|A Critique of the Efficient Performance Hypothesis|
|Jody S. Kraus, Sunday, 22 July 2007 [View as PDF]|
The classic economic justification of contract law’s default remedy of expectation damages is grounded on the efficient breach hypothesis: that promisors should be permitted and encouraged to breach when the net gains from breach exceed the net gains from performance. Expectation damages ensure that all and only efficient breaches will occur because promisors will find breach profitable only if its benefits exceed the value of performance to the promisee. The efficient breach hypothesis, and the defense of expectation damages that rests on it, has long been criticized for being inconsistent with the moral intuition that promisors necessarily forfeit their right to choose not to perform their promise. In his essay, The Efficient Performance Hypothesis, Richard Brooks claims that the theory of financial options can be used to identify a new contract remedy that respects the promisee’s moral right to performance without sacrificing the efficiency goal served by expectation damages. In this Response, I argue that options theory is irrelevant to the debate Brooks engages, that the moral objection motivating Brooks’s new remedy is itself unmotivated, and that Brooks’s remedy is likely to be less efficient than expectation damages.Introduction
At bottom, the moral critics of expectation damages reject Holmes’s famous claim that a promise imposes an obligation on the promisor to choose whether to perform or pay damages. On Holmes’s view, a promise to do X is interpreted, by default, as a promise to do X or pay damages. The moral critics, however, argue that there is a crucial moral difference between a promise to do X and a promise to do X or pay a sum of money. If a promisor makes the first promise, his failure to do X is morally wrong, even if he pays a sum of money to the promisee. The payment of money in this case at best constitutes damages for breach of promise. If a promisor makes the second promise, his failure to do X is morally permissible provided he pays the sum of money. Rather than constituting damages for breach, the payment of money in this case constitutes performance.
More than twenty-five years ago, Alan Schwartz presented the first sustained economic argument for an alternative to the expectation default remedy that would grant promisees the right to elect either specific performance or expectation damages. Although Schwartz was principally concerned to demonstrate that such a rule was more efficient than the expectancy default rule, he also noted that the same moral argument made against expectation damages argued in favor of his proposed rule:
[I]f the compensation goal rests on a moral notion that promises should be kept, that contract remedies should effectuate the state of affairs—performance—that the promisor has a duty to bring about and that the promisee has a right to have brought about, then specific performance is a preferable remedy to damages even though it might generate higher costs. These costs would be the price of achieving the moral goal of contract remedies.
In his essay, The Efficient Performance Hypothesis, Richard Brooks expands on Schwartz’s insights, arguing that the feature of Schwartz’s rule that explains its moral appeal is a formal one that defines an entire class of contract remedies overlooked before Schwartz’s article and under-appreciated since. Following a well-developed literature applying finance theory to contract law, Brooks explains that the expectation damages rule can be viewed as giving the promisor a “call option” because it entitles the promisor, rather than the promisee, to decide whether he will perform or breach and pay damages. Brooks argues that the novelty of Schwartz’s rule is that, unlike a call option, in the event of breach it entitles the promisee to choose between requiring the promisor to pay damages and forcing the promisor to perform. Brooks then claims that Schwartz’s rule can be understood as giving a “put option” to the promisee because it entitles the promisee to compel the promisor to purchase the promisee’s right to performance for an amount equal to the promisee’s expectancy damages. Brooks’s general thesis is that contract remedies that have the form of a put option, unlike those that have the form of a call option, are consistent with the moral intuition that the promisee, and not the promisor, should be entitled to choose whether the promisor will perform his promise. However, Brooks argues that Schwartz’s rule violates the ex post efficiency criterion of the efficient breach hypothesis because its damages component, expectation damages, does not force the promisee to internalize the opportunity costs of the promisor. As a result, the promisee might choose specific performance even when it would be more efficient for the promisor to pursue an exclusive alternative to performance.
Brooks claims, however, that the defect in Schwartz’s rule is not endemic to the entire class of contract remedies that have the form of a put option. Indeed, Brooks proposes a modification to the damage component (the put option) in Schwartz’s rule that he claims makes it just as ex post efficient as the expectation remedy. Instead of measuring damages by the promisee’s expectancy, Brook’s rule awards an amount equal to the difference between the promisor’s expected gains from its best alternative to performance and its expected gains from performance. Brooks calls the damage component of his modified rule a “disgorgement” remedy because it requires the promisor to disgorge any gains from nonperformance. But unlike the traditional disgorgement remedy, Brooks’s version does not require that the promisor actually benefit from breach and then disgorge his gains. Instead, it entitles the promisee to the gain the promisor would have realized had he breached and devoted his contractual resources to his best alternative to performance. For clarity, therefore, I will refer to Brooks’s proposed damage measure as “hypothetical disgorgement.” Since his rule assigns the right to performance to the promisee, Brooks rests its defense on the “efficient performance hypothesis,” which claims that, unlike the efficient breach hypothesis, his rule will satisfy the demands of morality as well as ex post efficiency.I. Brooks’s Moral Thesis: Old Moral Wine in a New Bottle
When Brooks first introduces and describes his proposed alternative to expectation damages, he appears to suggest that it responds to the moral objection against Holmes’s view by transferring the Holmesian choice from the promisor to the promisee, thereby creating a put option that is the mirror image of Holmes’s call option. If Brooks’s rule were a true mirror image of Holmes’s rule, it would treat a promise as creating a right of the promisee to choose whether the promisor must perform or pay damages. According to this interpretation of Brooks’s rule, a promise to do X not only deprives the promisor of the right to pay money instead of doing X, but also deprives him of the right to do X instead of paying money! Rather than having the Holmesian right to choose between doing X and paying money, promisors would have the obligation to comply with the promisee’s choice between having the promisor do X or pay a sum of money.
Of course, under this interpretation, Brooks’s rule would be no less morally objectionable than Holmes’s rule. The moral objection to Holmes’s rule is premised on the view that a promise to do X imposes on the promisor an obligation to do X and confers on the promisee a right to have the promisor do X. This view entails not only that promisors who fail to do X violate their moral obligation, but that promisors who do X thereby satisfy their moral obligation. Thus, if a promise to do X creates a moral obligation that cannot be satisfied by the payment of money instead of the performance of X, then it likewise creates a correlative moral right only to the performance of X and not to the payment of money instead. In short, when it comes to the moral logic of promising, what’s good for the promisor goose is good for the promisee gander.
Brooks makes it clear, however, that his proposed rule is not a mirror image of Holmes’s rule, and so does not confer on promisees any pre-breach right beyond performance. Instead, it preserves the promisor’s pre-breach right to satisfy his obligation by performing. But if the promisor breaches instead, Brooks’s rule gives the promisee a right to choose between performance and payment of money. Indeed, Brooks explains that his proposed rule “should be viewed, not as a pure put option for the promisee, but as what [Ian] Ayres calls a Dual Chooser Rule, whereby the promisor’s initial action (choice) can trigger the promisee’s put option. For example, the promisor might first cause the promisee to demand assurances or breach anticipatorily, which allows the put option to come about.” On this definition, however, the expectation remedy qualifies as a put option as well: once the promisor breaches, the promisee has the right to compel the promisor to buy its right to performance at an amount equal to the promisee’s expectation damages. The distinction between Brooks’s rule and expectation damages has nothing to do with the difference between call and put options— both provide the promisor with the pre-breach right to perform and the promisee with a post-breach put option.
Instead, the only difference is that Brooks’s rule sets a different exercise price for the post-breach put option and, like Schwartz’s rule, provides the promisee with the alternative of specific performance. Any moral advantage of Brooks’s rule over expectancy, therefore, has nothing to do with its character as a post-breach put option. Nor can it be attributed to its different measure of damages. Although Brooks’s defense of the ex post efficiency of his rule does turn on the specific measure he proposes for its damage component (hypothetical disgorgement), he is clear that its moral appeal turns solely on the fact that it deprives the promisor of his legal right to choose whether to perform following breach.
Once we clear away the rhetorical brush of options theory, Brooks’s moral thesis reduces to the claim that remedies that entitle the promisee to specific performance avoid the moral objection to expectation damages, whether or not they provide the promisee with the alternative of damages. But this is hardly a surprise. The premise of the moral objection is that promisees should be legally entitled to force their promisors to perform. Specific performance gives them that right. Indeed, as I indicated at the outset, Schwartz made precisely this axiomatic claim more than twenty-five years ago. Nor, it turns out, is options theory any more than window dressing for Brooks’s efficiency thesis. Brooks argues that Schwartz’s rule fails to satisfy the ex post efficiency criterion underlying the efficient breach hypothesis because its damages component, expectation damages, does not force the promisee to internalize the opportunity costs of the promisor. He argues instead that his hypothetical disgorgement damage measure would force the promisee to internalize the opportunity costs of the promisor, and so would induce the promisee to exercise her right of specific performance only when her benefits from performance exceed the promisor’s opportunity costs. Thus, Brooks argues, his rule would induce promisees to exercise their right of specific performance in exactly the same (socially efficient) instances in which promisors entitled to perform or pay expectation damages would choose to perform. Although I will question this argument below, my only claim here is that the difference between call and put options has no bearing on the argument’s merits. Brooks casts his claim in option pricing terms: the exercise price of the promisee’s put option in Schwartz’s rule induces the promisee to exercise that option even when it is not efficient to do so, while the exercise price of the put option in Brooks’s rule induces the promisee to exercise that option only when it is efficient to do so. In plain English: Schwartz’s rule can be made ex post efficient by changing its damage component from expectation to hypothetical disgorgement damages. The only thing lost in translation is technical cachet.
In the remainder of this Response, I therefore focus on the standard moral argument, to which Brooks’s moral thesis can be reduced, that specific performance is morally superior to expectation damages, and on Brooks’s economic thesis that his rule, unlike Schwartz’s rule, is ex post efficient.II. The Moral Argument for Specific Performance
As we’ve seen, the standard moral argument for specific performance rightly insists that a promise to do X is not morally equivalent to a promise to do X or pay a sum of money. Paying a sum of money at best compensates the promisee for breach of the first promise, while it constitutes performance of the second. Thus, if a promise is properly interpreted as not providing the promisor the alternative of paying a sum of money instead of doing the promised act, then expectation damages fail to hold the promisor to his promise. Instead, it requires the promisor to compensate the promisee for the loss caused by the promisor’s moral wrong of breaching his promise. Payment of compensation does not right this wrong. Specific performance would prevent the promisor from committing the wrong, and thus prevent the promisee from incurring a wrongful loss. Brooks concludes that morality requires specific performance, and not expectation damages, because “‘[b]reach of a legal contract without excuse is wrong conduct.’ No one can doubt this.”
Unfortunately, this argument is a non sequitur. First, the moral wrongness of breach, by itself, is insufficient to establish the moral permissibility of legally enforcing promises, whether by specific performance or otherwise. Only the strictest and least plausible form of legal moralism holds that the legal enforcement of morality writ large is justified.
Second, morality permits legal enforcement of a promise only if the promisor intends his promise to be legally enforceable. If promisors do not intend to be legally bound, the moral obligation to keep promises provides no ground for any legal enforcement, let alone specific performance.
Third, even when a promisor makes a promise that creates a moral obligation to do the promised act rather than pay money, he may nonetheless intend his promise to be legally enforceable only by means of damages, and not specific performance. Recent work in contract theory suggests that optimal contract design includes a mix of formal (legally enforceable) and informal norms, such as morality. Contractual partners might believe that the legal obligation to pay expectation damages, combined with an exclusively moral obligation to do the promised act rather than pay damages, provides the optimal mix of legal and non-legal incentives for the promisor to perform. By legalizing all moral norms governing a contract, including the moral obligation to perform rather than pay damages, the law would prevent parties from creating the optimal mix of norms for their transaction.
Fourth, even though promise-breaking is immoral, not all breaches of a legally binding contract are morally wrong. In some cases, contracting partners might want the promisor to be morally permitted either to do X or pay an amount equal to expectation damages. One way to legalize such an agreement is to make a so-called “alternative” contract in which the promisor is given the option to do X or pay an amount equal to expectation damages. Such a promise allows the promisor to avoid breach by paying a sum of money equal to expectation damages instead of doing X. Another way to legalize such an agreement is for the promisor to make a legally enforceable promise to do X that does not legally permit him to avoid breach by paying a sum equal to expectation damages. Given that contract law provides expectation damages as a default remedy, the resulting legal obligation will legally entitle the promisee to the promisor’s performance of X or payment of expectation damages, just as the disjunctive promise does. Although the promisor will legally qualify as a breacher, and therefore will be legally required to pay damages, the parties will understand that the promisor is morally permitted to breach provided he pays damages. In these cases, in which the parties have agreed ex ante that the promisor may choose to pay damages instead of performing, specific performance is not morally justified, let alone morally required.
Of course, if most contracting parties prefer specific performance, then moral theories that embrace positive individual rights might well permit—perhaps even require—the law to respect the majority’s preferences by adopting specific performance as the default remedy for breach. But this argument for specific performance is based entirely on the moral argument for accommodating party preferences and not on the premise that breaking a promise is morally wrong. Thus, the classic moral objection to expectation damages, which claims it is inconsistent with the moral obligation to keep promises, has no force. Brooks’s moral enterprise is therefore unmotivated.
III. The Ex Post Efficiency of Brooks’s Rule
Brooks motivates his hypothetical disgorgement remedy as a modification that cures the ex post inefficiency of Schwartz’s rule:
Schwartz’s put option for promisees—that is, giving the promisee the option of compelling performance or getting expectation damages—does not generate the same allocatively efficient outcomes as [a] call option for promisors (with expectation damages). Allowing the promisee to choose between performance and expectation damages does not achieve an efficient tradeoff . . . .
Brooks argues that the promisee will be indifferent between performance and expectancy, and therefore might insist on specific performance even when breach would be efficient. He then claims that the ex post inefficiency of Schwartz’s rule can be eliminated by substituting his hypothetical disgorgement measure for the expectancy measure in its damage component.
Brooks’s argument, however, fails to take into account the implications of Schwartz’s original efficiency analysis. The centerpiece of Schwartz’s argument is his refutation of the claim that the expectancy default rule is more ex post efficient than specific performance. Schwartz’s analysis begins by distinguishing between contracts for goods or services sold in competitive or “thick” markets, and those sold in uncompetitive or “thin” markets. Expectancy is the default rule only when the goods or services promised in the contract are sold in a thick market. In thick markets, a promisor need not breach to take advantage of a more profitable alternative to performance. Instead, the promisor can devote its own resources to the more profitable alternative and still perform its original contract by purchasing and delivering equivalent goods or services from another supplier. Thus, granting the right of specific performance to the promisee will prevent the promisor from taking the superior alternative only if the transaction costs of this “cover” transaction exceed the gains from the alternative. However, Schwartz persuasively argues that the litigation costs of securing an award of specific performance normally will greatly exceed the transaction costs of covering. Thus, in thick markets, a promisee typically will cover rather than specifically enforce her agreement in return for the promisor’s agreement to pay the contract-cover price differential. Since the expectancy remedy also leads the promisee to cover following the promisor’s breach, Schwartz concludes that in thick markets both specific performance and expectancy lead the promisee to cover, and neither remedy prevents the promisor from devoting his resources to significantly more profitable alternatives to performance.
In thin markets, neither the promisor nor promisee can cost-effectively cover. Under specific performance, the promisor would negotiate with the promisee for a release in return for a share of the profits from his alternative to performance. Under expectancy, the promisee would negotiate or litigate for compensation of her expected gain from performance. Since there is no cover or market price available to measure that expected gain, Schwartz argues that these negotiation and litigation costs are likely to be at least as high as the costs of negotiation occasioned by specific performance. Thus, Schwartz concludes that expectancy and specific performance are equally ex post efficient. His case for specific performance rests instead on the claim that expectancy is systematically more under-compensatory than specific performance.
Schwartz’s analysis raises a series of objections to Brooks’s analysis. First, Brooks’s analysis fails to consider the fact that, in thick markets, the promisor can bypass the promisee’s right to specific performance by either covering or negotiating with the promisee to cover. Thus, in thick markets, specific performance at worst may lead the promisor to cover when the promisee could do so at lower cost. However, this would prevent the promisor from taking a more profitable alternative only when the gains from the alternative are less than the difference between the promisor’s and promisee’s cover costs—hardly a significant allocative efficiency concern.
Second, Brooks assumes that the only value of specific performance to the promisee is the promisee’s gain from the promisor’s performance. Yet in thin markets, specific performance also provides the promisee leverage to bargain for a share of the potential gains from the promisor’s more profitable alternative. Brooks is therefore wrong that Schwartz’s rule leaves the promisee indifferent between performance and nonperformance. The promisee will always prefer the promisor to find and exploit a more profitable alternative to performance because the promisee can use its right to specific performance to bargain for a portion of that gain. Although the cost of such bargaining sometimes might prevent the promisor from exploiting the alternative, Schwartz argues that these costs are not likely to be greater than the litigation costs created by the expectancy rule. Schwartz concludes that neither remedy is more or less likely than the other to prevent the promisor from exploiting a more profitable alternative. Schwartz’s analysis therefore undermines Brooks’s claim that Schwartz’s rule is less ex post efficient than expectancy in either thick or thin markets.
Third, in thin markets, Brooks’s modification of Schwartz’s rule is almost certainly less ex post efficient than Schwartz’s rule. Simple information economics suggests that it will cost more for the promisee to prove the value of the breach to the promisor than to prove the value of her own expectancy. If the promisee chooses the damage option under Schwartz’s rule, she must prove her own expected net profit without the benefit of a market price for an equivalent performance. Although she knows her own expected net profits, her damage claim will succeed only if she can cost-effectively prove those profits to a court. In the parlance of information economics, the promisee can observe her own profits but can recover them in a law suit only if she can verify them as well. If the promisee chooses the damage option under Brooks’s rule, however, she must prove the promisor’s expected net gain from performance as well as that of his best alternative to performance. This requires the promisee to verify information she cannot even observe, and to do it twice.
Finally, consider whether Schwartz’s and Brooks’s rules will lead to socially optimal levels of searching for alternatives to performance of thin market contracts. The socially optimal search level is achieved when the marginal social costs of searching are equal to its marginal social gains. The promisor will invest in searching for alternatives to performance until his private marginal cost equals his private marginal gain from searching. Because the promisor’s private benefits from successful searches are equal to their social benefits (i.e., the promisor fully internalizes all gains from searching), the resulting total amount of searching is socially optimal. Schwartz’s and Brooks’s rules, however, give the promisee the right to specific performance, which allows the promisee to bargain for a share of gains from any alternatives to performance and therefore divides the benefits from search between the promisor and promisee. As a result, the promisor does not fully internalize the benefits of searching and so will engage in less than the socially optimal search level.
Although it is possible that total search levels will be the same even though the incentive to search is split between the parties, two reasons suggest that the total costs of searching will be greater under specific performance than expectancy. First, the portion of the gains from a search that each party can expect will depend on the outcome of bargaining that is likely to be less costly than any bargaining occasioned by the expectancy rule. Second, the promisor is likely to be a more efficient searcher of alternatives uses for his goods and services than the promisee. Brooks doubts this is true and suggests a case in which the buyer may be better situated than his seller to identify higher valued uses for the seller’s goods or services. The question, however, is not whether a promisee might ever be better than his promisor at identifying superior options for the promisor’s goods or services but whether that is likely to be true on average. A straightforward incentive analysis strongly suggests that promisors will be more efficient at this task. Under expectancy damages, promisors fully internalize all the costs and benefits of identifying the buyer who places the highest value on their goods and services both before and after they enter into contracts. Brooks’s and Schwartz’s rules completely transfer the incentive to identify superior alternatives from the seller to the buyer after the seller contractually commits them to a particular buyer. However, even under their rules, promisors remain the only ones that have an incentive to identify the highest paying buyers before they contractually commit those resources to a particular buyer. It seems clear that the expected gains to sellers from identifying buyers for goods and services before they contractually commit them are much higher than the expected gains from identifying buyers for goods and services during the relatively brief periods during which a buyer is contractually entitled to them. Therefore, sellers will invest more heavily in acquiring the expertise of identifying their highest paying buyers.
As a fallback position, Brooks grants that the promisor may be “the more efficient identifier of valuable opportunities” but then claims that the promisor could be induced to make socially efficient search investments by modifying the hypothetical disgorgement remedy to reimburse the promisor for any reasonable expenses it incurs in searching for superior alternatives to performance. This solution misses the mark, however, because the damage component of Brooks’s rule will have no effect on the promisor’s incentives to search for superior alternatives. Once a promisor knows that the promisee has the post-breach option of specific performance, he will not breach to exploit a superior alternative to performance unless the promisee has agreed to allow him to keep some portion of the resulting gains. This holds true for any remedy that gives the promisee the right to specific performance, whether or not it also includes a damage option. Thus, no modification to the damage component of Brooks’s rule will affect the promisor’s incentives to search for alternatives to performance.Conclusion
Brooks argues that contract law has overlooked the class of possible remedies that have the form of a put option; that these remedies, unlike remedies that have the form of a call option, avoid the classic moral objection to expectation damages; and that he has identified a particular version of a put option remedy that would also meet the demands of ex post efficiency underlying the efficient breach hypothesis. I have argued that Brooks’s proposed rule has the same form as expectation damages because both provide a post-breach put option for the promisee. As a result, any moral advantage of his rule over the expectation rule must be due to the fact that it provides the option of specific performance in addition to damages, not the fact that it provides a post-breach put option. I then argued that the classic moral objection to expectation damages has no force. It erroneously assumes either that the law should enforce all moral obligations or that contracting parties always intend all of their promissory obligations to be legally enforceable. Crude legal moralism is not defensible, and with good reason parties often do not intend all of their moral promissory obligations to be given legal effect.
Nor does the moral objection to expectation damages provide a convincing argument in favor of specific performance. The moral justification of contractual remedies, like the moral justification of contractual enforcement generally, turns on the parties’ intent. When the parties agree, at the time they form their contract, that the promisor is morally permitted to commit a legal breach provided he pays expectation damages, specific performance is not morally justified, let alone required. Although there is a compelling moral argument for respecting party preferences for specific performance, this argument is based on the moral case for providing positive legal rights to vindicate party autonomy rather than on the moral wrongness of breach. The moral wrongness of breaking a promise—the heart of the moral objection to expectation damages—has no bearing on the moral justification of the default remedy for breach of contract.
I also argued that Brooks fails to appreciate the implications of Schwartz’s original efficiency analysis for Brooks’s defense of his new remedy. Brooks’s defense does not take account of Schwartz’s well-known argument that specific performance is no less ex post efficient than expectancy in both thick and thin markets. In addition, Brooks’s damage rule is likely to impose much higher proof costs on promisees than expectancy because it conditions recovery on information that is far more costly to verify. Finally, although expectancy creates optimal incentives to search for superior alternatives to performance, the specific performance component of Brooks’s and Schwartz’s rules creates suboptimal search incentives.
Jody S. Kraus is the Robert E. Scott Distinguished Professor of Law, Professor of Philosophy, and Albert Clark Tate, Jr., Research Professor at the University of Virginia. The author thanks Paul Mahoney, Alan Schwartz, Robert E. Scott, and Steven D. Walt for very helpful comments.
Preferred Citatation: Jody S. Kraus, A Critique of the Efficient Performance Hypothesis, 116 Yale L.J. Pocket Part 423 (2007), http://www.thepocketpart.org/2007/07/23/kraus.html.