Fiduciary Political Theory: A Critique
abstract. “Fiduciary political theory” is a burgeoning intellectual project that uses fiduciary principles to analyze public law. This Essay provides a framework for assessing the usefulness and limitations of fiduciary political theory. Our thesis is that fiduciary principles can be fruitfully applied to many domains of public law. However, other domains are incompatible with the basic structure of fiduciary norms. In these domains, fiduciary political theory is less likely to be viable.
One contribution of this Essay is to describe the underlying structure of fiduciary norms. We identify three features of these norms that differentiate them from norms of contract, tort, and criminal law. First, fiduciary norms impose deliberative requirements: they make specific types of demands on an agent’s deliberation in addition to her behavior. Second, complying with fiduciary norms requires a special conscientiousness. Living up to a fiduciary obligation depends not only on how an agent behaves and deliberates, but also on whether she does so for the right reasons. Third, fiduciary norms impose “robust” demands, which require the fiduciary to seek out and respond appropriately to new information about the interests of her beneficiaries.
We use these insights to assess applications of fiduciary principles to theories of judging, administrative governance, and international law. A fiduciary theory of judging can explain certain aspects of the norms of judging better than alternative theories offered by Ronald Dworkin and Judge Richard Posner. The viability of a fiduciary theory of administrative governance is an open question. Whether this kind of fiduciary political theory is superior to alternatives (like the instrumentalist theory of administrative governance developed by Adrian Vermeule) turns on a deeper dispute about whether administrative law reflects a culture of justification. Finally, a fiduciary political theory of international law (like the one defended by Evan Fox-Decent and Evan Criddle) is unlikely to succeed. Fiduciary norms are structurally incompatible with the domain of international law because compliance with international-law norms is a function of how states behave, rather than how they deliberate or why they behave as they do.
authors. Ethan J. Leib is a Professor of Law, Fordham Law School. Stephen R. Galoob is an Assistant Professor of Law, University of Tulsa College of Law. We thank our home institutions for research support. Thanks to participants in the Fiduciary Law Workshop at McGill Law School, as well as audiences at the Law & Society Association Annual Meeting in Seattle, the Legal Ethics Schmooze at Stanford Law School, the Pace Law School faculty workshop, and the Oklahoma Junior Scholars Workshop for their reactions and comments. Thanks in particular to Craig Agule, Vincent Chiao, Evan Criddle, Seth Davis, Evan Fox-Decent, Andrew Gold, Abner Greene, Sam Halabi, Adam Hill, Chris Kutz, Thomas Lee, Janet Levit, Paul Miller, Aaron Saiger, Amy Salyzyn, Alex Sarch, William Simon, Gordon Smith, Steve Thel, Brad Wendel, and Ben Zipursky for their comments. Thanks also to the editors of the Yale Law Journal and several referees, whose careful engagement with our manuscript improved our arguments.
“Fiduciary political theory” is an intellectual project that uses fiduciary principles to analyze aspects of public law.1 The idea that fiduciary principles apply to public offices (rather than solely to relationships in private law, where fiduciary norms originate) has a long pedigree, with roots in the writings of Cicero, Grotius, Locke, and The Federalist Papers.2
In recent years, legal scholars and political philosophers from around the globe have revived this tradition. Several fiduciary political theorists address environmental and Indian law,3 where legal doctrines most explicitly invoke fiduciary concepts. Democratic theorists also invoke fiduciary principles to analyze the inevitability of discretion and the need for constraint that arise in basic questions of political representation and political legitimacy.4 More recent efforts of fiduciary political theorists investigate domains such as constitutional law,5 international law,6 administrative law,7 election law,8 the law governing public officials,9 and even the basic structure of political authority.10
This Essay provides a framework for analyzing the usefulness and limitations of fiduciary political theory. Our thesis is that fiduciary principles can be fruitfully applied to many domains of public law. However, other domains are incompatible with the basic structure of fiduciary norms. In these domains, fiduciary political theory is not viable. The main contributions of this Essay are to reveal the underlying structure of fiduciary norms and to show when fiduciary political theorizing is likely (or unlikely) to work.11
Toward these ends, we highlight three features of fiduciary norms that differentiate them from norms of contract, tort, and criminal law. First, fiduciary norms impose deliberative requirements: they make specific types of demands on an agent’s deliberation in addition to her behavior. Second, complying with fiduciary norms requires a special conscientiousness. Living up to a fiduciary obligation depends not only on how an agent behaves and deliberates, but also on whether she does so for the right reasons. Third, fiduciary norms impose what Philip Pettit calls “robust” demands,12 which require the fiduciary to seek out and respond appropriately to new information about the interests of her beneficiaries.
Fiduciary political theory is not viable in public-law domains where any of these core features of fiduciary norms are inapposite. In other words, fiduciary political theorizing is unlikely to work in legal contexts where behavior, rather than deliberation, is the coin of the realm; where any way of conforming to a norm counts as living up to it; or where norms do not impose robust demands.
Part I of this Essay develops the claim that fiduciary norms should be applied only in public-law contexts that are compatible with the basic structure of fiduciary norms. It then provides a framework for determining whether and when fiduciary political theorizing is likely to be viable.
Part II analyzes several recent efforts to apply fiduciary principles to domains of public law through the framework developed in Part I: judging,13 administrative governance,14 and international law.15 We conclude that fiduciary theories of judging explain certain aspects of judicial norms better than prominent theories offered by Ronald Dworkin16 and Judge Richard Posner.17 By contrast, the viability of fiduciary theories of administrative governance is an open question. Whether the fiduciary theory is superior to alternatives (like the instrumentalist theory of administrative governance developed by Adrian Vermeule)18 turns on a deeper dispute about whether administrative law reflects a culture of justification. Finally, our analysis suggests that fiduciary political theories of international law are unlikely to succeed. Fiduciary norms are structurally incompatible with the domain of international law because compliance with international-law norms is a function of how states behave, rather than how they deliberate or why they behave as they do.
This Part first offers a limiting precept for fiduciary theorizing about public law (in Section I.A) and then (in Section I.B) provides a framework to analyze when fiduciary norms are compatible with a domain of public law.
A fiduciary relationship traditionally emerges in contexts where one person (the fiduciary) has discretionary power over the assets or legal interests of another (the beneficiary).19 Standard private-law examples of fiduciary relationships include attorney-client, trustee-beneficiary, corporate officeholder-shareholder, and guardian-ward.20 In such relationships, the fiduciary has discretion to act on behalf of the beneficiary. The beneficiary is vulnerable to the fiduciary’s predatory or self-dealing actions within this discretionary sphere, yet must still repose her trust in the fiduciary. The fiduciary is obligated to prioritize the beneficiary’s interests over her own.21 At least three general indicia characterize fiduciary relationships: discretion, trust, and vulnerability. In relationships exhibiting these indicia, a fiduciary is subject to specific duties—usually, duties of loyaltyandcare—that govern her actions on behalf of the beneficiary.
There are several good reasons to interpret public-law relationships in light of fiduciary norms. First, there is considerable historical precedent for thinking about public-law relationships in this way.22 Second, the architecture of the fiduciary relationship often fits the obligations of public officeholders, allowing fruitful analogies from private law to public law. Third, fiduciary political theories are grounded in inherent features of authority, rather than the consent of the governed. Thus, the fiduciary political theorist can address fundamental questions about political authority while avoiding issues related to consent that have befuddled political theorists (particularly those in the social-contract tradition) for hundreds of years.23 In identifying what makes an exercise of power legitimate, the fiduciary political theorist focuses on how that power is actually used, rather than solely on the etiology of the institutions that purport to exercise it.
Despite these synergies, some scholars doubt the viability of fiduciary political theory on the basis of putative disanalogies between public and private law.24 However, this kind of skepticism reaches only some types of fiduciary political theory—namely, those that seek to analogize private-law fiduciaries with public-law actors. Such skepticism does not indict fiduciary political theory as such. Our focus here is on a broader, more structural concern. The most serious possible objection to fiduciary political theory—one that threatens the enterprise as such—is that private-law fiduciary norms are fundamentally incompatible with the structure of public-law norms.
What, then, are the features of fiduciary norms that determine the viability of fiduciary political theory? Attempts to answer this question have proven contentious. Scholars of fiduciary law disagree about the contours and content of fiduciary norms. For example, they disagree about the bases of fiduciary norms, what obligations they impose, and how fiduciary norms differ from nonfiduciary norms.25 Some contend that fiduciary norms have a uniform content or structure, while others argue that notions like loyalty and care vary substantially across contexts.26 Further, commentators disagree about whether the legal instantiations of fiduciary notions like loyalty resemble nonlegal analogues of those concepts.27 Our goal here is not to resolve these debates.
Rather, abstracting from disagreements about the substance of fiduciary norms exposes important structural features of fiduciary norms. In this Section we identify three such structural features that are crucial to understanding how fiduciary norms differ from other kinds of legal norms. Our analysis does not presuppose any particular substantive account of the grounds, contours, or content of fiduciary norms. As such, each of the features we identify can be appreciated by almost all fiduciary legal and political theorists.28
First, fiduciary norms govern deliberation in addition to behavior (Section I.B.1). An agent who does not deliberate in the way that a fiduciary norm calls for thereby fails to live up to that norm, no matter how she behaves. Second, fiduciary norms impose standards of conscientiousness (Section I.B.2). Some ways of conforming one’s behavior and deliberation to the requirements imposed by a fiduciary norm nevertheless violate that norm. As a result, fiduciary norms invite what are sometimes called “wrong kinds of reason” problems. Third, fiduciary norms are robustly demanding (Section I.B.3). The requirements they impose morph based on changes to the world and to the beneficiary’s circumstances. One implication of the robustness of fiduciary norms is that they impose an “updating” requirement: a fiduciary must be disposed to monitor changes to the world relevant to promoting a beneficiary’s interests or ends and also be disposed to revise her efforts in light of these changes.
Although some of these features characterize other types of legal norms, fiduciary norms are unique in being simultaneously characterized by all three. In the remainder of this Section, we explain each of these features and their implications. To demonstrate why the coincidence of these features is distinctive to fiduciary norms, we provide comparisons to other types of legal norms, particularly the norms of contract, tort, and criminal law.
Norms typically govern behavior.29 However, some norms are deliberation sensitive30—that is, they can “bear upon what goes on inside people’s heads” by “demand[ing] that we have or form certain attitudes and that we think or deliberate in certain ways.”31 When a norm is deliberation sensitive, whether someone lives up to it depends on whether she forms the attitudes, thinks, or deliberates in the ways that the norm requires.32
Several scholars have noticed that fiduciary norms are deliberation sensitive.33 For example, according to the “shaping account” of fiduciary loyalty that we have articulated in previous work, a fiduciary acts loyally only if she attributes nonderivative significance to the interests of her beneficiary.34 A fiduciary whose deliberation is not shaped by the beneficiary’s interests does not live up to the duty of loyalty, regardless of how she otherwise behaves.35 Likewise, what Paul Miller calls the “principle of prudence” construes the fiduciary duty of care as deliberation sensitive.36 According to Miller, this principle “requires that the fiduciary show due care when acting as fiduciary . . . in pursuing the objects which ground her authority,” which in turn imposes deliberative and attitudinal ideals that vary across fiduciary contexts.37 In Miller’s argument the fiduciary duty of care is equivalent to a duty to be careful; in failing to be careful, a fiduciary fails to live up to the norm.38
Other types of legal norms are not deliberation sensitive in the ways that fiduciary norms are. For example, tort-law norms are, in general, deliberation insensitive: violation of a tort-law duty of care is triggered by an action, and (for the purposes of tort law) an act is an “external manifestation of an actor’s will.”39 Likewise, default contractual norms are generally deliberation insensitive: usually, whether one lives up to her contractual obligations is a matter of how she behaves.40
Criminal-law norms, by contrast, are usually sensitive to deliberation in the form of mens rea. How an individual deliberates determines, in part, whether she is subject to criminal liability. However, the deliberation sensitivity of criminal norms differs from that of fiduciary norms. In judging whether someone has lived up to a criminal norm, behavior is a threshold issue. An agent’s deliberation is relevant only insofar as her behavior does not conform to that prescribed by the norm; deliberation is not relevant independently of behavior.41 The same conclusion does not apply to fiduciary norms, where deliberation is a freestanding requirement.
To illustrate the special deliberation sensitivity of fiduciary norms, consider a modified version of an example developed by Ken Simons42:
Operation: A medical procedure involves cutting a tendon. This procedure is highly risky: there is a thirty percent chance of injury to the patient even if the procedure is performed correctly, and a far higher chance of injury when the procedure is performed incorrectly. Danielle Doctor is a physician who has badly botched every such procedure that she has performed to date. Unaware of Danielle’s history of failure, Paul Patient asks Danielle to perform the procedure on him, and Danielle agrees. During the procedure, Danielle, by luck, guesses the correct tendon to cut and thus performs the procedure in exactly the way that a competent physician would. However, Paul is among the unlucky thirty percent of patients who sustain injury when the procedure is performed correctly.
The Operation case illustrates the deliberative aspects of criminal and fiduciary norms, regardless of whether either type of norm actually applies to physicians like Danielle in this (or any) legal system.
Simons contends that Danielle would not be subject to criminal liability in Operation. General incompetence like Danielle’s does not violate criminal norms, because a “free-floating incapacity or incompetence is never relevant to criminal liability.”43 To the extent that a physician is ever criminally liable for negligence, “it is not [her] general incompetence that justifies punishment,” so much as the “highly deficient skill revealed in [a] particular operation.”44 Because, as a matter of luck, Danielle’s performance in this operation did not manifest her lack of skill, criminal liability “would be unwarranted.”45
Simons’s comments suggest that criminal norms adopt what we will call a manifestation requirement: mental states (e.g., how an agent deliberates, what she intends, what she disregards) and their absence matter to criminal liability only insofar as they are connected with an agent’s behavior. As Gideon Yaffe puts it, “mens rea is essential” to judgments of liability for violating criminal norms, “but it isn’t relevant unless it’s manifested” in behavior.46 This requirement explains the criminal-law nostrum that punishment should be for acts, rather than for status: to punish in the absence of an action is to punish an agent for her presumed mental state, even though it has not caused her to behave in any particular way. It also explains the general disdain for so-called “thought crimes,” since these crimes punish for mental states that have not been manifested in an agent’s behavior.47
The Operation case illustrates that fiduciary norms are sensitive to deliberation in a different way than criminal norms are. If fiduciary norms applied to Danielle, then she would have failed to live up to them because her faulty pattern of deliberation is an instance of both carelessness and disloyalty. Fiduciary norms therefore reject the manifestation requirement. Disloyalty or carelessness can constitute a violation of these norms, regardless of whether or how these mental states are revealed in behavior.48
Consider an alternative version of the Operation scenario in which Paul does not suffer any injury from the procedure. Since fiduciary norms reject the manifestation requirement, it follows that Danielle would violate her fiduciary duties in this alternative scenario.49 Because of the way that fiduciary norms are sensitive to deliberation, someone can violate these norms solely through a faulty pattern of deliberation, regardless of whether this deliberation manifests in behavior.
Our analysis so far has concerned negligence, or the failure to appreciate a substantial and unjustifiable risk of which one should have been aware. However, our conclusions seem even stronger when applied to more involved mental states, like the mens rea of purpose that forms the core of attempt liability. There is no such thing as tort or contract liability for attempt. In general, failed attempts to harm someone do not violate tort norms. Likewise, someone who tries his best to breach a contract but winds up performing anyway does not necessarily violate contractual norms.
Criminal norms, of course, prohibit attempts. Someone who tries but fails to assault another person commits a crime—namely, the crime of attempt, rather than the crime of assault. Fiduciary norms seem to prohibit attempts as well. A fiduciary who tries to betray her principal has been disloyal, regardless of whether these efforts succeed.50 Thus, both criminal and fiduciary norms regarding attempts are deliberation sensitive. Here, too, there are differences in how deliberation matters, as demonstrated by the following scenario:
Wicked Operation: Assume the same medical procedure and protagonists as in Operation. Paul Patient asks Danielle Doctor to perform the operation on him. Danielle determines that she dislikes Paul, and she forms a plan to cut the wrong tendon during the operation, which will cause Paul excruciating pain. To remind herself of this plan (which she is likely to forget given her busy schedule), Danielle writes “Cut tendon to injure Paul during operation” into her notebook, and affixes her personal seal to the page. Danielle’s policy is to be fully committed to carrying out any plan to which she affixes her seal.
Has Danielle attempted to harm Paul in Wicked Operation? If traditional criminal norms applied in this case, then Danielle would almost certainly not have violated them at the point where the scenario cuts off. In general, criminal norms prohibiting attempts require not only that the defendant have the purpose to commit an object crime, but also that she take some action toward the commission of that crime.51 Danielle’s plan to injure Paul during the procedure would satisfy the mental-state requirement for attempt. Her behavior, however, would not satisfy any existing formulation of the act requirement for attempt. In the language of criminal law, writing the plan in the notebook and affixing the seal would likely be considered “mere preparation,” rather than behavior constituting a substantial step toward the result of injuring Paul.52 One explanation for this conclusion is that criminal norms embrace the manifestation requirement: Danielle has not attempted to injure Paul because her plan has not yet been manifest in her behavior.53 Merely planning to harm someone is not equivalent to trying to harm him.
Fiduciary norms do not support the same conclusion. Regardless of whether she has attempted to harm Paul, Danielle hasviolated a fiduciary obligation to him. More generally, evidence that you have “merely” planned to betray someone is sufficient to establish that you are disloyal toward her. It does not matter whether your intention is ever manifest in behavior leading toward a result. Intuitively, then, having a firm plan to harm someone who has trusted you is not merely an attempted betrayal; it is a betrayal. Beyond these intuitions, the rejection of the manifestation requirement coheres with several structural features of fiduciary norms. For example, because of the so-called “prophylactic” nature of fiduciary rules against conflicts of interest, a conflict of interest violates the fiduciary’s duty regardless of whether it has any causal effect on the actions that the fiduciary takes on behalf of the beneficiary.54
Therefore, fiduciary norms (like criminal norms, but unlike norms of contract and tort) are deliberation sensitive. They impose demands on both behavior and deliberation. Unlike criminal norms, however, fiduciary norms reject the manifestation requirement. Specific patterns of deliberation can violate fiduciary norms regardless of how (or whether) they are connected with behavior.
Norms have conditions of success. We use the term “compliance” to describe success in living up to a norm and “breach” or “violation” to describe failure to live up to a norm.55 There are several possible modes of complying with a norm. Two modes that are most relevant to our discussion are “following” and “conforming.”56 Someone follows a norm when she not only behaves or deliberates as the norm requires, but also justifies these actions by the fact that the norm requires these behaviors or deliberations.57 Someone conforms to a norm when she behaves or deliberates as the norm requires, “not because of the norm, but because of other considerations associated with the norm.”58Following a norm is more demanding, since it imposes second-order standards regarding how the norm figures into one’s practical deliberation. Because following is so demanding, some commentators see conformity as the default mode of complying with a norm.59
Different norms impose different standards for compliance. For certain types of norms, following is irrational or self-defeating, so conformity is the best (and perhaps only) way to comply.60 For other types of norms, following is the requisite mode of compliance.61 An agent would violate this type of norm if her behavior and/or deliberation matched the requirements of the norm, but she lacked the requisite practical orientation toward the norm. Still other types of norms are agnostic about compliance: conforming works just as well as following, and any route to compliance is just as successful as any other route.
Fiduciary norms impose standards of compliance. Complying with the fiduciary duty of loyalty requires a special conscientiousness regarding the interests or ends of the beneficiary. Patterns of behavior or deliberation that lack this conscientiousness breach fiduciary norms.62 Thus, not just any token of conformity counts as complying with a fiduciary norm. There are many possible ways to describe the conscientiousness that fiduciary norms require. In previous works, we argued that it is impossible to act loyally by accident. If an agent’s behavior and/or deliberation happen to match the pattern specified by a fiduciary norm, but the interests or ends of the principal do not influence the agent’s practical deliberation in the right way, then the agent has not complied with the fiduciary norm.63 Paul Miller contends that all fiduciary norms contain a “principle of fidelity,” according to which an agent must “manifest [a] commitment to the fate of the purpose or person to the extent that same is within the control of the fiduciary in the exercise of her powers.”64 By implication, to behave or deliberate in a way that does not manifest this commitment (or to lack the commitment altogether) is to breach the fiduciary norm. In the legal domain, the conscientiousness necessary to live up to a fiduciary duty is sometimes termed a requirement of “good faith.”65 Each of these formulations suggests that fiduciary norms are not agnostic about compliance: certain ways of conforming to fiduciary duties do not count as living up to fiduciary norms.66
By contrast, the other types of norms that we have considered so far do not impose standards of conscientiousness. For these norms, all modes of conforming are equally successful. Suppose that criminal norms prevent a particular course of action (Y-ing), but do not prohibit another course of action (X-ing). However, Adam thinks that X-ing is prohibited by criminal norms, and he forms and executes an elaborate plan to X. (Suppose also that Adam is disposed to break the law: if he knew that only Y-ing is criminally prohibited, he would Y.) According to the notion of “pure” legal impossibility, Adam does not breach any criminal norm by X-ing, even though he believes that he does.67 Adam is an accidental complier with criminal norms because his behavior happens to conform to the law. Yet this happenstance is sufficient for Adam to live up to the criminal norm.
Given these standards of conscientiousness, fiduciary norms are subject to what some philosophers call the “wrong kinds of reasons” problem.68 This problem arises, among other places, in discussions of “fitting attitudes” accounts of value, which define value as that which we have reason to value. Some reasons to value an object have nothing to do with the object’s value. These are not good reasons to value an object.69 The same basic argument applies to a variety of attitudes and beliefs. That someone is generous is the right kind of reason to admire her; that her benefactor will pay me to admire her is the wrong kind of reason.70
Mark Schroeder has argued that the “wrong kinds of reason” problem applies to any activity that is subject to standards of correctness.71 For Schroeder, these standards of correctness give rise to sets of reasons that are shared by all participants in the activity. Only the reasons within this set are the right kinds of reasons. Some (otherwise licit) considerations, however, might motivate performance of an action, despite being outside this shared set of reasons for the activity. Despite their motivational efficacy, these are nonetheless the wrong kinds of reasons.72
Fiduciary norms are (by Schroeder’s logic) vulnerable to the “wrong kinds of reasons” problem. Illicit or inconsiderate reasons might move a fiduciary to behave or deliberate in the way that fiduciary norms call for. Even so, these patterns of behavior or deliberation would be inappropriate because they would be outside the set of reasons that characterize fiduciary norms. For example, suppose that a fiduciary and a beneficiary were both members of a highly demanding religion, one that requires complete devotion to the faith.73 As a result of their common membership, the fiduciary might think about and act in a way that happens to advance the beneficiary’s interests or ends. If complying with a fiduciary norm required only that a fiduciary think about and advance his beneficiary’s interests or ends, then the co-religionist would live up to fiduciary norms. Yet, by definition, the co-religionist fiduciary is not loyal to the beneficiary. Rather, he is loyal to the religion, and in exhibiting this loyalty he happens to advance the interests or ends of the beneficiary. “That the beneficiary is a member of my religion” is certainly a licit reason for action, and it might be a powerful reason for action. Nevertheless, it is the wrong kind of reason for a fiduciary to be loyal to the beneficiary, at least as far as fiduciary norms are concerned.
Contract, tort, and criminal norms do not require conscientiousness, nor do they invite the “wrong kinds of reasons” problem. Someone who refrains from killing another person for morally inappropriate reasons (e.g., solely because he does not want to risk the chance of imprisonment) has complied with the criminal norm against homicide. Someone who fulfills his end of an agreement solely out of concern for his commercial reputation has lived up to the contractual norm regarding performance. Someone who does not injure another simply because it would not be fun has not violated any norm of tort.
In sum, fiduciary norms have standards of conscientiousness, while contract, tort, and criminal norms do not. Because of these standards, fiduciary norms invite “wrong kinds of reason” problems.
Fiduciary norms also involve a species of what Pettit calls “robustly demanding” values.74 According to Pettit, someone can enjoy a robustly demanding value “only insofar as [he] enjoy[s] a corresponding, intuitively thinner benefit . . . not just actually, but across a certain range of possibilities, where the extent of that range determines the degree of robustness with which [he] enjoy[s] it.”75 For example, Pettit sees friendship as a robustly demanding value. In order to be someone’s friend, you must favor her “reliably or robustly, and not just as a contingent matter: not just as luck or chance or a spasm of good will would have it.”76 Someone who is there for you in good times but would not be there for you in bad times is not your friend. This conclusion holds even if times never go bad.
The demands imposed by fiduciary norms are robust in all of these ways. Consider the paradigmatic fiduciary duty of loyalty. Whether an agent is loyal depends not only on how she acts to advance a principal’s interests or ends, but also on her disposition to act in circumstances where those interests or ends change. Part of the robustness of loyalty concerns how the beneficiary’s interests or ends must matter to the fiduciary. Loyalty inevitably has a counterfactual element: “regardless of what the beneficiary’s interests happen to be, if these interests were different, then the loyal fiduciary’s deliberative situation would be different as well.”77 If a fiduciary is disposed not to revise her deliberation in accordance with changes in the beneficiary’s interests or ends, then she is not loyal to the beneficiary.
The robustness of the demands on fiduciaries distinguishes these types of norms from norms of contract and tort. For example, because harm is an element of tort liability, an agent (T) does not necessarily violate a norm of tort law in circumstances where T’s conduct does not harm a victim (V), but would have harmed V under alternative facts that (unbeknownst to T) do not apply. In other words, as discussed above, tort norms do not usually assign liability for what in criminal law are called impossible attempts.78 Likewise, suppose that A and B have a contract regarding doing X. A will not violate any contractual norm if she does X but would not have done Y had the contract concerned Y-ing.
These conclusions are not true for fiduciary norms. Impossible attempts violate fiduciary norms. T would breach a fiduciary duty if he took an action whose purpose is to harm V, even if that action did not actually harm V. Likewise, A would breach fiduciary norms if she does X, but, in alternative circumstances in which doing Y was in B’s best interest, she would be committed not to do Y. Because fiduciary norms reject the manifestation requirement, how a fiduciary is disposed to behave or deliberate in nonactualized circumstances affects whether she lives up to her fiduciary duties.
A more concrete illustration of the robust character of fiduciary norms can be derived from Henry Richardson’s discussion of the responsibilities of medical researchers.79 By way of background, malaria and schistosomiasis are both parasitic diseases that are endemic in certain parts of Africa.80 Richardson notes that medical researchers working in these areas “will often confirm malaria diagnoses by checking fluid samples under the microscope. When they do so, they are likely to see, and so to diagnose infection by, the schistosomiasis parasite as well.”81 Further, “in many of the areas where such research is carried out, if the participants do not receive this medical care from the researchers, they will not receive it at all.”82 Consider the following example:
Case 1: As part of a public-health effort to combat malaria, Richard Researcher diagnoses and treats malaria in Preston Patient. Prior to receiving medical treatment, Preston signs a consent agreement that both waives his privacy rights and contains the following provision: “This study is only to diagnose and treat malaria.” During the study, Richard diagnoses Preston with schistosomiasis but does not convey this information to Preston or treat the disease.
The discovery of schistosomiasis in this example is what bioethicists call an “incidental finding,” or a “finding concerning an individual research participant that has potential health or reproductive importance and is discovered in the course of conducting research but is beyond the aims of the study.”83 Does Richard have a responsibility to convey these incidental findings to Preston, or to provide him with what Richardson calls “ancillary care” (that is, “medical care that . . . research subjects need but that is not required to make a study scientifically valid, to ensure a study’s safety, or to redress research injuries”)?84
Suppose that tort, contract, and fiduciary norms all apply to Richard in Case 1. It is unclear whether Richard’s actions would violate extant tort norms. The answer to this question depends on contingent facts about whether conveying incidental findings and providing ancillary care are part of the relevant standard of care, which in turn depends on how other medical researchers would have behaved in these circumstances.85 Likewise, Richard’s actions in Case 1 would probably not breach generally applicable norms of contract, since the disclaimer in the consent agreement is likely sufficient to relieve Richard of any responsibility for diagnosing or treating Preston’s schistosomiasis.
In contrast, Richard’s conduct would violate fiduciary norms. Richardson concludes that medical researchers have a general duty to warn and to treat in circumstances like Case 1, and his argument for that conclusion closely resembles an application of fiduciary principles. For Richardson, the medical researcher’s ancillary-care duties arise from the subject’s waiver of privacy rights. The volitional exercise involved in this waiver creates what Richardson calls a “moral entanglement,” whereby “special obligations unintendedly arise in a way that is ancillary to some other moral transaction.”86 However, the waiver does not fully explain the duties that attach to this relationship. According to Richardson, “The initial, mutually voluntary establishment of a privacy waiver—often a one-sided one—sets up a basic voluntary assumption of responsibility, one that potentially makes certain vulnerabilities of the offeror of special moral concern to the accepter.”87 This special moral concern, in turn, creates special responsibilities for a researcher (and for moral agents generally):
Once one has warned the other about a new problem that one discovers on the basis of private information (or has otherwise broached the issue), the morally best way to address these underlying concerns becomes not to duck out, but rather to help the other address the problem—assuming the other wants help. In this way, the duty to warn (or its cognates) further deepens the incipient relationship begun by the initial intimacy, providing a clear locus for a special obligation of beneficence.88
In contrast with norms of promise and contract, under which someone’s obligations are “largely under the control of [her] will,” entanglements like those involved in Case 1 “can extend our obligations in ways that we do not intend or foresee” because the researcher’s responsibilities to the subject “morph when one has warned the other about a danger discovered on the basis of private information.”89
Fiduciary responsibilities have exactly the “morphing” quality that Richardson attributes to entanglement-based responsibilities. Indeed, many commentators identify something like this “morphing” quality in describing fiduciary duties as “open-ended.”90 A fiduciary violates her responsibilities when, in light of changed circumstances, she fails to alter her behavior or deliberation in the way that is required to fulfill her mandate.91 In Case 1, then, fiduciary norms would impose responsibilities to convey incidental findings and provide ancillary care beyond those contemplated in the initial agreement. By contrast, responsibilities arising under contract and tort do not necessarily “morph” in this way.
The following example suggests another way that fiduciary norms are robustly demanding:
Case 2: As part of the same public-health effort in Case 1, Rhonda Researcher diagnoses and treats malaria in Pauline Patient. Rhonda could easily determine whether Pauline is infected with schistosomiasis; however, doing so would trigger ancillary-care responsibilities that Rhonda does not want to undertake. Therefore, Rhonda does not investigate whether Pauline’s sample is infected with schistosomiasis, nor does she provide treatment for this disease.
We think that norms of tort and contract reach the same conclusions in Case 2 as in Case 1. Because these types of norms do not (or, in the case of tort norms, do not necessarily) impose ancillary-care responsibilities, they do not necessarily impose any duty to seek out incidental findings or to investigate whether ancillary care is needed.
But if fiduciary norms applied, then Rhonda’s actions in Case 2 would likely violate them. The best explanation for this conclusion is that fiduciary norms not only impose responsibilities that “morph” but also (in light of these responsibilities) impose responsibilities to determine whether some morphing of responsibilities is called for in order to further the beneficiary’s interests or ends. We call this further responsibility an updating requirement. As part of this requirement, fiduciary norms demand that the fiduciary investigate changes in the beneficiary’s interest or ends. They also demand sensitivity to the results of the investigation: the fiduciary must be committed to revising her efforts on behalf of the beneficiary in light of changes to the beneficiary’s interests or ends. A failure to update in either of these respects (through, for example, a failure to investigate or an unwillingness to revise) is sufficient to violate fiduciary norms.
Our statement of the updating requirement so far has been abstract. The requirement is likely to be instantiated differently in different fiduciary contexts. For example, an attorney owes fiduciary duties to her client, part of which includes the duty to investigate matters that might advance the client’s legal interests. Although the rules regarding this duty are complex, at least some failures to investigate matters that would advance the client’s interests are violations of the lawyer’s fiduciary duty.92 Moreover, although the contours of the investigative duty differ depending on the context, the sensitivity requirement presumably does not. The fiduciary duty of loyalty would be breached if a fiduciary “adopt[ed] a[n] ‘[I] don’t care about the risks’ attitude” concerning a decision regarding the interests or ends of the beneficiary.93 A lawyer’s insensitivity to the fruits of investigation thus violates fiduciary norms because it amounts to an “I don’t care about the risks” approach to lawyering.
The application of this updating requirement further distinguishes fiduciary norms from other types of legal norms. Criminal norms impose no updating requirements: in order to comply with criminal law, one must only abide by the criminal norms that exist, rather than the ones that might exist. No citizen has a duty to investigate which criminal prohibitions would best serve the public interest or a duty to abide by criminal prohibitions that have not been formally enacted. Indeed, imposing such an updating requirement for criminal norms would be inconsistent with basic values of legality.94 Nor do contractual norms impose default updating requirements. Once an agreement is made, a party has no automatic duty to monitor changes to the counterparty’s interests or to unilaterally revise the terms of the agreement if doing so would advance the counterparty’s interests.95
Thus, fiduciary norms impose robust demands on agents. Unlike other types of legal norms, the requirements imposed by fiduciary norms necessarily “morph.” Among the demands unique to fiduciary norms is an updating requirement: the fiduciary must not only monitor changes in the beneficiary’s interests or ends but also modify her actions on behalf of the beneficiary as a result of these investigations.
To summarize the analysis set forth above, here is our framework for determining when fiduciary political theory is likely to work: because fiduciary norms are deliberation sensitive, have standards of conscientiousness, and impose robust demands, fiduciary political theory will only be capable of illuminating domains of public law where these features are at home. We use this framework to show how some efforts at fiduciary political theory illuminate, others appear to be incomplete, and still others are subject to criticism.
Toward this end, we consider three examples of fiduciary political theory: a fiduciary theory of judging advanced by Ethan Leib, David Ponet, and Michael Serota (Section II.A); a fiduciary theory of administrative governance articulated by Evan Criddle (Section II.B); and a fiduciary theory of international law defended by Criddle and Evan Fox-Decent (Section II.C). We mean these discussions to be case studies, rather than definitive engagements. Thus, we consider only the versions of these theories advanced by their authors, rather than all existing or possible versions. Our primary goal is neither to vindicate nor to condemn the substance of these particular theories.96 Rather, we aim to demonstrate whether any fiduciary theory is likely to be fruitful in these domains of public law.
To preview our conclusions, fiduciary theories of judging are capable of explaining important aspects of the judicial role that prominent alternative theories (like those advanced by Ronald Dworkin and Richard Posner) cannot easily explain. Fiduciary theories of administrative governance are intriguing, but they rest on premises that are largely undefended and would almost certainly be denied by those (e.g., consequentialists and pragmatists like Adrian Vermeule) who are most likely to reject fiduciary political theorizing. Finally, the domain of international law seems to us incompatible with several features of fiduciary norms. This incompatibility calls into question the viability of fiduciary political theory about international law, at least as that domain of law currently exists.
A theory of judging establishes how “judges [should] decide the controversies that are presented to them.”97 The fiduciary theory of judging advanced by Leib, Ponet, and Serota (LPS) contends that every judge has fiduciary obligations to the people who are bound by the law that the judge interprets and applies.98 LPS construe fiduciary norms as applying to any context in which “one person . . . has discretionary power over the assets or legal interests of another,” such that the legal interests of the latter are vulnerable to the discretionary decisions of the former.99 According to LPS, each of the indicia of fiduciary norms is realized in the judicial role: judges maintain wide discretionary authority over those who are subject to the law,100 judges are entrusted by those who are subject to the law with the authority to interpret it,101 and “the delegation to judges of substantial legal authority to apply or interpret the law leaves citizens vulnerable.”102 Therefore, LPS contend, fiduciary norms apply to judges.103 As such, judges owe “the people”—their putative beneficiary class—a duty of loyalty (a responsibility to be impartial toward the litigants in resolving a case),104 a duty of care (which requires the exercise of competence in saying what the law is and providing reasons for decisions),105 and a separate duty of candor (which “instructs that judges should say what they mean”).106 Consistent with the fiduciary theory of judging, the basic norms for judges impose each of these duties.107
From this brief sketch, it is possible to contrast fiduciary theories of judging with two prominent alternatives.108 An interpretivist theory like Ronald Dworkin’s “law as integrity” view grounds the judge’s role-based responsibilities on free-floating commitments to certain ideals.109 In Dworkin’s view, a judge should resolve cases based on morally weighty principles that fit (or explain) a sufficient number of existing legal materials. Dworkin’s theory of law holds that there is a right answer to hard legal questions,110 and “law as integrity” is an adjudicative methodology that allows the judge to reach this right answer.111 Dworkin concedes that his view is consequentialist in the sense that “each interpretive legal argument is aimed to secure a state of affairs that is superior, according to principles embedded in our practice, to alternatives.”112 The “law as integrity” view is thus, in Lawrence Solum’s phrase, “outcome-centered”; whether a judge has acted with integrity is determined by whether she reaches the correct resolution of the case before her.113
Another prominent theory of judging is the pragmatic theory of adjudication defended by, among others, Richard Posner. For Posner, legal pragmatism is, roughly, “a heightened concern with consequences or . . . ‘a disposition to ground policy judgments on facts and consequences rather than on conceptualisms and generalities.’”114 Posner contends that a judge should engage in pragmatic adjudication by “always tr[ying] to do the best he can do for the present and the future, unchecked by any felt duty to secure consistency in principle with what other officials have done in the past.”115 Pragmatic adjudication is outcome-centered in the same way that “law as integrity” is. Posner envisions a two-step process of judicial interpretation in which the judge first “infer[s] a purpose from the language and context” of the legal text or body of law, then “decide[s] what outcome in the case at hand would serve that purpose best.”116 For Posner, the outcome is what matters fundamentally; the judge’s reasoning is relevant only insofar as it is a reliable or useful guide to securing the best outcome.117 There is, Posner says, “no intrinsic or fundamental difference between how a judge approaches a legal problem and how a businessman approaches a problem of production or marketing.”118
The fiduciary theory seems better suited than these alternatives to explain the features of the norms for judges (including duties regarding loyalty, care, and candor) that LPS identify. However, assume for the sake of argument that both “law as integrity” and pragmatic adjudication could explain these basic tenets.119 In what follows, we argue that the norms for judges have each of the structural characteristics of fiduciary norms: they are deliberation sensitive (Section II.A.1), they impose standards of conscientiousness (Section II.A.2), and they apply robustly (Section II.A.3). Each of these features is widely recognized as an aspect of the norms for judges. Indeed, these features are perhaps so obvious as to be taken for granted. It would be difficult to imagine a judiciary worth having in which the norms for judges lacked any of these features. While outcome-centered alternatives (like those proposed by Dworkin and Posner) cannot easily accommodate these features, the fiduciary theory of judging can.
Yes. The norms for judges directly impose requirements on how judges deliberate. For example, Rule 2.2 of the American Bar Association’s Model Code of Judicial Conduct stipulates that a judge must “uphold and apply the law” and “perform all duties of judicial office fairly and impartially.”120Impartiality, in turn, requires the judge to be “objective and open-minded.”121A judge whose deliberation failed to be objective or who was not open-minded would violate this norm, regardless of how she resolved the cases before her. Moreover, a judge can breach the norms for judges by failing to live up to her deliberative responsibilities, regardless of whether this failure is manifest in her behavior. Existing statements of the norms for judges, particularly the prophylactic nature of the rules regarding conflicts of interest and prohibitions on ex parte contacts,122 reflect this possibility. It is also demonstrated by the following scenario:
Conspiring Judges123: Donor makes a $1,000 campaign contribution to Judge Rivers. Judge Rivers tells Donor, “You run into a problem with any of your people, you get ahold of me. Anything you need.”
Khan is a defendant in a firearm possession case before Judge Seagull. Donor tells Judge Rivers that Khan is one of her people and asks for his help.
Judge Rivers calls Judge Seagull and tells her that Khan is one of his friends. Judge Rivers asks Judge Seagull to “help” Khan, to which Judge Seagull replies, “Okay.” When Khan appears in Judge Seagull’s court several months later, she reduces the charges against him to a misdemeanor. Judge Seagull never takes anything from Judge Rivers.
We think that Judge Seagull violates a fiduciary norm in Conspiring Judges. This conclusion holds even if the decision to reduce the charges against Khan is socially optimal or consistent with existing legal materials. Likewise, Judge Seagull’s actions violate the norms for judges even if she would have reduced the charges without Judge Rivers’s exhortation, or if she never actually enters the order reducing the charges against Khan. Judge Seagull’s agreement with Judge Rivers is, at the very least, a violation of the deliberative norm requiring objectiveness and open-mindedness. These conclusions do not depend on how or whether Judge Seagull’s commitments are manifest in her behavior. Thus, the norms for judges appear to be sensitive to deliberation in the same way as fiduciary norms.
By contrast, neither Dworkin’s “law as integrity” view nor Posner’s pragmatic adjudication view can explain these results. On both theories, if the outcome of Judge Seagull’s decision is correct (that is, if it is the morally best interpretation of the extant legal materials or tends to produce socially optimal results), then she lives up to the norms for judges. In other words, both theories would attribute only derivative relevance to Judge Seagull’s deliberation: her deliberative process would matter only if she had not reached the correct decision.124 The problem with these alternative theories, then, is their outcome centrism. The norms for judges clearly impose requirements regarding outcomes. Conspiring Judges illustrates that these norms also impose freestanding deliberative requirements. Among the three theories of judging considered here, only the fiduciary theory captures this nuance.
Here, too, an affirmative answer seems clear. Not just any way of satisfying the behavioral or deliberative requirements of norms for judges will count as compliance. For example, Canon 1 of the Model Code of Judicial Conduct instructs that a judge must “uphold and promote the independence, integrity, and impartiality of the judiciary, and . . . avoid impropriety and the appearance of impropriety.”125 A judge whose actions have the tendency to promote the “independence, integrity, and impartiality of the judiciary” but who nonetheless fails to “uphold” these values would therefore fail to comply with this norm.
Another way to appreciate that the norms for judges impose standards of conscientiousness is to note that these norms, like fiduciary norms, are subject to “wrong kinds of reasons” problems. The Conspiring Judges scenario demonstrates this susceptibility. In that case, Judge Rivers asks Judge Seagull to reduce the charge against Khan. That your fellow judge asks you to reduce the charge can be a powerful reason to do so. However, this consideration is not within the shared set of reasons applicable to legal officials. It is, therefore, the wrong kind of reason for reducing the charge. In acting on it, Judge Seagull violates the norms for judges.
The alternative theories of judging fail to appreciate the conscientiousness standards implicit in the norms for judges. In the terms we introduced in Section I.B.2, pragmatic adjudication is generally agnostic about compliance. To the extent that outcomes are what matter fundamentally, any way that a judge reaches the correct decision is, in principle, as good as any other way. Posner’s agnosticism is both explicit and general: on his view, “[h]ow the judge arrives at his decision is . . . a ‘meta-legal’ question without interest” in establishing whether a judge has done his job.126
Dworkin’s view is also agnostic regarding much of the judge’s role, although this agnosticism is less explicit and more cabined. Dworkin’s position would allow that integrity imposes compliance standards (and invites “wrong kinds of reasons” problems), but only regarding what Dworkin calls “matters of principle,” or standards that are to be observed as basic “requirement[s] of justice or fairness or some other dimension of [political] morality.”127 By contrast, matters of policy, which set out “a goal to be reached, generally an improvement in some economic, political or social feature of the community,”128 do not call integrity into question and therefore do not impose standards of compliance.129 How the “law as integrity” theory would analyze the Conspiring Judges case depends on whether classifying criminal charges as a misdemeanor or felony is a matter of principle or a matter of policy. The latter classification seems better supported by Dworkin’s description of his view.130 If so, then the “law as integrity” theory would conclude that there are no conscientiousness standards for this aspect of the judge’s role, and that “wrong kinds of reasons” problems are inapplicable in this case.
To summarize, the norms for judges impose standards of conscientiousness. The fiduciary theory can explain these standards while alternative views cannot do so easily, if at all.
The norms for judges are also robust, in both the “morphing” and “updating” senses described in Section I.B.3. The morphing quality of these norms is captured by the widely acknowledged notion that judicial responsibilities are open-ended.131 The “updating” quality of these norms is evidenced in the idea that a judge’s decisions must be sensitive to new information. There are, of course, limits to (and debates about) the kinds of updating that judges must or may do. To wit, Rule 2.9(c) of the Model Code of Judicial Conduct states that a “judge shall not investigate facts in a matter independently, and shall consider only the evidence presented and any facts that may properly be judicially noticed.”132 Whatever the proper scope of judicial updating, however, it seems obvious that a judge can be required to conduct some sort of due diligence as part of the fulfillment of her official duties. The fiduciary theory of judging captures this requirement, although so can Dworkin’s and Posner’s theories.
However, appreciating the sensitivity of the norms for judges affords another way to appreciate what goes wrong in Conspiring Judges. To the charge that she has violated the norms for judges, Judge Seagull might respond that she resolved the case before her in exactly the way that a norm-abiding judge would have. Regardless of whether the decision was correct, however, Judge Seagull’s decision procedure was problematic because it was insensitive to the facts of the case, the content of the law, and the interests of the public. In other words, Judge Seagull would have decided the case in Khan’s favor regardless of the legal merits of his position; that the legal merits supported her decision was a happy accident. If the circumstances had been only slightly different—if, for example, Judge Rivers had asked Judge Seagull to increase the charges against Khan as a favor to his contributor—then this decision procedure would have yielded the wrong answer. Living up to the norms for judges requires more than that a judge’s decision have legal bases. In addition, the connection between these legal bases and the decision must be intentional.
Both “law as integrity” and pragmatism can explain some of the ways in which the norms for judges are robust. In particular, both approaches can appreciate how the requirements incumbent on a judge morph based on changes in circumstances.133 However, because these views are outcome centered, they cannot fully capture the updating requirement. In particular, they cannot easily condemn decisions made according to insensitive procedures that happen to succeed. Neither can they explain why the lucky success exhibited in Conspiring Judges is deficient. Fiduciary theory, on the other hand, can straightforwardly explain that an insensitive decision procedure is problematic because it exhibits neither the conscientiousness nor robustness that loyalty requires.
Thus, fiduciary political theory seems to illuminate the domain of judging. As articulated by LPS, the fiduciary theory of judging captures important aspects of the norms for judges that prominent alternative theories cannot easily explain. This is not to say that the fiduciary theory fully describes the normative constraints on judges, or that no version of the alternative theories could account for the core of the norms for judges.134 Rather, our claim here is that norms for judges are deliberation sensitive, impose standards of conscientiousness, and operate robustly. A theory of judging is incomplete if it neglects these features, and the fiduciary theory provides a straightforward way to both explain and justify them.
A theory of administrative governance provides standards for determining how governance by administrative institutions can (or cannot) be politically legitimate.135 Evan Criddle has offered the clearest interpretation of a fiduciary theory of administrative governance, as well as a fiduciary theory of administrative rulemaking (that is, how agencies should make rules to implement legislation).136 According to Criddle, fiduciary norms arise in all circumstances where an entrustment by a beneficiary leads to “substitution” (whereby an agent stands in as a steward “with discretion over an aspect of” the principal’s welfare)137 and “residual control” by the principal (who, “even after” entrustment, “reserve[s] the right to supervise fiduciary performance and, in appropriate circumstances, to take corrective action to remedy fiduciary malfeasance”).138 These circumstances create the need for fiduciary duties, which provide standards for determining that the agent’s “broad latitude to set discretionary policies . . . do[es] not transgress reasonable limits.”139 More broadly, Criddle argues that, whenever substitutional entrustment arises, fiduciary norms are necessary to prevent the principal from being dominated by the agent.140 For Criddle, whenever fiduciary norms apply, an agent is required to act both deliberately and deliberatively in order to advance the principal’s interests.141
Criddle contends that administrative governance exhibits each of the hallmarks of fiduciary norms. Because “the people as a whole” entrust authority to administrative institutions, these institutions “serve as stewards for the people.”142 Administrative agencies and public officials are therefore charged with advancing the interests of the people as a whole and are constrained by duties of loyalty and care that are owed to the populace.143 On Criddle’s logic, fiduciary norms apply to exercises of administrative rulemaking144 and adjudication,145 and more broadly to the structure of administrative institutions146 and the role-based responsibilities of administrators.147 The fiduciary theory thus supposes that administrative-governance norms are geared toward a “culture of justification,” in which “every exercise of power is expected to be justified” in terms of its advancement of the interests of the beneficiaries.148
Not everyone would accept the claim that administrative-governance norms reflect or aspire to such a culture of justification. Public-choice approaches see administrators as self-regarding, in contrast with the other-regarding orientation that the fiduciary theory presupposes.149 Pragmatic approaches are skeptical about the possibility of theorizing administrative governance, contending instead that the “structure and design of government institutions” should be evaluated “by the extent to which they help to mediate the conflicting values that make up our administrative legal culture.”150 Consequentialist approaches contend that the legitimacy of agency action should be appraised in terms of overall consequences: actions are legitimate to the extent that their benefits outweigh (or, in some versions, can be expected to outweigh) their costs.151
Adrian Vermeule’s work on administrative governance combines each of these types of arguments and provides perhaps the clearest contrast to the fiduciary theory. For Vermeule, the point of both agency actions and judicial review of those actions is the same: to secure the best outcomes, given institutional constraints.152 The kind of “optimization” that should guide administrative governance is a matter of “making incremental net-beneficial moves within the institutional space just up until the point at which the net benefits from further moves have diminished to zero—to the point where marginal benefits and costs are equal, or roughly equal as far as we can tell.”153 Furthermore, Vermeule contends that, in areas of administrative governance that are legal “black holes” (or “law-free zones that are themselves created by law”)154 and “grey holes” (or zones of law that “appear to comport with the rule of law but really do not”),155 there is no serious requirement that agencies justify their specific actions.156 Further, every standard-based stricture applicable to an administrative agency is, at least potentially, a legal grey hole.157
For Vermeule, then, the justification that an agency provides for its action is not categorically or fundamentally important to determining whether the agency’s action is legitimate. Of course, such public justification might be important in virtue of its tendency to promote good consequences. However, “[t]he question is always how much justification is enough [to achieve these consequences], and that is just another adjustable parameter.”158 By implication, some agency actions are legitimate even though they lack justification. For example, agencies acting within legal black or grey holes are presumptively legitimate.159 Moreover, Vermeule contends that categories of ostensibly abusive agency actions might nonetheless be legitimate if the costs of monitoring those abuses are not worth the benefits of preventing them.160 Likewise, for decisions about which the best outcome is uncertain, an agency action can be legitimate even though it ultimately lacks (consequentialist) justification, since the agency’s decisional process need not track the first-order justifications for its decisions.161
The divide between the fiduciary theory of administrative governance and an alternative like Vermeule’s parallels the contrast between the fiduciary theory of judging and its alternatives. In each of these domains, the fiduciary theory determines whether an agent has complied with a public-law norm in part by examining how the agent deliberates and how she (or it) justifies an action. By contrast, the alternative views are outcome centered, either deemphasizing deliberation or portraying it as only derivatively important. In Section II.A, we demonstrated the advantages of the fiduciary theory of judging by identifying several uncontroversial features of the norms for judges that the fiduciary theory could, but alternatives could not, easily explain. It is more difficult to perform this analysis for theories of administrative governance, since an advocate of a fiduciary theory (like Criddle) would presumably disagree with an advocate of an alternative view (like Vermeule) about the basic contours of the norms for administrative governance.162
At first blush, the answer to this question seems to be yes. A primary aim of judicial review of administrative decision making is to ensure that agencies comply with specific deliberative processes. Rulemakings must undergo “notice and comment” periods,163 which help ensure that agencies have conformed to the rules that should govern their conduct. Agency actions must either be supported by “substantial evidence,” or else not be “arbitrary” or “capricious.”164 In particular, the structure of so-called hard-look review165 seems geared to discern whether an administrative agency has deliberated in the right way. Under hard-look review, courts look to whether an agency has “examine[d] the relevant data and articulate[d] a satisfactory explanation for its action, including a rational connection between the facts found and the choice made.”166 As Criddle notes, this standard requires agencies to show that they
exercised due care by furnishing a full, contemporaneous administrative record, explaining in detail the rationale for their decisions, and validating departures from past decisions. Agencies’ explanations must address all salient aspects of a problem, including the relative costs and benefits of reasonable alternatives, and persuade courts that the final rule is not inconsistent with the empirical evidence before the agency.167
In other words, administrative-governance norms see results as important, but insufficient to determine the legitimacy of an agency’s actions. It is not enough for an agency to reach the correct answer or to arrive at the correct policy. The norms of administrative governance can be violated by deficient deliberative procedures alone.
On this construal, administrative-governance norms appear to impose freestanding deliberative requirements in exactly the same way as fiduciary norms. Consider a recent hypothetical example, contemplated by the Supreme Court in Judulang v. Holder,168 of an agency that enacts a policy of adjudicating cases before it by flipping a coin. According to the Court, an agency’s utilization of such a mechanism to decide which deportable aliens are eligible for discretionary relief would be inconsistent with the norms of administrative governance and “reverse[d] . . . in an instant” because it would be based on factors that are arbitrary and not even “loosely” tied to the broader purposes and concerns of the laws that the agency is charged with implementing.169 Moreover, such a policy would be illegitimate even if the results generated by the mechanism could have been justified on “other, more rational bases.”170
Yet the Court’s stated reasoning does not capture everything that is wrong with the use of such “[r]andomizing mechanisms”171 in administrative governance. Take the following example:
Dice Game: A statute instructs an agency to determine which claimants are eligible for discretionary relief. The agency (correctly) interprets the statute to favor the granting of discretionary relief in the majority of cases. The agency enacts a process of deciding claims through a roll of dice. Discretionary relief is denied if the dice roll turns up a prime number (2, 3, 5, 7, 11), and it is granted if the dice roll turns up composite numbers (4, 6, 8, 9, 10, 12).
The mechanism in Dice Game evades the Court’s stated criticisms in Judulang. The statutory goal of favoring the granting of discretionary relief can be said to influence the operation of the mechanism, since the interpretation of the dice roll makes a grant of relief more likely than a denial. Thus, the dice mechanism is (at least loosely) tied to the statutory purpose in favor of granting discretionary relief. Yet we think that the policy described in Dice Game is clearly illegitimate. This illegitimacy does not depend on whether (over the range of cases) rolling the dice yields the same pattern of decisions as would be secured by evaluating claims for relief on their merits. Part of the problem in Dice Game is explained by the deliberation sensitivity of administrative-governance norms. By definition, randomizing mechanisms like flipping a coin or rolling dice do not involve any adjudicative deliberation. This failure to deliberate seems objectionable in itself.
A critic of the fiduciary theory might deny that administrative-governance norms are actually deliberation sensitive, arguing that agency deliberation does not actually matter in the ways that administrative-law scholars posit that it does. For example, Vermeule argues that, in the grey holes that apply to large swaths of the administrative state, “hard look” review is illusory.172 In these areas, Vermeule contends, there is little actual scrutiny of how an agency decides how to act.173 This empirical contention, coupled with Vermeule’s conceptual claim that every standard-based administrative-governance norm is a potential legal grey hole, would seem to put pressure on the fiduciary theory by denying that an agency’s deliberation matters as such, rather than insofar as it happens to be connected to specific results. Of course, to refute the fiduciary theory, Vermeule would need to provide a link between the potential deliberation insensitivity of administrative-governance norms and the conclusion that these norms are, in fact and despite rhetorical evidence to the contrary, deliberation insensitive. This would amount to an argument that deliberation sensitivity is only real if it applies categorically to administrative-governance norms.174 However, if such an argument could be provided, then the deliberation sensitivity of administrative-governance norms might plausibly be denied.
In addition to his positive and conceptual arguments, Vermeule also makes the normative point that administrative-governance norms should be insensitive to deliberation. For Vermeule, deliberation-insensitive mechanisms can be expected to produce better consequences in a variety of contexts, including situations where first-order considerations do not unequivocally support any particular course of action.175 Along these lines, Vermeule would deny the Court’s hypothesis in Judulang, as well as the conclusion that we reach in Dice Game. According to Vermeule, when an agency is “focused entirely on the purposes of the laws it administers,” but “nonetheless reaches an uncertainty frontier at which first-order reasons for making choices in light of those purposes simply run out, yet choices must somehow be made,” randomization mechanisms (like flipping a coin or rolling the dice) “ought to be one perfectly acceptable mode of proceeding . . . .”176 Thus, the critic of fiduciary theory might grant that administrative-governance norms in the United States happen to be deliberation sensitive, but deny that they should be.
Under our framework, a fiduciary theory can explain administrative governance only if administrative-law norms impose standards of conscientiousness—that is, if not just any token of an agency’s conforming to a norm counts as complying with that norm. Here, too, fiduciary theory seems to describe extant administrative law. The main standard of judicial review of administrative decision making about matters of statutory interpretation—Chevron U.S.A., Inc. v. Natural Resources Defense Council, Inc.177—seems like a standard of compliance. Under Chevron, only some kinds of reasons qualify as good reasons. Chevron calls for review of an agency’s action based on the statute that delegates the relevant authority to the agency.178 As a result, the range of legitimate reasons for the agency’s action is set by the legislature, which (presumably) embodies the people’s objectives. An agency action that is not prompted or justified by this set of statutorily delineated reasons is illegitimate, regardless of whether that action is justified tout court.
The Dice Game scenario suggests how conscientiousness matters in administrative adjudication. As the Judulang Court hypothesized, the use of a randomization mechanism to resolve cases would be illegitimate even if the results of using such a mechanism were the same (on the whole) as those that would have been obtained through a legitimate decision-making procedure.179
Standards of conscientiousness also seem to apply to agency rulemaking. Recall that the intelligibility of “wrong kinds of reasons” problems in a domain provides strong evidence that the domain imposes standards of conscientiousness. “Wrong kinds of reasons” problems abound in administrative governance, and particularly in the agency rulemaking context.
The Supreme Court’s recent decision in Michigan v. EPA provides a real-world example of this phenomenon.180 At issue in this case was an EPA rule that regulated emissions of air pollutants from power plants. The EPA was authorized to make this rule by its mandate under the Clean Air Act to regulate if the agency finds that regulation is “appropriate and necessary.”181 In formulating its rule, the EPA did not consider the costs and benefits of regulation and, indeed, contended that these “should not be considered” when deciding whether power plants should be regulated under the hazardous air pollutants program.182 Yet the EPA’s regulatory-impact analysis (which the agency was required to prepare and which was not used in the rulemaking)183 estimated that the direct and ancillary benefits of the rule were between thirty-seven billion dollars and ninety billion dollars annually while the rule itself was estimated to require power plants to bear costs of $9.6 billion per year.184 Moreover, the EPA contended, while the agency did not consider costs in “first deciding whether to regulate power plants,” it would consider costs subsequently in implementing the rule.185
The Supreme Court struck down the EPA rule. Based on the “foundational principle of administrative law that a court may uphold agency action only on the grounds that the agency invoked when it took the action,”186 the Court held that the EPA rule was unlawful because of the agency’s failure to consider cost as part of its deliberations. According to the Court, it did not matter that the EPA’s regulatory-impact analysis showed that “once the rule’s ancillary benefits are considered, benefits plainly outweigh costs.”187 In other words, the agency’s failure to consider a relevant factor was sufficient to show that its action was illegitimate, independent of whether that factor actually would have supported the agency’s action.188
Taken at face value, Michigan v. EPA suggests that an agency’s compliance with administrative-governance norms turns on the reasons for which an agency acts, not merely the reasons that could justify its action. As such, the Court’s analysis invokes the “wrong kinds of reasons” problem: the reasons that motivated the EPA to promulgate the regulation diverged from the reasons that justified (or could have justified) its action, and this divergence ultimately compromised the legitimacy of the action.189 If “wrong kinds of reasons” problems arise in the administrative-law context, then administrative-governance norms appear to impose standards of conscientiousness. These conclusions are consistent with fiduciary theory.190
A critic of the fiduciary theory might deny that the norms of administrative governance impose standards of conscientiousness in this way. This denial might be stated at a very general level. Recall that the standards of conscientiousness contrast with agnosticism about compliance or the position that conforming to a norm is just as good as following it and that any route to conforming with a norm is just as successful as any other route. Agnosticism about compliance is an article of faith among many administrative-law scholars,191 although it is inconsistent with fiduciary theory192 and seems to be in tension with the holding of Michigan v. EPA.
More specifically, one could deny that there are conscientiousness standards for administrative governance in the following way, which parallels the Vermeule-inspired argument (described above) for denying that administrative-governance norms are deliberation sensitive. First, there are no conscientiousness standards in legal black and grey holes. The existence of such standards would conflict with the urgency and inscrutability that leads to their creation. To put this point less formally, it would be difficult to imagine the Court extending its reasoning in Michigan v. EPA to the military. Second, on Vermeule’s logic, every area of administrative law is potentially a grey hole. Third, every such grey hole is at least potentially agnostic about compliance. Fourth, conscientiousness standards are only real if they apply categorically within a domain.193 Thus, on this argument, the norms of administrative governance do not impose standards of conscientiousness, despite appearances to the contrary.
We do not think that this argument is sound. In particular, the first and second premises strike us as false, and the conceptual claim in the fourth premise seems unwarranted. Nonetheless, the argument is valid. Thus, it is at least plausible that the norms of administrative governance do not impose conscientiousness.
Vermeule might also offer a consequentialist argument for why administrative-governance norms should not have conscientiousness standards. In general, requiring conscientiousness makes it more difficult for an agent to comply with a norm and increases the costs of monitoring when an agent has complied with a norm. In a variety of contexts (especially those in which first-order reasons are indeterminate, yet some sort of decision must be made), it will be impossible for an agency to be conscientious. Vermeule might also contend that standards of conscientiousness produce no clear benefits.194 In most cases, the state of the world in which an agency acts for an inappropriate reason is exactly the same as the state in which that agency acts for an appropriate reason. By comparison, agnosticism about compliance can be expected to lower the costs of complying and of monitoring compliance, while producing the same or similar level of valuable activity. If what matters are the ways that an agency acts and not why it acts in those ways, then requiring agency conscientiousness seems both unnecessary and inefficient. Although we think this argument is inconsistent with core democratic values, it is at least plausible.
Administrative-governance norms seem robust in at least one sense: because they impose requirements to advance the interests of the public, the specifics of these requirements morph given changes in the circumstances of the people. Both Criddle and Vermeule would accept that the authority given to administrative agencies is and should be open-ended in this way,195 although some scholars of administrative law might not.196 Yet Criddle and Vermeule would disagree about the implications of these open-ended responsibilities. For Criddle, the requirement of public justification morphs along with changes in how the requirement that agency action advance the public interest is conceived.197 By contrast, on Vermeule’s theory, the open-endedness of agency responsibilities calls for de-emphasizing justificatory requirements, and perhaps abandoning them altogether.198 Vermeule sees such justificatory requirements as self-defeating: to require that an agency justify its action in terms of serving the public interest is to impede its capacity to serve those interests.
Whether the norms of administrative governance impose an updating requirement is a more contentious question. Recall that, if a norm imposes an updating requirement, living up to that norm requires an agent to seek out and respond appropriately to new information related to the advancement of the principal’s interests. Both of these requirements can be violated actually (if the agent does not seek out relevant information or does not consider it in making her decisions) or counterfactually (if the agent is disposed not to seek out new information, or would be disposed not to consider it if it were on offer).
Do administrative-governance norms impose an updating requirement? A fiduciary theorist like Criddle would likely say yes. Michigan v. EPA can be read to support the proposition that an agency’s failure to seek out or utilize relevant information in deciding how to act undermines the legitimacy of its action. Agencies could be understood under current law to have obligations to update and revise their rules in accordance with the public interest aims of the statutes that delegate rulemaking authority to them. Consider the Administrative Procedure Act (APA) petition process, which provides that “[e]ach agency shall give an interested person the right to petition for the issuance, amendment, or repeal of a rule.”199 If an agency wants to deny a petition, it also must furnish actual reasons for its denial.200
The APA further provides for judicial review to “compel agency action unlawfully withheld or unreasonably delayed,”201 suggesting that certain failures to update or revise regulations can be scrutinized by the judiciary. Indeed, the Supreme Court’s decision in Massachusetts v. EPA reinforces the claim that courts are willing to interrogate an agency’s decision not to regulate. In that case, the EPA had attempted not to decide whether greenhouse-gas emissions contribute to climate change. The Court found that the EPA’s inaction was not supported by a “reasoned justification.”202 New climate science essentially required the EPA to update and revise its regulatory approach to different kinds of pollutants.
To be fair, administrative law is equivocal on this point. Rarely do courts scrutinize an agency’s inaction or its failures to enforce, regulate, repeal, or amend its requirements.203 Even in Massachusetts v. EPA, the Courtacknowledged that the standard of its review of agency inaction is “highly deferential.”204 Furthermore, the state of administrative law on this point is not irrefutably consistent with the fiduciary theory. Although the fiduciary theory of administrative governance provides an argument for why agency inaction is problematic,205 a critic might contend that the fiduciary theory is revisionist because it attributes more significance to agency inaction than courts typically do. The fiduciary theory suggests that agency inaction should be scrutinized on par with agency action, but the latter is clearly policed much more stringently than the former.
Vermeule, for his part, would deny that the updating requirement applies to administrative-governance norms. Vermeule contends that, if the likely costs of inquiry exceed the likely benefits of the product of that inquiry, then an agency’s failure to seek out new information relevant to its decision does not affect the legitimacy of that decision.206 Vermeule would also likely point to the Court’s acknowledgment that the standards for reviewing agency inaction are “highly deferential”207 as evidence that failures of inquiry cannot, by themselves, compromise the legitimacy of an agency decision. Vermeule would also presumably deny that a violation of administrative-governance norms could be inferred from an agency’s use of an insensitive decision procedure, as suggested by his contention that randomization mechanisms for making agency decisions ought to be “one perfectly acceptable mode of proceeding.”208 Randomization mechanisms are, by definition, insensitive to new information, or any information for that matter. For Vermeule, as long as the benefits from utilizing an insensitive decision mechanism are expected to outweigh its costs, a decision made according to that mechanism is consistent with the norms of administrative governance.
Whether a fiduciary theory is viable in the domain of public administration depends on whether it is structurally compatible with the norms of administrative governance. Yet scholars dispute how to characterize the norms for administrative governance (and, in particular, whether these norms are deliberation sensitive, have standards of conscientiousness, and impose an updating requirement). Criddle would likely contend that each of these features characterizes the norms of administrative governance. We agree with this characterization, but perhaps that only means that we, like Criddle, are “romantics” about administrative law.209 Many scholars (particularly those of consequentialist, pragmatic, and public-choice stripes) would deny that these features characterize the norms of administrative governance, or that they should. These nonfiduciary views are not obviously incorrect or incomplete (as they appear to be in the domain of judging), and they can explain many of the same results that the fiduciary model explains without utilizing contested moral concepts or calling for greater policing of the intentions of political actors.
Our framework, then, leads to a modest conclusion: whether the fiduciary theory of administrative governance works depends on how one construes the fundamental norms of administrative governance. Absent a resolution of these issues, the fiduciary theory’s viability is an open question.
In a series of articles and books, Evan Fox-Decent and Evan Criddle have articulated a fiduciary theory of international law.210 Although Fox-Decent and Criddle focus on jus cogens and human rights, they contend that their insights apply more broadly to other topics in international law (such as jus ad bellum, belligerent occupation, and humanitarian law).211 Their fiduciary theory differs significantly from the historical invocations of the fiduciary concept in international law, in which fiduciary principles were prominently deployed to support unjust policies like expropriation and colonialism.212
For Fox-Decent and Criddle, the case for a fiduciary theory of international law arises out of the conjunction of two abstract principles of political morality, as well as a conceptual claim about political authority. The principle of nondomination holds that states must not “assum[e] arbitrary power over people’s legal and practical interests,” and the principle of noninstrumentalization holds that states must always treat those subject to their authority “as ends rather than mere means.”213 How can one person (or agent) exercise power over the legal and practical interests of another without dominating the other or treating her as an instrument? To answer this question, Fox-Decent and Criddle invoke the “fiduciary principle,” which “authorizes the fiduciary to exercise power on the beneficiary’s behalf but subject to strict limitations arising from the beneficiary’s vulnerability to the fiduciary’s power and his intrinsic worth as a person.”214 The fiduciary principle applies to all relationships in which one party “holds discretionary power of an administrative nature over the legal or practical interests of another party”215 and the latter party is vulnerable to the former party’s power (where vulnerability is defined as an inability “either as a matter of fact or law, to exercise the entrusted power”).216 The relationship between the state and the subject exhibits all of these hallmarks. According to Criddle and Fox-Decent, each state is therefore a fiduciary, and all those who are subject to its power are its beneficiaries. When a state’s action does not comport with fiduciary norms, the state breaches its fiduciary duties.
In what follows, we focus on Fox-Decent and Criddle’s discussion of human rights, which play a central role in their fiduciary theory of international law.217 For Fox-Decent and Criddle, respect for human rights is a “normative consequence of the state’s assumption of sovereign powers and [is] thus constitutive of sovereignty’s normative dimension. A state that fails to . . . respect human rights subverts its claim to govern and represent its people as a sovereign actor.”218 Human rights function to “protect persons subject to state power from domination and instrumentalization,” and the category of “genuine” human rights includes everything necessary to realizing nondomination and noninstrumentalization.219
Our framework suggests that Fox-Decent and Criddle’s fiduciary theory of human rights is problematic and, more broadly, that fiduciary norms are structurally incompatible with extant norms of international law. Our critique leaves open whether it would be a good idea to restructure international law to reflect fiduciary norms, since Fox-Decent and Criddle do not address this question.220
Consider the following scenario:
Solicitude: State A enacts a policy of not extraditing those within its territory to State B. The sole rationale for this policy is that State A wants to curry favor with State C (which is antagonistic to State B). Unbeknownst to the officials of State A, State B has a policy of torturing those in its custody. Thus, State A’s policy has the effect of protecting the human rights of those within its territory.
Does State A violate the norms of human rights law, specifically, the United Nations Torture Convention,221 in Solicitude? We think the answer is clearly no. Regardless of whether State A’s action breaches norms applicable to domestic policymaking, it does not violate any norm of international human rights.
Yet the fiduciary theory would reach the opposite conclusion. State A’s policy runs afoul of the principle of nondomination: any protection of human rights would be based on a factor that is not directly related to the interests of the beneficiaries, namely whether the policy pleased State C. Moreover, State A treats human rights as a “mere means” to curry favor with State C, in violation of the noninstrumentalization principle. This discrepancy between what fiduciary duty requires and how human rights norms actually work suggests that fiduciary theory is inapposite to this area of law.
In the remainder of this Section, we elaborate on this argument by analyzing in more granular detail whether the core features of fiduciary norms are compatible with the norms of international human rights law.
Human rights norms govern state behavior. The obligations they impose primarily concern how states behave.222 They do not, in general, impose freestanding requirements regarding how a state must deliberate.223 A state complies with human rights norms to the extent that human rights are realized in that state. On this logic, state action (or inaction) that has the effect of protecting human rights does not violate human rights norms, even if the action (or inaction) arises out of a deliberative process that is obviously deficient.
Of course, deliberation matters when determining whether human rights norms have been breached. For example, whether specific behavior counts as a violation of human rights can often depend on how specific actors deliberate.224 However, such examples indicate that the manifestation requirement applies to human rights norms: how a state (or other actor) deliberates is relevant only to the extent that this deliberation is manifest in behavior. In other words, human rights norms appear sensitive to deliberation in the same way as criminal norms, but not in a manner analogous to fiduciary norms.
Consider the human right to be free from torture. Article 3 of the United Nations Torture Convention states that no party to the Convention “shall expel, return . . . or extradite a person to another State where there are substantial grounds for believing that he would be in danger of being subjected to torture.”225 Compliance with this norm is primarily a matter of behavior. To be sure, deliberation matters to some questions related to compliance.226 However, deliberation is only relevant in virtue of a state’s behavior; the Convention does not establish freestanding deliberative requirements in the way that fiduciary norms do.
There are several good reasons why human rights norms are not deliberation sensitive. First, the most central goods secured by international law seem capable of being achieved solely through the realization of outcomes. One of the main aims of international law is to facilitate coordination in a way that “protects political communities from external aggression” and “protects citizens of those communities from domestic barbarism.”227 Achieving this aim does not necessarily require concern with how agents deliberate. Coordinated activity by agents who exhibit wildly divergent and even incompatible patterns of deliberation (for example, through an “incompletely theorized agreement”) would be sufficient to achieve the main purpose of international human rights law.228 A second explanation is the inscrutability of state intentions. It is very difficult to establish how a state deliberates, especially given the lack of shared ends and (on many issues) shared understandings of basic concepts that characterize the field of human rights. A third explanation is based on efficacy. Getting states to behave in ways that protect human rights is vexing enough.229 If human rights norms were deliberation sensitive, then the costs of complying with these norms would be far more burdensome (and compliance would be far more invasive of state sovereignty) than if compliance were construed primarily in behavioral terms. Thus, there are significant costs, but few obvious benefits, to imposing freestanding deliberation requirements on human rights norms.
Here, too, the answer seems to be no. To say that human rights norms primarily govern behavior is to imply that any way of conforming to these norms, however motivated, counts as compliance.230 As one commentator puts it, compliance with human rights norms “may thus take different forms, from coincidence to convenience (so as to cash in on potential incentives), obedience (so as to avoid sanctions) or internalization (when attitudes and beliefs change in line with international norms and become constitutive behaviour).”231 A state’s conscientiousness or lack thereof is generally irrelevant to establishing whether it has complied with human rights norms.232
The fiduciary theory reaches the opposite conclusion: a state whose behavior conforms to the requirements of human rights norms but whose motivations are inappropriate would nonetheless breach those norms. In Solicitude, for example, State A’s motivation for protecting the human rights of its population is not connected to the justification for human rights. Moreover, State A seems to actively violate Fox-Decent and Criddle’s principle of noninstrumentalization, since it protects human rights merely as a means of currying favor with State C. Fiduciary theory reaches an implausible conclusion in the Solicitude scenario, then, because it imposes standards of conscientiousness that do not characterize existing human rights norms.
There are also good reasons for international law’s general agnosticism about compliance. Evidentiary difficulties in assessing the reasons for state action weigh in favor of determining compliance through purely objective considerations.233 Moreover, given the divergent ends of states in international law, states can be expected to differ about which considerations justify the protection and promotion of human rights. As such, requiring that the reasons motivating a state’s action are congruent with the reasons that legally justify its action would invite discord in an already contentious policy space. Furthermore, requiring congruence would secure no definite benefit, since what matters most in the international sphere seems to be that human rights are protected, not why they are protected.
The notion that human rights norms impose standards of conscientiousness is also highly revisionist. Such standards would be anomalous on nearly every extant theory of international law. Other than Criddle and Fox-Decent, theorists of international law routinely presume that a state’s conforming behavior is sufficient to establish its compliance.234 The application of conscientiousness standards would also be inconsistent with the understandings of many human rights lawyers and activists.235 If international human rights norms imposed standards of conscientiousness, many fundamental tenets of international law (such as reciprocity and concern for reputation) would be unjustifiable or inadequate.236
Human rights norms are clearly robust in that the responsibilities that they impose may morph, given changes in the needs or interests of people.237 For example, the Torture Convention requires signatories to both criminalize torture and take “such measures as may be necessary” to establish criminal jurisdiction over instances of torture committed by its nationals or within its territories.238 The open-endedness of human rights norms is consistent with, but does not uniquely support, the fiduciary theory. A variety of other theories of human rights might also be able to capture this open-endedness.
Fiduciary theory would be directly supported if human rights norms imposed an updating requirement. Yet the updating requirement is inapposite in the context of international human rights law. While compliance with human rights norms often requires a state to investigate human rights abuses,239 these requirements are not capable of purely counterfactual violation: a state complies with human rights norms if it investigates abuses, even if it would not have done so under other, unrealized circumstances. Likewise, whether a state’s decision procedure is insensitive to the fruits of its investigation cannot, by itself, establish a breach of human rights norms, since deliberation has no independent relevance in determining a state’s compliance with these norms. In the Solicitude scenario, the flimsiness of State A’s commitment to protecting the rights against torture of those within its territory might be a ground for criticism, but it does not by itself indicate a violation of the Torture Convention.
Therefore, while the requirements imposed by human rights norms are robust in one sense, they are not robust in all of the ways that fiduciary norms are.
To summarize, fiduciary norms are structurally incompatible with the extant norms of international human rights law. Moreover, this incompatibility arises in other areas of international law to which Fox-Decent and Criddle apply the fiduciary theory.240 The fiduciary theory of international law thus does not provide an accurate picture of human rights law, or international law more generally.
Our arguments about the incompatibility between fiduciary norms and international law are empirical, rather than conceptual. It is possible to imagine a version of the international legal order that enshrines a robust “culture of justification” like the one that Criddle sees at the core of administrative law,241 although there are good reasons why (given our existing institutions) international-law norms do not police deliberation or impose standards for compliance or robustness. Nor do we think the search for a moralized theory of international law is futile.
The structural incompatibility we identify here, however, directly challenges whether any fiduciary theory can capture the extant structure of international-law norms. Our analysis suggests that one cannot straightforwardly derive the fundamentals of international law from the principles of nondomination and noninstrumentalization. Not every type of authority relationship (or every domain of public law) is concerned with domination and instrumentalization in the same way. Perhaps the world would be a better place if the same rigorous culture of justification applied to the international legal realm. But that is not the world we live in, nor is it the one that Criddle and Fox-Decent purport to describe.242
This Essay shows the promise and limits of fiduciary political theorizing, a burgeoning field of legal scholarship. One aim has been to provide new insights into the features that distinguish fiduciary norms from other types of legal norms, like the norms of contract, tort, and criminal law. Another aim has been to harness these insights into a framework for determining when fiduciary political theory is likely and unlikely to illuminate a public-law domain.
We demonstrated the utility of our framework by applying it to three case studies. The fiduciary theory of judging is successful, and likely superior to alternative theories of judging, in explaining and justifying the norms for judges. The viability of the fiduciary theory of administrative governance turns on fundamental questions about the structure of administrative law—questions that have not been satisfactorily resolved. Finally, the fiduciary theory of international law appears incompatible with the structure of the governing norms in human rights law and elsewhere in international law. These case studies demonstrate that fiduciary political theory is most appropriate in domains of law that enshrine what Criddle calls a “culture of justification” and least appropriate in areas where coordination (rather than justification) is what matters.
We consider the revival of fiduciary political theory to be one of the most exciting developments in legal and political philosophy in recent years. Fiduciary political theory offers common ground from which legal scholars and legal theorists can address many perennial questions about justice and politics. We hope the framework offered here clarifies the ambitions of this project and aids fiduciary political theorists in their efforts to explain and improve our lives under law.