Home | Databases | WorldLII | Search | Feedback University of Melbourne Law School Research Series

Harding, Matthew --- "Trust and Fiduciary Law" [2013] UMelbLRS 9

Last Updated: 27 June 2014

This paper was first published in the Oxford Journal of Legal Studies, Vol. 33, No. 1 (2013)

Trust and Fiduciary Law

Matthew Harding*

Abstract—How can it be that the fiduciary relationship has trust at its core if trust is neither a necessary nor a sufficient condition for the existence of such a relationship? My aim in this article is to make some arguments that I think might assist in solving that puzzle. First, I argue that fiduciary relationships are likely to be characterized by relatively ‘thick’ interpersonal trust. Secondly, I argue that moral duties referring to trust play a role in the justification of fiduciary duties, but that the role of trust in the underlying moral duties is contingent, yielding only a contingent connection between trust and fiduciary duties. Finally, I argue that a goal of fiduciary law should be enabling and supporting trusting relationships, but that this goal should be viewed within a broader liberal outlook according to which fiduciary law also enables and supports relationships on terms of detachment.

Keywords: equity, morality, normativity, discretion, private law, duty

1. Introduction

The fiduciary relationship has trust, not self-interest, at its core.

Norberg v Wynrib (1992) 92 DLR (4^th) 449, 488 (McLachlin J)

[A]n actual relation of confidence – the fact that one person subjectively trusted another – is neither necessary for nor conclusive evidence of the existence of a fiduciary relationship.

Hospital Products Ltd v United States Surgical Corporation [1984] HCA 64; (1984) 156 CLR 41, 69

(Gibbs CJ)

These two statements, considered side by side, reveal a puzzle in fiduciary law. How can it be that the fiduciary relationship has trust at its core if trust is neither a necessary nor a sufficient condition for the existence of such a relationship? My aim in this article is to make some arguments that I think might assist in solving that puzzle. The article is laid out in three parts. In

* Melbourne Law School, University of Melbourne. Email: m.harding@unimelb.edu.au. This paper had its origins in a DPhil thesis that I completed in 2006 at the University of Oxford and I must thank my supervisor Dori Kimel for all his advice and support throughout that project. Thanks also to Andrew Robertson for helpful advice, and to Michael Bryan, Richard Nolan and Lionel Smith for insightful and challenging comments on a draft. All errors are of course mine.

© The Author 2012. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

Section 2 I argue that, although trust is neither a necessary nor a sufficient condition for the existence of a fiduciary relationship, a fiduciary relationship is nonetheless likely to be characterized by trust on account of the nature of trust. I also argue that this trust is likely to be relatively ‘thick’ trust on account of the nature of fiduciary relationships. To this extent, it may be said that fiduciary relationships have a strong association with trust. In Section 3, I argue that moral duties referring to trust play a role in the justification of fiduciary duties. However, I also argue that trust figures only contingently in the justification of those underlying moral duties, yielding only a contingent connection between trust and fiduciary duties. In Section 4, I turn my attention away from the content and character of fiduciary relationships and duties to the broader question: in the service of what goals ought the state to subject fiduciary relationships to rules of law? I argue that once attention is paid to the goals that should be served by fiduciary regulation, the assertion that trust is at the core of the fiduciary relationship begins to make sense, notwithstanding that fiduciary relationships are not invariably trusting and that fiduciary duties are not justified by underlying moral duties grounded in trust.

2. Trust and Fiduciary Relationships

An understanding of the attitude of trust begins with an account of the optimism that characterizes trust, an optimism that depends in part on beliefs, commitments and background attitudes, and that is, in some sense, a response to risk. In light of the optimism of trust, a person who trusts tends to adopt a stance of positivity, and therefore is disposed to think favourably about, that with respect to which she trusts, notwithstanding the circumstances of risk in which her trust has formed. Trust is not the only attitude that is characterized by optimism in circumstances of risk. But trust is unique in that it entails optimism about choices that it is expected or anticipated that other people will make; the risk to which trust responds is the risk that human freedom presents in circumstances of human interaction. In this, trust may be distinguished from other attitudes of optimism in the face of risk: confidence in the workings of systems; faith in God; hope that events will play out one way or another.¹

When trust is understood in this way, even the most cursory reflection on the sorts of relationship that are undoubtedly recognized as fiduciary in law confirms that trust is neither a necessary nor a sufficient condition for a fiduciary relationship to form; it is not difficult, for example, to imagine a fiduciary relationship ‘gone sour’ in which the relationship continues to be

¹ Two points of clarification: (i) trust in oneself is not directed at the choices of other people, but rather at one’s own expected or anticipated choices; and (ii) the sort of background trust that is necessary to social living is directed at choice only in a highly generalized way. The trust that I discuss in this article is interpersonal trust, directed at the choices of other people in settings of particular human interactions; I take it as clear enough that when judges and scholars describe fiduciary relationships in terms of trust, it is interpersonal trust that they have in mind.

a fiduciary one even though trust has broken down completely. Nonetheless, I want to argue that fiduciary relationships are likely to be characterized by trust: in other words, such relationships are likely to entail at least one party to them forming an optimistic attitude about the choices that the other party will make. This is because of the character of trust itself. Moreover, I want to argue that because of a feature that fiduciary relationships all share, the trust that is present in fiduciary relationships is likely to be relatively ‘thick’ trust.

Like all attitudes, the attitude of trust usually comes combined with, and

lends character to, particular beliefs. In the case of trust, the beliefs in question are invariably beliefs about other people’s choices, but beyond that it is difficult to say anything with certainty about the content of trusting beliefs. Some writers have identified as the core case of trust the case where A has an optimistic belief that B will make a choice, most likely in some matter affecting A’s interests, based on the thought that A is ‘counting on’ B.² In this case A’s trust is combined with a belief about B’s good will. It is of course possible to isolate some normative dimension of trust, for example, its propensity to trigger and nourish human relationships of a certain type, and then posit trusting beliefs in good will as the core case of trust in light of that normative dimension. But this should not be allowed to obscure the fact that, conceptually, all that is required for a belief to be trusting is that it be combined with an attitude of optimism and directed at human choice. For example, A may have a trusting belief that B will choose in accordance with the requirements of some role that B occupies, or because B is A’s friend, or based on values or principles that B is known to adhere to, or even out of self- interest. Trust is a very widespread phenomenon, so widespread that its role in our epistemic lives is often taken for granted.

That said, I think it is possible conceptually to distinguish between what

might be called ‘thin’ and ‘thick’ forms of trust. The thickness of trust is to be assessed in light of three variables: (i) the range of expected or anticipated choices at which it is directed; (ii) the importance of those choices to the person doing the trusting; and (iii) the content of beliefs that it is combined with. This last variable suggests a distinction between beliefs in a person’s basic rationality or consistency of action on the one hand, and beliefs about a person’s character, ethics, relationships, roles or interests on the other. The profile of thin trust might best be illustrated by an example. A is going on holiday and asks his neighbour B to water his garden while he is away. A trusts B to water the garden in the sense that he optimistically believes that B will water the garden, but he believes that B will do so only because B is going away next month and is likely to want A to return the favour, and he does not care greatly whether or not B actually waters the garden. A’s trust is thin in

² See eg Philip Pettit, ‘The Cunning of Trust’ (1995) 29 P&PA 202, 204; Karen Jones, ‘Trust as an Affective

Attitude’ (1996) 107 Ethics 1, 4; Trudy Govier, Dilemmas of Trust (McGill-Queen’s University Press 1998) 6.

light of all three variables described above. By contrast, imagine that A, who is dying of cancer, asks his oldest and dearest friend B to take and raise his children after he dies. B promises to do so. A dies in peace, trusting B to carry out her promise. A’s trust is thick: it is directed at an open-ended set of choices that A expects that B will make in raising A’s children for years to come; these choices are of the utmost importance to A; and A’s trusting beliefs are informed by his longstanding friendship with B and the fact that B has made him a promise, and perhaps by other considerations as well, such as A’s perception of B’s moral integrity or A’s knowledge of the way in which B has skilfully and successfully raised her own children.

Given the breadth of trust, and the variety of trusting beliefs that might be formed in circumstances of human interaction, I do not think it is controversial to assert that in many if not most fiduciary relationships, trust is present in at least a thin form. For example, consider a fiduciary relationship that forms for a relatively brief period of time, pursuant to a contract, in the pursuit of some specific commercial purpose, say between joint venture partners. The parties in that relationship may have optimistic beliefs about each other’s likely choices within a limited range delineated by the shared purpose, based on nothing more than generalized expectations about rational behaviour in commercial settings. This is a thin form of trust, but trust nonetheless. Thicker forms of trust may arise where—as is often the case—a fiduciary occupies a role to which specific social meanings and expectations attach, or where a fiduciary relationship coincides with a family relationship, as might occur where the trustee of a family trust is herself a family member. Similarly, trust in fiduciary relationships is likely to be relatively thick where the relationships in question are enduring ones—think again of the family trust, or the relationship of partners—and it is likely to thicken over time just as it tends to do in respect of any enduring human relationship. But the point I want to emphasize for now is that fiduciary relationships may be characterized by even thin trust, and on that basis I think it may fairly be asserted that fiduciary relationships are typically— even if not invariably—characterized by trust of some sort.

Although I think this assertion is true, I also think it is rather uninteresting: it amounts to saying that human interaction is typically characterized by trust, that fiduciary relationships entail human interaction, and therefore that fiduciary relationships are typically characterized by trust. This sort of argument could hardly animate the claim that the fiduciary relationship has trust ‘at its core’; if it did, the proper response to that claim would be ‘so what?’ I take it that those who have claimed that the fiduciary relationship has in some way to do with trust have wanted to say not that fiduciary relationships, like most interpersonal dealings, are characterized by trust, but rather that the link between trust and fiduciary relationships is somehow special or significant. In the rest of Section 2, I want to argue that the link between trust and fiduciary relationships might be viewed as special or

significant in the sense that fiduciary relationships are likely to be characterized by relatively thick trust. This argument supports a claim that there is a strong association between trust and fiduciary relationships.

Isolating that which makes a relationship fiduciary is one of the great unfinished tasks of modern legal scholarship. Many have sought a distinctive fiduciary principle in the notion of an undertaking to act in a representative capacity;³ others in the concept of reasonable expectations;⁴ others still in the moral facts of vulnerability and dependence.⁵ Some have given up the search for a unifying feature altogether,⁶ and others have downplayed it, preferring to concentrate their attention on the nature and functions of fiduciary duties.⁷

The difficulties in identifying that which makes a relationship fiduciary are twofold: first, candidates for a distinctive fiduciary principle invariably fail to explain every type of relationship that has been or (according to a consensus of opinion) should be regarded as fiduciary in law; and secondly, such candidates invariably describe features of human interactions that are not and (again by consensus) should not be regarded as fiduciary relationships in law. Call these two problems the problems of underdetermination and overdetermination, respectively.

I want to focus on one feature that in my view is shared by all fiduciary relationships, and I want to argue that, in light of this feature, fiduciary relationships are likely to be characterized by relatively thick trust. However, before I do so, I want to make clear that the feature in question is not that which makes relationships fiduciary; in other words, I do not claim to have found the elusive fiduciary principle. Because the feature I will focus on is present in each and every fiduciary relationship, it escapes the problem of underdetermination, but it does not escape the problem of overdetermination as it may be found in a range of uncontroversially non-fiduciary human interactions as well. For present purposes, escaping the problem of underde- termination will be sufficient. To demonstrate that each and every fiduciary relationship has a propensity to generate or nourish thick trust because of a common feature is, to my mind, theoretically worthwhile in the setting of a wider consideration of the relation of trust and fiduciary law, and may indeed

³ Eg Austin Scott, ‘The Fiduciary Principle’ (1949) 37 Cal L Rev 539; Lionel Smith, ‘The Motive, Not the Deed’ in Joshua Getzler (ed), Rationalizing Property, Equity and Trusts: Essays in Honour of Edward Burn (LexisNexis 2003) 53.

⁴ Eg Paul Finn, ‘The Fiduciary Principle’ in TG Youdan (ed), Equity, Fiduciaries and Trusts (Carswell 1989) 1.

⁵ Eg Hospital Products Pty Ltd v United States Surgical Corporation [1984] HCA 64; (1984) 156 CLR 41 (HCA) 142 (Dawson J); Lac Minerals Ltd v International Corona Resources Ltd (1989) 61 DLR (4th) 14 (SCC) 68–69 (Sopinka J). See also the discussion in Leonard Rotman, ‘Fiduciary Law’s ‘‘Holy Grail’’: Reconciling Theory and Practice in Fiduciary

Jurisprudence’ (2011) 91 BU L Rev 932.

⁶ Eg John Glover, ‘The Identification of Fiduciaries’ in Peter Birks (ed), Privacy and Loyalty (Clarendon Press

1997) 269.

⁷ Eg Matthew Conaglen, Fiduciary Loyalty: Protecting the Due Performance of Non-Fiduciary Duties (Hart

Publishing 2010), although note ch 9. James Edelman, ‘When Do Fiduciary Duties Arise?’ (2010) 126 LQR 302 presents the question of what makes relationships fiduciary in a new way, arguing that whatever it is, it must justify the implication of fiduciary duties as terms of voluntary undertakings. However, Edelman too stops short of nominating a distinctive fiduciary principle.

form a component in a complete theory of fiduciary relationships even if it does not constitute a complete theory. At the very least, it will serve to distinguish fiduciary relationships from those interpersonal dealings that do not have the feature that all fiduciary relationships (and some other dealings) share, and this will focus attention on what distinguishes fiduciary relationships from the non-fiduciary dealings that do exhibit the feature in question. Such a narrowing of the inquiry is undoubtedly a theoretical gain.

The feature I want to focus on is this: in all cases where fiduciary

relationships arise, one person voluntarily allows another to exercise discretion in making choices that will affect the interests of the first person or at least interests that the first person cares about.⁸ I will call this feature of all fiduciary relationships ‘reliance on discretion’. Sometimes, each party to a fiduciary relationship relies on the discretion of the other—consider partners—but more often reliance on discretion is a one-sided affair. It must be acknowledged that, historically, fiduciary language has sometimes been used to describe interper- sonal dealings, such as those entailed in resulting, constructive and bare express trusts that are not characterized by reliance on discretion: this suggests that reliance on discretion may raise the underdetermination problem I alluded to above. However, the underdetermination problem arises only where a claimed feature of all fiduciary relationships turns out not to apply to relationships that are uncontroversially regarded as fiduciary, and the view that resulting, constructive and bare express trusts are fiduciary relationships can hardly be described as uncontroversial.⁹ So, assuming that reliance on discretion is a feature of all fiduciary relationships,¹⁰ let me turn to two reasons for thinking that reliance on discretion renders fiduciary relationships fertile ground for thick trust.

First, the voluntary yielding up of interests that characterizes an act of reliance is widely understood, according to a complex of social conventions and meanings, to be a fitting way to manifest relatively thick forms of trust; moreover, thick trust is widely understood to be a fitting attitude to manifest in circumstances of reliance. This is not to say that reliance and thick trust stand in any sort of necessary conceptual connection: it is perfectly possible to rely on another person in the sense of voluntarily letting her deal with one’s interests

⁸ See Ernest Weinrib, ‘The Fiduciary Obligation’ (1975) 25 U Toronto LJ 1. Note that it is not always the case that the person relying on the discretion of the fiduciary is the person to whom the fiduciary owes fiduciary duties: the beneficiaries of a discretionary trust may not even be aware of the existence of the trust and cannot be said to voluntarily allow anything to the trustee, but the settlor of such a trust invariably does rely on the trustee’s discretion in just the way described in the text.

⁹ See Lionel Smith, ‘Constructive Fiduciaries?’ in Peter Birks (ed), Privacy and Loyalty (Clarendon Press

1997) 249, 262–64; William Swadling, ‘The Fiction of the Constructive Trust’ (2011) 64 CLP 399, although Smith argues that constructive and resulting trustees are not fiduciaries because they have not consented to their role and not because they do not exercise discretion, and Swadling rejects the proposition that constructive trusts can give rise to fiduciary or any other duties because he rejects the proposition that there are constructive trusts.

¹⁰ If it turns out that reliance on discretion is not a feature of all fiduciary relationships, then it may still be said with confidence that it is a feature of almost all such relationships, so that the under-determination problem,

while not eliminated, is negligible.

even though one has only thin trust in her or, indeed, distrusts her. It is simply to say that reliance and thick trust are closely connected in our social world,¹¹ a connection that is reflected nicely in the English verb describing the act of reliance accompanied by thick trust: ‘to entrust’.¹² This connection is formalized especially clearly in the practice of promising, a practice that enables A to invite the reliance of B by drawing on social conventions and meanings whose currency is trust.¹³ Because reliance and thick trust are widely understood to be closely connected, the fiduciary relationship—as one invariably characterized by reliance—is fertile ground for thick trust in two ways: first, A, who is deliberating whether or not to rely on B and thereby form a fiduciary relationship with her, is more likely to do so if she already has thick trust in B than if she does not, given that reliance is typically thought to be an appropriate way to manifest thick trust; and secondly, A, who is already in a fiduciary relationship with B even though she lacks thick trust in B, may develop thick trust in B because of her sense that that is a fitting attitude to have, in view of her reliance on B.

The second reason why the fact that fiduciary relationships are characterized

by reliance on discretion renders such relationships fertile ground for thick trust has to do with the nature of discretion. Discretion entails the making of decisions that are, to some degree, matters of personal judgment and assessment.¹⁴ To the extent that a decision-maker exercises discretion, then, the decision-maker is in a position to engage with the reasons that bear upon the decision free from the constraints of higher-order norms—like those generated by rules, commands, threats, promises, etc—that require a decision one way or another or that require that relevant reasons be excluded or weighed one way or another.¹⁵ Given this feature of discretion, if A relies on B’s discretion in some matter, A is confronted with the risk associated with B’s freedom of choice in a relatively acute way.¹⁶ The sorts of beliefs that might enable A to respond to B’s freedom by exhibiting only thin trust—for example, the belief that B will choose x because B is required by law to do so and B, like most people, typically obeys the law—are simply not available to A. Instead, if A’s reliance on B’s discretion is to be characterized by trust, that trust will most likely have to be combined with beliefs about B that render that trust thicker: beliefs based, for example, on B’s perceived good will, intelligence, experience

¹¹ So much so that some scholars confuse the two: see Pettit (n 2) 204; Annette Baier, ‘Trust’ in Tanner

Lectures on Human Values, vol 13 (University of Utah Press 1991) 109, 117 and 120; possibly Richard Holton,

‘Fiduciary Relations and the Nature of Trust’ (2011) 91 BU L Rev 991, 992 (‘to trust is not to believe in performance; it is to act as if one believed’).

¹² On entrusting, see Matthew Harding, ‘Manifesting Trust’ (2009) 29 OJLS 245, 253–60.

¹³ ibid 255–57; Dori Kimel, From Promise to Contract: Towards a Liberal Theory of Contract Law (Hart

Publishing 2003) ch 1.

¹⁴ Denis Galligan, Discretionary Powers: A Study of Official Discretion (Clarendon Press 1986) 8.

¹⁵ On the nature of these higher-order reasons: Joseph Raz, Practical Reason and Norms (OUP 1999) ch 1; Joseph Raz, The Authority of Law (2nd edn, OUP 2009) ch 1.

¹⁶ See Tamar Frankel, Fiduciary Law (OUP 2011) 25–26.

or integrity. Of course, theoretically, A might form the trusting belief that B, as someone who may be expected reasonably to apprehend the requirements of rationality, will be appropriately responsive to all relevant reasons and choose accordingly, and this belief would be consistent with thin trust. But it is difficult to imagine why or how A would form such a belief in the absence of accompanying beliefs about B’s character, ethics or interests, unless A believed that B was performing a role one of the requirements of which is precisely to respond to reasons in this way. And a trusting belief of that sort is, all else being equal, relatively thick as well.

Reflection on the nature of discretion shows that fiduciary relationships, as

relationships in which reliance on discretion invariably occurs, are likely to be characterized by trust understood as relatively thick in light of the third variable I mentioned above when introducing the distinction between thin and thick trust, viz, the content of beliefs that trust is combined with. Something further may also be said about the thickness of trust in fiduciary relationships in light of the first two variables: the range of expected or anticipated choices at which trust is directed; and the importance of those choices to the person doing the trusting. Discretion is usually exercised over time in respect of a set of choices in pursuit of some purpose or end: this set of choices is sometimes referred to as the ‘sphere’ of discretion. In the case of fiduciary relationships, the sphere of discretion tends to be significant, both in the sense that it entails an open-ended set of choices in pursuit of a purpose, and in the sense that the purpose in whose service discretion is exercised is often an enduring one, and equally often an important one for at least one of the parties to the fiduciary relationship. To the extent that this is true, such trust as there is in fiduciary relationships is likely to be relatively thick in light of the first two variables I described above, because it is directed at an open-ended set of choices, and because it is directed at choices made in the pursuit of a purpose regarded as important and consequently themselves regarded as important. Not all fiduciary relationships entail a significant sphere of discretion—such a relationship may form in respect of a short-lived, specific and unimportant dealing—and so an argument appealing to the existence of a significant sphere of discretion in fiduciary relationships is susceptible to the underdetermination problem I referred to above. Nonetheless, if, as I believe, many fiduciary relationships entail a significant sphere of discretion, then the argument for regarding the association of trust and fiduciary relationships as a strong one

can only be bolstered.¹⁷

¹⁷ See also Roger Cotterrell, ‘Trusting in Law: Legal and Moral Concepts of Trust’ (1993) 46 CLP 75.

3. Trust and Fiduciary Duties

Establishing that relationships that entail reliance on discretion are likely to be characterized by relatively thick trust may support the claim that there is a strong association between trust and fiduciary relationships, but to my mind it does not do enough to support the claim that the fiduciary relationship has trust at its core. While the argument of Section 2 of the article proceeds from the assumption that fiduciary relationships are invariably characterized by reliance on discretion, and asserts that reliance on discretion is likely to be accompanied by relatively thick trust, it also acknowledges that A may rely on B’s discretion without trusting B, even in a thin sense. It is thus consistent with the proposition that there is no necessary connection between trust and fiduciary relationships, and to that extent it is inconsistent with the proposition that the core of the fiduciary relationship is trust. In Section 3, I want to consider whether a necessary connection between trust and fiduciary relation- ships might be established in light of the moral foundations of fiduciary duties. First I want to argue that the justification of some moral duties refers to trust: the fact that A trusts B may figure as a premise in an argument the conclusion of which is that B owes a moral duty to A, a duty that is grounded in the requirements of respect. However, the role that trust plays in such an argument is always contingent; B never owes a moral duty to A merely because A trusts her. Consequently, to establish that moral duties referring to trust play a role in the justification of fiduciary duties is to prove only a contingent, and not a necessary, connection between trust and fiduciary duties. I want to argue that, because of the strong association between trust and fiduciary relationships, and because the fiduciary duty of ‘no conflict’ is justified by moral duties grounded in the requirements of respect, there is a strong contingent connection between trust and fiduciary duties. However, I conclude by pointing out that, however strong it is, this connection, because of its contingency, cannot do the justificatory work needed to support the proposition that the fiduciary relationship has trust at its core.

The moral requirements of respect for others can come in weaker or stronger forms.¹⁸ In the weak sense, respect demands a general awareness of and

sensitivity to the interests of others when acting in ways that might affect them.¹⁹ This sort of basic respect is one of the building blocks of social living, and its justification is likely to be found in the individual and social values that it makes an instrumental contribution to supporting and promoting. In the strong sense, respect makes sharper, more rigorous, demands, demands that take the form of moral injunctions against neglecting, exploiting or using others and whose justification is to be found in the moral worth and dignity of

¹⁸ _{For further discussion, see Matthew Harding, ‘Responding to Trust’ (2011) 24 Ratio Juris 75.}

¹⁹ Joseph Raz, Engaging Reason (OUP 1995) 275.

individual persons.²⁰ The fact that A trusts B may figure as a premise in an argument that B owes some moral duty to A grounded in respect in either the weak or the strong sense. For example, if A is trusting B to pick him up from the bus station this evening, and B suspects that this is the case but does not know for certain, weak respect for A demands that B find out one way or the other, by phoning A or in some other way. And if A and B are married, and A trusts B absolutely in all matters relating to their shared life, but B no longer loves or even likes A, then strong respect for A demands (at the very least) that B refrain from abusing A’s trust by starting the extra-marital affair that he has toyed with for the past few months. These examples are intended to illustrate the point that respect may ground moral duties to respond to trust in one way or another, where trust is present.

The justification of some moral duties grounded in respect refers to trust, but it is important to note that the justificatory connection between these duties and trust is contingent. What gives normative force to moral duties grounded in respect is not trust, it is respect, although what respect requires may be informed by the fact of trust in cases where trust is present. Indeed, the fact that A trusts B, taken on its own, is never sufficient to establish that B owes a moral duty to A; in any argument purporting to establish the existence of such a duty, a further premise referring to the value of weak or strong respect must be introduced. In particular, there is no such thing as a moral duty to justify trust.²¹ Reflection on the practice of promising should assist in bringing out this point. Imagine that A promises B that he will take B out to dinner on Friday, and B trusts A to keep this promise. By making a promise, A has voluntarily undertaken an obligation to take B out to dinner on Friday, an obligation that derives normative force from the moral principle that one must keep one’s promises.²² This moral principle is grounded in the requirements of strong respect: to make a promise, thereby manipulating the beliefs (and possibly also influencing the action) of another, and subsequently to break the promise, is, all else being equal, using that other person.²³ This argument explains why A must do precisely what B trusts him to do, but it is not an argument that A must justify B’s trust. Instead, it is an argument that A must respect B in the way demanded by their circumstances, and this happens to entail doing what B trusts A to do, given the facts of A’s promise and B’s trust. If B did not trust A to keep his promise, the requirements of strong respect would still demand that A keep it. The value of the practice of

²⁰ For two different accounts of strong respect: Elizabeth Telfer and RS Downie, Respect for Persons (Allen and

Unwin 1969) ch 1; Stephen Darwall, The Second-Person Standpoint (Harvard University Press 2006) ch 6.

²¹ Of course, as I discuss in Section 4, trust has great moral potential even though it does not ground duties and in light of this it will often be the case that a failure to justify trust is a sort of ‘moral failure’ (a phrase I

borrow from Holton (n 11) 994).

²² See generally Joseph Raz, ‘Promises and Obligations’ in PMS Hacker and Joseph Raz (eds), Law, Morality and Society: Essays in Honour of HLA Hart (Clarendon Press 1977) 210; Joseph Raz, ‘Promises in Morality and

Law’ (1982) 95 Harv L Rev 916.

²³ See Kimel (n 13) 26.

promising is to be found in the special way it enables a person to invite the trust of another: this is why, as Charles Fried has said, promising gives trust ‘its sharpest, most palpable, form’.²⁴ Given this, breaking a promise is more likely than just about any other action to amount to using a person by manipulating her trusting beliefs.²⁵ In my view, if A cannot come under a moral duty to justify B’s trust where A has invoked a social practice whose very point is to enable A voluntarily to: (i) undertake an obligation the normative force of which is explained in large measure by the wrong of manipulating trusting beliefs; and (ii) invite B to trust A to carry out that obligation, then a fortiori A cannot come under such a moral duty in any other circumstances either.

To repeat: trust figures in the justification of moral duties grounded in respect, but only contingently. Obviously, this contingent connection between trust and moral duties grounded in respect is likely to be especially strong where moral duties grounded in respect arise in interpersonal dealings that are characterized by relatively thick forms of trust: as I have just discussed, promissory relations provide outstanding examples. What about fiduciary relationships? If it can be shown that moral duties grounded in respect arise in fiduciary relationships, then, because of the strong association between fiduciary relationships and trust, it may be asserted with confidence that there is a strong contingent connection between trust and those moral duties. And to the extent that the moral duties that arise in fiduciary relationships justify fiduciary duties, then the contingent connection between trust and the moral duties that arise in fiduciary relationships may be established between trust and fiduciary duties also. The key question, then, is whether fiduciary duties are justified by underlying moral duties grounded in respect.²⁶

As is the case with the fiduciary principle, the number and character of

fiduciary duties are matters of disagreement among judges and scholars. Among those who have accepted that fiduciary duties are those that are owed only by fiduciaries and by all fiduciaries, as opposed to those that fiduciaries and others owe or that only some fiduciaries owe,²⁷ there have been at least

²⁴ _{Charles Fried, Contract as Promise: A Theory of Contractual Obligations (Harvard University Press 1981) 8.}

²⁵ See Kimel (n 13) 22–27.

²⁶ There is a tradition of seeking the justification of fiduciary duties elsewhere than in underlying moral duties:

see eg Robert Cooter and Bradley Freedman, ‘The Fiduciary Relationship: Its Economic Character and Legal Consequences’ (1991) 66 NYU L Rev 1045; Frank Easterbrook and Daniel Fischel, ‘Contract and Fiduciary Duty’ (1993) 36 JL Econ 425; Robert Sitkoff, ‘The Economic Structure of Fiduciary Law’ (2011) 91 BU L Rev

1039; and, from a different perspective, Conaglen (n 7); Edelman (n 7). As should be clear from Section 4, I have much sympathy with such an approach. However, for a variety of reasons (including that a better understanding of the relation of trust and fiduciary relationships may be gained thereby) I think that it is worth

exploring the possibility that moral duties underpin fiduciary duties; moreover I do not think that the existence of

underlying moral duties ought to preclude the state moulding and shaping fiduciary law in the service of a variety of goals (including, of course, the vindication of the underlying moral duties).

²⁷ See Bristol and West Building Society v Mothew [1998] Ch 1 (CA) 16 (Millett LJ). But contrast JD Heydon,

‘Are the Duties of Company Directors to Exercise Care and Skill Fiduciary?’ in Simone Degeling and James Edelman (eds), Equity in Commercial Law (Lawbook Co 2005) 185; and see also Richard Nolan and Matthew Conaglen, ‘Good Faith: What does it Mean for Fiduciaries, and What does it Tell us about Them?’ in Elise Bant and Matthew Harding (eds), Exploring Private Law (CUP 2010) 319.

two broad disagreements: first, between those who think that only the so-called

‘proscriptive’ duties of ‘no conflict’ and ‘no profit’ are truly fiduciary duties,²⁸ and those who think that so-called ‘prescriptive duties’, like a duty of disclosure or a duty of care, are rightly regarded as fiduciary duties;²⁹ and secondly,

between those who think that the ‘no conflict’ and ‘no profit’ duties are one and the same,³⁰ and those who think that they are independent duties.³¹ For present purposes, it is not necessary to take sides in these debates. Instead, I want to concentrate on the one duty that everyone, everywhere, agrees is a truly fiduciary duty: the duty of ‘no conflict’. This is the duty that a fiduciary owes not to be in a position where there is a certain degree of possibility that the fiduciary’s non-fiduciary duties to her principal will come into conflict with the fiduciary’s own interests within the scope of the fiduciary relationship.³²

The duty of ‘no conflict’ also demands that a fiduciary not be a position where the non-fiduciary duties that he owes to one principal may come into conflict with the non-fiduciary duties that he owes to another principal.³³ For present purposes, though, it will suffice to concentrate on situations of conflict between non-fiduciary duties and interests: my arguments about those situations can be applied, mutatis mutandis, to situations of conflict between duty and duty as well. The prophylactic character of the duty of ‘no conflict’ has been explored in detail in the scholarly literature,³⁴ and I do not propose to consider it further here. Assuming that the duty of ‘no conflict’ aims to prevent conflicts from arising, the question is: why does the law regard conflicts as problematic in the fiduciary setting? The answer to this question promises to reveal a moral duty underlying the duty of ‘no conflict’.

To say that there is a conflict in the fiduciary setting is to say something about reasons that apply to a person in a fiduciary position. In a situation of conflict, a fiduciary’s non-fiduciary duties constitute reasons to choose x while her own interests constitute reasons to choose y. Why should such a situation be of concern to the law? The most plausible answer to that question points to the possibility that, in circumstances where non-fiduciary duties favour x but a fiduciary’s own interests favour y, the fiduciary will choose y over x. To my

²⁸ _{See eg Matthew Harding, ‘Two Fiduciary Fallacies’ (2007) 2 J Equity 1.}

²⁹ See eg Heydon (n 27) especially 233–34. Contrast William M Heath, ‘The Director’s Fiduciary Duty of

Care and Skill: A Misnomer’ (2007) 25 C&SLJ 370.

³⁰ See eg Paul Finn, Fiduciary Obligations (Law Book Co 1977) ch 21.

³¹ See eg Charles Harpum, ‘Fiduciary Obligations and Fiduciary Powers – Where Are We Going?’ in Peter

Birks (ed), Privacy and Loyalty (Clarendon Press 1997) 145, 146–48; (probably also) Conaglen (n 7) 114–20.

³² The degree of possibility is a matter of conjecture: see eg Boardman v Phipps [1966] UKHL 2; [1967] 2 AC 46 (HL) 88 (Viscount Dilhorne) (‘possibility’), 112 (Lord Hodson) (‘potential’), 124 (Lord Upjohn) (‘real sensible

possibility’); Chan v Zacharia [1984] HCA 36; (1984) 154 CLR 178 (HCA) 190 (Deane J) (‘significant’ possibility); Pilmer v

Duke Group Ltd (in liq) [2001] HCA 31; (2001) 207 CLR 165 (HCA) 199 (McHugh, Gummow, Hayne and Callinan JJ) (‘substantial’ possibility); Clay v Clay [2001] HCA 9; (2001) 202 CLR 410 (HCA) (the Court) (‘sensible real or substantial’ possibility).

³³ See Conaglen (n 7) ch 6.

³⁴ See eg Weinrib (n 8) 15; Peter Birks, An Introduction to the Law of Restitution (rev edn, Clarendon Press

1989) 332–33, 338–43; Smith (n 3) 75–76; John Langbein, ‘Questioning the Trust Law Duty of Loyalty: Sole

Interest or Best Interest?’ [2005] YaleLawJl 36; (2005) 114 Yale LJ 929, 945–47; Conaglen (n 7) 67–75.

mind, there are two ways in which one might argue for a moral duty demanding that a fiduciary not choose y over x in a situation of conflict. First, one might argue that choosing x will or is likely to yield better consequences than choosing y, for example because the fiduciary will generate more welfare by choosing x or because more welfare will be generated overall by consistent adherence to a rule that requires choosing x, even if within the parameters of the case at hand choosing x will not yield more welfare than choosing y. Secondly, one might argue that a fiduciary choosing y over x will violate some non-consequentialist constraint that applies to fiduciaries in situations of conflict.³⁵ As should be clear from Section 4, I think that consequentialist arguments may play a role in explaining and justifying fiduciary law, and I do not want to rule out the possibility that a moral duty grounded in consequentialist concerns might underpin the fiduciary duty of ‘no conflict’. However, for present purposes, I want to explore the notion that a non- consequentialist constraint prevents a fiduciary who has come to appreciate the reasons weighing in favour of x and y respectively from deliberately subordinating her non-fiduciary duties to her own interests in choosing y. I want to consider the idea that such a constraint takes the form of a moral duty that is grounded in the requirements of respect in the strong sense, and that this type of moral duty—a duty not to make choices deliberately for the wrong reasons—underpins the duty of ‘no conflict’ in fiduciary law.³⁶

As I pointed out above, the requirements of respect in the strong sense

forbid using other people as means to one’s own ends. One way of using a person is exploiting an opportunity that one would not have had were it not for that other person, so as to serve one’s own ends: this is often referred to in everyday language as ‘taking advantage of ’ another person. It must be emphasized that exploiting an opportunity that one would not have had were it not for another person only amounts to using that other person where certain conditions obtain: indeed, much economic and social life consists of people selfishly exploiting opportunities made available thanks to other people. Nonetheless, selfishly exploiting an opportunity that one would not have had were it not for another person amounts to using that other person in circumstances where one was given the opportunity for the purpose of advancing the interests of the other person. So, where B vacates the corner office and A moves in, A is not taking advantage of B in any sense even if it might be said in a general sense that A is taking advantage of the situation;

³⁵ Smith’s view—that fiduciary duties reflect a moral requirement that fiduciaries be motivated at all times by what they perceive to be their principals’ best interests—points to such a non-consequentialist constraint: see Smith (n 3) 67–72. Smith’s account of fiduciary duties, pointing to a constraint grounded in the requirements of loyalty rather than respect, seeks to justify a wider range of fiduciary duties (including prescriptive duties) than the account I present here does.

³⁶ I should note that I do not consider it necessary to reject either consequentialism or non-consequentialism for the purpose of making the arguments of this article. Indeed, at the deepest level it may turn out that there is

no choice to make between the two: see Derek Parfit, On What Matters (OUP 2011).

in these circumstances A is not expected to look after any interests but her own. By contrast, where B delegates to A the task of selecting the best office for B and the second best office for A, and A then deliberately chooses the best office for herself, leaving B with the second best office, A has taken advantage of B; the whole point of A selecting offices was to choose the best one for B, not to seize the chance to take the best one for herself. In a fiduciary relationship, it is invariably the case that the fiduciary is given discretionary powers for the purpose of advancing the interests of her principal: this is the reliance on discretion that I described above, now viewed in light of its purpose. These discretionary powers create opportunities for the fiduciary, opportunities that may be used for the purpose for which they were made available, or that may be used selfishly to serve the fiduciary’s own ends. To the extent that the opportunities afforded a fiduciary qua fiduciary are used selfishly, the fiduciary takes advantage of—uses—the person (principal) without whom the opportunities would not have been available and in whose interests the opportunities are meant to be used.

Therefore, in deciding what to make of an opportunity that is available to her only qua fiduciary, a fiduciary who deliberately subordinates her non-fiduciary duties to her own interests in her practical reasoning and who chooses accordingly has taken advantage of her principal in a way that violates the requirements of strong respect. It is precisely this subordination of the fiduciary’s non-fiduciary duties to her own interests in her practical reasoning that constitutes the selfish exploitation of the opportunity given to the fiduciary. For a fiduciary to choose for the reasons constituted by her own interests because she failed to understand her non-fiduciary duties would be a moral failing and perhaps even a violation of the requirements of strong respect, but it would not amount to using the principal. Only where the fiduciary apprehends that her non-fiduciary duties weigh in favour of x and that her own interests weigh in favour of y and deliberately chooses y over x has the sort of exploitation of opportunity occurred that compels the conclusion that the principal has been taken advantage of. Moral duties not to make choices for the wrong reasons in this way—duties grounded in strong respect—enliven the non-consequentialist constraint explaining why the law is concerned about conflicts in the fiduciary setting. It underlies, and justifies, the duty of ‘no conflict’ that characterizes fiduciary law: the duty of ‘no conflict’ seeks to ensure that circumstances do not arise in which a fiduciary’s non-fiduciary duties weigh in favour of x but her own interests weigh in favour of y, lest the fiduciary resolve that conflict in breach of moral duty by deliberately choosing y. In this way, it is worth noting, the duty of ‘no conflict’ is doubly prophylactic: it aims to prevent not only conflicts from arising, but also, at a deeper level, the resolution of conflicts in ways that are morally wrong.

Given that: (i) the fiduciary duty of ‘no conflict’ is justified by underlying

moral duties grounded in the requirements of respect; (ii) trust (where it is

present) figures contingently in the justification of moral duties grounded in the requirements of respect; and (iii) there is a strong association between trust and fiduciary relationships, it may be concluded that there is a strong contingent connection between trust and fiduciary duties. However, to my mind this conclusion does not suffice to establish the further proposition that the fiduciary relationship has trust at its core. Indeed, on this score it takes us no further than the conclusion of Section 2, that fiduciary relationships are likely to be characterized by relatively thick forms of trust, and it even gives us a reason to doubt a close connection between trust and fiduciary relationships, because it reminds us of the contingency of the relation of trust and the moral duties underlying the duty of ‘no conflict’. It might be thought that a stronger connection between trust and fiduciary duties could be identified by focusing on the moral foundations of fiduciary duties other than the duty of ‘no conflict’. For several reasons, I do not think that this strategy is likely to yield much: for one thing, there is disagreement about whether duties other than the duty of ‘no conflict’ are truly fiduciary; but more importantly, establishing a stronger connection between trust and fiduciary duties other than the duty of

‘no conflict’ requires establishing the sort of non-contingent connection that

cannot be established in the case of the duty of ‘no conflict’, and yet such a non-contingent connection can never be established by pointing to moral duties underlying fiduciary duties other than the duty of ‘no conflict’ because trust never grounds moral duties non-contingently. Could a non-contingent connection between trust and fiduciary duties be established by some argument that did not refer to moral duties underlying fiduciary duties? As I want to argue in Section 4 of this article, this is the most promising strategy available for showing why, and in what sense, the fiduciary relationship has trust at its core. But that strategy takes us beyond the moral justification of particular fiduciary duties, to broad themes that bear generally on the regulation of fiduciary relationships.

4. Trust and Fiduciary Regulation

In this last part of the article, I hope to show that the assertion that trust is at the core of the fiduciary relationship begins to make more sense in light of an account of the goals that the state ought to serve in the practice of subjecting fiduciary relationships to rules of law. At the centre of my argument is the proposition that fiduciary law provides guarantees that the conduct of fiduciaries will be consistent with the requirements of trustworthiness. This proposition is not a new one,³⁷ but I want to present it within a broader theory of the purposes of fiduciary law that I think is new. According to that broader

³⁷ See eg Tamar Frankel, ‘Fiduciary Law’ (1983) 71 Cal L Rev 795; Frankel (n 16) especially ch 7; Joshua

Getzler, ‘ ‘‘As If.’’ Accountability and Counterfactual Trust’ (2011) 91 BU L Rev 973.

theory, a main purpose of fiduciary law ought to be to enable people to form, maintain and develop relationships characterized by a cycle of trust and trustworthiness, relationships that are both instrumentally and intrinsically valuable. From a perspective informed by this theoretical account of the proper purposes of fiduciary law, it may be asserted that trust is at the core of the fiduciary relationship in the sense that a large part of the point of fiduciary law ought to be to facilitate trusting relationships. However, and crucially, fiduciary law should also serve the purpose of enabling people to rely on the discretion of others on terms that do not entail trusting, or even being likely to come to have trust in, those others. A theory of the purposes of fiduciary law according to which fiduciary law ought to facilitate trusting relationships and, at the same time, facilitate non-trusting reliance on discretion draws inspiration from liberal philosophical commitments.

Notwithstanding that it does not ground moral duties, trust has great moral

potential. Elsewhere, I have described in detail the ways in which A’s trust, if manifested in action, creates opportunities for B (the person trusted) to be trustworthy, as well as the ways in which, by manifesting the attitudes entailed in being trustworthy, B in turn may provide A with further reasons to trust, thus starting a cycle of trust and trustworthiness in the relationship between A and B that tends to broaden and deepen over time.³⁸ Where this cycle of trust and trustworthiness develops in the setting of a relationship that is either intrinsically valuable or formed in pursuit of valuable ends, the trusting character of the relationship is itself valuable, both instrumentally and intrinsically. It is in this propensity of trust to create value out of nothing, as it were, that the great moral potential of trust is shown. (Of course, the creative quality of trust can be harnessed in ways that generate disvalue—think, for example, of an instance of ‘honour among thieves’—but given that most human interaction tends to create either instrumental or intrinsic value, the creative quality of trust more often delivers value than disvalue.) The intrinsic value of trusting relationships (that are otherwise valuable) may be considered a basic truth about moral life. And a large part of the instrumental value of many trusting relationships may be understood in light of the cooperative action for valuable ends that trust makes possible. But trusting relationships are also instrumentally valuable because they make available moral understandings

of human action that would not otherwise be available; in this way trust not only contributes to the conditions for cooperative action, it also generates meaning that can be drawn on in interpreting such action. Thus, in the setting of a trusting relationship, it is possible both to fail morally in a distinctive way by proving untrustworthy, and to succeed morally in a distinctive way by proving trustworthy; at the outer ends of this spectrum of moral failure and success are the vice of betrayal and the virtue of loyalty.

³⁸ See Harding, ‘Manifesting Trust’ (n 12); Harding, ‘Responding to Trust’ (n 18).

So much for the moral value of trusting relationships. I want to argue that it ought to be a main purpose of fiduciary law to enable such relationships to form, persist, broaden and deepen in ways that generate the instrumental and intrinsic value that I have just been talking about. The first task is to deal with a general objection to this line of argument: according to the objection, the law can never hope to succeed in a purpose of enabling trusting relationships because such relationships cannot develop against the backdrop of the law’s coercion. So, the objection goes, if A trusts B to x in circumstances where the law requires B to x, then by x-ing, B will not prove trustworthy to A in the way necessary to contribute to the cycle of trust and trustworthiness in the parties’ relationship; rather, B will show A only that B understands and responds to coercion.³⁹ For two reasons this objection lacks force. First, it assumes that A is likely to interpret B’s actions as being motivated in some way by the legal requirement to x: there is no reason to assume that this will always, or even often, be the case, especially where A and B have a trusting relationship that ranges beyond their dealings over x, or where A and B view the legal regime within which their dealings over x are situated more as a facilitative regime than as a coercive one.⁴⁰ Whether or not a trusting relationship is likely to be formed or furthered by B’s x-ing depends on how A interprets B’s actions, and the fact that the law requires B to x cannot be regarded as determinative of the possible interpretations that will appeal to A.⁴¹ Secondly, the objection assumes that B cannot prove trustworthy to A by showing that he (B) is motivated to x because that is what the law requires. This assumption fails to consider the possibility that doing what the law requires because the law requires it may itself be a way of proving trustworthy. And yet having an attitude of respect for the law,⁴² and showing in one’s actions that one is motivated by that attitude, is a way of proving that one is worthy of trust. Manifesting an attitude of respect for law against a backdrop of just legal regulation may be exactly the sort of response to trust that invites further trust, and a possible consequent broadening or deepening of the cycle of trust and trustworthiness, on the basis that a person who exhibits a fitting attitude to law and its demands is likely to have appropriate attitudes in other settings as well.

The failure of the general objection to arguments that propose for law a

purpose of enabling trusting relationships shows that fiduciary law does not preclude the development of trusting relationships between fiduciaries and their principals. But how might fiduciary law contribute to the development of

³⁹ See Larry Ribstein, ‘Law v. Trust’ (2001) 81 BU L Rev 553. For a critique of Ribstein’s views, see Frank

Cross, ‘Law and Trust’ (2005) 93 Georgetown LJ 1457.

⁴⁰ A and B may well view a legal regime in this way where the purpose of the regime is facilitative rather than coercive: see HLA Hart, The Concept of Law (2nd edn, Clarendon Press 1994) ch 3 for the distinction. But even if Hart is wrong, and all law is coercive in purpose, A and B may nonetheless (mistakenly) view a legal regime as

facilitative.

⁴¹ ‘The mere existence of the law does not produce some sociological imperative to use it’: Cross (n 39) 1499.

⁴² On the attitude of respect for the law, see Raz, The Authority of Law (n 15) ch 13.

such relationships? In considering this question, I want to focus first on cases in which A does not trust B but nonetheless seeks to rely on B’s discretion in pursuit of some purpose that requires B’s cooperation in this way. One of the great contributions of law and economics scholarship to understanding fiduciary relationships is in its insight that, in this type of case, the law might assist A in overcoming the hurdle presented by her lack of trust in B, thus making it more likely that A will cooperate with B in pursuit of her purpose, in one of two broad ways. First, the law might empower A to negotiate specific legally enforceable undertakings from B and, with B’s consent, implement specific systems for monitoring B, so as to gain assurance that B’s discretion will be exercised only in certain ways and not in others. Secondly, the law might establish general default rules aimed at providing those in the position of A with the same degree of assurance about those in the position of B, rules whose operation is triggered, in the case of A and B, by A’s reliance on B’s discretion. Both strategies aim to give A the assurance that she needs if she is to rely on B’s discretion in pursuit of her purpose notwithstanding that she does not trust B; the first strategy aims to provide that assurance by enabling A to extract from B specific guarantees that B’s actions will be consistent with what trustworthiness requires in the circum- stances, while the second strategy aims to assure A by guaranteeing more generally that B’s actions will be consistent with the requirements of trustworthiness, but in a way that does not require A to deal directly with B about the matter. As law and economics scholars have pointed out, what sets these two strategies apart is the fact that the second one—the one establishing general default rules—is more economically efficient than the first because on the whole it entails fewer costs to parties in the positions of A and B. This more efficient strategy for guaranteeing conduct that is consistent with the

requirements of trustworthiness is the distinctive strategy of fiduciary law.⁴³

The guarantees of fiduciary law contribute to the development of a trusting relationship between a fiduciary and his non-trusting principal by giving the principal reasons to rely on the discretion of her fiduciary notwithstanding her lack of trust. The existence of these reasons increases the likelihood, all else being equal, that the principal will rely on the discretion of the fiduciary: in this sense they incentivize the principal. If the incentive is effective and the principal relies on her fiduciary’s discretion, the fiduciary is of course required by law to act consistently with what trustworthiness requires. However, and crucially, the fiduciary also has an opportunity to actually prove trustworthy by manifesting in his action attitudes that constitute an appropriate response to

⁴³ For arguments along these lines, see generally Cooter and Freedman (n 26); Easterbrook and Fischel (n 26). See also Richard Nolan, ‘The Legal Control of Directors’ Conflicts of Interest in the United Kingdom: Non-Executive Directors Following the Higgs Report’ (2005) 6 Theo Inq L 363, 369–77, arguing that the second strategy is not only efficient but also the only practically possible one, given the reliance on discretion that characterizes fiduciary relationships.

trust given the circumstances of the case; as I argued above, the existence of the guarantees of fiduciary law does not foreclose this possibility, and indeed those guarantees make available a way of proving trustworthy—manifesting an attitude of respect for the law—that would not otherwise be available. If the fiduciary takes up this opportunity and proves trustworthy, this then gives the principal a reason to trust that she previously lacked: the demonstrated trustworthiness of the fiduciary. Although one can choose to trust no more than one can choose to have any attitude,⁴⁴ trust is responsive to reasons and especially responsive to demonstrated trustworthiness. If the principal does come to trust in light of the fiduciary’s demonstrated trustworthiness, the cycle of trust and trustworthiness that characterizes trusting relationships is underway. Without the guarantees of fiduciary law, the reliance on discretion that created the opportunity for the fiduciary to prove trustworthy might never have occurred, leaving unrealized the creative potential of trust to that extent. In Section 2, I argued that reliance on discretion is likely to be accompanied by relatively thick forms of trust because, according to a complex of social conventions and meanings, thick trust is widely understood to be a fitting attitude to have in circumstances of reliance. This point should not be

forgotten when thinking about how the cycle of trust and trustworthiness might get going in a fiduciary relationship, against the backdrop of the guarantees of fiduciary law. If part of the social meaning of reliance is given by thick trust, then we might reasonably expect that those who perform acts of reliance even in the absence of trust are disposed generally to interpreting acts in response to reliance as manifesting trustworthiness and are thus disposed to seeking a rational basis for forming trust in the future. This is a reason to think that where a principal has available two possible interpretations of a fiduciary’s actions—first, that the actions were motivated by fear of sanction; and secondly, that the actions showed respect for the law—then all else being equal the principal may favour the latter interpretation.

In Section 2, I argued that there is a strong association between fiduciary relationships and trust in the sense that fiduciary relationships are likely to be characterized by relatively thick forms of trust. It might be thought that that argument sits uneasily with my present argument that it ought to be a main purpose of fiduciary law to enable the development of trusting relationships; if most fiduciary relationships are characterized by relatively thick forms of trust anyway, then any contribution that fiduciary law might make to the develop- ment of trusting relationships would surely be a marginal or insignificant one. This view of how the two arguments sit together is misguided. For one thing, it should not be forgotten that some fiduciary relationships are not characterized by trust of any kind: in the case of these relationships fiduciary law, by

⁴⁴ Although one may not choose to trust, one may choose to act as if one trusted, with a view to providing another person with an opportunity to prove trustworthy. For a discussion, and references to discussions in other literature on trust: Harding, ‘Manifesting Trust’ (n 12) 263.

guaranteeing trustworthiness, plays a role in creating conditions under which trust might develop. But more importantly, the guarantees of fiduciary law may be necessary to incentivize even those who already have some degree of relatively thick trust to rely on the discretion of others. Take, for example, a case where A wishes to engage a solicitor B, to act for her in some matter. A has generalized optimistic beliefs about solicitors, and she has a specific optimistic belief that B is likely to be true to type and act in the way that solicitors ordinarily act. To this extent A trusts B, and A’s trust is relatively thick, being combined with generalized beliefs about solicitors, rather than beliefs about basic rationality or consistency of action. However, A lacks specific trusting beliefs about B’s moral integrity, propensity to distinguish right from wrong, and so forth. In these circumstances, it is not implausible that A will refrain from relying on the discretion of B, notwithstanding her relatively thick trust, unless A receives guarantees that B will act in accordance with moral requirements that bear on their dealings. Such guarantees can be provided via fiduciary law: the duty of ‘no conflict’, which is justified by underlying moral duties grounded in the requirements of respect, is the clearest example. In view of such guarantees, combined with her trust (such as it is), A may have sufficient incentive to rely on B’s discretion, thus making it possible for B to prove worthy of trust that is thicker than the trust A currently has—say, by showing that he (B) really understands and responds to what strong respect demands in the circumstances—which in turn gives A a reason to have that thicker trust that she (A) currently lacks.

So far, I have been talking about the guarantees of fiduciary law as

guarantees that a fiduciary will act consistently with the requirements of trustworthiness, but I have not said anything about what trustworthiness requires in the setting of fiduciary relationships. Given that trust is situation specific, taking on a thinner or thicker quality according to the variables that I introduced in Section 2, it should not be surprising that what is required in order to be worthy of trust varies from case to case as well. As I have argued elsewhere, being trustworthy is a matter of being disposed to respond, and actually responding, to trust in an appropriate or fitting manner.⁴⁵ This may entail discharging moral duties and responsibilities, or justifying trust, or having and showing a certain attitude, such as respect, empathy, friendship or love. If a main purpose of fiduciary law ought to be to enable the development of trusting relationships, then fiduciary law should aim to guarantee trust- worthiness to the degree necessary to provide effective incentives to those who seek to rely on the discretion of others, because it is through incentivizing reliance on discretion that fiduciary law is able to create opportunities for the cycle of trust and trustworthiness in human relationships. In a community with

⁴⁵ Harding, ‘Responding to Trust’ (n 18) 81–82.

a broadly libertarian outlook, guarantees against basic sorts of moral wrong- doing, say in the form of the duty of ‘no conflict’, may be all that is required to supply the necessary incentives: further guarantees, say in the form of prescriptive duties, may be regarded as excessive and intrusive. But in a community that accepts the possibility of state-sponsored norms of interaction that go beyond a libertarian core, more may be demanded: fiduciary law might supply insufficient incentives by guaranteeing only against basic sorts of moral wrongdoing and might also have to guarantee fiduciary performance in accordance with other normative demands: at the highest level, the state might have to guarantee loyalty, requiring of fiduciaries that at all times they act, in their fiduciary capacity, in the best interests of their principals.⁴⁶

This brings me to one last point. By guaranteeing that fiduciaries will act in

accordance with the requirements of trustworthiness, fiduciary law is able not only to serve the purpose of enabling the development of trusting relationships; it is also able to serve the purpose of enabling cooperative pursuits, entailing reliance on discretion, in the absence of trust. It is possible (although unlikely) that A might desire to form a relationship with B, in the pursuit of some cooperative purpose that requires A to rely on the discretion of B, and yet not trust B even in a thin sense.⁴⁷ In these circumstances, the guarantees of fiduciary law—guarantees that B will act as a trustworthy person would act— may give A the incentives necessary to rely on B’s discretion notwithstanding that she does not trust B. And the guarantees mean that B may act as a trustworthy person would act even if B does not have the attitudes and beliefs necessary to prove worthy of A’s trust. If the incentives and the guarantees are effective in this way, fiduciary law enables A and B to mimic the normative world of trust (A acts as if she trusts, B acts as if he is trustworthy) without

having to create that normative world by actually having the required attitudes and beliefs, thereby reaping the benefits of trusting relationships but on terms of detachment. In a state committed to liberal philosophical principles, I think it is important that fiduciary law serve this purpose of enabling detached relationships that mimic trusting relationships in addition to its purpose of enabling trusting relationships themselves.⁴⁸ The normative framework within which people come together to cooperate in the pursuit of value should, in a liberal order, be a matter for them (subject to necessary constraints) and the liberal state should aim to facilitate a variety of such normative frameworks,

⁴⁶ For some relevant insights from a law and economics perspective: Michael Whincop, ‘Of Fault and Default: Contractarianism as a Theory of Anglo-Australian Corporate Law’ [1997] MelbULawRw 5; (1997) 21 Melbourne U L Rev 187, 207; Conaglen (n 7) 219–21; Getzler (n 37) 988–90.

⁴⁷ In this regard, it is worth noting that the functional equivalents of trust are not exhausted by distrust. There is also the neutral stance that Edna Ullmann-Margalit described as ‘neither-trust-nor-distrust’: see Edna

Ullmann-Margalit, ‘Trust, Distrust and In Between’ in Russell Hardin (ed), Distrust (Russell Sage Foundation

2004) 60.

⁴⁸ On the value of detachment, see Kimel (n 13) ch 5.

having regard to the value of human autonomy.⁴⁹ Certainly, building a trusting relationship as a normative framework for cooperative action is a powerful expression of the autonomy of the parties to the relationship, and to the extent that fiduciary law assists in making this possible it serves liberal ends. But equally expressive of autonomy is the choice of cooperative action on terms other than those offered by trust, and indeed on terms that reject the creative potential of trust. Given the creative potential of trust, a choice on terms of detachment is always, in a sense, a matter for regret; to my mind this is a reason for fiduciary law, where appropriate, to emphasize its goal of enabling trusting relationships over its goal of enabling detached cooperation. Nonetheless, a commitment to autonomy demands that the option of detached cooperation be facilitated by fiduciary law, if only because in some circum- stances, terms of detachment might be only ones realistically available.

⁴⁹ The version of liberalism that I have in mind is set out in Joseph Raz, The Morality of Freedom (Clarendon Press 1986). See also Joseph Raz, ‘Autonomy, Toleration and the Harm Principle’ in Ruth Gavison (ed), Issues in Contemporary Legal Philosophy: The Influence of HLA Hart (Clarendon Press 1987) 313.