I.

In the second part of the Discourse on the Origin and Foundations of Inequality among Men, Jean-Jacques Rousseau famously blames the invention of metallurgy and agriculture for putting an end to “the happiest and the most lasting epoch in the development of humankind.”Footnote ¹ Two thoughts lie behind this characteristically radical claim. The obvious one is that these new techniques allowed a previously inconceivable accumulation of wealth, so that inequalities among individuals became far more pronounced, with all the negative consequences Rousseau vividly depicts. The less obvious one concerns not the results that the new techniques made possible but rather how these results had to be brought about, and in particular the fact that individuals had to work together to achieve them. As Rousseau puts the point, in what must be the severest indictment of collective endeavor in print:

All that was lost, however, “the moment one man needed the help of another”Footnote ²—which is precisely what happened when metallurgy and agriculture became widespread.

The two thoughts are not independent: presumably, human beings came to need one another's help in part because they lost the psychological ability to do without the vast resources that could be produced through metallurgy and agriculture. But since my goal here is not to reconstruct Rousseau's position, I put these complexities to one side and focus exclusively on his most intriguing suggestion: that engaging in collective action jeopardizes an individual's freedom (leaving aside whether it also threatens her health, goodness, and happiness). As often with Rousseau, I suspect that he is on to an important truth but fails to articulate it with the required precision. One of my goals here is to provide a detailed account of the tension between collective action and individual freedom. A further goal is to defend a specific position as to how the tension can be resolved—a task that will take us well beyond the territory explored by Rousseau, closer to the ground covered by Immanuel Kant's discussion of contract in the Doctrine of Right.Footnote ³ My main claim will be that resolving the tension identified by Rousseau requires putting in place a system of contract rights. At a general level, this implies that the need for contract rights is grounded, at least in part, in the demands of freedom. The point also has specific ramifications for contract theory: the role that contract rights play in reconciling collective action and freedom turns out to be crucial to understanding how—and by whom—they can legitimately be enforced; it also explains why expectation damages should be the standard remedy for breach of contract.

I proceed as follows. I begin by explaining the general tension between collective action and freedom. I argue that it stems from the very nature of collective action and gives rise to a distinct problem—the normative problem of collective action—that finite rational agents need to address. I then show that addressing this problem is harder than it looks. In particular, I show that the problem cannot be solved merely in terms of moral obligations: individuals engaging in collective action will find themselves depending on one another's choices in a manner that undermines their freedom even if each is morally bound to do her share.

I then explain how putting in place a system of enforceable contract rights can solve the normative problem and what implications this has for how we should understand the enforceability of contract rights. Contracts are sometimes said to be enforceable in virtue of a need to ensure that certain promises are kept or in virtue of the binding force of the parties' consent, but I maintain that the issue is more complex.Footnote ⁴ If rational agents can create enforceable contract rights through acts of will, it must be because such rights are essential for freedom in the first place. As I go on to show, the point sets strict constraints on the kind of agent that can legitimately enforce contract rights and has illuminating implications for the type of damages appropriate for breach of contract.

II.

Let me begin with an outline of the problem that motivates Rousseau's worries. Clarifying its exact terms takes up much of what follows, but it will be useful at the outset to have a sense of where we are headed. Collective action poses two threats to individual freedom. The first is closely tied to the conception of freedom that Rousseau adopts—a conception according to which freedom demands not only the absence of interference on the part of others but also suitable independence from their choices.Footnote ⁵ The threat arises because a collective action can normally be carried out only if each member of the collective agent does her part; hence, by participating in a collective action, an individual comes to depend on whether others choose to do their part, which is straightforwardly in tension with freedom on Rousseau's view.Footnote ⁶ The second threat arises in the attempt to address the first. To avoid depending on one another, individuals acting together need a way of making each partner do her share of the collective action; in other words, they need an enforcement mechanism. Predictably enough, this gives rise to a second threat, since exerting control on individuals is not generally compatible with their freedom.

Two demands thus have to be met to resolve the tension between collective action and individual freedom. First, a demand of independence: each member of a group engaging in collective action must be suitably independent from the choices of other members. Second, a demand of legitimacy: the control exerted to satisfy the demand of independence must itself be consistent with individual freedom. Meeting both demands together constitutes what I call the normative problem of collective action. As I indicated above, my central aim here is to elucidate the nature of this problem and to determine the conditions under which it can be solved.

The recent literature on collective action is concerned mostly with a different problem: that of making sense of the possibility of action that is truly collective without committing oneself to the existence of mysterious collective entities.Footnote ⁷ This ontological query is undeniably important: any explanation of collective action taking the existence of collective entities for granted would clearly beg the question. But I want to suggest that the normative problem is no less fundamental. This is not to say that meeting its two constituent demands is essential to something's counting as collective action in the minimal sense. That may be the case for the demand of legitimacy—it would be odd to say that a group of individuals forced at gunpoint to transcribe five pages each from the Critique of Pure Reason should count as having engaged in collective action together—but it is not obvious that the same holds for the demand of independence. The point is rather that the normative problem must be solved for collective action to fulfill its intrinsic aim. Let me explain.

To grasp the normative problem's significance, we need to ask why it matters that there be such a thing as collective action in the first place. My suggestion is that the possibility of collective action matters because it is essential to our freedom. This follows from the fact that we are finite beings: we have limited capacities and resources, so that there are countless things that we can accomplish only with the help of others. Consequently, our freedom would be curtailed if acting with others was not possible, much in the way it would be if making use of external objects was not possible. The comparison may appear odd, since external objects seem essential to the very possibility of human projects, whereas there is a vast range of projects that one could pursue without the help of others. But the point is that for a finite being, the actions of others are potential means for the pursuit of projects; unjustified restrictions on the possibility of making use of those means curtail freedom, just as unjustified restrictions on the possibility of making use of external objects do.Footnote ⁸ And the point holds even if, as things turn out, I never actually need the help of others for the projects I elect to pursue.

The normative problem is thus intrinsic to the very nature of collective action. It does not take the form: collective action is a good thing, but it is in tension with other important values. It takes the form: collective action derives its importance from the fact that it is essential to the realization of a certain value, namely, individual freedom; yet the conditions of possibility of collective action appear to be in tension with that very value. In short, there is no escaping the normative problem: collective action can achieve its goal only if the two demands of individual freedom are met. If this is correct—the case remains to be made, of course—then the tension between collective action and freedom is deeper than Rousseau lets on. He assumes that human beings were perfectly free prior to the invention of techniques requiring them to work together. But no particular technological development is to blame for the fact that the actions of others can make a material difference to the pursuit of one's projects, which is ultimately what gives rise to the normative problem in my view. That is not a contingent fact about the techniques required to make the best of our environment but an inescapable fact about our nature as finite rational beings.

III.

So far I have discussed the normative problem in its general outline; it is now time to fill out the details. The first task is to explain exactly how the problem of dependence arises—how individuals who act together come to depend on one another's choices in a way that undermines their freedom. Consider a simple example. Suppose that, carelessly ignoring Rousseau's warnings about the perils of metallurgical endeavors, my friend Larry and I decide to build cars together. Assume that some part of the assembly process requires that we both be there, so neither of us can successfully carry out the task alone.Footnote ⁹ Also assume, at least for now, that we take on the project without incurring any moral obligation toward one another—say, by making it explicit that we each reserve the right to change our mind about the matter.Footnote ¹⁰ My contention is that, under these circumstances, taking on such a project undermines our freedom. The question is, how can this be the case? How can my willingly deciding to build cars with Larry be in tension with my freedom?

As I said above, the thought rests on a specific conception of freedom, following which a person's freedom consists in her ability to set and pursue ends for herself without depending on the choices of others. Providing a sustained argument for this view is beyond the scope of this paper, but I do want to mention an important underlying intuition, namely, that others can undermine my freedom not just by actively interfering with what I do but also by having the ability to interfere. To take an obvious example, if you have implanted a chip in my brain that allows you to control my movements, then you do not actually have to use it to undermine my freedom: the sheer fact that you could use it—the sheer fact that you have that kind of power over me—suffices to make me unfree. Indeed, it seems plausible to say that so long as my ability to set and pursue ends is subject to your choices in this manner—so long as I can act only by your leave—I am every bit as unfree as if you had actually prevented me from doing what I want.Footnote ¹¹

That there is at least a superficial tension between freedom so conceived and the kind of collective endeavor described above should be obvious. When I take on the end of building cars with Larry, I make myself dependent on his choices in a straightforward sense: I cannot assemble the car by myself; hence whether I can achieve my end depends on whether Larry chooses to stick with the plan. If he decides at any point that he would rather do something else—as he is entirely free to do in our scenario—then I am left hanging high and dry. Now, to be clear, the problem is not just that something essential to my project (Larry's help) is beyond my control. Countless things that are no less essential to my project—the absence of a violent earthquake, say, or the presence of relatively normal atmospheric conditions—are also beyond my control. Crucially, however, such factors are not under anyone's control. Indeed, it would be odd to call them means, since they are not things that human beings can make use of to pursue their ends: they are simply conditions that must obtain for the pursuit of certain ends. By contrast, what creates trouble in our example is that Larry's actions are means: they can be—and are—under someone's control. The problem is, that someone is Larry, not me.

That may seem puzzling. How could Larry's having control over his own actions be in tension with my freedom? Surely, if freedom demands that we not be subject to the choices of others, it must also demand that no one be required to take up an end simply because another person has adopted it—and hence that each person get to choose what projects she wants to get involved in. On what grounds could I then claim the right to decide when and to what end Larry has to lend me his help? But this is not what happens in our scenario. We are imagining that Larry has agreed to help me and that as a consequence I have taken on a project for which one of his actions is a means. My suggestion is that it is this fact—my taking on a project for which Larry's help is a means—that makes me dependent on his choices and hence jeopardizes my freedom. In other words, the problem is not that Larry has agreed to help me; if anything, that opens up new possibilities for me. The problem is that having agreed to help me—having made his help available to me as a means with which to pursue my project—it remains entirely up to him whether he is actually going to help me or not.

To emphasize, the point is not that I need to be certain to have all the means I need for a given project at my disposal—as if I could not take on a project without knowing for sure that all the help I need will be forthcoming. I could very well take on the project of building cars with someone, without being at all certain to find anyone willing to help. My freedom would not be compromised if I turned out to be unable to build cars because no one wanted to help me—no more than it would be if all the city's hardware stores mysteriously ran out of a part I need. This is a crucial point for Rousseau and Kant: as they understand it, freedom does not depend on one's being successful at pursuing any given end but on one's being able to set and pursue ends for oneself.Footnote ¹² After all, even if I cannot build cars, there are countless other things that I can do.

The problem in our scenario is thus not that I may or may not succeed at building cars but rather that someone else gets to decide whether I do or not. That is, the problem is not the uncertainty that Larry introduces in my project, as if he were an unpredictable summer day and my project a picnic. It is that one of the means for my project can be used only by his leave and that there is no reason for that to be the case once he has agreed to help me.Footnote ¹³ My suggestion is the following: if there is a way to remedy this problem—if it is possible to make it the case that Larry's doing his share is not left to his discretion without thereby infringing on his freedom—then freedom demands that it be done.

Although we have been focusing on a specific example, I want to stress that the problem of dependence is systematic: it springs from the general fact that those who engage in collective action inevitably come to rely on one another's contribution. I also want to stress that, although we have assumed so far that Larry's help is essential to my car-building endeavor, the assumption is not strictly speaking necessary: it merely serves to highlight the problem of dependence. In fact, the problem arises so long as Larry's help counts as a means for my project. This requires that his contribution be material, since I cannot be dependent on his performance if it would make no difference to my project. But so long as it does make a difference, and so long as it is my chosen means, the problem of dependence arises. Can this problem be solved—can individuals acting together retain suitable independence from one another's choices? Put more dramatically, can collective action be made fully consistent with the freedom of its participants and thereby be allowed to serve its intrinsic purpose? I now turn to that question.

IV.

There may seem to be an obvious solution to the dependence problem. We explicitly assumed above that Larry and I are in no way morally obligated to do our share. Might that be the cause of our troubles? After all, if Larry were morally obligated to do his share, it would not be entirely up to him whether to do it or not, and I would not be so straightforwardly dependent on his choices. The solution is all the more tempting given that people acting together normally do incur obligations toward one another or can easily do so by exchanging promises.Footnote ¹⁴ But can Larry's having a moral obligation really solve the dependence problem? At first glance, it does seem to help. If Larry promises to do his share, then he is no longer free, morally speaking, to do as he pleases. He is no longer allowed to opt out unilaterally: unless I release him, he cannot back out without wronging me. Doesn't that mark an improvement on a situation in which it is entirely up to him whether to do his share or not?

Whatever we make of that suggestion, it should be clear that the presence of a moral obligation leaves me dependent on Larry's choices in the sense that concerns us. For suppose that in spite of seeing himself as morally obligated to do his share, Larry still fails to show up for the task. That is clearly something he can do—not something he can do morally, granted, but that thought will be cold comfort if he chooses to do it anyway. The problem is that Larry remains free to do his share or not as he sees fit. If he decides to abandon me, for whatever reason, nothing will stop him. From my point of view, then, the situation remains essentially unchanged. If Larry's being morally obligated to do his share makes any difference, it is only that instead of depending directly on Larry's choice to do his share, I now depend on his choice to act morally.Footnote ¹⁵

The implication, I want to suggest, is that agents acting together will meet the demand of independence only if each can be forced to do her share when needed. Only in this way will the common endeavor not be left to depend on each person's choices, because only in this way will it not be up to each whether to do her share or not. But this brings us to the second problem I mentioned at the outset: that forcing people to do things is not in general a promising way to help them realize their freedom. The use of force may prevent those engaging in collective action from depending on one another's choices, but only at the risk of falling short of an even more obvious demand of individual freedom: the demand of legitimacy, which states that coercion should be used only in a manner compatible with freedom.

Once again, there may seem to be an easy fix: what if Larry and I explicitly consented to our agreement's being enforced if needed? Would that not suffice to meet the demand of legitimacy?Footnote ¹⁶ It is true that consent can, under certain circumstances, make actions compatible with an individual's freedom that would otherwise not be. For instance, your touching me does not restrict my freedom if you have my consent, since your action is made possible by and therefore ultimately results from my own choice. But the point does not transfer to the kind of situation we are considering. Any use of force grounded in an agreement will normally take place after the act of consent that is supposed to justify it; under such circumstances, it is unclear how consent by itself can reconcile the use of force with freedom.

Our scenario can be modified to illustrate the point. Suppose that Larry and I explicitly agree that whoever fails to show up for work will be forcibly dragged down to the plant by the other. And suppose that Larry still fails to show up. I go find him and remind him that he gave his word, to which he replies that he has changed his mind about the whole thing: he wants out of the plan. I then point out that he has not merely promised to do his share: he has agreed to be dragged down forcibly to work, should he fail to come on his own. To this, he rejoins that he has changed his mind about that as well: he no longer consents to being dragged down forcibly. The question is, can I nonetheless legitimately use force against Larry to make him come to work?

Let me be clear about what is at stake here. It is not just whether I have some justification for using force against Larry. That may well be the case, but the demand of legitimacy sets the bar higher than that. The question is whether the fact that Larry has previously given his consent suffices to make my present use of force against him compatible with his freedom. It does not. To the extent that actual and present consent indicates that a certain use of force against an individual results from her choices, it plausibly satisfies the demand of legitimacy. But it is not clear how the reasoning would go when an individual has previously consented to the use of force but no longer does. In such a case, the person does not choose to have force used against her; indeed, she explicitly chooses the opposite. Admittedly, she has formerly chosen to have force used against her in the future, should certain conditions obtain, but it is unclear why her former choice should determine what is compatible with her freedom now.

A further example may help sharpen the point. Consider the case of a boxing match. Boxers routinely engage in behavior that would normally count as battery. We do not commonly consider it as such, because the participants consent to being hit provided that certain rules are observed. However prejudicial the game may be to the participants' health, it is not contrary to their freedom precisely because of this consent. But note that the point is only plausible if the consent is actual and present—that is, if each boxer remains free to end the match at any moment. The cost of doing so may be significant—at the very least, the forfeiter will lose the match, with all the financial and professional consequences that entails. But it is crucial that each boxer get to decide at every moment whether force can be used against him. If a boxer who refuses to continue fighting is forced at gunpoint to stay in the ring, he is no longer engaged in a boxing match: he is undergoing battery, pure and simple.Footnote ¹⁷

Larry's case is more like that of a boxer who has agreed in advance that there could be no forfeiture and then changes his mind about the whole thing. But ultimately that makes no difference. Even on those terms, the boxer's former consent would not suffice to make the present use of force against him compatible with his freedom, because there is no reason the choice he made before the match should have priority over the choice he is making now; if anything, the reverse seems true. Of course, these complications will not arise if Larry still consents to our enforcement procedure when I use force against him. If that is the case, if his consent is actual and present, then the enforcement procedure will plausibly be consistent with his freedom. But satisfying the demand of legitimacy in this way simply sends us back to the dependence problem, since there is no guarantee of Larry's continued consent to the enforcement procedure. I have to depend on Larry's choosing to continue to consent, just as I previously had to depend on his choosing to do his share or to act according to his moral obligation. The normative problem remains unsolved.

V.

We seem to be in something of a bind. We have seen that if two individuals are to act together while meeting the demand of independence, then it must be the case that each of them can be forced to do her share. At the same time, the normative problem will be solved only if any potential use of force meets the demand of legitimacy—that is, only if it is itself consistent with freedom. The difficulty is that the most straightforward way of making the use of force consistent with freedom—obtaining the actual and present consent of those who are subject to the use of force—sends us back to the dependence problem. Despite what is sometimes claimed, consent cannot conjure enforceable obligations out of thin air. But it would be wrong to conclude that the normative problem cannot be solved or that consent plays no role in its solution. Obviously, if Larry can be forced to do his part in our endeavor, it is partly because he consented to the enforcement scheme. What our discussion shows is that this cannot be the whole story. We need to ask how the rest of it goes.

I propose turning to the normative problem itself for an answer. More specifically, I want to explore the possibility that it is precisely because agents engaging in collective action jeopardize one another's freedom that their agreements can legitimately be enforced and that any justificatory role played by consent should be understood against that background. Here it may prove useful to look at the matter in the terms I used when I originally introduced the normative problem. I said in Section II that the link between collective action and freedom boils down to a simple idea, namely, that the impossibility of having another's actions available as a means with which to pursue one's ends would constitute an unjustified restriction of freedom. That is the problem that our car-building example is meant to bring out; indeed, the whole point of the example is to show that Larry's action can be truly available as a means for me only if I have an enforceable right to his action. In other words, the point is to show that making another's actions into means for oneself is possible only if there can be enforceable contract rights.Footnote ¹⁸ Absent such rights, the choice whether or not to carry out an action necessarily remains at the agent's discretion, and the action accordingly fails to constitute a means with which others can set and pursue their ends.

It is against that background that we must assess the justifiability of forcing Larry to do his part. The issue is ultimately whether his actions must remain at his own discretion or whether they can, under certain circumstances, become means for another. If freedom demands that the latter be possible, as I have argued that it does, then it demands that Larry be forced to do his part if necessary. In other words, my suggestion is that the enforcement of contract rights can be reconciled with freedom in something like the manner in which coercion in general gets reconciled with freedom in Kant's Doctrine of Right: by showing that using force (in this case, to allow one person's action to become a means for someone else) simply amounts to the “hindering of a hindrance to freedom.”Footnote ¹⁹

Two things need to be established to apply this form of reasoning to our case. First, that the impossibility of having another person's actions as one's means (which is what the enforcement of contractual obligations is meant to hinder) constitutes a hindrance to freedom. I defended that idea in Section III. Second, that A's having an enforceable contractual right to B's action need not be incompatible with B's freedom. I now turn to that point. I want to suggest that A's having a right to B's action will be compatible with B's freedom provided that certain conditions are met. Given that these conditions must obtain for the demand of legitimacy to be met, I refer to them simply as conditions of legitimacy. As will become clear below, I take these conditions to be deeply entrenched in contract law as we know it.

Let me begin with an obvious condition of legitimacy concerning the manner in which A can come to acquire a right to B's action. If A's doing so is to be compatible with B's freedom, then it must be up to B to decide whether he will make his action available to A as a means with which to pursue her projects. As I said above, it would be plainly inconsistent with B's freedom if he could be dragged into A's projects despite himself. One should have appropriate control over whether one becomes bound by a contract or not. This corresponds to a fundamental tenet of contract theory: a contract is valid only if each party is free to take part in it or not, as she sees fit.Footnote ²⁰ It is in light of this idea that the role of consent in our broader story should be understood: securing B's consent is the obvious way to ensure that the decision to make his action available to A is his decision in the relevant sense.Footnote ²¹

This does not mean that consent is strictly necessary: giving B sufficient opportunity to avoid binding himself may be sufficient.Footnote ²² Furthermore, this is not to grant consent the kind of self-standing justificatory force that was found to be problematic in the previous section. If B gives his consent to helping A but then changes his mind about doing so, the justification for using force against him will not be that B bound himself through his previous act of consent alone. Rather, it will be that a system of enforceable contract rights is required by freedom and that B's previous act of consent ensures he was not bound despite himself. Consent can give rise to enforceable contract rights, the thought goes, but only in virtue of the deeper link between contract rights and freedom.Footnote ²³

The second condition of legitimacy I want to mention concerns a point seldom taken up in discussions of contract rights: that of knowing who has the authority to enforce such rights. It may seem surprising that the question should arise at all, since one may think that the parties to a contract are free to decide, at least in principle, who is to act as enforcer of their agreement. That would be correct if the enforceability of contracts were grounded in consent alone, but matters are more complex if their enforceability is grounded in freedom, for then we need an agent who can enforce rights without itself infringing on freedom. I take the main implication to be that the private enforcement of contracts is ruled out. The reason, in a nutshell, is that private arrangements necessarily make enforcement contingent on the choices of a private party—on his judgment that enforcement is required and on his decision actually to carry it out. And this just raises the dependence problem anew.Footnote ²⁴

Going back to our example will help illustrate the point. As before, suppose that Larry and I have agreed that if one of us fails to show up for the assembly the next day, then the other is authorized to use force against him (within reasonable limits) to make him do his share. When Larry fails to show up, I go and remind him that I have the authority to drag him to work. To this he replies that he no longer takes our agreement to be valid—his circumstances have changed, he explains, and he thinks that the change nullifies the agreement. Of course, it is possible for Larry to say all this in bad faith, but let us assume that such is not the case: he sincerely believes that the agreement no longer holds, while I sincerely believe that it does. He can then ask me what authority I have to use force against him. It may be tempting to say that I have authority in virtue of our prior agreement, but then (as we saw in the previous section) he can point out that once he has changed his mind, it is unclear how the prior agreement can bear the required justificatory weight. He can also point out that by using force against him, I impose my judgment concerning what the agreement requires of him and thus subject him to my choices precisely in the way that freedom demands that we avoid. Why, he may ask, should my judgment have priority over his in this manner? To this I have no answer.

The implication seems clear: to have enforceable contract rights against one another, we need to set up an agent whose choices are not private in the relevant sense. As I have explained elsewhere, this requires an agent that fulfills at least three conditions: it must aim to protect the rights of all who share a given territory; it must decide disputes in accordance with the rule of law; and it must be able to enforce rights with reasonable effectiveness.Footnote ²⁵ In a nutshell, the thought is that such an agent would not subject Larry to my choices, nor me to his, but rather subject both of us to a system of reciprocal coercion that is essential for our freedom. Now, it should be clear that any agent meeting these three conditions would count as a state in the minimal sense, with a legislative branch, an independent judiciary, and an effective executive branch. Thus we find that the state alone can legitimately enforce contract rights. To go back to our starting point, this means that collective action can be made fully consistent with freedom only within the state. In this way, the normative problem finds a fittingly Rousseauian solution. The dependence problem can be solved only through the state; insofar as freedom demands that the problem be solved, it also demands that there be a state—just as Rousseau (and Kant) thought it did.Footnote ²⁶

VI.

The picture I am drawing here may appear to have an odd implication. I argue that given the link between contract rights and freedom, it must be possible for one person to acquire an enforceable right to another's action. That may seem to entail that it must be possible to compel Larry specifically to perform his share of our car-building operation. Contract law, however, does not demand so much: it demands only that Larry either perform or compensate me. If Larry breaches our contract, the normal remedy is to make him pay monetary damages. Is this compatible with the argument I am putting forward? I believe that it is. Indeed, as I understand it, the approach I defend here not only is compatible with the option of compensation but actually requires it. Let me explain.

I claim that the enforcement of a contract is prima facie in tension with the freedom of the party against whom force is used but that acquiring enforceable contract rights must nevertheless be possible, since freedom itself demands that it be possible for an agent to have another's action as her means (provided that the other makes the action available to her). That may seem to imply that the standard remedy should be compelling specific performance. After all, if I have one of your actions as my means, am I not entitled specifically to your action? What should give us pause here is the visceral relation in which actions stand to the agents who perform them. Concretely, compelling specific performance is likely to mean forcing someone to do something at gunpoint. Do we want to say that freedom itself demands that it be possible to force Larry at gunpoint to show up for work? Not if there are less intrusive options. One such option is simply to seize some of Larry's property to compensate me. So long as doing so is equivalent to performance from my standpoint—that is, so long as my ability to pursue my end is not hindered by this way of proceeding—it is unclear why my freedom should demand more. And that being the case, it is hard to see how any greater interference with Larry's freedom could be compatible with the demand of legitimacy. Since compelling specific performance would obviously be a greater interference than the imposition of monetary damages, it follows that the latter's being the normal remedy is a condition of legitimacy.Footnote ²⁷

The view I defend also has implications for the type of damages that should be awarded. Following Lon Fuller's influential discussion, it has become customary to distinguish between three ways in which the law might react to a breach of contract.Footnote ²⁸ First, it might seek to restore the defaulting party to his precontract position by forcing him to return any gains he made at the other's expense; this would be to award restitution damages. Second, it might seek to restore the wronged party to her precontract position by compensating her for any expenditures she made or any opportunities she forewent as a result of her reliance on the contract; this would be to award reliance damages. Third, it might seek to place the wronged party in the position she would have occupied had the other party performed; this would be to award expectation damages. Fuller famously argues that there is something odd about the law's taking expectation damages to be the standard remedy, since its purported aim is to compensate the plaintiff, and the plaintiff never actually had what she is being awarded.Footnote ²⁹

The puzzle does not arise on the view I am putting forward. As I said above, if contract rights are ultimately grounded in freedom, then the law's aim in awarding damages must be to ensure that the wronged party's freedom is not hindered by the breach of contract. Compensation will therefore be adequate only insofar as it is equivalent to performance as a means with which the wronged party can pursue her end. In other words, to determine what constitutes appropriate compensation, we must ask what will put the wronged party in the position she would have occupied, as far as pursuing her end is concerned, had the other party performed.Footnote ³⁰ In normal circumstances, this requires damages to be based on the market value of the expected performance. A freedom-based view of contract rights thus entails that expectation damages should be the standard remedy and that such damages should be awarded independently of whether the wronged party has relied on the contract.Footnote ³¹ It also explains why other types of damages are inadequate. Reliance damages are inadequate because merely requiring Larry to restore me to my preagreement position would effectively leave him free to nullify our agreement at any time; the choice of whether to perform would remain entirely his, and hence the demand of independence would not be met. The same holds a fortiori for restitution damages.

VII.

By way of conclusion, I want to address one way in which the position I have defended here may seem in tension with common sense. I have argued for the strong claim that enforceable contract rights are necessary to make collective action consistent with individual freedom. But are there not perfectly unproblematic cases in which individuals act together without invoking contract rights? We do all sorts of things together informally—we paint houses, take walks, plan trips, and so on—without appearing to put our freedom at risk in the process. If anything, joint endeavors relying on trust rather than the possibility of enforcement seem to have special value precisely in virtue of that fact. Imagine what it would be like to take a walk with a loved one or to repaint one's house with a friend if each was contractually bound to do her share of the deal. Do we really want to say that such informal collective actions are in tension with their participants' freedom?

There is a danger here of romanticizing the value of relying on sheer trust. Enforceable rights do not always diminish the value of the relationships they regulate—surely, it would be odd to claim that marriage or common-law provisions for life partners have that effect. Still, the point is well taken: freedom cannot plausibly require that all interactions among individuals be grounded in contractual agreements. But that is not the conclusion we have reached. Freedom demands that it be possible for me to have the actions of others as my means; in other words, it requires that a system of contract rights be available, not that I make use of it at every turn. I can very well choose to paint the house with you informally, and thus allow you to retain discretion over my project, without thereby undermining my freedom. If I have a genuine choice about granting you that kind of discretion—if it is possible for us to enter into a contract should we so choose—then my decision to act informally with you does not threaten my freedom, since any dependence I find myself in results from my own choosing. By putting in place a system of contract rights, along with a state to enforce them, we reconcile both formal and informal collective action with freedom.