70 Am. U. L. Rev. F. 65 (2020).

* Professor, Brooklyn Law School. I am grateful to Evan Fox-Decent and Paul Miller for helpful suggestions on ideas in this Paper. I have also benefited greatly from prior conversations about justice with the late John Gardner. Any errors are my own.

From Loyalty to Justice

We all have good reason to act justly, but fiduciaries sometimes also have an obligation to act justly out of loyalty. Bruce Green and Rebecca Roiphe’s new paper on the prosecutor as a fiduciary is a powerful illustration of such loyalty.¹See Bruce A. Green & Rebecca Roiphe, A Fiduciary Theory of Prosecution, 69 Am. U. L. Rev. 805 (2020). It also provides a compelling account of the prosecutor’s role within the criminal justice system. On their view, prosecutors owe fiduciary duties that require them to pursue the abstract cause of justice.²See id. at 823. I wish to explore the implications of their approach for fiduciary theory by teasing out several strands of their argument.

To begin, however, I will offer a caution. We should recognize that violating a rule of loyalty does not inevitably equate to violating a rule of justice. Justice is often thought to govern how we allocate something,³See H.L.A. Hart, Prolegomenon to the Principles of Punishment, in Punishment and Responsibility: Essays in the Philosophy of Law 1, 21 (2d ed. 2008) (linking justice to an allocation); John Gardner, What Is Tort Law for? Part 1. The Place of Corrective Justice, 30 L. & Phil. 1, 6–7 (2011) (endorsing this approach). and if we adopt that view then acting loyally is not always a matter of justice. After all, loyalty is not always concerned with allocations. Loyalty may be concerned with keeping commitments, with partiality towards a loved one, with advancing the best interests of a beneficiary—and yet not be concerned with allocating anything.⁴See John Gardner, The Virtue of Justice and the Character of Law, in Law as a Leap of Faith: Essays on Law in General 253 (2012) [hereinafter Gardner, Law as a Leap of Faith].

John Gardner has offered a useful example to show how loyalty and justice may have different concerns:

If I betray a friend and not another, then the first may wonder what he did to deserve it, why he was picked out for bad treatment, etc. But the answer may be: nothing, nothing at all. When I betrayed him I wasn’t distributing the benefits and burdens of the rule ‘Don’t betray your friends’. I wasn’t distributing anything. I was just plain violating the rule, and if the friend I betrayed wants to complain about this, it is my disloyalty he should begin by complaining about, for the rule I violated is a rule of loyalty.⁵Gardner, Law as a Leap of Faith, supra note 4, at 253.

We might think Gardner is mistaken here. Isn’t there a kind of injustice in this case? Not necessarily, for as Gardner further explains:

If he [the betrayed friend] thinks there is an added insult—ie an injustice—in the fact that I didn’t betray my other friend instead, or as well, then he judges me by his rules, not mine. I am simply an ordinary moderately loyal soul aiming not to betray anyone, and occasionally failing. Whereas my aggrieved friend who complains of injustice mistakes me for some kind of allocation fanatic who spends time deciding whom he should betray, given that he is going to betray someone.⁶Id.

I think Gardner is right, and we should be careful not to blend loyalty questions and justice questions unless the two are genuinely related.

Yet, even if we start from this premise, loyalty and justice can still intersect. One context is arguably implicit in Gardner’s example above: sometimes we must allocate the benefits of our loyalty to multiple beneficiaries.⁷See id. at 253–54. If that allocation is handled unfairly by the loyal party (for instance, if the loyal party benefits one beneficiary at the expense of another without good reason), it may well be an injustice. Too much loyalty to one party and not enough loyalty to another is an allocation problem. Where loyalty’s benefits are being allocated among multiple parties, this inevitably raises justice concerns.

Nonetheless, I have a different context in mind: in some cases, the substantive content of our loyalty obligations may be concerned with justice. A loyal fiduciary may have an obligation to pursue justice because that is what it means to be loyal to her beneficiary, or because that is what it means to be loyal to a cause she has undertaken to advance.⁸Cf. Andrew S. Gold, The Internal Limits on Fiduciary Loyalty, 65 Am. J. Juris. 65, 65–66 (2020) (noting that loyalty may be concerned with the proper functioning of the legal system). Let’s call such a loyalty obligation a “justice-focused” loyalty obligation. Such obligations will be the topic of this Paper.

Justice-focused loyalty obligations take different forms. For example, in some cases a judge may have a fiduciary obligation to advance the cause of justice because the judge owes loyalty to that cause.⁹See Paul B. Miller & Andrew S. Gold, Fiduciary Governance, 57 Wm. & Mary L. Rev. 513, 570 (2015) (indicating that fiduciary loyalty may not be owed to any determinate beneficiary, and that it instead may involve advancement of an abstract goal or purpose concerning justice). The judge’s role may simply require the pursuit of justice without more, such that the obligation does not have any particular beneficiary; the duty at issue does not involve loyalty to any person or entity.¹⁰See id. at 562–63. To the contrary, it is the pursuit of justice in the abstract that counts as loyalty for this judge. In other cases, a fiduciary may owe her loyalty to a particular individual or group of individuals, and the content of this loyalty requires the fiduciary to pursue justice on behalf of the parties to whom she is loyal.¹¹See id. at 516 (describing service mandates which involve loyalty owed to determinate parties). I take Green and Roiphe to be suggesting that the second category governs prosecutors.¹²See generally Green & Roiphe, supra note 1, at 829.

What is especially interesting about cases of justice-focused loyalty is that complying with loyalty obligations in such cases can potentially alter an individual’s approach to pursuing justice. We might think that whatever our motivations, the pursuit of justice should come out the same way. Notice, however, that different norms of justice—some sound, some unsound—may be in conflict in a given fact pattern. They cannot all be pursued to the fullest extent, and pursuing one norm of justice could make it impossible to pursue another.

One way such conflicts may be resolved is through second-order norms of justice.¹³See Andrew S. Gold, Private Rights and Private Wrongs, 115 Mich. L. Rev. 1071, 1093 (2017) (discussing the role of second-order norms of justice in legal systems). There may well be norms of justice that tell us how other norms of justice should operate when they are in conflict. For example, second-order norms of justice may tell us that a norm of procedural justice should prevail over a norm of corrective justice.¹⁴See Andrew S. Gold, The Right of Redress 178 (2020) (noting that “a legal system might need to adopt second-order norms of justice that govern trade-offs between other norms of justice”). On the relation between equitable justice and the justice in redress, see id. at 199–202.

But it is not just second-order norms of justice that may operate in this fashion. Notably, it is also possible that loyalty obligations will tell us to prioritize one kind of justice over other kinds of justice. In that case, it is our concern with loyalty rather than directly with norms of justice that is doing the work, and the resulting outcome will bear the distinctive mark of whatever it is that loyalty requires. Justice may sometimes require us to be loyal in a certain way, but it is equally true that loyalty may sometimes require us to be just in a certain way.

For example, I have argued in The Right of Redress that the state as a fiduciary may owe its citizens private rights of action or an equivalent means for individuals to respond to the wrongs they have suffered.¹⁵See generally id. ch. 6. For discussion that is especially relevant here, see id. at 155–59. On this view, the state has a loyalty obligation to facilitate its citizens’ pursuit of a particular type of justice—the justice in redress—in light of the relationship between the state and its citizens.¹⁶Id. at 146–47, 156–57. This view could be justified on several grounds, including the reality that the state largely prohibits private acts of self-help.¹⁷Id. at 142–45, 157. While the state plausibly has a general responsibility to advance various types of justice (e.g., corrective, retributive, preventive, and distributive justice), it has loyalty obligations that call for it to provide a means for its citizens to obtain redress.¹⁸See generally id. at 139–59. Such loyalty obligations are not applicable to everyone, but instead the result of a distinctive state-citizen fiduciary relation with attached duties of loyalty.

A rather different concern arises if a prosecutor is acting on behalf of the public in Green and Roiphe’s sense.¹⁹See Green & Roiphe, supra note 1, at 829. One might think that a prosecutor as an agent should always be accountable to public preferences. On the other hand, one might not think this should hold true because prosecutors are not ordinary agents. Indeed, Green and Roiphe contend that prosecutors should not always follow the preferences of the public in pursuing justice.²⁰See id. at 813, 837­­­–38. Why not? There are various policy reasons, but it is in part because on their account the prosecutor is not that kind of fiduciary.²¹Id. at 843–44. Prosecutors owe a different kind of loyalty in their capacity as public fiduciaries; a loyalty that is not constrained in all cases by a beneficiary’s preferences.²²Id. at 837­­­–38. It bears noting that justice is very much at stake under any plausible interpretation of a prosecutor’s role. Notice, however, that loyalty obligations may determine the kind of justice to be pursued.

Still, we must understand the loyalty obligation at issue if it is to play a part in our legal analysis. With this in mind, a point of clarification is in order. At times, my work with Paul Miller on fiduciary governance mandates is cited in support of Green and Roiphe’s account.²³See, e.g., id. at 809 n.10 (describing fiduciary governance mandates in connection with Green & Roiphe’s account of a prosecutor’s fiduciary mandate). On fiduciary governance mandates generally, see Miller & Gold, supra note 9, at 523–24. Fiduciary governance mandates lack a determinate beneficiary.²⁴See Miller & Gold, supra note 9, at 523 (discussing the abstract nature of governance mandates). Green and Roiphe appear instead to be describing a service mandate—i.e., a fiduciary mandate that has a determinate beneficiary.²⁵See Green & Roiphe, supra note 1, at 843, 849 & n.157. While the definition of “the public” may be imprecise (and with good reason), the public is specifically noted as a beneficiary in Green and Roiphe’s work.²⁶See, e.g., id. at 809, 819. They also claim that “[i]f public officers are fiduciaries, then their actions must be made on behalf of the public.” Id. at 852. Depending on what is meant by “on behalf of,” this claim may be false. A public office holder might seek to advance the cause of justice without seeking to advance the public’s interests. If so, that public office holder could still be a fiduciary.

As noted, a fiduciary governance mandate lacks a determinate beneficiary, and this bears on the kind of loyalty a fiduciary owes.²⁷See Miller & Gold, supra note 9, at 523–24. A charitable purpose trust is a good illustration. As Miller and I note:

When a trustee for a charitable trust acts to advance the trust’s purposes, doubtless her success will redound to the benefit of the public. It does not follow that the public is a beneficiary of the trust. The trustee’s mandate is defined by the purposes established for the trust, and she owes her fidelity to those purposes rather than to any person or group of persons who hope to benefit from their fulfillment.²⁸Id. at 569.

In advancing the purposes at issue, there is no beneficiary at all. One must be loyal to certain charitable purposes rather than to anyone in particular.

This structural feature carries over (potentially) to public fiduciaries. As we further indicate: “The same is true of judges. Judicial offices are established for the benefit of the public, and their proper exercise predictably yields public benefits, but it does not follow that the public is a beneficiary of judicial offices in a formal legal sense.”²⁹Id.

One can think of public office holders as parties who must advance an abstract purpose without owing loyalty to anyone in particular—even if it turns out that members of the public will benefit as a consequence. Yet Green and Roiphe describe something different: prosecutors, on their view, “are fiduciaries who represent the public but are appointed or elected to pursue a particular abstract public interest, the interest in justice.”³⁰Green & Roiphe, supra note 1, at 809. They explicitly state that the “prosecutors’ beneficiary is the public.”³¹See id. at 813. They thus view “the public” as a beneficiary in the formal sense.

I do not generally write on criminal law, nor am I expert in the law’s understanding of prosecutorial discretion. Accordingly, I make no interpretive claim about the correct way to understand prosecutors as fiduciaries. That said, Green and Roiphe appear to be blending together two different accounts of public fiduciaries: (1) the account of judges as fiduciaries that was developed by Ethan Leib and co-authors (which emphasizes a fiduciary obligation to the public as a beneficiary),³²See Ethan J. Leib et al., A Fiduciary Theory of Judging, 101 Calif. L. Rev. 699, 720–21 (2013). and (2) the account of fiduciary governance developed by Miller and myself (which emphasizes a fiduciary obligation to advance an abstract purpose).³³See Miller & Gold, supra note 9, at 562–63.

This blending of types also has some practical implications. Because Green and Roiphe find that the public is a beneficiary, they seek to figure out whether the public’s wishes must be considered and even deferred to by prosecutors.³⁴See Green & Roiphe, supra note 1, at 840–41. It is worth noting that some private law fiduciaries must defer to a beneficiary’s wishes (e.g., agents), while others need not (e.g., trustees); deference to a beneficiary’s wishes is not inevitable in fiduciary settings.³⁵See Miller & Gold, supra note 9, at 559. Yet any puzzles concerning the public’s wishes should recede if there is no determinate beneficiary. If, for example, prosecutors were tasked with advancing justice much like the trustees of a charitable purpose trust are tasked with advancing a charitable purpose, that would be a true fiduciary governance mandate. There would be no identifiable beneficiary with wishes to consider. As far as I can tell, however, Green and Roiphe are not describing a fiduciary governance mandate.

I think this puzzle—the proper degree of deference to a beneficiary’s wishes—is one of the most difficult and important questions in public fiduciary theory. Theorists have paid much attention to the problem of identifying the beneficiary in public fiduciary settings, and a great deal of emphasis has also been placed on the objectives a public fiduciary should have. Yet, the question of paternalism versus obedience to instructions is just as fundamental. In some contexts, like trust law, fiduciary principles may call for a degree of paternalism.³⁶See Miller & Gold, supra note 9, at 559. In other cases, like pure agency law cases, the nature of the undertaking at issue may call for obedience to instructions.³⁷See id. at 559. Still, many fiduciary relationships are not so easily identified as one or the other type, and public fiduciary settings often face indeterminacy on this question. While fiduciary governance mandates may fall outside of this debate, there is much still to be said about how paternalistic—if at all—a public fiduciary may be.

Let’s bracket these conceptual concerns and proceed to some additional questions raised by the conception that Green and Roiphe adopt. They claim that there is a fiduciary duty of loyalty to the public that is to be pursued by advancing the cause of justice.³⁸See Green & Roiphe, supra note 1, at 834–35. That is a very interesting claim with a rich set of implications, and I will consider some further inquiries that this claim invites.

First, what kind of justice does this loyalty involve? One might think that prosecutors should be concerned with a kind of justice that is specific to a particular prosecution. Compare the following language from Green and Roiphe’s article: “Because, as fiduciary theory helps clarify, the prosecutor’s duty is to guard justice, not the public’s interest in general, however, it follows that considerations intrinsic to the justness of a case ought to take precedence over any other public value.”³⁹Id. at 835 (emphasis added). Yet not every approach to justice in a legal system requires doing justice between the parties in particular cases.⁴⁰Cf. Benjamin C. Zipursky, Civil Recourse Theory, in The Oxford Handbook of the New Private Law 52, 68 (Andrew S. Gold et al. eds., 2020) (linking civil recourse to the conditions of a just liberal democracy). Does Green and Roiphe’s account focus on the pursuit of justice between the parties with respect to particular cases?

Green and Roiphe also call for prosecutors to advance justice norms based on understandings that evolve across time.⁴¹See Green & Roiphe, supra note 1, at 813 (“[W]hile prosecutors’ beneficiary is the public, prosecutors serve the public not by satisfying the preferences of an amalgam of citizens at a particular moment in time but by pursuing the abstract public interest in justice that is, and ought to be, elaborated within prosecutors’ offices over time.”); see also id. at 818 (indicating that prosecutors “discern, and contribute to developing, the collective understanding of justice as they implement it in any given case”). A prosecutor concerned with this kind of justice may not just be worried about justice in any one case, for she will presumably care about a series of disputes that will be litigated over future decades. Moreover, this picture sounds like it is tethered to the understandings of justice that have been elaborated by prosecutors in the past.⁴²See id. at 819 (“[W]ithin the confines of the law, prosecutors have vast discretion. . . . Professional tradition or consensus and office policies may also work to fill in the gaps.”). There is no reason why such an approach must coincide with an overarching concern for justice in a particular case, even though it is possible that doing case-specific justice will advance and contribute to such evolving norms.

There is also another potential relationship between justice and a prosecutor’s fiduciary loyalty. A prosecutor might be loyal to the public, at least indirectly, by providing the kind of justice required by a criminal law statute—and this might qualify as loyal conduct even if the result is not in the public’s best interests. Alternatively, a prosecutor might instead be loyal to the public by providing the kind of justice that is in the public’s best interests, or that a prosecutor believes is in the public’s best interests. And there are other possibilities as well. With such possibilities in mind, is the fiduciary prosecutor someone who must pursue the public’s best interests? If so, how is that to be squared with the justice-focused loyalty that prosecutors owe? At several points, Green and Roiphe refer to the public’s interest in justice, and it is unsurprising that a loyal fiduciary who owes loyalty to the public would take this interest into account in some way.⁴³See, e.g., Green & Roiphe, supra note 1, at 809, 819. Yet, not all types of justice advance the best interests of the public—indeed, some types of perfectly legitimate justice may actually leave the public as a whole worse off on various measures (e.g., in terms of economic hardship, or in terms of social stability).⁴⁴See id. at 846, 857–58. Much depends on how “best interests” are defined, but there is a reason for the saying: fiat justitia ruat coelum.⁴⁵Loosely translated as: “[L]et justice be done, though the sky should fall.” Andrew S. Gold, Pernicious Loyalty, 62 Wm. & Mary L. Rev. (forthcoming Mar. 2021) (quoting Philip Pettit, The Robust Demands of the Good: Ethics with Attachment, Virtue, and Respect 223 (2015)). An overly zealous adherence to the fiat justitia rule by a loyal fiduciary may also cause problems. See id. The pursuit of justice could sometimes cause severe harm to the public, all things considered.

As far as I can tell, Green and Roiphe do contemplate an obligation to pursue justice in the best interests of the public, at least some of the time. As they state: “[t]he discretionary power of prosecutors at the core of their fiduciary mission derives from making these sorts of calculations in the best interest of the public rather than at its behest.”⁴⁶Green & Roiphe, supra note 1, at 812. How does a prosecutor pursue justice such that it is in the best interests of the public? Is there a certain type of justice that corresponds to this goal?

To be clear, I am not suggesting that fiduciary loyalty always requires someone to act in the best interests of a beneficiary. Some think that being a fiduciary inevitably calls for acting in the best interests of a beneficiary.⁴⁷See Lionel D. Smith, Contract, Consent, and Fiduciary Relationships, in Contract, Status, and Fiduciary Law 135 (Paul B. Miller & Andrew S. Gold eds., 2016) (“If the result of the interpretive exercise is the conclusion that the powers are held managerially, the relationship is fiduciary. This means, first, that the powers must be exercised in what the fiduciary believes are the best interests of the beneficiary.” (footnotes omitted)). I think this view is false, and not just because some fiduciary mandates lack beneficiaries. Loyalty does not always answer to a best interests standard, and fiduciary law is replete with types of loyalty that diverge considerably from what a beneficiary’s best interests require. An agent’s duty of obedience provides salient examples (if obedience is viewed in loyalty terms), as does loyalty that requires a fiduciary to “be true” to her beneficiary.⁴⁸Andrew S. Gold, Purposive Loyalty, 74 Wash. & Lee L. Rev. 881, 888 (2017) (describing a conception of loyalty as “being true”). On the gap between being true and best interests approaches, see id. On the gap between obedience-centered loyalty and best interests approaches, see Gold, supra note 8, at 69.

Nevertheless, let’s assume that prosecutors should advance the public’s best interests. Once we recognize that loyalty and best interests can diverge, the challenges ramify quickly. Could it not turn out that the best interests of the public will require less complete justice in a particular case? Or that the best interests of the public will require a kind of justice that deviates from the evolving norms of justice that Green and Roiphe endorse?⁴⁹See Green & Roiphe, supra note 1, at 820, 836. Perhaps the answer to these concerns is to understand “best interests” as a term of art, but in that case we need to know more about this term of art and its content.

I will not seek to resolve any of these puzzles here, as Green and Roiphe know better than I do which kind of loyalty is involved in prosecutorial settings. What I do hope to suggest is that the puzzles that arise are potentially complex. If these puzzles are complex, however, that is also to be expected. Fiduciary theorists are only just beginning to think through how loyalty and justice interrelate in the public fiduciary setting. Green and Roiphe’s account is a major step forward in that it suggests one, and possibly several, new ways to conceptualize a fiduciary’s justice-focused loyalties.