. Nonacquiescence is thus a possible response to decisions from the lower federal courts but likely not to decisions of the Supreme Court.

Congress divided the federal courts of appeals into thirteen different circuits, twelve of which have jurisdiction over particular regions of the United States. (The Third Circuit's jurisdiction extends to Pennsylvania, New Jersey, and Delaware, for example.⁵⁵) The Court of Appeals for the District of Columbia Circuit [D.C. Circuit] bears special mention. Although it is a regional circuit court like the other eleven circuits, because the District of Columbia is the seat of the federal government, decisions of the D.C. Circuit often bind an agency nationwide. Various statutes also limit appeals from particular agencies to the D.C. Circuit only.⁵⁶ In those cases, the D.C. Circuit effectively has nationwide jurisdiction.

Figure 1. Geographic Boundaries of the Federal Courts

Source: https://www.uscourts.gov/sites/default/files/u.s._federal_courts_circuit_map_1.pdf. Notes: District court boundaries are delineated by gray lines within state boundaries.

A result of the structure of the federal court system is that when one circuit resolves a question of law, the other circuits are free to adopt or reject that determination.⁵⁷ The Judiciary Act of 1891 (commonly referred to as the "Evarts Act") created the modern federal circuit court system.⁵⁸ Section 6 of the Evarts Act provides that "the judgments or decrees of the circuit courts of appeals shall be final in all cases," but it does not specifically address how circuit courts should treat the decisions of their sister circuits.⁵⁹ Whether intended by Congress or not, early decisions of the newly created circuit courts of appeals started to assert a duty of "independent judgment in cases of first impression" in a particular circuit.⁶⁰ What ultimately arose is a system of intercircuit dialogue where multiple circuit courts may have the opportunity to analyze the same legal issue, learning from the decisions of other circuits.⁶¹ As noted above, the Supreme Court has explained that this process of percolation helps improve the quality of legal decisionmaking at both the Supreme Court and the courts of appeals and helps signal to the Supreme Court which legal issues may warrant its attention.⁶²

Although a circuit is not bound by the decisions of its sister circuits, earlier decisions of a court of appeals are generally binding precedent in future cases on the same question of law within a circuit.⁶³ This feature of the federal courts is known as the law of the circuit doctrine.⁶⁴ By statute, circuit courts usually hear appeals in three-judge panels.⁶⁵ To prevent the same panel from being overwhelmed by hearing every appeal, Congress created multiple judgeships for each circuit court.⁶⁶ The First Circuit, for example—the smallest circuit—has six judgeships, while the Ninth Circuit—the largest circuit—has 29 judgeships.⁶⁷ Decisions of these panels are binding on the decisions of future three-judge panels on the same question of law.⁶⁸ A circuit court can overturn its own precedent, however, through a special circuit court procedure known as a rehearing en banc.⁶⁹ A rehearing en banc is a hearing before a larger panel of appellate judges. In smaller circuits, all active judges on the court of appeals participate in en banc proceedings, while in larger circuits a subset of active judges may participate.⁷⁰

For agency nonacquiescence, the absence of intercircuit stare decisis combined with the inapplicability of collateral estoppel to the federal government means that if one circuit rules that an agency's interpretation of a statute is wrong, the agency might have the option to continue to apply its preferred interpretation in other circuits that either have not ruled on the issue or have approved the agency's interpretation. In other words, the agency may be able to engage in intercircuit nonacquiescence because the circuit court's decision overturning the agency's statutory interpretation is binding precedent only in that circuit's geographic region.⁷¹

Due to the law of the circuit doctrine, however, it is controversial for an agency to engage in nonacquiescence within the circuit that has ruled against it (intracircuit nonacquiescence). The government is not bound by collateral estoppel to accept a past decision, but the court itself is bound by its own past decisions in most circumstances. Intracircuit nonacquiescence therefore requires the agency to rely on the limited mechanisms that allow a circuit to revisit a past decision. Whether this practice is legally permissible has generated a robust debate among scholars. Some argue that intracircuit nonacquiescence violates the separation of powers because an agency encroaches on the judiciary's authority to "say what the law is" when it decides for itself not to follow the precedent of the circuit in which it takes action.⁷² Others argue that agencies are not bound by precedent the way district courts in a circuit are bound by circuit precedent.⁷³ Rather, agencies are entrusted by Congress to develop nationally uniform policies.⁷⁴ While agencies are bound to follow circuit court decisions with respect to the parties to a specific case, these scholars argue, unless subject to an injunction requiring it, internal agency practices may not need to conform to circuit court decisions, and the agency should be free to relitigate the issue it lost until the legal system settles on a nationally uniform rule.⁷⁵ Despite the lingering questions of its legality, agencies have engaged in this kind of nonacquiescence from time to time, citing the need for uniform national policy.⁷⁶

Statutory Venue Provisions

The term venue refers to the court or courts in which it is appropriate for a lawsuit to proceed.⁷⁷ The question of which court will hear an appeal from an agency also bears on whether an agency decides to engage in nonacquiescence. As discussed in more detail below, statutory venue provisions often direct challenges to agency actions to a particular court or courts and thus may permit an agency to predict which circuit would hear challenges to its actions. If a venue provision directs an appeal from an agency to a circuit that has already ruled against the agency, the agency may choose not to engage in intracircuit nonacquiescence because the agency's chances of prevailing are slim (although, as will be discussed below, some agencies knowingly engage in intracircuit nonacquiescence). Conversely, if the appeal will be heard in a court that has not yet ruled on the issue, the agency may wish to continue to press its preferred legal interpretation even if another circuit has ruled against the agency.

Whether the appeal will be in a different circuit is determined in part by venue, which is defined by statute. Outside of the original jurisdiction provisions in Article III of the Constitution, which apply only to the Supreme Court,⁷⁸ Congress has authority to decide which courts will hear what cases.⁷⁹ Unless a more specific statute applies, the general venue statute provides in relevant part that a plaintiff can bring a case against an agency in a judicial district (i.e., district court) where a defendant resides, where a substantial part of the events giving rise to the suit took place, or where the plaintiff resides.⁸⁰ For a challenge to an agency action, the general venue provision would permit suit either where the agency is headquartered (possibly Washington, D.C.), in the judicial district that encompasses the office that took the action (if different from the headquarters), or in the judicial district where the plaintiff lives.

Congress has enacted numerous agency-specific statutes that include venue provisions for challenges to particular agencies' actions. For example, the Administrative Orders Review Act (commonly known as the Hobbs Act) directs challenges to orders from certain agencies—such as the Federal Communications Commission—to the circuit court of appeals in which the petitioner resides or has its principal office or to the D.C. Circuit.⁸¹ Other examples include the Clayton Act, which permits those subject to cease-and-desist orders issued by certain agencies to challenge those orders in the circuit where the alleged violation occurred or where the target of the order resides or carries on business.⁸² Unlike the general venue provision, the Hobbs Act and the Clayton Act permit appeals from certain agencies to go directly to the courts of appeals, bypassing the district courts. Whether an appeal will go to a district court or an appellate court, however, is immaterial to the nonacquiescence analysis. Both district courts and appellate courts are bound by prior decisions of the court of appeals in the circuit to which a venue provision directs the appeal. Finally, some statutes vest a single court with jurisdiction over challenges to certain agency actions, essentially providing that court with nationwide jurisdiction in those appeals.⁸³ Statutes that limit appeals to a single court prevent both intercircuit dialogue and also intercircuit nonacquiescence. Statutes limiting venue to a single court may also limit the frequency of nonacquiescence, as every instance of nonacquiescence would result in the more controversial intracircuit nonacquiescence. Further, a single court may be able to engage in more effective oversight of an agency's nonacquiescence policies (to the extent it has them) than would be possible if an appeal could be heard in multiple courts.

Narrow venue provisions, such as those that direct challenges to a single court, permit agencies to predict which court will hear an appeal and thus permit them to choose whether to engage in inter- or intracircuit nonacquiescence. Venue provisions that permit challenges to be brought in multiple courts, however, limit an agency's ability to predict what court might hear an appeal. When an agency takes an action contrary to a prior circuit court ruling but, due to a broad venue provision, is unsure whether a challenge to the action would be heard by that court, this is sometimes referred to as venue choice nonacquiescence.⁸⁴ An agency in this situation may hope that the case will be heard in a circuit that has not yet weighed in on the question or has ruled in its favor. Because an agency cannot predict which court will hear an appeal at the time it acts, the agency's action may be reviewed in a court that had previously ruled against the agency on that issue. In that case, although the agency may intend to engage only in intercircuit nonacquiescence, its action would look to the reviewing court like the far more controversial intracircuit nonacquiescence.⁸⁵ Further, broad venue provisions may mean that some of the courts where the agency action could be challenged have ruled against the agency, while others have ruled in its favor.⁸⁶ Under this scenario, no matter what action the agency takes, it will engage in nonacquiescence. Again, depending on where a challenge is brought, it may appear to the court that the agency was engaging in intracircuit nonacquiescence when it was in fact the agency's intent to continue to administer a nationally uniform policy in the face of conflicting judicial decisions.

Agency Actions and Judicial Review Under the Administrative Procedure Act (APA)

The features of the federal system discussed up until this point have affected the precedential value of a reviewing court's decision—whether the reviewing court has identified a legal rule that applies to the agency's actions. The following section discusses the interaction between the type of agency action at issue and the remedy applied by the reviewing court—whether the reviewing court in a particular case has issued an order that the agency must follow in its future decisions. The type of agency action and the remedy applied by the reviewing court both bear heavily on whether the agency has the option to engage in nonacquiescence in a future action. Rather than limiting the precedential value of the reviewing court's legal determination in future proceedings, the type of agency action and the remedy imposed by the court often determines what entities are bound by an order from the first court to evaluate the agency's action.

Agency Action Under the APA

The APA classifies agency action in several ways, two of which are relevant to this report: rulemaking and adjudication. First, the APA defines rulemaking as the process by which the agency formulates, amends, or repeals a rule.⁸⁷ A rule under the APA is "the whole or a part of an agency statement of general or particular applicability and future effect designed to implement, interpret, or prescribe law or policy."⁸⁸ An adjudication, in contrast, is the agency process for formulating an order—that is, "the whole or a part of a final disposition . . . of an agency in a matter other than rule making."⁸⁹ A common way to differentiate between rules and orders is that rules regulate future conduct, while orders remedy past conduct.⁹⁰ Another way to differentiate between the two that is more relevant to the nonacquiescence discussion is that orders apply to the parties to the adjudication, while rules usually apply generally.⁹¹ As will be discussed below, the difference between the particular applicability of orders and generally applicable rules plays an important role in an agency's ability to engage in nonacquiescence.⁹²

Judicial Review Under the APA

The APA provides for judicial review of final agency action.⁹³ Section 706 of the APA permits a reviewing court to "hold unlawful and set aside" agency action if it violates one of six standards of review.⁹⁴ The most commonly invoked standard is the "arbitrary and capricious" standard of review, which permits a court to set aside agency action if it is "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law."⁹⁵

The type of agency action a reviewing court sets aside (i.e., an order or generally applicable rule) determines whether an agency will have the option to engage in nonacquiescence in future actions. When a court sets aside a rule, the rule no longer has legal effect.⁹⁶ The agency cannot then rely on that rule for subsequent actions, such as enforcement actions, against any party nationwide.⁹⁷ If the agency were to carry on as if the rule were still valid, it would be directly disobeying a court order and could be subject to contempt sanctions.⁹⁸

When a court sets aside an agency order, just as with a rule, the order no longer has legal effect. The difference is that the order touches only the parties to the agency adjudication. The effect of a court setting aside an order is accordingly much narrower than when a court sets aside a rule. Further, although agency adjudications may implicate generally applicable rules or policy, each adjudication is a discrete action, and it is that discrete action that the court sets aside—not the generally applicable rule. The upshot is that unless the court issues an injunction, an agency may be able to continue to apply its preferred policy in future adjudications.

The Effect of Injunctions on Nonacquiescence

Another factor that affects whether an agency can engage in nonacquiescence is whether a reviewing court issues an injunction and, if so, the scope of that injunction. An injunction is a court order to either take an action or refrain from taking an action.⁹⁹ A court can issue an injunction in addition to deciding the legal and factual merits of a case.¹⁰⁰ For instance, if a court finds that a company violated a pollution control statute, it might order the company to stop discharging the pollution into the environment or to clean up the pollution that it had discharged. In addition to ordering a party to do or not do something, an injunction can vary in scope. For instance, a court could order a defendant from acting against the plaintiff only, or it could bind the defendant's future actions as they apply to any person.¹⁰¹ An injunction can accordingly help clarify which parties are bound by a court's decision. Injunctions also permit a court to more closely manage a party's compliance with its decision. If an agency refuses to act in accordance with a precedential legal rule, it is at risk of losing its next case in which that rule applies. If an agency refuses to obey a court-ordered injunction, in contrast, it is at risk of immediate contempt proceedings before the same court.

Returning to the nonacquiescence context, the issuance of an injunction may affect the ability of an agency to engage in nonacquiescence.¹⁰² If a court sets aside an agency's adjudication order, the effect of the court's decision is to nullify the legal effect of the order, which applies only to the parties to that order. This limited application arguably leaves the agency free to continue to apply its preferred legal interpretation in future actions outside of the geographic jurisdiction of the issuing court, or, put another way, to engage in intercircuit nonacquiescence. An agency could still engage in intracircuit nonacquiescence in this scenario but would likely be unsuccessful in any future court challenge in the same circuit, because precedent would bind a future court to adopt the earlier court's legal determination. An injunction, however, could bind the agency even in future cases that involve other parties and thereby limit the agency's ability to apply its preferred legal interpretation in future actions.¹⁰³

For example, a court could enjoin an agency from enforcing an order against the parties to the underlying agency adjudication. This limited injunction would still leave open the possibility that the agency could engage in inter- or intracircuit nonacquiescence in future actions. A court could also issue an injunction that bars the agency from issuing an order based on the same legal grounds to anyone within the geographic jurisdiction of the court or, more controversially, to anyone nationwide. In the case of an injunction that extends to the geographic jurisdiction of the issuing court, the agency would be unable to engage in intracircuit nonacquiescence without violating a court order and subjecting itself to contempt sanction, but it could still engage in intercircuit nonacquiescence. Injunctions that prohibit an agency from applying certain rules or policies in any case thus limit the agency's ability to engage in nonacquiescence. In the case of a nationwide injunction, the agency could not engage in nonacquiescence in any jurisdiction. Were it to do so, the agency could be subject to contempt sanctions in the court that issued the injunction.

Examples of Agency Nonacquiescence

Social Security Administration: Intracircuit Nonacquiescence

Perhaps the most notable example of agency nonacquiescence occurred between the 1960s and 1980s at the Social Security Administration (SSA) in its disability benefits program.¹⁰⁴ The SSA administers a disability benefits program pursuant to the Social Security Act of 1935.¹⁰⁵ The Social Security Act permits a claimant to appeal a benefits determination to federal district court.¹⁰⁶ The Act's venue provision provides for district court review in the district "in which the plaintiff resides or has his principal place of business."¹⁰⁷

Until 1985, the SSA maintained a public policy that directed administrative law judges not to consider lower federal court decisions (trial or appellate) that were inconsistent with the SSA's interpretation of federal law as binding on future cases.¹⁰⁸ The SSA supported its position by arguing that "[t]he federal courts do not run SSA programs, and [SSA adjudicators] are responsible for applying the [SSA's] policies and guidelines regardless of court decisions below the level of the Supreme Court."¹⁰⁹ The agency went on to explain that its "policy of nonacquiescence is essential to ensure that the agency follow its statutory mandate to administer the Social Security program in a uniform and consistent manner."¹¹⁰

Judicial Reaction to the SSA's Nonacquiescence

The SSA's intracircuit nonacquiescence policy came to a head in a series of court decisions condemning the practice. During the 1980s, the SSA adopted a policy allowing it to reduce Social Security benefits to a particular beneficiary even in the absence of any evidence that the beneficiary's condition had improved.¹¹¹ The Ninth Circuit condemned that practice on two occasions, ruling that the SSA could not reduce benefits unless SSA relied on evidence that the beneficiary's medical condition had improved.¹¹² Following these two decisions, the SSA issued a formal notice explaining that it would comply with the Ninth Circuit's decision as it applied to the parties to the case but would continue to apply its review standard in future cases, including those that would be appealed in the Ninth Circuit.¹¹³ A short time later, an SSA beneficiary appealed his termination of benefits to the district court in the Central District of California, one of the districts within the Ninth Circuit.¹¹⁴ The case was brought as a class action that included all SSA claimants who resided in the Ninth Circuit.¹¹⁵ Because the case was brought as a class action, all the members of the class were parties to the case.¹¹⁶ Further, because the case was brought in a district court in the Ninth Circuit, the district court was bound to follow Ninth Circuit precedent that required evidence of medical improvement before the SSA terminated benefits. The district court found for the plaintiffs, and the district court also entered a class-wide injunction requiring the SSA to follow Ninth Circuit precedent in any benefits determination for any member of the class residing in the Ninth Circuit.¹¹⁷ The district court, relying on Marbury, declared that "governmental agencies, like all individuals and other entities, are obliged to follow and apply the law as it is interpreted by the courts."¹¹⁸ The court further declared that "for the Secretary to make the general assertion that a decision of the Court of Appeals is not to be followed because she disagrees with it is to operate outside the law."¹¹⁹

The district court also stressed the unequal treatment introduced by the SSA's policy for claimants that do not have the funds or inclination to appeal their benefits determinations to federal court. The court called the SSA's policy "prejudicial and unfair."¹²⁰ The Ninth Circuit affirmed the injunction except as applied to benefits denied before the Ninth Circuit held that medical improvement was a necessary condition to the termination of benefits.¹²¹ In upholding the injunction, the Ninth Circuit held that "the Secretary, as a member of the executive, is required to apply federal law as interpreted by the federal courts cannot seriously be in doubt."¹²²

Courts in the Second Circuit rejected the SSA's intracircuit nonacquiescence policy on similar separation of powers grounds and concerns that the SSA's policy treated claimants who did not appeal their benefits determinations differently from those who did.¹²³ Echoing his Ninth Circuit colleagues, one district court judge noted that the SSA's nonacquiescence policy "was inconsistent with the constitutionally required separation of powers."¹²⁴

Congressional Reaction to the SSA's Nonacquiescence Policy

The SSA's nonacquiescence policy also received sustained attention from Congress. In 1984, both the House and the Senate debated including provisions in the Social Security Disability Benefits Reform Act that would limit or prohibit the SSA's nonacquiescence policy. The House version of the bill would have prohibited the SSA from engaging in intracircuit nonacquiescence unless the SSA sought Supreme Court review of an adverse ruling.¹²⁵ If the Supreme Court declined to hear the appeal or the SSA's appeal was dismissed, it would have to acquiesce to the lower court's ruling.¹²⁶

The Senate version of the bill took a different approach. The Senate version would not have banned nonacquiescence outright.¹²⁷ Rather, it would have required certain procedural safeguards, such as publishing a notice in the Federal Register and sending a report to certain Senate committees if the agency intended to engage in nonacquiescence.¹²⁸ Both the House and the Senate provisions were stripped from the bill in conference.¹²⁹ The conference report, however, stated that the decision to eliminate the nonacquiescence provisions should not "be interpreted as approval of 'non-acquiescence.'"¹³⁰

Environmental Protection Agency: Intercircuit Nonacquiescence

Beginning in 2012, the Environmental Protection Agency (EPA) adopted an explicit policy of intercircuit nonacquiescence under the Clean Air Act.¹³¹ EPA's policy grew out of litigation in the Sixth Circuit over the interpretation of its own regulations that determined when certain sources of emissions could be treated as a single source.¹³² The regulations required sources to be "adjacent" to be considered a single source.¹³³ EPA interpreted adjacent to include sources that were functionally related, irrespective of the distance between the sources.¹³⁴ EPA applied this interpretation to a natural gas refinery in Michigan, determining that the refinery and nearby natural gas wells owned by the refinery were a single source and were therefore required to obtain an emissions permit from EPA.¹³⁵ The refinery challenged EPA's interpretation of adjacent, and the Sixth Circuit found in favor of the refinery, holding that adjacent required physical proximity.¹³⁶

After the Sixth Circuit's decision, EPA issued a memorandum to its regional offices stating that outside of the Sixth Circuit, EPA had no intention to change its interpretation of adjacent.¹³⁷ A trade association that represented companies in areas outside the Sixth Circuit challenged EPA's policy in the D.C. Circuit, arguing that EPA's nonacquiescence policy violated its regulations that require EPA to "assure fair and uniform application by all Regional Offices of the criteria, procedures, and policies employed in implementing and enforcing [the Clean Air Act]."¹³⁸ The D.C. Circuit agreed and vacated EPA's memo.¹³⁹

In response to the D.C. Circuit's ruling, EPA amended its regulations to no longer require it to uniformly apply the Clean Air Act in response to an adverse court ruling.¹⁴⁰ The regulations also require a regional administrator to follow the decisions of a circuit court or district court with jurisdiction over that region, even if doing so would result in inconsistent application of EPA's Clean Air Act programs across circuits.¹⁴¹ The D.C. Circuit upheld EPA's amended regulations, finding that the Clean Air Act did not foreclose EPA's ability to engage in intercircuit nonacquiescence.¹⁴²

National Labor Relations Board (NLRB): Venue Choice Nonacquiescence

The NLRB has a long history of engaging in venue choice nonacquiescence.¹⁴³ The judicial review provision of the National Labor Relations Act (NLRA) permits a challenge to an order issued by the NLRB in multiple circuits, making it difficult for the NLRB to predict which circuit (if any) will hear a challenge to one of its orders.¹⁴⁴

The NLRA authorizes the NLRB to enforce the unfair labor practices provision of the Act.¹⁴⁵ Once the NLRB issues a final order, any aggrieved party can challenge the order (1) in the circuit where the unfair labor practice occurred, (2) where the person resides or transacts business, or (3) in the D.C. Circuit.¹⁴⁶ If the order creates multiple aggrieved parties—an employer and a union, for example—the number of circuits where the order can be challenged increases.¹⁴⁷

Possibly as early as the 1940s, but at least by 1957, the NLRB engaged in nonacquiescence.¹⁴⁸ In its 1957 decision in Insurance Agents International Union, the NLRB declared that administrative law judges adjudicating unfair labor practices cases must follow the directives of the Board and are not bound to follow circuit precedent.¹⁴⁹ The Board has expressed that it has congressionally delegated authority to ensure a nationally uniform administration of the NLRA and therefore to pursue its vision of national labor policy at the administrative level except where the Supreme Court has announced a different rule.¹⁵⁰ Additionally, because of the broad venue choice available under the NLRA, the Board cannot be certain which court of appeals will hear the case.

Judicial reactions to the NLRB's policy have been mixed. Earlier cases tended to look unfavorably on the NLRB's nonacquiescence practices, while some later cases have taken account of the NLRB's inability to predict in which court an appeal will be filed. For instance, in one early case addressing the NLRB's nonacquiescence, the Seventh Circuit compared the NLRB to a district court and held that the agency's decisions "are subject to review and consequent approval or disapproval by the reviewing body," ultimately holding that intracircuit nonacquiescence usurps the proper role of the courts.¹⁵¹ The Third Circuit took a similar view, stating that Congress "vested superior power for interpretation" in the courts, not the NLRB.¹⁵² As the junior tribunal, the court explained, the NLRB does not have the authority to disagree "respectfully or otherwise" with the pronouncements of the courts.¹⁵³ These early courts, however, assumed that the NLRB was engaging in intracircuit nonacquiescence without first evaluating the effect of the NLRA's venue provisions. More recent cases have taken a nuanced approach to the NLRB's nonacquiescence in light of venue uncertainty. While condemning intracircuit nonacquiescence, the D.C. Circuit called the NLRB's nonacquiescence "a function of ignorance, not defiance" due to the NLRA's broad venue provisions.¹⁵⁴ The court went on to explain that, although venue uncertainty can excuse what ultimately appears to the reviewing court to be intracircuit nonacquiescence, when the agency has good reason to know where a case will be filed, the agency engages in "improper" nonacquiescence.¹⁵⁵

Criticism and Support of Nonacquiescence

Intercircuit Nonacquiescence

Agency nonacquiescence has generated criticism in all of its forms. Although intracircuit nonacquiescence is by far the most controversial form, intercircuit nonacquiescence also has its critics. Critics of intercircuit nonacquiescence generally raise the point that intercircuit nonacquiescence undermines uniform application of the law by federal agencies.¹⁵⁶ This criticism, however, is difficult to square with Congress's choices both to create regional courts of appeals with geographically limited jurisdiction and to establish broad venue provisions for a number of agencies. It is also difficult to square with the Supreme Court's praise of intercircuit dialogue. Intercircuit dialogue would be impossible if agencies were required to adopt as a national rule the decision of the first court of appeals to decide on an issue.¹⁵⁷

Intracircuit Nonacquiescence

Separation of Powers

Intracircuit nonacquiescence has generated considerable debate primarily because it raises serious separation of powers concerns and charges of unfair treatment of those who cannot or choose not to pursue their claims in federal court.

Starting with the separation of powers concerns, critics of intracircuit nonacquiescence argue based on Marbury and Cooper that the federal judiciary has the constitutional authority to render authoritative interpretations of federal law that bind all entities within their geographic jurisdiction.¹⁵⁸ Further, they argue that the Take Care Clause of the Constitution, which states that "the President shall take care that the laws are faithfully executed," binds federal agencies to follow the judicial interpretation of federal law. As one scholar put it: "Because laws assume meaning only through interpretations, and because the judiciary has the 'peculiar province' of providing those interpretations, those 'faithfully executing' the laws would seem bound to heed legal interpretations provided by article III courts."¹⁵⁹

Scholars making the argument that intracircuit nonacquiescence violates the separation of powers often compare agencies to district courts in their duty to comply with precedent.¹⁶⁰ These scholars see statutory judicial review provisions as good evidence that, although Congress may have entrusted an agency with authority to administer a statute, agencies are subordinate to the federal courts in questions of the meaning of federal law.¹⁶¹

The existence of deference doctrines such as Chevron deference may limit the force of this argument, some scholars argue.¹⁶² Chevron deference, named for the Supreme Court decision in which it was articulated, is a judicially created doctrine that requires a court to defer to an agency's reasonable interpretations of ambiguous federal law it administers.¹⁶³ One core assumption of Chevron is that ambiguity in a statute is an implicit delegation of interpretive authority from Congress to the agency.¹⁶⁴ That Congress from time to time delegates authority to an agency to issue authoritative interpretations of a statute it administers indicates that Congress intended for the agency to be the primary policymaker, with the courts' role limited to ensuring that agencies are acting reasonably.¹⁶⁵ Some further argue that the fact that the judiciary itself has created and applied these deference doctrines, combined with the authority granted to agencies to administer federal statutes through rulemakings and adjudications, demonstrates that agencies are not simply district courts.¹⁶⁶ Rather, they occupy a different role as a coordinate branch of government that, at least in some circumstances, has the authority to interpret federal law for itself.¹⁶⁷

As at least two scholars have argued, federal agencies' location in a coordinate branch of government entrusted with statutory duties to administer regulatory programs indicates Congress's preference for administrative, rather than judicial, government.¹⁶⁸ These scholars point to federal agencies' "statutory duty to develop coherent, nationally uniform policies under their statutes" as a justification for nonacquiescence.¹⁶⁹ According to this view, agencies are not servants of the federal courts, and to see them that way ignores the "vital differentiations between the functions of judicial and administrative tribunals."¹⁷⁰ While the Supreme Court has not weighed in on the constitutionality of intracircuit nonacquiescence, it has come close to endorsing the view that agencies are not equivalent to district courts. As Justice Frankfurter's majority opinion in FCC v. Pottsville Broadcasting Co. explained:

The technical rules derived from the interrelationship of judicial tribunals forming a hierarchical system are taken out of their environment when mechanically applied to determine the extent to which Congressional power, exercised through a delegated agency, can be controlled within the limited scope of "judicial power" conferred by Congress under the Constitution.¹⁷¹

Supporters of the limited availability of intracircuit nonacquiescence have relied on Justice Frankfurter's opinion to support their contention that courts should not treat agencies like district courts.¹⁷²

There are few (if any) scholars, however, who endorse intracircuit nonacquiescence without reservation. Scholars who endorse the practice recognize the constitutional tension raised by intracircuit nonacquiescence but argue that in certain circumstances, other considerations weigh in favor of preserving the practice.¹⁷³ The most complete study of the practice proposed several limitations on the use of intracircuit nonacquiescence.¹⁷⁴ These limitations include a requirement that to engage in intracircuit nonacquiescence, an agency must make reasonable attempts to persuade the judiciary to validate its position nationwide through Supreme Court Approval.¹⁷⁵ This kind of limitation would likely rule out an agency consistently declining to appeal its losses to the Supreme Court to preserve its ability to engage in nonacquiescence. Other potential limitations include requiring a public reasoned justification for engaging in nonacquiescence and subjecting instances of intracircuit nonacquiescence to strict judicial scrutiny.¹⁷⁶

Practical Considerations: Uniformity and Fairness

Practical considerations also play a role in support for some uses of intracircuit nonacquiescence. Some have argued that agencies should be able to engage in intracircuit nonacquiescence in situations where the first circuit to address an issue has ruled against an agency's position but other circuits have found in the agency's favor.¹⁷⁷ Without the ability to relitigate the issue in the circuit that ruled against it, the agency will not have the opportunity to attempt to secure a nationally uniform rule.¹⁷⁸ Further, the first court may wish to revisit its earlier decision in light of the subsequent decisions in the agency's favor.¹⁷⁹ Intracircuit nonacquiescence in certain circumstances might further the percolation of legal issues across the federal court system and possibly help facilitate the selection of a nationally uniform rule.

The claim that an appellate court may want to revisit a previous adverse ruling against an agency, however, is debatable. Some have suggested that appellate courts rarely revisit prior rulings regardless of how other circuits have ruled.¹⁸⁰ If courts rarely revisit their prior rulings, there is little dialogue to be preserved in the intracircuit nonacquiescence context, and intracircuit nonacquiescence may just be a "vehicle for delaying ... justice."¹⁸¹

Further, intracircuit nonacquiescence is not without practical costs. While it may help facilitate a nationally uniform rule in some circumstances, some have argued that intracircuit nonacquiescence places those without the means or the inclination to appeal their cases to federal court at a disadvantage.¹⁸² For these scholars, the fact that those with the fewest resources are likely the ones that bear the costs of agencies' decisions to engage in intracircuit nonacquiescence weighs in favor of abandoning the practice altogether on fairness grounds alone.¹⁸³

Considerations for Congress

Congress has the authority to limit or bar agencies from engaging in nonacquiescence of any kind.¹⁸⁴ Under its constitutional power to define the powers and authority of administrative agencies, Congress can define in what situations (if any) nonacquiescence is permissible.¹⁸⁵ Agencies are "creatures of statutes," and, accordingly, Congress retains the power to limit or prohibit agencies engaging in nonacquiescence regardless of the inapplicability of collateral estoppel to the federal government and regardless of the lack of intercircuit stare decisis across the federal court system.¹⁸⁶

Congress has on occasion debated whether to limit or bar nonacquiescence. As noted above, both the House and the Senate debated whether to limit or bar entirely the SSA's ability to engage in intracircuit nonacquiescence.¹⁸⁷ Ultimately, the conference committee chose to leave out a specific provision addressing nonacquiescence at the agency, but it noted in its report that the absence of such a provision was not an endorsement of the practice.¹⁸⁸

In 1975, Congress created a commission dubbed the Commission on Revision of the Federal Appellate System to study the federal appellate court system, including agency nonacquiescence, and provide recommendations for improvement.¹⁸⁹ The commission was known informally as the "Hruska Commission" for its chairman, Senator Roman L. Hruska.¹⁹⁰ Nonacquiescence in both its inter- and intracircuit forms received extended unfavorable treatment by the commission, which noted that nonacquiescence contributed to uncertainty and confusion in areas of law in which an agency has engaged in nonacquiescence.¹⁹¹ The commission further expressed that nonacquiescence encourages forum shopping and permits differential treatment of similarly situated people.¹⁹² To remedy these issues, the commission recommended creating a new court with national jurisdiction that could issue nationally binding decisions.¹⁹³

Congress could approximate the recommendation of the Hruska Commission by limiting appeals of agency action to a single court. As described above, Congress has at times limited the venue of certain appeals to a single court, ensuring national uniformity without the creation of a new court.¹⁹⁴ Congress could limit the venue for appeals from particular actions, particular agencies, or for all actions taken by all agencies. Some scholars have advocated for limiting venue choice as a way to reduce some of the negative costs of nonacquiescence.¹⁹⁵