The Yale Law Journal

October 2014

Self-Help and the Separation of Powers

abstract. Self-help doctrines pervade the law. They regulate a legal subject’s attempts to cure or prevent a perceived wrong by her own action, rather than through a mediated process. In their most acute form, these doctrines allow subjects to take what international lawyers call countermeasures—measures that would be forbidden if not pursued for redressive ends. Countermeasures are inescapable and invaluable. They are also deeply concerning, prone to error and abuse and to escalating cycles of vengeance. Disciplining countermeasures becomes a central challenge for any legal regime that recognizes them.

How does American constitutional law meet this challenge? This Article contends that a robust set of unwritten, quasi-legal norms shapes and constrains retaliation as well as cooperation across the U.S. government, and it explores how these conventions of self-help correspond to regulatory principles that have emerged in public international law. Re-envisioning intragovernmental conflict through the lens of self-help gives us new descriptive and critical purchase on the separation of powers. By attending to the theory and practice of constitutional countermeasures, the Article tries to show, we can advance familiar debates over legislative obstruction and presidential adventurism, and we can develop richer models of constitutional contestation within and beyond the branches.

author. Associate Professor, Columbia Law School. For especially helpful exchanges, I thank Akhil Amar, Will Baude, Jessica Bulman-Pozen, Josh Chafetz, Adam Cox, Richard Epstein, David Fontana, Michael Gerhardt, Bernard Harcourt, Judge Brett Kavanaugh, Jeremy Kessler, Harold Hongju Koh, David Law, Daryl Levinson, Peter Margulies, Bill Marshall, Richard McAdams, Tom Merrill, Gillian Metzger, Henry Monaghan, Eric Posner, Jeff Powell, Joseph Raz, Daphna Renan, Anthea Roberts, Kim Roosevelt, Steve Sachs, Adam Samaha, Fred Schauer, Reema Shah, Jamelle Sharpe, Chris Sprigman, Peter Strauss, Adrian Vermeule, Justin Wert, Ryan Williams, and David Zionts, as well as workshop participants at Chicago, Colorado, Columbia, Hastings, NYU, Temple, and Washington University in St. Louis law schools. For financial support, I thank the William S. Friedman Faculty Research Fund.


Sometimes people are allowed to take the law into their own hands. They may bypass the courts and the cops and take unilateral measures to cure or prevent misconduct in their midst. Across the United States, doctrines have been developed to regulate such “self-help” behavior in criminal justice, property, contracts, torts, and other areas of private law. In public international law, a whole subfield is devoted to the self-help issue.

And so one might wonder: when may a U.S. government institution “attempt to redress a perceived wrong” by another U.S. government institution through its “own action,” rather than through a third-party process?1 More specifically, when may officials in one branch of the federal government attempt to redress another branch’s perceived wrong through means that, but for that wrongdoing, would be impermissible?2

The question goes to the core of the separation of powers; both the separateness and the balance of powers among the branches depend upon its answer. Rising levels of partisanship lend it new urgency. Yet the question never seems to get asked, at least not in these terms. American lawyers have not developed a framework for analyzing or administering self-help remedies in constitutional law. Nor have they given much attention to the unwritten practices that shape interbranch struggle more generally. The result, this Article explains, has been an imbalanced discourse around constitutional conflict and constraint—an obsession with the Constitution’s formal allocation of authorities, and relative neglect of the informal norms that determine how those authorities are wielded and disputes about them settled.

The issues are abiding, but recent events help to underscore the stakes. Consider three legislative-executive clashes that have generated heated debate:

1.      In January 2012, President Obama made recess appointments to top posts at the Consumer Financial Protection Bureau and the National Labor Relations Board.3 The Senate had been holding regular pro forma sessions during its holiday break in order to foreclose this option, but the Justice Department’s Office of Legal Counsel (OLC) opined that the appointments were lawful under “a practical construction” of the Recess Appointments Clause.4

2.     In June 2012, the Department of Homeland Security announced a policy to stop deportation of, and give work authorization to, some one million undocumented immigrants who came to the country as children.5 This “Dreamers policy” mirrors the DREAM Act that successive Congresses had failed to pass. The immigration statutes do not exempt this class of individuals from deportation (hence the push for the DREAM Act), but the Justice Department defended the policy as within the scope of delegated prosecutorial discretion.6

3.      Over the past several years, the Department of Education has granted more than forty states the “flexibility” to pursue educational reform plans that do not comply with central requirements of the No Child Left Behind Act of 2001 (NCLB).7 The Administration acknowledges that these waivers have effectively rewritten the regulatory framework devised by Congress, but it insists that they are justified by “unintended consequences” of that framework as well as by explicit statutory language.8

On one prevalent view, the common thread linking these cases is the disdain they show for constitutional boundaries. The President determines to pursue a legally dubious course of action; he finds executive branch lawyers who will bless his preferred approach; and he forges ahead, heedless of the limits that Congress has placed on him. The episodes, accordingly, “suggest that this president lacks a proper respect for constitutional checks and balances.”9 Abstracting from particulars, they reveal a deep continuity between the Obama Administration and its predecessors in the contingent, instrumental approach taken to the law when important political objectives are at stake.

This narrative is powerful. But like the Administration’s official legal analyses, it reflects a dichotomous approach to parsing presidential initiative—as either wholly consistent or inconsistent with separation-of-powers principles, supportive or corrosive of checks and balances—that obscures the dynamism and complexity of interbranch conflict. It also misses some of the most interesting features of the current constitutional period. A more refined set of conceptual tools is needed.

Another reading of these cases is available, and it points toward a more nuanced perspective on the President’s relationship to law. On this alternative account, President Obama responded in measured terms to a profound breakdown of the policy process that had come to jeopardize the integrity of representative government. Congress was the constitutional villain. According to the Administration’s public narrative, Senate Republicans conceded the suitability of the recess appointees, yet they nevertheless stonewalled the nominations in an unprecedented effort to prevent disfavored agencies from exercising their statutory responsibilities.10 A majority of both houses voted for the DREAM Act, yet a runaway filibuster nevertheless doomed its passage in the Senate.11 Congressional leaders from both parties supported overhauling NCLB, yet a “dysfunctional” legislative environment made that goal unattainable.12

Underlying these outcomes were the familiar drivers of today’s pathological politics: the exceptional levels of internal coherence and ideological purity within the Republican and Democratic camps, the elevation of partisanship above institutional or geographic identity as the defining ethic, the relentless minoritarian blocking tactics, the permanent campaign. President Obama was faced with a Congress whose top Senate Republican had announced numerous times that “[t]he single most important thing we want to achieve is for President Obama to be a one-term president,”13 and whose top House Republicans reportedly gathered on the night of his inauguration to devise a plan to “mortally wound” him through “united and unyielding opposition.”14 This sort of maximalist obstructionism, the President has repeatedly suggested, is incompatible with the traditions of interbranch practice and the assumptions of separation-of-powers theory.15 It works a serious wrong on the American people and our scheme of governance. At the extreme, it triggers a limited right of reprisal.

On a sympathetic reading, then, President Obama’s maneuvers can be seen as a species of constitutional self-help—attempts to remedy another party’s prior wrong rather than to ignore inconvenient legal barriers.16 His actions were meant to be preservative, not usurpative,in nature. The key to unlocking this understanding is the observation that President Obama’s denunciations of Congress consisted of more than policy critique. President Obama accused Congress of contravening not only the electorate’s political preferences but also basic constitutional conventions, unwritten quasi-legal norms that allow the branches to function. Take this reading far enough, and the President’s “We Can’t Wait” mantra17 comes to look less like a populist repudiation of legal limits (as in, “We can’t wait for the statutes to say what I want them to say!”)18 and more like an expression of intent to redeem the constitutional order (as in, “We can’t wait for Congress to start policing itself and stop destroying our government!”).

There are numerous difficulties with this view, both as a description of President Obama’s behavior and as a prescription for executive practice. Some unilateral measures have no remedial value. Some are illegal regardless. Nevertheless, this reconceptualization of recent events helps to illustrate this Article’s principal claim, which is that many of the most pointed ways in which Congress and the President challenge one another can plausibly and profitably be modeled as self-help rather than self-aggrandizement, as efforts to enforce constitutional settlements rather than to circumvent them. The claim is less radical than it might seem. Even if certain of these efforts ultimately prove inconsistent with the best reading of the Constitution—as some surely will—this does not mean that as a class they defy legal rationalization. Our understanding of such constitutional contestation, I will propose, can be clarified by an analogy to the law of self-help and above all to the international law doctrine of countermeasures.

This study has three main goals. The first is to introduce the concept of self-help into separation-of-powers analysis, with special reference to executive power. Although the concept of self-help plays a critical role in numerous bodies of law, it barely figures as such in constitutional doctrine or scholarship. Part I defines interbranch self-help and identifies it as a significant feature of our constitutional design, as well as a plausible component of a regime committed to the rule of law. In developing these points, I aim to remain as agnostic as possible about the best abstract theory of constitutional interpretation or the separation of powers.19 The methodology in this Part and throughout is to take constitutional law as it is currently practiced, and to try to reconstruct aspects of its immanent normative structure.

The second goal is to connect the idea of constitutional self-help to the idea of constitutional conventions, and thereby to enable a richer and more realistic portrayal of both. These ideas can illuminate, for example, how and why legislative obstruction triggers executive branch reprisals. Against the backdrop of a constitutional text that assigns Congress hardly any affirmative responsibilities, constitutional conventions arguably impose on House and Senate members a much larger set of duties to the executive and to the polity more broadly. In the gridlocked, agonistic “Age of Dysfunction” that we now inhabit,20 Part II explains, questions about these duties have taken on additional prominence. Much of today’s most vexing political behavior challenges not the interpreted Constitution, but the unwritten norms that facilitate comity and cooperation in governance. Moreover, these developments highlight the instability of the line between the two.

The third and final goal is to draw on self-help law and theory to gain critical purchase on the U.S. government’s workings—to introduce an analytical framework that enhances our ability to interpret and assess constitutional conflict in general and the current political moment in particular. The basic dilemma posed by self-help is pervasive throughout public and private law, as Part III explains. Recognizing the distinctive features of other legal regimes should not stop constitutional lawyers from tapping their intellectual resources.

The leading constitutional law treatments of interbranch conflict instead abstract away from remedial practices and justifications, and turn away from legal modes of reasoning altogether. They envision the branches as engaging in “constitutional showdowns”21 or playing “constitutional hardball.”22 These accounts valuably situate intragovernmental disputes within the framework of welfare economics (showdowns) and American political development (hardball). As suggested by the bellicose sports analogies, however, these accounts have little to say about the remedial framework in which such disputes do or could take place. They are attentive to the potential for legislative-executive struggle to generate new legal and quasi-legal understandings, but not to the ways in which unwritten norms may bear on the struggle itself. The same goes for many of the broader studies that have navigated “the web of documents, practices, institutions, norms, and traditions that structure American government” and comprise what is known as the “small-c” constitution.23 By allowing us to interpret interbranch conflict in more law-like terms, a self-help perspective allows us to subject this conflict to closer theoretical and institutional scrutiny.

After Part III canvasses regulatory strategies that have been used to discipline self-help, Part IV considers how these strategies map onto interbranch relations. Particular attention is paid to the most developed and apposite body of self-help law, the international law of countermeasures. Applying its standards, we can see more clearly what it is about President Obama’s approach to congressional obstruction that is problematic. We can also begin to refine our models of constitutional change and to re-envision other domains of constitutional contestation, such as state-federal conflict and intrabranch conflict. Indeed, the lines between these domains are somewhat artificial; self-help behaviors across the political branches not infrequently respond to self-help behaviors within Congress, and vice versa.

Not only can the law of countermeasures serve as a benchmark for U.S. practice, Part IV explains, but its core prescriptions also align with the conventions that shape and constrain retaliation within the U.S. system. It is not fanciful, then, to imagine that Americans would draw on international law to fashion parallel principles for domestic self-help, as we have already groped our way to a reasonable approximation thereof. The project of elaborating and enforcing these principles will only become increasingly important if the Age of Dysfunction persists. There is ample reason to expect it will,24 which makes this an auspicious time to take up the subject of constitutional countermeasures.

I. constitutionally derived tools and types of interbranch self-help

A. Definitional Preliminaries

A study of interbranch self-help faces a threshold issue of definition. What counts? In ordinary language, “self-help” may refer generically to “providing for or helping oneself without dependence on others,” or more specifically to “the act or right of redressing or preventing wrongs by one’s own action . . . without recourse to legal process.”25 In legal discourse, the meaning of self-help is no less fluid. Some formulations require a unilateral “attempt to redress a perceived wrong,”26 while others extend to such speculative ex ante measures as locking the door of one’s car or walking home on a well-lit street.27 Another strain of the private law literature construes self-help more narrowly, as the option “to do something that would otherwise be legally actionable in order to prevent or cure a legal wrong.”28

The fundamental divide concerns whether the self-help label should be reserved for rights and remedies that may be exercised only to cure a wrong that has been or will imminently be done by another. Let us refer to these rights and remedies as conditional self-help powers, and to redressive tools that are not so limited as general self-help powers. Verbally criticizing someone who harms you is a general self-help measure. Physically restraining that person is a conditional one.29 As Richard Epstein has observed in a discussion of private law, the “question of self-help . . . becomes considerably more difficult” with respect to the conditional category.30 The commission of an act that constitutes a prima facie violation of the law, on the theory that it is justified in response to someone else’s violation, is inherently more fraught than the commission of an act that would be lawful regardless.

Translating these points to the separation-of-powers field, a broad definition of interbranch self-help might include any attempt to resolve another branch’s wrongdoing in lieu of or prior to third-party dispute resolution.31 Although this approach is appealingly simple and worth keeping in view, I will focus on conditional self-help as the category of greatest theoretical interest. As Part III explains, the distinctive promise and peril of the self-help concept lie in its capacity to transform what under normal circumstances is impermissible behavior—throwing a punch, breaching a contract, invading another’s land—into an acceptable method of enforcement.

Accordingly, I will emphasize an alternative definition of interbranch self-help as the unilateral attempt by a government actor to resolve a perceived wrong by another branch, and thereby to defend a perceived institutional prerogative, through means that are generally impermissible but that are assertedly permitted in context. This definition is not meant to do justificatory work. I say “perceived wrong” and “assertedly permitted” because these self-helpers, like all self-helpers, may misjudge the legality or propriety of the other party’s conduct or of their own response.32 Moreover, as we will see, these self-help efforts are problematic formally as well as substantively, insofar as they involve modes of legislative, executive, and judicial action that are typically thought to be the exclusive province of a different branch.

The extension of self-help to public law arguably raises additional complications. The “self” in interbranch self-help is an aggregate: the interests and entitlements at stake are organizational rather than personal in nature. There is nothing mysterious in this. Collective agents such as corporations routinely engage in activities described as self-help,33 and the anthropomorphization of the branches is a standard move in constitutional theory.34 I will consider in Part IV how the branches might be pulled apart,35 but it is useful and parsimonious to treat them as unitary actors for the purpose of introducing the basic framework.

Is there something nonetheless contrived about characterizing the use of state authority as self-help? The currency of the concept in public international law suggests not.36 It would indeed be strange to characterize, say, a police officer’s power to make arrests as a self-help tool, even if this power is designed to redress legal wrongs. Yet the reasons for this strangeness do not necessarily carry over to other government contexts. Police officers do not make arrests to defend their institution against the encroachments of another institution with which they are in privity or competition; the notion of a self being helped is especially obscure. Making arrests, furthermore, is one of a police officer’s core duties; the sense of a deviation from ordinary roles or ordinary law enforcement is entirely absent. The government behaviors highlighted here, in contrast, involve extraordinary exercises of public power by and among coordinate actors.

In short, the idea of interbranch self-help is plausible. But how is it actually carried out in the United States? Before drilling down further into the idea itself, I will provide some context by reviewing ways in which the branches seek to remedy each other’s constitutional transgressions. Parts I.B, I.C, and I.D catalog tools derived from the written Constitution that may be deployed for conditional self-help, general self-help, or both.37 Part I.E introduces additional refinements to the concept of interbranch self-help, setting the stage for Part II’s consideration of its role in the murkier yet critically important realm of conventions.

B. Congressional Tools

Congress’s remedial toolkit is particularly extensive. When individual members or committees believe that the President has committed a constitutional wrong, they may seek rectification by denouncing her actions under the protective umbrella of the Speech or Debate Clause,38 blocking or delaying her nominations in the Senate,39 conducting hearings and investigations,40 or introducing restrictive legislation and condemnatory resolutions.41 When one or more houses agrees, Congress may pass those bills and resolutions (subject to potential presidential veto);42 hold executive officers in contempt;43 defund particular agencies, programs, and positions;44 and, in extreme cases, exercise the impeachment power.45

Congress’s remedial options vis-à-vis the courts are similarly broad, at least in theory. The Article III guarantees of life tenure and fixed compensation protect judges from only the crudest forms of reprisal.46 When working majorities within Congress believe the federal judiciary has acted wrongfully, they may deny salary increases and cost-of-living adjustments;47 withhold funding for administrative and institutional supports;48 expand or contract the courts’ jurisdiction;49 increase or decrease their membership;50 revise statutes to repudiate interpretations; and, in certain circumstances, exercise the impeachment power51 or resist the full implications of a ruling. After the Supreme Court invalidated a state criminal conviction for flag-burning on First Amendment grounds in Texas v. Johnson,52 Congress responded by passing a comparable flag-burning statute and “invit[ing] [the Court] to reconsider.”53 After the Court invalidated “legislative vetoes” in INS v. Chadha,54 Congress kept using them—and executive branch officials largely, if begrudgingly, kept complying—on an informal basis.55

All of the aforementioned tools are made available to Congress by constitutional text and structure, as conventionally understood. Modern commentators may debate their legality at the margins. For instance, many would question Congress’s authority to demand certain categories of executive branch information relating to core presidential deliberations, or to strip certain categories of federal court jurisdiction relating to core judicial functions. But as a general matter, few doubt that these tools are legitimate means for Senators and Representatives to try to enforce their vision of the Constitution.

C. Presidential Tools

The President has several undisputed remedial tools of her own. She may use the power of the pulpit to criticize Congress and the courts56 and to pressure them into adopting reforms.57 She may use the veto to quash legislation—and the threat thereof to induce Congress to revise or abandon disfavored bills.58 She may grant pardons or reprieves to individuals convicted under laws or circumstances believed to be unconstitutional.59

Beyond rhetoric, vetoes, pardons, and reprieves, presidential efforts to resist the other branches quickly become more controversial. Constitutional lawyers vigorously debate whether and under what conditions executive branch officials may decline to comply with federal statutes, subpoenas, and judicial rulings or decline to defend federal laws in court. History furnishes prominent examples of each mode of self-help. President Thomas Jefferson refused to enforce the Sedition Act, leading to his canonization in First Amendment lore.60 President Andrew Johnson refused to enforce the Tenure of Office Act, leading to his impeachment.61 Having unilaterally suspended habeas corpus, President Abraham Lincoln ignored Chief Justice Taney’s order granting habeas relief in Ex parte Merryman.62 President Franklin Roosevelt declined to defend the constitutionality of a law barring salaries for officials thought to be communists,63 and prepared to defy a Supreme Court ruling (which never came) that would have vitiated his efforts to forestall a wave of bankruptcies.64 Numerous high-level officials have disobeyed congressional subpoenas under a claim of executive privilege; when those officials subsequently received a criminal contempt citation, the Justice Department refrained from prosecuting.65

Recent years have witnessed a surge of interest in presidential action contrary to Congress’s will: the “lowest ebb” category of cases identified by Justice Jackson’s Youngstown concurrence.66 While modern Presidents may occasionally depart from the Justices’ constitutional reasoning,67 they never flout Supreme Court orders directed at them, and few lawyers question this obeisance.68 Yet even as they have ceded constitutional ground to the Court, Presidents have been forcefully asserting their interpretive authority against that of Congress.69 Vetoes and pardons are just the lower bound. As set forth in a series of OLC opinions, executive branch legal policy now incorporates a general, perpetual claim of right to defy congressional enactments believed to be unconstitutional, including an “enhanced responsibility to resist unconstitutional provisions that encroach upon the constitutional powers of the Presidency.”70 “The Constitution demands of the executive and legislative branches alike an ethic of institutional responsibility,” OLC proclaims in its most ambitious statement on the separation of powers, according to which each branch must actively and continuously protect its role “in the overall constitutional structure.”71

This view has been operationalized in numerous high-profile cases. Within the last decade, President Bush drew significant attention to nonenforcement questions in claiming a preclusive power to disregard the Foreign Intelligence Surveillance Act, the federal anti-torture statute, and other laws seen as impinging on his national security prerogatives.72 President Obama drew significant attention to nondefense questions by continuing to enforce a central section of the Defense of Marriage Act (DOMA) at the same time that he sought its judicial invalidation.73

Presidents also seek redress in subtler ways, without conceding that their actions are at odds with any legislative enactment. Whether or not they conceive of their handiwork in these terms, executive branch lawyers routinely deploy interpretive strategies designed to neutralize congressional conduct. They may invoke the canon of constitutional avoidance to bypass the most natural reading of legal texts that Congress has drafted.74 Citing resource constraints or case-specific factors, they may invoke principles of enforcement discretion to similar effect.75 And they may try, more generally, to dissolve apparent inconsistencies between the demands of legality and their moral or practical assessments of Congress’s behavior, by incorporating moral or practical considerations into their very conception of law.76 Of particular note, the turn to functional and custom-based modes of legal reasoning—ubiquitous in the separation-of-powers field—allows criticisms of “unwarranted” or “unprecedented” congressional behavior to be leveraged in support of assertions of presidential power.77 President Theodore Roosevelt, for example, relied largely on functional considerations in developing the view that the Senate has a duty to approve treaties except in rare situations—and hence that in light of the Senate’s failure to provide its prompt advice and consent on an important treaty he had negotiated with the Dominican Republic, he was justified in executing the agreement through a modus vivendi.78

D. Judicial Tools

In comparison to its congressional and presidential incarnations, judicial self-help plays only a modest role in our constitutional system. Through the practice of judicial review, the federal courts, of course, periodically determine that legislative and executive actions lack constitutional authority and therefore refuse to give them effect. And from the perspective of the courts, these judgments might be seen as attempts to resolve a perceived wrong by another branch through means that would otherwise be impermissible, albeit in a process that is initiated by a litigant rather than by the court itself. But it would be quite odd to view all these episodes as instances of self-help, the practice of which is identified in contradistinction to judicial dispute resolution.79 Adjudication is the paradigmatic “nonself” option—the remedial mechanism set up precisely to dispense impartial justice rather than to defend one party’s prerogatives—that self-help bypasses. To conceive of interbranch self-help in a manner that swallows judicial review would be to ignore what makes self-help a distinctive mode of enforcement, if not to render the concept all but meaningless.

This is not to say that judicial self-help is an oxymoron. The federal courts may respond to congressional or presidential wrongdoing through less conventional, less disinterested means that plausibly merit the self-help label. Judges might be seen as engaging in self-help, for instance, when they hold government officials in contempt, or when they criticize the political branches to some ameliorative end in extra-adjudicatory speeches, articles, or congressional testimony.80 It might further be seen as self-help when judges figuratively depart from their assigned role of deciding cases and controversies81 and level similar criticisms in dicta; or when they bend standard principles of adjudication or interpretation (including justiciability and deference doctrines) in an effort to resist external threats to their own authority.82 On this understanding, judicial review exercised to consolidate the very power of judicial review, from Marbury v. Madison83to Boumediene v. Bush,84 might be assessed in self-help terms.

Yet while there may be an intriguing debate to be had about the contours of judicial self-help, it is likely to remain a rarefied debate so long as we limit ourselves to irregular or judge-initiated practices and exclude the bulk of judicial review. This Article will focus on the forms of interbranch self-help that are most readily identifiable as such: those that begin, and often end, outside the courts.

E. Additional Distinctions

When convinced that another branch has gone too far, the preceding sections explained, U.S. government actors possess a variety of tools with which to try to restore the constitutional status quo ante. Now that we have a panoptic view, we are in a position to revisit the concept of interbranch self-help and the general/conditional divide highlighted above.85 Doing so further clarifies why certain maneuvers tend to arouse special concern.

In contemporary practice, each of the political branches enjoys a mix of general and conditional self-help powers. The President’s enumerated powers are understood to be general ones. She may veto any bill, pardon any convict, or criticize any congressional act, for any reason or for no reason at all, without running afoul of the written Constitution.86 In contrast, the President may not decline to enforce a duly enacted law at least unless and until she has determined that enforcing it would be unconstitutional; in the absence of such a determination, nonenforcement would manifestly violate her obligation to “take Care that the Laws be faithfully executed.”87 The only possibly permissible use of nonenforcement is as a mode of self-help, to cure or prevent congressional wrongdoing in the form of a constitutionally defective instruction.88 The same goes for the nondefense of statutes, noncompliance with subpoenas, and use of the constitutional avoidance canon.89

Legislative self-help tends to be less legally constrained, with the important exceptions of impeachment and contempt. Congressional bodies have a number of generally available means with which to resist presidential abuses, including oversight and legislation. But they may not impeach an executive officer or hold her in contempt unless she has committed certain serious offenses90 or obstructed the legislative process.91

As these examples suggest, the conditional self-helper frequently undertakes a type of action associated with a different type of legal actor.92 The separation of powers becomes less separate. When the President declines to enforce a statute or to apply the most natural reading of it, she may effectively refashion or repeal law in a legislative manner. When members of the House and Senate pursue an impeachment, they exercise traditional forms of executive and judicial power in charging and trying individuals, respectively. When judges inveigh against politicians outside of the adjudicatory process, they themselves act like politicians.

This observation about conditional self-help may generalize beyond the branches. Throughout private law, the iconic forms of legalized self-help tend to involve behaviors, such as violence and trespass, that are thought to be the exclusive province of state officials. The transgressiveness of conditional self-help lies not only in the flouting of generally applicable legal norms, but also in the crossing of generally applicable roles and boundaries.

Some “departmentalists,” it should be noted, take the view that the President has no lawful option other than to refuse to enforce measures she deems unconstitutional (and irredeemable through the avoidance canon).93 From this perspective, it may seem odd to characterize nonenforcement as a conditional self-help power, when it is at bottom a constitutional duty. There is some terminological tension but no contradiction here. Legal powers to exercise a certain special form of authority can also entail duties. Just think of jury duty, or a fiduciary’s responsibility to manage another’s property, or the various ways in which police officers are charged with using force as part of their job. It may not be accurate on the departmentalist account to conceptualize nonenforcement as a self-help privilege,94 because of the President’s lack of discretion, but it remains possible to conceptualize nonenforcement as a self-help power.

While conditional self-help powers are especially concerning as a matter of law, they are more transparent as a social practice. The contingent nature of their legality generates information about their users’ beliefs and motivations. Because presidential noncompliance with duly enacted laws may be permissible if and only if those laws are constitutionally defective, the very act of noncompliance, if acknowledged, reveals the President’s underlying critique. Furthermore, because our political culture prizes government legality, at least as a rhetorical matter,95 such contingently lawful acts are typically accompanied by public explanations. General self-help powers, on the other hand, may conceal their etiology. Congress has near-plenary authority under the Constitution to fund or defund any program run by the Department of Agriculture. If Congress stops paying for a program, the move might represent an effort to redress a perceived constitutional wrong, or it might represent nothing more than a change in policy priorities. Or it might be a hybrid, inasmuch as different members have different motivations for supporting the legislation. The fact of defunding does not itself tell us whether any sort of self-help has occurred.

Beyond the general/conditional divide, other distinctions bear on the legibility and legality of interbranch self-help. Some self-help tools are clearly conferred by constitutional text or rooted in Founding-era practices, whereas others have a more ambiguous constitutional basis. No reader of the Constitution, for instance, could deny the President’s power to veto bills presented to her, but serious scholars can and do deny the existence of a presidential power to decline to execute any (or nearly any) duly enacted law.96 In general, Congress’s self-help powers are on more solid originalist footing than the President’s. Most of Congress’s powers are enumerated, and the two main ones that are not—contempt and oversight—have deep common law roots and are widely seen as implicit in the Framers’ design.97

Some acts of interbranch self-help respond to encroachments on the self-helper’s own powers, whereas others target more diffuse wrongs to the constitutional system or a segment of society. When President Obama asserted executive privilege over deliberative documents relating to Operation Fast and Furious, he emphasized his desire to avert “significant, damaging consequences” to the executive branch.98 When President Obama declined to defend part of DOMA, he emphasized his desire to vindicate equal protection values.99 We might call the former type of act direct self-help and the latter proxy self-help. Most of the political branches’ remedial tools may be used in either fashion. The only tools that necessarily entail direct self-help are the contempt power and executive privilege. Under black-letter law, Congress may not issue contempt citations and the President may not assert executive privilege except to remedy threats to each branch’s own processes or prerogatives.100

Proxy self-help is recognized in numerous legal contexts.101 I may use otherwise forbidden force to prevent imminent bodily harm to the person standing next to me.102 The United States may use otherwise forbidden force to defend its allies against armed attack.103 The immediate beneficiary of proxy self-help is not the self that does the helping, but none of the prevailing definitions of self-help require otherwise.104 Moreover, the very distinction between the helper’s and the helpee’s interests—and therefore between direct and proxy self-help—is problematic in the separation-of-powers context, given that the branches are engaged in repeat-play relationships with each other and are steered by officials who have taken an oath to support the Constitution.105 Even if President Obama’s DOMA decision was made for the benefit of same-sex couples, it may have also advanced a more general principle of presidential interpretive discretion and reduced the executive’s complicity in a disfavored scheme. Even when a President acts to prevent direct harm to the presidency, she may simultaneously claim that she is defending the entire Constitution or the popular will that underwrites her office.

Finally, some forms of interbranch retaliation may be subject to eventual court review, whereas others can be expected to evade judicial resolution. Under current doctrine, for instance, it seems very likely that the Senate’s failure to provide advice and consent on a presidential nomination,106 or its actions in trying the impeachment of a federal officer, raise no justiciable controversy.107 For its part, presidential nonenforcement of statutes does not always insulate them from judicial review and may in some cases even facilitate it, as when the law’s alleged defect is the manner in which it limits executive power.108 In other cases, however, nonenforcement will preclude adversity and so deny courts the opportunity to resolve the constitutional question. The nondefense option does not carry a similar consequence (at least not at the district court level)109 and for that reason is seen by many commentators as less threatening to rule-of-law values.110

A table may help to distill these various distinctions. It is the set of conditional powers, once again, that is of greatest interest and that tracks the narrower definition of interbranch self-help advanced above.111

Table 1.

taxonomy of constitutional self-help powers112


II. self-help and constitutional conventions

Defined in conditional terms as the unilateral attempt to redress another branch’s wrongdoing through otherwise impermissible means, interbranch self-help thus plays a prominent role in our constitutional system. Defined to include generally permissible behaviors, it plays a pervasive role. These observations prompt a number of normative as well as conceptual concerns. As Part III will explain, conditional self-help opens the door to error and abuse by interested actors, instrumentalization of the law, and escalating cycles of vengeance. Disciplining conditional self-help becomes a central challenge for any legal regime that recognizes it.

The difficulty goes still deeper in the separation-of-powers context. Interbranch self-help attaches not only to legal rules grounded in the Constitution’s text but also to a broader set of emergent, quasi-legal norms that organize the workings of government. These latter norms are sometimes referred to as constitutional conventions.113 To grasp the full scope and significance of inter-branch self-help, it is necessary to integrate conventions into the analysis. 114

This Part explores the intersections among constitutional conventions, interbranch self-help, and the separation of powers. After laying some general groundwork, I will explain how the recent rise in partisan conflict underscores the importance of both conventions and the self-help that emerges when conventions come under strain. Appreciating these dynamics, I will further suggest, can enrich understandings of constitutional construction and small-c constitutionalism generally.

A. Conventions in Constitutional Theory

On the standard view, constitutional conventions (1) are norms of domestic governmental behavior (2) that emerge from decentralized processes, (3) are regularly followed (4) out of a sense of obligation, and (5) are not directly enforceable in court but rather (6) are enforced by political sanctions, if not also by “the internalized sanctions of conscience.”115 Intragovernmental self-help and elections, not judicial review, are the institutional mechanisms to curb violations. Most commentators appear to regard these six conditions as jointly necessary and sufficient, although the literature is rarely clear on the point.116 Of particular note, there is a broad consensus that political practices lacking any regulative or prescriptive dimension—say, the habit of holding certain meetings on one day of the week versus another—fail to satisfy condition (4). Whether out of sincere normative conviction or a more instrumental concern to avoid backlash, actors not only must follow a convention but also must believe that it ought to be followed; or at least, they must believe that others in the system so believe.

Constitutional conventions are often analogized to the rules of the game. They are rules that distribute responsibilities and facilitate cooperation among “the major organs and officers of government.”117 (This might be considered a seventh necessary condition.118) They regulate the “machinery of government.”119 In so doing, they supplement the organic laws’ allocation of powers and duties by determining how officials “should actually apply” the discretion they have been delegated.120 Conventions are often further described as unwritten. Although they may be memorialized in writing, nothing critical turns on this, as conventions generally owe their existence, maintenance, and influence to practical dynamics and evolutionary forces. The fact of memorialization may turn out to be otiose.

Depending on one’s conception of politics and law, constitutional conventions might be considered a special class of political norms (that give rise to extrajudicial sanctions when breached) or a special class of legal norms (that do not give rise to judicial sanctions).121 Or they might be thought to sit somewhere in between the two.122 On the one hand, constitutional conventions resemble legal norms in their regularity, normativity, and disciplining effects on public actors. Like laws, conventions, where they obtain, are typically thought to provide content-independent reasons for compliance.123 On the other hand, conventions differ from most domestic legal norms in their autonomy from authoritative texts, interpreters, and enforcers. Like principles of political morality, no designated actor can create or vindicate them.

Conventions, moreover, may vary in clarity, scope, and strength, both relative to each other and over time.124 The practice of Presidents’ limiting their tenure to two terms was, for many decades, seen as “central to the maintenance of the U.S. constitutional project.”125 By the end of President Grover Cleveland’s second, nonconsecutive term, questions about the meaning and wisdom of this practice began to arise, culminating in its apparent override by President Franklin Roosevelt (who won a third consecutive term in 1940) and subsequent codification in the Twenty-Second Amendment.126 The convention of a two-term presidency helped coordinate the highest levels of U.S. politics for over a century; underwent a period of contestation and deterioration from the late 1800s to 1940; and ultimately emerged, stronger than ever, as formal law.

Constitutional conventions are primarily associated with the “unwritten constitutions” of Commonwealth countries, following the pioneering work of A.V. Dicey,127 but analogous ideas have played a recurring role in American constitutional theory. In the late nineteenth and early twentieth centuries, a number of prominent U.S. authors probed “the real constitution” that had developed organically to supplement (or replace) the canonical document.128 A generation ago, commentators such as Jesse Choper, Louis Henkin, and Louis Pollak criticized government practices as “anticonstitutional in tradition, if not unconstitutional in law,” in terms evoking the law/convention distinction.129 Within the last decade, a small group of public law scholars, including Adrian Vermeule and Keith Whittington, has begun to mine the Commonwealth literature on conventions.130 As Whittington has observed, American constitutional conventions might be viewed as a mode of constitutional construction. Whereas constitutional interpretation, for Whittington, seeks to recover the semantic content of the Constitution’s text, constitutional construction “seeks to identify how constitutional meaning and practices are developed in the interstices of the constitutional text, where discoverable meaning has run out.”131 Conventions help to organize public life in what Lawrence Solum calls “the construction zone,” or the vast domain in which the text underdetermines outcomes.132 They help to shape a normative order in which representative politics is transacted. When they are violated, they trigger responses and counter-responses that ultimately stabilize or destabilize that order.

Understood this way, it quickly becomes apparent that conventions, and the intragovernmental self-help that backs them up, ought to be of interest to a much larger group of scholars. Some of the most ambitious projects in constitutional theory today explore the United States’ small-c constitution: the relatively stable set of rules, practices, and arrangements that are not housed in the constitutional text but nonetheless are thought to serve a constitutional function “because they are important to the structure of government or because they reflect fundamental American values.”133 While small-c theories come in many different stripes, they share a commitment to assimilating into constitutional analysis “all rules which directly or indirectly affect the distribution or the exercise of the sovereign power in the state”134—and therefore a commitment, at least in principle, to thinking systematically about constitutional conventions.

B. Separation-of-Powers Conventions

In the United States, many constitutional conventions serve to organize relations and promote cooperation among the coordinate branches.135 We might call these separation-of-powers conventions.136 A narrow view of this category would include only those conventions that impose linked obligations on actors from multiple branches. A broader view—which I will adopt—could also embrace those non-judicially-enforceable norms, such as Senate filibuster norms, that regulate behavior within a certain institution of government, yet in so doing substantially shape the way that institution interacts with another branch.

Some additional examples will help to crystallize the basic idea. When modern Presidents believe the Supreme Court has misconstrued the Constitution, they nonetheless acquiesce in its judgments;137 Presidents seek to control the courts, instead, through the appointment process.138 This is part of the convention of judicial supremacy over constitutional interpretation. When working majorities within Congress disapprove of the federal courts’ rulings, they too exercise restraint, hardly ever using their powers over court size, structure, administration, and procedure in a punitive fashion.139 “Court packing” is especially out of bounds. This is part of the convention of judicial independence. When Presidents are deciding on certain nominations, including district court judgeships and other local positions, they consult with the home-state Senators from their party.140 This is the convention of senatorial courtesy. When Presidents intend to make recess appointments, they inform Senators beforehand,141 and they do not install individuals in posts for which they have previously been voted down (not just filibustered) by the full Senate or the relevant committee.142 Senators, in turn, show substantial deference to the President’s selections for executive offices, particularly in cases of cabinet nominations and nominees who have served in Congress.143

The budget process incorporates a number of interlocking conventions. At a global level, Congress ensures that the other branches have sufficient funds to function. Budget negotiations may generate political brinksmanship, but major appropriations cutoffs are not threatened or used absent exceptional circumstances.144 Executive agencies treat as authoritative the spending instructions contained in the committee reports that accompany “lump-sum” appropriations and authorization bills, even though these reports lack the force of law.145 Agencies also typically seek informal approval from an appropriations committee before they move unobligated funds from one activity to another within a given account, even when they possess statutory discretion to do such “reprogramming.”146 The framework statutes that govern the various phases of the budget’s preparation have been characterized as “codifying norms and practices that first developed informally.”147

More diffusely, a slew of conventions regulate the flow of information between the political branches. Agencies are expected to communicate with designated congressional committees about notable developments and activities. Some agencies and their congressional overseers have developed protocols for sharing discrete types of information: for instance, the State Department periodically furnishes the House Foreign Affairs and Senate Foreign Relations Committees with a list of significant international agreements that have been cleared for negotiation.148 When a congressional body seeks sensitive documents from an executive branch component, both sides engage in an “accommodation” process of consultation and compromise before threatening to issue a subpoena or to assert a nondisclosure privilege, respectively.149 Executive branch officials who testify before Congress respond, in writing, to the “questions for the record” that they receive following the hearing.150 Within the first two months of each legislative session, the President comes to the Capitol to deliver a State of the Union Address that is pitched, at least superficially, in nonpartisan or bipartisan terms.151

On certain readings of the Constitution, some of these examples may seem better characterized, not as conventions, but as obligations that are implicit in the document’s design or otherwise legally binding. There is no escaping this ambiguity: the line between convention and law is destined to remain hazy under prevailing norms of constitutional argumentation.152 This is especially true in the separation-of-powers field, given the relative dearth of judicial enforcement153 and the plasticity of leading interpretive approaches. When in a functionalist mode, the Supreme Court asks (and, following the Court, government lawyers ask) whether a development “‘impermissibly undermine[s]’ the powers” of a branch154 or “‘disrupts the proper balance between the coordinate branches.’”155 Lacking any clear conception of “balance” to undergird it,156 this inquiry potentially legalizes all manner of interbranch grievances. When in a customary mode, the Court asks whether a practice has been “systematic, unbroken,” and “long pursued” with the knowledge and acquiescence of the relevant branches.157 Lacking any clear conception of what counts as historical practice or institutional acquiescence,158 this inquiry is no less slippery. “Ethical” appeals to the character of the American polity,159 although less common, may further obscure the line between Constitution and convention.

Moreover, even as executive and congressional officials operate in such an accommodating interpretive culture, they simultaneously inhabit a sociocultural environment that prizes notions such as tradition, prudence, and checks and balances. Ideals of cooperation and comity suffuse their sense of role morality. Beliefs about legal obligation, consequently, may be all the more apt to merge with beliefs about political obligation in the domain of interbranch interaction. The effort to distinguish “legally normative” constitutional conventions from “other normative” constitutional conventions may be not just conceptually fraught but psychologically naïve.160

Whatever the best way to characterize these practices, the examples above suggest that they are closely bound up with interbranch self-help. More specifically, self-help not only polices violations of convention but is also itself subject to conventional constraints. The two are partially co-constitutive. Senators who conclude that the President has failed to abide by the norm of senatorial courtesy may vote against the President’s nominations. The threat of retaliation sustains the norm.161 However, Senators who conclude that the Supreme Court has acted wrongfully do not (in this era) threaten to create more justiceships. The convention against court packing is, among other things, a limit on how legislative and executive self-help may be exercised.

As the examples further suggest, constitutional conventions do more than generate limits on discretion or entitlements to act in certain ways. They also impose some affirmative obligations.162 The President must ensure that the Senate is apprised of anticipated recess appointments and that congressional committees are kept informed about the agencies they oversee. Congress must do its part to ensure that those agencies are adequately staffed and supported. More nebulously, members of both branches seem to accept (if not always honor) the principle that they are to engage each other in a spirit of constitutional good faith, pursuant to an übernorm of comity and forbearance.

The imposition of affirmative obligations on Congress is especially notable, because the text of the Constitution is all but silent in this regard. The text sets forth numerous legislative powers and limits on those powers. But except for the largely inert Guarantee Clause,163 the Constitution’s plain language does not instruct Congress to pursue or accomplish any particular ends, much less to facilitate the presidency.164 The written Constitution gives Congress the option to do many things—regulate commerce, levy taxes, establish lower courts, and so on. The unwritten Constitution, in the form of constitutional conventions, gives Congress the responsibility to exercise some of those options, with due regard for the other branches.

C. Obstruction, Retaliation, and Construction

Most of the separation-of-powers conventions described in the preceding section remain as sturdy as ever. Other conventions centered in Congress, however, have unraveled to some extent in recent years. Our politics seems increasingly riven by partisan forms of constitutional hardball.165

The basic story is well known. In the Senate, filibusters were traditionally understood as an exceptional measure, “the tool of last resort.”166 By the early twenty-first century, they had become standard for many nominations and virtually all significant pieces of legislation.167 “Holds” on nominations and bills likewise appear to have become much more common and much longer in duration.168 Once used primarily for expressive purposes, the hold has become a de facto veto. Senators from both parties historically deferred to the President’s nominations for sub-cabinet executive offices and (to a somewhat lesser degree) judgeships. Opposition-party Senators are increasingly likely to contest these selections, even where the individual’s competence and integrity are conceded.169 Senators also used to vote on a larger share of nominations. Rather than thwart them through inaction, the Senate held up-or-down votes on candidates for senior executive branch positions who had made it out of committee.170 The deterioration of conventions transcends Senate procedure. For instance, it had long been taken for granted that the debt ceiling would be raised by statute as necessary, until House Republicans threatened in 2011 to force the government to default on its existing obligations unless President Obama acceded to future deficit reductions.171

These developments have led some to conclude that we have entered an Age of Dysfunction, in which gridlock has destroyed legislative capacity and “[t]he nation’s political system seems completely incapable of solving, or even grappling with, its most pressing problems.”172 This conclusion is disputable. The origins and effects of each development are complex, and congressional conflict and partisan rancor are not new phenomena. As described above, numerous separation-of-powers conventions appear to remain intact, while others—such as the norm against Presidents’ concealing the very fact of statutory noncompliance or the norm against partisan impeachments for non-felony offenses—are plausibly characterized as having been tested by recent challenges but ultimately reaffirmed.173 Senate Democrats have now revised the chamber’s rules to limit filibusters for executive branch and lower-court judicial nominations.174 The convention against holding the debt ceiling hostage appears to have been restored.175 Current levels of gridlock might not be so dire in light of historic baselines or popular preferences.176 We may be in a temporary moment of transition.177

What is not much in dispute, however, is that the Senate has become increasingly polarized, individualistic, and time-constrained; that these dynamics have exacerbated collective action problems in that body and in Congress more generally; and that the resulting forms of obstructionism strike many observers as a threat to the efficacy and integrity of the political process. There is, in short, a widespread fear that the breakdown of certain separation-of-powers conventions is contributing to a breakdown of our system of representative government.

Enter self-help. The President has not sat idly by as these conventions have deteriorated and elements of Congress have “engage[d] in strategies of obstruction and confrontation that well-socialized politicians might not have attempted in the recent past.”178 To the contrary, he has made increasingly bold unilateral moves. From the outset, President Obama determined to follow his predecessors in refusing to enforce appropriations riders that would have limited his diplomatic flexibility.179 He has asserted executive privilege over deliberative documents sought by the House Committee on Oversight and Government Reform.180 After some hesitation, he invoked the canon of constitutional avoidance to preserve181—and then exercise182—an option to bypass Guantánamo detainee transfer restrictions.

Above all, though, President Obama has employed self-help tactics that are not predicated on a showing of a large-C constitutional violation by Congress.183 Faced with a Republican Party that “has, at every conceivable juncture, frustrated the [Affordable Care Act’s] implementation” and that would undoubtedly block new legislation to clarify or fortify its terms, his Administration has used “myriad delays, waivers, and creative reinterpretations” to salvage the statute.184 Beyond health care, numerous executive agencies have used statutory waiver provisions to effect “nearly wholesale administrative revision” of major regulatory initiatives in fields ranging from the budget to education to welfare policy.185 Other agencies have “pooled” their powers “to do things they could not otherwise do” in the absence of new legislation.186 In the area of foreign policy, the Administration has increasingly bypassed Congress through “stealth multilateralism,” pursuing nonbinding international agreements that do not need legislative approval and participating in international institutions tied to treaties that the Senate will not ratify.187 The Department of Homeland Security has overseen an “extraordinary” mass release of immigration detainees in advance of sequestration—ostensibly to save money but perhaps also to counter House Republicans’ efforts to trim the federal budget and thwart comprehensive immigration reform.188 As noted at the Article’s outset, President Obama has also relied on highly contestable interpretations of the Recess Appointments Clause and the immigration laws in installing blocked executive branch nominees and implementing parts of his domestic policy agenda.189 The President’s entire “We Can’t Wait” campaign190 can be seen as an advertisement for executive self-help in response to a Congress that, according to Obama, will not “do its job.”191

The general pattern transcends this President and these examples. Obstructionism does not simply paralyze politics in a system of separated powers. It also generates its own correctives, through interbranch (and intrabranch) self-help.192 Aggrieved officials cease to follow ordinary norms of cooperation and constraint. Whether or not it is lamentable, this dynamic is perfectly predictable once we attend to the tools and incentives of the actors within each branch, as well as the role of constitutional conventions. As Jon Elster notes, if a violation of a putative convention “causes no reaction,” then the convention “never existed in the first place.”193

President Obama’s congressional antagonists do not necessarily concede that they have violated any operative conventions or otherwise triggered his conditional self-help powers. Indeed, taking a longer view, one might characterize the rise of obstructionist tactics within Congress as an emergent form of legislative self-help, in response to an antecedent rise in executive power that itself threatens the interbranch balance. There is no value-neutral baseline from which to assess competing charges of constitutional aggrandizement or abdication.194 There is no value-neutral baseline from which to assess self-help claims more generally: the validity of a conditional self-help measure always depends upon an evaluation of the other party’s conduct as well as the measure itself, and the two sides may hold very different views as to the former. They may disagree about the underlying law or facts, or about the relevant time frame for considering the facts.195 A self-help lens may be able to facilitate positive and normative analysis, but it cannot in itself justify President Obama’s—or anyone else’s—actions.

Congressional Republicans have, in turn, greeted some of President Obama’s self-help measures with retaliatory efforts of their own. To list just a few such efforts, members of both houses have made countless speeches assailing the President for overstepping the bounds of his constitutional role.196 The Senate Foreign Relations Committee reported out a resolution repudiating the President’s position that U.S. armed forces deployed to Libya were not engaged in “hostilities” under the War Powers Resolution.197 Senator Ted Cruz and others have employed a battery of unorthodox procedural maneuvers in a campaign to defund “Obamacare.”198 And the House Financial Services Committee refused to accept testimony from Richard Cordray on the ground that his recess appointment to be Director of the Consumer Financial Protection Bureau was constitutionally invalid.199 Congress may be hampered in its use of certain self-help tools by the same collective action problems that contribute to gridlock in the first instance,200 or by an ethic of court-centrism that contributes to a preference for judicial dispute resolution.201 But legislators who are not from the President’s political party have been actively deploying those tools that do not require bipartisan support.

As recent events underscore, while self-help may be a conservative practice inasmuch as it seeks to reestablish some prior equilibrium, it can also be an engine of legal and political creativity. Government actors may reconceive of their traditional habits, authorities, and obligations in light of their counterparts’ perceived transgressions. In 2007, Senate Democrats pioneered the use of pro forma sessions to block recess appointments during intrasession breaks; the move was necessary, the Senate Majority Leader said, to get the nominations process “back on track.”202 In 2010, President Obama’s Solicitor General appeared to concede the constitutional efficacy of this tactic.203 Less than two years later, OLC took a contrary position in the case of Cordray, based on a “practical construction” of the Recess Appointments Clause and its role in the separation of powers.204

The details here may be surprising, but OLC’s bottom line is not. As explained above, under prevailing norms of constitutional argumentation in the separation-of-powers area, allegations that another branch has engaged in unprecedented or destructive behavior can almost always be mobilized to support a claim of enhanced discretion.205 And so President Obama’s defenders could insist that his recess appointments struck “a badly needed blow for checks and balances,” as the Constitution “is not blind to the threat of Congress’s extending its internal squabbles into a general paralysis of the entire body politic.”206 Through self-help, politicians may advance novel constitutional propositions or practices without conceding any break with the past. They innovate under the mantle of restoration.

The creativity unleashed by self-help can be more radical still. Matthew Stephenson has recently argued that, in response to “[e]xcessive Senate obstructionism,” the President ought to construe the Senate’s failure to act on certain executive branch nominations as implied consent.207 Under current understandings of constitutional law, this argument is outlandish.208 Yet if Senate obstructionism persists and comes to be seen by many as an urgent threat, the argument may gain traction.209 Few issues are ever definitively settled in the “massively iterated—indeed, endless”—game that is the separation of powers.210 I noted above that constitutional conventions can be understood as a mechanism or modality of constitutional construction.211 Interbranch self-help offers an additional, overlapping tool of construction. Even as the prospect of retaliation serves to entrench certain constitutional norms, the actual exercise of self-help may disrupt others.

If this characterization of American constitutional practice is accurate, or substantially so, then there is cause for concern. It is not at all obvious a priori that interbranch self-help will turn out well. Perhaps such self-help will tend to under-protect customary norms that embed the accumulated wisdom of many actors and ages, serve a valuable settlement function, or reflect important tenets of political morality. Perhaps it will generate avulsive tit-for-tat behavior. Perhaps it will systematically advantage the President over Congress. Perhaps it will breed more, not fewer, violations of law and convention. In the absence of second-order norms to govern how the branches may retaliate against one another, self-help may be just as likely to unsettle as to secure rule-of-law values.

Does the United States have any such second-order norms? Could they really curb institutional aggrandizement and tame constitutional conflict? To understand how self-help may be aligned with systemic goals, it is useful to turn to legal regimes that have engaged these sorts of questions more deeply.

III. the unity of self-help: second-party enforcement in law and theory

"In many social contexts, self-help, when rendered promptly and in proper amounts, is one of the most indispensable and effective methods of social control."

—Robert C. Ellickson212

"Self-help may well be the first step toward anarchy."

—Idaho Supreme Court Justice Stephen Bistline213

We have seen that in the separation-of-powers field, self-help plays a vital, inescapable, and at times deeply troubling role. The same could be said of many fields of law: the basic dilemma posed by self-help is widespread, if not universal. Institutional and doctrinal designers have struggled for centuries to harness the benefits of second-party enforcement while keeping anarchy and abuse at bay, to mediate “the longstanding tension between the imperatives of an established system of laws and the individual needs and desires to avoid and remedy injury as effectively and efficiently as possible.”214

This Part considers solutions that have emerged from this struggle. The goal is not to provide a satisfying account of all self-help law, but rather, and more modestly, to identify some common strategies and generalizable insights that might shed light on the interbranch context. I will focus on the international law doctrine of countermeasures, which allows states to take certain actions that would otherwise be unlawful in response to lawbreaking by another state. This doctrine offers not only the most similar case to the domestic separation of powers—it, too, regulates conflict among formally equal government institutions—but also the most reticulated body of rules on conditional self-help. In distilling these rules and connecting key principles to private law analogues, I hope primarily to facilitate analysis of U.S. constitutional practices. Much as Part II sought secondarily to advance the study of constitutional conventions, I hope also to contribute to the burgeoning literature on “self-help as a unified theme.”215

A. The Dilemma of Self-Help

Self-help would not pose such a knotty problem for legal designers if it did not yield valuable benefits. But it does, sometimes. When individual actors are allowed to take unilateral measures to remedy the wrongdoing of others, in advance or in lieu of a mediated process, it may serve to deter such wrongdoing from occurring in the first place, reduce administrative costs, promote autonomy- or sovereignty-related values, and facilitate speedier redress. Less obviously, decreased reliance on third-party dispute resolution might serve to facilitate the maintenance of cooperative relations, mitigate feelings of alienation from the law, or generate deeper internalization of first-order legal norms. From any number of nonconsequentialist as well as consequentialist perspectives, self-help holds theoretical appeal. It certainly holds pretheoretical, atavistic appeal. “Self-help of some kind has always been with us,”216 and so long as human (and institutional) nature remains fundamentally self-regarding and self-protective, it presumably always will.

Against these potential benefits, the normalization of self-help carries tremendous risks, especially in the case of conditional privileges to act in ways that otherwise would be unlawful. In taking it upon themselves to rectify the misdeeds of others, self-helpers effectively act as judges of their own cause. There is ample reason to worry that they will misconstrue the law along the way—not just, or even primarily, on account of bad faith, but on account of motivated cognition217 and reliance on congenial interpretive methods or theories of law.218 That is, self-help increases the likelihood not just of legally abusive remedial determinations but also of legally erroneous ones. In so doing, self-help may tend to exacerbate asymmetries of power between the parties, favoring “the strong over the weak.”219 Self-help can also generate negative spillovers, paradigmatically in the form of escalating cycles of recrimination, retaliation, and violence. These overlapping concerns, together with the potential shift in decisional authority away from the courts, are enough to menace most any conception of a well-ordered society. Like bounty hunters, self-help is a mechanism for enforcing law that contains the seeds of a greater lawlessness.

Importantly, this dilemma is not confined to legal regimes in which centralized enforcement is utterly lacking or radically defective. Quite the opposite. No regime depends entirely on third-party policing or adjudication to enforce its rules and administer remedies. To do so would be grossly inefficient, if not also incompatible with people’s basic desires to defend themselves and to get even. It is “a ubiquitous, if underappreciated, feature of all legal systems, ancient and modern,” that “heavy reliance” is placed “on the use of self-help to enforce legal commands” and to produce social control.220 Constitutional law is no different, as has already been shown.

Still, one might wonder whether other regimes could have much meaningful to say about constitutional self-help. In most areas of domestic law, the exercise of conditional self-help is subject to potential judicial oversight. If I take it upon myself to redress a perceived breach of my contract or property rights—say, by deducting damages from what I owe a seller221 or by trespassing to reclaim a personal chattel222—then a court may eventually be engaged to review the legality of my actions. In the separation-of-powers context, by contrast, the shadow of adjudication is fainter. Legal disputes are less apt to be justiciable.223 Courts play a comparatively minor role in supervising self-help, whereas mechanisms like elections and public opinion play a much larger role. Actual violence, moreover, is not such a pressing concern. It is a convention of interbranch practice, we might say, that feuding government officials refrain from physically attacking one another.224 (This is another example of how self-help not only polices violations of convention but is also itself partly conventional in nature.225)

These discrepancies counsel against any strong form of comparativism in theorizing constitutional self-help. They do not necessarily undermine more limited forms of intellectual arbitrage, however, such as the effort to extract organizing principles from different areas of law so as to consider how government practices stack up against those principles. And in any event, even if private law is placed to the side, there is another legal regime that more closely parallels the separation of powers with regard to managing conflict: public international law. It should go without saying that the comparison is inexact. Notably, nation-states may use military force against one another, which generates special pressure to regulate self-help ex ante. I will consider the implications of this and other points of disanalogy in Part IV.226

But the two systems share important features, such as the dual role of public institutions as the primary authors and addressees of law, the tension between these institutions’ formal equality in law and the potential for significant power imbalances, the prominence of customary norms as against codified rules, the broad scope for reciprocity and repeat play, and the modest scope of judicial review. Both systems are at the same time deeply dependent on, and vulnerable to, self-help by state actors.

This observation would not have surprised the Framers, who appear to have drawn on international balance-of-power theories in developing the domestic separation of powers.227 As recent scholarship has underscored, international law and constitutional law are beset by fundamentally similar problems of substantive uncertainty, democratic legitimacy, and—for lack of a centralized compliance authority standing above the state—enforceability against government officials.228 Their conceptual and practical connections on these axes are foundational, so much so that it makes sense to speak of “public law” as a domain that straddles the two.229

In pursuit of interbranch self-help insights, then, it seems appropriate to look not just beyond constitutional law but also beyond our borders. This is especially the case because international law, unlike constitutional law, has developed explicit, trans-substantive rules for regulating conditional self-help.

B. The International Law Solution: Countermeasures Doctrine

Among its so-called “secondary rules of responsibility,” public international law contains a host of remedial doctrines that regulate self-help by states above and beyond the “primary” norms that dictate what they may and may not do ab initio. The principle of states’ “untrammeled right to self-help” was a pillar of classical international law,230 along with stringent notions of national self-determination and control over a defined territory. International lawmakers in the post-World War II era have endeavored to retain ample space for horizontal (state-to-state) enforcement, while limiting its negative externalities, enhancing formal dispute resolution mechanisms, and preserving the primacy of the United Nations system.

Most famously, the U.N. Charter recognizes an “inherent right of individual or collective self-defence” for states subject to actual or imminent armed attack, “until the Security Council has taken measures necessary to maintain international peace and security.”231 The right of self-defense may permit forcible responses, or reprisals, which otherwise would be unlawful.232 International law also recognizes necessity as a basis for failing to fulfill certain obligations, in cases where that failure is “the only way a State can safeguard an essential interest threatened by a grave and imminent peril.”233 More mundanely, the international system allows a wide scope for retorsions: nonforcible acts of lawful retaliation, such as the limiting of diplomatic relations.234 Although some have suggested that retorsions are subject to principles of proportionality, necessity, or good faith, the mainstream view is that any such constraints are not legal but political in nature.235 Retorsions are an example of what I have been calling general self-help.236

Between the life-and-death imperatives of self-defense and the diplomatic rituals of retorsion sits the law of countermeasures. Countermeasures are nonviolent “measures that would otherwise be contrary to the international obligations of an injured State vis-à-vis the responsible State, if they were not taken by the former in response to an internationally wrongful act by the latter.”237 Which is to say, they are conditional self-help privileges.

Countermeasures have a long history in international law.238 By the late twentieth century, they had been accepted, in some form, by the International Court of Justice in cases such as Gabčíkovo-Nagymaros Project239 and by distinguished arbitral tribunals in cases such as the Air Service Agreement award.240 In the latter case, the tribunal approved the United States’ cancellation of Air France’s Paris-to-Los Angeles route, which in normal circumstances would have constituted a clear violation of the countries’ air service agreement, in response to France’s disruption of Pan America’s London-to-Paris route.241 Similar logic also plays a prominent role in instruments such as the Vienna Convention on the Law of Treaties, which provides that a material breach of a bilateral treaty entitles the injured party “to invoke the breach as a ground for terminating the treaty or suspending its operation in whole or in part.”242

The most important statement of the law of countermeasures now appears in the Articles on Responsibility of States for Internationally Wrongful Acts, issued by the International Law Commission (ILC) in 2001.243 Although they remain “soft law” (rather than a binding treaty) and have proven controversial in certain respects, these Articles are regarded by many as authoritative.244 They elaborate a general regime of legal responsibility to govern the rights and duties of states in situations where an international obligation has allegedly been breached. This regime is meant to be independent from, and neutral on, questions concerning primary norms, and to “apply to the whole field of the international obligations of States.”245 Instead of tailoring remedial rules by subject matter, the Articles “presume that international law is a unified body of law,” at least at a high level of abstraction.246 The Articles function as “residual,” or default, principles, which may be modified or superseded by more specialized rules in particular treaty regimes or customary domains.247 The General Agreement on Tariffs and Trade, for instance, has almost entirely displaced the Articles with its own regime of responsibility, in which a Dispute Settlement Body of the World Trade Organization (WTO) must authorize all countermeasures in advance.248 Even in this context, however, adjudicators have increasingly looked to the Articles to inform and confirm interpretations of WTO law.249

In developing the Articles on countermeasures, the ILC wrestled with the fundamental dilemma of self-help outlined above. Skeptical states emphasized countermeasures’ potential to generate abuse, destabilize international law, and exacerbate inequalities of power, providing “cover for cowboy diplomacy by rich, powerful states.”250 Against these concerns, it was noted that countermeasures serve a valuable enforcement function and that the articulation of limiting principles could provide a check on aggressive applications. The ILC ultimately chose to acknowledge that “countermeasures are justified under certain circumstances,” while subjecting them to explicit “conditions and limitations” to keep them “within generally acceptable bounds.”251

These conditions and limitations are stringent. Countermeasures may be directed only against states that are responsible for prior, internationally wrongful acts252: “A State taking countermeasures acts at its peril, if its view of the question of wrongfulness turns out not to be well founded.”253 Except in urgent cases of necessity, countermeasures must be preceded by notice, demand for cessation or reparation, and an offer of negotiation.254 They may not be taken if the responsible state’s wrongful act has ceased and a dispute settlement procedure is underway.255 They must be “commensurate with,” or proportionate to, “the injury suffered.”256 Although they are not required to be “reciprocal” (that is, identical or closely related in kind to the offending behavior of the responsible state), reciprocity increases the likelihood of proportionality.257 Countermeasures must also be terminated as soon as the responsible state has complied with its legal obligations,258 reversible in their effects as far as possible,259 and deployed “instrumental[ly],”260 for the purpose of securing compliance with the responsible state’s international obligations rather than exacting punishment.261 Finally, countermeasures may never disturb obligations arising under ongoing dispute settlement procedures, nor may they contravene principles of diplomatic and consular inviolability; fundamental human rights; the U.N. Charter’s restraints on the use of force; or peremptory norms such as the prohibitions on genocide, slavery, and torture.262

General international law, as reflected in the ILC’s Articles, has thus come to recognize the legitimacy of conditional self-help, but only when cabined by an array of substantive and procedural constraints. In many cases, no centralized body will be in a position to enforce these constraints. The lawfulness of countermeasures will be judged by the parties, expert bodies, and the international community, just as the lawfulness of the alleged initial breach will be.

C. Organizing Principles

Like all legal regimes that provide for second-party enforcement, the Articles on countermeasures can be understood, in schematic terms, as an effort to manage the dilemma of self-help. The overarching goal is to harness the benefits of self-help—in particular, enhanced deterrence and reduced administrative costs and delay—while minimizing the likelihood of negative spillovers, errors, and abuses of power and procedure.263 The Articles do not speak in cost-benefit terms, and any aspiration to “optimize” such a vast swath of public law may well be quixotic. But the Articles’ approach to countermeasures is sensible in light of these policy considerations.

The notice and demand requirements, for instance, may reduce error and abuse by slowing down the pace of self-help, publicizing the injured state’s legal views and intentions, and affording alleged wrongdoers a grace period in which to change or explain their conduct. The limitations on the use of countermeasures during the pendency of a dispute settlement procedure channel states away from self-help when the administrative costs of third-party resolution are relatively low. The prohibitions on use of force, human rights violations, and targeting of third parties work to contain the risk of violence and cycles of escalation; so do the requirements of proportionality, prompt termination, and proper motive (redress rather than retribution). And the mere existence of a doctrine that purports to regulate an area once thought to be beyond law provides a focal point for negotiation and criticism, enhances the reputational risk for states that use aggressive self-help tactics, and lends plausibility to charges of unjustified retaliation made by otherwise non-credible lawbreaking states. The law of countermeasures decreases the odds, however marginally, of a vicious regress whereby basic norms of fair play lose all purchase whenever the primary legal norms are believed to have been breached.

Although the point is not critical for this Article’s purposes, much of the domestic private law on self-help appears to share a broadly similar normative structure.264 A Vanderbilt Special Project on self-help observed in 1984 that the two “principal factors which indicate that self-help will be acceptable are that the available judicial remedies are somehow inadequate and the threat of a self-help remedy to society’s interests in law and order is minimal.”265 American legislatures and courts have, over time, significantly curtailed many of the conditional self-help privileges available at common law. To take just one prominent example, when seeking to repel invasions of one’s real property, owners today generally must limit themselves to reasonable, or “proportionate,” measures addressed to imminent threats, give advance warning before applying any force, and refrain from applying deadly force.266 These limitations parallel the requirements of proportionality, notice and demand, and imminence under general international law.

In line with the Vanderbilt Special Project’s observations, Richard Epstein has emphasized that “reasonableness” limitations permeate self-help doctrine and has suggested that they act as a brake on negative spillovers.267 Across substantive areas, Epstein further contends, self-help law “pairs a quick, cheap and reliable remedy with incomplete relief . . . which by definition and design does not leave the aggrieved party as well [off] as he would have been if the other party had faithfully performed its obligations in the first place.”268 Lawmakers may affirmatively deny the possibility of complete, state-enforced relief in situations where a party uses self-help. For instance, in some U.S. jurisdictions, a mortgagee who resells a foreclosed property is prohibited from subsequently seeking a deficiency judgment against the mortgagor;269 if the mortgagee wishes to exercise the self-help privilege of power-of-sale foreclosure, then she must forfeit a valuable remedial option. Extending Epstein’s insights, Adam Badawi has recently argued that in fields such as contracts and property, self-help law can “be understood as a mechanism to force investments in accuracy” by funneling mistake- and spillover-prone cases to state authorities and relaxing the rules of engagement where legal rights are clearer.270 Thus, efforts to recapture one’s chattels are treated much more leniently if the owner is in hot pursuit of an item taken from her without any claim of right.271

Synthesizing and simplifying these observations, we can provisionally identify a handful of generic strategies that have been developed to tame conditional self-help. Mutatis mutandis, these strategies recur throughout domestic and international law.

1.      Proportionality as a global norm. Virtually every self-help regime insists that conditionally lawful measures be proportionate to the other party’s alleged wrong. Alternatively phrased in terms of reasonableness, commensurateness, or the like, this is the basic substantive standard governing all self-help law.

2.     Notice and demand requirements. Virtually every self-help regime also insists on some form of notice of intent or demand for cessation in advance of conditionally lawful measures, with exceptions for cases of urgency. This deliberation- and publicity-generating mechanism is the basic procedural tool of self-help law.

3.      Incentivizing adjudication. In order to channel parties into formal dispute resolution in situations where the risks of self-help are seen as especially high (though not high enough to ban it outright), the relative cost of self-help may be raised. For instance, lawmakers may subject self-helpers to ex post discipline, permit courts to award extra-compensatory damages, or withhold certain benefits from parties who have previously pursued self-help.

4.     Categorical prohibitions. Self-help regimes frequently forbid certain inherently dangerous, costly, or irreversible measures, such as the use of deadly force or the targeting of third parties. These same measures may be especially likely to reflect a retributive purpose and especially unlikely to serve a restorative or corrective function.

5.     Time limitations. Self-help regimes frequently restrict the window of time in which conditionally lawful measures may be taken after the other party’s prior or impending breach.

Although I cannot begin to prove the point here, the discussion in this Part is sufficient, I think, to support the positive hypothesis that every mature system of self-help regulation relies on a combination of some or all of these strategies. Does constitutional law?

IV. coping with constitutional countermeasures

The answer to the question just posed is a qualified “yes.” In the balance of the Article, I will try to demonstrate that U.S. legislative-executive conflicts are regulated—already—by a rough set of constitutional conventions analogous to the international law doctrine of countermeasures.

Attending to these conventions of self-help, I will further argue, has the potential to yield a range of analytical and practical payoffs. It enables us to move beyond metaphors like hardball and showdowns in explaining how government actors register and manage disagreement—to return to the precincts of law an area that constitutional scholars have abandoned to game theory and grand strategy. In particular, a self-help perspective may help us to refine as well as to generalize criticism of President Obama’s approach to congressional obstruction; to rejuvenate the strain of separation-of-powers theory that emphasizes functionality alongside friction; to disambiguate constitutional law by making it more acceptable for officials to confess violations of first-order norms; and to develop richer models of constitutional conflict and constitutional change within and beyond the branches.

A. The Latent Doctrine of Constitutional Countermeasures

In their influential treatments of intragovernmental conflict, Mark Tushnet and coauthors Eric Posner and Adrian Vermeule appear to assume that outside of judicial review, “constitutional hardball” and “constitutional showdowns” occur in an essentially unregulated zone.272 The branches fight over their legal and quasi-legal authorities, but the fight itself is not governed by any higher-order normative principles.273 The gloves come off.

This assumption is plausible. As a null hypothesis, we might imagine that there are no identifiable unwritten, non-judicially-enforceable norms that constrain interbranch conflict generally or retaliation specifically. Rather, elected officials are constrained in these matters only by “politics” and realize that, if the public disapproves of how they act, they may pay a political price. The strident rhetoric of hardball and showdowns is entirely appropriate, because the currency of this realm is political power, the prevailing ethic one of antagonism and opportunism.

However, this assumption need not hold true as a logical matter—as we saw in the previous Part, it is perfectly possible to have “rules of responsibility” that fetter government disputants274—and indeed it does not appear to hold true in practice. Against the null hypothesis, I submit that there is in fact a set of identifiable, second-order norms that constrains legislative-executive conflict. The set includes expectations of proportionality, categorical prohibitions, notice and motive requirements, and the partial privileging of judicial review.

Taken individually, these conventions of self-help play an important role in structuring various sites of contestation, such as the nonenforcement of statutes. Taken together, they amount to a latent doctrine of constitutional countermeasures: an inchoate, imperfect version of the principles for regulating conditional self-help that one finds in international (and private) law,275 embedded within our separation-of-powers conventions. If I am right about this, then constitutional scholars may be well served not just by a new conceptual framework but also by a new linguistic register in which to consider interbranch conflict, one that sounds more in responsibility and restraint than in domination and violence.276 Even as the conventions of self-help flesh out the picture of constitutional contestation suggested by “hardball” and “showdowns,” they may simultaneously recast this picture in less menacing terms.

As in all self-help regimes, proportionality is the basic substantive standard governing irregular attempts to redress perceived wrongs by and within the political branches. In the public sphere, this can be seen most vividly in the way that charges of disproportionality are used to delegitimize acts of self-help.277 Thus, in response to efforts last summer to limit Senate obstructionism, defenders of the status quo countered that the plan to change filibuster rules without the usual two-thirds supermajority vote represented a “threat to blow the Senate up.”278 Previous plans to tighten the filibuster rules for judicial nominations were similarly tarred by opponents as the “nuclear option.”279 Shorn of its apocalyptic imagery, this rhetoric amounts to a claim that the remedial measures under consideration were excessive in light of the alleged wrong they were targeting.

Conversely, government officials who seek to exercise self-help take pains to demonstrate that their tactics are proportionate—and furthermore that the situation is urgent (“We can’t wait!”) and that their motives are appropriately corrective rather than retributive or exploitative. All recent Presidents, for instance, have declared that they would assert executive privilege over materials requested by a legislative body only “in the most compelling circumstances.”280 Senators pushing to reform the filibuster through a simple majority vote have repeatedly emphasized the “modest” nature of the changes they seek, in relation to the gravity of the problem.281 Statements such as these may be self-serving justifications, to be sure, but that in itself does not rob them of regulative force or of evidentiary value for determining the existence and content of self-help norms.282

In addition, the standard remedies associated with the branches’ conditional self-help powers283 are reciprocal, in the sense that they are closely bound to the motivating wrong, and hence proportionate in more concrete, formal terms.284 Under prevailing norms, the President does not decline to enforce or defend statute A,or to comply with subpoena B,because she believes that congressional behavior C is improper and wants to pressure Congress into ending C. The scope of the President’s main self-help powers is defined by, and extends no further than, the particular legislative measure deemed to raise constitutional problems. Impeachment and contempt, for their part, are openly predicated on discrete offending acts. According to Supreme Court dicta from 1821 that has hardened into an “accepted maxim,”285 congressional contempt sanctions must also be limited to “the least possible power adequate to the end proposed.”286

The pull of a proportionality constraint, along with the associated expectations of proper motive and prompt termination,287 has been evident at some of the most critical junctures in our constitutional development as well as in the ordinary course of governance. Congress’s determination in the wake of the Civil War to exclude Southern representatives, with ratification of the Fourteenth Amendment set as the price of readmission, was arguably the most radical instance of conditional self-help in the nation’s history. Yet extraordinary as this measure was, it was defended by its Republican sponsors not only as a legitimate consequence of war but also as the least draconian, least federalism-flouting means available—as a constitutionally conservative remedy compared to “alternatives that posited the destruction of the Southern states and their remission to territorial status, or that authorized permanent federal supervision of state political institutions under the Guarantee Clause.”288 The exclusion remedy was proportionate, its sponsors insisted.289 The constitutional wrong it was designed to rectify was just very, very grave.

General self-help powers do not generate the same intrinsic pressure for proportionality, given that their validity does not necessarily depend upon another’s wrong. Yet even in their case, we observe a bias against disproportionate and nonreciprocal remedies. It is all but unheard of for the President to veto a bill concerning issue D out of a desire to end congressional practice C, or for a congressional committee to investigate one matter in retaliation for perceived executive misconduct in another matter. The Advice and Consent power is now a significantoutlier in this regard, as it has become fairly common for Senators to hold or filibuster nominations or (less often) legislation for reasons that lack a direct connection to the nominee or the bill.290 These unusually attenuated maneuvers are subject to intense criticism on this basis,291 however, and the actors involved continue to abide by certain outer limits of proportionality. When Senator Richard Shelby tried in early 2010 to place a “blanket hold” on all of President Obama’s pending nominations, he was roundly denounced—even by Republican colleagues—for transgressing all such limits,292 and in short order he backed down.293

At a higher level of generality, an ethic of proportionality helps to maintain the boundary between the “large-C” Constitution and the “small-c” constitution. Recall that the former consists of norms that are grounded in the canonical document and viewed as legally binding, while the latter consists of norms that guide the processes of government but are not so pedigreed.294 Interbranch self-help polices both sets of norms, as discussed in Parts I and II. A convention of self-help polices the divide itself, by forbidding noncompliance with large-C norms in response to breaches of small-c norms. No President ever contends, for example, that lawful but awful behavior by Congress liberates her to treat a duly enacted statute as void. The remedy would be seen as out of line with the critique.

This convention can exert justificatory pressure in two directions: pressuring government actors (i) to justify their self-help measures as consistent with large-C norms, as President Obama has done in the cases listed in the Introduction, or (ii) to characterize the other branch’s alleged misdeeds as themselves violations of large-C norms. These pressures, in turn, help explain how arguments and practices move over time from “off-the-wall” to “on-the-wall,” through the efforts of entrepreneurial Presidents and legislators to leverage critiques of their rivals to reconfigure the boundaries of fair play and constitutional meaning.295 If specific cases of intragovernmental self-help can propel constitutional innovations,296 then conventions of self-help delimit their argumentative form. A meta-principle of proportionality not only reflects but also secures the basic normative structure of separation-of-powers law.

Table 2.

meta-proportionality in interbranch self-help


Beyond proportionality, a slew of conventions impose the equivalent of notice or demand requirements on executive self-help.297 As already noted, Presidents acknowledge an obligation to seek a negotiated solution with congressional overseers before asserting executive privilege over requested documents.298 The President is further expected to inform relevant members of Congress in advance of all intended recess appointments,299 international agreements,300 vetoes or constitutional signing statements,301 and determinations not to enforce or defend statutes.302 Whatever the scope of the President’s constitutional authority to unilaterally change the course of policy or violate the will of Congress, convention appears to dictate that such moves not be made in deep secrecy, without a prior or at least prompt explanation given to lawmakers.303 Some of the bipartisan outrage inspired by President George W. Bush’s defiance of laws like the Foreign Intelligence Surveillance Act traded on this second-order norm of how self-help is to be exercised, not necessarily on any substantive disagreement with the Administration’s constitutional claims.304

Other conventions moderate executive self-help by facilitating judicial review. For at least two decades, OLC has taken the position that the President should “base his decision to comply (or decline to comply)” with a statute he deems unconstitutional “in part on a desire to afford the Supreme Court an opportunity to review the constitutional judgment of the legislative branch.”305 This convention has resulted in numerous “enforcement-litigation gaps,”306 whereby the executive continues to enforce a law against third parties and then declines to defend the law when they challenge it in court. So strong is the underlying norm, according to Neal Devins and Saikrishna Prakash, that while executive branch officials “express fealty to the duties to enforce and defend, they often act as if they only have a duty to create a justiciable controversy.”307 Another line of OLC doctrine suggests greater scope for Presidents to decline to enforce laws that are indisputably invalid.308 This line of doctrine reinforces the norm against nonenforcement in the modal case of disputable constitutionality, and serves the important function of funneling the most mistake-prone cases into court.309

Still other conventions impose categorical constraints on interbranch self-help. Government self-helpers may not engage in physical violence, disregard judicial judgments or what is seen as clear constitutional text,310 or violate the rights of private parties. These norms have been so deeply internalized that they are almost never broached, much less breached. They go without saying. In this regard, it is important to keep in mind those self-help moves that officials do not make—the countless times that an aggrieved President does not, say, try to dissolve Congress, or install a new Supreme Court Justice by fiat, or for that matter urge Congress to increase the Court’s size.311 Conventions of self-help may have the most bite in rendering politically untenable, taking off the table, retaliatory measures that are never made or even contemplated.312 To focus only on observable self-help is to overstate the permissiveness of the system.

In sum, a close look at U.S. government practices suggests that numerous conventions regulate constitutional self-help, particularly the core case of conditional self-help; that these conventions echo many of the key principles used to discipline countermeasures in international law; and that their combined effect is to make intragovernmental conflict substantially more orderly and civil—more norm-bound—than is often assumed. The latent doctrine of constitutional countermeasures shapes and stabilizes the varied mechanisms of “checking and balancing.” Although I cannot conclusively establish these points in the space I have here, I believe the foregoing examples make out a strong prima facie case. If anything, I suspect this discussion has understated the breadth of the phenomenon. There may well be many more conventions of self-help in the U.S. system that deserve acknowledgement and attention.

B. The Dignity of Retaliation: On Taking Self-Help Seriously

By attending to conventions of self-help, then, we can begin to reinterpret intragovernmental conflict in a more sympathetic and nuanced manner, to resubmit the study of this conflict to the language of the law. Even so, the existence of these conventions may not seem all that impressive. The basic substantive standard, proportionality, is a famously fuzzy concept in international law.313 Perhaps identifying it as a constraint in this context is just a fancy way of saying that politicians do not wish to be seen as “going too far.” All of the above-mentioned norms may be susceptible to manipulation and violation, given that judicial enforcement does not back them up. Moreover, a great deal of political behavior, as well as constitutional law, likely rests at bottom on reciprocity, reputation, and repeat play;314 some of the conventions that contribute to the stability of our system may be epiphenomenal expressions of these underlying dynamics.

These are all reasons to develop, not to discount, the idea of constitutional self-help. Even if some of the conventions just described seem unremarkable at a retail level, the structural logic that connects them is compelling. It would be odd, from an institutional design perspective, to impose strict threshold conditions before certain self-help powers may take effect, but then to allow unknown or unbridled uses of those powers once triggered. The preceding section showed how the separation of powers “in action” avoids this mismatch between restricted right and unrestricted remedy, even though the separation of powers “in the books” by no means guarantees it.315 Only when we take the conventions of self-help as an integrated suite can we see this feature of our government practices.

Proportionality can indeed be a fancier way of saying that one may not go too far.316 But that is a momentous thing to say. The fact that American politicians defend their retaliatory actions as proportionate reflects and reinforces a consensus that intragovernmental conflict is not a free-for-all. It allows self-helpers to be judged and criticized even when their accusations against the other side are valid. It may also trigger the “civilizing force of hypocrisy”317—requiring politicians to be more measured in their self-help than they might otherwise prefer, if they wish to maintain consistency with their prior statements about the importance of the norm. The open manner in which proportionality is debated, both inside and outside the Beltway, facilitates the checking and legitimating forces of publicity and public opinion more generally.

Proportionality rhetoric is not just cheap talk. The cost to the speaker of trumpeting an ethic of restrained retaliation is greater than zero. As in international law, furthermore, the mere existence of proportionality talk may generate “compliance pull,”318 progressively enhancing the status or salience of the norm and thereby narrowing the range of retaliatory moves that are seen as acceptable.319 The more refined our discourse of proportionality—the better our ability to pinpoint what it requires in specific contexts—the more constraint may be exerted.

Just as some peripheral vagueness does not imply that conventions of self-help will be indiscernible, a lack of judicial enforcement does not imply that they will be inefficacious. These conventions are presumably followed (by and large) for many of the same reasons that all constitutional rules and conventions are followed (by and large), including equilibrium political and institutional forces; officials’ professional socialization and reputational concerns; and constraints associated with the party system, independent media, civil society, and reverence for checks and balances.320 Across the U.S. government, the breadth, frequency, and overlapping nature of interactions among the branches may broadly support norms of moderation by generating ongoing “multiplex” relationships that guarantee to each “a rich menu of future opportunities to render self-help sanctions.”321

This is an important point of potential contrast with the international system, as well as certain other national systems. Our domestic conventions of self-help are generally less fleshed out, less stringent, and less prominent than the international rules on countermeasures. But this makes sense, because we do not need to constrain retaliating branches as much as the international system needs to constrain retaliating nation-states. Spongier norms will generally serve in the constitutional context, given the various institutional structures, cultural characteristics, and ideals of solidarity that curb the worst abuses of self-help. In a fragile democracy, concerns about self-help conventions’ falling apart, about cycles of retribution leading to a collapse of law and order, have greater bite. The president who feels aggrieved might just go ahead and break up the parliament. At the international level, the mightier state might just go ahead and cut off all intercourse with its weaker rival. Many factors combine to make analogous moves largely unthinkable in the United States.

Of course, intragovernmental conflict does occur here, frequently sparked by one side’s perceived violation of a constitutional rule or convention. Yet if government actors are willing to violate those first-order norms, why wouldn’t they also be willing to violate second-order norms of permissible retaliation? Why wouldn’t the sociological conditions that produce a breakdown at the first level tend to produce a breakdown at the second, or third, or nth level? Whatever the answer, such vicious cycles appear to be the exception rather than the rule as a descriptive matter; the conventions of self-help appear to be highly resilient. This resilience might reflect the fact that government actors have particularly strong endogenous incentives to abide by norms of moderation and fair play, insofar as these norms “effectively bundle numerous prospective policy outcomes” or help preserve the larger separation-of-powers “game.”322 Or it might reflect the fact that even if government actors have no clue what the equilibrium strategy is in any given conflict, they may be aware that such norms have served their institution reasonably well in the past and therefore rely on them as decisionmaking heuristics.

Alternatively, conventions of self-help may exert an especially strong normative influence because of the sense in which they implicate basic fealty to the constitutional order or basic tenets of morality.323 The core principles of self-help doctrine—retaliate only in a proportionate manner, try to work things out first, spare innocent parties—are both easily comprehensible and consonant with ethical ideals such as the Golden Rule. It takes little constitutional sophistication to grasp the threat to good governance posed by a “nuclear option.” Plausible accusations about breaches of these conventions, accordingly, may tend to resonate with important nongovernmental audiences, in particular median and independent voters, in ways that bolster their constraining force on government actors.

In any case, norm-following always gives rise to an infinite regress concern. The more rules that one has to follow, however, the harder it is to violate them all. This is in the nature of failsafes. The very existence of an additional layer of norms applicable to intragovernmental conflict should tend to curb efforts to vanquish the other side.

The possibility that many of our separation-of-powers conventions have a game-theoretic underpinning of some sort—so that senatorial courtesy, for instance, persists because of the expectation of retaliation or reciprocity arising out of a repeated game324—hardly means that they should not be of interest to lawyers. At worst, this possibility means that we might wish to turn to disciplines like political science or philosophy to understand these conventions better. The same is true of every governmental norm. Game theory has little if anything to say about the substantive content of these constitutional arrangements. Nor does it have much to say about where government actors’ preferences come from, or why they would seek to safeguard the prerogatives of their office beyond their own tenure. What is distinctive about the field of interbranch conflict is not that it is amenable to formal modeling. What is distinctive is that we do not even have a legal discourse to which such reductions could be applied.

Even if one could demonstrate that a certain convention has its origins in certain equilibrium forces, it would be a fallacy to conclude that this undermines the convention’s normative character.325 And virtually all of the conventions of self-help outlined above, along with the vast majority of separation-of-powers conventions more broadly, appear to be at least partially “thick” in our constitutional culture: they have been internalized to some extent by officials and opinion-makers “as a rule of political morality.”326 Instrumentalism aside, executive branch self-helpers genuinely seem to believe that they should act proportionately and non-vindictively, give advance notice, support judicial review, not target third parties, and so forth. Or at least, they seem to recognize that others hold these beliefs, and therefore they purport to do likewise. Anyone who wishes to grasp the phenomenology or the practical morality of the separation of powers—not just its written rules and observable outputs but also the way it is experienced by government participants—has reason to care about the conventions of self-help.

By forsaking the language of law in favor of sports metaphors and social science, constitutional scholars have thus blunted their ability to assess constitutional practice from an internal as well as an external perspective.327 A self-help framework cannot tell us which side will or should prevail in any given interbranch conflict. But it nonetheless gives us a richer set of resources with which to investigate this conflict’s legal and normative character; it allows us to see retaliation as both a subject and a source of constitutional regulation. To the extent that countermeasures principles can promote stability and fairness in a domestic system as in the international system, a self-help framework may also provide government actors with a more satisfying set of reasons for accepting these principles as obligatory standards of conduct.

Moreover, self-help offers an especially apt lens through which to view our current constitutional period, with its twinned anxieties regarding legislative dysfunction and executive aggrandizement. As applied to government gridlock and efforts to circumvent that gridlock, a self-help framework has deep resonance with the Founding vision. The doctrine of constitutional countermeasures provides a potential bridge between the two major accounts of the Framers’ separation of powers: the view that their fundamental purpose in adopting this principle was to prevent tyranny, andthe view that their fundamental purpose was to create a government that is able to get things done.

It has become commonplace to observe, as Justice Brandeis once put it, that the aim of the separation of powers is “not to promote efficiency but to preclude the exercise of arbitrary government.”328 We have already seen how the conventions of self-help can serve the Brandeisian vision by disciplining interbranch struggle. Yet as many have argued, the Framers did not accept Justice Brandeis’s antinomy. Prominent among their reasons for embracing the separation of powers was the promotion of energetic and responsible governance in the common interest.329

This is the strain of separation-of-powers theory that best aligns with the use of otherwise impermissible measures in response to partisan obstructionism. (Whether it ultimately justifies these measures is another matter, inescapably dependent on one’s theory of constitutional legitimacy and the particulars of the case at hand.) On the efficiency account, there are not just discrete clauses of the Constitution that authorize discrete types of self-help. There is also a structural principle of workable government that may inform and support presidential adaptation in the face of an uncooperative Congress, and vice versa. Those who would defend such adaptation have no need to reconceive the Constitution for the Age of Dysfunction. We already have a prominent theory of the separation of powers that identifies policy paralysis as a constitutional evil—and hence a theory that might be mobilized to try to legitimate some of the most controversial forms of contemporary self-help, even as the theory of countermeasures might be mobilized to ensure the President’s goals are pursued in a law-like manner. Self-help doctrine can render less arbitrary the struggle for efficiency.

C. Why Can’t We Wait?: The Missing Discourse of Constitutional Self-Help

Seen in this light, the jarring thing about the American system of separated powers is not that there is so much retaliation or quasi-constitutional custom, but rather that there is so little acknowledgement of either. Government officials hardly ever invoke constitutional “self-help” or “conventions” as such. Yet if the descriptive claims advanced in this Part are valid (or largely so), then the absence of self-help and conventions from our constitutional discourse is not evidence of their insignificance so much as it is evidence of the poverty of the discourse. An appreciation of these concepts may allow us to cultivate more honest and productive dialogue around the separation of powers.

For instance, if we were to follow through on the idea of conventions as distinct from politics, then we could not just dismiss out of hand the view that the Senate has a meaningful obligation—albeit not a strictly legal one—to restore its historic practice of giving all senior executive branch nominees who have made it out of committee a floor vote.330 We would have to think much more carefully about possibilities for congressional misbehavior outside of passing unconstitutional statutes. When the President responds, we would likewise have to think much more carefully about possibilities for executive misbehavior above and beyond the written constitutional floor. Whatever the best view of conventions’ normative status, simply conceiving of the branches as conventional duty-bearers would enable more nuanced conversations about how and why the branches invite as well as engage in reprisals.

If officials and observers were self-consciously to adopt a theory of countermeasures, it could go substantially further toward disambiguating constitutional law. By helping to distinguish between more and less lawful remedial approaches, self-help theory can also help to distinguish between norms regarding the content of the branches’ generally applicable obligations and norms regarding how violations of those obligations are to be judged and corrected. Countermeasures principles, recall, are superimposed on a set of “primary” rules.331 Instead of always pretending that their actions are consistent with these rules, government actors in an explicit regime of constitutional countermeasures may be liberated to acknowledge facial inconsistency—and then defend their behavior as conditional self-help. Legal disagreement would be displaced to some extent from the first-order norms of constitutional conduct to the second-order norms of acceptable retaliation, creating greater rule-ness and clarity as to the former.

President Obama’s approach to congressional resistance illustrates the problem. Through his “We Can’t Wait” campaign,332 the President has been constructing an extended moral and consequentialist argument about how Congress has been acting in deeply irresponsible ways, ways that threaten constitutional values. Lacking an account of constitutional self-help on which to draw, however, he has been unable to link this argument to an appeal for enhanced discretion to pursue his objectives. His searing indictments of legislative obstructionism have remained legally inert.

In its lengthy opinion on Richard Cordray’s intrasession recess appointment, for example, OLC mentions “practical” considerations no fewer than eleven times, without once mentioning that Senate Republicans blocked a floor vote on Cordray in an arguably unprecedented manner, or that his recess appointment was an inherently time-bounded, reciprocal response.333 In its lengthy briefs filed in court in defense of the Dreamers policy, the Justice Department similarly declines to note the use of the filibuster in derailing the DREAM Act.334 Legislative gridlock goes unmentioned in the various documents that defend the NCLB waivers.335

The Administration’s official legal scripts have effaced the actual motivations and developments that seem to be driving some of its most controversial behaviors. In so doing, they have obscured these behaviors’ real sources of legitimation. The absence of a vocabulary of constitutional self-help has driven a wedge between the Administration’s professional and political discourses, degrading both.

A self-help framework might put President Obama’s campaign on a firmer footing.336 Barred from claiming that congressional Republicans’ violations of small-c constitutional norms entitle him to violate large-C norms because of the meta-proportionality principle of interbranch self-help,337 the President’s lawyers (like many executive branch lawyers before them) have resorted to subtler forms of interpretive self-help.338 The cases listed in the Introduction and Part II.C are prominent examples.339 The basic pattern that has emerged in these cases and others combines a latitudinarian reading of statutory and constitutional restrictions with a suggestion, typically made in a less formal setting, that Congress’s institutional pathologies have enhanced the President’s discretion as a matter of policy if not also of law.340 The implicit theory seems to be that when the lawmaking branch of government is broken, the law-executing branch must enjoy greater freedom to utilize the laws with which it is stuck. Dysfunction yields discretion.

No Administration official, however, has publicly articulated this theory, or any other self-help theory. No official has acknowledged the looseness of the President’s readings of legal texts and then endeavored to justify that looseness as an unconventional but not strictly unconstitutional response to the failure of congressional lawmaking.341 Although this is the self-help move to which the President’s lawyers seem most drawn, they have been unwilling to embrace it on these terms. The failure to acknowledge and defend conditional self-help as such is itself contrary to basic principles of self-help law.342

It may also be poor legal strategy. The one time that a top Administration lawyer hinted at the idea of interpretive self-help in court—when the Solicitor General suggested in the Noel Canning oral argument that the recess appointment power “may now act as a safety valve given [current levels of Senate] intransigence”—he was rebuked by Justice Ginsburg for departing from his written submissions, and he ultimately found no supporters.343 Perhaps this particular self-help plea was destined to fail before a Court predisposed to “institutional formalism,”344 but the Solicitor General had not laid the groundwork necessary to give it a chance.345

There are many other self-help moves that might be made. Consistent with the self-help principles sketched above, a less perilous alternative would be for the Administration to limit itself to reciprocal violations of convention—refusing to give advance notice of international agreements, reprogramming unobligated funds (as allowed by statute) without checking with appropriators, or the like. A more ambitious tactic would be to try to convert President Obama’s critique of legislative obstructionism into a large-C critique. In his push for filibuster reform, for instance, the President could have argued that the modern, routinized version of the filibuster had come to violate the structure of the written Constitution, and hence that Senate Democrats were not legally obligated to comply with the relevant procedural rules.346

Any given argument for conditional self-help may not succeed in the end, of course. The President’s lawyers would have a heavy burden to show why contemporary forms of obstructionism deserve to be seen as constitutional evils that justify an exceptional response, as under no plausible reading of the large-C or small-c constitution is the President entitled to have his preferred policies adopted by Congress. To the extent that courts play a reviewing role, they may not be persuaded by such arguments, as Noel Canning exemplifies. So too with the public, the press, and the professoriate.

If these efforts to vindicate executive self-help were to fail, however, at least they would do so with transparency and integrity—forthright about the contingent nature of their legitimacy, openly tied to and limited by the congressional misconduct that allegedly potentiates them. The language of self-help would both enrich the terms and clarify the stakes of the debate between President Obama and his congressional critics. A self-help framework offers the most principled basis for the President, and any successor who feels similarly stuck in an Age of Dysfunction, to promote a vision of the separation of powers that links constitutional responsibility to constitutional redress.

D. Anxieties and Extensions

Applying a self-help lens, this Part has suggested, we may find that some of what looks pathological or lawless in our constitutional order is in fact compatible with well-known regulatory models. This observation does not amount to a defense of existing practices. As a justificatory framework, self-help theory is inherently limited: the validity of any given act of retaliation always depends upon an assessment of the other side’s alleged wrong, and that assessment always requires a substantive theory of the relevant laws or conventions.347 While identifying behavior such as President Obama’s “We Can’t Wait” campaign as an attempt at self-help can guide both justification and critique, mere identification cannot resolve whether the behavior is warranted or desirable.

It bears emphasis, more generally, that even if the United States does have a latent doctrine of constitutional countermeasures that negotiates between efficiency and anti-tyranny principles, this negotiation is unlikely to yield ideal outcomes. Longstanding interbranch norms such as the ones highlighted in this Article may be workable and attractive to a Burkean traditionalist, yet suboptimal from any number of perspectives.348 On some readings of the Constitution, they may even be unlawful. Conditional self-help measures, as explained above, pose problems for formalists as well as functionalists insofar as they involve one branch’s exercise of powers typically associated with another branch.349 Those who believe that the Take Care Clause requires the President to enforce each and every statute, lest she usurp the legislative power to make law, will be dissatisfied with any scheme that countenances nonenforcement.350

It is also very possible that the actors within each branch will have mismatched incentives or capacities to act in the interests of their branch. Countermeasures regimes generally assume that participants will be motivated to police and retaliate against initial violations. This assumption is plausible in international law, where nation-states’ identical legal authorities may also make breaches easier to identify and reciprocity easier to attain. When it comes to the branches and especially to Congress, however, institutional self-interest will often be complicated by both partisan and personal incentives. As scholars such as Daryl Levinson and Richard Pildes have highlighted, the recent (re)emergence of “exceptionally strong and polarized parties” puts significant pressure on the Madisonian model of inherently and continuously rivalrous branches.351

Although party competition now suffuses numerous aspects of legislative-executive interaction, it does not undercut the utility of a self-help perspective. On the one hand, partisanship clearly cannot explain all of the retaliation that occurs within government.352 On the other hand, even when partisanship does play a driving role, officials still must channel many of their retaliatory behaviors through the Constitution’s institutional framework.353 Anyone interested in why the branches check and balance one another needs to consider partisanship; anyone interested in how they do so ought to consider self-help. Partisan dynamics interact in complex ways with the quality and quantity of interbranch self-help,354 and the remedial tools wielded by individual Senators ensure that some amount of it is kept in constant supply.355 But at least as a first approximation, it is plausible to worry that such self-help will occur too frequently or vigorously in times of divided government, or too infrequently or fecklessly in times of unified government (when the same political party controls Congress and the presidency).356

Beyond the contingencies of partisan politics, a Madisonian might worry further that the executive is structurally better organized than Congress for the consistent practice of self-help, for all the reasons that the executive is better organized to take decisive action and push legal boundaries generally. As a result of collective action problems and informational deficits as well as partisan dynamics, it is now widely believed that “Congress as a body does not systematically seek to protect its prerogatives against presidential encroachment.”357 One might consequently be concerned that even as individual members or committees of Congress will tend to over-provide certain forms of partisan self-help (for example, punitive oversight or routine use of holds by a Senator from the opposition party), Congress as a whole will tend to under-provide certain other forms of collective self-help (for example, legislation designed to redress presidential overreach), in ways that do not net out. Depending on one’s assumptions and priors, this concern might seem severe.358

In calling attention to the role of constitutional countermeasures, then, I do not mean to tell a celebratory story. There is no shortage of reasons to lament the status quo. What I do mean to affirm is that we go wrong when we read retaliatory interbranch activity as necessarily destabilizing of good governance or rule-of-law values. There is no a priori reason to assume that all such activity will undermine rather than shore up the separation of powers.359 As in international law, what we need are principles to distinguish the more destructive kinds of self-help from the more constructive ones. I have also tried to show that constitutional self-help can be valuable for the same reasons that self-help can be valuable in other contexts. Notwithstanding the risk of error and abuse, constitutional self-help holds out the promise of efficient deterrence, reduced transaction costs, and greater internalization of stability-inducing norms. Accordingly, the simple observation that our system contains large amounts of interbranch self-help, including conditional self-help, should not frighten us. It would be far more frightening if our system lacked robust second-party enforcement mechanisms to safeguard constitutional compliance.360

Against the potential theoretical and practical benefits of admitting self-help into our constitutional lexicon, what might be the costs of doing so? The most acute concern, I think, is that the very project of developing a framework for constitutional self-help will tend to facilitate opportunistic behavior—in particular, presidential power grabs. Would a theory of constitutional countermeasures only end up abetting aggrandizement?

This is the same perverse-consequences concern that the ILC faced in developing the Articles on State Responsibility.361 Countermeasures doctrine is meant to be a tool of limitation, not of authorization or exculpation. It layers “secondary rules” on top of preexisting “primary” norms. Yet simply by bringing such dangerous self-help behavior under a legal rubric, some argued, the ILC ran the risk of legitimating that behavior and thereby advantaging the most aggressive actors.362 It is notable in this regard that “countermeasures” and “conditional self-help” sound fairly technical and antiseptic. These labels might mask to some extent just how troubling governmental retaliation can be. It would sound rather different if, following Michel Foucault, we instead referred to this doctrine as an economy of “illegalities,” in which officials constantly negotiate relations of power by pushing the edges of law and trying to get away with it.363

This perversity objection, however, may be less forceful than it seems. It cannot be ruled out that adding another arrow to the quiver of arguments that might be deployed in defense of constitutionally questionable practices could lead to greater presidential adventurism. But the project of developing a vocabulary of constitutional self-help has no necessary pro-executive valence, any more than the project of developing a vocabulary of residential self-help has a pro-landlord valence.364 There are far too many contingencies in play to predict with confidence whether crisper rules in this area would ultimately generate more or less constraint. Moreover, and critically, all of the forms of self-help reviewed in this Article are already occurring, just on more opaque terms. And they are already occurring in a legal and political context that tends to be extremely favorable to activist Presidents.365 Against this backdrop, greater attention to interbranch self-help has nontrivial checking (as well as illuminating) potential. “[T]he elaboration of a balanced regime of countermeasures,” the ILC concluded, is “more likely to be of use in controlling excesses than [is] silence.”366

As the ILC did with the intergovernmental system, American lawyers and lawmakers can work to improve our system of intragovermental self-help to the extent that they find it structurally too President-friendly367 or otherwise misguided. Adapting Vincent Blasi’s “pathological perspective,” for instance, they might try to frame domestic countermeasures doctrine for the worst of times, by curtailing some of the most alarming moves that could be made by strong Presidents, even at the cost of simultaneously blocking some salutary applications.368 They could press Congress to refuse to confirm nominees who were previously given intrasession recess appointments. They could urge a categorical prohibition on use of the avoidance canon in executive branch interpretation, or a categorical obligation on the President to pair nonenforcement of statutes and subpoenas with the facilitation of judicial review.369 And they could ask courts to enforce or develop any or all of these principles (or, conversely, to stay even further away from legislative-executive conflict). More radically still, they could try to revamp the structure of constitutional conflict on the model of the WTO, so that a designated dispute resolution body authorizes countermeasures ex ante, in lieu of a decentralized regime of tit-for-tat sanctioning.370

These efforts might fail. The political will may not be there. Conventions cannot always be changed, much less fine-tuned, through critical argumentation, although it is sometimes possible.371Only by recognizing the existence of conventions of self-help, however, can we begin to see the regulative work that they collectively do, and so envision them as a potential site of reform.

Instead of shrinking from the idea of constitutional self-help, we should push the inquiry further. This Article has focused on legislative-executive interaction as the most prominent context for constitutional countermeasures. A self-help perspective potentially opens up a number of additional avenues for research and reconceptualization, to which I can only gesture here.

Intrabranch self-help. Most obviously, a self-help framework could be applied to interactions within as well as among the branches. As a practical matter, there is no clean line separating the two domains. On account of partisanship, conflict in Congress often stimulates and responds to legislative-executive conflict;372 some of this Article’s examples, such as the presidentially supported push for filibuster reform in the Senate, highlight the porousness of the inter- and intrabranch categories. Within the legislative process, the prevalence of informal norms is well established.373 What is far less clear, because the question has not been posed in these terms, is whether some of these norms specifically regulate efforts to correct violations of the written rules of procedure or the unwritten first-order norms of cooperation. To the extent that such self-help conventions exist, even the most bitterly partisan, legally unconstrained struggle in Congress may be more principled, and less pathological, than it first appears.

Within the executive branch, retaliatory measures are less likely to revolve around partisanship than around rivalries spanning agencies and their White House overseers, agencies and other agencies, civil servants and political appointees, and other bureaucratic divides. Constitutional scholarship is not the only field that could profit from focused inquiry into how executive branch officials set out to redress each other’s perceived wrongs through and beyond the law. If the study of interagency coordination is “all the rage in administrative law,”374 then the study of interagency conflict management presumably merits some enthusiasm as well. Is leaking confidential information to the press, for example, a recognized form of intra-executive self-help? And if so, does even this rogue behavior—a paradigm case of seeming lawlessness and venality in government—abide by norms of proportionality and reciprocity, categorical subject matter limitations, and the like?375

State-national self-help.A self-help framework lends itself to the study of vertical as well as horizontal contestation in government. As every student of American constitutional law knows, our history is replete with conflict between state and federal officials. The most spectacular examples of state self-help continue to haunt the constitutional imaginary: nullification and secession. The states’ position lost decisively in these cases, so that the large-C Constitution is now widely (though not universally) viewed as denying states the authority, on structural and Supremacy Clause grounds, to abrogate federal law or withdraw from the union in response to allegedly unconstitutional action by the national government.376 It took a civil war, and later the civil rights movement, to vindicate this principle. The basic self-help norm in the federalism sphere is a categorical bar on nullification and secession.

Yet unilateral efforts by state officials to remedy perceived wrongs by the national government have hardly gone away. They just tend to take subtler forms. As federal regulatory schemes have increasingly encroached upon traditional zones of state authority, states have found new ways to resist. Recent work by Jessica Bulman-Pozen and Heather Gerken suggests that states routinely “use regulatory power conferred by the federal government to tweak, challenge, and even dissent from federal law,”377 often along partisan lines.378 Some of these uncooperative efforts are perfectly lawful, as when states exploit regulatory gaps to challenge federal policy.379 Some of these efforts are more legally dubious, as when states go beyond what federal law requires in pursuing unlawfully present aliens, on the view that the federal executive branch has failed “to enforce the immigration laws as written.”380 We are beginning to appreciate that decentralized self-help practices are not marginal but pervasive in the modern administrative state.381

We do not yet have a detailed understanding of the nature and scope of many of these practices, however, or of the informal norms associated with them. We have not yet laid all the groundwork for reconciling state self-help with the triumph of federal constitutional supremacy. Important open questions include whether and to what extent state self-helpers abide by—or should be made to abide by—secondary rules of retaliation, such as an obligation to renounce certain especially worrisome tactics, defend their actions as proportionate, give advance notice of their intentions, or channel their dissent into a judicial forum. Self-help theory may give scholars and policymakers additional tools with which to identify and secure a cooperative core to uncooperative federalism.

Self-help and the secondary virtues. The value of a self-help framework potentially extends beyond law. By abandoning so much of the field of intragovernmental conflict to political pundits and social scientists, constitutional scholars may have constricted their capacity to see its ethical as well as legal dimensions. This Article has suggested that even the most rancorous confrontations in American politics generally retain an overarching, law-like integrity. They generally adhere to conventions of self-help that make them far more predictable, limited, and normatively laden—and far closer to the international law of countermeasures—than a purely partisan or game-theoretic account could hope to show. An additional cost of adopting a lawless theoretical perspective, then, may be an unwarranted invitation to constitutional cynicism and alienation. A self-help lens may be useful not only for reconceptualizing the separation of powers but also for reclaiming a sense of identification with it.

In his famous 1986 Harvard Law Review Foreword, Frank Michelman argued that the Supreme Court should aspire to model, in its own behaviors, ideals of practical reason and republican self-government that are unattainable on a national scale.382 In a similar spirit, we might want the political branches to model, in their interactions, an ideal of restrained, respectful, civic-minded conflict management. We might want them to display what the philosopher Alasdair MacIntyre once called the “secondary virtues,” such as “a pragmatic approach to problems, co-operativeness, fair-play, tolerance, [and] a gift for compromise”383—virtues that concern not which projects or goals are to be pursued, but rather howthey are to be pursued.

Here is the thing: Congress and the President already model this ideal. Their interactions already conform, imperfectly but substantially, to a multitude of unwritten norms facilitative of pragmatism, cooperation, and fair play. These norms have been tested in recent years, and there is no guarantee that they will not deteriorate further. Our political discourse may at times seem distant from MacIntyre’s standards. But the actual workings of the federal government—some of them, at least—have not been so terribly far off. A trace of the secondary virtues can still be glimpsed in the conventions of interbranch self-help.


The cases listed at the start of this Article are so vexing, not only because of the difficult issues they present on the merits, but also because of the dramatic manner in which the President resolved his own critique of Congress. We worry that brute politics may have triumphed over law. Yet unilateral reprisals of this sort do not necessarily entail lawlessness. To the contrary, they are one of our basic tools for securing both law and convention, both the “large-C” and the “small-c” constitutions. Accustomed to associating constitutional enforcement with the courts, legal scholarship currently lacks an account of these tools. It lacks a framework to parse or police the intricate terrain of intragovernmental self-help.

The development of such a framework, this Article has tried to show, can shed light on high-level questions pertaining to constitutional commitment, contestation, and change as well as on intensely practical concerns of governance. In an era marked by legislative obstructionism and “an executive style of governing that aims to sidestep Congress more often,”384 it becomes all the more vital to think systematically about the functions and limits of interbranch retaliation—and thus all the more appropriate to admit self-help concepts into the constitutional lexicon. The doctrine of countermeasures provides a model. Following the example set by their international law counterparts, American public law scholars may wish to set themselves the task of theorizing, and domesticating, constitutional countermeasures.

The place to begin is with the recognition thatintragovernmental self-help is all around us, for good and for ill. This Article is an attempt to identify, sort, and evaluate it as a unified phenomenon, a cohesive remedial logic at the heart of the American constitutional project. Above and beyond any of the specific arguments the Article has advanced, I hope the analysis can stimulate fruitful lines of inquiry—on constitutional self-help, constitutional conventions, the general theory and practice of self-help, and the intersections and interactions among them.