The Yale Law Journal

October 2020

The Attorney General’s Settlement Authority and the Separation of Powers

Administrative LawCivil ProcedureConstitutional Law

abstract. This Note presents a novel defense of the Attorney General’s authority to settle litigation against the United States and, in the process, make policy commitments. After canvassing the existing law and critiques of the settlement authority, the Note argues that the constitutional separation of powers does not forbid entering into policymaking settlements. The Note then proposes (1) new doctrine to make these settlements more consistent with administrative law norms, and (2) institutional norms and statutory changes that would improve the administration of settlement agreements.

author. Law Clerk, U.S. Court of Appeals for the Ninth Circuit; J.D. Yale Law School, 2020. My deepest thanks to Cristina Rodríguez for supervising this project, and to Jade Ford, Alex Lichtenstein, and Nicholas Parrillo for their help and insights. I am also grateful to Edward Pickup, Joshua Feinzig, Alexander Nabavi-Noori, and the editorial staff of the Yale Law Journal for their suggestions throughout the publication process. This Note reflects only my personal views and not the views of the U.S. Court of Appeals for the Ninth Circuit or any member thereof.


In 2018, protests erupted against the Trump Administration because of a new family-separation policy.1 The Administration (erroneously) blamed the Clinton-era Flores settlement agreement, publicly highlighting the obscure world of these agreements in federal policymaking.2 In fact, the Flores settlement emerged from a lawsuit in which the Department of Justice (DOJ) agreed to settle claims regarding the treatment of noncitizen children detained by immigration authorities.3 It “establishes a ‘nationwide policy for the detention, release, and treatment of minors’ in immigration custody,” which “binds the parties until the federal government promulgates final regulations implementing the agreement.”4 This settlement—like many others between DOJ and private parties—prescribes the policies to which the U.S. government must adhere, regardless of changing circumstances or political judgments.

The Trump Administration promulgated a Flores-settlement rulemaking that purports to “implement” the settlement and return control over this aspect of immigration policy to the executive branch.5 The rulemaking is of only dubious consistency with the underlying agreement,6 and the Trump Administration likely intended to overturn, in effect, the settlement agreement. The Ninth Circuit now may need to decide the critical question: whether preexisting settlement agreements can prevent agencies from promulgating new, otherwise-legal regulations.7

The Flores settlement highlights the importance of understanding the settlement power. By settling either constitutional or statutory claims against the government, DOJ can make policy for both current and future administrations. For example, in settling a case about whether a statute is constitutional, the United States might agree to refrain from enforcing certain aspects of the statute. Doing so effectively overturns part of the law without congressional action. Alternatively, the government might commit to implementing or to interpreting the law in particular ways. Such a commitment could take the form of a mandatory rulemaking or the forbidding of certain kinds of enforcement action. In effect, the administration could make policy without any of the traditional trappings of administrative law.

The settlement power raises fundamental questions about the policymaking relationship between the judiciary and the executive branch: To what extent can an administration bind itself and its successors to particular policies or actions that would otherwise remain discretionary? How can long-term judicial oversight of federal policy be consistent with the executive branch’s duty to faithfully execute the law? Do policymaking settlements unduly transfer federal power to private plaintiffs, who can “collude” with friendly administrations to enshrine favorable approaches to huge swaths of policy entrusted to the executive branch?

Opposing views have developed on this topic. Under one theory, the unbridled power of the executive branch to settle litigation and constrain future exercises of discretion abrogates the prerogatives of future executives and potentially undermines democracy itself.8 On another theory, Congress has imposed few external constraints on the Attorney General’s control of litigation, including the settlement power.9 As a result, the Attorney General retains broad discretion to settle litigation on terms she deems appropriate, including limits on future executive-branch action. Potentially implicit in this understanding are countervailing separation-of-powers concerns that imposing limits on the settlement authority would infringe the executive’s authority to execute the law.10 Neither account is complete, necessitating a new separation-of-powers perspective on this timely topic.

After demonstrating the insufficiency of prior separation-of-powers accounts of the settlement power, I present a novel argument for how policymaking settlements can be reconciled with both the constitutional separation-of-powers and traditional administrative law principles. The constitutional separation of powers clearly permits the executive branch to execute its vision of the law, including its view of the illegality of current practices and policies, by settling lawsuits. The judiciary should avoid ex ante review of these settlements because approval decisions hinge on quintessentially executive judgments about litigation strategy and the government’s position. The judiciary should, however, continue to enforce settlements. And Congress can always override policymaking settlements through its legislative authority. But unilateral executive decisionmaking can and should be checked through traditional administrative law mechanisms of judicial review. Whenever DOJ consents to relief in a way that makes policy, the resulting policy should be subject to the same judicial scrutiny as though it were formulated outside the shadow of the settlement. Moving judicial review to the ex post position ensures that private parties and the government retain the principal benefits of settlements without sacrificing transparent, reasoned decisionmaking.

This Note analyzes policymaking settlements in which the executive branch surrenders a legal right to pursue certain policies or actions specified by an agreement to resolve litigation.11 By limiting the breadth of future decisionmaking capacity, the executive branch can effectively constrain the range of permissible future policies even in the absence of new legislation or a judicial decision. For example, the Flores settlement thwarted both the Obama Administration’s and the Trump Administration’s efforts to impose certain policies related to detaining immigrants.12 But unlike traditional forms of executive-branch policymaking, such as notice-and-comment regulation, few process safeguards exist for settlement agreements.13 As a result, these types of settlements might be conceptualized as an undertheorized form of “unorthodox policymaking,”14 where the executive branch binds itself and others (including future administrations) through litigation.

As in the Flores case, these settlements are most likely to arise in litigation against federal agencies by private parties seeking some kind of policy development or implementation. These might be statutory or constitutional challenges, and may arise under the Administrative Procedure Act’s (APA) waiver of sovereign immunity.15 Empirical evaluation of the quantity, type, and partisan valence of settlements is nearly impossible without a catalogue of policymaking settlements from DOJ, which has proven challenging to obtain.16 This Note therefore focuses on the legal concerns raised by these settlements.

The argument proceeds in five parts. Part I describes existing doctrine and DOJ policy with respect to settlements. Part II analyzes shortcomings in prior critiques of the settlement power and reframes the separation-of-powers debate in terms of executive control over litigation strategy. Part III explains how ex post judicial review of settlements can recognize the legitimacy of executive control over litigation while preserving the values of transparency, reasoned decisionmaking, and judicial review. Part IV offers other, constitutional alternatives for Congress to exercise its authority to limit or direct the settlement authority. Finally, Part V offers a good-governance perspective on the limits of DOJ’s institutional role and competencies. In light of these governance critiques, the Note concludes with guiding principles for use of the settlement authority.