There is a prominent tradition, sometimes called “contractarianism,” that reduces fiduciary obligations to promissory obligations.Footnote 1 Debates over the contractarian approach raise questions about the nature of fiduciary obligations. Does the structure of fiduciary obligation ultimately differ from that of promissory obligation? Contractarianism succeeds as an account of fiduciary obligation only if fiduciary and promissory obligations have the same underlying structure.
We argue that there is at least one important structural difference between fiduciary obligations and promissory obligations: fiduciary obligations have intentional requirements that promissory obligations do not necessarily have. A fiduciary obligation characteristically shapes the fiduciary's deliberation vis-à-vis her beneficiary in a particular way. A fiduciary whose deliberation is not shaped in this way does not live up to her fiduciary obligation, no matter what else she does. Thus the fiduciary duty of loyalty does not admit of what we call accidental compliance: it is not satisfied by actions done for inappropriate or illicit reasons.
By contrast, promissory obligations do not have the same intentional requirements. There is nothing incoherent about inadvertently keeping a promise. Your promise need not shape your deliberation in any particular way in order for your action to count as promise-keeping. Therefore, intentions matter to the satisfaction of fiduciary obligations in a way that they do not matter to promissory obligations. You can accidentally keep a promise, but you cannot accidentally be loyal.
Our analysis proceeds in three parts. Section II explores the asymmetry between promissory obligations and the fiduciary obligation of loyalty with respect to accidental compliance. As normative concepts, loyalty does not admit of accidental compliance, while promise-keeping does.
This asymmetry indicates that loyalty has a unique intentional dimension, which we describe in more detail in Section III. There, we provide a model of the intentional dimension of fiduciary loyalty (which we call the “shaping” account) and demonstrate its superiority over two alternatives, the “adopting” and “aiming” accounts.
Section IV utilizes the “shaping” account to critique two prominent strategies that reduce fiduciary obligations to other normative phenomena. The contractarian approach, which conceives of fiduciary obligations as a species of contractual obligations, is misguided. Fiduciary obligations cannot be reduced to promissory obligations because the former have at least one structural feature that the latter do not have. A second strategy, which we term the “proscriptivist” approach,Footnote 2 reduces loyalty to the legal “no conflict” and “no profit” rules that constrain fiduciaries. This approach also fails to capture fiduciary obligations as normative phenomena. One can successfully (but accidentally) comply with these proscriptive rules, while fiduciary loyalty does not admit of accidental compliance. In other words, some actions that do not violate these rules nevertheless fail to be loyal.
II. ACCIDENTAL COMPLIANCE: AN ASYMMETRY BETWEEN PROMISSORY AND FIDUCIARY OBLIGATIONS
Subsection A first lays out some preliminaries of our analysis. Subsection B argues that accidentally complying with a promissory obligation is plausible, while accidentally complying with a fiduciary obligation is not.
Before offering our argument, we should clarify our terminology and methodology, then motivate our inquiry. Our main aim is to account for promises and fiduciary obligations conceptually, as features of our “normative landscape.”Footnote 3 Although we offer doctrinal examples in our discussion, it is entirely possible that the juridical treatment of these phenomena deviates from their more fundamental normative structure. Doctrine sometimes only hints at the underlying normative terrain. There are rule-of-law values and pragmatic considerations of institutional design that weigh against the perfect correspondence between legal doctrine and the underlying normative phenomenon. To wit, it is routine to think of fiduciary law as exhorting moral behaviorFootnote 4 or involving the underenforcement of moral norms.Footnote 5 Contract law similarly invokes norms of promise-keeping without necessarily being coextensive with promissory morality. One such example is that promises can be unilateral and gratuitous, but the law does not enforce promises without bilateral consideration.Footnote 6
Ultimately, the prospect of a discrepancy between legal doctrine and the core features of normative phenomena should not be surprising. However, there would be plausible reason for concern if the juridical instantiation of the normative phenomena did violence to the underlying normative phenomena—for example, if consistently adhering to the norms of contract made it difficult or impossible for one to be a promise-keeper.Footnote 7 Our analysis supposes that there is and should be a connection between the institutional and noninstitutional versions of certain normative concepts. However, we do not offer a sustained defense of this assumption nor do we think that this relationship between normative and juridical phenomena need be one of identity. While doctrine only hints at an underlying normative terrain, there is a normative terrain of obligations that arise out of relationships of entrustment and vulnerability. Our project is to explore this terrain.
Promises and fiduciary relationships both characteristically generate obligations. Fiduciary obligations arise out of fiduciary relationships, such as the trustee-beneficiary relationship or the attorney-client relationship. While there are many aspects to fiduciary obligation, the core duties are generally seen as loyalty and care. The duty of loyalty is usually thought to require the fiduciary to pursue the interests of her beneficiary above her own.Footnote 8 The duty of care requires acting with reasonable diligence and prudence on behalf of the beneficiary.Footnote 9 Most fiduciary scholars identify loyalty as the paradigmatic fiduciary obligation;Footnote 10 we adopt this convention and focus our attention on the criteria for and structure of fiduciary loyalty.Footnote 11
Many commentators analyze the connection between promissory and fiduciary obligations in terms of formation, content, or available remedies. In what follows, however, we mostly ignore these issues. Instead, we focus on satisfaction, or the conditions under which someone can be said to have lived up to or discharged the obligations that apply to her. Thus we use the phrase “A kept her promise to B” instead of the awkward (but more accurate) “A satisfied her promissory obligations to B regarding Φ’ing.” Likewise, we take the phrase “A acted loyally toward B” as equivalent to “A, in Φ’ing, satisfied her duty of loyalty toward B.”
Judgments about satisfaction and nonsatisfaction can reveal the contours of promissory and fiduciary obligations in the same way that evidence about “reactive attitudes” (like resentment or indignation) are said to reveal the content of moral principles.Footnote 12 Moral prohibitions are characteristically associated with reactive attitudes, which both respond to and criticize the violation of moral norms. Thus, examining when people react with resentment or indignation to the conduct of others provides some insight into what kinds of actions are morally prohibited. Similarly, violating an obligation invites a distinctive form of criticism of the violator.Footnote 13 That a particular mode of criticizing someone's actions seems appropriate suggests that these actions violate her obligations. If such criticism seems inappropriate, then it seems plausible to say that she has lived up to her obligation.Footnote 14
B. Conformity and Compliance; Or, Why Loyalty Is Not Promise-Keeping
The requirements for satisfying a fiduciary obligation differ from the requirements for satisfying a promissory obligation. In particular, there is an asymmetry between promise-keeping and loyalty regarding the possibility of what we call accidental compliance.
To demonstrate what accidental compliance means, consider Joseph Raz's distinction between conformity and compliance. For Raz, a person conforms to an obligation if she does what the obligation requires—that is, if her behavior matches the pattern specified in the obligation. By contrast, someone complies with an obligation when she acts for it. Compliance, then, not only requires conforming behavior but also has a further intentional dimension.Footnote 15 We use the term “accidental compliance” to describe a subset of actions that conform to but do not comply with an obligation; namely, ones where an agent's behavior matches the pattern specified by the obligation, but the obligation does not have the requisite influence on her deliberation to count as compliance.
We use the term “accidental” in the sense that some philosophers use the notion of deviant or “wayward” causation.Footnote 16 One way to interpret Gettier problems,Footnote 17 for example, is as ruling out that justified true belief entails knowledge in cases where the connection between the justification and the formation of the belief is “accidental,” in particular, where an agent's arrival at a correct belief does not proceed through the justification for that belief. Such cases are often said to exhibit “deviant” causation. “Accidental” for us has a similarly broad usage. It describes both the unwitting agent who does not specifically aim to comply with the obligation, as well as cases when an agent's conforming behavior reflects bad faith toward either the principal or the obligation itself. We focus on the case of the unwitting agent, since if the asymmetry is reflected here, then it is presumably also reflected in the case of an “illicit” agent.
Promissory obligations seem capable of satisfaction through accidental compliance. To see how, consider the following example:
Lucky Payer: Abner promises to pay Barry five dollars on Tuesday. On what Abner believes to be Wednesday, he gives over the contents of his wallet (which contains a single bill that Abner believes to be a one-dollar bill) to Barry. Unbeknownst to Abner, it is actually Tuesday, and the bill in Abner's wallet is actually a five-dollar bill.
We think Abner has kept his promise even though he didn't mean to do so. Accidental compliance with a promissory obligation seems intelligible because promise-keeping is typically about conformity rather than compliance.Footnote 18 If you promise to Φ, then you usually keep your promise if your behavior counts as Φ’ing. In general, how your behavior comes to overlap with the content of your promise does not matter to the question of whether you have kept your promise.
This is not to say definitively that promise-keeping lacks an intentional dimension. Intentions might be part of the content of a promise. If someone promises to look after your best interests in making a decision, then determining whether she has kept her promise will require some reference to her intentions as well as her behavior. Rather, our point is that promises are not always like this. They do not necessarily or routinely have intentional requirements. Where they do, they do so as part of the content of what is promised rather than because of a structural feature of promises.
If promises do not necessarily have an intentional component, then promissory obligations can be satisfied through either conformity or compliance. Consider the Lucky Payer case again. There are many ways to criticize Abner. We might think him inconsiderate, unreliable, a lout. But we do not think it is appropriate to say that Abner has violated his core promissory obligation. Abner's actions seem to be promise-keeping because his behavior (paying five dollars on Tuesday) matches the content of his promise to Barry.
Our argument that accidental compliance satisfies promissory obligations is largely intuitive, although it has theoretical plausibility.Footnote 19 However, some theorists might dispute this argument and contend instead that there is an irreducibly intentional dimension to promise-keeping. For example, the defender of an instrumental account of the normativity of promises (which sees promises as significant insofar as they serve some other basic interest) might see an intentional dimension to promise-keeping as necessary to facilitate the central values served by promises. Something like this instrumentalist argument undergirds David Owens's contention that nonconscientious conformity with the terms of one's promise is not really promise-keeping.Footnote 20
Consider also a noninstrumentalist account of promises such as T.M. Scanlon's, which establishes the normativity of promises in terms of the “value of assurance.”Footnote 21 On such an approach, it is unclear whether behavior that conformed to but did not comply with the terms of a promise would be compatible with Scanlon's account of the principle of fidelity (what he calls “Principle F”), which explains both how promises can provide assurance and why they have normative significance.Footnote 22 Along these lines, Michael Pratt sees promises as characteristically forming a valuable kind of relationship, and “[m]erely to do as you said you would do is not to effect a relationship of any special value with me, except insofar as your doing what you said counts as the discharge of a duty that you owed to me.”Footnote 23
In addition to the intuitive and theoretical considerations in favor of our case, we think it is possible to answer the arguments that posit an intentional dimension to promise-keeping. However, one might say that it takes a theory to beat a theory, and we do not have a fully developed theory of promising to offer here. Thus we are reluctant to assert that those who deny the intelligibility of accidental promise-keeping (and by implication attribute an irreducibly intentional dimension to promissory obligations) are incorrect. Even if there is an intentional dimension to promise-keeping, though, our main claim about the asymmetry between promissory and fiduciary obligations can go forward. The asymmetry would exist (and the reductionist strategies we discuss in Section IV would still be problematic) if the intentional dimension of promise-keeping differed from the intentional dimension of loyalty.
In contrast with promissory obligations, the fiduciary's obligation of loyalty is not satisfied by accidental compliance. In other words, it does not seem possible to act loyally by accident. Consider the following hypothetical example:
Dartboard Trustee: Tammy Trustee is considering three courses of action on behalf of Bernice Beneficiary. Through her investigation, Tammy has learned that Option 1 would best advance Bernice's interests, while Options 2 and 3 would substantially set back Bernice's interests. To decide which course of action to take, Tammy fastens three sheets of paper on a dartboard, marked 1, 2, and 3. Tammy will undertake whichever course of action corresponds with the sheet on which the dart lands. Tammy's thrown dart hits the sheet marked 1, which corresponds to Option 1. Because of this, Tammy undertakes Option 1 on Bernice's behalf.
Has Tammy Trustee acted loyally toward Bernice Beneficiary? We think not. However, the Dartboard Trustee case parallels the Lucky Payer in several respects. In both scenarios, the protagonist's pattern of behavior conforms to the substantive requirements of the obligation. Abner takes the course of action that a promise-keeper would, and Tammy Trustee takes the course of action that a loyal trustee would. Yet both cases involve a deviant form of causation. There is a missing or nonstandard linkage between the protagonist's deliberation and his or her behavior. Yet we reach different verdicts: Abner keeps his promise, but Tammy Trustee does not act loyally.
There is doctrinal support for the view that loyalty has an intentional dimension, that a fiduciary does not satisfy her duty of loyalty by accidentally complying with it. As one court put it, the heart of loyalty is “an attitude, not a rule.”Footnote 24 Another court found that a corporate director violated his fiduciary duty of good faith by “intentionally act[ing] with a purpose other than that of advancing the best interests of the corporation.”Footnote 25
These doctrinal tidbits raise a more fundamental issue: namely, that the notion of loyalty seems to have an intentional dimension “off the shelf.” Loyalty is in part a term used to describe a set of behaviors. Certain patterns of behavior can by their very nature be incompatible with acting loyally.Footnote 26 However, loyalty also has an irreducibly intentional component. As Simon Keller notes:
The fact that you act towards something in the way that you would if you were loyal to it does not establish that you really are loyal to it. Whether or not you are loyal to something depends not only on what you do, but on how you are motivated.Footnote 27
One cannot, then, unwittingly act loyally. Unwitting action is action done for no reason, and the unwitting agent would fail to act for the reasons (whatever they are) that ground loyalty. “Illicit” loyalty is also incoherent. If your behavior on behalf of a principal mirrors that of a loyal fiduciary, but you are motivated by a desire to subvert the interests of the principal, then you do not act loyally. Otherwise, the double agent who feigns allegiance to one principal in order to act on behalf of another would act loyally toward both.
Accordingly, there is an asymmetry between promise-keeping and loyalty. Acting loyally necessarily has an intentional dimension to it. As such, accidental compliance is insufficient to satisfy the fiduciary obligation of loyalty. Keeping a promise, by contrast, does not necessarily have an intentional dimension. Someone can keep her promise when her behavior conforms to its terms regardless of her intentional stance. Alternatively, if both promises and loyalty have intentional dimensions, then the intentional requirements for acting loyally are different from (and more exacting than) those for keeping a promise.
III. THE INTENTIONAL DIMENSION OF FIDUCIARY LOYALTY
In this section, we describe the intentional dimension of fiduciary loyalty. Subsection A articulates a view that we term the “shaping” account. The shaping account both explains the asymmetry regarding accidental compliance that we identified in Section II and hones in on the structural differences between fiduciary and promissory obligations. Subsection B compares the shaping account to two extant candidates, which we term the “adopting” and the “aiming” accounts. The shaping account captures the strengths of these alternatives without inviting the objections against them.
A. The Shaping Account of Loyalty
Just what are the intentional requirements for fiduciary loyalty? We contend that fiduciary loyalty has a characteristically “shaping” function: in deliberating, a loyal fiduciary robustly attributes nonderivative significance to her beneficiary's interests.Footnote 28 The exact shaping required for loyalty varies across fiduciary contexts. However, if a fiduciary's deliberation is not shaped in appropriate ways, then she does not satisfy her obligation of loyalty.
The core of the shaping account is the robust attribution of nonderivative significance to the beneficiary's interests. That a particular course of action would affect the beneficiary's interests must matter to the fiduciary's deliberation. It is, of course, possible to assign significance to a consideration without assigning it preemptive significance or incorporating it into one's ends. This is the crux of Rawls's notion of seeing people as “self-originating sources of valid claims.”Footnote 29 One need not take the internal point of view regarding the claims of another in order to recognize the validity of those claims. Further, the question of a claim's validity is distinct from the question of its weight. The shaping account requires that the beneficiary's interests necessarily matter to the fiduciary's practical deliberation, but it does not impose any particular weight that these interests must have (beyond having priority over the fiduciary's self-interests).
Moreover, the beneficiary's interests must matter to the fiduciary solely in virtue of their connection with the beneficiary. The importance of attributing nonderivative significance can be seen in the following example adapted from Bernard Williams.Footnote 30 After a shipwreck, a number of people are drowning, including an agent's spouse. Various impartial moral schemes might justify the husband's saving his wife rather than a stranger. For example, partiality towards one's wife in a life-and-death scenario might contribute to the impartial value of marital relationships. Suppose that the agent saves his wife because of the impartial justification for spousal favoritism rather than because she is his wife. Here the husband's pattern of behavior (i.e., rescuing his spouse) conforms to the demands of spousal loyalty because the husband assigns significance to his wife's interests. Yet it is difficult to say that the husband has acted loyally. This difficulty arises because the husband attributes only derivative significance to his wife's interests. Her interests matter only in virtue of the broader utilitarian justification for marital partiality. The husband would have, in Williams's memorable phrase, “one thought too many,” since “it might have been hoped by someone (for instance, by his wife) that his motivating thought, fully spelled out, would be the thought that it was his wife, not that it was his wife and that in situations of this kind it is permissible to save one's wife.”Footnote 31 Likewise, on the shaping account, loyalty requires not only that the beneficiary's interests matter, but also that these interests matter because of their connection with the beneficiary and not in virtue of some other consideration.
Shaping also 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. In other words, the beneficiary's interests shape the fiduciary's deliberation only if the fiduciary is disposed to revise her deliberation in accordance with changes in the beneficiary's interests. This robustness requirement rules out cases of accidental compliance in which the beneficiary's interests (formulated in a particular way) happen to matter to the fiduciary but would not matter if formulated in any different way. Loyalty is more robust than this kind of accidental compliance would allow.
Thus our shaping account posits three elements for fiduciary loyalty.Footnote 32 A fiduciary acts loyally toward a beneficiary only if (1) the beneficiary's interests matter to the fiduciary's deliberation; (2) these interests matter solely because they are the interests of the beneficiary rather than for some other reason; and (3) the fiduciary is disposed to revise her deliberation based on changes to the beneficiary's interests.Footnote 33
The shaping account explains why Tammy Trustee does not act loyally, even though her pattern of behavior matches that of a loyal trustee. Tammy violates element (1) because Bernice's interests do not matter to Tammy's deliberation at all. Tammy's deliberation is structured by the vicissitudes of a dart throw, which has nothing to do with Bernice's interests. This account also explains why the double agent does not act loyally toward her false principal. Although her behavior otherwise mirrors that of a loyal agent, the false principal's interests only matter to the double agent derivatively in virtue of their capacity to advance the true principal's interests. This violates element (2).Footnote 34
Why must loyalty shape a fiduciary's deliberation? We think the answer lies in a variety of aspects of paradigmatic fiduciary relationships. Fiduciary relationships characteristically involve discretionary powers and high levels of entrustment. Given imbalances of expertise and power, these relationships also make it difficult for the beneficiary to monitor the fiduciary's actions. As a result, the beneficiary in a fiduciary relationship is subject to the predation of her fiduciary in a way that people generally are not. Each of these aspects (discretion, trust, vulnerability, difficulties of monitoring and accountability) contributes to the explanation of why loyalty has an intentional component. Given a beneficiary's vulnerability to the fiduciary, the beneficiary can be expected to be concerned with not only the fiduciary's behavior but also her faithfulness. Ensuring that the beneficiary's interests shape the fiduciary's deliberation in a certain way is essential to ensuring that the beneficiary will not be exploited or have unnecessary risks imposed upon her.
This functionalist explanation for the connection between shaping and loyalty makes the shaping account ecumenical. The requisite weight of the beneficiary's interests on the fiduciary's deliberation will vary across fiduciary contexts in virtue of the beneficiary's vulnerability or level of entrustment. The greater the discretion, trust, or vulnerability, the tighter the intentional requirements for acting loyally. Thus the shaping account can explain why obligations of loyalty are calibrated differently across different kinds of fiduciary relationships, which is often said to be a desideratum of any viable theory of fiduciary obligations.Footnote 35
B. Alternatives: Adopting and Aiming
The shaping account is one of many ways to specify the intentional dimension of fiduciary obligations. Two alternative proposals are the “adopting” account and the “aiming” account. Both of these alternatives recognize that fiduciary obligations have an intentional dimension, so both are on the right track. However, both alternatives are more demanding than the shaping account, and as a result, each faces difficulties that the shaping account avoids.Footnote 36
The adopting account articulated by Arthur Laby (among others) posits that “the irreducible core of the fiduciary relationship is the fiduciary's obligation to adopt the principal's goals, objectives, or ends.”Footnote 37 Laby formulates the adopting account in Kantian terms. Fiduciary loyalty is open-ended, so it is a species of what Kant calls “imperfect” duties.Footnote 38 Disloyalty is a similarly open-ended inquiry, involving the fiduciary's “adopting of an end indifferent to, or in opposition to, the obligatory end”—that is, the ends of the beneficiary.Footnote 39 The adopting account can get the intuitively right answers in both the Dartboard Trustee and double agent cases. In neither case does the protagonist satisfy the obligation of loyalty because neither protagonist adopts her beneficiary's ends as her own.
We see two main problems with the adopting account, neither of which seriously wounds the shaping account. The first problem is that the adopting account cannot easily recognize the limits of fiduciary obligations. Consider Lord Brougham's statement that:
[The] advocate, in discharge of his duty, knows but one person in all the world, and that person is his client. To save that client by all means and expedients, and at all hazards and costs to other persons . . . is his first and only duty; and in performing this duty he must not regard the alarm, the torments, the destruction which he may bring upon others.Footnote 40
One way to interpret the adopting account is as generalizing Brougham's dictum, substituting “fiduciary” for “advocate” and “beneficiary” for “client.” As Laby sees it, in adopting the ends of the beneficiary, the fiduciary forsakes the ends of anyone else (including herself).Footnote 41 Yet this reading of the adopting account would have the implausible implication that loyalty could justify or require the fiduciary's adoption of the beneficiary's morally illicit ends.
To see why, imagine an attorney for Tony Soprano. Tony has a variety of ends, most of them illicit. Attorneys characteristically have fiduciary obligations toward their clients, including obligations of loyalty. On the adopting account, Tony's attorney could act loyally toward Tony only by adopting Tony's illicit ends. We find this implication implausible. Loyalty is often taken to have normative force: that an action is required as a matter of loyalty can change the deontic status of doing it.Footnote 42 Yet Tony's attorney would not have any reason to further Tony's illicit ends. The generally applicable prohibitions on this kind of facilitation apply to efforts on Tony's behalf.Footnote 43
If fiduciary loyalty could justify fiduciaries in adopting the illicit ends of their principals, then the fiduciary relationship would implicate a form of bootstrapping. By forming fiduciary relationships, one could generate moral permissions or requirements to act in otherwise impermissible ways.Footnote 44 Bootstrapping is problematic on a variety of theoretical grounds, not least of which is that it denies the overridingness of moral verdicts.Footnote 45 Loyalty is a form of partiality, a way of treating someone's interests differently from the (potentially conflicting) interests of others. To the extent that partiality can be justified, this justification has limits. Generally applicable principles of partiality seem capable of requiring or forbidding actions that are otherwise optional but not of permitting or requiring actions that are otherwise morally prohibited.Footnote 46 Thus, if loyalty is taken to have normative significance, then the adopting account (at least as Laby formulates it) cannot easily recognize the limits of loyalty. This account cannot explain why loyalty does not provide a justification for Tony Soprano's attorney to adopt her client's illicit ends.
The shaping function avoids these difficulties. One can attribute significance to a beneficiary's interests (in accordance with element (1) of the shaping account) without adopting those interests as one's own. Of course, adopting the ends of a beneficiary is sometimes an effective way of ensuring that a fiduciary makes the kinds of attributions constitutive of loyalty. It is sometimes easier to see how to advance the ends of another when you take those ends to be your own. But taking this stance is not a necessary condition for acting loyally.Footnote 47 Moreover, one can attribute nonderivative significance to a consideration without seeing it as preeminent. On the shaping account, then, Tony Soprano's lawyer can act loyally and within the bounds of partiality by taking Tony's licit ends to have significance in virtue of their being his ends; taking Tony's illicit ends to be normatively inert in virtue of being illicit, based on general constraints concerning partiality; and being disposed to advance Tony's ends if they were to become licit. The legal limits on what loyalty can permit or require thus cohere with loyalty's deeper normative structure.
A second (and more fundamental) problem with the adopting account is that it admits of accidental compliance. Laby's version of the adopting account is concerned primarily with the requirement that the fiduciary adopt the beneficiary's ends. Laby does not discuss the process by which this adoption must occur.Footnote 48 Yet this leaves the adopting account susceptible to fulfillment through deviant causation—that is, to cases where a fiduciary adopts a beneficiary's ends in a way that does not seem to exhibit loyalty toward the beneficiary. Say that a fiduciary and beneficiary are both members of a highly demanding religion. By virtue of their common membership in this religious group, the fiduciary might adopt the beneficiary's ends. However, the beneficiary would not figure necessarily in this adoption story: that these ends were the ends of the beneficiary would not explain why the fiduciary adopts them. This scenario does not indicate loyalty to the beneficiary so much as loyalty to a group of which the beneficiary happens to be a member. Yet Laby's version of the adopting account cannot capture this result. On his view, the fiduciary who adopts the beneficiary's ends based on considerations unrelated to the beneficiary acts just as loyally as the fiduciary who adopts these ends because they are the beneficiary's.
The shaping view avoids the prospect of accidentally complying with obligations of loyalty. If the interests of the beneficiary matter to the fiduciary solely in virtue of their membership in a common religion, then the fiduciary attributes only derivative significance to the interests of the beneficiary. This would violate element (2) of the shaping standard. Moreover, the beneficiary's interests might (ex hypothesi) cease to matter to the fiduciary if they deviated from the ends set by the common religion. This would violate element (3) of the shaping standard.
Another way of specifying loyalty's intentional dimension is what might be called, after Lionel Smith, the “aiming” account.Footnote 49 On this view, fiduciary loyalty requires aiming to act in beneficiary's best interest. Loyalty is thus largely a question of motivation: “Whatever powers a fiduciary has, he must exercise (or not exercise them) with a particular motive. He must act (or not act) in what he perceives to be the best interests of the beneficiary.”Footnote 50 For Smith, this intentional requirement differentiates the fiduciary duty of loyalty from the duties of care and diligence: the latter depend on whether “a particular objective standard was not met. Motive is irrelevant.” By contrast, “the exercise of a power can be set aside for a breach of the duty of loyalty if the power was exercised (and presumably if a decision was made not to exercise a power) with the wrong motive.”Footnote 51
On the aiming account, then, loyalty is determined by whether consideration of the beneficiary's interests figured causally into the agent's behavior. If a concern for advancing these interests caused the fiduciary to act, then the fiduciary has acted loyally. Where this concern does not play this causal role, the fiduciary has not acted loyally. As Smith puts it, fiduciary loyalty is about the “justiciability of motive.”Footnote 52 Chancellor Leo Strine and co-authors contend that Delaware corporate law embodies something like the aiming account:
The term good faith has long been used as the key element in defining the state of mind that must motivate a loyal fiduciary. To wit, the duty of loyalty most fundamentally requires that a corporate fiduciary's actions be undertaken in the good faith belief that they are in the best interests of the corporation and its stockholders.Footnote 53
Aside from its power to explain a major source of fiduciary jurisprudence, the aiming account has important strengths. Like the adopting account, the aiming account can explain why neither the Dartboard Trustee nor the double agent acts loyally. Neither protagonist is moved to act by the interests of the beneficiary. Further, unlike the adopting account, the aiming account does not admit of accidental compliance. To wit, the coreligionist fiduciary does not act loyally toward her beneficiary because she does not act in order to advance the beneficiary's interests so much as to advance interests that the beneficiary happens to have.
Despite these strengths, however, we think the aiming account fails to capture the intentional dimension of fiduciary loyalty for at least two reasons. First, the aiming account raises concerns about fairness. It is generally thought to be unfair to hold someone responsible for something she cannot control or help.Footnote 54 If “ought” implies “can,” then the aiming account relies on the voluntarist assumption that people can control the reasons for which they act. It would not be fair to hold a fiduciary to account for an action done for an illicit reason while holding him harmless for doing that same action for a licit reason (as the aiming account does) unless the fiduciary were capable of acting for one of these reasons but not the other. Furthermore, the aiming account assumes self-knowledge, or that people have privileged access to their own motivational states and know why they do what they do.
Yet both these voluntarist and self-knowledge assumptions are (at best) controversial. Smith acknowledges that the motives of others “are inscrutable at the best of times”; this inscrutability, he claims, grounds the no-profit and no-conflict legal rules that govern fiduciaries.Footnote 55 Some philosophers (influenced by developments in neuroscience and psychology) suggest that our own motives are similarly inscrutable to introspective analysis.Footnote 56 Furthermore, there is good reason to deny that people have the power to intend at will.Footnote 57 We do not mean to resolve whether either the voluntarist or self-knowledge assumption is true. However, if either assumption is false, then the aiming account is unfair, assuming that “ought” implies “can.”
By contrast, the shaping account does not require that fiduciaries have privileged access to their mental states or voluntary control over the reasons for which they act. Rather, it requires only that the beneficiary's interests structure the fiduciary's deliberation in the specific way described above: namely, that the beneficiary's interests have nonderivative significance and that the fiduciary be disposed to update her deliberative orientation given changes in the beneficiary's interests. Without fully elaborating an argument (which would require, among other things, a complete account of practical deliberation), it seems much more likely that these requirements do not ask too much. It is more plausible that you know how, why, and how much something matters to you than that you know why you act.Footnote 58 It is also more plausible that you can control the significance of a consideration in your deliberation (say, by adjusting your reasoning in order to make the requisite attributions) than that you can control your motivations for acting.Footnote 59
A second problem is that the aiming account deems loyalty impossible in cases where we have good reason to think it is possible and, indeed, to be expected. In particular, the aiming account denies that a fiduciary who is alienated from the beneficiary's interests could act loyally toward that beneficiary. Yet, to the extent that some form of alienation appears in many fiduciary relationships (and is perhaps inevitable in the modern world), this conclusion would be highly revisionist.
Consider the case of a disaffected attorney who knowingly undertakes the best course of action for her client as part of going through the motions. We think it is possible for such a disaffected attorney to live up to her fiduciary obligations, to act loyally toward her client.Footnote 60 Yet the aiming account could not reach this verdict; because the disaffected attorney is “just doing her job” rather than being moved by the goal of advancing her client's interests, she cannot act loyally toward her client.
By contrast, on the shaping account, the intentional dimension of loyalty imposes deliberative requirements rather than motivational ones.Footnote 61 Motives are one way of specifying loyalty's intentional dimension, but not the only (or even the best) way to do so. Many phenomena that have intentional components do not impose specific or strict motivational requirements. Take the example of instrumental rationality. Acting rationally has intentional components (e.g., the requirement to structure deliberation and respond to evidentiary considerations in certain ways) but no explicit motivational requirement. Why you act rationally is not necessarily connected with whether you act rationally.Footnote 62
Similarly, on the shaping account, whether you act loyally is not solely (or necessarily) a function of why you act. We agree with Smith's more recent argument that the intentional component of loyalty is not solely a function of motive.Footnote 63 A wide variety of motivations seem compatible with acting loyally.Footnote 64 Both the alienated attorney and the true believer can act loyally, since both are capable of making the requisite attributions and having the appropriate dispositions regarding their clients’ interests. However, we disagree with Smith's further conclusion that motive is irrelevant to determining whether one has satisfied her duty of loyalty. Rather, certain motives can preclude the possibility of loyal action by negating intentional requirements of loyalty.Footnote 65 The double agent comes to mind: her motives for furthering the false principal's interests rule out the possibility that she could attribute the kind of nonderivative significance to the interests of the false principal that would be required of loyal action.
In sum, we agree with Laby's and Smith's (erstwhile) positions that fiduciary loyalty has an intentional dimension. However, neither the adopting nor the aiming standard best describes this dimension. The shaping account is superior because it can resolve cases that undermine these other standards. Fiduciary loyalty requires robustly attributing nonderivative significance to a beneficiary's interests in virtue of the beneficiary's having them. If the fiduciary fails to attribute this significance to the beneficiary's interests and ends, then no matter what else we can say about her actions, she does not act loyally.
IV. AGAINST REDUCING FIDUCIARY OBLIGATIONS
Our analysis of the intentional dimension of fiduciary obligation, if sound, has important implications for ongoing theoretical debates. Both “contractarianism” and a view that we call “proscriptivism” are attempts to reduce fiduciary obligations to some other normative phenomenon. Debates about contractarianism, discussed below in Subsection A, concern whether fiduciary obligations can be reduced to contractual obligations. Debates about proscriptivism, discussed below in Subsection B, concern whether fiduciary obligations can be reduced to the legal rules that prohibit fiduciaries from conflicts of interests and profiting from beneficiaries. Neither of these reductionist strategies captures the intentional dimension of fiduciary obligations, the way that loyalty must shape the deliberation of the fiduciary. Therefore, neither of these positions describes the normative structure of fiduciary obligations.
A. Against Contractarianism
The so-called contractarian position sees fiduciary obligations as a species of contractual obligations. The contractarian asserts that it is possible to redescribe fully the former in terms of the latter. Contractarianism is in part an argument about how fiduciary obligations arise—namely, from voluntary arrangements, just as contractual obligations do.Footnote 66 As Larry Ribstein puts it, “one becomes a fiduciary only by contract, including by contracting for a relationship in which the law says fiduciary duties arise.”Footnote 67 The contractarian also contends that fiduciary and contractual obligations have a common structure and content: the content is fixed by agreement, with doctrinal provisions serving as default rules.Footnote 68 As such, fiduciary obligations (like contractual obligations) can be modified and disclaimed by the agreements of the parties.Footnote 69 These commonalities in origin and structure indicate a common instrumental goal: the point of both fiduciary and contractual obligations is to maximize the utility of the parties to these obligations.
Because the contractarian position has its roots in the law-and-economics tradition, many contractarians utilize (sometimes unwittingly) an additional premise that contractual performance is solely about behavior. The promises underlying contracts concern behaviors (“pay or perform”), and you satisfy your contractual obligation by behaving in the way specified by your contractual obligation or by paying expectation damages.Footnote 70 By implication, the contractarian is committed to the position that fiduciary obligations are also about behavior: you satisfy your fiduciary obligation by behaving in the way that loyalty specifies, or else by paying damages.
Several commentators have argued against the contractarian position, denying that contractual and fiduciary obligations share a common originFootnote 71 or a common structure.Footnote 72 However, the contractarian can evade many of these criticisms by recasting fiduciary obligations as based on a hypothetical bargain, reflecting the terms that the fiduciary and beneficiary would have agreed upon if they had considered them. Thus the contractarian can explain any putatively distinct elements of fiduciary obligations as a species of promissory obligations.
The contractarian position has many strengths. It can explain some core features of fiduciary law, as well as its variation across contexts. However, our analysis points to a deep problem with the contractarian view, one that cannot be evaded by the legerdemain of appeal to hypothetical bargain. In general, it is possible to reduce A to B only if every element of A is also an element of B. Thus the contractarian's reduction is possible only if every element of fiduciary obligation is also an element of contractual obligation. Yet, as the arguments of Sections II and III indicate, these two phenomena have different intentionality requirements. Promissory obligations can be satisfied through accidental compliance; given the intentional dimension to fiduciary loyalty, fiduciary obligations cannot. Thus the contractarian position is conceptually mistaken. Acting loyally has requirements that promise-keeping does not.Footnote 73
The contractarian might respond to our critique by denying that there is any asymmetry between contractual and fiduciary obligations. For example, the contractarian might relax the behaviorist assumption described above and posit that fiduciary loyalty involves both a promise to behave in a particular manner and a promise to regard the interests of the beneficiary in certain ways. Alternatively, the contractarian might contend that promissory obligations (like fiduciary obligations) have an intentional dimension that is captured in the requirement of “good-faith performance.”Footnote 74 If either of these efforts to deny the asymmetry succeeds, then the contractarian could still redescribe fiduciary obligations as promises or concatenations of promises.
Before examining these rehabilitative efforts, it is important to note how much they would concede. Broadening the notion of promise-keeping to include more than mere behavior (“pay or perform”) would, as noted above, undermine a fundamental tenet of the law-and-economics tradition out of which the contractarian position arises. Arguably, the analytic tractability of the behaviorist assumption is what inspired the effort to reduce diverse legal phenomena to contract. Thus, giving up behavior-based criteria for promise-keeping is no small concession. Likewise, many law-and-economics scholars deny that good-faith requirements are either useful or coherent tools for analyzing contractual obligations. Thus the contention that these standards form part of the core of two different (and prominent) types of obligations would not sit well with many contractarians.
In any event, it is not clear that either of these attempts to rehabilitate contractarianism by denying the asymmetry succeed. Seeing fiduciary obligations as constellations of promises to behave and to consider interests would not change the fact that all of the component promises can be satisfied through accidental compliance. Consider the following scenario involving a promise to consider someone's interests:
Big Decision Hypnosis: Annie promises Betty to consider Betty's interests in making a big decision that will affect Betty (among others). Prior to making the decision, Annie is hypnotized such that, in making the big decision, she will consider the interests of the next person she sees after awaking from the hypnotic state. After awaking from the hypnotic state, the first person Annie sees is Betty. Thus Annie considers Betty's interest in making the big decision.
We think that Annie keeps her promise to consider Betty's interests. If so, then promises to consider someone's interests can be satisfied through accidental compliance. What matters is that interests are considered, not how this consideration comes about. However, given the contingency of Annie's consideration of Betty's interests, it seems that Annie has not acted loyally toward Betty.
The shaping account can explain why Annie has failed to act loyally. While Annie attributes some significance to Betty's interests, this attribution is derivative (in virtue of seeing Betty after being hypnotized) rather than nonderivative (in virtue of the interests being Betty's). Thus it violates element (2) of the shaping account. Moreover, the attribution is not robust. Although Annie happens to consider Betty's interest in this scenario, she might not have done so if she had come across someone other than Betty after awaking from hypnosis. Thus her actions also violate element (3) of the shaping account. In sum, the asymmetry regarding accidental compliance remains, and the redescription of fiduciary obligations in terms of contractual obligations fails.Footnote 75
Likewise, appealing to a good-faith requirement for the satisfaction of contractual obligations would not save the contractarian position. For the contractarian's reductionist strategy to succeed, it is not enough that both promissory and fiduciary obligations have some intentional requirements. Rather, they must have the same intentional requirements—the conditions for keeping a promise must be identical to those for acting loyally. One prominent view construes contractual good faith in terms of opportunistic behavior.Footnote 76 Under this standard, there is nothing necessarily defective about accidental compliance. For example, the Lucky Payer is an example of accidental compliance without opportunism at all. Even on a broader construction of good faith, examples of accidental compliance would not necessarily constitute promise-breaking.Footnote 77
This argument can be framed in more positive terms. The broadest construction of contractual good faith (which, incidentally, most contractarians would reject for its breadth) requires the promisor to attribute significance to the interests of the promisee. This requirement parallels element (1) of the shaping account of fiduciary loyalty. However, contractual good faith imposes no counterpart to element (2) of the shaping account, which requires that the interests of the beneficiary have nonderivative significance to the fiduciary. Nor is there any counterpart to element (3), which requires that the beneficiary's interests matter robustly to the fiduciary. Indeed, such a requirement would seem absurd in the context of promises, tantamount to the notion that where A promises B to Φ, A keeps this promise to Φ only if she would have been similarly disposed to γ had B asked (or bargained) for it. Even if good faith imposes an intentional dimension to promise-keeping, the requirements of acting loyally are far more exacting than the requirements of performing in good faith, so the contractarian's reduction does not succeed. The “good faith” required of contractual performance (which is usually considered in objective terms) is not coextensive with the intentional dimension of fiduciary loyalty, which has an undeniably subjective component.Footnote 78
B. Against Proscriptivism
Our analysis in Sections II and III also casts doubt on efforts to redescribe fiduciary obligations in terms of the proscriptions applicable to fiduciaries, a strategy we call proscriptivism. Some advocate for proscriptivism directly.Footnote 79 Others embrace it indirectly by distinguishing between “minimal” and “maximal” notions of loyalty and arguing that both are legally cognizable.Footnote 80
Matthew Conaglen offers the most developed version of the proscriptivist position.Footnote 81 Conaglen seeks to identify what is distinctive about fiduciary obligations under British Commonwealth law. He argues that the “no conflict” rule (which prohibits fiduciaries from “acting with a conflict between duty and interest”) and the “no profit” rule (which prohibits fiduciaries from “making a profit off of the fiduciary position”) both “form the core of the fiduciary doctrine and . . . constitute its conceptualization of loyalty.”Footnote 82 Conaglen contends that the conjunction of these rules facilitates the fiduciary's discharge of his nonfiduciary duties toward the beneficiary (which may be promissory in nature). Every other aspect of fiduciary obligations (like the duty of care and the duty not to act for improper purposes) has a parallel in nonfiduciary obligations. So, Conaglen argues, the fiduciary's duty of loyalty can be completely redescribed as the conjunction of the “no conflict” and “no profit” rules.
Conaglen's analysis is incomplete as a description of positive law and conceptually confused as a description of loyalty as a normative concept (although, to be sure, Conaglen does not aspire to the latter task). First, the “no conflict” and “no profit” rules are not the only ways to conceptualize fiduciary loyalty. For example, the shaping account of loyalty offers an alternative way of appreciating what is distinctive about fiduciary obligations. On this view, fiduciary obligations differ from other types of obligations based on the specific intentional requirements that they impose on fiduciaries.
Second, Conaglen's proscriptive account is problematic because it neglects the intentional dimension of fiduciary loyalty. In life (if not in law), it seems possible for a fiduciary to satisfy both the “no conflict” and “no profit” rules while failing to act loyally. For example, in circumstances of accidental compliance, a fiduciary could meet the “no conflict” and “no profit” rules without acting loyally.
Consider the Dartboard Trustee case once again. Tammy Trustee has no conflicts of interest in her representation of Bernice Beneficiary. Nor does Tammy reap unwarranted profit at Bernice's expense. Therefore, Tammy does not run afoul of either the “no conflict” or “no profit” rules. Conaglen's view, then, implies that Tammy acts loyally, although he might contend that Tammy's actions run afoul of one or more of her nonfiduciary duties. But Tammy does not seem to act loyally. Her actions violate elements (2) and (3) of the shaping account. Because proscriptivism cannot capture the intentional dimension of loyalty, it cannot account for the structure of fiduciary obligations. To cite a real-world example, in a traditional hedge fund operation the asset managers’ fee of “two and twenty” (two percent as a set management fee plus 20 percent of the profits of the fund) might be earned without conflict or extra profit, with managers and investors all receiving a substantial return on investment. Yet if the managers’ “proprietary software” that makes the trading decisions in the fund were to invest based on a random algorithm, then the managers would seem to run afoul of their requirement of loyalty. In this scenario, the “two and twenty” fee would seem to be susceptible to disgorgement, the standard remedy for a breach of the legal duty of loyalty.
The proscriptivist might try to evade our critique by casting her position solely in descriptive terms, eschewing any attempt to describe loyalty as a normative phenomenon. (The contractarian might try a similar move.) On this gambit, proscriptivism (and contractarianism) attempts only to describe how judges decide cases involving alleged breaches of fiduciary duties. We think that this strategy is mistaken. For example, epistemic difficulties in determining subjective mental states might prompt legal institutions to utilize specialized tests for measuring whether a fiduciary's actions were in good faith. However, this utilization would not actually establish whether good faith (as a normative phenomenon) is ultimately subjective or objective. What loyalty is and whether someone has acted loyally seem to have answers that are independent of judicial determinations.Footnote 83 The utilization of objective standards might make the inquiry manageable, but this concern about administrability does not mean that the result of the inquiry is the unvarnished truth about the underlying normative phenomenon. Indeed, administrability concerns might preclude an inquiry from tracking the normative truth. Thus, if proscriptivists and contractarians recast their projects as simply describing how judges deploy a purely juridical concept, then they must relinquish any ambition to explain the underlying normative phenomena. We think that a project that aims to describe legal concepts while eschewing the broader normative landscape is problematic, even on its own terms.Footnote 84
Loyalty makes demands on us. Some of these demands concern our actions, and others concern our deliberation. Fiduciary loyalty's deliberative demands are best captured by the “shaping” account. When you act loyally toward someone, she should matter to your deliberation in specific ways. If someone does not matter to your deliberation in these ways, then you do not act loyally toward her.
Promises also impose demands, but not the same kinds of demands as loyalty. The demands of promises concern behaviors. They do not center on intentions. How you deliberate about a promise (or a promisee) is usually irrelevant to whether your actions count as promise-keeping. Thus promises and loyalty are not the same thing, and the latter cannot be accounted for solely in terms of the former. Those (like the contractarian or the proscriptivist) who attempt to reduce fiduciary loyalty to something else miss the essence of what loyalty is and why it matters.
Our account of the structure of fiduciary obligations has other important implications that we cannot explore at length here. For example, an ongoing debate in fiduciary theory concerns the extension of fiduciary concepts to various areas in public law, where public offices are often considered to be public trusts.Footnote 85 Yet efforts to extend fiduciary principles may vary in their success, and the intentional dimension of fiduciary obligations can explain why. In general, extensions of fiduciary concepts are most likely to be fruitful in areas of law and politics where the imposition of intentional requirements is independently plausible, and unfruitful in areas of law where there is independent reason to believe that intentions do not or should not matter. The normative structure of loyalty constrains these and other efforts to theorize fiduciary obligations.