I. The subjective view of loyalty

The orthodox account of the fiduciary relationship places loyalty at its heart. On this view, not merely do all fiduciaries, by virtue of being fiduciaries, owe a duty of loyalty; but that duty is the “foundational” feature of the fiduciary relationship.Footnote ¹⁶ This suggests two things. First, that most (if not all) of the other distinctive features of the fiduciary relationship are to be explained by reference to it. Second, that the explanation for why any particular relationship is “fiduciary” ought, at least in central cases, turn upon facts that justify the need for loyalty.Footnote ¹⁷ On this view, therefore, loyalty is at the analytical core of what it means to be a fiduciary. Yet, what does it mean for a fiduciary to be “loyal”? What is the content of loyalty?

The paradigmatic formulation is that the duty of loyalty requires that the fiduciary “act in the best interests of the beneficiary” (the “best interests” formula).Footnote ¹⁸ However, literally interpreted, this cannot be true. It would require that, where the duty of loyalty operates, the fiduciary must act in what objectively is in the best interests of the beneficiary. So, for example, ceteris paribus, a trustee charged with investing assets on behalf of a beneficiary would be obliged to make the investments that, in truth, will bring the beneficiary the highest financial return, regardless of the obvious difficulty for anyone in predicting exactly what will do so. If the fiduciary failed to do so, however, then the fiduciary would fail in their duty of loyalty, even if their decision was prudent, reasonable, or even undeniably the strategy with the best expected return. Further, and perhaps most damningly, as a result, the duty of loyalty would also be inconsistent with another “primary” fiduciary duty, that is, the “duty of care,” since pursuing what is, in truth, in the beneficiary’s interests would not be consistent with what the fiduciary’s prudent and reasonable judgment determines.Footnote ¹⁹

Yet if the duty of loyalty, under its paradigmatic formulation, cannot literally require the fiduciary to act in the best interests of the beneficiary, what then does it mean? Broadly speaking, the extant debate offers two kinds of answer: the subjective prescriptive view and the proscriptive view.Footnote ²⁰

The subjective prescriptive view of the best interests formula of loyalty (or just “subjective view,” for short) is most closely associated with the work of Lionel Smith.Footnote ²¹ As Smith puts it, “the requirement of loyalty is subjective, in the sense that it requires the fiduciary to exercise judgement in what they subjectively believe to be the best interests of the beneficiary.”Footnote ²² As Smith often expresses the point, on this view it is the motive of the fiduciary that is the subject of the duty, not the deed itself. Or, more precisely, what matters are the reasons that the fiduciary takes as determining the basis of their decision as to how to act in their fiduciary capacity.Footnote ²³ Those reasons have to be “what he perceives to be the best interests of his principal.”Footnote ²⁴

Smith concedes that, prima facie, by regulating subjective motive rather than objective results this principle might seem to run against the grain of typical principles of private law.Footnote ²⁵ Nonetheless, Smith makes two main arguments in favor of it.

First, Smith argues that the duty of loyalty explains principles regulating the exercise of “fiduciary powers,” that is, those powers that a fiduciary has, by virtue of their fiduciary position,Footnote ²⁶ such as a trustee’s power to dispose of trust property; or a director’s power to issue shares. Smith argues that the case law demonstrates a consistent aversion by the courts to review the exercise of such powers by their outcome (or substantive results), but a willingness to review motive.Footnote ²⁷ For example, the sale of property by a trustee is reviewable for improper motive, even absent any violation of any other fiduciary dutyFootnote ²⁸; so too the actions of company directors.Footnote ²⁹

Second, in a more defensive vein Smith turns to the two “prophylactic duties” considered by many to be equally characteristic of fiduciary relationships: the no-conflict and no-profit rules. Both rules explicitly regulate outcomes, not motive. The no-conflict rule excludes the fiduciary from acting in their capacity as fiduciary with a conflict of interest, and the no-profit rule excludes the fiduciary from making a profit from the powers or resources that they have by virtue of being a fiduciary. In both cases, motive is irrelevant, such that even if the fiduciary was acting in what they perceived to be the best interests of the beneficiary, indeed even if they did in fact act in what was in the best interests of the beneficiary, the existence of conflict or profit would still be a breach of fiduciary duty. Smith argues, however, that we can still justify such “outcome”-related principles on the basis of the need to promote adherence to a “subjective” duty of loyalty. This is because they relieve the beneficiary of the hardest element of enforcing any duty of loyalty, that is, having to prove what is often inscrutable—motive. Instead, they need only prove the existence of conflict or profit. As a result, he argues: “the prophylactic rules entitle the beneficiary not to have to wonder about the fiduciary’s motive. Understood in this way, it seems to me that prophylaxis can be seen to grow properly out of the voluntarily undertaken duty of loyalty.”Footnote ³⁰

There are a number of objections to this subjective view of loyalty, but I shall just focus upon what I take to be the deepest theoretical problem.Footnote ³¹ This problem may be stated by way of a simple question: what value grounds the duty of loyalty so construed? Putatively the duty of loyalty is owed to the beneficiary and, thus, breaching it without due justification is a wrong perpetrated upon that beneficiary.Footnote ³² So one might presume that some interest of the beneficiary must be at stake that not merely explains the value of such loyalty to the beneficiary but also why it is always owed to them; why they ought to have a legal claim to loyalty. Yet, what interest could that be? What interest could the beneficiary possibly have in the fiduciary pursuing what they perceive to be in the beneficiary’s best interests, independent of what in fact is in the beneficiary’s best interests?

As Stephen Smith puts much the same point: “[I]t would be surprising if the law cared whether fiduciaries acted as [subjectively] loyal fiduciaries would act. … [M]ost fiduciary relationships are strictly commercial and in these cases what beneficiaries care about are outcomes, not motives. If I transfer funds to a trustee to invest on my behalf, what I care about is whether the trustee invests the assets productively—not whether she invests with the right motive. … Loyalty has little, if any, inherent value.”Footnote ³³

Indeed, to make the point more acute: consider where the fiduciary’s subjective perception of the beneficiary’s best interests is incorrect, and indeed counterproductive to the beneficiary’s actual best interests. On the subjective view of loyalty, in such a scenario, the fiduciary is not merely permitted or excused in acting contrary to the beneficiary’s objective best interests—they owe it to the beneficiary to do so, and, indeed, the beneficiary is wronged somehow if the fiduciary does not. What interest could the beneficiary have in that?

One might respond to this objection by conceding that there is no inherent value to subjective loyalty—and, thus, nothing inherently wrongful, morally speaking, in a fiduciary failing to be subjectively loyal—but maintain that this is not necessary to explain the legal duty of subjective loyalty. It is sufficient that this legal duty itself will, generally, best serve the relevant ends of the fiduciary relationship. Specifically, it will tend to best promote beneficiaries’ interests objectively, at least compared to relevant alternative possibilities.Footnote ³⁴

However, this response suffers from two problems. First, simply as a matter of doctrine, such an instrumental justification of the fiduciary duty of loyalty would fail to account for the many moralizing statements of law itself to the effect that disloyalty in a fiduciary relationship is a wrong, regardless of such consequences.Footnote ³⁵

Second, even assuming that the fiduciary’s subjective loyalty is in fact the best means to the end of best promoting the beneficiary’s interests in fiduciary relationships, without an underlying moral duty it is not clear why the law would dictate that fiduciary relationships ought to have that overriding end. After all, typically, like contracts, fiduciary arrangements are voluntary arrangements between consenting parties. Like contract, the fiduciary arrangement involves an agreement to share costs and benefits (and risks and opportunities) in certain ways as one establishes a set of duties. So why, unlike contract, would the law always, as a default, predetermine a particular sharing of costs and benefits that prioritizes the interests of the beneficiary over those of the fiduciary in this particular way? Why not, as a default, just let the distribution be a matter of agreement among them—as in contract? To be sure, many fiduciary relationships involve vulnerable parties, but not all; and even where they do, contract law has its own principles to account for such a fact, including the duty of good faith.Footnote ³⁶ But in general, if the duty of loyalty is not justified by some underlying moral duty grounded in some relevant interest of the beneficiary that always needs to be protected, it is unclear why the law would then predetermine any such thing about the fiduciary relationship, rather than prioritizing the agency of the parties, and their ability to distribute its costs and benefits as they wish, on whatever terms they wish.Footnote ³⁷

In short, the duty of loyalty is supposedly meant to be the defining feature of fiduciary relationships—that which distinguishes it from other relationships, and particularly mere contractual relationships—but if it lacks a deep moral justification, then our ability to maintain such a principled delineation of fiduciary law in general comes under great pressure.Footnote ³⁸ This ultimately gives succor to an entirely different account of fiduciary loyalty altogether: the so-called “proscriptive view.”

While the subjective view of the standard “best interests” loyalty formula may be thought to read down its facially objective content, the proscriptive view effectively reads it away. On this view, a fiduciary is strictly speaking under no general duty to act in the best interests of their beneficiary—either subjectively or objectively construed.Footnote ³⁹ Instead, any fiduciary will have a set of non-fiduciary duties owed to the beneficiary, that is, duties that arise for reasons that are not peculiar to the fiduciary relationship. They arise either as a matter of general law, or as in effect a matter of contract: that is, as what the fiduciary explicitly or implicitly undertook to do when becoming a fiduciary.Footnote ⁴⁰ These duties shape what might be termed the fiduciary’s “mandate.”Footnote ⁴¹ Hence, for example, a trustee will have a duty to act in accordance with the terms of the trust they have undertaken—including, say, to only act in pursuit of certain authorized purposes—but not because they are a fiduciary per se, but because that is what they have agreed to do.Footnote ⁴² On this view, then, the only obligations that are “peculiarly fiduciary” are the no-conflict and no-profit rules, that is, specific proscriptions against certain behaviors (conflicts, profits), rather than a prescription (to act in the best interests of the beneficiary).Footnote ⁴³ In turn, so the argument goes, these proscriptions are not justified as instrumental to the best interests of the beneficiary per se, but rather the performance of the other non-peculiarly fiduciary duties owed to them. This is to say that the two peculiarly fiduciary duties are putatively justified because they tend to promote the fulfillment of the non-peculiarly fiduciary duties (which are presumed, in turn, to promote or respect some set of interests of the beneficiary).Footnote ⁴⁴

What then of “loyalty” on this view? Well, the “duty of loyalty” is either reduced to being simply the performance of these other non-peculiarly fiduciary duties consistent with the peculiarly fiduciary duties of no-conflict and no-profit,Footnote ⁴⁵ or indeed, according to some, best eliminated altogether. Hence, in this latter vein, Stephen Smith proclaims: “not only is loyalty not a central concern of fiduciary law; it is not a concern at all. Nor should it be. What matters in fiduciary relationships is that fiduciaries fill their mandates (eg, executing a trust, promoting the principal’s business, providing legal services).”Footnote ⁴⁶

The proscriptive view might be thought to explain loyalty only by explaining it away. But its proponents can well respond that whether you call it loyalty or not, at least the theory identifies something that is truly distinct about the fiduciary relationship—the no-conflict and no-profit rules—and grounds them on something of relevant value, that is, the fulfillment of the obligations of the parties. This may imply that the ultimate grounds of fiduciary relationships are contractual, resting on the agreement of the parties to a particular kind of fiduciary protection for particular terms of their agreement—but for many this is a theoretical price they are willing to pay.Footnote ⁴⁷ Of course, this proscriptive view is not without its own problems.Footnote ⁴⁸ But without a strong defense of the value of loyalty, as a subjective prescription, the subjective view is poorly placed as an alternative. Let us now turn to existing attempts.

II. Smith’s arguments for the value of loyalty

Lionel Smith provides two arguments that may be cast as responses to our problem. Underlying both is a natural strategy. Since the subjective duty of loyalty putatively arises because one is a fiduciary—to regulate the exercise of power that one thereby has as a fiduciary—then presumably the explanation of that duty must lie in the nature of that power.Footnote ⁴⁹ But what, then, is “fiduciary power”? Consistent with many others, Smith holds that the distinctive feature of fiduciary power is that it is the power to make decisions “for” (or “on behalf of,” or indeed “as”) another person.Footnote ⁵⁰ But what precisely does that mean? Smith’s two arguments for the value of loyalty derive from two different answers to this question.

First, Smith glosses the relation of acting “for” another in terms of the fiduciary acquiring control over that other’s “autonomy”:

Given this characterization of what fiduciary power necessarily is, Smith then immediately makes the following argument for subjective loyalty:

The problem with this argument is twofold. First, the argument is built on a contradiction. Smith states: “[i]n every fiduciary relationship, the fiduciary acquires control over a part (or in some cases, all) of another person’s autonomy.” But a fiduciary, F, cannot control beneficiary B’s autonomy. Instead, F’s control over B in a certain domain displaces B’s autonomy in that domain (if B ever had autonomy in that domain in the first place). In such a case, B is no longer “controlling what happens in his or her life.” F is instead. There is an exception to this principle, that is, where B controls F’s control of B, which is the defining feature of one kind of fiduciary relationship: the agency relationship (to which I turn in Section VI). But it is notably absent in most others.

However, even if we accept that in some sense we might be able to respect someone’s autonomy when it is absent, we still arrive at a second problem. It is simply a non sequitur that the proper form of such respect for such autonomy is “best interests” loyalty subjectively construed. It does not follow that the proper way to respect another’s ability to determine for themselves what to do with their own lives is to exercise one’s ability to determine what they ought to do in accordance with one’s own view of what is best for their lives.Footnote ⁵³ Indeed, one might think the opposite would hold: in order to respect the value of autonomy, a fiduciary ought to either do their best to restore the other’s autonomy, or at least seek to act in accordance with what they believe would have been its effects, that is, to act in the way they believe the other would have chosen to act autonomously.Footnote ⁵⁴

Having said all this, it is worth noting one claim that Smith makes in the midst of his argument above: without the duty of loyalty a fiduciary “could use their powers … as if they were nothing more than pecuniary assets, like money in the bank” and consequently “using [those powers] as pecuniary assets would be using another person as if he or she were a pecuniary asset.” As I shall explain below,Footnote ⁵⁵ I agree with Smith that this kind of proprietary relationship does constitute the wrong perpetrated against the beneficiary. However, unlike Smith, it cannot be wrong because it disrespects another’s (often absent) autonomy.Footnote ⁵⁶

In more recent work, Smith has now adapted his conceptions of “fiduciary power” and “loyalty.” First, Smith now holds that, in the most general formulation, a fiduciary (or “administrator”)Footnote ⁵⁷ is “a person has been authorized and empowered to carry out a mission, an assignment, a function, a task, a charge for or on behalf of someone else.”Footnote ⁵⁸ On this view, what defines fiduciary power is actually that it always has a particular kind of purpose, that is, to achieve some “other-regarding mission.”Footnote ⁵⁹ The question as to the existence, and particular nature of this mission is to be answered by interpretation of the particular grant of power at issue.Footnote ⁶⁰ Upon such interpretation, the “other regarding mission” of the grant of power may still be the advancement of the interests of a beneficiary (or set thereof); but often not: it turns upon the terms of the grant of power.Footnote ⁶¹ Second, given that fiduciary power is now defined as power held for an other-regarding mission, Smith adapts his view of loyalty accordingly: a fiduciary is loyal where they act in what they perceive to be the best pursuit of that other-regarding mission.Footnote ⁶² With these adaptations, we might think that the value that grounds the duty of loyalty (or, indeed, as Smith now prefers, “requirement” of loyalty)Footnote ⁶³ can be found in the mission it promotes. At minimum, it shows a kind of respect for that mission to act in what one perceives to be its best pursuit, and that it is reasonable to assume such respect is owed by agreement in taking up the power.Footnote ⁶⁴

This new purposive conception of loyalty, however, has at least one major problem: it fails to explain how in many contexts, far from requiring pursuit of the particular other-regarding mission of the power at stake, the duty of loyalty must regulate those domains of discretion left underdetermined by that very mission.

To illustrate: take a trust that has a number of explicit purposes set out by the settlor and agreed to by the trustee, e.g., the educational, cultural, and sporting advancement of the beneficiary. In this case, the fiduciary has been granted their powers specifically to pursue these purposes. On Smith’s new view, then, it would seem, these purposes define the “other-directed mission” of the grant of power, and the duty of loyalty is a duty to act in what the fiduciary perceives to be the best pursuit of those specific purposes. Doctrinally, of course, the courts might still identify this as a case of “best interests loyalty” qua a duty to pursue the best interests of the beneficiary, but Smith appears to be committed to reading “best interests of the beneficiary” in such a case as “pursuing the specific purposes of the trust.”

Now this does contort the ordinary meaning of the concepts at play. But even putting this aside, the bigger problem is that it leaves unexplained what loyalty means and requires, if anything, in this context, when such a mission underdetermines the fiduciary’s discretion. So, to illustrate, assuming the terms of the trust are silent on the proper balance between the educational, cultural and sporting purposes, what is the loyal way of resolving such purposes? If we take Smith’s view literally, then loyalty would have no answer except to restate the problem: pursue those purposes! Instead, most obviously, the answer is to resolve them in the way that the fiduciary perceives to best serve the interests of the beneficiary where such “best interests” are to be defined in the capacious, orthodox sense. But if that is the case, then loyalty cannot just mean “act in what one perceives to be the best pursuit of one’s other-regarding mission” qua the specific purposes of the trust. This is not an isolated problem for Smith, but in fact an endemic one: since the specific purposes of any fiduciary relationship, if any, will always tend to underdetermine the action in a range of ways given, say, competing interpretations of those purposes; uncontemplated circumstances; scenarios where the relevant purpose(s) just do not apply, and so on. In each case, the role of the duty of loyalty is precisely to guide the fiduciary’s discretion where the clear demands of the relevant “mission” run out.Footnote ⁶⁵ But if that is the case, then it cannot be identical to pursuing that very mission.

To be sure, at points, Smith provides grounds for a different interpretation of his own new theory: that is, regardless of their particular explicit purposes, we should take the general mission of almost all private law fiduciary relationships to be at least the service of the best interests of their beneficiaries, once again now in the original, orthodox sense.Footnote ⁶⁶ On this view, we would then get a stipulated alignment, at least, between “other-directed mission” and the paradigmatic “best interests” formula, and the problem above disappears; but only for two further problems to emerge.

First, facially, it is just not true that the mission of many private law fiduciary relationships is the best interests of the beneficiary simpliciter. The mission of the trust sketched above, for example, is defined by the specific purposes set out in that trust—the educational, cultural or sporting advancement of the beneficiary. To be fair, the settlor may have set out such specific purposes with the overall implicit aim of best serving the interests of the beneficiary overall. But they may not have: she may primarily be seeking to motivate the beneficiary to be more like her, independent of any assessment as to what would be best for the beneficiary. Either way it does not matter. The mission of the trust is clear as a matter of interpreting the grant of power,Footnote ⁶⁷ and it is to pursue these particular purposes; it is not to promote the best interests of the beneficiary in general.

Second, retreating to this more general assumption about fiduciary relationships in effect just begs the question at hand. Consistent with Smith’s original theory of loyalty, on his new view we are still to identify fiduciary relationships by the existence of “fiduciary power,” but now we are to identify such fiduciary power by the fact that it has an “other-directed mission” (rather than, alternatively, inferring that from the fact that the power is fiduciary first, as defined by some other condition). But if having an “other-directed mission” is not to be identified with the explicit purposes of particular grants of power, but instead some general implicit mission to best serve the interests of another, then how do we identify which relationships have that general mission? Why do some kinds of power have it and others do not?Footnote ⁶⁸ And what value, if any, lies in the subjective pursuit of such a mission by fiduciaries, even when they are wrong about what best advances that mission? We are, in effect, back to where we started.

III. Two freedom-based arguments for loyalty

Smith’s own arguments, therefore, for the value of subjective loyalty—both under his original and new formulation—do not hold. There are, however, two other arguments elsewhere in the literature that, prima facie, might. They are developed separately, but mutually endorsed, by Evan Fox-Decent and Evan Criddle, as two strands of one freedom-based argument for the fiduciary duties in general: “[t]he best justification for fiduciary duties, we believe, is that these legal obligations are necessary to provide for the beneficiary’s secure and equal freedom vis-à-vis the fiduciary.”Footnote ⁶⁹

The first strand of this argument, developed by Fox-Decent specifically, aims to generalize an argument made by Kant in the Metaphysics of Morals for the duty of parental care. As Fox-Decent construes this duty, it is fiduciary in nature and, akin to Smith’s own subjective view of loyalty, an obligation upon parents to exercise their legal powers over the child “in what they reasonably regard as the child’s best interests.”Footnote ⁷⁰ What grounds such a duty for Kant, according to Fox-Decent, is that it is necessary to respect the child’s “dignity as a person,” in a context where the parent has unilateral power over the child’s interests. As Kant says:

This imperative to respect personhood, according to Fox-Decent, generalizes to ground the duty of loyalty in fiduciary relationships: “Legal personality and the idea of dignity intrinsic to it supply the moral basis of the beneficiary’s right to the fiduciary obligation.”Footnote ⁷² But what exactly is it about being a “person” that demands loyalty subjectively understood in a fiduciary context? This, unfortunately, is not entirely clear.Footnote ⁷³ I think we might place two interpretations on Fox-Decent’s interpretation of Kant.

The first interpretation is that it is the fact that all (moral) persons, including children, have what Kant terms an “innate right” to external freedom.Footnote ⁷⁴ As Fox-Decent explains in detail,Footnote ⁷⁵ Kant holds that each person, by virtue of their personhood, has one and only one “innate right” and that is to equal freedom with each other, where “freedom” is understood as “independence from being constrained by another’s choice.”Footnote ⁷⁶ All other rights—“acquired rights” such as property and contractual entitlements—are derivative of this innate right via the exercise of the acquired right. In order to respect such an innate right, and the acquired rights derived from its exercise, therefore, law is needed to secure them in “a regime of equal freedom in which no party can unilaterally impose terms of interaction on another.”Footnote ⁷⁷ Fiduciary principles, including the duty of loyalty, then are ultimately needed as part of that regime—or so Fox-Decent’s argument seems to go.Footnote ⁷⁸

But how do we get from this Kantian premise of equal freedom to the duty of loyalty as an implication? After all, as Fox-Decent accepts, a fiduciary can “unilaterally impose terms of interaction upon another”—that is, the beneficiary.Footnote ⁷⁹ So why does compliance with the duty of loyalty preserve and/or respect the latter’s equal freedom qua “independence from being constrained by another’s choice” in doing so? There is no such independence in a fiduciary relationship; indeed that seems inherent to the idea, and the parent-child relationship exemplifies this. The parent is constantly imposing their will upon the child’s normative and descriptive condition: not merely, physically constraining them, but making decisions about their property (if any), or status under law (e.g. citizenship), or liabilities under contract (e.g. at a later age, employment). The fact that these impositions might be made loyally—that is, in the child’s interests as the parent sees them—might make them justified for some as yet undisclosed reason; but they would still come at the cost of the child’s freedom qua independence. Hence, if loyalty does not promote or preserve freedom as independence, freedom as independence cannot be the reason that justifies loyalty.Footnote ⁸⁰

Given this objection, one might be tempted to head in a different interpretative direction in attempting to understand Fox-Decent’s argument. It is not respect for the fact that the beneficiary’s personhood grounds an innate right to external freedom that justifies the duty of loyalty, but instead another fact about personhood as such. Persons have internal freedom.

Once again, following Kant,Footnote ⁸¹ Fox-Decent notes that a person is “a subject whose actions can be imputed to him,” which in the case of human beings is grounded in our having moral (or “internal”) freedom, that is, being self-determining agents responsible for our actions. Or as Kant says, “[m]oral personality is therefore nothing other than the freedom of a rational being under moral laws.” A thing, by contrast, is “[a]ny object of free choice which itself lacks [such] freedom,” and thus “is that to which nothing can be imputed.”Footnote ⁸²

However, this putative fact about being a person cannot explain the value of loyalty either. This is for two reasons. First, for some paradigmatic beneficiaries in fiduciary relationships, the relevant attribute of moral freedom is simply absent, or at least not sufficient for justifying agency and responsibility (that is, having action “imputed” to one). Children, especially very young children, and the severely mentally impaired are obvious examples. Second, even if one might be tempted to argue that on some conception of “capacity” such agents still have the capacity (if unrealized) for such moral freedom, it is still not clear why the fact that a person has such agency (or the capacity for agency)—and hence, the ability to set ends for themselves, and be responsible for them—is a reason, within the fiduciary relationship, for the fiduciary to be obliged to set those ends for them on their behalf. Both these things may be true, but this is a non sequitur. Indeed, this just returns us to the objection made against Lionel Smith’s argument from autonomy above: why does respect for the beneficiary’s agency demand, in effect, substituting it with the fiduciary’s agency rather than the most obvious response which is doing everything one can to bring the beneficiary’s agency into existence, and take direction from it once it does exist? This might well come at the expense of what the fiduciary thinks is in the beneficiary’s best interests overall, but that in part is what it means to respect the beneficiary’s agency.

As I shall come to in the next section, I do think there is something in the idea that loyalty is required to respect the beneficiary’s personhood in a fiduciary relationship. But the Kantian version of this argument, at least as drawn specifically from the Metaphysics of Morals, cannot explain this inference either as necessary to respect external freedom or as necessary to respect internal freedom. However, Evan Criddle has also developed an argument for loyalty that draws upon a different republican conception of freedom instead. On the republican view, freedom is neither non-interference nor Kantian independence. Instead, as Criddle explains, it is freedom from “domination,” where “domination” is subjection to another’s arbitrary power.Footnote ⁸³ Largely following the work of Philip Pettit, another can be said to have such power with respect to us, in the relevant sense, where they have the ability to interfere with us (even if they do not choose to do so). That ability to interfere is “arbitrary” where it is not “forced to track the avowed or avowal ready interests of the interferee.”Footnote ⁸⁴ Criddle’s argument is that the fiduciary duties, and loyalty in particular, are justified because they ensure that liberty in this sense—freedom from domination—is not compromised by fiduciary power:Footnote ⁸⁵

The problem with this republican argument for loyalty is twofold. First, and most importantly, the duty of loyalty, or at least the duty of loyalty subjectively construed—that is, to pursue the best interests of the beneficiary as one perceives them to be—is not equivalent to a duty to follow the avowed or avowal-ready interests of that beneficiary. The former requires that the fiduciary make and act upon their own judgment as to the interests of the beneficiary, independent of whether the beneficiary has expressed a view on the matter; or whether they are capable of expressing a view; or even, in some cases, in a manner openly inconsistent with such a view if they do express it. The latter requires, by contrast, that the fiduciary take the beneficiary’s view on the matter as decisive (“avowed, or avowal-ready”).Footnote ⁸⁷

Second, even if one adjusted one’s definitions of “loyalty” and/or “nonarbitrary” interference to somehow cohere, the resulting republican value of non-domination could not explain the duty of loyalty per se, but rather only the value of its enforcement. To explain: on the republican view, we are dominated if another has the mere descriptive ability to interfere with us arbitrarily; and we are not dominated when another does not have this descriptive ability. However, a fiduciary can be loyal—whether that be in the sense of only pursuing their beneficiary’s interests as they perceive them, or as they are “avowed”—but still have the descriptive ability to act disloyally—that is, not to act in pursuit of such interests, or as they are avowed. This would arise where the fiduciary’s compliance with the duty of loyalty is not exhaustively “controlled” qua invigilated and enforced—as, indeed, surely most are not. But in this case, such a loyal fiduciary would still “dominate” the beneficiary, on the republican view, because their choice to be loyal or not would, descriptively speaking, remain “up to them.” They would retain the ability to interfere in an arbitrary (qua “non-loyal”) manner, even if they reliably choose not to. Furthermore, and most importantly, the republican cannot explain the value of such “uncontrolled” loyalty; the mere fact that the fiduciary has the ability to be loyal or not—to interfere arbitrarily or not—would still constitute domination. Of course, one might think it odd to consider such a loyal fiduciary to be “dominating” simply because their loyalty is not assured by the control of some other actor, but instead moved, say, by a desire to adhere to duty itself. Yet, I suggest, that is all the worse for the republican account of “domination” rather than the value of loyalty.

IV. An aside: ordinary loyalty

At this point, it may strike the reader as odd that none of the preceding accounts of the value of the legal “duty of loyalty” appeal to the value of “loyalty” in its ordinary sense as a virtue in non-legal life. After all, we might think, we tend to see loyalty as praiseworthy in general; why not then in fiduciary relationships in particular?

There are two issues that frustrate such an argument. The first is alignment in meaning. Is “fiduciary loyalty,” in accordance with the subjective best interests standard, recognizably a kind of loyalty—in the ordinary sense—at all? Some argue not. Ordinary loyalty cannot be instantaneous, yet fiduciary loyalty is required at the moment a fiduciary takes up their role; ordinary loyalty requires emotional attachment to the object of loyalty, yet fiduciary loyalty—as most obviously manifest in professional financial fiduciaries—requires no emotional attachment at all; ordinary loyalty involves identification with one’s object, fiduciary loyalty does not; and so on.Footnote ⁸⁸ However, this objection seems to assume more conceptual unity to the ordinary concept of loyalty than it actually has. As Simon Keller argues, there is a multiplicity of different “ways” of being loyal in ordinary life with barely more in common than “an individual’s taking something’s side, and doing so with a certain sort of [partial] motive” with respect to that something.Footnote ⁸⁹ Fiduciary loyalty, construed as pursuing the best interests of one’s beneficiary as one perceives them to be, has enough in common with this thin notion at least to claim to be a kind of loyalty, or at least a close juridical cognate to (or evolution) of it.Footnote ⁹⁰

The more pressing issue is that, upon closer inspection, it is not clear that ordinary loyalty—at least when taken at the relevant level of generality—itself is inherently valuable. After all, the value of “taking something’s side, and doing so with a certain sort of [partial] motive” with respect to that something would seem to turn upon the value of that something, and/or what we are doing for their sake.Footnote ⁹¹ Loyal Nazis and mafiosos are classic counterexamples. To be sure, some kinds of valuable relationships—friendship or the parent-child relationship—demand specific kinds of ordinary loyalty, and hence in those relationships loyalty of these kinds may be a necessary means to such pro tanto valuable ends.Footnote ⁹² But this just illustrates the need to seek the inherent value, if any, of a particular kind of loyalty (e.g., best interests fiduciary loyalty) in the relevant particular kind of relationship (i.e. fiduciary relationship).Footnote ⁹³ Let us now turn to our argument for that very thing.Footnote ⁹⁴

V. DEFINING FIDUCIARY POWER

So the value of best interests loyalty, subjectively construed, remains unresolved. How then, if at all, might we progress? Let us return to Lionel Smith’s underlying strategy: the explanation, if any, of the value of subjective loyalty must lie in the peculiar nature of fiduciary power itself, and that is the power to act “for” (or “as” or “on behalf of”) another in some sense. Echoing Fox-Decent, fiduciary power is the ability, in some sense, to wield another’s legal “personality.”Footnote ⁹⁵ It must be something about holding this kind of power, where the beneficiary’s personhood is at stake, that triggers the correlating duty of loyalty. But what? I shall proceed to answer this question in two steps: first, I shall set out a conception of “personality” that gives a sense of how one individual (a fiduciary, F) might wield the power inherent in another’s personality (the beneficiary, B). I shall then offer, second, an argument for why the fiduciary’s subjective loyalty is required to respect the beneficiary’s personhood in just such a relationship.

This is no place to offer a full-fledged account of what it is for an individual to be a “person” at law.Footnote ⁹⁶ However, let me simply assume the following two necessary conditions. First, B is only a person if B is the kind of entity that not merely has interests but also a claim to means that can be used to further those interests. Such means are to be set aside as B’s—as “theirs”—rather than others’ (or no one’s). Hence, such a claim must be recognized in terms of rights (and/or other empowering normative incidents qua permissions, normative powers and immunities).Footnote ⁹⁷ Second, B is only a person if B also has a moral claim against others that some agent has the power to determine if and how B’s said rights (qua means) are to be dealt with (or not) to pursue B’s interests (qua ends). We might call this power to undertake such dealings (or not) with B’s rights the “personal power” of B.Footnote ⁹⁸

Presumptively, of course, we might assume B ought to hold their own personal power, and indeed B will have a further claim against others that they respect this fact. But, it is not in itself inconsistent with B’s status as a person if, for some reason, it is another, F, who wields B’s personal power so defined. And this is precisely what arises in a fiduciary relationship. In such a case, F’s agency “personates,”Footnote ⁹⁹ “substitutes,”Footnote ¹⁰⁰ or “transubstantiates,”Footnote ¹⁰¹ for that of B in dealing with B’s rights.

Having said this, what does it mean exactly for F to hold the power to “deal with” B’s rights in such a case? F might deal with B’s rights descriptively (e.g. literally using a tool owned by B to achieve some end), or normatively (e.g. selling B’s right to such another party C). But either way, the mark of such dealing qua exercise of B’s personal power is that B is left with the normative consequences of the action (or, indeed, inaction), in effect as if they had performed the action themselves. So, even if there is only a descriptive use of B’s rights (as with the tool) by F, but B bears any arising liabilities (and/or other disempowering normative incidents qua no-rights, duties, and disabilities), then F is wielding the personal power of B. By contrast, if F is liable to others for the consequences (e.g. some resulting harm to C), then they are not wielding B’s personal power. F merely has a permission to use something that is B’s by rights (e.g. the tool). In short, the upshot is that: if F holds the personal power of B in some respect, then F de jure or de facto controls the normative position of B in relation to others in that respect.

This is, one might think, relatively obvious—as it should be—in the paradigmatic case of trust. In a trust, F holds property for B, in the sense that while F holds the legally recognized right over the property, B has the beneficial interest in it—that is, the property is set aside as a means for B against the possible claims of others.Footnote ¹⁰² As F wields that right, however, B must bear the consequences of what is done: for example, if F sells trust property, then another party C comes to have a right to the property, and B loses their beneficial interest in it.

Consider too the case of parent and child, introduced by Fox-Decent above. As a person, the child, B, has certain means set aside for them qua rights, including most obviously over their own body, such that others have a duty not to interfere with them. However, a parent, F, has the power to determine that others may do so, for example, for the purposes of medical care. In this case, the child bears the consequences, by way of the resulting “no right” against the doctor’s touch.Footnote ¹⁰³

Take also the prima facie somewhat harder case of an adviser and their client.Footnote ¹⁰⁴ An adviser, F, may not hold any normative power de jure with respect to the rightful means of the one they are advising, B. But an adviser, F, at least presumptively, has sufficient control over their client B’s actions, by way of the latter’s trust, vulnerability, or incapacity, that they are able—at least, in part—to determine B’s actions resulting from such advice. F will, therefore, in that part be morally responsible for what B does. However, when B acts (or not) on the basis of such advice by F in this way, others can still rightfully treat B as if she were fully responsible for her resulting actions. B cannot disclaim responsibility for such actions as, in the relevant part, “not her own”; and thus will be subject to any resulting duties (and/or other disempowering normative incidents). The result is that, even without any normative power, de jure, with respect to B’s means, F still has such power, de facto, because F has the power to change B’s rights (and/or other empowering incidents) in ways that place B under new duties (and/or other disempowering normative incidents qua no-rights, liabilities, and disabilities).Footnote ¹⁰⁵

Now, if this is an adequate definition of the kind of power that fiduciaries hold, what then is the value in F manifesting subjective loyalty in its exercise? What is the value in loyalty qua F seeking to pursue the best interests of B as F perceives them, even if F gets those interests wrong?

Consider the absence of any duty regulating F’s use of B’s personal power. In this case F could exercise (or not) that power as they wish. Or to be precise: F would have (to the relevant extent) the recognized power to deal with B’s rights (and/or other empowering incidents) in ways that place B under new liabilities (and/or other disempowering normative incidents)—at their own whim. At whim, F—depending on the scope of the particular power—could place B under terms of contract; F could sell, lease, or grant permission to use B’s property; F could grant themselves or another permission to use B physically; and so on. And, in each instance, B would have to accept such actions as normatively effective, with their disempowering corollaries (liabilities, duties, no-rights and disabilities), and with no claim for redress. It would be as if she had performed the actions herself, but they were not for herself or even determined by herself.

The result, I suggest, would be that B would find herself—at least within the relevant scope of the personal power—to be in a status akin to being F’s property. B would find her legal personality—qua her personal power—available for use in service of another’s—F’s—ends. This ownership relation follows most clearly from a “normative standing” conception of property. If we assume that the relation of owner to property is one in which the owner has the “power to change the situation of others in relation to a particular object,”Footnote ¹⁰⁶ then we find that, by virtue of holding the personal power of B, F has just such a power to change the situation of others in relation to B at whim, empowering them with respect to B. F would have the power to change, at their own “say so,” depending on the scope of the power, what rights, permissions, privileges, immunities and/or powers they may have in relation to B—including both her physical person and other means.

Importantly, such a relation is best construed as an appropriation of the beneficiary herself as property, not merely the appropriation of some normative power that was her property but now is not.Footnote ¹⁰⁷ This is because it is an appropriation by F of the power to change the normative position of the beneficiary, B, herself as a particular object in relation to others—to control B’s liabilities, duties, immunities and no-rights, rather than the mere acquisition by F of a power—that was previously B’s by right—to change the normative position of some other particular object, X, in relation to others. Although, of course, by holding the former power, F might well execute the latter appropriation as a form of self-dealing.

This proprietary relationship between one agent, F, and another, B, I call “despotism.”Footnote ¹⁰⁸ In its primary Classical usage, “despotes” means “lord” or “master” with respect to a household, but unlike its cognate “kyrios,” emphasizes an ownership, rather than merely hierarchical relation of subjection.Footnote ¹⁰⁹ Thus, the paradigmatic despotic relationship is where one agent owns another: the master-slave relationship. But, once again, the emphasis is on the slave as property rather than the slave as subject; something that can be bought, sold, and used at the owner’s whim rather than someone who can merely be directed to perform certain acts. Of course, in any particular case, F might not hold all of B’s personal power, only some part of it. F may not own B in toto. But so long as they can wield some part of such personal power at whim, they maintain a proprietary interest in B: a despotic claim upon them.

Now consider, alternatively, what arises where F holds the personal power of B but is obliged to pursue B’s best interests as F perceives them. In this case, F is excluded from using such power at whim for their own ends, or desires. Indeed, instead, F is obliged to put their own agency at the service of B’s interests, whether they want to or not. As a consequence, they are prohibited from treating B in a proprietary fashion. Instead, F is required to make himself an instrument of B as an end. Of course, in pursuing B’s best interests as they perceive them, F may get those interests incorrect. But, in itself, this is not despotic. B’s personal power is used in pursuit of B’s interests as ends, even if incorrectly; and B’s personal power remains prohibited for use for anyone else’s ends, including F’s.Footnote ¹¹⁰

So, if the argument above is correct, then the duty of subjective loyalty excludes despotism in a fiduciary relationship. My final claim is that this is valuable; indeed, for two reasons: first because of what it avoids (its “negative” value)—the wrong of despotism—and, second, because of what it does—shows respect for the person of the beneficiary in doing so (its “positive” value).Footnote ¹¹¹

First, I take the wrong of despotism—holding another in the status of “property” rather than “person”—to be, in itself, decisively intuitive. But can we say any more as to why? Well, first, reflecting on the previous arguments for loyalty above, we can clarify why not. Despotism is not necessarily wrong because it disrespects the beneficiary’s agency (or autonomy, or “internal freedom”). After all, once again, a person need not have such agency. Further, holding another as property is, indeed, inconsistent with the person’s “external freedom,” that is, the person setting their own ends and pursuing them with their own means. But this will also be true in a non-despotic loyal fiduciary relationship too. Similarly, holding another as property is also inconsistent with the person’s freedom as “non-domination,” that is, freedom from the alien control of another. But once again, that too will be true in a non-despotic loyal fiduciary relationship. So neither the preservation of external nor republican freedom can explain the wrong of despotism (at least in contrast to loyalty).

A more promising explanation, instead, lies not in Kant’s Metaphysics of Morals, but in his Groundwork, in the prohibition on treating any person “merely as a means” in his Formula of Humanity: “So act that you treat humanity, whether in your own person or in the person of any other, always at the same time as an end, never merely as a means.”Footnote ¹¹² Despotism is precisely a relation where one agent is normatively positioned as a mere means for another as end—qua property—rather than an end in oneself. By virtue of this fact, therefore, one might say that it constitutes a kind of disrespect for the moral personhood of another; disrespect of one’s status as an end in oneself not a mere means. It is wrongful to be in this kind of relation with another; and a wrong that grounds a right in any moral person not to be in such a relation.

That, I suggest, is the wrong in despotism, but now turning to how to avoid it, on many interpretations, the Formula of Humanity just leads us back to the kind of freedom-implying argument in the Metaphysics of Morals. That is, we can only act in ways that do not treat another “merely as means,” if and only if we treat them in ways consistent with their equal freedom, and hence their consent. However, this is not correct: we also treat another as an end, and not merely a means if (1) we recognize them as having means protected by rights, and (2) we determine the use (or not) of those means with the exclusive aim of pursuing that person’s interests, that is, pursuing that person herself ‘as an end’Footnote ¹¹³ This does not entail the external freedom of that other, but it is still an alternative manner of adhering to the Formula of Humanity with respect to them.Footnote ¹¹⁴ It is a way in which that other may not exercise an ability to determine their own ends and pursue them with their own means but still is treated by us as an end not merely a means because they are treated by us as the sole end of their own means. In this way, we do not merely avoid the wrong of despotism—after all that could be achieved by simply divesting oneself of the personal power of another—but do so in a valuable way, a way that constitutes a form of ongoing expressive respect paid to the personhood of that other.

So to conclude, the value of subjective loyalty then is twofold: first, it avoids the disrespect of treating another as mere means—as a non-person; and it is also a particular way of achieving such avoidance that expresses respect for that other as a person—as an end worthy of their own dedicated means. But is it the only way?

VI. Varieties of service loyalty

Our argument so far has been focused upon defending the subjective prescriptive construal of the duty of loyalty as it is paradigmatically formulated, that is, as a “best interests” standard: the fiduciary is obliged to pursue the best interests of the beneficiary. In so doing, I have, by implication, sought to strengthen its hand in contrast to the opposing “proscriptive” construal of this same paradigmatic standard (and, hence, as a result the risk of the collapse of fiduciary law effectively into contract). We now can point to an underlying moral duty to avoid despotism that the subjective duty of loyalty discharges in a valuable manner that manifests respect for the moral personhood of the beneficiary.

However, somewhat regardless of the correct construal of the “best interests” standard itself as the requirement of fiduciary loyalty in general, recent commentators have stressed that in some relationships, at least in some jurisdictions, loyalty would appear to demand departures from it.Footnote ¹¹⁵ In these particular contexts, loyalty requires the pursuit of a principal’s wishes, or conformity with moral duties, or the pursuit of some impersonal purpose—not simply the pursuit of the beneficiary’s best interests (however construed). So how then does our argument above relate to these departures from the paradigmatic “best interests” standard of loyalty? I shall first address cases that still involve specific identified beneficiaries (“service relationships”) in this section, and those that prima facie do not (“governance relationships”) in the next.Footnote ¹¹⁶

In the case law, there are at least two kinds of departure from the “best interests” standard for loyalty while still clearly in service of particular identifiable beneficiaries.

The first, and most obvious, is where loyalty requires that the fiduciary seek to pursue the wishes (or will or preferences) of the beneficiary instead of their best interests. The central case is the agency relationship.Footnote ¹¹⁷ The defining feature of the agency relationship, as distinct from other fiduciary relationships, is that the principal (qua “beneficiary”) retains control over their agent (qua “fiduciary”); and the primary means of such control is the agent’s duty of obedience, that is, the agent is obliged to comply with the principal’s instructions. In this respect, the agent is clearly obliged to follow the principal’s will. However, in and of itself, this requirement is not inconsistent with the agent’s duty of loyalty being to pursue the principal’s best interests, nor even is it inconsistent with construing such compliance as loyal behavior by the best interests standard. This is because such instructions could be construed simply as constraints on the agent’s power, channeling the pursuit of best interests in particular ways rather than displacing it.

The real issue arises where such instructions underdetermine action, that is, where instructions are unclear, or conflict, or there is a material change of circumstances, or they are simply silent. In such scenarios, no loyalty-constraining duty of obedience can be at stake, because there is no relevant instruction per se to obey. So, just as discussed above when the purposes of a trust underdetermine action, the duty of loyalty in the agency relationship must operate to govern the residual discretion of the agent where instructions underdetermine action. However, instead of requiring that such an agent pursue the principal’s best interests as the agent perceives them to be, the duty of loyalty in an agency relationship requires that the agent seek to give effect to what the principal’s wishes would be with respect to the issue if she were to express them, that is, the agent ought to continue to “aim at accurate replication of how the principal would have acted at the time the agent determines what to do.”Footnote ¹¹⁸ And indeed, the agent ought to do what she believes the principal would have wished, even if this is contrary to what the agent believes would have been in the principal’s best interests.Footnote ¹¹⁹

The second kind of departure from the best interests standard is where that standard (or indeed the agency will standard) operates in general but is subject to what Andrew Gold has termed “internal constraints.”Footnote ¹²⁰ These are constraints on the pursuit of such ends, but instead of being construed by the courts as constraints on loyalty (“external constraints”), they are in fact construed as demands of loyalty itself (“internal constraints”). There are two kinds of internal constraint at play. The first are other legal (or ethical) obligations owed to the beneficiaries themselves. For example, the duty of a director not to lie to shareholders. Delaware judges have held that the fulfillment of such a duty is not merely a constraint on the director’s duty of loyalty to pursue the best interests of the company but indeed a requirement of such loyalty itself.Footnote ¹²¹ The second kind of “internal constraints” are other legal (or ethical) obligations owed to others. For example, as Andrew Gold argues a lawyer’s loyalty to her client is not constrained by but includes her duty to uphold the rule of law, even if this comes at the expense of her client’s interests.Footnote ¹²²

So, given these doctrinal claims, how do they square with our anti-despotism argument above? Our two kinds of divergent standards of service loyalty can be understood as two alternative ways of avoiding the wrong of despotism.

First, the law can avoid despotism in a fiduciary relationship by placing the fiduciary under a duty to comply with the beneficiary’s (qua “principal”) will, as the fiduciary perceives that will to be. After all, it is hardly an ownership-like relationship with respect to another to be obliged by law to pursue the wishes of that other. Indeed, consistent with our critique of various attempts to ground best interests loyalty above (Sections II–III), it is plausibly a way of showing respect for the agency of the principal. This is to say, unlike the best interests standard, it does not show respect for the latter’s personhood per se—that they are an end to which means are dedicated—but instead their moral agency (their “internal freedom”)—that is, that they are a rational being that has the ability to set the ends for their means.

Second, the law can also avoid despotism by requiring the fiduciary’s compliance with the kinds of “internal constraints” detailed above. In such cases, the fiduciary F continues to be excluded from using their power at whim for their own ends, or desires; and F remains obliged in general to put their own agency at the service of B, that is, in pursuit of B’s interests or wishes. F only departs from that pursuit, either in aid of fulfilling some other obligation to B or to another D. But such a departure is not to instrumentalize B, treating her as a means to someone else’s end. First, by fulfilling obligations owed to B herself—like not lying to her—F merely offers respect due to B as a kind of person (or agent). And, by fulfilling obligations to others D, such as the obligations under positive law, or preserving the rule of law, F merely generalizes that kind of respect to others—in a way that B, indeed, herself owes to those others.Footnote ¹²³ Neither act is a proprietary—despotic—use of B.

So both the will-compliance standard and the internal constraints standard of loyalty are as consistent with the anti-despotism principle as the best interests standard (indeed, perhaps other standards could be too).Footnote ¹²⁴ But is this a problem then for our overall argument? We might formulate the putative objection this way: if the law can satisfy the requirement to make fiduciary power safe from despotism just as well by any of the above means, then what, if anything, is so special about the best interests standard that leads it to being the “general” standard to which the others are departures? Could not will-compliance instead just be the general standard? Or could not internal constraints loyalty be the norm rather than the exception?

If our argument above is correct, then the most important thing for fiduciary law—as the law governing relations where one agent holds the personal power of another—is that despotism is avoided. As a rule, this is the peculiar, unifying risk of all fiduciary relations, and it can only be avoided in a fiduciary relationship by obliging the fiduciary to be loyal in some sense. Footnote ¹²⁵ But which sense, in particular, is underdetermined by the need to avoid despotism itself. The law, therefore, is best construed as resolving such under-determination in each kind of fiduciary relationship by selecting the sense of loyalty that tends to have the most apposite value. Hence it selects the best interests standard when personhood is most apposite, will-compliance when agency is more apposite, internal constraints standard when certain modes of respect are particularly apposite, and so on.Footnote ¹²⁶ Indeed, it is entirely consistent with my argument that, at this stage, the underlying normative reasoning of law may indeed be quite instrumental, that is, while the need for some variety of duty of loyalty is justified non-instrumentally by the anti-despotism argument, the choice between such varieties to fulfill that need may be made on the basis of what will tend to best promote relevant interests in aggregate.

Having said all this, however, it would be inconsistent with the anti-despotism principle to leave the particular way of being loyal up to the discretion of the fiduciary. This is because that would still leave the fiduciary to choose between how they are loyal, on the basis of their own whim.Footnote ¹²⁷ Anti-despotism precludes any such discretion.

So why, then, is the best interests standard the general standard of loyalty to which the others are best considered departures?

We might infer then that this reflects the judgment that respect for the personhood of the beneficiary is generally of most apposite value in fiduciary relationships. For example, the best interests standard makes far more sense in a range of contexts than the will-compliance standard, such as where the beneficiary is liable to lack agency, or the will of that beneficiary will tend to be hard to construe, or is likely to be misinformed, or ill-equipped to make a decision; or, indeed, is a collective liable to have no clear will; or the purpose of the trust was never to serve the will of the beneficiary but instead to serve that of the settlor; and so on. Conversely, the agency relationship may be construed as an exception where respect for agency is the more salient value at play. But, having said this, fiduciary law could be quite different, or even different across jurisdictions, reflecting different judgments on such matters, and yet still be recognizably fiduciary law, so long as when presented with a case of one agent wielding the personal power of another it required some kind of loyalty from the former agent that excluded despotism.Footnote ¹²⁸

VII. Varieties of governance loyalty

If the duty of loyalty, regardless of its prescriptive variety, is to be grounded in the anti-despotism principle, then it would appear to presuppose a defined beneficiary who is at risk of being placed in a despotic, property-like relationship. However, another subset of common fiduciary relationships might be thought to fail this presupposition, specifically, public officials, corporate directors, and charitable trustees. Prima facie, one might think in each case the fiduciaries’ duty of loyalty is owed to no identifiable beneficiaries, but rather an impersonal purpose. But if that is so, then what of the anti-despotism argument? Let us deal with each case in turn.

First, a growing number of legal and political theorists have come to endorse the view that public offices should be construed as fiduciary roles and that the duty of loyalty, operating with respect to any discretion left by the other constraints on such roles, is to be construed as a duty to pursue the public interest, as the public officer perceives it to be.Footnote ¹²⁹ I take my analysis above to imply the following justification of this view: public officials hold “public power,” where such “public power” is to the public as, in effect, a private law fiduciary “personal power” is to a specific person (or set of persons). It is via public power that the public agent acts for (as, or on behalf of) the public. Hence, the public agent owes a subjective duty of loyalty to that public to exclude the possibility of being in a despotic relationship with respect to that public. This reasoning applies not merely to any particular public officer but to any public agent, including public organizations. No doubt in many cases, such public agents will find their powers regulated by purposes in much the same way as other fiduciaries, but the underlying duty of loyalty to the public remains to resolve under-determination left by any such purposes qua discretion.

Having said this, do public agents really wield the “personal power” of the public at large? My response is yes, quite literally. To hold that F holds the “personal power” of B is simply to say that F holds the power to deal with B’s rights (and/or other empowering normative incidents—permissions, normative powers, immunities, and the like)—to use them, change them, exchange them, extinguish them—in ways that leave B responsible for the consequences. This is, I suggest, an entirely accurate description of the power that the state wields vis-à-vis its people, not merely individually, but collectively. And hence, on my view, it follows that this power, if unregulated by loyalty, is despotic.Footnote ¹³⁰ The state not merely changes our normative position by law, most obviously; it is also able to act with our resources (e.g., our property, our sovereign powers internationally, etc), in ways that, ultimately, leave us liable for the consequences—as if we had performed the relevant act ourselves. Indeed, there is nothing conceptually incoherent, or indeed historically unprecedented, in having such a slave-like “despotic” relationship to the state, something the ancient Greeks themselves well appreciated when they first expanded the usage of the term “despotismós” from master–slave relationships to name the form of government they associated with the “barbarians”—a form of government where the ruler effectively held their subjects like property.Footnote ¹³¹

Second, prima facie company directors would appear to fit neatly with my argument: they are fiduciaries of the companies they govern; the company is a person at law; the company directors act on its behalf—wielding its “personal” power, as defined above; and in so doing, the company directors owe a duty of loyalty to the company, typically construed subjectively as a duty to serve the best interests of the company as they perceive those interests.Footnote ¹³² The problem, however, is that when we turn to explaining the latter duty, we can no longer appeal directly to the wrong of despotism—as a form of disrespect to the moral personhood; nor the value of subjective best interests loyalty—as a form of respect for such moral personhood. This is because the company is not a moral person—it is merely an artificial person.Footnote ¹³³ And, of course, companies are in fact owned (by their shareholders), and there is nothing at all wrong with that. So what then can explain the value of subjective loyalty in such a case?

I suggest the answer is a cognate anti-conversion principle—rather than anti-despotism principle—that is, loyalty is required to avoid company directors standing in an improper ownership-like relation with respect to something (that is, the company) that is in fact owned by others (that is, the shareholders).Footnote ¹³⁴ The corporate duty of loyalty then avoids this wrong, since being obliged to act in the best interests of the company is not an ownership-like relation with that company.Footnote ¹³⁵ This is the negative value of having such a corporate duty of loyalty. But a further advantage of this analysis is that just as with the anti-despotism principle, it also permits a variety of conceptions of what corporate loyalty actually requires of directors that may all satisfy the anti-conversion principle, although with different positive values. So, it might be satisfied by identifying the company’s interests with its shareholders’ interests,Footnote ¹³⁶ with its purposes,Footnote ¹³⁷ or indeed both;Footnote ¹³⁸ and it might be satisfied if the directors pursue such aims subject to “internal constraints,” such as not intentionally violating positive law.Footnote ¹³⁹ Different jurisdictions may take such different views in different kinds of case for different reasons, but they would all be unified by the same basic anti-conversion principle: the company is the property of its shareholders, not its directors.

Finally, the hardest case for my argument—I think at least—is charitable trusts. At common law, charitable trusts have no specific beneficiaries—no one with any legally recognized claim to benefit from the trust. Nor can the charitable trust itself be the true object of the trustees’ duty of loyalty, since it has no legal personality. Instead, the charitable trust is established for the promotion of a charitable purpose. The trustee’s duty of loyalty is to pursue such purpose; and to do so, subjectively construed, in the best way as they perceive it to be.Footnote ¹⁴⁰ So how do we justify this kind of duty of loyalty without a beneficiary?

First, one might seek to ground the justification of such a duty in the value of the trust’s purpose itself. Indeed, at least at common law, the purpose has to have some value qua public benefit to qualify as a valid purpose.Footnote ¹⁴¹ However, the duty of loyalty is, once again, subjectively construed. Thus, as per our original objection (Section I) doing what one perceives to be in the best pursuit of such a valuable purpose may not itself be at all valuable; indeed, it may be quite counter-productive to such value.

In response, one might then hold that while subjective loyalty may not necessarily promote the value of the purpose, it at least shows respect for that purpose. This may be true. But why would charitable trustees necessarily owe such respect for any particular purpose? The most obvious answer would be that it is the purpose they agreed to promote when becoming trustees.Footnote ¹⁴² But then (a) the justification for the duty of charitable loyalty is at risk of collapsing into purely a contractualist one; (b) the duty would be logically obviously owed to the settlor, yet without any express term in the trust, the law does not recognize the settlor as having any correlative rights; and (c) if the trustees’ duty of loyalty is only grounded in respect for the specific purpose of the specific trust, then—repeating cognate issues about purposes above—it is hard to explain how loyalty might regulate behavior where such purposes underdetermine action, or indeed where they become illegal, impossible, or impractical.Footnote ¹⁴³

Tentatively, therefore, I suggest an alternative argument. The morally defining feature of a charitable trust is not its specific purpose, but rather the act of charity itself, that is, the alienation of means available to oneself as an end and their dedication instead to an end of public benefit. I submit that such an act itself, the act of charity—of dedicating one’s means to a valuable end (or ends) other than oneself—is itself inherently valuable, somewhat independent of its consequences, manifesting a kind of moral worth due respect.Footnote ¹⁴⁴ A trustee’s duty of loyalty is then grounded in respect for such an act. Such respect is not owed to the settlor per se, but to their act. In the same way that, for example, a work of art may be due respect because of its worth, distinct from any respect due to its artist.

It follows then that, in general, the most apposite manifestation of such respect is to seek to serve as best as possible the specific charitable intention of the charitable act qua the charitable trustee’s duty of loyalty. But because this duty is ultimately grounded in respect for the settlor’s establishment of the charitable trust as an act of charity rather than the respect for the original specific end of the trust thereby established, it is not incoherent that where this end for some reason underdetermines the trustees’ action (or otherwise loses normative force due to becoming illegal, impossible or impractical), then the trustees may still remain loyal by seeking to alter the terms of the trust to pursue some new end that approximates the settlor’s charitable intent. That would still respect the act of charity itself that established the trust, even if not the original purpose of the trust so established.