A. Normative Uncertainty About Justice

Let justice describe what persons owe one another morally and thus what moral rights they have against one another. If a person, P, has a moral right against D that D make a certain choice, then D infringes P’s right—commits an injustice against P—when D fails to do so. Thus, “justice” in the sense I use that word here has a broad domain. It includes perfect and imperfect duties and many duties that we do not think ought to be coercively enforceable.

To be more precise, I mean to say that a problem of “justice” arises whenever two persons’ rational standpoints offer conflicting prescriptions about a choice that is to be made, where a person’s rational standpoint consists of the self- and other-regarding reasons that apply to them, setting aside reasons of justice. Standpoints will inevitably conflict insofar as morality is compatible with partiality. Because we each pursue our own plans and projects and give special weight to the interests of those with whom we stand in special relationships, some of the reasons that define one person’s standpoint will not be shared by all others.Footnote ¹⁰ Given that standpoints conflict, we need a further set of normative considerations to tell us whose standpoint governs each choice—considerations of justice. By assigning standpoints to choices, the balance of these considerations gives rise to a system of rights at the moral level: a person has a moral right to have a choice made in accordance with her standpoint when justice assigns her standpoint to that choice.

Thus, while I agree with John Gardner’s high-level observation that justice addresses allocative questions, the objects to be allocated in my framework are not, in the first instance, morally significant assignable goods as they are for GardnerFootnote ¹¹ (though my framework will have downstream implications for the allocation of such goods), but rather fundamental normative entities—persons’ standpoints. Thus, in my framework, unlike Gardner’s (and Smith’sFootnote ¹²) justice constitutes primary rights, and not merely what persons may owe one another should they infringe a primary right.Footnote ¹³ My picture entails, contra Gardner, that the moral right not to be tortured should be understood as a right of justice.Footnote ¹⁴

In this respect, my conception of justice aligns with the insistence of the corrective justice theorists of the Toronto School that corrective justice determines the content of primary rights alongside the secondary rights that arise from their infringement. On the Toronto School Kantian picture, each of us must be free to set our own purposes independently of the will of particular others, which in turn requires that our rights be omnilaterally defined and enforced to ensure that we are not at the mercy of the will and judgment of particular others.Footnote ¹⁵ The contours of our private rights, moreover, cannot require anyone to serve another’s needs and purposes because that would violate the requirement that “no person is in charge of another.”Footnote ¹⁶ Thus, private rights must not do more than instantiate the procedural freedom of each to do what they wish with their means compatible with a like freedom of all.

While I agree with the Toronto School that justice defines persons’ primary rights, I reject its austere picture of its content. Justice in my framework is something that exists apart from the efforts of an appropriately constituted state to define and enforce it. And justice may reflect substantive considerations arising from persons’ needs and purposes.

There is, however, a further parallel between my framework and that of the Toronto School alongside an important difference. Normative uncertainty plays an essential role in the corrective justice framework, but, contrary to my picture, its focus is on normative uncertainty of an exclusively conceptual or metaphysical kind. On this view, there are no determinate answers to questions about the content of persons’ private rights and duties, and the law provides an essential role in constituting those rights and duties by resolving this indeterminacy. Thus, if the principles of justice instructed us determinately how to relate to one another, the role of the state in constituting our private rights would be significantly diminished. We might still need a centralized enforcement apparatus, but the content of private rights could be determined without state action because no exercise of will or judgment would be required to figure out their content. But if there is indeterminacy about the content of our rights, there is no truth of the matter prior to the resolution of the indeterminacy. Thus, making such choices involves a significant moment of collective self-definition, and such decisions must be made omnilaterally to avoid the risk that others unilaterally define what justice means for us.Footnote ¹⁷

But uncertainty about justice seems to be at least in part epistemic rather than metaphysical, once we conceptualize justice as the set of substantive criteria that mediate among the competing partial perspectives that are embodied in agents’ rational standpoints. Such partial perspectives inevitably reflect persons’ particular needs, circumstances, plans, and projects such that mediating among them will mean being attentive to a multitude of morally relevant variables arising from those considerations. There will likely be more or less just ways of mediating among these variables, even if justice does not pick out a single allocation in every case. At the same time, figuring out how to evaluate the possibilities is an epistemically demanding problem that is likely to engender disagreement even among reasonable people.

The existence of significant epistemic uncertainty about justice is, of course, compatible with there being some clear answers about what justice requires. It seems clear, for example, that D may not without provocation punch P in the face. In terms of my framework, P’s standpoint clearly controls D’s choice whether to punch P in the face clearly dictating that D should not punch P.

But much of what people owe one another is normatively uncertain. How much care does D have to take to reduce the chance that D accidentally injures P? To what extent and how does it depend on D’s or P’s circumstances or characteristics? Must D use resources at D’s disposal to assist P or avoid setting back P’s interests? Does it matter to what extent resources have been justly distributed? To what extent must D compensate P for harm D causes P even when D appeared to take appropriate precautions to prevent injury to P? Although we often discuss such questions as if they have right answers, the answers to such questions are far from obvious. And the answers will likely change considerably depending on the circumstances of the parties and the details of their relationship.

But once we allow that there are procedure-independent truths about justice about which we are epistemically uncertain, the elegance of the Toronto School’s picture is disrupted because there may be a cost in terms of justice associated with selecting one way of resolving it over another. When the uncertainty is entirely metaphysical, there can be no such cost because all that matters is that the selection be made through the appropriate (omnilateral) procedure. When there is epistemic uncertainty about justice, by contrast, we must reckon with the possibility that the best possible procedure will select a substantively unjust outcome. It is thus no longer obvious that there is a purely procedural solution to the problem, and the fact that the law may purport to give us solutions does not, without more, entail that we must defer to those answers.

Epistemic normative uncertainty about the content of persons’ rights presents a moral problem for the parties, even when the law has done the best it can to define what justice between them requires. This is because it is unlikely that the law can successfully converge on the true set of rights. And even if the law is more likely to get things right than the parties in the spirit of Joseph Raz’s normal justification thesis,Footnote ¹⁸ it does not follow that the parties ought to defer to the law’s judgment in every case. Sometimes they will come to a different plausible understanding about how to resolve uncertainty about justice between them that is not obviously inferior to the court’s resolution. The parties do not have reason to regard the law as offering the final word in such cases just because in a range of other cases deference to the law is the epistemically superior course.

Perhaps the answer is more procedure of the right sort. Someone might think, for example, that the democratic credentials of the legal system give the parties intrinsic reasons to defer to a publicly determined judgment not obviously superior to the parties’ own. The authority of representative institutions is sometimes justified along such lines. Representative institutions arguably instantiate principles of equal respect for the judgments of all on matters of public importance or help to constitute egalitarian relations between persons.Footnote ¹⁹

I believe that this suggestion is along the right lines with two important caveats. First, such principles grant representative institutions authority that must be substantively bounded: if those institutions make plainly unjust decisions that fail to instantiate principles of equal respect, it is hard to see why that same principle would require deference to those decisions. Second, while liberal egalitarian principles may give us compelling reasons to regard democratically made decisions on matters of collective importance as authoritative within substantive bounds, similar intrinsic considerations of equal respect for the agency of each might be marshaled in favor of a principle of deference to the parties’ joint judgment when the normatively uncertain question of justice implicates the parties’ relationship alone. If the parties have arrived together at a view about what justice between them plausibly requires, and the matter does not infringe on the rights of third parties, respecting the parties’ joint judgment honors the joint agency of the parties and deliberative ideals better than imposing the community’s judgment on them. As I argue in greater detail elsewhere, considerations such as these suggest that the moral authority to resolve normative uncertainty about questions of (first-order) justice that pertain only to the relationship between parties is allocated to the parties as a matter of (second-order) justice. Because it instantiates a liberal democratic principle of equal respect for agency, I name my conception the democratic conception. Footnote ²⁰

The resulting picture of morally valid agreements occupies middle ground between accounts of promises and agreements according to which they are valid insofar as they stand in the right relationship to more fundamental moral considerations—in my framework, the considerations of first-order justice—and accounts that view valid agreements as autonomous generators of obligations and thus perhaps independent determinants of duties of justice. The former camp includes accounts like those of Rawls and Hume who offer conventionalist accounts of promising as arising from a valuable social practice.Footnote ²¹ The latter camp includes libertarian accounts, including Kantian accounts, according to which agreements are morally valid insofar as they express the parties’ equal procedural freedom.Footnote ²² My account resembles the former accounts in requiring that agreements must stand in the right relationship to underlying first-order considerations of justice in order to be morally valid. But it resembles the latter in that it entails that agreements can have freestanding significance when they are arrived at through the right procedures and do not run contrary to clear requirements of justice.

One might object that an appeal to democratic values entails only that institutions of the collective that are designed to accord equal respect to the views of each ought to make the relevant determinations. I contend, on the contrary, that an allocation of authority to the parties is required when the matters of justice are local to the parties’ relationship. This is easy to see when there is metaphysical indeterminacy about the correct choice: in such cases, allocating authority to the affected parties allows the parties to engage in a joint act of collective self-determination—defining what it is that justice in their relationship requires. Involving others in that constitutive act gets in the way of that expression of collective autonomy by those whose rights will be constituted by it. It might seem that the same is not true when the source of uncertainty is epistemic rather than metaphysical given that justice ranks the options, raising the possibility that the resolution the parties might choose will be less just than that of the collective. But my claim is not that the parties may make plainly unjust decisions so long as they do so together. Rather, my contention is that we should respect choices of the parties that are aimed at realizing justice when they cannot easily see their way to the truth. Given that the parties’ epistemic predicament means that they cannot see through the normative uncertainty to the truth, the best they can do is deliberate together in good faith about what justice requires—in effect, defining what justice means to them given their epistemic predicament. In this sense, the parties exercise joint agency in making their choice much as they do when they resolve metaphysical indeterminacy. The important difference between an epistemically uncertain choice and a metaphysically indeterminate one is that the former, unlike the latter, must be regarded as provisional—subject to revision as the parties’ epistemic situation improves. Thus, the parties may find themselves under a duty to reconsider their epistemically uncertain choice.

What about the rule of law? Doesn’t a moral delegation to the parties to resolve normative uncertainty about justice between them threaten values we associate with that ideal? It entails that the norms that regulate relationships depend on private agreements and so will inevitably vary in accordance with the idiosyncratic judgments of the parties. Thus, the norms one set of parties articulates for their relationship may not be amenable to principled reconciliation with the norms other parties articulate to govern theirs.

My suggestion here, however, is that the parties’ determinations are morally authoritative only when they resolve matters of justice that do not implicate the rights of nonparties in a substantively plausible way, and reflect their good faith, joint deliberations about what justice between them requires. In this sense, parties’ morally authoritative determinations are public: they embody collective judgments made through fair deliberations by all whose rights they implicate. The requirement that they be arrived at jointly in good faith helps to ensure they are oriented toward justice in the right way by ruling out resolutions that are arrived at through self-interested bargaining or the imposition of one party’s views on the other or influenced by false consciousness and the like. The parties’ resulting determinations are also only morally valid when they plausibly settle moral uncertainty about what justice between the parties requires. Agreements to implement plainly unjust results are substantively invalid even when they reflect the parties’ considered joint views about justice.

There therefore remains an essential role for public oversight. The law must set out general standards to regulate the parties’ relationship when agreements about justice that meet these demanding substantive and procedural constraints are not reached. The law also has an important role to play in facilitating good faith deliberations between the parties about what justice between them requires. And the law must police the agreement process to ensure that only morally valid agreements displace publicly determined norms. But contrary to corrective justice accounts, the role of legal institutions is often restricted to promoting the right kind of deliberations between the parties rather than having the final say about what justice between them requires. In cases of clear-cut injustice, the community ought to respond to the injustice regardless of what the parties themselves decide to do about it, and the community ought to override agreements that plainly run contrary to what justice requires.Footnote ²³ But when things are less clear, courts properly play a facilitative rather than determinative role, so long as the parties are willing to discharge their responsibilities to deliberate carefully about what justice between them requires.

B. Public Default Rules

On the picture of private law just described, we should not think of remedial orders as having the final word about what the parties owe one another, except when the dictates of justice are clear. So long as there is significant normative uncertainty about what justice between the parties requires, remedial orders do something more modest. They specify legal consequences that are apt only because the parties have failed to articulate their own joint vision of justice for their relationship. Courts should accordingly view their role as setting default rules that will facilitate good-faith deliberations about normatively uncertain questions. In this way, we can make sense of the ways in which remedial law looks more like a public backstop rather than a direct determinant of the parties’ rights and duties.Footnote ²⁴

Some idealizing assumptions lie behind this conception of the law’s role. In a world in which parties simply cannot be trusted to set aside their self-interest and deliberate in good faith, there is no point in attempting to facilitate deliberations about justice between the parties. Private agreements might relieve burdens on the system by saving public and private dispute resolution costs. But they will reflect only the parties’ predictions about the possible content of anticipated remedial orders and their estimates of the costs of continuing to litigate, mediated by strategic considerations that affect the bargaining power of each relative to the other. They will not reflect plausible good faith joint determinations of what the parties owe one another as a matter of justice.

When enough parties are willing to exercise their moral authority in good faith, however, remedial law can construct default rules that will help the parties draw their own conclusions about what justice between them requires. This does not mean that remedial law ought to be insensitive to considerations of justice between the parties. Deliberating about what justice requires may be a burdensome task. Courts can ease these deliberate burdens by developing default rules and principles that reflect their expertise. So long as parties and courts both understand that courts are there primarily to play a deliberation-enhancing role, the anticipated content of the court’s remedial orders will then influence the parties’ deliberations not by serving as a threat point when negotiations fail; after all, when the parties take their deliberative obligations seriously, they will not be focused on maximizing their own advantage. That anticipated content will instead influence the parties by reflecting normative considerations of which the parties ought to take heed in their deliberations with one another.

How exactly should courts construct these defaults? Because courts see many similar cases, they will tend to have epistemic advantages over the parties about matters of justice that apply generally. But courts lack such expertise when it comes to evaluating the moral considerations that bear on the parties’ specific relationship given their needs and circumstances. By making their remedial orders reflect moral considerations that apply generally, courts can ease the deliberative burdens on the parties by providing parties with a plausible starting point for their joint deliberations about what justice between them truly requires.

Because, however, the moral authority to resolve normative uncertainty about what justice between the parties requires ultimately lies with the parties not the court, courts must take care not to articulate principles that take them beyond the domain of the general. Although disputing parties could always reflect a different view in their agreement, the court’s view is likely to exert significant influence on the deliberations of parties who are deliberating in good faith. This is in part because the parties have reason to adopt a deferential stance toward the principles courts apply insofar as courts are operating within their domain of expertise, and it might not always be obvious to the parties that the court has strayed beyond that domain. It is also because the playing field between alternative views of justice is no longer level when one view is reflected in a court order because that view will prevail in the event of disagreement. Even parties who are transparently committed to articulating their own vision might find it psychologically difficult to ignore a baseline that advantages their own view. Even if they can resist that temptation, they may pay heed to the baseline defensively whenever they are not confident in their deliberating partner’s willingness or ability to refrain from exploiting the baseline to advance their own preferred vision of justice.

Even when courts are operating within their domain of expertise, moreover, they ought to articulate the applicable normative considerations in a way that is as neutral as possible among different conceptions of justice that the parties might apply to their relationship. Adopting a controversial conception of justice risks biasing how the parties deliberate about the relationship-specific matters that lie within the parties’ expertise if parties strive to achieve consistency between their resolution and the principles of remedial law. All else equal, then, midlevel principles that can rendered compatible with multiple theories of justice are preferable to principles that depend on controversial moral foundations.Footnote ²⁵

For remedial law to play a deliberation-enhancing role, moreover, parties must be able to anticipate its content and reflect it in their own deliberations insofar as they find it applicable to their relationship and circumstances. While no duty to do as the court declares arises until the court declares it, the considerations that underlie that declaration ought to shape the parties’ deliberations before a court is called on to decide the case, thus influencing the content of the parties’ agreement. The order steps in to define the parties’ duties, however, in the event the parties’ deliberations fail. Thus, while remedial law should try to abstain from controversy, it must do so in a way that is intelligible to disputing parties. It is thus important that the reasons that underpin the content of the remedial order can be reconciled with the reasons that underpin the primary legal right, even though strict continuity between right and remedy is not required.Footnote ²⁶

Indeed, on my conception, even primary legal rights are at best approximations of what parties owe one another morally speaking. Thus, private liability often fails to track genuine wrongdoing.Footnote ²⁷ Consider, for example, the way some primary rights in tort law are defined with reference to a hypothetical reasonable person of average intelligence and capabilities.Footnote ²⁸ It is unlikely that ensuing legal duties of care describe what the parties owe others by way of care at the moral level given that they are much harder for some to discharge than others.Footnote ²⁹ The law at the level of both right and remedy is there to facilitate the management of moral uncertainty that is the primary responsibility of the parties to figure out except in clear-cut cases where the dictates of justice are clear.

In sum, it is desirable on the democratic conception that legal doctrine have its own internal structure that has a freestanding logic relative to the considerations of justice that ultimately ground and justify it. This is because responsibility for reflecting those considerations of justice is not concentrated in the courts as representatives of the community but is allocated to the parties whenever there is normative uncertainty about their content and implications. Legal doctrine ought not to reflect all-things-considered judgments about what justice requires. Rather, it should be structured to ensure that the right kind of deliberations about justice occur.

Thus, any resemblance between my framework and a Coasean picture according to which private ordering is the mechanism via which we achieve just (or, in the economic framework, social welfare maximizing) outcomes so long as transaction costs are low enough is superficial rather than deep.Footnote ³⁰ Nor does my framework resemble an administrative model of private law whereby private parties exercise authority to answer questions of justice that have been delegated to them by the state because, say, they are better situated to evaluate what justice requires (in the way economists suppose that parties are better suited than government actors to identify mutually beneficial exchanges). Rather, my central contention is that the parties are morally authorized to settle relationship-specific matters of justice between them because such an allocation of authority is a morally justified response to the fact of moral uncertainty about justice because it instantiates democratic egalitarian values of equal respect for agency. To validly exercise this authority the parties must engage in good faith joint deliberations with one another about what justice between them requires, and the products of their deliberations must be substantively plausible resolutions of the applicable normative uncertainty. Hence, the need for public oversight and other institutional mechanisms to ensure that the right kind of deliberations between the parties occur.

A. General Damages

Given that courts’ comparative epistemic advantages pertain to normative considerations that generally govern rather than considerations that arise from the needs and circumstances of the parties, remedial orders ought to be fashioned in accordance with the former not the latter. Call this the principle of generality. It applies both to the ways primary legal rights get defined in the first place and how courts fashion remedial orders in response to breaches of or threats to primary rights. With respect to the former, the principle aligns with the ways in which primary rights in tort law are often defined with reference to a hypothetical reasonable person of average intelligence and capabilities.Footnote ³¹ At the remedial stage, it helps to justify the pervasive use of market prices to determine general damages when such measures are available along with some guiding principles when they are not.

Consider general damages for trespass or conversion. Damages may be given by the market rent for the property trespassed upon or converted, even when the trespass was harmless.Footnote ³² Such damages reflect the idea that pervades the substantive law of property that property rights confer on owners the authority to decide who may make use of the property.Footnote ³³ Trespassers and convertors interfere with that authority even if they cause no setback to the right holder’s interests because they did not interfere with anything the right holder would have otherwise done with the property.Footnote ³⁴ The remedy accordingly tracks in a transparent way the law’s conception of the nature of the rights infringement by reflecting the value of the set of opportunities conferred upon the right holder, even when it does not reflect the value of the opportunity the right holder likely would have pursued absent the infringement.

The market rent is a reasonable way of measuring this value in a way that also accords with the principle of generality. If a court were seeking a party-specific measure, it would have to imagine what the deliberations of the parties would have looked like had the trespasser or converter sought permission in advance of using or occupying the right holder’s property, constructing a remedy reflecting the price the parties would have decided upon as part of any morally valid agreement they would have reached. But any such agreement between the parties, were it morally valid, would reflect the parties’ settlement of morally uncertain questions of relationship-specific justice, questions that lie squarely within the parties’ domain of moral authority. The market rent, by contrast, is a measure of the price a typical owner and renter reasonably would have set. In this sense, it embodies a normative judgment about what a plausibly just transaction between typical parties would look like—the output of an idealized negotiation process, but one that importantly abstracts from the details of the parties’ relationship about which the court has neither the moral authority nor special expertise to determine.Footnote ³⁵ It thus sets a fair default for the parties’ deliberations about what justice in their relationship specifically requires.

Of course, the market price is not always a good approximation of the just price between typical parties, even assuming that the market is reasonably competitive. It is less likely to serve as a just default when, for example, the market has been distorted by background injustice, as when participants on one side of the market are predominantly beneficiaries of background injustice and/or participants on the other side of the transaction are predominantly victims.

Yet, it is not obvious that courts ought to reflect such systemic considerations in their remedial orders. Courts are institutionally poorly situated to evaluate the import of such matters for the relationship between the parties. Should courts attempt to do so, they risk imposing a controversial vision of justice on the parties about a matter on which they have no special institutional expertise, distorting the parties’ deliberative processes and making it less likely that any agreement between the parties reflects a genuine joint vision of what justice between the parties requires. Legislatures may be institutionally better positioned than courts to evaluate the import of such systemic concerns and thus come up with appropriately neutral default rules. Even so, insofar as the remedy is supposed to reflect the way in which those systemic considerations bear on the parties’ relationship, legislative intervention still carries the risk that it will prevent the parties from articulating their joint vision of what justice between them requires.

Even in markets that have not been distorted by systemic injustice, the market rate seems like the wrong default in cases where the trespass or conversion is motivated by profit. Profiteering conduct does not merely deprive the victim of the opportunity to decide how their property is to be used. It shows an essential disregard for the principles that ought to guide private law on the democratic conception. A person who wants to profit by using another’s property ought to engage in good faith deliberations with the latter about whether justice between the parties would allow for such use and on what terms. Instead of initiating such deliberations, the profiteer unilaterally decides to do what he wants with the property. The correct default, therefore, ought to force the profiteer into negotiations with the victim about the benefit the profiteer derived from its unilateral act by making the profiteer liable for the full benefit he derived from the tort.Footnote ³⁶ A market price default might more closely approximate the agreement that the parties would have reached had they deliberated about what justice between them required in advance. But they might have negotiated a price that would have given the victim a greater share of the profits to be had from the use of the property. Given the profiteer’s demonstrated disregard for the deliberative processes the democratic conception calls for, it makes sense for the court to set a default that emphasizes the latter possibility rather than the former, thereby empowering the victim to negotiate about what justice demands with the profiteer up to that amount. While the victim may end up with more than they would have received had proper deliberative procedures been followed, such a default compensates the victim for the profiteer’s disregard of the victim’s procedural right to fair deliberations about what justice between them required rather than conferring a windfall on the victim.Footnote ³⁷

When it comes to general damages for breach of contract, courts can sometimes rely on the contract itself to provide the measure of lost value rather than looking to market prices. This is because the contract itself serves as a natural, neutral starting point for the parties’ deliberations, assuming that the contract itself is morally valid and so reflects the vision of justice that the parties constructed when the contract was formed. By reflecting the parties’ agreement in their remedial orders, therefore, courts set neutral defaults for the parties to deliberate around. Enforcing payment of the contract price where the contract has been fully performed, for example, is neutral between the parties precisely because they both agreed to it.Footnote ³⁸ In this sense, it is a less intrusive default than a market measure of the value of a performance owed by the breaching party. Of course, unless the parties included a remedial clause in their contract, the parties’ agreement sets out performance terms not the remedy for breach. Thus, except where performance involves the payment of money, market prices will inform the measure of the difference in value between the performance provided and the agreed-upon performance. Market measures are used, for example, when the breach consists of delayed performance or defectively rendered performance.Footnote ³⁹

Even when the contract includes a specification of damages for breach, it does not follow that the specification ought to inform the court’s remedial order. Courts should only honor such specifications when they reflect the parties’ considered joint judgment about what justice between them requires in the event of a breach. Remedial clauses fashioned with other objectives in mind—deterrence, for example—are not morally valid on my conception and so ought not to receive judicial deference, as I have argued elsewhere.Footnote ⁴⁰

What about intentional, profitable breaches of contract? Should the breaching party be required to disgorge its gains from breach in the same way that the profiteering trespasser ought to disgorge gains reaped from the trespass? At first glance, a similar logic seems to apply. The breaching party ought to have renegotiated the terms with its contracting partner to get permission to “breach.” But there is a way in which breaching a contract is less of an imposition than a trespass. Whereas the trespasser infringes with the owner’s decision with whom to deal and on what terms, the contract breacher is only imposing terms on their partner. Where, moreover, the breach responds to circumstances that were not anticipated by the parties when they entered the contract, it becomes less likely that their original agreement captures what justice between the parties requires, in which case the parties ought to renegotiate their agreement in response. There is thus often an onus on both parties, breaching party and victim alike, to renegotiate. When the infringement is a trespass, by contrast, the parties generally have no prior agreement covering the conduct at issue and, indeed, there is a good chance that the victim would have declined to deal with the trespasser had the trespasser asked. It is therefore less clear that a disgorgement remedy ought to be available for a profiteering breach of contract as it is for a profiteering trespass. A disgorgement remedy seems inapt, for example, where the breach victim was unwilling to make a good faith effort to renegotiate terms in response to changed circumstances or where renegotiation in advance of the breach was, for practical reasons, impossible. Under such conditions, it is fairer to make a measure of the value of the lost performance the baseline for potential settlement negotiations between the parties. A disgorgement remedy for breach of contract remains appropriate, however, when a profitable, intentional breach also appears designed to subvert any possibility of renegotiation or otherwise take advantage of the breach victim.Footnote ⁴¹

When the legal right that has been infringed is a right not to have one’s person or property carelessly harmed, courts should seek an objective measure of the harm. When the harm is to tangible property, market measures often exist and can serve as an appropriately objective measure of damages.Footnote ⁴² But market measures simply do not exist when the harm is physical injury to someone’s person. How then should a fair measure of harm for pain and suffering and loss of enjoyment be determined? On the one hand, a fixed schedule would not be able to capture relevant differences across cases. On the other hand, having juries decide creates tremendous unpredictability for the parties. Juries decide damages for a single case and so lack the perspective gained from deciding similar cases. Given that attempting to put a monetary measure on the suffering of a badly injured victim is a difficult task, the judgment of one jury may differ sharply from the judgment of another. That type of uncertainty about the remedy that a court might order seems likely to create fear that could impede good faith deliberations between injurer and victim about what justice between them requires. If so, judicial oversight of jury awards with an eye toward creating consistency across cases is desirable.Footnote ⁴³

B. Special Damages

We have just been discussing how courts ought to determine the right remedial response to the direct rights infringement that the victim has suffered. But rights infringements may also make the victim worse off than the victim would have been but for the wrong because of its interaction with the victim’s particular circumstances and plans. To what extent ought the victim be made whole for these counterfactual losses? Scholars and courts often emphasize the centrality of the make-whole or rightful-position principle: remedial rules ought to strive to restore claimants to the position they would have been in but for the defendant’s wrong.Footnote ⁴⁴ Here I will argue that whatever its value as a way of making sense of the doctrine, a counterfactual make-whole principle ought not to have the status of a fundamental principle of remedial law.

The primary problem with indexing remedies to a prediction about where the claimant would have been but for the defendant’s legal wrong—even a very plausible prediction that would have been within the contemplation of both parties in advance of any transaction between them—is that it assumes the normative validity of that counterfactual. It is easy to come up with scenarios in which that assumption is highly questionable. Suppose I breach a contract to deliver new computers to a business that engages in predatory lending practices with the result that the business is unable to initiate new loans for a month reducing the business’s profits. There is something questionable about the notion that I ought to compensate the business for profits lost because it was unable to make predatory loans for a month even if this was an entirely foreseeable consequence of my breach at the time of contracting and it is straightforward to estimate what the lost profits are.

A defender of the make-whole principle might appeal to an institutional division of labor to deflect an objection along these lines. It is the business of other areas of law to police predatory conduct. The law of remedies is entitled to help itself to the assumption that the rest of the legal system is doing its job and focus on the remedial task at hand.

But the problem goes beyond counterfactual predictions about likely earnings from plainly unjust agreements such as predatory loans to counterfactual predictions about earnings from contracts that are not obviously unjust. The problem is that contracting parties, on the democratic conception of morally valid agreements, ought to be articulating a joint vision of what justice between them demands. Normative uncertainty makes it uncertain from the perspective of an outsider (and the parties themselves before they arrive at their vision) what their vision will be—assuming that the parties are taking seriously their duties to articulate a vision of justice. Indeed, the fact that the content is uncertain to an outsider is an entailment of justice because second-order justice allocates moral authority to the parties to settle the extant moral uncertainty about first-order justice.

None of this is to say that lost economic opportunities are irrelevant from a remedial standpoint. But it is not clear that the just response is to try to match an empirical prediction about what would have happened absent the wrong. What remedial law ought to be doing, on my conception, is providing the parties with a fair baseline for their deliberations about what a just response to such lost opportunities looks like.

What does that baseline look like? Determining damages for the loss of economic opportunities arising from agreements that would have arisen in the future involves speculation that is different in kind from that involved in determining damages for lost opportunities arising from agreements that existed at the time of a wrong. This is because there is normative uncertainty about the shape of those future opportunities that ought to be settled by those involved. A positive prediction, even a very accurate one, about the shape those opportunities would have taken may not supply the correct normative baseline. By contrast, determining the losses that flow from existing agreements is primarily an empirical matter.

We can draw on this distinction to give a more coherent shape to remedial law’s certainty principle—the requirement that losses be proven with reasonable certainty in order to be recoverable—than is offered by current law.Footnote ⁴⁵ The origins of the certainty principle lie in attempts during the eighteenth century to control the jury’s assessment of damages, suggesting that the rule ultimately has a procedural justification.Footnote ⁴⁶ But contemporary courts do not apply the principle consistently sometimes invoking destabilizing counter principles, such as the principle that doubts ought to be resolved against the wrongdoer whenever the wrongdoer’s conduct created the uncertainty about damages—a principle that threatens to swallow the certainty principle in its entirety given that it is always the case that the wrong created uncertainty about what would have otherwise occurred.Footnote ⁴⁷ Melvin Eisenberg concludes that the principle is unstable and thus unsound.Footnote ⁴⁸

On my democratic conception, we can reconceptualize the certainty principle as tracking instead a coherent substantive concern with ordering a defendant to compensate a claimant for losses the estimation of which will involve speculation about how normative uncertainty would have been resolved—in particular, losses arising from a claimant’s lost opportunities to deliberate with others about what justice would have required. Empirical predictions about what would have happened—about how those deliberations would have in fact turned out—do not capture the essence of what has been lost—namely, opportunities to realize gains following deliberations about what justice requires. And it is not obvious what the just response to losses of the latter kind ought to be, except that the just response will be highly context specific given that the losses depend on how the injured party would have exercised their agency to deliberate with others about what justice requires. Applying the principle of generality, then, remedial law should direct defendants to compensate only for losses connected to existing agreements, thus leaving it to the parties to determine the just responses to losses connected to future agreements given their circumstances and relationship. Where profits from existing contracts are uncertain, these losses might be measured (at least in part) by the costs the promisee incurred or expected to incur in order to make those existing contracts profitable.Footnote ⁴⁹ Whether other more speculative losses associated with existing contracts ought to be compensated, though, would be a separate question that depends on the burden of proof courts think apt for civil suits generally.

The same distinction between current and future economic opportunities also suggests a different way of understanding the economic loss rule. The rule often takes on an absolute form: absent a special relationship, there is no legal duty to take care to protect others against pure economic losses—losses deriving from lost economic opportunities that are not associated with a negligently caused injury to the claimant’s person or property.Footnote ⁵⁰ The corrective justice theorists regard the rule as an implication of the more general absence of liability for nonfeasance—in turn grounded in the principle of independence—which entails a general absence of private legal duties to serve the needs and purposes of others.Footnote ⁵¹ At the same time, they endorse recovery for economic damages that are indirectly caused by wrongful injury of the claimant’s property or their person or infringement of their contractual rights because such damages reflect “how you would have used the means to which you have a right.”Footnote ⁵²

My account, by contrast, supplies an argument for the symmetric treatment of pure economic losses and consequential economic damages resulting from breach of contract or wrongful damage of the claimant’s property. There should not be recovery for negligently caused losses of future economic opportunities. But there should be recovery for negligently caused economic losses arising from disruption to existing agreements.

When it comes to economic losses that result from serious personal injury a counterfactual make-whole principle does not supply the right default either. But neither does a default that would simply decline to force the defendant to compensate the claimant for such losses. The reason is that a debilitating physical injury does not merely deprive the injured person of particular future opportunities; it diminishes the victim’s capacity to decide how they will participate in social and economic life more generally. The ensuing losses are thus akin to direct damages for damage to property in depriving the victim of the capacity to deliberate about how they will use their person to participate in the economy and society. But unlike property damage cases, objective measures of such losses are not readily available because they cannot be quantified in market-based terms. The counterfactual make-whole principle suggests that we ought to make an empirical prediction about what the claimant would have earned but-for the defendant’s wrong and order the defendant to pay the claimant that amount. But an accurate empirical prediction about the claimant’s future earnings cannot supply the correct normative baseline, not just because it involves projections about how moral uncertainty about justice will be resolved, but also because it will reflect injustice in the distribution of opportunities, including along the lines of race, gender, and social class. It seems we need a standardized measure of lost opportunities indexed to characteristics of the claimant that will not reflect or compound unjustly earned privilege or unjust disadvantage. Perhaps the measure could reflect characteristics such as the claimant’s age or number of dependents. It clearly ought not to reflect the claimant’s race or gender, nor, perhaps, indicators of the claimant’s socioeconomic class, such as their education level.Footnote ⁵³

Should consequential losses arising from breaches of contract be further restricted to those that were reasonably within the contemplation of the parties at the time of contracting?Footnote ⁵⁴ The standard economic justification for this restriction is that it gives a party an incentive to communicate unusual consequences of a breach to their counterparty, thus enabling the latter to take appropriate precautions or adjust the terms of their relationship.Footnote ⁵⁵ But this explanation does not explain why consequential losses that were in the contemplation of neither party ought not to be recoverable. According to my conception, that restriction makes sense because the moment of contracting has special significance as a moment when the parties are deliberating about what justice between them requires. They can only deliberate about what is reasonably within their contemplation and thus as a general matter they ought not to be held liable for consequences of breach they could not reasonably have anticipated when the contract was formed, at least when it comes to breaches of contract that were not willful or in bad faith. Of course, it may turn out that justice does require the breaching party to compensate the victim for the unexpected losses the breach caused given the particularities of their relationship. But the parties ought to be the ones who make such determinations, except where a party exhibits contempt for the deliberative process through a willful breach.

C. Avoidability, Offsetting Benefits, and Collateral Sources

Considerations that restrict when courts ought to order recovery for consequential damages support restrictions on the types of losses for which damages may not be ordered on the grounds that the losses were avoidable by the breach victim. Losses that can reasonably be mitigated through unilateral action by the victim, for instance by stopping work on a project that the breaching party has indicated they no longer want, ought not to be recoverable.Footnote ⁵⁶ But when losses can be mitigated only when others also get involved, the determination that they are reasonably avoidable involves making a projection about the way in which the breach victim will settle normative uncertainty about justice with the persons willing to help to reduce the losses. Courts ought to be hesitant about premising remedial orders on predictions about what such settlements would look like.

That courts should be hesitant does not mean that courts should never premise remedial orders on such predictions. Under some conditions, it is clear what the victim reasonably ought to do to minimize their losses. For example, when a breach of contract consists of the failure to deliver a fungible good and there is a well-functioning market for substitutes, preventing the victim from recovering losses that could have been avoided as a result of the victim’s purchase of a substitute good at the market price does not involve the court imposing its judgment about what justice requires because the victim and breaching party determined that justice between them required the exchange of such a good for the contract price when they entered into the contract, and the nature of the transaction is such that the identity of the seller likely is not of significance to the buyer.

It is more problematic to deny recovery for avoidable losses when these conditions do not hold. When the identity of the party with whom the contract was made mattered to the breach victim, declining to compensate the victim for losses that the victim could have recouped by contracting with a similar party imposes a burden on the victim to enter into an agreement with such a party or else suffer those losses. This is a significant deliberative burden to place on the victim, once we understand a morally valid contract to represent a determination by the parties of what justice between them requires. It becomes more of a burden the thinner the market for such parties as then the victim’s choice as to who to contract with to minimize their losses is significantly constrained.

These considerations suggest that breaches by employers of employment agreements in particular ought to be subject to, at most, a very limited duty to mitigate.Footnote ⁵⁷ This is not to say, of course, that the parties might come to their own determination about what damages are required in light of alternatives available to the victim. Indeed, an ex ante stipulation of damages for an employer’s breach may reasonably take the likelihood of such opportunities into account, and a court ought to respect such a stipulation so long as it looks like the parties were deliberating in good faith about what justice between them requires in the event of such a breach.Footnote ⁵⁸ But in this case, both parties have had a say about the impact of alternatives that are or may be available to the employee in such an event rather than having a view about which such alternatives the employee ought to pursue being imposed on them by an outsider. It also follows that duties to mitigate ought to be less stringent in the realm of tort where the parties lacked the opportunity to deliberate in advance about consequences of the wrong and there is no agreement from which we can infer the preferences of the victim.

As for when a plaintiff’s court-ordered damages ought to be reduced to reflect an offsetting benefit—a benefit that the plaintiff would not have realized but for the wrong—my conception suggests that this generally ought not to happen unless the benefits are realized pursuant to the plaintiff’s duty to mitigate. Avoidable losses are losses deriving from the defendant’s wrong that the plaintiff has a presumptive responsibility to minimize in virtue of the general characteristics of the defendant’s wrong against the plaintiff. What else the plaintiff does in response to the wrong they have suffered might have a bearing on the just resolution of the dispute between the defendant and the plaintiff. But whether and how it does so will be dependent on peculiarities of their relationship and circumstances, given that the plaintiff was not under a duty to pursue those benefits merely in virtue of the fact of the defendant’s wrong against the plaintiff. Thus, my conception endorses the collateral source rule, which prevents money damages from being reduced by an amount that was reimbursed from an independent source, such as an insurance company.Footnote ⁵⁹ The plaintiff’s independent relationships with third parties have no general relevance to the dispute between the plaintiff and defendant, even if they might in particular cases.Footnote ⁶⁰

D. Injunctive Relief

Injunctions order a defendant to take or refrain from certain actions in the service of preventing a future or continuing legal wrong or repairing losses the plaintiff has suffered on account of a past wrong. When should a court issue injunctive relief? The answer, I will suggest, depends on the existence and extent of the normative uncertainty concerning what justice between the parties requires not only about the remedy but also about the underlying right. Substantive questions about the nature of any underlying wrong thus end up intruding on the question of remedy in a way that they do not when the question is one about damages alone.

The justification for this substantive turn arises from the asymmetry of benefits and burdens that is more likely to be associated with an injunctive remedy than a damages remedy. Whereas a damages remedy purports to impose a monetary burden on the defendant that is commensurate with the plaintiff’s gain, injunctions sometimes very clearly impose disproportionate burdens on the defendant.

Sometimes, of course, they do not impose disproportionate burdens. Consider an order to pay the contract price or an order to hand over identifiable property where the defendant has not changed its position significantly based on a reasonable assumption that the property is theirs. Even in these cases, there clearly will be an asymmetry between benefits and burdens when the defendant is very poor and the plaintiff very rich. But such asymmetries, which also exist for damages remedies, depend on the needs and circumstances of the parties. Thus, in keeping with the principle of generality courts ought not to make determinations about their existence and import, leaving those matters to the parties to figure out.

In the case of other kinds of orders, it will more often be true that a rough symmetry between benefits and burdens is lacking. At the same time, the moral import of those asymmetries for the parties’ relationship will often be uncertain and dependent on the circumstances of their relationship. Thus, attending to the asymmetry risks intruding on the parties’ ability to determine for themselves what justice between them demands. It therefore makes sense that there would be a presumption against injunctive relief such as that embodied in the adequacy rule: injunctions may not be awarded when there is adequate remedy at law in the form of a damages remedy.Footnote ⁶¹

When the inadequacy of damages can be established, courts may then go on to balance the hardships to determine whether the injunction should nonetheless not be awarded.Footnote ⁶² In keeping with the principle of generality, the inquiry ought to focus on general considerations that avoid inquiring too much into the circumstances of the parties. It is at this point that substantive questions about what has transpired between the parties become relevant. Consider a trespass case involving the encroachment of a structure on the defendant’s parcel of land onto the plaintiff’s neighboring property. Suppose that when each party purchased their neighboring lots, each was mistaken about where the boundary between the parcels lay and thus unaware that the structure was encroaching. Even though the law would find a trespass here, it is morally uncertain whether the legal boundary describes what the parties owe one another as a matter of justice. Given, moreover, that both parties were content with the situation prior to the discovery of the true legal boundary between their lots, it is likely that ordering the defendant to remove the structure would impose significant hardship on the defendant relative to the plaintiff’s gain from the removal of the encroachment.Footnote ⁶³ The threat of such a remedy would also give a bad faith plaintiff a tremendous upper hand in negotiations with the defendant about what a settlement ought to look like.Footnote ⁶⁴ Even if both parties were willing to deliberate in good faith about what justice between them requires, the anticipation of such an order would likely influence the outcome so long as they perceive the courts to have relevant epistemic expertise. In effect, the denial of the injunction in favor of damages that compensate the plaintiff for any decrease in the market value of the lot caused by the encroachment seems appropriate in part because of uncertainty about whether the defendant’s structure really infringes the plaintiff’s rights at the moral level.

We should take a different view had the defendant built the structure knowing that it would intrude on the plaintiff’s lot. Such an act by the defendant is a clear procedural violation, at least so long as the existing property line constitutes a reasonable determination by the legal authorities about what line between the lots justice requires. The defendant ought to have negotiated with the plaintiff prior to erecting the structure rendering the defendant’s objection to the burdens of the injunction hollow. (The defendant’s action might also be substantively unjust if the defendant in fact owed it to the plaintiff not to encroach on the plaintiff’s land as a matter of substantive justice, though, assuming the reasonableness of the legal determination of the original boundary, moral uncertainty about justice will generally make it unclear where the just boundary lies.)

An appeal to underlying moral uncertainty about the parties’ rights also helps to explain why injunctions against negligence are generally unavailable. A different explanation for their unavailability might begin with the formal definition of the tort as a harm-inclusive wrong: a defendant has not committed negligence against a plaintiff by acting carelessly unless physical harm to the plaintiff or the plaintiff’s property results.Footnote ⁶⁵ On this conceptualization, negligence could not be enjoined in advance of harm occurring because the tort is not complete until the harm materializes, except perhaps in unusual cases where a person was “continuously injuring [another] by continuously failing to take care.”Footnote ⁶⁶ But others view the wrong as the failure to take reasonable care to avoid injuring others with harm serving merely as a predicate to actionability rather than part of the definition of the wrong.Footnote ⁶⁷ If the wrong is understood in the latter terms, there is no reason, in principle, why various forms of unreasonably risky conduct might not be enjoined prior to harm being realized. Perhaps we do not see such injunctions because of practical difficulties surrounding specifying what counts as reasonable care in advance of harm being realized.Footnote ⁶⁸ But it is not obvious that such an explanation can account for all cases. Sometimes it is obvious that the law would declare certain conduct unreasonable. In such cases, a better explanation for the unavailability of injunctive relief is that the law’s judgment that conduct is unreasonable speaks generally not to the specifics of the parties’ relationship, and thus, given moral uncertainty about justice, does not entail that such conduct is, in fact, a wrong against a potential victim. Enjoining such conduct in advance, however, expresses the law’s view that it is impermissible, thus interfering with the parties’ ability to conclude otherwise in light of their own reflections about what justice between them requires. (In principle, it would not preclude the parties from coming to a different conclusion before an injury occurs, but parties to tort disputes typically have limited opportunities to negotiate in advance.)

Finally, consider a nuisance case like Boomer v. Atlantic Cement Co. where the court denied the plaintiffs an injunction against a nuisance caused by a cement plant because the economic harm that would have resulted from closing the factory was disproportionate to the inconvenience caused by the dust and other particles that the factory was omitting.Footnote ⁶⁹ Stephen Smith criticizes the remedial decisions in Boomer and similar cases on the grounds that they seem to depend on the court’s view of the underlying wrong (or lack of it).Footnote ⁷⁰ But nuisance cases often involve scenarios that seem especially morally uncertain at the level of the parties’ underlying rights. The intrusion of substantive concerns on remedial determinations may just reflect the underlying moral complexity of the problem. These are also often cases where the rights of nonparties are implicated, such that on the democratic conception they are not apt for private resolution by the disputing parties. There is, therefore, less of a worry if courts intervene to balance all the equities when issuing their remedial orders in such cases because the parties to the dispute ought not to be the ones to resolve moral uncertainty about the rights of third parties.

E. Access to Justice, Attorneys’ Fees, and Punitive Damages

We have seen that the democratic conception of remedial law starts from an idealized picture of the private agreements that remedial law ought to be designed to facilitate. An agreement is morally valid when it reflects the parties’ good faith joint determination of what justice between them plausibly requires given what transpired between them. This is a normative claim about the type of agreement that remedial law ideally ought to have in view rather than a descriptive claim about how parties settling a dispute behave. Settling parties need not always be perfectly motivated to relate to one another justly for my conception to have bite, though widespread unwillingness to do justice might mean that remedial law ought to take a different tack, grounded in a less ideal justificatory theory. At least so long as enough parties are motivated to conform to justice, however, remedial rules should be designed primarily with good faith parties in mind. Such parties will take those rules as a starting point for their deliberations about what justice between them requires.

An important reformist implication of my account, therefore, is that there ought to be much more oversight of settlement and litigation processes to encourage disputants to engage in dispute resolution in good faith and with an eye to articulating a reasonably just resolution of their controversy.Footnote ⁷¹ The system also needs to ensure a level playing field. Everyone should have access to civil council, for example, no matter their means.

My conception also points to an important place for the recognition of procedural wrongdoing. Agreements are morally valid not merely when they articulate a substantively plausible vision of justice between the parties; the vision must also be the output of good faith deliberations between the parties about what justice requires. When a party to a dispute fails to deliberate in good faith or otherwise makes opportunistic use of the dispute resolution process that party wrongs the other in an important sense regardless of the outcome. This suggests a possible role for punitive damages. Punitive damages are aptly awarded against a party who engages in such procedural wrongdoing. Understood in these terms, they are not like criminal punishment, as they purport to compensate the victim for a procedural wrong that has been committed against them.Footnote ⁷²

My conception also sheds light on the proper allocation of dispute-resolution costs across the parties. When one party has clearly committed an injustice against another and it is clear what that party ought to do in response, that party ought not to resist that outcome and to the extent that they do, they ought to compensate the victim for the victim’s dispute resolution costs including reasonable attorneys’ fees. But pervasive moral uncertainty about justice means that such cases will be the exception rather than the rule. When the demands of justice are uncertain, both parties are under a duty to deliberate with one another about what justice requires and so ought to bear their own costs, at least so long as they are both willing to deliberate together in good faith. When one party refuses to engage in good faith deliberations or otherwise disrupts the process, however, that party ought to pay the costs of the other side.Footnote ⁷³ At the same time, too much policing of the settlement process may end up undermining the free deliberations of well-motivated actors. Courts may accordingly have reason to underenforce this duty of good faith in practice.Footnote ⁷⁴