In Federalist 51, James Madison offered what has become the canonical account of how the separation of powers would pit branch against branch for the greater good. The officials of an institution would act on behalf of their institution for the Constitution to function properly. In Madison’s account, ensuring the presence of the right quantum of institutional loyalties would serve as a durable and plausible mechanism for enforcing institutional boundaries and ensuring a stable constitutional order. But modern scholars take a more skeptical view of his theory. This Article reconsiders the Madisonian concept of institutional loyalty as an object of analysis for constitutional scholars and jurists. Our core thesis is that institutional loyalty can be identified, evaluated, and even elicited through conscious and careful institutional design. We first provide a definition of institutional loyalty and situate the concept in the American constitutional past and present. We then marshal evidence that institutional loyalty may well have been decisive in some contemporary interbranch dynamics, even if its effects are inconstant and asymmetrical. In particular, we suggest that loyalties’ effects in the executive and judiciary are greater than their effect in the legislative context. We caution, however, that it is a mistake to view institutional loyalties as a constitutional end in themselves. Rather, institutional loyalty can promote or undermine structural constitutional goals, depending on the circumstances. Calibrating the appropriate mix of such loyalties across the branches therefore presents a considerable, if unavoidable, array of challenges. To that end, the Article offers a taxonomy of causal mechanisms by which institutional loyalty can be generated within each of the three branches. Working branch by branch, the Article identifies examples of institutional reforms capable of modifying institutional loyalty in ways that could promote widely shared constitutional ends.


The Constitution’s separation of powers implies the existence of three distinct and separate branches.1 Each was initially imagined to act “as a self-executing safeguard against the encroachment or aggrandizement of one branch at the expense of the other.”2 In a famous passage in Federalist 51, James Madison amplified this pivotal causal mechanism. “Ambition must be made to counteract ambition,” he explained, and “[t]he interest of the man must be connected with the constitutional rights of the place.”3 In this key passage, Madison invoked the institutional loyalty of officials—their tendency to identify with and to act in ways that promote their home institution—as a central dynamo of branch autonomy and healthy interbranch friction. Relying on these loyalties, he predicted that fractious interactions between branches fomented by this institutional loyalty would, in net, enhance individual liberty.4 At the same time, Madison recognized that voters would at times be driven by partisan, ideological, or even material “passions” that clouded their respect for these institutional boundaries.5 In these moments, he suggested, officials’ loyalty to their home institutions would shelter valued institutional norms against the fickle tides of popular sentiment.6

A recent wave of empirically informed and theoretically sophisticated scholarship has challenged the significance of this optimistic Madisonian equilibrium. This scholarship has powerfully questioned the Framers’ optimistic account of rivalrous branches led by zealous empire builders. Particularly now that our two political parties are ideologically homogeneous, the modern position contends, our national political-party system has “tied the power and political fortunes of government officials to issues and elections” and thereby fostered “a set of incentives that rendered these officials largely indifferent to the powers and interests of the branches per se.”7 They predict that officials will have an “array” of interests, but rarely will these interests “strongly correlate[ ] with increasing the scope or wealth of government institutions.”8 The result is that partisan and ideological loyalty often eclipse institutional loyalty both as a practical and as an analytic matter.9

The ascendancy of this important and insightful body of work means that the idea of distinctively institutional loyalties receives short shrift in the constitutional-law literature. Attention to the effects of intense partisan and ideological loyalties (which we do not deny) have crowded out descriptive questions of why institutional loyalties might persist and why they matter (whether for good or ill) when they do persist, as well as the normative question of how to generate appropriate institutional loyalties when they are desirable.10 It is against this backdrop that we aim to reevaluate institutional loyalty—the psychological cornerstone of a larger Madisonian political logic—as an object of sustained analysis in constitutional scholarship. In the service of that larger project, we advance here three points—one descriptive, one analytic, and finally a normative claim.

First, as a descriptive matter, we argue that the behavior of federal officials cannot always be explained simply by partisan or ideological motives. The current working of our constitutional system evinces the lingering influence of institutional loyalty of the kind Madison anticipated, particularly in the executive and judicial contexts. Officials may variously support an increase or a decrease in the power of their institution but often enough be motivated by a loyalty to the best interests of their institution. We do not claim, to be clear, that such loyalties are the most important or consequential element of our constitutional system. More modestly, we suggest that they persist to an extent that warrants more careful theorizing. Our aim here is not to measure their pervasiveness: it is to show that they operate at least occasionally in important policy consequence—and as such are worth identifying, defining, and exploring in terms of the institutional design of our constitutional system.

This descriptive claim can be illustrated with three examples, each drawn from a different branch, in which officials’ behavior is difficult to explain exclusively by partisan or ideological motivations.

First, faced with a politically polarizing challenge in a presidential election year to President Barack Obama’s signature healthcare legislation, Chief Justice John Roberts is alleged to have shifted his vote to support the legislation.11 Glossing his switch, journalist Jan Crawford observed that the chief justice “is keenly aware of his leadership role on the court, and he also is sensitive to how the court is perceived by the public.”12 Standard ideological or attitudinal models of judicial behavior do not offer a straightforward explanation of his vote, or his alleged shift.13 While secure conclusions are difficult to reach, it is at least plausible to think that concern about the Court as an institution figured large in the chief justice’s reasoning.

       Second, during the presidencies of George W. Bush and Obama, many lawyers serving as cabinet officials and senior political appointees resisted White House initiatives in favor of positions motivated by allegiance to their agencies’ or offices’ legalistic institutional agenda. Attorney General John Ashcroft and his deputy James Comey, for example, resisted White House pressure to authorize a surveillance program they believed ultra vires.14 Similarly, senior lawyers in the Office of Legal Counsel (OLC) and the Department of Defense resisted Obama’s 2011 military intervention in Libya on legalistic grounds, while State Department and White House lawyers defended it.15

       Despite the different administrations involved, these examples involve officials resisting ideological or partisan ambitions on legalistic grounds when their home agency or department has an interest in maintaining a legal constraint on the presidential agenda. Generalizing about these lawyers’ actions, former OLC lawyer Jack Goldsmith has explained, “A political appointee is a temporary steward in the institution in which she works, and is often moved to preserve the values and reputation of that institution,” even at the cost of compromising an administration’s immediate policy goals.16 We take Goldsmith as a credible source for the idea that officials believe themselves to be “stewards” of their institution.

       Third, over the course of the twentieth century, Congress has created a number of durable institutional structures that are not well explained in terms of the partisan or ideological interests of members. Foremost among these is the Legislative Reorganization Act of 194617 (LRA). This Act “vigorously reasserted congressional oversight power over the Executive Branch, and it remains the statutory basis for a great deal of contemporary oversight activity.”18 It reorganized the unwieldy congressional committee system, reducing the number of committees from forty-eight to nineteen in the House and from thirty-three to fifteen in the Senate.19 In addition, it defined committee jurisdictions in clear and systematic ways that allowed legislators to specialize in the oversight of specific elements of the executive.20 Congressional supervision of the administrative state of the kind familiar today would simply not exist without the LRA. The leading historical accounts of the LRA’s legislative passage emphasize that Congress intended to bolster its institutional capacity to act.21 More generally, Professor Eric Schickler’s study of every major institutional design change within Congress over the past century found “several major reforms,” including the LRA, were motivated in important part by “Congress-centered interests” that were distinct and different from partisan or personal careerist interests alone.22

The official action at stake in each of these three contexts was not only materially significant, but also hard to explain in purely partisan or ideological terms. Rather, in each case, a pivotal decisionmaker made a costly investment that advanced institutional interests in a way that was likely at odds with (or at least orthogonal to) the optimal pursuit of partisan or ideological goals.23 These actions are at least suggestive evidence of the continuing salience of institutional loyalty across the federal government (although it is quite telling that to find a legislative example, we are forced to retreat some seventy years back in time). They do not imply that institutional loyalty matters always, or even a majority of the time. Rather, the examples suggest that on some key policy questions, institutional loyalties can influence the shape and nature of federal action.

With this descriptive claim in hand, our second, analytic contribution is to define with precision the potential mechanisms through which “institutional loyalty” can operate. This analytic project has several elements. To begin with, we define “institutional loyalty” to mean an individual official actor’s psychological proclivity to perceive his or her proper course of behavior in terms of, or as incorporating, what he or she perceives to be the best interests of his or her home institution, and to behave in accordance with the interests of his or her home institution. This includes not only loyalty to advance a branch’s interest in the separation-of-powers context, but also positive loyalty toward agency- and department-level interests. We then suggest that the Constitution contains several mechanisms with the potential to generate institutional loyalty at the branch level. To establish a more complete accounting of relevant mechanisms, we identify further examples from the agency and legislative design contexts. These examples, we readily concede, involve loyalties to a subbranch level—but they are helpful to our analytic project nonetheless. The ensuing taxonomy illuminates untapped options for recalibrating institutional loyalties, and thereby enabling institutional retrenchment against potentially destabilizing partisan and ideological forces. We hence conclude that the modern position is right to posit that institutional loyalty does not emerge naturally or inevitably. We resist, however, the unspoken (if fairly plain) implication of the modern position that our constitutional system is bereft of mechanisms to induce and reinforce such loyalties.24

To demonstrate the potential for institutional redesign, we explore two distinct mechanisms that constitutional designers can use to elicit, or tamp down on, institutional loyalties. First, constitutional designers can manipulate the selection of officials who populate branches in ways that render them more or less likely to be institutionally minded. There are two relevant design decisions with such “selection effects.”25 These are rules that govern entrance to a branch and rules concerning exit. Scholars have previously explored selection rules’ use to promote a range of other constitutional goals, such as democratic accountability, transparency, and the minimization of democratic agency costs.26 But their underappreciated effect on institutional loyalty, we submit, rewards renewed attention.

Our second design margin hones in on the effect of organizational socialization on officials’ proclivity to align themselves with an institutional mission.27 The three branches of the federal government are bordered by a “thick political surround” of internal and external entities and interest groups.28 Against that backdrop, constitutional designers can advance or limit branch-level incentives by fostering an institutional mandate that ousts attachments to competing elements of the thick political surround, and thereby provides a crisper focal point for institutional loyalty. We explore the ways in which a constitutional designer (or a legislator) can “slice up” institutions of government in order to induce beliefs in a mandate. Alternatively, designers can harness, or even create, social networks to strengthen or undermine institutional loyalty. These networks, which emerge within and also cut across branches, are often ignored because of their informal, unstructured operation. But they too importantly promote (or undermine) ideological and partisan interests that compete with institutional loyalty.

Finally, in addition to these descriptive and analytic points, we aim to make a distinct normative contribution. We argue that constitutional designers usefully take account of institutional loyalty as part of their efforts to craft a desirable separation of powers. In our view, institutional loyalties are not in and of themselves intrinsically desirable ends. Rather, they play a foundational role in sometimes helping, and sometimes hindering, the realization of otherwise normatively desirable constitutional ends.

Nevertheless, because institutional loyalties can play a pivotal role in shaping how structural constitutional law operates at times, the situations in which these loyalties promote desirable constitutional goals ought to be considered and embraced. In contrast, when loyalties undermine those goals, they should be avoided. So the task of the institutional designer is complex: it is to calibrate the appropriate mix of such loyalties across the branches by estimating when they will advance needful constitutional ends, and when they will retard them. Of necessity, this task requires some estimation and informed prediction. It is not one that can be executed with mathematical precision given the vagaries of national political life. But this does not distinguish it from most other elements of constitutional and institutional design, which must be accomplished in the teeth of substantial uncertainty about the future.

Our reckoning of institutional loyalties, in sum, is more nuanced than Madison’s. Consistent with this subtler approach, we aim here to identify conditions under which institutional loyalty might motivate constitutional compliance, counteract disabling partisan polarization, and dampen the agency costs of representative democracy. We think that institutional loyalty should be cultivated to these ends. On the other hand, we flag instances in which such loyalty undermines the rule of law, thwarts the vindication of constitutional rights, and destabilizes the deliberative, polyarchic29 form of governance sought by the Framers. Not surprisingly, we think institutional loyalty in the latter cases should be titrated with greater caution.

In terms of specific reforms, our suggestions are branch specific. We think that useful institutional reform efforts focus now on increasing institutional loyalty within the legislature while diminishing it within the judiciary. The executive branch presents a subtler question. In some contexts, the executive is powerfully motivated by institutional loyalty in ways that redound to the public good. This may be especially so when elected actors press agendas that are directly disruptive of longstanding democratic or institutional practice. But in other regards, there is a case for diluting their effects in ways that protect the rule of law from potentially corrupting and distorting influences.

Our focus on these interbranch relations means we must sideline the related but distinct question of federalism as a cockpit in which institutional loyalty also plays a potentially salient function. The question whether state officials advance the institutional interests of states when lobbying Congress,30 participating in cooperative federalism programs,31 contributing to administrative agency rulemaking,32 or advancing structural constitutional arguments in the Supreme Court33 is an important and fascinating one. We hope our analysis shows the utility of an institutional loyalty–focused framing. Perhaps this lens is especially useful at a time when not merely individual fidelities to institutions, but even the stable and predictable operation of national institutions themselves, appear to be subject to pressure of sorts from populist political movements on all sides seeking to disrupt the institutional status quo.34

The Article proceeds in three parts. Part I sets the conceptual groundwork by defining and historicizing the concept of institutional loyalty. We demonstrate the past and present importance of institutional loyalty as both a complement to and substitute for other mechanisms to safeguard the separation of powers. Part II then introduces a typology of four mechanisms whereby institutional loyalty can be cultivated. For each pathway, we carefully examine necessary assumptions and prerequisites. In Part III, working across all three branches, we consider the extent to which institutional loyalty can be identified at work in each of the three branches. We further adduce suggested reforms for strengthening or rechanneling institutional loyalty based on the insights gained through Part II’s typology.

I.  Institutional Loyalties (and Their Critics)

We begin our analysis by clarifying the idea of institutional loyalty. We first offer a definition of the concept. We then trace its historical and contemporary importance to constitutional law. While the idea has deep roots and foundational importance to constitutional law, prevailing legal scholarship is largely hostile to the concept. We then develop a range of motivating examples to demonstrate the continued prevalence of institutional loyalty.

A.    Defining Institutional Loyalties

An institutional loyalty is an individual official actor’s psychological proclivity to perceive his or her proper course of behavior in terms of, or as incorporating, what he or she perceives to be the best interests of his or her home institution, and to behave in accordance with the interests of his or her home institution. To have an institutional loyalty is thus to maintain a stable conception of how best an institution’s mandate—the core purposes or functions it aims to achieve—can be promoted and to act in accordance with that conception. Competing loyalties and interests arise, of course, but loyalty to the institution helps ensure that behavior consistent with the interests of the institution persists in the face of this competition.

While the idea of an institutional loyalty has fallen out of current constitutional jurisprudence, it is a familiar one from our daily lives. Most of us belong to a team, a religious or civic institution, or an organized association (or even a law school). We necessarily decide when and whether to align our individual sentiments with the apparent needs of the institutions with which we affiliate. Within such institutional contexts, it is common to observe that some individuals identify and behave more consistently with the institution’s shared interests, while others hew to a more narrowly defined, individual conception of self-interest. Given the pervasiveness and obvious salience of such loyalties in ordinary life, we think it is at least worth asking whether analogs exist in public law.

Institutional loyalty is also familiar to scholars outside of constitutional law. Within the rational-choice tradition, institutional allegiances are evaluated in terms of the costs and benefits to individuals of participation in a group. The central collective-action problem that interest groups face, most famously identified by Professor Mancur Olson, turns on the incentives that individuals face to act in their own interests and thereby free ride on and undermine their institution.35 Following Olsen, public-choice scholars have written extensively about the conditions in which institutional loyalties arise and overcome individual incentives to free ride.36 But in so doing, political scientists have departed somewhat from the standard motivational premises of rational-choice theory. Some, in a tenor that is relevant to our project here, have contended that institutions are often constructed to have “purposes” and be “carriers of ideas,” such that their “norms and values affect their members” to act on behalf of their institutions.37

An institutional loyalty rests, whether explicitly or implicitly, on a contestable judgment about how to conceptualize a branch’s best interests. The right way to be loyal, say to Congress or the executive, cannot be identified mechanically ex ante. Institutional loyalists may therefore disagree about the precise demands imposed by their fidelity. That said, we think that institutional loyalty is often characterized by long time horizons. Given the durability of the branches, their interests are more likely to be understood in a longer rather than a shorter time frame. In contrast, the ideological and partisan loyalty identified by the modern position may evince a wider variety of time horizons, ranging from brief to long.

Institutional loyalty might also exist in two subtly different forms, but both yield similar behavioral effects. First, an individual official can perceive an institution’s interests as her interests: there is no gap between individual sentiment and institutional loyalty. An institutional loyalty, in other words, can be sincere.38 Alternatively, an individual official might disagree in whole or in part with the institution’s goals, but nonetheless decide to treat the institution’s interests as her own. She might do so for strategic reasons (for example, career advancement) or out of a sense of role morality.39 For the purposes of our analysis, we largely lump together sincere and strategic forms of allegiance.

Institutional loyalty is not necessarily identical to what Professor Daryl Levinson has called “empire-building government.”40 Consider again our threshold examples concerning Chief Justice Roberts and the executive-branch lawyers in OLC.41 In both, there is evidence that institutional loyalty was at work and that loyalty yielded careful thought about how best to advance an institution’s interests. Both Roberts and the executive-branch lawyers advanced an institutional agenda by trimming their home institution’s powers. To Roberts, it was (arguably) obvious that avoiding some divisive rulings would bolster the Article III judiciary’s reputation. To the executive-branch lawyers, it was obvious that their home department would be strengthened in the long term by advancing more limited legal claims (and hence perhaps a means of credible commitment that would reduce congressional resistance to delegation and judicial skepticism of executive action).

On both points, Roberts and the lawyers might have been wrong as a matter of fact: their chosen actions might have weakened their respective branches in unforeseen ways. Or they might have been wrong as a normative matter: the “best” way for an institution—be it a branch, a business, a family, or a nation—to prosper is rarely beyond dispute. But the complexity and nuance of their judgments suggest that what we call institutional loyalty need not be “empire-building” in character.42

B.    Historicizing Institutional Loyalties

Institutional autonomy is an important goal of American constitutional design. Democratic control over institutions has long been an important mechanism to generate this autonomy. Democratic control, though, can often be insufficient to protect institutions. From the Framers’ vantage point, ensuring that those working for institutions are loyal to these institutions was an important complement to democratic control and a necessary foundation for institutional autonomy.

The idea of institutional loyalty enters American constitutional law in James Madison’s account of the separation of powers in Federalist 51.43 The question of what would preserve institutional boundaries between the three branches was pressed actively in the constitutional ratification debates. Antifederalist critics of the 1787 Constitution were alarmed by what they perceived as deficiencies in the Constitution’s separations between powers.44 Many, such as the pseudonymous Pennsylvania Officer in the Late Continental Army, worried that the 1787 proposal had simply “not kept separate” different governmental powers.45 A related concern was that institutional barriers would not prove stable. The Philadelphia-based Antifederalist Centinel, for example, doubted the viability of “three balancing powers [that is, branches], whose repelling qualities are to produce an equilibrium of interests.”46

The most celebrated response to these arguments is found in a series of essays, beginning with Federalist 47, in which Madison, writing pseudonymously as Publius, defended the proposed “constitutional equilibrium” between the three branches.47 Madison argued that “dependence on the people” would be the “primary controul on the government.”48 Madison turned to Thomas Jefferson’s proposal of popular enforcement of separation-of-powers constraints via periodic constitutional conventions tasked with resolving interbranch contentions.49

But such popular control, Madison explained, would be inadequate for several reasons. To begin with, frequent conventions would imply “defect[s]” in government, sapping the “veneration, which time bestows on every thing.”50 Madison also worried that partisan “passions” would cloud popular judgment about the importance of institutional boundaries.51 Even when popular judgment recognized these boundaries’ value, the people might be handicapped because they lacked information about what their agents were doing.52 To be sure, Madison thought that the government’s relationship with the people “ought to be marked out, and kept open,”53 but he also knew that the aspiration toward transparency would on occasion be thwarted.54

These concerns were grave enough, Madison thought, to make a “necessity of auxiliary precautions.”55 If “better motives”56 protecting institutional boundaries would not always be found among the people, they could be supplied via the mechanism of institutional loyalty. Federalist 51 introduces the concept in a famous passage:

Ambition must be made to counteract ambition. The interest of the man must be connected with the constitutional rights of the place. . . . This policy of supplying by opposite and rival interests, the defect of better motives, might be traced through the whole system of human affairs, private as well as public. We see it particularly displayed in all the subordinate distributions of power; where the constant aim is to divide and arrange the several offices in such a manner as that each may be a check on the other; that the private interest of every individual, may be a centinel over the public rights.57

Madison’s argument has several elements, which can be usefully disaggregated: (1) the alignment between officials’ and offices’ interests (which we label institutional loyalty) (2) that will be “opposite and rival” to each other (3) so as to shield public rights and stabilize a “constitutional equilibrium.” We are interested here in the first element. This element, we note, is conceptually distinct from the other pieces of Madison’s argument.

In Madison’s account, institutional loyalty plays the part that “self-love” has in Adam Smith’s famous economic theory, a private vice that can be set in dynamic interaction against itself to promote the public good.58 Institutional loyalties lead officials to “resist encroachments of the other[ ]” branches.59 Official loyalty to one’s institution means that their resistance does not fluctuate along with popular “passions.”60 In this sense, institutional loyalty complements democratic control as a mechanism to protect branch-level boundaries, especially at moments when passions sweep the populace.

Madison limns three paths by which this mechanism has an effect. First, institutions can operate as “a check on [each] other.”61 If the executive branch is generating “encroachments” on Congress, for example, officials loyal to the legislative branch expend time and effort to “resist” the executive.62 Second, Madison also thought a third party (such as a federal court, or perhaps the several states) might identify and “resist encroachments.”63 The legislative branch can resist the executive branch by initiating or organizing challenges to the executive branch in federal court, for instance. Third, those within an institution have a comparative advantage in identifying and resisting encroachments on their authority.64

Writing as Publius, Madison and Alexander Hamilton both expressed their expectation that institutional loyalties would generate good government. Addressing the Senate’s willingness to punish executive-branch wrongdoing through the impeachment process, Hamilton in Federalist 66 hence avers to that chamber’s “pride, if not . . . [its] virtue,” as a spring of action.65 And, discussing relations between the national government and the several states, Madison hypothesized “motives on the part of the State governments, to augment their prerogatives,” and conjectured that even officials elected to federal office would have “prepossessions . . . generally . . . favorable to the States.”66 These pro-institutional inclinations, he suggested, would be so strong that even if the national government had “an equal disposition with the State governments to extend its power beyond the due limits,” the latter were likely to “have the advantage.”67

There is also a trace of an institutional loyalty–based argument in Hamilton’s defense of the federal judiciary in Federalist 78.68 On Hamilton’s account, members of the federal judiciary would be steeped in a dense network of “strict rules and precedents,” which require hard study to master.69 Judges’ behavior, Hamilton argued, would be oriented and shaped by organizational socialization on the branch. By analogy to then-contemporary models of human psychology, Hamilton seems to have supposed that the legalistic loyalty inculcated in Article III judges would orient the federal bench toward acting as the “conscience” of the federal government.70 It requires only a small step to hypothesize on this basis that judges’ guild loyalty will lead them to value and protect their branch’s distinctive institutional culture and role.

In developing this Madisonian vision, we think it is important to observe that institutional loyalty, while a crucial design dimension of constitutional law, is also a normatively complicated one. Institutional loyalty is not identical to constitutional loyalty. An institution’s best interests can be served by behaviors not contemplated or allowed by constitutional law. Moreover, the boundaries of an institution are already sufficiently robust that officials’ zeal and loyalty can disserve the stability of the constitutional system.

But this gap between institutional and constitutional loyalty is left largely unaddressed by Madison and his contemporaries. The Federalist Papers separately address institutional loyalty (in places like Federalist 51) and constitutional loyalty. Hamilton, for instance, defended the Electoral College in Federalist 68 as a means of ensuring loyalty to constitutional principle given its tendency to select for “characters pre-eminent for ability and virtue,” and not only those with a talent for “the little arts of popularity.”71 He did not connect this theory (however sound it might be in practice) to his accounts of institutional loyalties, or to the Madisonian vision of ambition checking ambition. Nor does Publius ever explain why a national representative process that generates officials inclined to pursue “the common good of the society”72 would also throw up institutionally disposed officials. In short, even when reconstructed with a friendly eye, the Madisonian account of institutional loyalties is characterized by gaps and discontinuities with the balance of Publius’s theory of constitutional design.

This institutional loyalty understanding of official motivation has not fallen completely out of the jurisprudence. In construing the Federal Vacancies Reform Act73 in National Labor Relations Board v SW General, Inc,74 for example, Roberts explained Congress’s decision to alter that statutory scheme in 1998 as motivated by a “[p]erceiv[ed] threat to the Senate’s advice and consent power.”75 At least in the Court’s view, therefore, it is still sensible to gloss congressional action in terms of the durable institutional prerogatives and powers of Article I institutions. In a similar vein, the District of Columbia courts have recognized that when Congress as “[an] institution [ ] files suit, it can obtain a remedy for the ‘institutional’ injury.”76 The recognition for Article III standing purposes of a legislature’s ability to vindicate “institutional” interests by seeking injunctive or declaratory relief is a formal legal recognition of the idea that the officials within one of the three branches can and do act on the basis of institutional, rather than ideological or partisan, grounds.77

C.    Questioning Institutional Loyalties

Current legal scholarship tells a different story about institutional loyalties. Leading voices in law reviews appear to be skeptical that institutional loyalties even exist and are skeptical of the notion that they might be sufficiently reliable and regular to explain institutional behavior.78 In consequence, such motivations do not play a large role in current constitutional scholarship on the separation of powers. To be clear, few scholars reject the bare possibility of an institutional loyalty. It is rather that the latter concept plays a relatively marginal role in their accounts. This is complemented by an embrace of an alternative strain of fidelities—partisan and ideological, in the main—that are cast as more acutely motivating than institutional loyalty. Behavior might be consistent with institutional loyalty, on this view, but is never caused by it. We term this skeptical approach the “modern position” on institutional loyalty. We set forth this skeptical story in general terms, and then consider its roots. This serves as a prelude to Parts II and III, and it suggests that institutional loyalty not only exists but can be analyzed in parsimonious and rigorous terms as part of constitutional law. To be clear, while we admire much of the scholarship that the modern position comprises, and find much to praise in it, we think its most ambitious variants sweep too far. We offer a modest course correction here—a measure of supplemental theorizing of a term that can otherwise easily drop out of institutional analysis.

The modern position is based on three interrelated descriptive claims about other affiliations that crowd out institutional loyalties. First, some scholars argue that partisan loyalties dominate and lead officials to act inconsistently with institutional interests.79 Officials rely on parties to win elections for Congress or the White House and to secure political appointments to the executive branch and the federal bench; they subsequently maintain their fealty to their partisan patrons.80 Officials also rely on parties to exercise significant powers while in office and pursue partisan agendas when interacting with other governmental bodies.81 As a result of these observed regularities, “realist claims about legal indeterminancy [sic] and the relation of law and politics are widely accepted in the academy.”82 To those who subscribe to the overwhelming power of partisan loyalty, it is a historical irony that Madison and his contemporaries (many of whom would go on to create national parties) did not realize that political parties would exist and would generate such powerful incentives that can in practice counteract institutional loyalty.83

Second, other scholars contend that ideological allegiances conflict with and overwhelm institutional loyalties. An official’s underlying sentiment about policy motivates their behavior, even when in conflict with institutional interests.84 This assumption dominates political science models of judicial behavior.85 It is commonplace in that literature to assert that “the Supreme Court decides disputes in light of . . . the ideological attitudes and values of the justices. Simply put, Rehnquist vote[d] the way he [did] because he [was] extremely conservative; Marshall voted the way he did because he [was] extremely liberal.”86 While measures of ideology have become more nuanced of late, many hew to the belief that the ideological loyalties of the justices motivate decisions, not their loyalty toward the Court qua institution. Building on this attitudinal model, positive political theorists have observed that even ideologically motivated judges must account for the likely strategic responses to their interventions and tailor their actions accordingly.87 While such models yield more nuanced predictions, they are characterized by the same a priori marginalization of institutional loyalty.88

Ideological loyalty is correlated with, but not identical to, partisan loyalty. Being motivated by a belief about what furthers a policy goal is different from being motivated by a belief about what serves a political organization. But presidents of one party sometimes, either on purpose89 or as an unintentional side effect of another goal, nominate an official with a differing ideological perspective, in some instances “using the agent’s known enmity to the principal’s benefit.”90 For example, presidents might nominate judges with different ideological preferences from their own as a way of credibly signaling the judges’ competence, or alternatively in order to impose costs on the partisan opposition. President Obama’s 2016 nomination of the eminent moderate jurist Merrick Garland to the Supreme Court, we think, has this flavor. At other instances, partisan affiliation may simply be a bad proxy for ideological preferences. Famously, President Dwight Eisenhower nominated both Justice William Brennan (a Democrat) and Chief Justice Earl Warren (a Republican) to the Supreme Court.91 He later expressed regret about both because of their liberal tilt.92 All that said, we note that the two major political parties have become increasingly ideologically homogeneous and polarized as party and ideology have become more strongly correlated.93

In addition, it is worth noting that it is quite possible for a partisan or an ideological loyalty to bleed into an institutional loyalty. An ideological commitment to certain policy goals may conduce to a belief in the primacy of one particular branch; for instance, a strong concern with national security might conduce to a preference for executive-branch primacy.94 In such instances, the line between ideological and institutional loyalties may be blurred. But the existence of ambiguous cases at the margin does not rob these categories of their utility as a general matter.

Third, the modern position argues that personal, materialistic loyalties conflict with and overwhelm institutional loyalties.95 A loyalty to maximize power (in current office or by seeking a higher one) could conflict with institutional loyalty. A desire for the immediate or longitudinal acquisition of wealth could also conflict with institutional loyalty. The federal criminal offense of honest services fraud, which penalizes naked self-dealing in the performance of “official act[s],”96 is one mechanism for mitigating such conflicts. The recent Supreme Court decision in McDonnell v United States97 narrowly construed that prohibition, effectively allowing personal pecuniary interests greater leeway to displace both ideological and institutional concerns.98

The modern position supplements these arguments by insisting on the impossibility of defining institutional powers. Madison might have been a theorist of institutional loyalty, but he famously wrote that “no skill in the science of government has yet been able to . . . define, with sufficient certainty, [government’s] three great provinces, the Legislative, Executive and Judiciary.”99 Legal scholars have likewise argued that defining what it would mean to protect “Congress” is a difficult enterprise.100 Absent a cogent account of institutional perimeters, it is assumed, there is no way to maintain a loyalty toward an institution.

Legal scholars articulating the modern position hence construct an account of governmental behavior in which there is scant space for institutional loyalty. When a president disagrees with large numbers of his partisan or ideological allies in the Congress, scholars are not only reluctant to ascribe this to a favorable loyalty toward the executive branch but instead offer the alternative explanation that it could “reflect[ ] the divergent preferences of the different temporal and geographical majorities that the House, Senate, and President represent (as opposed to the institutional interests of the branches as such).”101 Scholars writing about the persistence and predictability of institutional loyalties are therefore left playing defense, having to discount possible alternative loyalties that could explain a behavior that appears to be based on an institutional loyalty.

Official claims on behalf of institutional interests may also be discounted as a form of “cheap talk.” The modern position is that such claims are ex post rationalizations, not ex ante loyalty. Publicly asserting an alternative loyalty—instead of an institutional one—would be politically disastrous.102 For example, consider the congressional reaction to the equally divided Supreme Court’s affirmance of the Fifth Circuit’s decision that had earlier invalidated Obama’s immigration deferred-action programs,103 which included Republican Speaker Paul Ryan issuing a widely noted statement proclaiming his institutional loyalty. Ryan said that he supported the lawsuit to ensure that “Article I of the Constitution was vindicated” and that this success was a “major victory in our fight to restore the separation of powers.”104 The modern position is that Ryan talked a good institutional game but had a distinct and different underlying loyalty. As one of the major national leaders of the Republican Party, his partisan loyalty led him to challenge the actions of a president of the other party. As an ideologically conservative elected official, he did not believe in creating a legal status for undocumented immigrants. As an ambitious young politician, placing himself in front of a major national issue promised more power and prominence. Taking Ryan as a model actor in the constitutional system, it is a short step to a general rejection of the possibility of an institutional loyalty.

In short, the modern view of the motivations of institutional actors within our separation of powers leaves little space for institutional loyalties, even if it does not reject as a categorical matter the possibility of such fidelities. It is this lacuna in the literature that we aim to explore here.

D.   Identifying Institutional Loyalties

Institutional loyalty can be observed among contemporary officials in all three branches. Although we can identify instances of institutional loyalty in quite disparate circumstances, we make no claim here about their relative frequency in comparison to other motivations (for example, partisan, ideological, and so on). More modestly, we think institutional loyalty is not a marginal phenomenon. It has played meaningful roles in many important constitutional-law disputes. We offer a range of illustrations drawn from each of the three branches, extending our discussion in the Introduction.105 We err on the side of numerosity given that the idea of institutional loyalty is broad—for reasons explored at the end of this Part.

1.   The executive branch.

We begin by offering instances of institutional loyalty at work within the executive. We start with high-profile post-9/11 claims of executive authority in the Bush administration. We then turn to equally controversial debates about executive authority in the foreign policy area during the Obama administration.

In the wake of the September 11 attacks, the Bush White House articulated a “concede[dly] . . . aggressive” view of executive authority that would have preempted or narrowed congressional directives on military deployment, detention, torture, and electronic surveillance.106 Other policy paths of less resistance existed but were rejected because of a loyalty to a particular constitutional role for the executive. Many (but not all) of the policies pursued under this Article II flag could have been supported by “creative[,] . . . perhaps even tendentious,” interpretations of federal statutes.107 Republicans controlled the House until 2007, and the Senate after 2002, and likely would have supported new statutes authorizing much of what the Bush administration argued that Article II granted it the power to do.108

But members of the Bush administration, and in particular Vice President Richard Cheney, offered a different account. On their view, decisions were motivated by a personal commitment to establishing an expansive constitutional role of the executive, which in their view had been unduly cabined since the 1970s.109 In their own words, President Bush and Cheney had a long-standing desire “to leave the presidency stronger than they found it.”110 This assertion of this view of executive authority nevertheless ran the risk of “achiev[ing] the opposite” by cultivating “a harmful suspicion and mistrust.”111 Cheney, at a minimum, was quite aware that “personal leadership, public education, political support, and interbranch comity” all might be determinants of executive power.112 But he nevertheless advanced Article II grounds, even though they raised unnecessary, costly, and divisive objections to immediate policy choices, because they advanced the institutional authority of the presidency.113

The example of Cheney is an interesting one for our purposes because his fidelity to the executive branch was evident both when he sat in the White House and also while he was a member of the House of Representatives. It is worth remembering that Cheney started his career in the White House as an assistant to Donald Rumsfeld in the Office of Economic Opportunity in 1969–1970. He rose to the position of White House chief of staff for President Gerald Ford before he ever set foot in Congress.114

It is consistent with our view of institutional loyalties that Cheney seems to have formulated strong views about the executive branch as an employee of the executive branch and then maintained those views even as he moved to a different branch of government. Hence, the fact that Representative Cheney held the same views as Chief of Staff Cheney and Vice President Cheney does not undermine our claim; rather, it shows the potential for institutional loyalties to stick notwithstanding transitions between different branches of government.

Of course, this need not always be the case. There are other famous instances of interbranch transition—think of Chief Justice John Marshall, who served in the House, the Adams administration, and on the Court,115 or Chief Justice Roger Taney116—not characterized by the same tenacity of initial institutional fidelity.

While Cheney and his colleagues were motivated by an institutional loyalty to expand the power of the executive branch, other cabinet officials and senior political appointees resisted some of their efforts because of their institutional loyalties. Attorney General Ashcroft was a former Republican governor and senator who was “controversial because of his impassioned advocacy of conservative causes.”117 As he lay in an intensive care unit in the hospital on the night of March 10, 2004, though, Ashcroft summoned the energy to resist White House Counsel Alberto Gonzales and Chief of Staff Andrew Card when they asked him to reauthorize a domestic surveillance program. Ashcroft later testified and wrote that he refused to sign because he believed the program to be legally flawed. He was supported by Comey, who in his own testimony indicated that he believed the White House to be usurping the institutional role of the Department of Justice in deciding what is legal.118

Institutional loyalty, moreover, is not the preserve of one administration or one party. In 2011, Obama determined that he did not need congressional authorization under the War Powers Resolution119 to continue air strikes in Libya.120 This conclusion was resisted by two political appointees—Jeh Johnson, the general counsel of the Department of Defense, and Caroline Krass, the acting head of the OLC at the Department of Justice.121 Johnson had been a political appointee during the Clinton administration, had raised money and campaigned for past Democratic presidential nominees,122 and would later serve as secretary of Homeland Security for Obama. Krass had previously served as a special advisor to Obama,123 and would later become his general counsel for the Central Intelligence Agency (CIA).124 Despite these partisan commitments, though, both Johnson and Krass argued to Obama that his legal conclusions were incorrect, and a former head of the OLC, Walter Dellinger, criticized the president’s “unusual process” of rejecting OLC and other legal advice from within the executive branch as problematic.125 In so doing, they reflected an institutional alignment at odds with traditional stories of partisanship and ideology.126

2.   The judiciary.

Two recent examples highlight the possibility of institutional loyalty among federal judges. A first high-salience example of a justice acting against perceived ideological preferences as a result of what apparently were institutional concerns is, as noted previously, Roberts’s alleged switch to vote in favor of upholding the Patient Protection and Affordable Care Act127 (ACA) in National Federation of Independent Business v Sebelius128 (“NFIB”). Commentators immediately characterized Roberts’s decision as “a brilliant act of judicial statesmanship” that shielded the Court, as an institution, from accusation of partisan bias.129 Roberts’s vote, we think, is hard to explain purely in terms of his known ideological preferences.130 Consistent with that view, many conservative commentators subsequently condemned the chief justice in no uncertain terms as an ideological turncoat.131 Although it is not possible to say with certainty what motivated Roberts, we think that institutional loyalties are plausibly thought to have played a role.

Similarly, in the wake of Justice Antonin Scalia’s untimely death in 2016, the chief justice, and other justices, made seemingly concerted efforts to “find consensus whenever possible” by avoiding 4–4 splits and reducing the number of disputes in its pipeline.132 In Zubik v Burwell,133 the justices were asked to decide whether regulations issued pursuant to the ACA, mandating that employers provide contraception coverage to women violated the statutory rights of nonprofit religious employers.134 Just two years earlier, the Court divided 5–4 in Burwell v Hobby Lobby Stores, Inc135 about the related issue of whether the mandate unduly burdened the religious freedoms of for-profit corporations.136

By the time oral argument transpired on March 23, 2016, Scalia had died, leaving a likely 4–4 split on the Court.137 Obama had also nominated Garland of the US Court of Appeals for the District of Columbia Circuit to replace Scalia, leading to Republican opposition and placing the Court in the midst of the heated political debate in a presidential election year yet again.138 Less than one week after oral argument, the eight justices unanimously “issued an unusual order”139 directing the parties to file supplemental briefs finding a compromise outcome that both sides would find agreeable.140 In May, the Court issued an unsigned unanimous opinion remanding the case to the lower courts to find an acceptable compromise order.141

As in the NFIB case, the justices appeared to behave in ways inconsistent with their ideological preferences (that is, their sincerely held opinions about the substance of the law), and instead consonant with the interests of the judiciary as an institution. A polarized and deadlocked Court avoided dividing on ideological or jurisprudential grounds in order to preserve the public perception of the Court as above and independent of politics. This was certainly the interpretation of the opinion by many of those who follow the Supreme Court most closely.142 Indeed, Justice Elena Kagan confirmed this interpretation during a subsequent public appearance.143

3.   Congress.

Examples of institutional loyalty influencing the collective action of either Congress or a single chamber are more difficult to discern than their judicial- or executive-branch analogs. A threshold problem is that in many instances, Congress asserts its interests by what Professor David Mayhew calls “non-levering”—that is, exercising a veto by nonaction.144 There are many examples, to be sure, of legislative assertions of institutional prerogative as a means toward a partisan end. For example, the Senate’s refusal to hold confirmation hearings for Obama’s nomination for the Supreme Court vacancy left by Scalia’s sudden demise sounds in institutional prerogative. Similarly, a lawsuit filed by the House of Representatives challenging expenditures on the ACA has resulted in a district-court opinion that endorses a legislative power to challenge violations of the Appropriations Clause.145 But we think both are better understood as partisan initiatives with small positive institutional spillovers.146

Better examples of institutional loyalty on Congress’s part focus on costly legislative acts that have little immediate partisan or electoral payoff, but that have enabled the legislative branch as a whole to pursue its identified interests in the long term. That is, when a legislative action produces little or no certain short-term policy effect, but in the long term enables Congress, it is more likely to be explained by institutionally oriented motives.

The leading comparative study of institutional design changes within Congress, by political scientist Professor Schickler, highlights the plurality of motives that underscore many legislative acts.147 Notwithstanding that finding, Schickler identifies several instances in which what we call institutional loyalties were decisive (if not uniquely at play). The first of these is the LRA, which “sought to enhance Congress’s position relative to the executive by strengthening congressional committees and by providing new integrative devices, such as party policy committees and a centralized budget process, to coordinate committee activities.”148 As we have noted, the LRA slimmed the number of committees in both the House and the Senate as a means to empowering more effective congressional oversight. To that end, it also authorized additional staff to help professionalize legislators’ offices.149 Congressional debates show that legislators were concerned by the disjunction between a rapidly expanding executive branch and a Congress that had “relatively stood still.”150 This “remarkably consistent message from Democrats and Republicans across the ideological spectrum” about Article I power and “prestige” helped enact a measure that stripped many members of the perquisites of committee leadership and membership—that is, diluted their personal power without advancing an ideological or partisan goal.151 The LRA’s history is also striking in that it is characterized by a bipartisan recognition of the positive role that government plays—in contrast to the relentless attacks on government from Republicans in particular since the 1980s.

But the 1946 Act is not the only instance of Congress acting in its own defense that Schickler identifies. Between 1889 and 1990, he finds that at least nine out of forty-two institutional reforms in Congress were motivated in part by what he calls “Congress or chamber-oriented interests” as distinct from party interests, policy interests, and reelection-related interests.152 In addition to the LRA, Schickler identifies appropriations and Senate committee reform in the 1920s, the Joint Committee on Atomic Energy of 1946, and the Stevenson committee reforms of 1977 as relevant examples.153 The Joint Committee on Atomic Energy, for instance, was created after World War II as a joint body with members from both the House and Senate exercising sole jurisdiction, sole power to report relevant legislation, and important oversight powers over the vital and growing postwar policy question of how to manage atomic energy.154 Like the LRA, the Joint Committee emerged from a bipartisan consensus that Congress needed the institutional capacity to keep up with regulatory growth within the executive.155 In short, the “middle to late twentieth century” was a particularly fruitful period for such reforms, as Congress endeavored to respond systematically to increasing executive-branch authority.156

In addition to Schickler’s examples, Professor Josh Chafetz’s recent survey of congressional powers identifies the creation of the General Accounting Office and the Senate’s defense of its prerogatives in relation to treaties as instances of legislative self-assertion that are not well explained by partisan or ideological motives.157 The former, created in 1920, was intended to fashion a new accounting department that would be accountable “only to Congress” and that would give Congress “the very facts that Congress ought to be in possession of.”158 Building on Chafetz’s list, we think that the Congressional Budgeting Office, which is tasked with issuing estimates of how much proposed legislation will cost the federal fisc, is an institutional innovation that reflects institutional rather than partisan or ideological interests.159

Some scholars have identified the federal budgeting process as a “centralizing change intended to safeguard congressional power.”160 In 1974, provoked by President Richard Nixon’s aggressive use of impoundment authority, Congress passed the Congressional Budget and Impoundment Act of 1974,161 which provided a framework for coordinated committee consideration of expenditures, and a predictable set of mechanisms for legislative deliberation over the final packet of budget proposals.162 At its heart was the new annual congressional budget resolution, which set forth “overall national fiscal policies” at the beginning of the budgeting process, thereby enabling coordination toward a final appropriations measure.163 Although these new procedures can be glossed as means of empowering Congress as an institution, they can also be seen as a mechanism for empowering members of budget committees while increasing the time in which members could seek payoffs consistent with game-theoretic models of self-serving legislative behavior.164 Although Schickler does find evidence of institutional motives in relation to the 1974 Budget Act, we think the force of parochial interests may be stronger here than in the other cases mentioned above.

Finally, Professors Adrian Vermeule and Cass Sunstein have recently suggested that the 1946 Administrative Procedure Act165 (APA) is similarly a historical “compromise” between proregulatory and antiregulatory forces, which is not “generally and systematically progressive, or proregulatory, or anything else.”166 If they are correct, the APA should be understood as a farsighted investment in a legal framework that enabled Congress to achieve its shifting regulatory goals without having to create a basic legal framework from the ground up each time it did so. Professor Elizabeth Garrett has identified other forms of “framework legislation” that Congress has enacted and still enforces that could be described in a similar fashion to budgetary legislation or the APA.167

Although we have assembled examples of institutional loyalty manifesting within Congress, we do not wish to exaggerate the force of our claim. It may once have been the case that Congress engaged in robust self-defense of its institutional prerogatives168 by (among other things) resisting judicial orders of which it disapproved.169 But it is striking that all of our examples are relatively removed from the contemporary moment. This suggests that institutional loyalties have eroded over time as a result of many forces, both within and outside Congress.170 Today, only a pale shadow of their former force may be felt.

* * *

Institutional loyalty can be observed motivating recent actions of all three of the branches. Each of our examples highlights an instance in which loyalty prevailed against countervailing partisan or ideological concerns. Our examples not only provide evidence of institutional loyalty but also show how the latter influences officials’ ultimate actions, notwithstanding partisan or ideological preferences. Because we have selected relatively high-profile examples, we are confident that institutional loyalties cannot be written off as a marginal phenomenon: they are instead a meaningful element of contemporary separation-of-powers dynamics.

Our examples, however, leave open the questions of why and how the strength of institutional loyalties has fluctuated over time and how such changes have interacted with shifts in partisan and ideological attachment. It seems likely that the core claim of the modern position—that ideological and partisan loyalties dominate—is itself a contingent historical artifact of the ebbing strength of institutional loyalties. Indeed, one benefit of bringing to bear the concept of institutional loyalties is that it casts light on the process of motivational change within the political elites charged with managing the three branches of government.

For example, rates of both partisan polarization among political elites and party discipline within Congress have changed markedly over time171 in ways that may have influenced the strength of institutional loyalties. Polarization, for example, arguably undermines legislators’ willingness to pursue institutional rather than partisan ends as party leadership exercise greater agenda control in pursuit of distinct and ideologically incompatible agendas.172 On the other hand, the decline of institutional loyalties—say as a result of increasingly lucrative exit options for former members of Congress—may have contributed to the growing force of partisan and ideological preferences. The interaction between partisan or ideological motives and institutional loyalties provides a motor driving changes to institutional behavior over time. It would thus be a mistake to describe the motivations of key actors in our constitutional system as static, rather than as dynamic and evolving over time.

II.  The Sources of Institutional Loyalty

If institutional loyalty is neither impossible nor inevitable, how does it come to be in the first instance? This Part focuses on the question of how institutional loyalties arise. Our aim here is to identify and loosely categorize the discrete choices made by the designer of a constitution, or of important national institutions, that make officials either more or less likely to identify with and seek to promote the goals of their institutional home. We identify two relevant margins of design. Each can be manipulated in two different ways.

First, we explore the influence of institutional design on the selection of institutional personnel. Institutional designers have a large measure of control over entrance and exit rules. The terms of entrance and exit by officials into an institution, therefore, select not only on the kind of officials who opt into the institution in the first instance, but also the extent to which they remain long enough to develop institutional loyalty.173

Second, we posit that officials’ preferences are shaped and regulated by organizational socialization, and in particular the social context in which they operate. Social context, operating both within and around an institution, can foster “durable, transposable dispositions . . . [that in turn will] generate and organize practices” by which branches, agencies, and organizations implement missions.174 We explore two mechanisms through which institutional designers can use organizational socialization to promote loyalty and identification with an institution. First, the choice of institutional mandate can influence the extent of such identification. Second, we explore how institutional designers can harness, and even seed, wider social networks that surround an institution, generating or reinforcing positive (or negative) loyalty toward an institution.

All these mechanisms can operate at the level of constitutional design, understood as the written work product of the Philadelphia Convention and successive amendment. Because the 1787 Constitution does not comprehensively describe the full institutional landscape of the federal government, however, it also leaves ample room for legislators and presidents to develop the “small-‘c’” constitution, which comprises “the fundamental political institutions of a society, or the constitution in practice.”175 As a result of the Constitution’s incompleteness, many possible avenues for institutional reform (some of which are identified in Part III) remain open. We hence sketch here mechanisms that are available both to constitutional drafters and more mundane administrators of government.

A.    Entrance and Promotion Rules

A first means to elicit institutionally disposed officials is to employ devices for regulating the threshold choice of official hires in ways that sort the institutionally minded from those with self-serving or ideological or partisan motives, and to ensure that only the former are promoted within the organization.176 Some (but not all) of the design options described in this Section operate through a demand for a costly signal. Such measures leverage the insight that when the cost of a qualification “is negatively correlated with the unseen characteristic that is valuable to the employers,” it can be used to distinguish “good” types (who easily acquire it) from “bad” types (who do not find getting the qualification worthwhile).177 In addition to devices that select for more institutionally minded candidates, an institutional designer can sculpt criteria for promotion—that is, the shape of the job ladder—to select right-minded individuals. We begin by exploring selection mechanisms in the constitutional context, before using statutory examples to illustrate other mechanisms not identified or used in the Constitution’s text.

As noted, Madison, in Federalist 51, asserted that “[t]he interest of the man must be connected with the constitutional rights of the place.”178 The balance of The Federalist Papers reveals scant attention, however, to the important question of how this connection would be made. In contrast, Publius pays close attention to how selection rules for the branches conduce to virtuous officials. In Federalist 10, most famously, Madison developed a theory of the “extended” republic, in which the process of representation would dilute local factions and “refine and enlarge the public views.”179 As we discussed in Part I, though, virtuous officials are not necessarily those with institutional loyalty.180

At best, Publius’s argument offers mere hints of how institutional loyalty might be produced by selection. In Federalist 57, Madison suggested that an elected official’s “pride and vanity [would] attach him to a form of government which favors his pretensions, and gives him a share in its honors and distinctions,”181 perhaps including those of an institutional character. Madison in Federalist 62 glosses the pre–Seventeenth Amendment regime of state legislative power to appoint senators not only as a way of “giving to the state governments . . . an agency in . . . the federal government”182 but also as a means of picking senators with “due acquaintance with the objects and principles of legislation.”183 The latter phrase might be glossed (with some difficulty) as an inclination to attend to the rights and interests of a legislating institution. Finally, Hamilton emphasized in Federalist 78 that Article III judges would be appointed via the same method as principal officers of the executive branch—and also praised judges for their expected quality of “judgment.”184 Although this again might be read as evidence of selection for an institutional loyalty, it also suggests an optimistic view of the filtering power of presidential nomination and Senate confirmation.185

We do not think that these hints add up to a complete account of the Constitution’s selection mechanisms as a means for promoting institutional identification. Nor do we think such a comprehensive theory, explaining how the Constitution selects for both public virtue and institutional loyalty, can be developed from the constitutional text alone. As Professor Joanne Freeman has explained, it is not obvious that the Framers’ ideas about republican virtue can be reconciled entirely with their commitment to representative democracy.186 Trying to knead their invocations of public virtue into claims about institutional loyalty seems even more implausible. As a theorist of institutional loyalty as a matter of constitutional design, therefore, Publius falls far short.

Other constitutional designers have done better. In many other jurisdictions, the judiciary is organized as a form of civil service with lifetime career paths. Career judiciaries rely on costly signals insofar as they demand that judges renounce potentially lucrative private careers early in their professional lives. By building job ladders within the judiciary, career judiciaries also align the professional ambitions of judges with the goals of the court system as a whole. In Germany, for example, Professor John Langbein characterized the judiciary as a prestigious career choice, open only to those with the “best” academic credentials in their study in programs dedicated to the law, with partisan considerations playing a “very subordinate” role.187 Promotion depended on meritocratic evaluation by peer judges, with partisan considerations playing a much lesser role.

Langbein, the leading American commentator on the German judicial system, reaches an unequivocal judgment: its career path attracts the “very able” to the bench and then assures that “career advancement [is] congruent with the legitimate interests of the litigants.”188 The German judiciary, in short, selects for and cultivates an institutional loyalty by orienting its officials toward excellence in the performance of the institution’s central task: settling disputes fairly. Of course, mere installation of a career judiciary is no panacea.189 Rather, we flag the German example as evidence that other constitutional designers may have identified ways of eliciting desirable institutional loyalty among civil servants.

Institutional loyalty might also be cultivated, or diminished, through statutory selection mechanisms. The modern American civil service has a series of rules that embody both costly entrance mechanisms and also promotion schemes that entangle individual and institutional ambitions. Contrary to the standard “starting point in analyzing politicians’ behaviors[,] that they are socially motivated,”190 careful attention to these schemes shows how extant statutory frameworks carefully cultivate institutional loyalty.

Starting with the Pendleton Act of 1883,191 statutory civil-service laws have contained filtration mechanisms, such as competitive entrance exams.192 The current merit hiring process for the federal civil service requires that career positions are filled based on objective skill and experience.193 These threshold mechanisms make long-term government service less attractive for those with strong partisan or pecuniary motives, but lacking in relevant skills or subject-matter expertise. As a result, these screens narrow the pool of applicants to those most likely to internalize an institutional loyalty, even if applicants themselves are unlikely to have such loyalty before taking on a government position.

In addition to navigating these screens, professional civil servants must forgo the greater compensation typically available in the private sector when choosing to enter public service. Lawyers, engineers, scientists, and accountants all earn far less than their counterparts in the private sector.194 This public/private salary differential has two effects. Like competitive exams, it again selects against certain types (for example, those primarily motivated by personal pecuniary motives). And by ousting high-powered incentives,195 it also preserves space for other forms of motivation. Institutional ambitions and norms are, as a result, far more likely to infuse the preferences and behavior of officials than would be the case absent the civil-service regime’s constraints on political and pecuniary motives.196

These threshold mechanisms can be complemented by a promotion structure that elicits institutional loyalty even more actively. Promotions depend on “seniority and the passage of time, rather than on productivity,” or evidence of skill.197 Partisan and pecuniary motives are excluded as grounds of official action by the civil-service laws. The 1939 Hatch Act198 ensured that “employment and advancement in the Government service [does] not depend on political performance,”199 just as the honest services law rules out self-interested pecuniary motives among elected officials.200 Legal historian Nicholas Parrillo has charted the demise of profit as a motivating force within the federal bureaucracy as fees and bounties gradually fell into desuetude, to be supplanted by fixed, outcome-independent salaries.201 To the extent that salary competition does occur now, it is channeled through a centralized system of classifications of different bureaucratic positions for salary purposes—a context in which institutional priorities are highly influential.202

Finally, once partisan and pecuniary motives are taken off the table, civil servants also tend to have long careers closely tied to their home institutions. Turnover in federal employment is low. In each year between 2004 and 2012, data from the Office of Management and Budget (OMB) shows that about 0.7 percent of federal employees were fired, between 2.4 to 3 percent resigned, and between 2.5 and 3.6 percent retired.203 While OMB does not retain distinct data on rates of internal promotion and hiring, this exceedingly low rate of exit (around 7 percent) suggests that promotion from outside the federal bureaucracy is relatively rare. The sheer expected durability of federal employment tends to deepen identification with a home institution.204 Simply put, officials who anticipate that their career will be entangled with a specific institution have a reason to advance the interests of that institution qua institution.205

B.    Exit Rules

Institutional loyalty can be elicited by imposing costs on officials’ exit from an institution. Costly institutional exit arises when officials incur high opportunity costs if they depart an institution or obtain low or artificially suppressed returns from external opportunities. As a result of the benefits of staying put and the costs of departure, officials are more likely to remain in their home institution, and concomitantly more likely to identify their career goals with the larger goals of the institution. Costly exit has been identified as an important mechanism in both the labor economics and the public administration literature.206 Its links to constitutional and statutory rules, not to mention its larger constitutional function, have so far received insufficient attention.

Costly exit mechanisms operate in two ways. First, high returns to serving in an institution increase the odds of longer terms of service. Longer civil-service tenures function in this regard akin to life tenure for federal judges. In expectation, such tenures increase the degree of identification between an official and the institution with which her career is entangled. Regardless of the length of intended institutional service, moreover, high returns from continued service can immediately benefit institutions. An official receiving large returns to institutional service, like any well-compensated employee,207 is more likely to be well disposed to her institution, and hence more likely to formulate and adopt a view of what is in its best interest.

High returns, moreover, need not take a monetary form. Public officials can instead be motivated by the opportunity to shape policy or by reputational gains.208 High-ranking officials in all three branches of government are likely to have forsaken greater financial rewards in the private sector in order to receive returns from policy influence. Chief Justice Roberts, for instance, was earning more than $1 million every year in private practice in 2003 when he accepted President Bush’s nomination to the US Court of Appeals for the District of Columbia Circuit, and thereby embraced a position with roughly one-sixth the salary.209 His tenure protection and opportunities for policy-related discretion are one form of compensation for the opportunity costs of forgoing private practice. Indeed, to the extent that this effectual wage is greater than Roberts’s market wage, life tenure may comprise an implicit efficiency wage.210

A constitutional designer can use policy influence as a compensatory incentive in lieu of pecuniary rewards, and thereby elicit institutional loyalty. Article III judges, for example, cannot have their salary reduced while in office. Their salary is essentially a lifetime guaranteed annuity so long as the judge is not impeached.211 This guarantee is justified as a (surely partial and imperfect) way of ensuring “an independent Judiciary”212 that is “free from control by the Executive and the Legislature.”213

An institutional function is thus pursued through the shaping of individual incentives. By assigning judges the opportunity to influence the path of the law as an important form of implicit compensation, the Constitution gives individual judges a stake in the judiciary as a whole. Their policy discretion depends on the judiciary’s continued prestige and legitimacy. As a result, judges have a strong incentive to maintain the institutional predicates of policy influence. Further, the expectation of a durable judicial career means that even if a judge’s influence at a given moment in time is minimal, her opportunities for shaping the law can reoccur over an extended period of time—resulting in an eventually significant amount of policy influence.

Statutory regimes can also use the promise of policy influence to generate institutional loyalty. Agency designers can allocate either substantive powers or formal titles as means to assign both responsibility and influence to specific personnel.214 A recent example involves the reorganization of the intelligence services after the terrorist attacks of September 11, 2001, to include a new leadership position, a “Director of National Intelligence” (DNI).215 The DNI was made by statute the head of the intelligence community, the primary adviser to the president on matters of intelligence related to national security, the authoritative voice on the intelligence budget, and the hiring authority for key officials in the intelligence community.216 Whereas the head of the intelligence community had formerly been the CIA chief, the new DNI is independent of any specific component of the intelligence bureaucracy.217 By augmenting the powers of the intelligence leadership position, while detaching it from any specific agency, members of Congress hoped to instill a larger, government-wide sense of mission in the office—one not dogged by the parochial concerns of a particular agency.218 Implicit in the DNI’s new powers, moreover, is the possibility that the office’s occupant would be blamed politically if the intelligence community failed to prevent another spectacular terrorist attack akin to 9/11. In this fashion, legislators may have hoped to align future DNIs’ personal interests with the executive’s larger mission of mitigating national security risks rather than more parochial institutional concerns.

The force of such bureaucratic incentives, however, often depends on the strength and durability of the underlying agency. Implicit compensation in the form of policy discretion for bureaucrats works better with robust and enduring agencies than with weak and transient agencies. All else being equal, officials exercising policy discretion as a form of implicit compensation thus have good reason to identify with and promote their institution’s persistence. There is a strong empirical connection between the extent to which an institution is insulated from presidential control and its durability.219 This suggests that an agency is less likely to attract and cultivate expertise if it is under close presidential supervision because the marginal official will be uncertain whether any context-specific expertise they accrue will be rendered valueless by the dissolution of the agency. Institutional loyalty toward a specific agency or department—and hence to the mission embodied in the statutes that agency is charged with enforcing—is undermined by structural controls that increase a president’s control and influence.220 The weaker presidential removal authority, the more durable an agency and the more likely its officials are to invest in expertise.

The second species of costly exit mechanism takes the form of rules that inhibit or even bar exit, especially to positions that provide higher returns than continued government service. Officials are more inclined to identify with their institutions if there is no alternative career pathway, such that they are more likely to pursue a longer career within government. Lower returns from exit can result if officials have institution-specific investments that are not transferable either to another institution within government or to the private sector.221 Low returns to exit can also result from formal, legal prohibitions to exit. We consider each of these possibilities in turn.

Most obviously, institutional designers make exit costly by simple prohibitions on the utilization of any capital obtained while an official within an institution. More institution-specific investments are encouraged, by contrast, when substantial portions of capital obtained from working within the institution cannot be monetized on departure. The Ethics Reform Act of 1989,222 for example, bars former members of Congress from lobbying for a year;223 a measure recently extended this prohibition to two years for senators.224 Congress’s failure to bar lobbying of agencies supervised by an individual legislator, however, may render these prohibitions functionally ineffectual.225 On his first day in office in 2009, President Obama issued an executive order barring federal employees from a wide range of lobbying activities for two years.226 The following year, he barred lobbyists from serving on federal boards.227 Such measures aim to gum up the “revolving door” between government and the private sector.228 When successful, they lower the opportunity cost of remaining in public service by lowering the expected payoffs of exit. When discarded, they diminish institutional allegiance in favor of baser concerns.

The second mechanism is less obvious. It relies on a connection between expertise acquisition and costly exit. Federal officials develop institution-specific human capital by maintaining responsibility for matters not routinely addressed by other public or private institutions.229 The design of federal civil-service laws often facilitates the acquisition of context-specific human capital by employees, that is, expertise and knowledge tightly wound into the specific aims and policy goals of their home agency. Civil servants generally do not enter their positions with the knowledge and skills needed to pursue institutional goals. Even a system of merit selection will not select for the “most meritorious” employees in the potential applicant pool unless the applicants “believe that their efforts and expertise will be applied to pursue goals” that they share.230 Such expertise is often “relationship specific” and “specifically tailored” to the operational environment and goals of a particular agency.231 Tenure protection induces officials to invest in that bespoke human capital.232 By inducing the acquisition of institution-specific expertise, the civil-service regime thus helps create a cadre of officials whose professional standing, and whose specific expertise, is tightly linked to the specific agenda and policy goals of their institution. These officials are, all else being equal, likely to rank the institution’s goals very high.

Consider, for instance, the Office of the Legal Adviser in the Department of State, or “L,” an office renowned for the institutional orientation of its lawyers, even in comparison to some other legal offices within the executive branch.233 Many attorneys there work on matters of public international law, a field with little substantial presence in other parts of the federal government, state governments, or the private sector. Lawyers in L also invest in social networks related to public international law through organizations like the American Society for International Law. As we explain below,234 this generates human capital within that office that is not as easily transferable even compared to that acquired in other forms of government legal service. The result is longer tenures within L,235 and deeper identifications with that office and the executive branch’s foreign policy missions more generally than otherwise might be the case.236

Finally, the development of institution-specific human capital helps elucidate why political appointees tend over time to develop agency-level institutional loyalty that can overpower partisan loyalty. Presidential appointees rely on agency staff for knowledge and expertise on policy questions.237 By becoming immersed in and acquainted with a complex body of rules, those appointees also make personal investments in an epistemic resource that is specific to a particular institution (the judiciary or an agency).238 This asset-specific investment then ties the official’s interests to those of the agency in which he or she is embedded. As a result, political appointees have been known to defect from the partisan or ideological agenda of the official’s putative executive-branch sponsor.239

C.    Institutional Mandates

We turn next to the first of two ways in which constitutional designers can take advantage of organizational socialization to generate institutional loyalty. Unlike entrance and exit rules, organizational socialization operates as a “treatment effect” rather than a selection effect.240 In the following two Sections, that is, we are concerned with ways in which the design of an institution can influence an individual’s perception of, and tendency to identify with, her home institution. Our basic claim is that an institution’s dominant architectural elements can elicit an internal culture in which institutional loyalty will thrive.

We begin by isolating a simple element of institutional design—institutional mandate—as a significant determinant of institutional loyalty. The stronger an institution’s mandate, we suggest, the more likely institutional loyalty will emerge. We then explore how different design choices observed in either constitutional or statutory contexts either diminish or augment relevant social networks.

1.   Defining and cultivating institutional mandates.

Governmental institutions—whether a branch or an entity within a branch—are typically understood by participants and observers alike to have a set of purposes or functions. These constitute the institution’s mandate. An institution’s mandate need not be articulated in a constitution or an organic statute, although they often are. Institutional loyalty can be understood in terms of officials’ endorsement of those purposes and functions into their own preference sets.

An institutional mandate can be created by the constitutional provision or statute that creates an institution. Alternatively, it might be a result of policy entrepreneurship by agency leaders with strong policy agendas.241 Institutional leadership can expend resources priming new officials on the importance of particular functions or aims. They can equip new appointees with the epistemic resources and practical capability to understand and execute a mandate. Even when not backed by such investments, both written and informal mandates can serve as “focal points,” helping to coordinate actors’ expectations and behavior in light of some institutional ends and not others.242 Common knowledge of the agency’s mandate, for instance, may provide a basis for coordination among officials who otherwise have little knowledge of peers in physically remote offices or functionally separate divisions of the agency. Such coordination, Professors Tiberiu Dragu and Mattias Polborn have recently demonstrated, is particularly important in maintaining the rule of law against the efforts of potentially autocratic leaders.243 When administrators expect each other to resist unlawful policies, they show formally, legal constraints on political leaders are more likely to be self-enforcing.244

Of importance to our analysis, an institution’s mandate need not be unitary. By substantive command or via the imposition of procedural obligations,245 an agency can be tasked with a plurality of goals. There is not a sharp divide between unitary and plural mandates, but rather a spectrum. At one end of the spectrum, an entity might be assigned a substantively narrow obligation that requires little by way of discretionary judgment or expertise. The institution’s objections also may be connected to official action by a relatively short causal chain. A DMV tasked solely with examining potential drivers and distributing permits to them has something of this character.

At the other end of the spectrum, an entity might have several, potentially conflicting goals that relate to official actions through long and uncertain causal chains.246 The Food and Drug Administration (FDA), for example, is “charged both with ensuring that new drugs placed on the market are safe and effective (a task that generally requires cautious and deliberate action) and with speedily granting access for doctors and patients to those new, safe, and effective drugs (a task that requires expeditious review of those drugs).”247 The FDA, when pursuing these conflicting mandates of safety and public health, necessarily makes compromises between incommensurable ends.

The key point for (big-c or small-c) constitutional designers is that the internal heterogeneity of an institutional mandate influences the extent to which officials are likely, or even able, to formulate institutional loyalty. The more plural and the more abstract an entity’s goals, the more disagreement there is likely to arise even among insiders about how best to carry out its mandate. In the stylized examples we have just offered, it is much easier (and hence less costly) to discern the institutional mandate for the DMV branch office than for the FDA. Entities with more diverse institutional mandates, accordingly, are less likely to be populated by officials with institutional loyalty than entities with singular and unitary mandates.248

The effect of institutional architecture on mandates is evident both at a constitutional level and a statutory level. With the Constitution, both single- and multiple-mandate branches can be observed. Textualist constitutional scholars draw a distinction between a “unitary” executive and the “plural” judiciary and legislature.249 Although the force of this textual argument can be disputed,250 the simple contrast between unitary and plural branches plainly has some force. The relatively hierarchical structure of the executive juxtaposes with the relatively flat structure of the legislative branch, in which each legislator’s vote formally has the same weight within a given chambers. It is thus no surprise we commonly talk of the “Obama” or “Trump” presidency, and not (usually) the Ryan/McConnell Congress.251 Those labels suggest a widely shared belief that (1) there is a singular measure of branch-level performance for Article II and (2) successes and failures can be attached to a specific officeholder for the executive, but not Congress.

Even if two institutions have the same functions, the way in which they execute that function can influence the extent of relevant officials’ loyalty and zeal in pursuing that function. Consider the distinct ways in which the executive and the legislative mediate political competition. The quadrennial presidential election provides for the diachronic alteration of power between partisan factions. Effectual political power under Article II comes in a unitary package, which changes hands periodically. By contrast, while Congress too has periodic elections, its plurality and heterogeneity invite an additional element of synchronic political competition. The simultaneous possession of political power by plural, adverse factions generates barriers to the identification of a shared Article I mandate precisely because Congress folds in partisan divisions in a way the White House does not. This is no accident: the housing of political contestation is now recognized as a central function of democratic legislatures.252 Hence, even though Congress and the presidency have a similar democratic mandate—channeling political conflict into formal, legal outcomes—differences in how that mandate is configured over time radically influence their expected institutional loyalty.

Once more, it would be a mistake to think that the federal constitution exhausts the range of possible configurations that might elicit strong institutional mandates. Consider the possibility of a single-mission branch. Absent from our national organic document, this idea is pursued with vigor at the state constitutional level. In that context, executive power is often “unbundled” into mandate-specific offices, each of which is subject to separate election.253 A majority of states thus directly elect an attorney general, a lieutenant governor, and a secretary of state, as well as a governor.254 This unbundling of functions into distinct executive bodies will in expectation amplify institutional loyalty: for it is much easier for an official within an unbundled executive to identify and align herself with a singular goal (for example, crime control or the efficient provision of social services) than it is for an official within a bundled executive to assemble an institutional loyalty.

In the administrative-law context, in which both single- and multiple-mandate agencies abound, there is ample anecdotal evidence that single-mandate agencies tend to foster a “dedicated but zealous” culture that is somewhat tone deaf to “the arguments and ideas of policymakers in other agencies as well as in the White House.”255 Hence, a quantitative study of the Environmental Protection Agency (EPA) elicited the view from within the agency that EPA staff tend to be focused “narrowly on environmental interests,” in contrast to the “broader perspective” taken by the White House.256 Even multiple-mandate missions tend to focus on one mission at the expense of others.257 Officials who favor the losing mandate are hardly inclined to hew unreservedly to the institution’s subsequent path. They are thus unlikely to evince institutional loyalty to the same extent.

Finally, an institutional mandate may have a dynamic effect on the selection into and out of the institution. If an institution is renowned for its mandate, not only are officials with a prior commitment to that mandate likely to select in, but also, once embedded in the institution, they are more likely to make nontransferable mandate-specific investments in expertise. This makes exit costlier. For example, if the EPA is known to maintain a commitment to combating climate change, rather than fostering polluting industries, it will be less likely to attract staff who are climate change skeptics.258 And vice versa. Officials hired will then redouble their epistemic investments to make the EPA’s mission (however conceived) a success. The result is a positive-feedback mechanism by which the strength of an institutional mandate increases over time.

In summary, how institutions are sliced up, whether by function or by subject matter, directly shapes the extent to which their officials tend to develop institutional loyalty. Insulation from synchronic partisan conflict, pursuit of an indivisible mission, and the elimination of plural conflicting mandates—all these are likely to conduce to sharper institutional loyalty.

2.   Diluting institutional mandates.

Cultivating an institutional mandate, however, is not costless. The channeling of political conflict into a single legislative forum, for example, is a central element of constitutional design.259 Even if Congress is, as a result of playing this function, unable to muster the same level of institutional loyalty as the executive, its role as a forum for routine partisan contestation is sufficiently important not to be derogated. There are a number of other design elements, however, that fragment institutional mandates and so undermine institutional loyalty in pursuit of other public values. The resulting trade-offs have not yet been identified. We consider the interaction of mandates with the “internal separation of powers”260 to illustrate such conflicts.

Skepticism about the constraining effect of the separation of powers has induced some scholars to advocate alternatively for an “internal” separation of powers.261 The idea of an internal separation of power has an oxymoronic aspect. One prominent commentator suggests that they can be understood as mechanisms that “seek to achieve [the interbranch separation of powers’] goals by operating within the confines of a single branch.”262 Leading examples include the separation of adjudication from rulemaking or prosecutorial functions within administrative agencies, and the creation of “independent” agencies that exercise a measure of policy discretion free of presidential control.263

Whether or not these design elements have independent justifications,264 it seems likely that their installation would dilute branch-level (and sometimes agency-level) institutional loyalty. Functional or policy mission–based distinctions within a branch foster plural, rather than unitary, understandings of a branch’s mandate. The very function of independent agencies, such as the Federal Reserve, is to create acoustic separation between short-term partisan interests and longer-term systemic goals.265 Similarly, “dissent channels,” which allow career bureaucrats to voice frustrations about an administration’s policy and which have also been championed as form of internal separation of powers, might have benefits in defeating groupthink.266 But because they have costs in terms of diluting institutional loyalty at the level of the branch, there is a trade-off between fostering loyalty to the branch as opposed to a subunit, such as an agency.

Consider by way of example the decision of career diplomats within the State Department recently to use a dissent channel to challenge the Obama White House’s limited military deployment in the Syrian conflict.267 Their memorandum aired, and hence likely compounded, “deep rifts and lingering frustration” within the executive branch.268 The mere availability of a dissent channel undercuts pressure toward conformity on a single executive-branch position.269 It hence acts as a friction on the formation of executive-oriented institutional loyalty.

D.   Social Networks

Institutional loyalty is generated when officials are part of “social networks”270 that generate and reinforce commitments to an institution. That social networks influence the behavior and preferences of political officials and citizens is now “well established empirically.”271 This is because “[p]eople frequently think and do what they think and do because of what they think relevant others think and do.”272 This Section renders this intuition in slightly more formal terms, and integrates it with the literature on constitutional design, which has largely ignored the topic.273

The social networks of government officials influence behavior and preferences through at least two pathways. First, networks shape the “epistemic community”274 in which officials operate.275 Social networks help define the “argument pools” to which officials are exposed.276 Information diffuses quickly through social networks. The stronger the network, the faster the information diffuses, and the stronger its influence on argument pools.277 Empirical studies of official behavior thus identify strong “peer effects” by which “bureaucrats’ responses to uncertainty turn less on supervisory instructions and more upon what they perceive peer bureaucrats to be doing.”278

Second, internal social networks generate reputational costs for officials straying from an institutional orthodoxy. Networks reduce monitoring costs. It is cheaper to obtain and evaluate information about close colleagues or friends. These networks have proven important. The leading study of bureaucratic autonomy, by Professor Daniel Carpenter, identified such social networks as a critical tool for professional administrators, who used their alternative power base “in political and social networks” to reduce “their dependence on elected officials.”279 Social networks thus provide a way of collectively amassing and exercising power on something other than an ideological or partisan basis. In the early Republic, Chief Justice Marshall showed a canny awareness of that possibility by insisting that the justices all reside together in a boardinghouse. Marshall aimed “to use the camaraderie of boarding-house life to dispel dissent and achieve a one-voiced Opinion of the Court.”280

With these mechanisms in hand, it is possible to pick out a number of ways in which institutional designers exploit or resist social networks to generate institutional loyalties. To begin with, numerous constitutional rules, conventions, and statutory rules are usefully understood as efforts to cultivate intra-institutional networks and tamp down on cross-institutional networks. The Incompatibility Clause of Article I, for example, states that “[N]o Person holding any Office under the United States, shall be a Member of either House during his Continuance in Office.”281 As Hamilton wrote in Federalist 76, if legislators were serving in the executive branch, there would be a constant source of “executive influence upon the legislative body.”282 Analogously, the Supreme Court has recently implied that the Due Process Clause bars state and federal judges alike from deciding cases under the excessive influence of either legislative- or executive-branch actors.283

Textual anti-networking rules are supplemented by weaker conventions.284 When Justice Abe Fortas was nominated to become chief justice, he was defeated in part because of revelations of his close relationship with President Lyndon B. Johnson.285 This generated an informal convention against the justices being an essential and close part of the network of senior officials within the other branches of government, motivated in part by a concern that such ties would compromise the institutional loyalty of the justices.286 This convention, though, is not as strong as a constitutional rule. Hence, in 2004, when Vice President Cheney went hunting with Justice Scalia even as a case denominated with his name was before the Court, loud objections were raised by commentators287 and by litigants.288 In contrast to Fortas, however, Scalia publicly defended his decision not to recuse himself by pointing out the mundaneness of cross-branch social networking. “Social contacts with high-level executive officials (including cabinet officers),” he explained in a letter to the Los Angeles Times, “have never been thought improper for judges who may have before them cases in which those people are involved in their official capacity, as opposed to their personal capacity.”289 It may well be that this successful defense has eroded the post-Fortas convention somewhat.

Consonant with Scalia’s defense, there are many ways in which the Constitution remains open to interbranch networks. For example, there is no Due Process Clause bar to executive or legislative influence on a coordinate political branch.290 If anything, the Framers seemed more concerned about preventing officials sharing social networks with those in power in state governments than preventing officials sharing social networks with those in other branches of the federal government. The creation of a new national capital beyond the control of state governments, for example, was meant to prevent federal officials from becoming too intertwined with state officials.291

Other constitutional rules generate social networks that reinforce rather than reduce institutional loyalty. Providing the heads of institutions with strong control over their institutions can lead them to surround themselves with members of their social networks.292 Consider again Cheney’s defense of a unitary and powerful executive branch.293 Cheney himself wielded unprecedented control over the operations of the Office of the Vice President, using this power to surround himself with friends and allies.294 This personnel-related authority allowed him to translate ideas diffusing through his external social network about the merits of a powerful executive295 into institutional practice operationalized through an internal network of fellow travelers.296 External social networks, including law professors and other legal experts, helped aggrandize the branch’s authority.

Institutions can also be part of wider professional social networks that cut across the public–private divide, and that police adherence to institutional norms while at the same time providing legitimation and public support. Elements of the “thick political surround” can also feature institution-specific investments that would be endangered by official behavior that defies loyalty. Because officials are networked with these actors, they are exposed to pressure to conform to institutional norms, but also benefit from an extraneous source of political capital.

Such networks are particularly robust in respect to the Article III judiciary; they help explain one of our motivating examples, Roberts’s decisive vote in NFIB.297 The chief justice, and the Court as a whole, is embedded in a larger social network of commentators, think tanks, scholars, and lawyers, largely located inside the Beltway.298 This network has been reinforced by the recent development of a powerful “Supreme Court Bar” comprising many leading national law firms.299 This diffuse network is an important source of criticism, and hence social pressure toward certain sorts of institutional behavior, sometimes epitomized (or caricatured) as the so-called “Greenhouse effect.” This is the alleged phenomenon “in which some Supreme Court Justices have drifted away from the conservatism of their early votes . . . towards the stated preferences of cultural elites, including left-leaning journalists and the ‘liberal legal establishment that dominates at elite law schools.’”300 Although this purported dynamic has a partisan flavor at present (at least in the accounts that circulate in the media), it is also possible that the justices’ integration into this network reinforces professional norms respecting legal craft and precision in ways that are nonideological in character.

This social network also provides an important source of validation and political support for the Court as an institution. The Court derives legitimacy and strength from this network—from the commentators on the left and right who routinely identify the Court as a vital national institution, from the many elite law firms that rely on it for prestige (and even business), and from the media that pay obsessive attention to the justices and their doings. In short, the Court’s social network provides ballast that both strengthens and also roots it to particular institutional practices and norms.

* * *

To summarize, some basic architectural choices about how branches and agencies are set up strongly shape the possibility of institutional loyalty. As we explore further below, institutional loyalty may or may not be desirable. To the extent it is thought desirable, though, the design decisions that conduce to shared institutional mandates should take it into consideration.

III.  Recalibrating Institutional Loyalties

We shift in this Part from description and analysis to a more normative stance. Having isolated the nature and importance of institutional loyalties, and their causal origins, we consider how individual officials’ loyalty within each branch might be usefully recalibrated. Our aim in this Part is to be illustrative, offering suggestions, rather than firm or final diagnoses. We hope to demonstrate that the recalibration of institutional loyalty is a valuable, yet to date underappreciated, mechanism for attaining larger constitutional goals.

A measure of caution is counseled by the background complexity of the separation of powers. That constitutional tradition is animated by multiple normative ambitions, including liberty, democratic accountability, and rule-of-law promotion.301 These goals can and do clash.302 We do not aim to settle the profound normative questions of how to optimize these goals over competing constitutional ends across different contexts here. Nor do we tackle in this Article the difficult questions of which kinds of branch-level reforms are incentive compatible.303 Because our ambition here is to show the conceptual utility of institutional loyalty, we will put to one side process-related questions of how one gets from “here to there.”304 In short, the branch-by-branch illustrations of pathways for institutional reform that we offer below should be taken as a “proof of concept,” not a strict agenda for institutional reform.

Our analysis, moreover, makes no strong assumptions about the motives of an institutional designer. This raises an important criticism: Are we answering the question of how institutional dispositions arise by simply shifting the problem from the level of design to designer? What good does the identification of a toolkit do if adequate motivational foundations for designers are unavailable? There are, in fact, several reasons why institutional designers have good cause to consider how their choices influence officials’ dispositions. Most importantly, most designers want their institutional progeny to succeed. Official dispositions keyed to a particular institution are often necessary to the latter’s successful operation. Explaining the importance of a shared “corporate culture” among managers, Professor David Kreps has pointed out that culture furnishes focal points for the resolution of unexpected contingencies “in the minds of its hierarchical inferiors.”305 Investment in a joint (institutional) project also means that external actors can confidently develop “stable expectations” about institutions’ likely behavior.306

In addressing the utility of reforms that leverage institutional loyalty for each branch in turn, we begin by offering a hypothesis as to whether the branch is appropriately characterized by excessive or insufficient incentives to heed the branch’s interests. We recognize that these baselines are controversial. Our purpose in specifying them, however, is to facilitate an analysis of the feasible design margins that can be recalibrated to generate institutional loyalties. Readers with a different normative prior should therefore attend to the lessons for institutional design, rather than our stipulated baselines. Having set a (provisional) normative baseline, we then work from the causal pathways identified in Part II to recommendations for how the surfeit or deficiency of institutional loyalty respectively might be mitigated.

A.    The Executive Branch’s Incentives Refocused

Today, the executive branch exercises extensive policy authority regardless of the party or the president in power.307 This authority is mediated through entities, such as the National Security Council and the Office of Information and Regulatory Affairs, that are able to cultivate high degrees of institutional loyalty via selective entrance rules, a complex civil-service regime, and the existence of durable career paths wholly within the federal government. The net result is that high-level executive personnel are generally characterized to a far greater degree by institutional loyalty relative to their congressional analogs. Hence, Article II–related loyalties are in no immediate risk of being crowded out entirely by partisan or ideological loyalty. If they are threatened at all, it is by loyalties to more granular units, such as departments and agencies.308

We assume an executive characterized by strong institutional loyalty toward the branch as a whole, and even stronger loyalties to agencies and departments. Given this rough baseline, we consider how institutional loyalties might be refocused toward the branch, rather than the agency or department, level, and, alternatively, how more government-wide preferences might be cultivated. We examine closely the merits of expanding on the rotation system that authorizes executive-branch officials to work outside of their usual institutional setting for a period of time.309

Job responsibilities for all employees, including those in the executive branch, can be sliced and allocated using many different tools. Executive-branch responsibilities can be bundled into permanent parcels (called “portfolios” in Washington speak) that are assigned to particular executive-branch officials. This is the normal practice in the executive branch, in which officials work for many years in the same offices with roughly the same portfolios—with small exceptions.310 Civil servants in the executive branch tend to be promoted within the same agency or department in the executive branch, rather than moving between them. Political appointees tend to go “in and out” of the executive branch, rather than moving between agencies or departments within the executive branch.311 All this entrenches agency-level loyalties.

In contrast, rotations within the executive branch mitigate loyalties that stop at the agency or department door. Federal law now recognizes executive job rotations but imposes numerous limitations on them.312 The most reliable data produced by the federal government has also found a very small number of “detailees,”313 the technical term for those detailed to another part of the executive branch or another branch entirely.314 The alumni of these rotations are not only likely to generate complicated institutional loyalties, but also to be effective in pursuing those complicated loyalties. A number of empirical studies suggest that accrual of “specific human capital” in the form of relationships with other officials predicts “effectiveness in office.”315

Job rotations within and between branches expand loyalties to the branch by leveraging several of the mechanisms identified in Part II. Entrance mechanisms, for example, can select for the detailees most likely to generate and manage loyalties to both their home and detailed institution. Detailees can be assigned to a new institution only if their home institution—usually at the highest level—agrees that a temporary assignment would be useful after the detailee returns.316 The new, temporary institutional employer must also approve the detailee, and sometimes must pay them out of their limited budget.317 These entrance mechanisms are essentially screening tools to identify those capable of the more complicated institutional loyalties that flow from detailing. An alternative approach is to vary employee assignments across time. Temporary task bundling (also known as “job rotation”318 or using a “Type Z organization”319) means assigning officials for short periods of time to distinct tasks.320 These distinct and temporary tasks can be allocated to officials while these officials remain in their current positions.321

Exit mechanisms can also encourage asset-specific investments in the home institution of the detailee, as well as in the potential future or actual present institution to which the official is detailed. Detailees invest in their home institution because they have a position there. Investing in that position will yield continued and potentially improved employment prospects there in the future. The cost of exiting a home institution, though, is mitigated by job rotations. Rotation creates new social networks and employment opportunities, mitigating the costs of exiting a home institution. Once a detail is arranged, moreover, the actual experience of working in a new institutional setting makes the possibility of exit more concrete, and hence less costly.

The social context of the detailee is also transformed by their temporary assignment. There is a strong relationship between those with whom officials work and those who are the strongest nodes in their social networks.322 After a civil servant works for years for an executive-branch agency or department, these connections within their agency or department can become even stronger. Political appointees, for instance, receive their executive-branch positions because of their ties to other political elites who have been, are, or will be political appointees.323

Rotations mean that networks are expanded to include those outside of their home institution. Upon returning to their home institutions, detailees are likely to have internalized many of the perspectives of the officials they added to their network from their temporary institutional homes. Scholars of employees in the public sector have often found that “employees adapt their behavior consistent with the norms and expectations of people around them,”324 in “profound” and persistent ways.325 Going forward, newly networked officials become regular parts of the personal and professional network of the detailed official, generating loyalties to their temporary institutional home even after an official returns home.

Job rotations within the executive branch can therefore be a tool used by those arguing for the need for a more holistic Article II perspective. The OLC, for instance, has been criticized by scholars like Professor Bruce Ackerman for being unduly loyal to the power of the president.326 OLC lawyers tend to move to other positions outside of the executive branch, such as becoming law professors, or move into the White House.327 An OLC lawyer who spent months working for the equally elite Office of the Legal Advisor in the Department of State might for instance be less skeptical of international-law constraints on presidential power.328

Reformers could also use rotations to dilute an Article II orientation. Executive-branch officials can be temporarily detailed to another executive-branch agency or department, or even to the legislative or judicial branch. Job rotations from the executive branch to the other branches in particular sensitize the executive branch to the legislative and judicial branches. The executive branch dominates the federal government, with more than 2.6 million employees.329 It should not be surprising that it is quite common in Washington to find executive-branch officials sharing social networks with entirely or almost entirely other executive-branch officials330—they are the overwhelming majority of officials living in Washington, after all. If officials are working on more, and more varied, issues, they are more likely to build more heterogeneous social networks that incorporate Congress, the federal courts, and even the private sector. These wider social networks may deflate past institutional loyalties not only because they change officials’ sources for information, but also because they serve as pathways to alternative, nonexecutive positions.

There is evidence that earlier professional experiences in another branch of government generate loyalties to that branch that endure even after an official has moved on to another branch. Empirical studies have found, for instance, that federal judges are more likely to rule in favor of the executive branch if they have served in the executive branch earlier in their careers.331 Courts have also noted the complex loyalties held by those who have moved offices, characterizing such personnel as a blend of employees of their old institutions, employees of their new institutions, and independent contractors.332 There is every reason to believe that a recent job rotation would have roughly the same—or greater—loyalty-generating power as a professional experience decades earlier would have. The OLC lawyer who spends several months working for the Senate Judiciary Committee would have a different sense of presidential power than the OLC lawyer who never left the executive branch.333

Job rotations across branches, though, are rare. When they do happen, detailee traffic tends to be from the legislature (and sometimes the courts) to the executive—and rarely the other way around. Federal law authorizes detailing of executive-branch officials to House or Senate committees, but does not do the same for the personal staffs of members of the House or Senate.334 The relatively formidable power of the modern executive branch means that the attraction of working in the executive branch is substantial. The polarized and despised Congress has a harder time attracting talented staffers.335 Job rotations across branches right now, therefore, might well serve to spread the gospel about the importance of executive-branch loyalty, rather than tempering it.336

B.    Judicial Ambition and the Rule of Law

Institutional loyalties within the federal judiciary might be supposed an unfettered good insofar as they conduce toward “judicial independence” from the Congress and the executive branch.337 The implicit assumption of this view is that by negating the influence of coordinate branches, structural protections of judicial independence enable judges to exercise their independent judgment about what law (and law alone) requires. But this need not be so. Over the course of the twentieth century, the federal judiciary developed an unprecedented institutional heft and has successfully secured “important authority over key jurisdictional and administrative powers.”338 In a vivid display of an institutional loyalty at work early in the twentieth century, Chief Justice William Howard Taft extensively lobbied Congress on behalf of the federal courts and secured to the Supreme Court almost unfettered discretion over its caseload, near-plenary authority to set its own agenda, and freedom to determine how and why it would intervene on matters of national salience.339

Yet it is far from clear that the fruits of the discretion achieved by Taft and other advocates for the institutional judiciary necessarily promote useful constitutional ends. As Professor Pamela Karlan succinctly explains, “Judges should be independent, not so much so that they can conceive goals and policies of their own and realize them, but so that they can enforce the goals and policies embodied in the Constitution and the laws enacted by the democratic branches.”340 But, as one of us has argued, bipartisan coalitions of the Supreme Court have narrowed dramatically the range of constitutional remedies available to criminal defendants, prisoners challenging their convictions via habeas corpus, and ordinary citizens engaged in retail encounters with the state.341 The same trend can be observed in the context of statutory civil-rights remedies.342 That these restrictive transformations have a “long and bipartisan pedigree”343 is suggestive of their institutional roots: they reflect the judiciary’s institutional interest in stanching the flow of certain kinds of litigation.344

As a consequence of institutional loyalties, individual rights’ holders are no longer able to vindicate entitlements through ex post remedies in federal court.345 Viewed in this light, institutional loyalties in the Article III context are, paradoxically, inconsistent with promotion of the constitutional rule of law and at odds with conventional notions of corrective justice and deterrence.

What might be done, then, to mitigate forms of judicial self-regard with unwelcome implications for constitutional rights’ holders? Setting aside the pressing question of how such reform might occur, we develop one suggestion here in some detail.

Since 1886, Congress has fashioned separate adjudicative mechanisms “for the bringing of suits against the Government of the United States.”346 In respect to remedies that can be granted through a freestanding adjudicative process—rather than, say, as incidental proceedings embedded within a criminal prosecution—it would be desirable to create a specialized court staffed by an institutionally distinct cadre of judges. The advantage of such a bespoke cadre of judges is the fostering of a discrete and separate institutional mandate from that of Article III. Legislators might leverage the focal-point effect provided by a new jurisdictional statute, assigning cases to a discrete pool of judges separately and distinctly charged with the vindication of constitutional and civil rights.

Congress might model this new judicial arm on the bankruptcy bench, although a new “remedies bench” should have plenary Article III protection. The bankruptcy courts have successfully exhibited “Article III values” as a consequence of a social-network effect—a “continued connection to an audience[,] the bankruptcy bar”—that “holds in high esteem professional, creative, and non-ideological resolution of complex disputes.”347 To leverage the same sort of social-network effect, Congress might employ an entrance rule—a requirement that appointees to this have litigated in the past on behalf of a constitutional right or a statutory civil right. In contrast to the current federal courts, which are dominated by former federal prosecutors,348 the resulting bench would come to public service already aligned to the institutional goal of providing remedies to individuals harmed by the government’s unconstitutional or unlawful actions. It would also be entangled from the start in social networks in which the social value of individualized remediation for constitutional wrongs is well understood.

We are, of course, under no illusions that Congress as currently constituted is about to expend effort on behalf of a dispersed and hardly high-status population of constitutional rights’ holders (although we are also of the view that legal scholars should not self-censor based on present political realities).349 Our point here is rather to demonstrate that the causal mechanisms we have identified in Part II—entrance and exit rules, institutional mandates, and social networks—have a continuing role to play in refining our federal institutions toward public-regarding constitutional ends.

C.    Congress “Redivivus”?

It is common ground that Congress today is characterized by a “radical separation between the two major political parties.”350 Regardless of whether ideological bifurcation within Congress reflects changes in public preferences,351 there is a measure of consensus today that a polarized Congress does not, and cannot, serve the nation well.352 Similarly, there is a widespread view that lobbyists possess too much influence in Congress, with many believing that special-interest bribery is just “[t]he way things work in Congress.”353

These two popular diagnoses are not wholly consonant with one another.354 Nonetheless, they suggest broad agreement on one point: legislators are excessively motivated by partisan and ideological355 interests, perhaps as well as by pure pecuniary ones. Concomitantly, institutional loyalty seems in short supply on Capitol Hill. Scholars since Professor John Hart Ely have doubted Congress’s ability “to pull up its socks and reclaim its rightful authority” against the executive branch.356 Our analysis in this Section stipulates, as a baseline matter, an insufficiently institutionally loyal legislative branch. How might the repertoire of design modifications canvassed in Part II inform efforts to reform a Congress of that sort? We develop a potential avenue of reform that employs many of the mechanisms we identified in Part II, although we are under no illusions about the intractability of the task.

We focus on congressional staff, not legislators themselves. The legislature’s entrance and exit rules for such staff might be reformed to create an executive-branch-style congressional civil service durably disposed toward the institution of Congress.357 Legal scholars focus on the optimal distribution of civil servants versus political appointees in the executive branch. The parallel question in the legislative-branch context receives scant attention. Scholars like Professors Abbe Gluck and Lisa Bressman have started to fill this void by offering thick descriptions of congressional staffers’ beliefs and behavior.358 We build here on Gluck and Bressman’s impressive work by considering ways to strengthen staffers’ institutional loyalty.

Constitutional doctrine already recognizes that congressional staff matter. As the Supreme Court has recognized, senators’ and representatives’ aides in particular possess a measure of authority that rivals, and perhaps sometimes even surpasses, that of executive-branch officials.359 Elected officials in Congress spend large portions of their time on election-related tasks.360 They spend much of their time outside of Washington and hence away from the daily legislative activities in Congress.361 They hence depend on staffers, who often act as their agents during intrabranch negotiations362 and as their conduit to constituents and lobbyists.363 Because there are comparatively few staffers in Congress, each staffer can wield substantial power.364 The increasing complexity of policy, moreover, means that legislators as principals can imperfectly monitor the myriad choices staffers as agents are constantly making.

Empowerment, however, does not translate into an institutional orientation. Congressional staffers are usually divided into one of two categories. There are “personal staff” who work for the representative or senator directly, and “committee staff” (often called “professional staff”)365 who work for the House or Senate committee. To begin with, the selection and promotion rules for staffers—including so-called professional staffers—do not conduce currently to positive loyalty toward Congress as an institution.

Neither competitive examinations nor objective merit-related criteria are employed in hiring most congressional staffers akin to those used in the civil service.366 Aides instead are hired by representatives and senators based on their loyalty toward their immediate boss, their political party, and their ideology. Often, staffers are evaluated on the basis of past performance in the quite different context of a congressional campaign.367 The number of professional staff has been declining substantially over the years.368 Professional staff are usually hired through partisan networks anyway, with few exceptions. And while executive-branch civil servants are promoted based on criteria favoring the more institutionally disposed,369 most congressional staffers do not last long enough to be meaningfully promoted.370 An important exception, we note, is the Congressional Budget Office (CBO), one of the products of the 1974 framework budget legislation,371 and an institution expressly modeled on the executive branch’s operation.372

Further, exit-related dynamics do not foster meaningful institutional loyalties to Congress. There are no intrinsic rewards to identifying with Congress as an institution. Congress remains much maligned by public opinion.373 Material returns to congressional service are also low. Congressional staffers are poorly paid, work long hours, and usually can be removed at any time—most importantly when their elected boss retires or loses reelection.374

Instead, staffers have rich exit options in both the executive branch and the private sector. For those driven by partisan or ideological goals, there are enormous returns to going to work for the executive, especially in comparison to the contemporary Congress.375 An executive-branch position may allow the former staffer to help draft regulations or litigate path-making cases. Both may carry more policy heft, and more intrinsic appeal, than legislative trench warfare. Private-sector lobbying positions, in contrast, offer not only the potential for influence, but also substantially higher salaries. Staffers routinely double their salary by defecting to the private sector.376 For staffers whose former bosses remain in office, lobbying work is especially lucrative.377 This creates an incentive for earlier rather than later exit from public service.

Compounding the institutional-loyalty deficit, congressional staffers do not organize around the singular institutional mandates that generate loyal officials in some agencies and departments in the executive branch.378 Congress has a broad and diffuse portfolio of responsibilities, not the focused portfolio of a single-mission institution.379 Members of Congress do not usually divide portfolios within their offices along discrete policy lines. Personal staff will cover a range of political and policy portfolios, and given the shortage of staff members, those dedicated to policy matters will usually have multiple and diverse policy issues in their portfolios.380

Committee staff are unified by their nominal commitment to the mandate of their committee. Committee mandates, though, are (perhaps purposefully) broad.381 They might perceive themselves as part of the policy community that is within their committee’s jurisdiction—a lawyer for the Senate Judiciary Committee, for instance, feels part of the legal community. This is limited, though, because committee staffers that are hired and fired by majority or minority members will feel that their “mandate” is ultimately to serve whatever their member says, rather than to serve or shape a discrete policy mandate or professional community.

Stronger institutional loyalty among congressional staff might be cultivated in several different ways. At the entry stage, the democratic accountability of members of Congress means that significant numbers of staffers are and should be hired by the members that will employ them. A larger number of positions, though, can be filled by professional staff working for committees than presently is the case. Committees do not feature staffers with sufficient institutional loyalties to power the institutional autonomy of Congress, but committee staffers are still at the greatest remove of any current staffers from the sway of elected officials.382 Civil-service examinations akin to those required for many executive-branch positions can be required to join committee staff.

Alternatively, the relevant professional community can offer its evaluation of the professional competence of the potential staffer. For instance, the American Bar Association plays an important role evaluating the credentials of judicial nominees.383 It might also evaluate the credentials of those applying for a position on the Senate Judiciary Committee. Committee staffers might be approved only by a vote of either the other staffers and/or the members of the committee. This could lead to the selection of committee staffers on a partisan basis, just like other nominees that come to a committee vote. However, inducing partisan warfare over committee staff generates an opportunity cost that prevents committees from dealing with more substantial issues engulfed by partisan warfare, such as judicial nominees. The more time spent debating the next committee counsel for the Senate Judiciary Committee, the less time available for debating the next nominee to the US Court of Appeals for the District of Columbia Circuit.

Modifying exit incentives for staffers would also help generate institutional loyalty. Stronger returns to remaining in Congress can also make a difference for congressional staffers. Congress enjoys the power of the purse, but opens that purse more to similarly situated executive-branch officials than to their congressional counterparts. A lawyer working for the Securities and Exchange Commission (SEC)—even if that lawyer is not in a leadership position at the SEC—can make over $200,000 a year.384 The senior Democratic lawyer on the Senate Judiciary Committee, a veteran of several decades, made roughly 80 percent of that working on the nomination of then-Judge Neil Gorsuch to succeed Justice Scalia and determine the future of the Supreme Court.385

Exit-related constraints on postemployment lobbying in the private sector, in particular in relation to issues on which a former staffer worked, would also shift the expected payoffs of exit versus continued service in Congress.386 The draw of working for the executive branch can be reduced by constraints on post­employment opportunities in the other branches of government. Rather than staffers contemplating what will make them an appealing candidate to a K Street lobbying firm or a presidential candidate or president, constraints on these opportunities upon leaving Congress will direct their attention toward the legislative branch.

This reimagined congressional staffing model would also generate more of a focused institutional mandate. Staffers hired to pursue a discrete policy agenda for a committee would think of themselves as part of a legislative epistemic community. They would be evaluated and approved by their fellow committee members. They would also be evaluated and approved by related professional communities. Rather than considering what serves the electoral interests of the Chair of the Senate Judiciary Committee, the staffer can consider what serves the interests of the rule of law and Congress’s role in defending it.

Finally, the institutional dispositions of legislators and their staff might be promoted by narrowing the available scope for acting on pecuniary motives. Strikingly, the Supreme Court has recently expanded the scope for legislative self-serving. In its 2016 McDonnell decision, the unanimous Court narrowed the scope of the “honest services” statute,387 which criminalizes fraudulent deprivations of the “intangible right of honest services.”388 Honest services charges are common in public-corruption cases.389 The McDonnell Court held that the statute applied only when a defendant official “formal[ly] exercise[d] governmental power . . . similar in nature to a lawsuit before a court, a determination before an agency, or a hearing before a committee.”390 In part, the Court justified this narrowing gloss with the concern that a broad public-corruption law would “chill” interactions between constituents and their elected representatives that ought to be protected by the First Amendment.391 McDonnell undermines efforts to cultivate institutional dispositions because it widens the domain of activity in which representatives can act on pecuniary, rather than institutional or public-regarding, reasons.392 In contrast, representatives who are insulated from strong external pulls have a better chance to focus on institutional concerns. Our analysis thus provides a new ground for critiquing the McDonnell decision and favoring a broad public-corruption prohibition.393


Every organization struggles with how to ensure that its members act on behalf of its collective or corporate interests. The molding of atomized, selfish individual actors into committed institutional loyalists is therefore a problem of organizational design across academic fields and across our public life. In no domain of organizational life is this problem more important to consider, and more difficult to solve perhaps, than in the context of federal governmental design. When Madison anticipated the “interest of the man must be connected with the constitutional rights of the place,”394 therefore, his aspiration was not distinctive or different from that of many institutional entrepreneurs.

Nevertheless, scholars have struggled to visualize how a Constitution devised before the existence of—and without regard to—cohesive and polarized political forces and figures could continue to operate. The ensuing modern position is pessimistic. It implies that when the massive power of the federal government is in play, partisan, ideological, or selfishly material motives will dominate. Institutional concerns will trail in their wake. Our constitutional system works as it was supposed to only when what really motivates officials coincidentally overlaps with constitutional principles. When partisan, ideological, or even crudely materialistic interests conflict with constitutional principles, damage is done to these principles. At a singular, extreme moment, or with gradual deterioration, this could mean the disappearance of these ancient and valued principles.

We think this modern position can be supplemented. Branches, no less than private associations, can and do cultivate loyal, well-disposed officials capably motivated to act on the institution’s behalf. These “constitutional rights of the place” must be consciously cultivated by careful institutional design, and not simply assumed (or assumed away).395 To that end, we have identified and taxonomized four mechanisms capable of nurturing institutional loyalty. Properly deployed, these tools of institutional loyalty have the potential to play a meaningful role in ensuring that our national institutions operate as more than a blind crashing together of conflicting partisan forces.

Madison was right that “parchment barriers”396 will not protect us when our “dependence on the people”397 produces the forces or figures that the modern position fears and that threaten institutional boundaries. Our Constitution is not a “machine that would go of itself.”398 But Madison was also right that we are not without hope when ideological or partisan passions overwhelm fragile institutional boundaries.399 It is at those moments that our system includes within it other tools to ensure that institutional boundaries persist and persevere. The ambition of this Article has been to demonstrate the importance and utility of institutional loyalty as one of these tools to ensure the constitutional machine still works even in moments of stress.

  • 1. See, for example, Humphrey’s Executor v United States, 295 US 602, 629 (1935) (“The fundamental necessity of maintaining each of the three general departments of government entirely free from the control or coercive influence, direct or indirect, of either of the others, has often been stressed and is hardly open to serious question.”).
  • 2. Buckley v Valeo, 424 US 1, 122 (1976) (per curiam). See also Commodity Futures Trading Commission v Schor, 478 US 833, 860 (1986) (Brennan dissenting) (“In order to prevent [ ] tyranny, the Framers devised a governmental structure composed of three distinct branches.”).
  • 3. Federalist 51 (Madison), in The Federalist 347, 349 (Wesleyan 1961) (Jacob E. Cooke, ed) (“[T]he great security against a gradual concentration of the several powers in the same department, consists in giving to those who administer each department, the necessary constitutional means, and personal motives, to resist encroachments of the others.”).
  • 4. See id at 348 (“[S]eparate and distinct exercise of the different powers of government . . . is admitted on all hands to be essential to the preservation of liberty.”). See also Bond v United States, 564 US 211, 222 (2011) (“[S]eparation-of-powers principles . . . protect each branch of government from incursion by the others . . . [and] protect the individual.”); Boumediene v Bush, 553 US 723, 742 (2008) (“The Framers’ inherent distrust of governmental power was the driving force behind the constitutional plan that allocated powers among three independent branches. This design serves not only to make Government accountable but also to secure individual liberty.”). For a similar recent statement to the same effect, see National Labor Relations Board v Noel Canning, 134 S Ct 2550, 2559 (2014) (“We recognize, of course, that the separation of powers can serve to safeguard individual liberty.”).
  • 5. Federalist 49 (Madison), in The Federalist 338, 343 (cited in note 3) (“The passions . . . of the public, would sit in judgment. . . . [M]ere declarations in the written constitution, are not sufficient to restrain the several departments. . . . [O]ccasional appeals to the people would be neither a proper nor an effectual provision, for that purpose.”).
  • 6. See  Gillian E. Metzger, Ordinary Administrative Law as Constitutional Law, 110 Colum L Rev 479, 498 (2010) (lauding such separations as ways of enhancing legality and rights); Neal Kumar Katyal, Internal Separation of Powers: Checking Today’s Most Dangerous Branch from Within, 115 Yale L J 2314, 2317 (2006).
  • 7. Daryl J. Levinson and Richard H. Pildes, Separation of Parties, Not Powers, 119 Harv L Rev 2311, 2323 (2006).
  • 8. Daryl J. Levinson, Empire-Building Government in Constitutional Law, 118 Harv L Rev 915, 920 (2005). See also Jack Goldsmith and Daryl Levinson, Law for States: International Law, Constitutional Law, Public Law, 122 Harv L Rev 1791, 1832 (2009) (“Not all of the structural and political mechanisms Madison envisioned have worked in the ways he anticipated or hoped.”); Daryl J. Levinson, Parchment and Politics: The Positive Puzzle of Constitutional Commitment, 124 Harv L Rev 657, 670 (2011) (“Madison never explained why the branches of government, or the state and federal governments, would reliably have political incentives at odds with one another—why they would tend to compete rather than cooperate or collude.”). For endorsements of this view by other scholars, see Kate Andrias, Separations of Wealth: Inequality and the Erosion of Checks and Balances, 18 U Pa J Const L 419, 430 (2015) (“[A] system intended to channel competition through the political branches actually channels it through the political parties.”); Bruce Ackerman, The Living Constitution, 120 Harv L Rev 1737, 1809 n 222 (2007) (describing the parties-not-powers theory as a “breakthrough” and an “essential reference point”).
  • 9. See, for example, Curtis A. Bradley and Trevor W. Morrison, Historical Gloss and the Separation of Powers, 126 Harv L Rev 411, 443 (2012) (“[T]he Madisonian model of interbranch rivalry is especially inaccurate during times of unified government.”); Levinson and Pildes, 119 Harv L Rev at 2329 (cited in note 7) (“[W]hen government is unified and the engine of party competition is removed from the internal structure of government, we should expect interbranch competition to dissipate. Intraparty cooperation (as a strategy of interparty competition) smoothes over branch boundaries and suppresses the central dynamic assumed in the Madisonian model.”).
  • 10. The scholars that have remained loyal to institutional loyalty are therefore often left playing defense against the skeptical modern position that institutional loyalties are relics of our constitutional past. Compare Josh Chafetz, Congress’s Constitution: Legislative Authority and the Separation of Powers 28–35 (Yale 2017) (criticizing Levinson and Pildes on the ground that “party discipline is by no means absolute in the American system”), with Jon D. Michaels, Of Constitutional Custodians and Regulatory Rivals: An Account of the Old and New Separation of Powers, 91 NYU L Rev 227, 227 (2016) (“[L]eading accounts . . . fail to capture the multidimensional nature of administrative control in which the constitutional branches (the old separation of powers) and the administrative rivals (the new separation of powers) all compete with one another to influence administrative governance.”). But see Josh Chafetz, Congress’s Constitution, 160 U Pa L Rev 715, 715, 774 (2012) (“Congress has significantly more constitutional power than we are accustomed to seeing it exercise. . . . [A] possible explanation for congressional underutilization of its powers is that members of Congress are largely unconcerned with congressional power; their primary loyalty is to their party, not their branch.”). Our aim here is to systematize these hints.
  • 11. Jan Crawford, Roberts Switched Views to Uphold Health Care Law (CBS News, July 2, 2012), online at
    -views-to-uphold-health-care-law/2/ (visited Oct 19, 2017) (Perma archive unavailable).
  • 12. Id.
  • 13. See Jan Crawford, Discord at Supreme Court Is Deep, and Personal (CBS News, July 9, 2012), archived at (noting that Roberts broke with those who had assumed he would be an ideological ally).
  • 14. See Preserving Prosecutorial Independence: Is the Department of Justice Politicizing the Hiring and Firing of U.S. Attorneys?, Hearings before the Senate Committee on the Judiciary, 110th Cong, 1st Sess 216–17 (2007) (statement of James B. Comey, former deputy attorney general, Department of Justice).
  • 15. See Charlie Savage, 2 Top Lawyers Lost to Obama in Libya War Policy Debate (NY Times, June 17, 2011), archived at
  • 16. Jack Goldsmith, Lawyerly Integrity in the Trump Administration (Lawfare, May 14, 2017), archived at
  • 17. 60 Stat 812.
  • 18. Jonathan G. Pray, Congressional Reporting Requirements: Testing the Limits of the Oversight Power, 76 U Colo L Rev 297, 305 (2005).
  • 19. Roger H. Davidson, The Advent of the Modern Congress: The Legislative Reorganization Act of 1946, 15 Legis Stud Q 357, 365 (1990).
  • 20. Id.
  • 21. See id at 360. See also Karla W. Simon, Congress and Taxes: A Separation of Powers Analysis, 45 U Miami L Rev 1005, 1027 n 81 (1991) (“The Legislative Reorganization Act of 1946 formalized the oversight function of Congress.”).
  • 22. Eric Schickler, Disjointed Pluralism: Institutional Innovation and the Development of the U.S. Congress 8, 14 (Princeton 2001).
  • 23. There are other examples in which institutional interests and partisan motives clearly align. Consider, for example, the Senate’s refusal to hold hearings on Obama’s Supreme Court nominee, Judge Merrick Garland. Carl Hulse, Supreme Court Showdown Could Shape Fall Elections (NY Times, Mar 16, 2016), online at
    2016/03/17/us/politics/supreme-court-nomination-obama-congress.html (visited Oct 19, 2017) (Perma archive unavailable). We largely discard these examples because they do not provide unambiguous evidence of institutional loyalty at work.
  • 24. See Levinson and Pildes, 119 Harv L Rev at 2318 (cited in note 7) (“Madison’s will-based theory of separation of powers would seem to require government officials who care more about the intrinsic interests of their departments than their personal interests or the interests of the citizens they represent. Democratic politics is unlikely to generate such officials.”).
  • 25. Adrian Vermeule, Selection Effects in Constitutional Law, 91 Va L Rev 953, 953 (2005) (using the term and noting that such effects flow from “the question of which (potential) officials are selected to occupy those posts over time”).
  • 26. Constitutions are typically shaped by a range of goals, including the creation of channels for peaceful political contestation, the enabling of public-good creation, the fostering of legitimacy (democratic or otherwise), and the minimizing of agency costs. See Tom Ginsburg and Aziz Z. Huq, Assessing Constitutional Performance, in Tom Ginsburg and Aziz Z. Huq, eds, Assessing Constitutional Performance 3, 14–23 (Cambridge 2016) (setting out four criteria for the evaluation of a constitution’s success: legitimacy, channeling political conflict, limiting agency costs, and creating public goods). The same sort of ends-related pluralism characterizes the separation of powers. See Aziz Z. Huq and Jon D. Michaels, The Cycles of Separation-of-Powers Jurisprudence, 126 Yale L J 346, 382–91 (2016).
  • 27. See Daniel Carpenter and George A. Krause, Transactional Authority and Bureaucratic Politics, 25 J Pub Admin Rsrch & Theory 5, 13 (2014) (discussing the relationship between organizational socialization and institutional loyalty). See also John Brehm and Scott Gates, Working, Shirking, and Sabotage: Bureaucratic Response to a Democratic Public 3 (Michigan 1999) (summarizing historical research that shows the influence of “a bureaucrat’s own preferences, peers, supervisors, and clients” on work decisions).
  • 28. Huq and Michaels, 126 Yale L J at 391 (cited in note 26).
  • 29. Robert A. Dahl, Polyarchy: Participation and Opposition 8 (Yale 1971) (“[P]olyarchies are regimes that have been substantially popularized and liberalized, that is, highly inclusive and extensively open to public contestation.”).
  • 30. See Aziz Z. Huq, Does the Logic of Collective Action Explain Federalism Doctrine?, 66 Stan L Rev 217, 280–88 (2014) (analyzing difficulties states’ officials have in advancing state interests in Congress, and suggesting generalizations about when that might occur).
  • 31. Erin Ryan, Negotiating Federalism, 52 BC L Rev 1, 31 (2011) (“State and federal actors also negotiate over enforcement policy and individual enforcement actions arising within cooperative federalism programs.”).
  • 32. See Miriam Seifter, States as Interest Groups in the Administrative Process, 100 Va L Rev 953, 961, 970 (2014) (arguing that “state interest groups’ advocacy efforts were initiated to create a voice for states qua states—a voice for the institutional interests of state governments rather than the varied political preferences of state constituents or individual state officials”—and noting that “[t]o the extent federal law has directed agencies to engage states in federal decision making, it has done so largely by giving state interest groups a central role”).
  • 33. See, for example, Reply Brief of State Petitioners on Medicaid, Florida, South Carolina, Nebraska, Texas, Utah, Louisiana, Alabama, Colorado, Pennsylvania, Washington, Idaho, South Dakota, Indiana, North Dakota, Mississippi, Arizona, Nevada, Georgia, Alaska, Ohio, Kansas, Wyoming, Wisconsin, and Maine; Schuette; Branstad v United States Department of Health and Human Services, Docket No 11-400, *11 (US filed Mar 12, 2012) (available on Westlaw at 2012 WL 864598) (challenging the Affordable Care Act’s requirement that states expand their Medicaid programs as unconstitutionally coercive).
  • 34. For more extended consideration of the interaction of populism with constitutional democracy in the US context, see Aziz Huq and Tom Ginsburg, How to Lose a Constitutional Democracy, 65 UCLA L Rev *19 (forthcoming 2018), archived at
    -8ZAH; Aziz Z. Huq, Book Review, The People against the Constitution, 116 Mich L Rev *5 (forthcoming 2018), archived at
  • 35. Mancur Olson, The Logic of Collective Action: Public Goods and the Theory of Groups 165–67 (Harvard 1971) (stating that large institutions face difficulties in convincing individual members to act in the collective institutional interest).
  • 36. The empirical literature refuting or at least complicating Olson’s simplified claim is extensive. For a few of the best summaries of this literature, see Ernst Fehr and Simon Gächter, Fairness and Retaliation: The Economics of Reciprocity, 14 J Econ Persp 159, 162–63 (Summer 2000); Dan M. Kahan, Reciprocity, Collective Action, and Community Policing, 90 Cal L Rev 1513, 1516 (2002).
  • 37. Karen Orren and Stephen Skowronek, The Search for American Political Development 82–83 (Cambridge 2004). Our reading of the rational-choice tradition in legal scholarship is that there is a focus generally on individual costs and benefits, although sophisticated theorists are clear that they understand individuals to have other preferences and values. See, for example, Guido Calabresi, The Future of Law and Economics: Essays in Reform and Recollection 136–41 (Yale 2016) (describing this phenomenon and noting in particular that “economists have sometimes ignored the desire for altruism and beneficence that in fact many people have”). We aim here to take one subset of such goals and values seriously on their own terms.
  • 38. See, for example, Micah Schwartzman, Judicial Sincerity, 94 Va L Rev 987, 992 (2008) (defining sincerity in a similar context as “correspondence between what people say, what they intend to say, and what they believe”).
  • 39. See Arthur Isak Applbaum, Ethics for Adversaries: The Morality of Roles in Public and Professional Life 63 (Princeton 1999) (describing role morality as the “particular moral reasons for action that others, outside the [institutional] role, do not face”); W. Bradley Wendel, Book Review, Professional Roles and Moral Agency, 89 Georgetown L J 667, 673 (2001) (discussing role morality in terms of “the power of roles to permit an agent to deliberate on the basis of a restricted set of reasons—leaving out, for example, considerations of harm to third parties”).
  • 40. Levinson, 118 Harv L Rev at 917 (cited in note 8) (defining “empire-building” as governmental behavior that seeks to maximize power and/or wealth “at the expense of competing government bodies—and, ultimately, at the expense of the citizenry”).
  • 41. See text accompanying notes 11–16.
  • 42. See Levinson, 118 Harv L Rev at 928 (cited in note 8).
  • 43. Federalist 51 at 349 (cited in note 3).
  • 44. See Herbert J. Storing, What the Anti-Federalists Were For 54 (Chicago 1981).
  • 45. Id at 60 (quoting an Antifederalist tract by the “Officer of the Late Continental Army”). See also Cecilia M. Kenyon, Men of Little Faith: The Anti-Federalists on the Nature of Representative Government, 12 Wm & Mary Q 3, 23 (1955) (describing the Antifederalists’ demand for a “more rigid” separation of powers).
  • 46. Centinel, To the Freemen of Pennsylvania, in Herbert J. Storing, ed, 2 The Complete Anti-Federalist, 136, 138 (Chicago 1981). See also Kenyon, 12 Wm & Mary Q at 23 (cited in note 45) (discussing Antifederalist demand for “more effective checks and balances”).
  • 47. Federalist 49 at 341 (cited in note 5). Because our analysis is an effort to clarify the argument offered by Madison, we do not offer any larger claim about belief in institutional loyalty among the Founding generation.
  • 48. Federalist 51 at 349 (cited in note 3).
  • 49. Federalist 49 at 338–39 (cited in note 5) (quotation marks omitted):

    One of the precautions . . . as a palladium to the weaker departments of power, against the invasions of the stronger, is . . . that whenever any two of the three branches of government shall concur in opinion, each by the voices of two thirds of their whole number, that a convention is necessary for altering the constitution, or correcting breaches of it, a convention shall be called for the purpose.

  • 50. Id at 340.
  • 51. Id at 342–43.
  • 52. For a helpful discussion of this problem of institutional clarity in constitutional design, see G. Bingham Powell Jr, Elections as Instruments of Democracy: Majoritarian and Proportional Visions 61–64 (Yale 2000).
  • 53. Federalist 49 at 339 (cited in note 5).
  • 54. Id at 339–40 (noting that the direct appeal to the people would not always be a means of resolving conflict).
  • 55. Federalist 51 at 349 (cited in note 3).
  • 56. Id.
  • 57. Id.
  • 58. See Adam Smith, An Inquiry into the Nature and Causes of the Wealth of Nations 16 (Chicago 1976) (Edwin Cannan, ed) (originally published 1776):

    But man has almost constant occasion for the help of his brethren, and it is in vain for him to expect it from their benevolence only. He will be more likely to prevail if he can interest their self-love in his favour, and shew them that it is for their own advantage to do for him what he requires of them.

    For Madison’s reading of Smith, see Samuel Fleischacker, Adam Smith’s Reception among the American Founders, 1776–1790, 59 Wm & Mary Q 897, 905–15 (2002); David Prindle, The Invisible Hand of James Madison, 15 Const Polit Economy 223, 231–34 (2004).

  • 59. Federalist 51 at 349 (cited in note 3).
  • 60. Federalist 49 at 343 (cited in note 5) (“The passions ought to be controuled and regulated by the government.”).
  • 61. Federalist 51 at 349 (cited in note 3) (emphasis added).
  • 62. Id.
  • 63. Id at 349, 351.
  • 64. Id at 349. As an aside, we think Madison missed a complication here: if officials change course only when another institution resists, their reliance on others’ policing efforts might crowd out any internalization of the Constitution’s structural norms. See Bruno S. Frey, A Constitution for Knaves Crowds Out Civic Virtues, 107 Econ J 1043, 1044–45 (1997). See also Adrian Vermeule, Hume’s Second-Best Constitutionalism, 70 U Chi L Rev 421, 424 (2003) (“[T]he self-interest assumption may crowd out public-spirited motivations.”). The possibility of a feedback effect of this sort suggests a need to consider how the design of interbranch interactions dynamically influences the incentives and beliefs of those within the branches.
  • 65. Federalist 66 (Hamilton), in The Federalist 445, 451 (cited in note 3).
  • 66. Federalist 46 (Madison), in The Federalist, 315, 317–19 (cited in note 3).
  • 67. Id at 319.
  • 68. See Federalist 78 (Hamilton), in The Federalist 521, 526–30 (cited in note 3).
  • 69. Id at 529.
  • 70. Daniel W. Howe, The Political Psychology of The Federalist, 44 Wm & Mary Q 485, 500 (1987).
  • 71. Federalist 68 (Hamilton), in The Federalist 457, 460–61 (cited in note 3).
  • 72. Federalist 57 (Madison), in The Federalist 384, 384 (cited in note 3).
  • 73. Pub L No 105-277, 112 Stat 2681 (1998), codified at 5 USC § 3345 et seq.
  • 74. 137 S Ct 929 (2017).
  • 75. Id at 936.
  • 76. United States House of Representatives v Burwell, 130 F Supp 3d 53, 72 (DDC 2015) (authorizing suit by the House to challenge violations of the Appropriations Clause), quoting Kucinich v Bush, 236 F Supp 2d 1, 7 (DDC 2002). The Supreme Court has recognized an analog institutional interest in state legislatures. See Arizona State Legislature v Arizona Independent Redistricting Commission, 135 S Ct 2652, 2664 (2015) (authorizing a challenge under Article I to an independent redistricting commission on the ground that “[t]he Arizona Legislature . . . is an institutional plaintiff asserting an institutional injury”).
  • 77. Of course, the actual decision to initiate a lawsuit may be better explained on partisan grounds. See text accompanying notes 145–46.
  • 78. Part of the reaction of modern scholars to institutional loyalty is to question whether there has ever been a clearly articulated theoretical account of the mechanisms—if any—that lead to institutional loyalty. See, for example, Levinson and Pildes, 119 Harv L Rev at 2317 (cited in note 7) (“[I]t has never been clear exactly how the Madisonian machine [of competitive branches] was supposed to operate.”); M. Elizabeth Magill, The Real Separation in Separation of Powers Law, 86 Va L Rev 1127, 1158 (2000) (“Just how tension and competition [between the branches] are created and maintained is never clearly spelled out by courts or commentators.”). The modern position has been to criticize the absence of mechanisms for institutional loyalty just as much as the accuracy of any mechanisms that could be offered.
  • 79. See, for example, Levinson and Pildes, 119 Harv L Rev at 2323 (cited in note 7) (“[E]lectoral and policy interests of politicians have become intimately connected to political parties.”). See also Jacob E. Gersen, Unbundled Powers, 96 Va L Rev 301, 330–31 (2010) (“Individuals have party loyalties and personal interests, however, which may systematically diverge from institutional interests.”).
  • 80. See, for example, Neal Devins and David E. Lewis, Not-So Independent Agencies: Party Polarization and the Limits of Institutional Design, 88 BU L Rev 459, 490 (2008) (noting that cross-party commission appointees are loyal to their party, not the president who appointed them).
  • 81. See, for example, Jessica Bulman-Pozen, Executive Federalism Comes to America, 102 Va L Rev 953, 975 (2016) (“The partisan genealogy of executive federalism . . . is striking . . . . [The] national party system generates ties among state and federal actors.”). See also generally Jessica Bulman-Pozen, Partisan Federalism, 127 Harv L Rev 1077 (2014) (exploring how partisanship has transformed federalism and interactions between governments). For other scholarship that discusses the dynamic between partisan and institutional motives, see, for example, Eric A. Posner and Adrian Vermeule, Constitutional Showdowns, 156 U Pa L Rev 991, 1036 (2008); Mark Tushnet, Constitutional Hardball, 37 John Marshall L Rev 523, 529–30 (2004).
  • 82. Alice Ristroph, Is Law? Constitutional Crisis and Existential Anxiety, 25 Const Commen 431, 434 (2009).
  • 83. See, for example, Bruce Ackerman, The Failure of the Founding Fathers: Jefferson, Marshall, and the Rise of Presidential Democracy 16–26 (Harvard 2005); Bruce Ackerman and David Fontana, Thomas Jefferson Counts Himself into the Presidency, 90 Va L Rev 551, 557–67 (2004).
  • 84. The precise form of ideology generating the ideological loyalty can be disputed. The various forms of ideological loyalty all return to some comprehensive worldview regarding the proper behavior of government motivating official behaviors. See David Fontana and Donald Braman, Judicial Backlash or Just Backlash? Evidence from a National Experiment, 112 Colum L Rev 731, 749–51 (2012).
  • 85. See Thomas J. Miles and Cass R. Sunstein, The New Legal Realism, 75 U Chi L Rev 831, 836 & n 22 (2008) (“To date, the question that has received the most attention from the New Legal Realists is the influence of a judge’s political ideology or attitudes.”).
  • 86. Jeffrey A. Segal and Harold J. Spaeth, The Supreme Court and the Attitudinal Model 65 (Cambridge 1993). This is the classic statement of attitudinalism, one that has been modified and expanded over the past generation. An exception to this view is Judge Richard Posner’s influential 1993 article, which emphasizes leisure and income but not ideological “power trip[s].” Richard A. Posner, What Do Judges and Justices Maximize? (The Same Thing Everybody Else Does), 3 S Ct Econ Rev 1, 3, 31 (1993). Neither in this nor in later work does Posner underscore institutional loyalty.
  • 87. See, for example, McNollgast, Politics and the Courts: A Positive Theory of Judicial Doctrine and the Rule of Law, 68 S Cal L Rev 1631, 1634 (1995) (arguing that accounting for strategic context yields “a better understanding of when and how partisan politics will affect judicial doctrine”).
  • 88. Id at 1637 (stipulating that judges have exogenously given preferences over policy). See also Lee Epstein and Tonja Jacobi, The Strategic Analysis of Judicial Decisions, 6 Ann Rev L & Soc Sci 341, 343 (2010) (“Strategic accounts, in contrast, belong to a class of nonparametric rational-choice models, as they assume that goal-directed actors—
    including judges—operate in a strategic or interdependent decision-making context.”).
  • 89. Eric A. Posner and Adrian Vermeule, The Credible Executive, 74 U Chi L Rev 865, 900–02 (2007).
  • 90. Jacob E. Gersen and Adrian Vermeule, Delegating to Enemies, 112 Colum L Rev 2193, 2196 (2012).
  • 91. Dwight D. Eisenhower, President’s News Conference (American Presidency Project, Sept 30, 1953), archived at Some commentators argue that the Warren appointment was the result of a political deal. See Ed Lazarus, The Truth about Justices (Time, May 26, 2009), online at
    packages/article/0,28804,1900851_1900850_1900845,00.html (visited Oct 19, 2017) (Perma archive unavailable).
  • 92. Bernard Schwartz and Stephan Lesher, Inside the Warren Court: 1953–1969 92 (Doubleday 1983) (noting that Eisenhower called the Warren appointment the “biggest damn-fool mistake I ever made”).
  • 93. In past decades, ideological loyalty and partisan loyalty have increasingly overlapped because of sorting by ideologically homogenous political parties. For empirical evidence, see Morris P. Fiorina, Samuel J. Abrams, and Jeremy C. Pope, Culture War?: The Myth of a Polarized America 61–63, 78 (Pearson 3d ed 2011); Matthew Levendusky, The Partisan Sort: How Liberals Became Democrats and Conservatives Became Republicans 126–28, 133 (Chicago 2009).
  • 94. One of us, however, has argued that the association of executive-branch primacy and national security success is premised on erroneous factual premises. Aziz Z. Huq, Structural Constitutionalism as Counterterrorism, 100 Cal L Rev 887, 906–09 (2012) (describing parochial interests of security agencies like the CIA).
  • 95. See, for example, Posner, 3 S Ct Econ Rev at 31 (cited in note 86).
  • 96. United States v Kemp, 500 F3d 257, 279–81 (3d Cir 2007) (holding that bribery in the honest services context requires “a specific intent to give or receive something of value in exchange for an official act”) (emphasis omitted), citing United States v Sun-Diamond Growers of California, 526 US 398, 404–05 (1999).
  • 97. 136 S Ct 2355 (2016).
  • 98. Id at 2364–65 (construing narrowly the honest services statute in political-corruption cases).
  • 99. Federalist 37 (Madison), in The Federalist 231, 235 (cited in note 3).
  • 100. See Martin S. Flaherty, The Most Dangerous Branch, 105 Yale L J 1725, 1735–36 (1996); Magill, 86 Va L Rev at 1158 (cited in note 78); Victoria Nourse, The Vertical Separation of Powers, 49 Duke L J 749, 754 & n 21 (1999). See also Gary Lawson, The Rise and Rise of the Administrative State, 107 Harv L Rev 1231, 1238 n 45 (1994) (“The problem of distinguishing the three functions of government has long been, and continues to be, one of the most intractable puzzles in constitutional law.”).
  • 101. Levinson and Pildes, 119 Harv L Rev at 2324 (cited in note 7).
  • 102. For empirical evidence that the public is convinced by public-regarding explanations, such as institutional loyalty, rather than appeals to narrow and personal self-interest, see Eitan D. Hersh and Brian F. Schaffner, Targeted Campaign Appeals and the Value of Ambiguity, 75 J Polit 520, 532 (2013).
  • 103. See United States v Texas, 136 S Ct 2271 (2016) (per curiam).
  • 104. Speaker Ryan on United States v. Texas Decision (Speaker Ryan Press Office, June 23, 2016), archived at
  • 105. See text accompanying notes 11–22.
  • 106. David J. Barron and Martin S. Lederman, The Commander in Chief at the Lowest Ebb—Framing the Problem, Doctrine, and Original Understanding, 121 Harv L Rev 689, 706–12 (2008).
  • 107. Id at 715.
  • 108. Id at 713–14.
  • 109. For a discussion of the origins and evolution of the Cheney view about executive power, see Scott Shane, Recent Flexing of Presidential Powers Had Personal Roots in Ford White House (NY Times, Dec 30, 2006), online at
    12/30/washington/30roots.html (visited Oct 19, 2017) (Perma archive unavailable).
  • 110. Jack Goldsmith, The Terror Presidency: Law and Judgment inside the Bush Administration 140 (Norton 2007).
  • 111. Id.
  • 112. Report of the Congressional Committees Investigating the Iran-Contra Affair, S Rep No 100-216, HR Rep No 100-433, 100th Cong, 1st Sess 438 (1987).
  • 113. See, for example, Barton Gellman, Angler: The Cheney Vice Presidency 99–102 (Penguin 2008); Frederick A.O. Schwarz Jr and Aziz Z. Huq, Unchecked and Unbalanced: Presidential Power in a Time of Terror 187–200 (New Press 2007); Jane Mayer, The Hidden Power (New Yorker, July 3, 2006), archived at
  • 114. Dick Cheney and Liz Cheney, In My Time: A Political and Personal Memoir 92, 130 (Threshold 2011).
  • 115. David McCullough, John Adams 485–86, 539, 560 (Simon & Schuster 2001).
  • 116. Charles M. Ellis, Roger B. Taney and the Leviathan of Slavery (The Atlantic, Feb 1865), archived at
  • 117. Alison Mitchell, Ashcroft Debate Shows Deep Rifts (NY Times, Feb 1, 2001), online at (visited Oct 19, 2017) (Perma archive unavailable).
  • 118. James Risen and Michael S. Schmidt, 2004 Showdown Shaped Reputation of Pick for F.B.I. (NY Times, June 21, 2013), online at
    obama-says-comey-will-strike-balance-on-security-and-privacy-at-fbi.html (visited Oct 19, 2017) (Perma archive unavailable).
  • 119. Pub L No 93-148, 87 Stat 555 (1973), codified at 50 USC § 1541 et seq.
  • 120. See Savage, 2 Top Lawyers Lost (cited in note 15).
  • 121. Id.
  • 122. See Eric Konigsberg, In Clinton’s Backyard, It’s Open Season as an Obama Fund-Raiser Lines Up Donors (NY Times, Feb 24, 2007), online at (visited Oct 19, 2017) (Perma archive unavailable). One of us has previously explored the effects of such preferences among executive-branch actors on the scope of executive power in these pages. See generally Aziz Z. Huq, Book Review, Binding the Executive (by Law or by Politics), 79 U Chi L Rev 777 (2012).
  • 123. President Obama Announces Key Additions to the Office of the White House Counsel (White House Office of the Press Secretary, Jan 28, 2009), archived at
  • 124. See Manu Raju and Josh Gerstein, Senate Confirms New CIA Lawyer (Politico, Mar 13, 2014), archived at
  • 125. See Savage, 2 Top Lawyers Lost (cited in note 15).
  • 126. See Curtis A. Bradley and Trevor W. Morrison, Presidential Power, Historical Practice, and Legal Constraint, 113 Colum L Rev 1097, 1133 (2013) (noting that executive-branch “lawyers are socialized in an ethos associated with the American polity and the American style of law and government” and that “OLC has developed a range of practices and traditions—including a strong norm of adhering to its own precedents even across administrations”).
  • 127. Pub L No 111-148, 124 Stat 119 (2010).
  • 128. 567 US 519 (2012).
  • 129. Bradley Joondeph, A Marbury for Our Time (SCOTUSblog, June 29, 2012), archived at See also Daniel Epps, In Health Care Ruling, Roberts Steals a Move from John Marshall’s Playbook (The Atlantic, June 28, 2012), archived at
  • 130. See Adam Liptak, Chief Justice John Roberts Amasses a Conservative Record, and Wrath from the Right (NY Times, Sept 28, 2015), online at (visited Oct 19, 2017) (Perma archive unavailable).
  • 131. Linda Greenhouse, Blaming the Chief Justice (NY Times, May 12, 2016), online at (visited Oct 19, 2017) (Perma archive unavailable) (collecting sources).
  • 132. Ariane de Vogue, Down a Justice, John Roberts Looks to Find Compromise, Avoid 4–4 Ties (CNN, May 10, 2016), archived at
  • 133. 136 S Ct 1557 (2016).
  • 134. Id at 1559.
  • 135. 134 S Ct 2751 (2014).
  • 136. Id at 2759. See also Adam Liptak, Justices Seem Split in Case on Birth Control Mandate (NY Times, Mar 23, 2016), online at (visited Oct 19, 2017) (Perma archive unavailable) (noting how Zubik “builds on” Hobby Lobby).
  • 137. See, for example, Liptak, Justices Seem Split (cited in note 136).
  • 138. Ariane de Vogue, How McConnell Won, and Obama Lost, the Merrick Garland Fight (CNN, Nov 9, 2016), archived at
  • 139. Adam Liptak, Supreme Court Hints at Way to Avert Tie on Birth Control Mandate (NY Times, Mar 29, 2016), online at (visited Oct 19, 2017) (Perma archive unavailable) (“The Supreme Court . . . issued an unusual order indicating that the justices are trying to avoid a 4-to-4 deadlock in a case pitting religious freedom against access to contraception.”).
  • 140. Zubik v Burwell, 2016 WL 1203818, *2 (US) (“The parties are directed to file supplemental briefs that address whether and how contraceptive coverage may be obtained by petitioners’ employees through petitioners’ insurance companies, but in a way that does not require any involvement of petitioners beyond their own decision to provide health insurance without contraceptive coverage to their employees.”).
  • 141. Zubik, 136 S Ct at 1560 (“[T]he parties on remand should be afforded an opportunity to arrive at an approach going forward that accommodates petitioners’ religious exercise while . . . ensuring that women . . . receive full and equal health coverage.”) (quotation marks omitted).
  • 142. See Adam Liptak, Justices, Seeking Compromise, Return Contraception Case to Lower Courts (NY Times, May 16, 2016), online at
    2016/05/17/us/supreme-court-contraception-religious-groups.html?mcubz=0 (visited Oct 19, 2017) (Perma archive unavailable) (stating that the Court “is exploring every avenue to avoid 4-to-4 deadlocks, even if it does not decide the question the justices have agreed to address”).
  • 143. See Jesse Byrnes, Kagan: Eight-Judge Supreme Court “Working Really Hard” to Avoid Deadlocks (The Hill, Apr 5, 2016), archived at (quoting Kagan as stating that “[a]ll of us are working hard to reach agreement” and all are “especially concerned” about reaching consensus).
  • 144. David R. Mayhew, The Imprint of Congress 98 (Yale 2017) (emphasis omitted).
  • 145. See United States House of Representatives v Burwell, 130 F Supp 3d 53, 81 (DDC 2015) (“The House of Representatives has standing to pursue its allegations that the Secretaries of Health and Human Services and of the Treasury violated Article I, § 9, cl. 7 of the Constitution when they spent public monies that were not appropriated by the Congress.”); United States House of Representatives v Burwell, 185 F Supp 3d 165, 168 (DDC 2016) (finding a violation of the Appropriations Clause in the same suit).
  • 146. In a similar fashion, many of the public “actions” that Professor David Mayhew famously charted in his landmark history can be explained by individual-level prestige. See David R. Mayhew, America’s Congress: Actions in the Public Sphere, James Madison through Newt Gingrich 37 (Yale 2000). At the same time, Mayhew emphasizes that Congress “has absorbed a large share of the country’s most talented and ambitious politicians” into its “opportunity structure.” Id at 218. An implication of this observation is that the future career prospects of many pivotal national figures turn on the institutional fortunes of the legislative branch—thus entangling the institutional and the personal.
  • 147. Schickler, Disjointed Pluralism at 12–13 (cited in note 22) (noting the presence of “multiple interests” behind many institutional changes).
  • 148. Id at 14.
  • 149. Id at 146–50 (explaining why this was not a “cartelistic” move by the majority).
  • 150. Id at 141–42, citing S 2177, 79th Cong, 2d Sess (May 13, 1946), in 92 Cong Rec 6558 (June 10, 1946) (statement of Sen Bridges).
  • 151. Schickler, Disjointed Pluralism at 142–43 (cited in note 22).
  • 152. Id at 256.
  • 153. Id at 257.
  • 154. Id at 150–51.
  • 155. Schickler, Disjointed Pluralism at 152–53 (cited in note 22). Indeed, Schickler finds that the Joint Committee was the “only formal, floor-approved change [to Congress’s processes] that is explained by a single interest,” that is, in congressional capacity. Id at 249.
  • 156. Id at 257. See also Mayhew, The Imprint of Congress at 115 (cited in note 144) (“Congress can also put its feet down, taking authority back or saying to the White House that enough is enough.”).
  • 157. Chafetz, Congress’s Constitution at 31, 63 (cited in note 10).
  • 158. Thomas D. Morgan, The General Accounting Office: One Hope for Congress to Regain Parity of Power with the President, 51 NC L Rev 1279, 1281 (1973), quoting 66th Cong, 1st Sess (Oct 17, 1919), in 58 Cong Rec 7085 (remarks of Rep Good).
  • 159. For details of the Congressional Budget Office’s operation, see Philip G. Joyce, The Congressional Budget Office: Honest Numbers, Power, and Policymaking 224 (Georgetown 2011); Walter J. Oleszek, et al, Congressional Procedures and the Policy Process 66–76 (CQ 10th ed 2016).
  • 160. Schickler, Disjointed Pluralism at 8 (cited in note 22).
  • 161. Pub L No 93-344, 88 Stat 297, codified at 2 USC § 601 et seq.
  • 162. 2 USC § 601 et seq. See also Kate Stith, Rewriting the Fiscal Constitution: The Case of Gramm-Rudman-Hollings, 76 Cal L Rev 593, 615–17 (1988) (discussing circumstances in which the 1974 Budget Act was passed).
  • 163. Stith, 76 Cal L Rev at 617 (cited in note 162).
  • 164. See Daniel Diermeier and Roger B. Myerson, Bicameralism and Its Consequences for the Internal Organization of Legislatures, 89 Am Econ Rev 1182, 1188–92 (1999). We are grateful to Brian Feinstein for discussion on this point.
  • 165. 60 Stat 237 (1946), codified as amended in various sections of Title 5.
  • 166. Cass R. Sunstein and Adrian Vermeule, Libertarian Administrative Law, 82 U Chi L Rev 393, 465–66 (2015).
  • 167. See Elizabeth Garrett, The Purposes of Framework Legislation, 14 J Contemp Legal Issues 717, 718 (2005) (“Framework legislation creates rules that structure congressional lawmaking; these laws establish internal procedures that will shape legislative deliberation and voting with respect to certain laws or decisions in the future. They are laws about the congressional lawmaking process itself.”).
  • 168. See Neal Devins, Why Congress Does Not Challenge Judicial Supremacy, 58 Wm & Mary L Rev 1495, 1501 (2017) (“The early Congresses were vigorous defenders of legislative prerogatives.”).
  • 169. See Louis Fisher and Neal Devins, Political Dynamics of Constitutional Law 25–28 (West 5th ed 2011) (discussing Congress’s express disapproval of the Marbury v Madison litigation, its resulting cancellation of the Supreme Court’s 1802 term, and its refusal to honor a Court order to turn over documents concerning the Marbury appointment).
  • 170. See, for example, Neal Devins, The Constitutional Politics of Congress, in Mark Tushnet, Mark A. Graber, and Sanford Levinson, eds, The Oxford Handbook of the U.S. Constitution 155, 170–75 (Oxford 2015).
  • 171. For information on political polarization among political elites, see Richard H. Pildes, Why the Center Does Not Hold: The Causes of Hyperpolarized Democracy in America, 99 Cal L Rev 273, 287–97 (2011). For information on party discipline over time, see James M. Snyder Jr and Tim Groseclose, Estimating Party Influence in Congressional Roll-Call Voting, 44 Am J Pol Sci 193, 198–99 (2000) (Figure 1).
  • 172. See Neal Devins, Party Polarization and Congressional Committee Consideration of Constitutional Questions, 105 Nw U L Rev 737, 762–68 (2011).
  • 173. Many legal rules “produce feedback effects that, over time, bring new types of government officials into power.” Vermeule, 91 Va L Rev at 953 (cited in note 25). We are concerned with a narrower class of selection-related rules.
  • 174. Pierre Bourdieu, The Logic of Practice 53 (Polity 1990) (Richard Nice, trans). Pierre Bourdieu called this a “habitus.” Id (emphasis omitted). Our analysis is consistent with Bourdieu’s frame although we do not employ that relatively unfamiliar terminology.
  • 175. David A. Strauss, The Irrelevance of Constitutional Amendments, 114 Harv L Rev 1457, 1459–60 (2001).
  • 176. See Vermeule, 91 Va L Rev at 956–57 (cited in note 25).
  • 177. Michael Spence, Signaling in Retrospect and the Informational Structure of Markets, 92 Am Econ Rev 434, 437 (2002) (emphasis added). See also Michael Spence, Job Market Signaling, 87 Q J Econ 355, 361–68 (1973) (presenting the original model).
  • 178. Federalist 51 at 349 (cited in note 3).
  • 179. Federalist 10 (Madison), in The Federalist 56, 62–63 (cited in note 3).
  • 180. See text accompanying notes 168–70 (discussing the lack of institutional loyalty in today’s Congress when compared with other branches).
  • 181. Federalist 57 at 386 (cited in note 72).
  • 182. Federalist 62 (Madison), in The Federalist 415, 416–17 (cited in note 3).
  • 183. Id at 419.
  • 184. Federalist 78 at 522–23 (cited in note 68).
  • 185. See Federalist 76 (Hamilton), in The Federalist 509, 513 (cited in note 3) (arguing that the president will not be tempted by “private inclinations and interests” or “a spirit of favoritism” because of the specter of reputational harms from a Senate rejection of a nominee).
  • 186. Joanne B. Freeman, Affairs of Honor: National Politics in the New Republic 37 (Yale 2001) (describing a “fundamental contradiction of republican politics[:] When both personal reputations and political careers rested on popular approval, what was the distinction between public-minded lawmaking and demagoguery?”).
  • 187. John H. Langbein, The German Advantage in Civil Procedure, 52 U Chi L Rev 823, 849–53 (1985).
  • 188. Id at 848.
  • 189. During the second half of the twentieth century, the dominant Liberal Democratic Party in Japan used its control over the judiciary’s administrative apparatus to promote partisan ends. J. Mark Ramseyer, The Puzzling (In)Dependence of Courts: A Comparative Approach, 23 J Legal Stud 721, 725 (1994).
  • 190. Rafael di Tella and Raymond Fisman, Are Politicians Really Paid like Bureaucrats?, 47 J L & Econ 477, 478 (2004).
  • 191. 22 Stat 403 (1883).
  • 192. The Pendleton Act’s competitive exam structure, however, was not mandatory. See Pendleton Act § 2, 22 Stat at 403–04.
  • 193. See 5 USC § 2301(b) (setting forth merit-related criteria and distinguishing impermissible criteria). See also Civil Service Reform Act of 1978 § 3(2), Pub L No 95-454, 92 Stat 1111, 1111–12.
  • 194. Congressional Budget Office, Comparing the Compensation of Federal and Private-Sector Employees *6–7 (Jan 2012), archived at
  • 195. High-powered incentives flow from market transactions, in which efficiency gains flow to the transacting parties. Oliver E. Williamson, The Economic Institutions of Capitalism: Firms, Markets, Relational Contracting 90–91 (Free Press 1985).
  • 196. See David E. Lewis, The Politics of Presidential Appointments: Political Control and Bureaucratic Performance 30 (Princeton 2008) (noting that civil servants “often feel bound by legal, moral, or professional norms to certain courses of action and these courses of action may be at variance with the president’s agenda”); M. Todd Henderson and Frederick Tung, Pay for Regulator Performance, 85 S Cal L Rev 1003, 1007 (2012) (“In the absence of high-powered incentives, it is assumed that those individuals who self-select into regulatory jobs will value public service and will do the work of aligning performance with desired social welfare outcomes.”). Professors M. Todd Henderson and Frederick Tung raise the possibility that regulators may also simply prefer leisure to the harder work of the private sector but give no reasons for thinking this is a more significant dynamic in the public sector than in the private sector. See Henderson and Tung, 85 S Cal L Rev at 1007 (cited in note 196).
  • 197. Ronald N. Johnson and Gary D. Libecap, Courts, a Protected Bureaucracy, and Reinventing Government, 37 Ariz L Rev 791, 822 (1995).
  • 198. 53 Stat 1147 (1939).
  • 199. Civil Service Commission v National Association of Letter Carriers, 413 US 548, 566 (1973) (discussing the Hatch Act § 9(a), 53 Stat at 1148).
  • 200. See text accompanying notes 96–98.
  • 201. Nicholas R. Parrillo, Against the Profit Motive: The Salary Revolution in American Government, 1780–1940 1–4 (Yale 2013) (tracing the process by which American lawmakers went from paying public officials by fees for service and bounties to paying them fixed salaries).
  • 202. See Burke D. Grandjean, History and Career in a Bureaucratic Labor Market, 86 Am J Sociology 1057, 1058–59 (1981) (discussing how “occupational prestige” influences classification of bureaucratic positions for salary purposes).
  • 203. See US Government Accountability Office, Recent Trends in Federal Civilian Employment and Compensation *17 (Jan 2014), archived at
  • 204. See James L. Perry, Bringing Society In: Toward a Theory of Public-Service Motivation, 10 J Pub Admin Rsrch & Theory 471, 479–83 (2000).
  • 205. See Vermeule, 91 Va L Rev at 955–64 (cited in note 25) (identifying mechanisms whereby longitudinal effects can generate new officials with power).
  • 206. See Anthony M. Bertelli and David E. Lewis, Policy Influence, Agency-Specific Expertise, and Exit in the Federal Service, 23 J Pub Admin Rsrch & Theory 223, 235–39 (2012); Sean Gailmard, Politics, Principal-Agent Problems, and Public Service Motivation, 13 Intl Pub Mgmt J 35, 40–42 (2010); Sean Gailmard and John W. Patty, Slackers and Zealots: Civil Service, Policy Discretion, and Bureaucratic Expertise, 51 Am J Polit Sci 873, 880–86 (2007).
  • 207. See Carl Shapiro and Joseph E. Stiglitz, Equilibrium Unemployment as a Worker Discipline Device, 74 Am Econ Rev 433, 435–41 (1984).
  • 208. For empirical studies demonstrating a motivation to shape policy by federal officials, see generally Lewis, The Politics of Presidential Appointments (cited in note 196); Anthony M. Bertelli, Determinants of Bureaucratic Turnover Intention: Evidence from the Department of the Treasury, 17 J Pub Admin Rsrch & Theory 235 (2007).
  • 209. Jeffrey Toobin, No More Mr. Nice Guy (New Yorker, May 25, 2009), archived at
  • 210. See Jonathan S. Masur and Eric A. Posner, Regulation, Unemployment, and Cost-Benefit Analysis, 98 Va L Rev 579, 623, 625–26 (2012) (discussing efficiency wages).
  • 211. US Const Art III, § 1, cl 2 (“The Judges . . . shall hold their Offices during good Behaviour, and shall . . . receive for their Services, a Compensation, which shall not be diminished during their Continuance in Office.”).
  • 212. United States v Will, 449 US 200, 217, 220–21 (1980) (stating that salary protections and tenure “ensure[ ] a [ ] judge . . . the compensation” that will overcome the temptations of the “more lucrative” world of “private practice”).
  • 213. Id at 218.
  • 214. See Paul C. Light, The Tides of Reform: Making Government Work, 1945–1995 60–65 (Yale 1997).
  • 215. See National Commission on Terrorist Attacks upon the United States, The 9/11 Commission Report 411–15 (2004), archived at
  • 216. Intelligence Reform and Terrorism Prevention Act of 2004 § 1011(a), Pub L No 108-458, 118 Stat 3638, 3648–50, codified at 50 USC § 3024(f).
  • 217. Anne Joseph O’Connell, The Architecture of Smart Intelligence: Structuring and Overseeing Agencies in the Post-9/11 World, 94 Cal L Rev 1655, 1732–33 (2006).
  • 218. See, for example, Huq, 100 Cal L Rev at 908 (cited in note 94).
  • 219. David E. Lewis, Presidents and the Politics of Agency Design: Political Insulation in the United States Government Bureaucracy, 1946–1997 157–58 (Stanford 2003) (finding that insulated agencies are substantially more durable than noninsulated agencies). Lewis includes location outside the cabinet, specific qualifications for appointment, commission structure, and fixed terms as measures of political insulation. Id at 62–63.
  • 220. The relationship between presidential control, in particular through the removal power, and White House influence is, however, contingent and uncertain. See Aziz Z. Huq, Removal as a Political Question, 65 Stan L Rev 1, 24–52 (2013).
  • 221. See Bertelli and Lewis, 23 J Pub Admin Rsrch & Theory at 224 (cited in note 206) (“Our evidence suggests that acquiring [federal] agency-specific human capital—knowledge and skills that are nontransferable to employers outside the agency—drives down turnover intention.”).
  • 222. Pub L No 101-194, 103 Stat 1716.
  • 223. Ethics Reform Act § 101, 103 Stat at 1719, 18 USC § 207(e).
  • 224. Honest Leadership and Open Government Act of 2007 § 101, Pub L No 110-81, 121 Stat 735, 737–38. See also Richard L. Hasen, Lobbying, Rent-Seeking, and the Constitution, 64 Stan L Rev 191, 200–03 (2012).
  • 225. See Anita S. Krishnakumar, Towards a Madisonian, Interest-Group-Based, Approach to Lobbying Regulation, 58 Ala L Rev 513, 532 n 86 (2007).
  • 226. Executive Order 13490 § 2 (2010), 3 CFR 193, 198.
  • 227. The White House, Lobbyists on Agency Boards and Commissions, 75 Fed Reg 35955, 35955 (2010).
  • 228. Id.
  • 229. See Bertelli and Lewis, 23 J Pub Admin Rsrch & Theory at 231 & n 11 (cited in note 206) (noting that officials in the Department of Homeland Security and the National Labor Relations Board self-reported the highest amount of institution-specific capital, and officials in the Federal Trade Commission and the Department of Education self-reported the lowest amount).
  • 230. Sean Gailmard and John W. Patty, Learning while Governing: Expertise and Accountability in the Executive Branch 130–31 (Chicago 2013).
  • 231. Id at 33.
  • 232. Gailmard and Patty, 51 Am J Polit Sci at 881 (cited in note 206).
  • 233. See David Fontana, Executive Branch Legalisms, 126 Harv L Rev F 21, 40–41 (2012).
  • 234. See text accompanying notes 270–73.
  • 235. See Harold Hongju Koh, The State Department Legal Adviser’s Office: Eight Decades in Peace and War, 100 Georgetown L J 1747, 1749 (2012) (“[T]he heart of L has been the dedicated career lawyers. . . . One measure of the relative importance of these two sets of positions is that . . . the longest-serving Legal Adviser [ ] served fifteen years . . . [b]ut many career attorneys . . . have served in L for years longer than that.”). See also Bruce Ackerman, The Decline and Fall of the American Republic 95–110 (Belknap 2010) (discussing the OLC’s incentives).
  • 236. Fontana, 126 Harv L Rev F at 40–41 (cited in note 233).
  • 237. David B. Spence, Administrative Law and Agency Policy-Making: Rethinking the Positive Theory of Political Control, 14 Yale J Reg 407, 431 (1997) (“[P]olitical scientists have long noted as political appointees ‘grow’ into the role of agency head and acquire additional expertise in the relevant policy arena, they often adopt the preferences and perspectives of agency careerists on policy issues.”).
  • 238. See, for example, Cornelius M. Kerwin and Scott R. Furlong, Rulemaking: How Government Agencies Write Law and Make Policy 129–64 (CQ 4th ed 2011) (discussing complex and resource-intensive processes for creating and managing internal agency rulemaking).
  • 239. E. Donald Elliott, TQM-ing OMB: Or Why Regulatory Review under Executive Order 12,291 Works Poorly and What President Clinton Should Do about It, 57 L & Contemp Probs 167, 176 (Spring 1994) (“[I]n most administrations, after a few years, the OMB and White House ‘managers’ generally come to hold in contempt their erstwhile colleagues in the agencies, believing that they have ‘gone native’ and adopted the characteristic values of their agencies.”).
  • 240. See Vermeule, 91 Va L Rev at 953 (cited in note 25) (highlighting treatment effects or “incentive-based” effects as those focused “on the creation of optimal incentives for those who happen to occupy official posts at any given time”).
  • 241. See Daniel P. Carpenter, The Forging of Bureaucratic Autonomy: Reputations, Networks, and Policy Innovation in Executive Agencies, 1862–1928 367 (Princeton 2001) (“[B]ureaucratic power grew not at the expense of democratic participation but in a symbiotic relationship with it. Precisely because federal agencies shrewdly orchestrated the participatory energies of American politics, they cleared a unique policymaking place for themselves in the American institutional order, at least for a time.”).
  • 242. For an explanation of how focal points influence behavior absent sanctions, see Richard H. McAdams, Beyond the Prisoners’ Dilemma: Coordination, Game Theory, and Law, 82 S Cal L Rev 209, 233–34 (2009) (suggesting that a focal point can provide a salient means for “individuals [to] coordinate in a particular way, and thereby create self-fulfilling expectations that the recommended behavior will occur”).
  • 243. Tiberiu Dragu and Mattias Polborn, The Administrative Foundation of the Rule of Law, 75 J Polit 1038, 1047 (2013).
  • 244. Id at 1040–44.
  • 245. J.R. DeShazo and Jody Freeman, Public Agencies as Lobbyists, 105 Colum L Rev 2217, 2219 (2005) (noting the possibility of “substantive or procedural” mandates; the latter “impose obligations to consider additional factors, perform a particular analysis, or consult with specific players”).
  • 246. Id at 2220 (“Congress can create the potential for interstatutory conflicts where the agency must balance multiple and potentially competing obligations arising from different statutes usually passed at different times [b]y different enacting majorities.”).
  • 247. Eric Biber, Too Many Things to Do: How to Deal with the Dysfunctions of Multiple-Goal Agencies, 33 Harv Envir L Rev 1, 7 (2009).
  • 248. Another factor that might sharpen the felt pull of an institutional mandate is the degree to which an entity is the sole actor tasked with pursuing a policy goal. Congress, for example, can task more than one agency with a policy goal as a way of stimulating interagency competition. See William A. Niskanen, Bureaucrats and Politicians, 18 J L & Econ 617, 636–38 (1975). But it is far from clear whether a plurality of mandates will in practice be successful in generating increased competition (and hence increased internalization of institutional aims), or whether it will induce a slackening of effort, if officials believe that their role in the achievement of policy is no longer critical. See, for example, Aziz Z. Huq, Forum Choice for Terrorism Suspects, 61 Duke L J 1415, 1462 (2012) (discussing this moral hazard problem in the national security context).
  • 249. Steven G. Calabresi and Kevin H. Rhodes, The Structural Constitution: Unitary Executive, Plural Judiciary, 105 Harv L Rev 1153, 1179 (1992).
  • 250. For objections to the centrality of the Constitution’s text more generally, see Strauss, 114 Harv L Rev at 1464–67 (cited in note 175).
  • 251. Well, not always. See Thomas O. McGarity, Still Free to Harm: A Response to Professor Farber, 92 Tex L Rev 1629, 1629 (2014) (speaking of the “104th (‘Gingrich’) Congress”). This nomenclature, though, speaks to the extraordinary influence of then-Speaker Newt Gingrich, as well as the extent of his ambitions at the time. It is also noteworthy that we refer seamlessly to the “Roberts Court” or the “Rehnquist Court,” despite the “relative insignificance of the office” of chief justice. Ronald J. Krotoszynski Jr, The Unitary Executive and the Plural Judiciary: On the Potential Virtues of Decentralized Judicial Power, 89 Notre Dame L Rev 1021, 1024 (2014).
  • 252. Professor Jeremy Waldron has recently underscored the “institutional responsibility” of an opposition party to “oppose, to scrutinize the government, to hold them accountable for their decisions.” Jeremy Waldron, Political Political Theory: Essays on Institutions 101 (Harvard 2016). Many constitutions outside the United States formalize this role through what are termed “government in opposition rules.” David Fontana, Government in Opposition, 119 Yale L J 548, 563 (2009). When those are adopted, we suspect that institutional loyalty in the legislative branch is correspondingly more difficult to get off the ground.
  • 253. Gersen, 96 Va L Rev at 309–10 (cited in note 79).
  • 254. Christopher R. Berry and Jacob E. Gersen, The Unbundled Executive, 75 U Chi L Rev 1385, 1433 (2008) (documenting state-level practice).
  • 255. Sierra Club v Costle, 657 F2d 298, 406 (DC Cir 1981).
  • 256. Lisa Schultz Bressman and Michael P. Vandenbergh, Inside the Administrative State: A Critical Look at the Practice of Presidential Control, 105 Mich L Rev 47, 89 (2006).
  • 257. See, for example, Rachel E. Barkow, Prosecutorial Administration: Prosecutor Bias and the Department of Justice, 99 Va L Rev 271, 312 (2013) (finding that in the Justice Department, “the law enforcement mission will trump all others”); DeShazo and Freeman, 105 Colum L Rev at 2221 (cited in note 245).
  • 258. There is also an ideological dimension to this dynamic. A mandate for a part of the federal government to do something can generate progovernmental loyalty that makes federal agents themselves more likely to select for and generate incentives toward liberal worldviews. See, for example, Joel D. Aberbach and Bert A. Rockman, Clashing Beliefs within the Executive Branch: The Nixon Administration Bureaucracy, 70 Am Polit Sci Rev 456, 466–68 (1976); Joel D. Aberbach and Bert A. Rockman, The Political Views of U.S. Senior Federal Executives, 1970–1992, 57 J Polit 838, 842–49 (1995).
  • 259. Ginsburg and Huq, Assessing Constitutional Performance at 18–20 (cited in note 26) (defining constitutional performance by the extent to which quotidian political conflict is channeled in nonviolent, discursive forums, among other goals).
  • 260. See, for example, Gillian E. Metzger, The Interdependent Relationship between Internal and External Separation of Powers, 59 Emory L J 423, 426–29 (2009); Katyal, 115 Yale L J at 2319–25 (cited in note 6).
  • 261. See Metzger, 59 Emory L J at 427–37 (cited in note 260) (describing examples of administrative structures that operate as an “internal” separation of powers); Katyal, 115 Yale L J at 2316–25 (cited in note 6) (advocating greater internal separation of powers in the face of increasing congressional abdication).
  • 262. Metzger, 59 Emory L J at 427–28 (cited in note 260).
  • 263. Id at 429–30.
  • 264. See Posner and Vermeule, 74 U Chi L Rev at 898 (cited in note 89) (arguing that an internal separation-of-powers proposal is “self-defeating”).
  • 265. The Federal Reserve’s relationship to the White House is rather more complicated than the conventional story would suggest. See George A. Krause, Federal Reserve Policy Decision Making: Political and Bureaucratic Influences, 38 Am J Polit Sci 124, 135–40 (1994) (finding “significant evidence for presidential influence on consensual decision making” by the Reserve’s Board of Governors).
  • 266. See, for example, Katyal, 115 Yale L J at 2328–29 (cited in note 6).
  • 267. Mark Landler, 51 US Diplomats, in Dissent, Urge Strikes on Assad (NY Times, June 16, 2016), online at
    -obama-airstrikes-diplomats-memo.html (visited Oct 28, 2017) (Perma archive unavailable).
  • 268. Id.
  • 269. For example, the Syria dissent did not address the potential costs and risks of state collapse in Syria, a risk the US military prioritized. See id.
  • 270. See Betsy Sinclair, The Social Citizen: Peer Networks and Political Behavior xi (Chicago 2012) (defining a social network as “the complex collection of relationships that arise from each person’s geography, work, and leisure activities”).
  • 271. James H. Fowler, et al, Causality in Political Networks, 39 Am Polit Rsrch 437, 438 (2011). See also David A. Siegel, Social Networks and Collective Action, 53 Am J Polit Sci 122, 122 (2009) (“Across social science, a wealth of empirical evidence illustrates the ways in which social interactions can alter choice.”).
  • 272. Cass R. Sunstein, Deliberative Trouble? Why Groups Go to Extremes, 110 Yale L J 71, 77 (2000).
  • 273. For an exception, see Jessica Bulman-Pozen and Heather K. Gerken, Uncooperative Federalism, 118 Yale L J 1256, 1271 (2009) (discussing social networks in the cooperative federalism context).
  • 274. For applications to government, see Harold Hongju Koh, Why Do Nations Obey International Law?, 106 Yale L J 2599, 2648 (1997); Peter M. Haas, Introduction: Epistemic Communities and International Policy Coordination, 46 Intl Org 1, 4 (1992).
  • 275. See, for example, Robert M. Cover, The Supreme Court 1982 Term—Foreword: Nomos and Narrative, 97 Harv L Rev 4, 42 (1983).
  • 276. See Sushil Bikhchandani, David Hirshleifer, and Ivo Welch, A Theory of Fads, Fashion, Custom, and Cultural Change as Informational Cascades, 100 J Polit Economy 992, 995 (1992) (explaining the implications of limited argument pools). See also Timur Kuran and Cass R. Sunstein, Availability Cascades and Risk Regulation, 51 Stan L Rev 683, 715–35 (1999) (discussing how limited argument pools shape preferences).
  • 277. By the “strength” of a network, we mean a number of variables that characterize relationships. See Daniel J. Brass, Kenneth D. Butterfield, and Bruce C. Skaggs, Relationships and Unethical Behavior: A Social Network Perspective, 23 Acad Mgmt Rev 14, 17 (1998) (“The strength of a relationship refers to the frequency, reciprocity, emotional intensity, and intimacy of that relationship. Casual acquaintances, represented by infrequent interaction and indifferent affect, are characterized by weak ties.”) (citation omitted). Information traveling between stronger ties is more likely to be duplicative. See Mark Granovetter, The Strength of Weak Ties: A Network Theory Revisited, 1 Sociological Theory 201, 204 (1983); Gabriel Weimann, The Strength of Weak Conversational Ties in the Flow of Information and Influence, 5 Soc Networks 245, 260–63 (1983).
  • 278. Huq, 65 Stan L Rev at 48 (cited in note 220), citing John Brehm and Scott Gates, Working, Shirking, and Sabotage: Bureaucratic Response to a Democratic Public 73–74 (1997). See also Damon Centola, The Spread of Behavior in an Online Social Network Experiment, 329 Science 1194, 1197 (2010) (“Whereas locally clustered ties may be redundant for simple contagions, like information or disease, they can be highly efficient for promoting behavioral diffusion.”) (citations omitted).
  • 279. Carpenter, The Forging of Bureaucratic Autonomy at 15 (cited in note 241).
  • 280. Ruth Bader Ginsburg, The Supreme Court: A Place for Women, 32 Sw U L Rev 189, 191 (2003). Marshall also provided his colleagues with port and madeira, which cannot have hurt his cause. Id.
  • 281. US Const Art I, § 6, cl 2.
  • 282. Federalist 76 at 514 (cited in note 185).
  • 283. See Williams v Pennsylvania, 136 S Ct 1899, 1905 (2016) (“The Court now holds that under the Due Process Clause there is an impermissible risk of actual bias when a judge earlier had significant, personal involvement as a prosecutor in a critical decision regarding the defendant’s case.”). See also In re Murchison, 349 US 133, 136 (1955) (“A fair trial in a fair tribunal is a basic requirement of due process.”).
  • 284. For a definition and discussion of conventions, see Adrian Vermeule, Conventions of Agency Independence, 113 Colum L Rev 1163, 1166, 1181–94 (2013) (“[U]nwritten political norms within relevant legal and political communities impose sanctions for perceived violations of agency independence or create internalized values or beliefs.”).
  • 285. Laura Kalman, Abe Fortas: A Biography 337–55 (Yale 1990).
  • 286. See Gerard N. Magliocca, The Legacy of Chief Justice Fortas, 18 Green Bag 2d 261, 261 (2015) (“Nonetheless, Fortas’s defeat was a watershed that redefined the norms governing how the Justices interact with the White House and relate to presidential politics.”).
  • 287. See Amanda Frost, Keeping Up Appearances: A Process-Oriented Approach to Judicial Recusal, 53 U Kan L Rev 531, 573 n 205 (2005) (collecting examples). For the case in question, see generally Cheney v United States District Court for District of Columbia, 542 US 367 (2004).
  • 288. See generally Cheney v United States District Court for District of Columbia, 541 US 913 (2004) (regarding motion to recuse Scalia).
  • 289. David G. Savage, Trip with Cheney Puts Ethics Spotlight on Scalia: Friends Hunt Ducks Together, Even as the Justice Is Set to Hear the Vice President’s Case (LA Times, Jan 17, 2004), online at (visited Oct 21, 2017) (Perma archive unavailable).
  • 290. We have almost nothing like a “due process of lawmaking.” See generally Hans A. Linde, Due Process of Lawmaking, 55 Neb L Rev 197 (1976). At the agency level, though, doctrines governing formal adjudications and notice and comment provide something at least comparable to due process for most executive actions. Kevin M. Stack, Agency Statutory Interpretation and Policymaking Form, 2009 Mich St L Rev 225, 234 (“[A] legality principle associated with due process imposes . . . constraint[s] on the considerations an agency may take into account in formal adjudication [and] in notice-and-comment rulemaking.”).
  • 291. Federalist 43 (Madison), in The Federalist 288, 289 (cited in note 3):

    Without [exclusive jurisdiction at the capital], not only the public authority might be insulted and its proceedings interrupted, with impunity; but a dependence of the members of the general Government, on the State comprehending the seat of government for protection in the exercise of their duty, might bring on the national councils an imputation of awe or influence, equally dishonorable to the Government, and dissatisfactory to other members of the confederacy.

  • 292. See Miller McPherson, Lynn Smith-Lovin, and James M. Cook, Birds of a Feather: Homophily in Social Networks, 27 Ann Rev Sociology 415, 435 (2001).
  • 293. See notes 109–14 and accompanying text.
  • 294. See Joel K. Goldstein, The Rising Power of the Modern Vice Presidency, 38 Pres Stud Q 374, 384–87 (2008) (discussing Cheney’s institutional claims and central role in staffing the administration).
  • 295. For the history of the unitary executive, see generally Amanda Hollis-Brusky, Helping Ideas Have Consequences: Political and Intellectual Investment in the Unitary Executive Theory, 1981–2000, 89 Denver U L Rev 197 (2011). See also Stephen Skowronek, The Conservative Insurgency and Presidential Power: A Developmental Perspective on the Unitary Executive, 122 Harv L Rev 2070, 2096–2103 (2009).
  • 296. See Gellman, Angler: The Cheney Vice Presidency at 156 (cited in note 113) (discussing long-term social and professional connections between Cheney and his staff).
  • 297. See text accompanying notes 11–13.
  • 298. See Lawrence Baum and Neal Devins, Why the Supreme Court Cares about Elites, Not the American People, 98 Georgetown L J 1515, 1537–44 (2010) (documenting elite influences on the Court).
  • 299. See Richard J. Lazarus, Advocacy Matters before and within the Supreme Court: Transforming the Court by Transforming the Bar, 96 Georgetown L J 1487, 1491–1502 (2008).
  • 300. Baum and Devins, 98 Georgetown L J at 1518 (cited in note 298), quoting Evan Thomas and Stuart Taylor Jr, O’Connor’s Rightful Heir? (Newsweek, Jan 30, 2006), archived at
  • 301. Huq and Michaels, 126 Yale L J at 382–91 (cited in note 26) (analyzing the multiplicity of goals in the separation-of-powers context).
  • 302. Id at 382.
  • 303. See, for example, Adrian Vermeule, Contra Nemo Iudex in Sua Causa: The Limits of Impartiality, 122 Yale L J 384, 397 (2012) (noting the difficulty of legislative redistricting proposals that are incentive incompatible because their implementation is blocked by the very problem of legislative entrenchment they aim to cure). But see Heather K. Gerken, Getting from Here to There in Redistricting Reform, 5 Duke J Const L & Pub Pol 1, 6–12 (2010) (suggesting pathways around this incentive incompatibility problem). There is a sense in which incentive-incompatibility objections are often better understood as failures of political imagination.
  • 304. Gerken, 5 Duke J Const L & Pub Pol at 5 (cited in note 303). In any case, legal scholars are not necessarily in the best position to judge the political feasibility of a specific proposal. See generally, for example, Ryan Bubb and Richard H. Pildes, How Behavioral Economics Trims Its Sails and Why, 127 Harv L Rev 1593 (2014).
  • 305. David M. Kreps, Corporate Culture and Economic Theory, in James E. Alt and Kenneth A. Shepsle, eds, Perspectives on Positive Political Economy 90, 93 (Cambridge 1990).
  • 306. Russell Hardin, Trust 147 (Polity 2006). See also Thomas C. Schelling, The Strategy of Conflict 134–35 (Harvard 1960) (“Trust is often achieved simply by the continuity of the relation between parties.”).
  • 307. See, for example, Matthew Crenson and Benjamin Ginsberg, Presidential Power: Unchecked and Unbalanced 11 (Norton 2007) (“Sometime in the second half of the twentieth century, the president moved into the driver’s seat of our political system.”).
  • 308. See text accompanying notes 229–31 (discussing the costs of excessive focus on a single agency’s mission).
  • 309. See, for example, DAO 202-334: Details (Office of Privacy and Open Government, June 22, 1971), archived at (discussing the terms and operation of detailing within the Commerce Department).
  • 310. See Lewis, The Politics of Presidential Appointments at 47–48 (cited in note 196).
  • 311. See Carl Brauer, Tenure, Turnover, and Postgovernment Employment Trends of Presidential Appointees, in G. Calvin Mackenzie, ed, The In-and-Outers: Presidential Appointees and Transient Government in Washington 174, 182 (Johns Hopkins 1987) (“[T]he general rule among presidential appointees is ‘in and out and never in again.’”).
  • 312. See, for example, 2 USC § 4301(f) (listing “[l]imitations” on those “detailed or assigned” from the executive branch to the legislative branch); 3 USC § 113(b)(4) (requiring that the president annually report to Congress on the usage of detailees).
  • 313. US General Accounting Office, Report to the Chairman, Subcommittee on Human Resources, Committee on Post Office and Civil Service, House of Representatives: Personnel Practices: Schedule C and Other Details to the Executive Office of the President *3 (Nov 1992), archived at (finding anywhere from 115 to 342 detailees to high-level executive-branch positions during different calendar years).
  • 314. Id at *27.
  • 315. Sarah A. Binder and Frances E. Lee, Making Deals in Congress, in Nathaniel Persily, ed, Solutions to Political Polarization in America 240, 255 (Cambridge 2015).
  • 316. See, for example, 3 USC § 112 (providing that for intra-executive detailees to the “White House Office, the Executive Residence at the White House, the Office of the Vice President, the Domestic Policy Staff, and the Office of Administration,” the new institution “shall reimburse the detailing department, agency, or establishment”); US Department of Homeland Security, Coast Guard Division, Commandant Instruction 5703.3 *2 (Apr 18, 2003), archived at (“[C]ongressional detail provisions shall be approved by the Vice Commandant, U.S. Coast Guard based on the recommendation of the Assistant Commandant for Governmental and Public Affairs developed in consultation with the Chief, Office of Congressional Affairs.”).
  • 317. See Judicial Watch, Inc v Department of Energy, 412 F3d 125, 132 (DC Cir 2005).
  • 318. Metin M. Coşgel and Thomas J. Miceli, Job Rotation: Cost, Benefits, and Stylized Facts, 155 J Inst & Theoretical Econ 301, 301 (1999) (describing job rotations as not assigning each worker “to a single and specific task but [instead] to a set of several tasks among which he or she rotates with some frequency”).
  • 319. William G. Ouchi, Theory Z: How American Business Can Meet the Japanese Challenge 32 (Addison-Wesley 1981) (“[L]ifelong job rotation [means] . . . a technician may work on a different machine or in a different division every few years, and all managers will rotate through all areas of the business.”).
  • 320. This is sometimes called “job enlargement” in the labor economics literature. Eaton H. Conant and Maurice D. Kilbridge, An Interdisciplinary Analysis of Job Enlargement: Technology, Costs, and Behavioral Implications, 18 Indust & Labor Rel Rev 377, 377 (1965) (describing this as “the practice of restoring to jobs some of the skill, responsibility, and variety that have been lost through work simplification”).
  • 321. An example, by way of analogy, is found in L, which is the State Department Legal Adviser’s Office. L requires lawyers to rotate among issue divisions within that office. Lawyers might work first on a narrow question of international environmental law of concern to few within the executive branch, but then develop knowledge of international commercial arbitration, gaining expertise in a new issue and exposure to a distinct social-network-sector salience. See Koh, 100 Georgetown L J at 1773 (cited in note 235) (“[O]ur [office’s] system of attorney rotation, which keeps lawyers in L for long tenures while allowing them to work on an ever-expanding set of issues, thus avoid[s] intellectual calcification while maintaining institutional continuity and knowledge.”).
  • 322. See McPherson, Smith-Lovin, and Cook, 27 Ann Rev Sociology at 426–27, 431 (cited in note 292). See also Jone L. Pearce and Amy E. Randel, Expectations of Organizational Mobility, Workplace Social Inclusion, and Employee Job Performance, 25 J Org Behav 81, 86 (2004) (arguing that social inclusion in the workplace has a positive relationship to individual workplace performance).
  • 323. See Gabriel Horton and David E. Lewis, Turkey Farms, Patronage, and Obama Administration Appointments *20–28 (Vanderbilt Public Law & Legal Theory Working Paper No 09-19, 2009), archived at
  • 324. Donald P. Moynihan and Sanjay K. Pandey, The Ties That Bind: Social Networks, Person-Organization Value Fit, and Turnover Intention, 18 J Pub Admin Rsrch & Theory 205, 210 (2007).
  • 325. Herbert A. Simon, Administrative Behavior: A Study of Decision-Making Processes in Administrative Organization xvi (Free Press 3d ed 1976).
  • 326. See Ackerman, The Decline and Fall of the American Republic at 152 (cited in note 235) (describing the “lawlessness” generated by OLC deference to presidential priorities).
  • 327. See id at 97; Fontana, 126 Harv L Rev F at 25–27 (cited in note 233).
  • 328. See Fontana, 126 Harv L Rev F at 40–41 (cited in note 233).
  • 329. See US Office of Personnel Management, Federal Civilian Workforce Statistics: The Fact Book *8 (2007), archived at
  • 330. See generally Ashley Parker, All the Obama 20-Somethings (NY Times, Apr 29, 2010), online at (visited Oct 21, 2017) (Perma archive unavailable) (reporting on several apartments and homes in Washington shared by high-level Obama administration officials).
  • 331. See Michael C. Dorf, Does Federal Executive Branch Experience Explain Why Some Republican Supreme Court Justices “Evolve” and Others Don’t?, 1 Harv L & Pol Rev 457, 459–66 (2007); Rob Robinson, Executive Branch Socialization and Deference on the U.S. Supreme Court, 46 L & Society Rev 889, 907–13 (2012).
  • 332. See, for example, Judicial Watch, 412 F3d at 131–32:

    Although the district court acknowledged that, during their detail, the employees worked exclusively on NEPDG matters, were supervised by the Office of the Vice President, and did not occupy an office at the DOE, it appeared not to consider those factors in deciding whether the detailees remained DOE employees; instead the district court appeared to rely solely upon the agency having paid their salaries.

    See also Schedule C and Other Details to the Executive Office of the President at *10–17 (cited in note 313) (noting various factors to consider in evaluating whether a detailee works for a past or current institution).

  • 333. Variation in institutional experience of this kind, indeed, might be made an informal criterion for hiring to the federal judiciary, especially at the appellate level, to promote a bench without asymmetrical sympathies.
  • 334. See 2 USC § 4301(f). See also Members’ Handbook and Committees’ Handbook, Detailees (House Administration, 2016), archived at
  • 335. See Lee Drutman and Steven Teles, A New Agenda for Political Reform (Wash Monthly, Mar 2015), archived at (“The federal government across all its branches has experienced a deterioration in its ability to acquire, process, and analyze information. But the problem is especially urgent in Congress, which is at the center of America’s current governing crisis.”).
  • 336. See Detail Opportunity: OIRA’s Regulatory Exchange and Training Program (OMB), archived at (“The purpose of this program is to help develop a cadre of agency experts in Executive Order regulatory review and planning, and to foster better cooperation among the agencies and OMB.”).
  • 337. See Plaut v Spendthrift Farm, Inc, 514 US 211, 222 (1995) (“[T]he need for separation of legislative from judicial power was plain [to the Framers].”); Northern Pipeline Construction Co v Marathon Pipe Line Co, 458 US 50, 59 & n 10 (1982) (“The independence from political forces . . . helps to promote public confidence in judicial determinations.”).
  • 338. Aziz Z. Huq, Judicial Independence and the Rationing of Constitutional Remedies, 65 Duke L J 1, 54 (2015).
  • 339. Edward A. Hartnett, Questioning Certiorari: Some Reflections Seventy-Five Years after the Judges’ Bill, 100 Colum L Rev 1643, 1660–84 (2000) (documenting these changes). See also generally Justin Crowe, The Forging of Judicial Autonomy: Political Entrepreneurship and the Reforms of William Howard Taft, 69 J Polit 73 (2007) (emphasizing the role of Taft’s personal efforts).
  • 340. Pamela S. Karlan, Judicial Independences, 95 Georgetown L J 1041, 1043 (2007).
  • 341. For synoptic accounts of remedial contraction, see Huq, 65 Duke L J at 12–40 (cited in note 338); Sarah Staszak, No Day in Court: Access to Justice and the Politics of Judicial Retrenchment 31–34 (Oxford 2015).
  • 342. See Samuel R. Bagenstos, The Structural Turn and the Limits of Antidiscrimination Law, 94 Cal L Rev 1, 23 (2006) (noting “judges’ unwillingness to engage in rigorous scrutiny of [subjective-practices] business structures” in employment discrimination cases).
  • 343. Samuel R. Bagenstos, Who Is Responsible for the Stealth Assault on Civil Rights?, 114 Mich L Rev 893, 904 (2016).
  • 344. Huq, 65 Duke L J at 63–67 (cited in note 338).
  • 345. Ex ante remedies are unavailable for most retail constitutional wrongs, because most citizens cannot predict official action against them.
  • 346. Tucker Act, 24 Stat 505 (1887). See also United States v Mitchell, 463 US 206, 213–15 (1983) (discussing the history of the Tucker Act, and treating it as a waiver of sovereign immunity); Richard H. Fallon Jr, et al, Hart and Wechsler’s The Federal Courts and the Federal System 896–904 (Foundation 7th ed 2015) (cataloguing various statutory waivers of federal immunity).
  • 347. Troy A. McKenzie, Judicial Independence, Autonomy, and the Bankruptcy Courts, 62 Stan L Rev 747, 752 (2010).
  • 348. See Bob Egelko, Obama Nominations Heavy on Ex-Prosecutors (SF Chron, Feb 3, 2013), archived at (noting that former prosecutors constituted 45.7 percent of President Obama’s judicial nominees, 44.7 percent of President George W. Bush’s judicial nominees, 40.7 percent of President Bill Clinton’s judicial nominees, 37.3 percent of President George H.W. Bush’s judicial nominees, and 40.8 percent of President Ronald Reagan’s judicial nominees).
  • 349. Bubb and Pildes, 127 Harv L Rev at 1678 (cited in note 304) (noting the “risks” of such an approach to scholarship).
  • 350. Pildes, 99 Cal L Rev at 276 (cited in note 171).
  • 351. Compare Alan I. Abramowitz, The Disappearing Center: Engaged Citizens, Polarization, and American Democracy 49–57 (Yale 2010) (arguing that polarization of elites drives legislative polarization), with Fiorina, Abrams, and Pope, Culture War? at 19 (cited in note 93) (rejecting the popular polarization thesis).
  • 352. For a representative version of this complaint in the legal academy, see Gillian E. Metzger, Agencies, Polarization, and the States, 115 Colum L Rev 1739, 1749 (2015) (decrying the “growing range of contexts in which a majority of legislators would prefer to alter the policy status quo but lack the numbers to overcome the objections of the President or a Senate minority”). For a related criticism of the accompanying gridlock, see Michael J. Teter, Gridlock, Legislative Supremacy, and the Problem of Arbitrary Inaction, 88 Notre Dame L Rev 2217, 2218–19 (2013) (decrying legislative inaction in the face of a pressing need to make substantive policy decisions).
  • 353. See CBS News/NY Times Poll—Congress, the Abramoff Scandals, and the Alito Nomination (CBS News, Jan 26, 2006), archived at; Lydia Saad, Americans Decry Power of Lobbyists, Corporations, Banks, Feds (Gallup, Apr 11, 2011), online at (visited Oct 21, 2017) (Perma archive unavailable).
  • 354. On the one hand, Congress is viewed as excessively influenced by a single cabal of moneyed interests. On the other hand, Congress is perceived not as unified by any one interest, but as divided by the two-party system. If you think that Congress is both hyperpartisan and also in the grip of special interests, that is, your account of Congress is somewhat internally contradictory.
  • 355. But see Josh Chafetz, The Phenomenology of Gridlock, 88 Notre Dame L Rev 2065, 2076–77 (2013) (arguing that parties have incentives to moderate positions over time).
  • 356. John Hart Ely, War and Responsibility: Constitutional Lessons of Vietnam and Its Aftermath 52 (Princeton 1993).
  • 357. The call for something like a congressional civil service is at least seventy years old. See Thomas I. Emerson, Book Review, Reviews: Congress at the Crossroads. By George B. Galloway, 56 Yale L J 1094, 1094 (1947) (describing “well-handled hearings and [a] report” calling for more professional congressional civil service leading to some, smaller reforms).
  • 358. See generally Abbe R. Gluck and Lisa Schultz Bressman, Statutory Interpretation from the Inside—an Empirical Study of Congressional Drafting, Delegation, and the Canons: Part I, 65 Stan L Rev 901 (2013); Lisa Schultz Bressman and Abbe R. Gluck, Statutory Interpretation from the Inside—an Empirical Study of Congressional Drafting, Delegation, and the Canons: Part II, 66 Stan L Rev 725 (2014). For an earlier, related article, with a smaller sample size, see Victoria F. Nourse and Jane S. Schacter, The Politics of Legislative Drafting: A Congressional Case Study, 77 NYU L Rev 575, 582–90 (2002). Political scientists have been more interested in the congressional civil service. See, for example, Michael J. Malbin, Unelected Representatives: Congressional Staff and the Future of Representative Government 11–24 (Basic 1980); Lee Drutman and Steven Teles, Why Congress Relies on Lobbyists Instead of Thinking for Itself (The Atlantic, Mar 10, 2015), archived at
  • 359. Gravel v United States, 408 US 606, 616–17 (1972):

    [I]t is literally impossible, in view of the complexities of the modern legislative process, with Congress almost constantly in session and matters of legislative concern constantly proliferating, for Members of Congress to perform their legislative tasks without the help of aides and assistants; that the day-to-day work of such aides is so critical to the Members’ performance that they must be treated as the latter’s alter egos.

  • 360. See Philip M. Stern, The Best Congress Money Can Buy 118 (Pantheon 1988) (quoting a member of Congress stating that around the times of elections he spends “eighty percent of [his] time” at fundraisers).
  • 361. One report found that members were spending just 56 percent of their time in Washington. See Lisa Desjardins, Congress in D.C. Far Less Than It Used to Be (CNN, Aug 1, 2013), archived at Congress meets only about 140 days per calendar year. Members of Congress have fewer staff members in Washington relative to in their district offices than they used to previously. See Keeping Congress Competent: Staff Pay, Turnover, and What It Means for Democracy (Sunlight Foundation), archived at
  • 362. See Gravel, 408 US at 616–17.
  • 363. Hasen, 64 Stan L Rev at 219 (cited in note 224); Barbara S. Romzek and Jennifer A. Utter, Congressional Legislative Staff: Political Professionals or Clerks?, 41 Am J Polit Sci 1251, 1271 (1997).
  • 364. There are over two million executive-branch civil servants. See Recent Trends at *16 (cited in note 203). For every 200 to 250 executive-branch officials, there is 1 congressional official. See Brookings, Vital Statistics on Congress *4 (AEI 2013), archived at
  • 365. See Bressman and Gluck, 66 Stan L Rev at 728–29 (cited in note 358) (referencing the existence of “nonpartisan, professional staff” as a discrete category).
  • 366. See notes 191–96 and accompanying text.
  • 367. See Terry Carmack, How to Get a Job on Capitol Hill (Politico, Aug 4, 2015), archived at
  • 368. See Drutman and Teles, Why Congress Relies on Lobbyists (cited in note 358) (“[Speaker of the House] Gingrich cut committee staff by a third, reduced the legislative support staff by a third, and killed the Office of Technology Assessment.”).
  • 369. See text accompanying notes 197–205.
  • 370. Senate staffers work for the Senate for an average of five years. See Senate Staff Employment Data: 1991–2001 (Congressional Management Foundation), archived at House staffers last even less time than that. See Craig Schultz and Richard Shapiro, 1994 U.S. House of Representatives Employment Practices: A Study of Staff Salary, Tenure, Demographics and Benefits *5 (Congressional Management Foundation, 1994), archived at Recall again that—with a real civil-service structure—executive-branch civil servants last several times longer than that. See Recent Trends at *17 (cited in note 203).
  • 371. See text accompanying notes 160–64.
  • 372. Mark S. Kamlet and David C. Mowery, The First Decade of the Congressional Budget Act: Legislative Imitation and Adaptation in Budgeting, 18 Polit Sci 313, 315–16 (1985).
  • 373. For aggregate data, see Congress and the Public (Gallup), online at (visited Oct 21, 2017) (Perma archive unavailable).
  • 374. See Schultz and Shapiro, 1994 U.S. House of Representatives Employment Practices at *5 (cited in note 370).
  • 375. See David J. Samuels, Presidentialized Parties: The Separation of Powers and Party Organization and Behavior, 35 Comp Polit Stud 461, 462 (2002) (“Under presidentialism, the possibility of capturing the executive branch . . . may become parties’ driving goal.”).
  • 376. See Drutman and Teles, Why Congress Relies on Lobbyists (cited in note 358).
  • 377. Jordi Blanes i Vidal, Mirko Draca, and Christian Fons-Rosen, Revolving Door Lobbyists, 102 Am Econ Rev 3731, 3739–46 (2012) (finding that lobbyists who are former staffers lose, on average, about 25 percent of their salaries when their former legislative boss leaves office).
  • 378. See Sierra Club v Costle, 657 F2d 298, 406 (DC Cir 1981).
  • 379. See Gersen, 96 Va L Rev at 303 (cited in note 79) (“There is one Congress that exercises the legislative function for all policy domains rather than two Congresses, one for foreign affairs and one for domestic affairs. Functions are separated and substantive powers are bundled together within each function.”).
  • 380. See, for example, Office of Representative Keith Ellison, Legislative Assistant Opening (Democratic Whip), archived at; Office of Congressman Al Green, Job Announcement: Legislative Assistant (Democratic Whip), archived at
  • 381. See, for example, Jurisdiction (Senate Committee on the Judiciary), archived at (“In addition to its critical role in providing oversight of the Department of Justice and the agencies under the Department’s jurisdiction, including the Federal Bureau of Investigation, and the Department of Homeland Security, the Judiciary Committee plays an important role in the consideration of nominations and pending legislation.”).
  • 382. See Bressman and Gluck, 66 Stan L Rev at 729 (cited in note 358) (stating that “nonpartisan, professional staffers [ ] are not directly accountable to members”).
  • 383. See Charlie Savage, Ratings Shrink President’s List for Judgeships (NY Times, Nov 22, 2011), online at
    -rejects-many-obama-picks-for-federal-judgeships.html (visited Oct 21, 2017) (Perma archive unavailable).
  • 384. See August Jackson, SEC Enforcement Lawyers’ Salary (Chron), archived at
  • 385. Senate Judiciary Committee Salaries (Legistorm), archived at (noting the salary of Bruce Cohen, a veteran of the Senate Judiciary Committee).
  • 386. For a useful taxonomy of potential lobbying reforms, see Hasen, 64 Stan L Rev at 237–40 (cited in note 224).
  • 387. McDonnell, 136 S Ct at 2365.
  • 388. 18 USC § 1346.
  • 389. A state or federal official can be charged with using the mail or wires to further a scheme to defraud citizens of their right to the fair, honest, and impartial services of their public officials. See Randall D. Eliason, Surgery with a Meat Axe: Using Honest Services Fraud to Prosecute Federal Corruption, 99 J Crim L & Crimin 929, 932 (2009) (explaining the statutory framework). See also Lynne Marek, DOJ May Rein in Use of ‘Honest Services’ Statute (National Law J, June 15, 2009), online at (visited Oct 29, 2017) (Perma archive unavailable) (“[Honest services fraud] was the lead charge lodged by U.S. attorney offices against 79 suspects in fiscal year 2007, up from 63 in 2005 and 28 in 2000.”).
  • 390. McDonnell, 136 S Ct at 2372.
  • 391. Id.
  • 392. We bracket here the application of honest services fraud to gifts to a representative’s campaign or PAC, and focus on gifts that enrich a representative in her personal capacity.
  • 393. Although the Court struck a First Amendment note, we are not convinced this should be preclusive. To make its First Amendment argument, the Court endorsed a model of democracy—direct lobbying of elected officials by the public—that key Framers soundly rejected. McDonnell, 136 S Ct at 2372. See also Bernard Manin, The Principles of Representative Government 94 (Cambridge 1997) (noting Madison’s reliance on a “principle of distinction” to justify representative democracy, and his rejection of direct democracy). There is therefore little reason to think that lobbying is at the center of the First Amendment. See Maggie McKinley, Lobbying and the Petition Clause, 68 Stan L Rev 1131, 1136 (2016) (concluding that the Petition Clause was not intended to cover activity akin to contemporary lobbying). There is also some reason to think the decision rests on an infirm understanding of how quid pro quos distort public action. Carl Hulse, Is the Supreme Court Naïve about Corruption? Ask Jack Abramoff, NY Times A15 (July 6, 2016) (quoting former lobbyist Jack Abramoff: “When somebody petitioning a public servant for action provides any kind of extra resources—money or a gift or anything—that affects the process”).
  • 394. Federalist 51 at 349 (cited in note 3) (advocating “giving to those who administer each department the necessary constitutional means, and personal motives, to resist encroachments of the others”).
  • 395. Id.
  • 396. Federalist 48 (Madison), in The Federalist 332, 333 (cited in note 3).
  • 397. Federalist 51 at 349 (cited in note 3).
  • 398. See generally Michael Kammen, A Machine That Would Go of Itself: The Constitution in American Culture (Routledge 2017).
  • 399. See Federalist 51 at 349 (cited in note 3).