Contract Law: Property Rules vs. Liability Rules and the Choice for Contract Rights, Summaries of Law

Download Contract Law: Property Rules vs. Liability Rules and the Choice for Contract Rights and more Summaries Law in PDF only on Docsity!

Specific Performance

Anthony T. Kronmant

In an important article,' Calabresi and Melamed distinguish two different techniques for protecting legal entitlements. One they call a "property" rule and the other a "liability" rule. According to Calabresi and Melamed, a right or entitlement is protected by a property rule (^) when it can be appropriated by a non-owner only if he first purchases^ permission to^ do^ so^ from^ the^ owner^ of^ the right.

2

When a right is protected by a rule of this sort, one who appropriates it without the owner's permission will always be subject to a special sanction-typically, a fine or imprisonment.' If a right is protected by a liability rule, in contrast, a non-owner who unilaterally appro- priates it need only compensate the owner, after the taking, for any loss the owner suffers.^4 The compensatory amount which a non- owner must pay for taking a right protected by a liability (^) rule is set by a representative of the state rather than by the owner of the right in a voluntary transaction between owner and taker. Calabresi and Melamed attempt to explain why some legal entitlements are protected by a property rule and others by a liabil- ity rule. They suggest that in certain cases the cost of negotiating the voluntary transfer of a right may be sufficiently high to frustrate the transfer. Where this is so, a property rule, which is intended to encourage transfers of this sort, is likely to promote an inefficient allocation of resources. This point is illustrated by automobile acci- dents and pollution torts.^5 In both cases, a voluntary transfer of entitlements is almost certain to be prohibitively expensive: in the case of an automobile accident because of the cost of identifying the victim beforehand, and in the case of pollution torts because of free-

t Assistant Professor of Law, The University of Chicago. I would like to thank Gerhard Casper, Walter Hellerstein, Edmund Kitch, Thomas Jackson, William Landes, Richard Pos- ner, George Priest, Antonin Scalia, and Franklin Zimring for their helpful comments on an earlier draft of this article. I would also like to thank Ms. Brigitte Bell, a second-year student at the University of Chicago Law School, for her valuable research assistance. I Calabresi & Melamed, Property Rules, Liability Rules and Inalienability:One View of the Cathedral, 85 HARv. L. REV. 1089 (1972). They also discuss a third technique for protect- ing entitlements-inalienability. The law restricts or forbids the sale of inalienable rights. Id. at 1111-15. For a treatment of property and liability rules from a philosophical perspective, see R. NOZICK, ANARCHY, STATE AND UTOPIA 54-87 (1974). 2 Calabresi & (^) Melamed, supra note 1, at 1092. 3 Id. at 1126. See also R. (^) NoZICK, supra note 1, at 57. Calabresi & Melamed, supra note 1, at 1692. Id. at 1108-09, 1115-24.

The University of Chicago Law Review

rider and hold-out complications which are likely to make any nego- tiated settlement enormously difficult and time-consuming. Where the costs of voluntarily transferring a particular entitle- ment are low, Calabresi and Melamed argue, economic considera- tions strongly support the use of a property rule to protect that entitlement. This argument is illustrated by the use of property rules in the criminal law: "[T]he thief or rapist... could have negotiated [a voluntary transfer of what he takes] without undue expense (at least if the good was one which we allowed to be sold at all) because we assume he knew what he was going to do and to whom he would do it." '^7 In such cases, liability rules are inappro- priate because they "represent only an approximation of the value of the object to its original owner and willingness to pay such an approximate value is no indication that it is worth more to the thief than the owner." 8 In their discussion of property and liability rules, Calabresi and Melamed do not consider one very important species of legal right: the kind of right that is created by contractual agreement, (^) the right to the performance of a promise. All of the examples in their article are drawn from the law of torts, crimes, or real property. Since contract rights have special features that distinguish them from the various entitlements created and protected by these other branches of the law, it is appropriate to ask whether contract rights should be protected by a property rule or a liability rule. In contract law, a liability rule permits a promisor to breach his promise provided he compensates the other party by payment of money damages. The fundamental alternative to money damages, in the law of contracts, is specific performance. A promise may be said to be specifically enforceable when the law gives its owner, the promisee, a right to require the actual (or "specific") performance of the promise. The right to positively enjoin a promise, like the right to negatively enjoin a nuisance, may be viewed as an entitle- ment protected by a property rule. In both cases, the owner of the right is in a position to force the would-be taker to negotiate a voluntary transfer of the particular entitlement. If the taker acts unilaterally (by simply refusing to perform, or by continuing to pollute), he can be compelled by an injunctive order to honor the owner's entitlement; and if he then refuses to honor the injunction itself, he may be forced to make a payment (not necessarily pecuni-

' Id. at 1126-27. 7 Id. at 1127. Id. at 1125.

[45:

Specific Performance

ary) to the state or the promisee greater than that required to com- pensate the promisee for his loss.^9 Moreover,^ if^ performance^ is^ still possible, a supplemental injunction mandating performance will likely issue, again backed up by civil and criminal contempt sanc- tions. If one approaches the question from the^ theoretical^ perspective developed by Calabresi and Melamed, there are two considerations suggesting that all (or most) contract rights should be protected with a property rule. First, a contract typically^ involves^ only^ two parties.'" Where only two parties are involved, the special hold-out and free-rider difficulties that^ plague^ multi-party^ negotiations^ do not arise." Second, and more important, the parties to a contract already know one another and so need not worry about the special problems of identification arising, for example, in the case of^ auto- mobile accidents. These considerations suggest that the costs of

negotiating a voluntary transfer of^ contract^ rights^ are likely^ to^ be low. Following Calabresi and Melamed, this should be^ regarded^ as a reason for protecting rights^ of^ this^ sort^ with^ a^ property^ rule.'^ 2

Instructive cases are collected in 0. Fiss, INJUNCTIONS 714-814 (1972). " This is, of course, not true in^ every case.^ Perhaps^ the^ most^ important^ exception^ is^ the third-party beneficiary contract. In this article, I ignore the complications posed by these more elaborate contractual arrangements. " This distinguishes a contract negotiation from the pollution case^ discussed^ in^ Cala- bresi & Melamed, supra note 1, at 1106-08, and from^ at^ least^ some^ eminent^ domain^ proceed- ings. For a discussion of the conflicting philosophical ideals that have informed judicial interpretation of the compensation clause, see B.^ ACKERMAN,^ PRIVATE^ PROPERTY^ AND^ THE CONSTITUTION (1977). 12 There is^ an^ additional^ consideration^ strengthening^ the^ case^ for^ protecting^ contractual entitlements with a property rule. According to Calabresi & Melamed, supra note 1, at 1108- 09, 1119, the use of a property rule in both automobile accidents and^ pollution^ torts^ is^ likely to inhibit an efficient allocation of the resources involved. It is not obvious that specifically enforcing all contractual entitlements would have similar misallocative consequences. Suppose that^ A^ contracts^ with^ B^ to^ buy^ B's^ piano.^ Suppose,^ in addition,^ that^ A^ has^ the right to specifically enforce B's promise. If C values the piano more than^ A,^ he^ will^ offer^ to pay B a premium for breaking his contract with A,^ and^ if^ the^ premium^ is^ large enough,^ B will be able to buy his way out^ of^ the^ contract, and^ the^ piano^ will^ go^ directly^ to^ C.^ Of^ course, the premium may be too small to cover both the release payment A demands and the costs of negotiating a settlement. If so, the piano will go to A, who will in turn sell it to C. Once again, the piano ends up in the hands^ of^ C,^ the^ higher-valuing user,^ but^ this^ time^ after^ two transfers rather than one. The allocative outcome is the same in both cases; the only differ- ence is a^ distributional^ one. The result, under a money damages rule, should be identical. If the difference between C's offer and the original contract price exceeds what B must pay A in damages, B will breach and the piano will go to C. However, the piano will remain in C's hands only if he values it more than^ A.^ If^ A's actual^ loss-the^ amount^ he^ would^ have^ demanded^ for^ relinquishing^ his right to B's performance in the first place-exceeds the value C places on the piano, A will now contract to buy the piano from C. Of course, this will only happen if A's money damages are undercompensatory (if they do not reflect his actual loss). But there is always a risk that the representative of the state^ who^ determines^ the amount^ of^ the^ payment^ will^ underestimate

1978]

The University of Chicago Law Review

This view, however, appears to have had little influence in shaping our law of contract remedies.^1 3 The normal remedy for breach of contract is, of course, money damages. 14 Specific perform- ance is exceptional. 5 The Anglo-American law of contracts protects most contract rights with a liability rule, only a few with a property rule. It is natural to wonder whether the peculiar mix of property and liability rules in the law of contracts can be explained on economic grounds. Although a great deal has been written about the efficiency of our law of contract damages, this more basic question has been largely ignored.' 6 The first two parts of this article argue that, in

the extent of the harm suffered by the injured party. The important point is that the piano will go to the higher-valuing user, whether or not the damages awarded the original promisee are compensatory. One might argue that under a money damages rule the cost of moving resources to their ultimate consumers will be less than what it would be if all contract rights were specifically enforceable. It is true that a promisor who must perform or pay a penalty will be more likely to attempt to buy his way out of a contract before breaching than a promisor who is only required to pay damages if he fails to perform. In some cases, this will mean an additional transaction which could be avoided under a money damages rule. In other cases, however, a property rule may prevent the transfer of a particular resource to a lower-valuing user and thus eliminate the necessity of an additional exchange shifting the resource back to the original promisee. Furthermore, the onus of a property rule might give promisors an increased incentive to carefully identify their various opportunities before committing themselves con- tractually-with the result that resources would be more likely to flow directly to higher- valuing users. On balance, it is certainly not obvious that a decision to protect contractual entitlements with (^) a property rule would increase the total cost of moving resources to their most efficient uses. ,1 See generally Barbour, The History of Contract in Early English Equity, in OXFORD STUDIES IN SOCIAL AND LEGAL HISTORY (1914); Washington, Damagesin Contract at Common Law (pts. 1-2), 47 & 48 LAw Q. REv. 345, 90 (1931-1932). " 11 S. WILLISTON, CONTRACTS § 1338 (3d ed. W. Jaeger 1968); 5 A. CORBIN, CONTRACTS § 993 (1960); Farnsworth, Legal Remedies for Breach of Contract, 70 COLUM. L. REv. 1145- (1970). ,1 The limited use of specific performance is a fairly recent historical development. See G. TREITEL, THE LAW OF CONTRACT 834-41 (3d ed. 1970); Dawson, Specific Performance in France and Germany, 57 MICH. L. REv. 495, 532, 537-38 (1959). This development has been criticized as inconsistent with natural justice. Union Pacific Ry. Co. v. Chicago Ry. Co., 163 U.S. 564, 600 (1896) (Fuller, C.J.). Not all legal systems draw the line between specific performance and money damages in the same way. See R. POUND, AN INTRODUCTION TO THE PHILOSOPHY OF LAW 240 (1922); Beardsley, (^) Compelling Contract Performance in France, 1 HASTINGS INT'L & COMP. L. REv. 93 (1977); Dawson, supra; Grossfeld, Money Sanctions for Breach of Contract in a Communist Economy, 72 YALE (^) L.J. 1326, 1333 (1963); Treitel, Remedies for Breach of Contract, in VII INTERNATIONAL ENCYCLOPEDIA OF COMPARATIVE LAW (ch. 16) 2 (1976). 16 See, e.g., Barton, The Economic Basis of Damages for Breach of Contract, 1 J. LEG. STUD. 277 (1972); Birmingham, Damage Measures and Economic Rationality: The Geometry of Contract Law, 1969 DUKE L.J. 49; Birmingham, Breach of Contract, Damage Measures, and Economic Efficiency, 24 RUTGERS L. REv. 273 (1970); Posner & Rosenfield, Impossibility and Related Doctrines in ContractLaw: An Economic Analysis, 6 J. LEG. STUD. 83 (1977); Schwartz, Sales Law and Inflations, 50 S. CAL. L. REv. 1 (1976). Professor Posner briefly

[45:

1978] Specific (^) Performance

general, the combination of property and liability rules employed in the law of contracts makes economic sense. (^) In Part III conventional explanations for the courts' refusal to enforce private contractual provisions purporting to grant the promisee a right to compel spe- cific performance are skeptically examined. In the final part of the paper, (^) I criticize judicial willingness to permit a promisor to defeat his promisee's property rule protection by transferring the promised goods or services to a good faith purchaser. It is suggested that economic considerations support the constructive trust approach to this problem adopted by some courts.

I. THE "UNIQUENESS" TEST

Specific performance is an equitable remedy 7 which a court, in its sound discretion," may grant a promisee whose money dam- ages remedy is inadequate." The situations in which courts are prepared to order specific performance are heterogeneous. Typical situations include contracts for the sale of land;" contracts for the

discusses the more fundamental (^) issue-whether there are economic consideratons that ex- plain why the law of contracts provides for specific performance in some cases and money damages in others. R. POSNER, ECONOMIC (^) ANALYSIS OF LAw §§ 4.2, 4.12 (2d ed. 1977). '1 E. FRY, A TREATISE ON THE SPECIFIC PERFORMANCE OF CONTRACTS § 3 (6th ed. 1921); J. POMEROY, A TREATISE ON THE SPECIFIC PERFORMANCE OF (^) CONTRACTS § § 1-3 (3d ed. with J. Mann 1926); 11 S. WILLISTON, CONTRACTS § 1418 (3d ed. W. Jaeger 1968). See, e.g., Klein v. Shell Oil Co., 386 F.2d 659 (8th Cir. 1967). ,8 E.g., Lee v. Crane, 270 Ala. 651, 653, 120 So. 2d 702, 703 (1960). Although certain kinds of contracts (such as (^) contracts for the sale of land) are, as a general rule, specifically enforced, courts do not feel bound by traditional categories and will sometimes exercise their discretion to deny specific performance of an agreement that would normally be specifically enforceable. See, e.g., Paddock v. Davenport, 107 N.C. 710, 12 S.E. 464 (1890) (specific performance of a contract for the sale of an interest in land denied on the ground that money damages would adequately compensate the vendee). See also cases cited at note 20 infra. ' "[We do not give specific relief ordinarily but only exceptionally where pecuniary relief is considered inadequate." R. POUND, supra note 15, at 240. Money damages are gener- ally regarded as inadequate when they are too difficult to assess. (^) See City Stores Co. v. Ammerman, 266 F. Supp. 766 (D.D.C. 1967), aff'd per curiam, 394 F.2d 950 (D.C. Cir. (^) 1968) (contract for a lease in a shopping center). The difficulties of assessing money damages are likely to be especially acute in cases involving long-term output and requirements contracts. See, e.g., American Smelting & Ref. (^) Co. v. Bunker Hill & Sullivan Mining & Concentrating Co., 248 F. 172 (D. Ore. 1918). See generally E. FRY, supra note 17, (^) at §§ 49-90; J. POMEROY, supra note 17, at §§ 28-34, 47-50; RESTATEMENT OF CONTRACTS § 358 (1932); 11 S. WILUSTON, supra note 14, at § 1418. 1 Contracts for the transfer of real property have traditionally been specifically enforced. "Where land, or any estate therein, is the subject-matter of the agreement, the equitable jurisdiction is firmly established." J. PoMERoY, supra note 17, at § 10. Not only are contracts for the sale of land specifically enforced, covenants running with the land and options to purchase land are specifically enforceable as well. See Mobil Oil Corp. v. Brennan, 385 F.2d 951 (5th Cir. 1967) (covenant running with the land); Abdallah v. Abdallah, 359 F.2d 170 (3d Cir. 1966) (option to purchase real property); McCullough v. Newton, 348 S.W.2d 138 (Mo.

1961. (sale of land).

The University of Chicago Law Review

sale of heirlooms, antiques,^2 '^ and certain^ licenses^22 and^ patent rights2 that can only be obtained from^ the^ promisor;^ contracts^ for the sale of^ a^ majority^ of^ shares^ in^ a^ particular^ corporation;^24 and long-term output and requirements contracts. 25 Occasionally,^ an

More recently, however, courts have exhibited greater willingness to deny specific per- formance of^ contracts^ for^ the^ transfer^ of^ real^ property,^ on^ the^ ground^ that^ the^ plaintiff^ has an adequate remedy at law. See Watkins v. Paul, 95 Idaho 499, 511 P.2d 781 (1973); Suchan v. Rutherford, 90 Idaho 288, 295-96, 410 P.2d 434, 443 (1966); Duckworth v. Michel, 172 Wash. 234, 19 P.2d 914 (1933). In the two Idaho cases, the court asserted that the vendee's purpose in entering the contract was not to obtain the land and put it to a specific use, but rather pecuniary profit. For a general discussion of the problem, see J. DAwsON & W. HARVEY, CASES ON CONTRACTS 180 (3d ed. 1977); Bird & Fanning, Specific Performance of^ Contracts^ to Convey Real Estate, 23 Ky. L.J. 380 (1935) (approves the tendency of modern courts to assess more carefully the adequacy of money damages in land cases). 21 Contracts involving heirlooms and antiques are specifically enforceable on the theory that the article involved typically has sentimental significance and value over and above its pecuniary worth. The classic case is Pusey v. Pusey, 23 Eng. Rep. 465 (1684), in which specific performance was granted for the transfer of an ancient horn given the Pusey family by the Danish King Canute. The horn had more than sentimental value, however. It signified con- veyance of certain realty. See also Burr v. Bloomsburg, 101 N.J. Eq. 615, 33 A. 962 (1927) (sale of a diamond ring); Falcke v. Gray, 62 Eng. Rep. 250 (1859) (sale of two china jars). The court in Falcke, while denying specific performance on other grounds, said: "In the present case the contract is for the purchase of articles of unusual beauty, rarity and distinc- tion, so that damages would not be an adequate compensation for non-performance.... 62 Eng. Rep. at 252-53. 2 Cf. Nelson v. Richia, 232 F.2d 827 (1st Cir. 1956) (sale of a business with a licensed trade name, buyer required to perform). 2 Cf. Patent & Licensing Corp. v. Olsen, 188 F.2d 522 (2d Cir. 1951) (employee ordered to assign patents on process developed in course of employment to employer); McFarland v. Stanton Mfg. Co., 53 N.J. Eq. 649, 33 A. 962 (1895) (patented improvements on a process ordered handed over to transferee of original patent); Whitcomb v. Whitcomb, 85 Vt. 76, 81 A. 97 (1912) (partner ordered to assign patent to partnership that^ had^ equitable^ ownership of the patent). See generally J. POMEROY, supra note 17, at § 20. [I]t is clear that specific performance is a particularly appropriate remedy for enforcement of "buy-sell" agreements of shares of stock of closely-held corporations. Money damages would not be adequate. It is extremely difficult to determine the value of stock in a closely-held corporation and money damages would not accomplish the primary purpose of such.. agreements-to prevent outsiders from entering the busi- ness. In re Brown's Estate, 446 Pa. 401, 409, 289 A.2d 77, 81 (1972) (footnotes omitted). Cf. King v. Stevenson, 445 F.2d^ 565,^572 (7th^ Cir.^ 1971)^ (specific^ performance^ of^ an agreement^ to^ allow the president of a corporation to purchase sufficient stock to retain control of the business); Bumgardner v. Leavitt, 35 W. Va. 194, 203, 13 S.E. 67, 71 (1891) (specific performance would be granted to defendant to prevent a takeover by antagonistic interests, therefore plaintiff is entitled to specific performance on ground of mutuality of remedy). See generally E. FRY, supra note 17, at §§ 1496-1529;^ J.^ POMEROY,^ supra^ note^ 17,^ at^ §§^ 17-19. 2 See, e.g., Laclede Gas. Co. v. Amoco Oil Co., 522 F.2d 33 (8th Cir. 1975) (supply of gas to residential subdivisions); Campbell Soup Co. v. Wentz, 172 F.2d 80 (3d Cir. 1948) (sale of carrots, specific performance denied on other grounds); Hunt Foods v. D'Odisbo, 98 F. Supp. 267 (N.D. Cal. 1951) (peaches); American Smelting & Ref. Co. v. Bunker Hill & Sullivan Mining & Concentrating Co., 248 F. 172, 182-83 (D. Ore. 1918) (lead-silver ore); Eastern Rolling Mill Co. v. Michlovitz, 157 Md. 51, 65-69, 145 A. 378, 383-85 (1929) (scrap iron); Michigan Sugar Co. v. Falkenhagen, 243 Mich. 698, 220 N.W. 760 (1928) (sugar beets);

[45:

(^19781) Specific Performance

employment 28 or construction2 contract (^) will also be specifically en- forced. The most important common feature (^) of these diverse cases is the central role played by the idea (^) of "uniqueness. ' '2 1 If the "subject

St. Regis Paper Co. v. Santa (^) Clara Lumber Co., 173 N.Y. 149, 65 N.E. 967 (1903). In many cases, the grant of specific performance is coupled with an injunction forbidding transfer (^) of the goods (^) to third parties. See generally 11 S. WILLISTON, supra note 14, at § 1419B. 26 Specific performance of employment contracts has traditionally been denied on three grounds: (1) the presumed adequacy of the (^) plaintiff's legal remedy; (2) the difficulty of supervision; and (3) the aura of (^) involuntary servitude associated with the compulsion of services. 11 S. WILISTON, supra note 14, (^) at § 1423. See, e.g., Tucker v. Warfield, 119 F.2d 12 (D.C. Cir. (^) 1941) (contract to take care of plaintiff and provide her with all the necessities of life). (^) Nevertheless, employment contracts are sometimes indirectly enforced through injunc- tions forbidding the defendant to perform similar work for anyone else. (^) Lumley v. Wagner, 42 Eng. Rep. 687 (Ch. 1852). In one interesting case, the reason given for considering (^) specific enforcement of an employment contract was that (^) the interest of the employee-who was to be paid in stocks-was inextricably bound up (^) with that of his employer. McCutcheon v. National Acceptance Corp., (^143) Fla. 663, 197 So. 475 (1940). Cf. In re Staklinski & Pyramid Elec. Co., 6 N.Y.2d 159, 160 N.E.2d 78, 188 N.Y.S.2d 541 (1959) (arbitration award granting specific (^) enforcement of employment contract against employer upheld as not contrary to public policy). See generally E. FaY, supra note 17, at §§ 110-15, 852-54; J. (^) POMEROY, supra note 17, at § 24; RESTATEMENT, supra note 19, (^) at § 379; 11 S. WILLISTON, supra note 14, at §

E.g., City Stores Co. v. Ammerman, 266 F. Supp. 766, 776-80 (D.D.C. 1967), af'd per curiam, (^) 394 F.2d 950 (D.C. Cir. 1968). Construction contracts have generally been thought to create special problems of judicial supervision. Nevertheless, construction contracts have been specifically enforced when the construction (^) materials were of a special type and could be obtained only through the builder, Rector of St. David's v. Wood, 24 Ore. (^) 396, 34 P. 18 (1893), or when it was held to be in the public interest to (^) compel completion of the construc- tion, Gas Securities Co. v. Antero (^) & L.P. Reservoir Co., 259 F. 423 (8th Cir.), cert. denied, 250 U.S. 667 (1919) (a reservoir construction project secured through public (^) bonds and par- tially completed at (^) the time of breach). See generally J. POMEROY, supra note 17, at § 23; 11 S. (^) WILLISTON, supra note 14, at § 1422A. The uniqueness test is explicitly incorporated in (^) the Uniform Commercial Code, § 2- 716(1). (^) According to the Official Comment, this section introduces a "new concept of what are 'unique' goods." Under the rule stated in § 2-716(1): Specific performance is no longer limited to goods which are already specific (^) or ascer- tained at the time of (^) contracting.... Output and requirements contracts involving a particular or peculiarly available source or market present (^) today the typical commercial specific performance situation, as contrasted (^) with contracts for the sale of heirlooms or priceless works of art which were usually involved in the older (^) cases. U.C.C. § 2-716, Comment 2. If the Comment ended there; it could be concluded (^) that the Code has merely adopted a well-established principle, and (^) given it a broader and economically more sophisticated interpretation. But the Comment continues: (^) "However, uniqueness is not the sole basis of the remedy under this section, for the relief [specific (^) performance] may also be granted 'in other proper circumstances' (^) and inability to cover is strong evidence of 'other proper circumstances.'" Id. This suggests that the draftsmen contemplated (^) a second, inde- pendent basis for awarding specific performance in particular cases. It is unclear, (^) however, what (^) this independent basis might be. The problem of construing "other proper circumstan- ces" may be avoided by reading (^) it as nothing more than a restatement of the proposition which (according to the Comment) is implicit in (^) the Code's notion of uniqueness-that uniqueness can only be determined by looking at "the total situation which characterizes (^) the contract." Id. Read in this way, (^) § 2-716(1) states only one test, not two. Although this reading treats "other proper circumstances" as (^) nothing but a clarification of the uniqueness test, it

The University of Chicago Law Review

matter of [a] contract is unique in character and cannot be dupli- cated" or if obtaining "a substantial equivalent (^) involves difficulty, delay, and inconvenience," a court will be more apt to compel specific performance. "The fact that such a duplicate (^) or equivalent cannot be so obtained does not necessarily show that money dam- ages are not an adequate (^) remedy, but is a fact that tends strongly in that direction."^3 Conversely, if the subject matter of a contract is such that "its substantial equivalent for all practical purposes (^) is readily obtainable from others than the defendant in exchange for a money payment, this fact will usually in the absence of other factors be sufficient to show that money damages are an adequate remedy for^ breach.^ ' 31

As the cases illustrate, the subject (^) matter of a particular con- tract (^) may be thought unique for a variety of reasons.^3 2 Nevertheless, courts often use the concept of uniqueness in a way which suggests that it has some relatively fixed and well-recognized meaning. An economic analysis of the law of specific performance (^) must begin with a workable conception of uniqueness. In common disc6urse "unique" means without a substitute or

is fairly consistent with the developing caselaw. See Laclede Gas Co. v. Amoco Oil Co., 522 F.2d 33, 40 (8th Cir. 1975); Kaiser Trading Co. v. Associated Metals & Minerals Corp., 321 F. Supp. 923, 932-33 (N.D. Cal. 1970). 29 5 A. CORBIN, supra note 14, at § 1142. 30 Id. 3 Id. (emphasis added). 12 See, e.g., Campbell Soup Co. v. Wentz, (^172) F.2d 80 (3d Cir. 1948) (seeds for special variety of carrots had been supplied by buyer); Huddleston"v. Williams, 267 Ala. 447, 103 So. 2d 809 (1958) (papers necessary for poodle registration); Elliott v. Hones, 11 Del. Ch. 343, 101 A. 872 (1917) (horse had particular qualities giving it promise of development (^) into a valuable race horse); Thompson v. Commonwealth, 197 Va. 208, 89 S.E.2d 64 (1955) (contract to build and deliver legislative electronic voting machines not readily available on the open market). Under the Uniform Commercial Code, "the (^) test of uniqueness... must be made in terms of the total situation which characterizes the contract." U.C.C. § 2-716, Comment 2. Thus, in De Moss v. Conart Motor Sales, Inc., 34 Ohio Op. 535,72 N.E.2d 158 (Ct. C.P. 1947), aff'd on other grounds, 149 Ohio St. 299, 78 N.E.2d 675 (1948), the court granted specific performance for the sale of a car because a manufacturing shortage had caused extensive delays. Specific performance was also held appropriate where (^) the plaintiff sought to purchase a fiberglass (^) boat manufacured only by the defendant, Gay v. Seafarer Fiberglass Yachts, Inc., 14 U.C.C. Rep. 1335 (N.Y. Sup. Ct. 1974); where the parties to the contract had stipulated that the cotton in question was unique, R.L. Kimsey Cotton Co., Inc., v. Ferguson, 233 Ga. 962, 214 S.E.2d 360 (1975); where the competitor's toner and developer for a copying machine were "distinctly inferior," Copylease Corp. v. Memorex Corp., 408 F. Supp. 758 (S.D.N.Y. 1976); and where it was unlikely that the plaintiff could locate a "substantially identical piece of used (^) equipment [transformer]... in order to perform its contract with [a] third party," Ace Equipment Co., Inc. v. Aqua Chem., Inc., 73 Pa. D. & C. 300, 302 (Ct. C.P. 1975). However, a substantial change in the price of substitutes is not itself ground for awarding specific performance. Duval Co. v. Malcolm, 233 Ga. 784, 214 S.E.2d 356 (1975); Hilmor Sales Co. v. Helen Neuschaefer Div. of Supronico, Inc., 6 U.C.C. Rep. 325 (N.Y. Sup. Ct. 1969).

[45:

Specific Performance

equivalent.Y In the framework of conventional economic analysis, however, the concept of uniqueness is troublesome. Although it might seem reasonable to define the economic uniqueness of a good in terms of its attributes or properties, this is not the definition economists employ. Economists recognize this sort of unique- ness-they call it "technological" uniqueness-but they do not de- fine the substitutability of goods in these terms. 4 For the purposes of economic theory, the substitutability of a particular good is deter- mined by observing consumer behavior, not by cataloguing the var- ious properties of the good. If an alteration in the relative price of one good affects the demand for another, then these two goods are said to be economic substitutes. The degree of their substitutability is called^ the^ "cross-elasticity^ of^ demand.^

'3 5

On this view, every good has substitutes, even if only very poor ones. Because all goods compete for consumer attention, a substan- tial change in the relative price of any good always affects (^) the con- sumption of other goods. Economists are interested in determining how great a change in the price of one good is required to effect a change of given magnitude in the consumption of certain other goods. But these are really questions of degree, resting on the under- lying assumption-fundamental to economic theory-that all goods are ultimately commensurable. 6 If this assumption is accepted, the idea of a unique good loses meaning. This point may be illustrated by a case that under present law would almost certainly be held to involve a unique good.3 7^ Suppose that A contracts with Sotheby's to purchase the handwritten (^) manu- script of Hobbes's Leviathan. If Sothby's refuses to perform- perhaps because it has a more attractive offer from someone else- A will undoubtedly be disappointed. Yet no matter how strong his affection for Hobbes, it is likely there are other things that would make A just as happy as getting the manuscript for the contract

13 See, e.g., 11 OXFORD ENGLISH DICTIONARY 235 (1961) (defining "unique" as "unequalled"). G. STIGLER, THE THEORY OF PRICE 25-26 (3d ed. 1966). Id. at 31-33. This proposition has most often been explained and defended in utilitarian terms. See, e.g., J. HICKS, VALUE AND CAPITAL 42-52 (2d ed. 1946); P. SAMUELSON, FOUNDATIONS OF Eco- NOMIC ANALYSIS 90-117 (1947). Recently, however, one economist has attempted to demon- strate that the law of the negatively sloped demand (^) curve (on which the notion of universal substitutability depends) may be "derived fundamentally from scarcity alone rather than from an assumption [of the sort made in all utility theories] that behavior is 'rational,"' G. BECKER, ECONOMIC THEORY 11 (1971). See id. at 11-23. See also Becker, IrrationalBehavior and Economic Theory, 70 J. POL. ECON. 1 (1962). 2,Cf. Lowther v. Lord Lowther, 33 Eng. Rep. 230 (1806) (terms of sale of a painting by Titian held proper issue for equity).

1978]

The University of Chicago Law (^) Review

price. For example, A may be indifferent between purchasing the manuscript at the specified price and having twenty-five hours of violin lessons for the same amount.3 If so, then A will be fully compensated for the loss he suffers by Sotheby's breach 9 upon re- ceiving the difference between the cost of twenty-five (^) hours worth of violin lessons and the contract price. However, despite the fact that the manuscript has an economic substitute, a court would be likely to order specific performance of the contract (assuming Soth- eby's still had the manuscript in (^) its possession)^4 " on the ground that the subject matter of the contract is unique. Pursuing the matter further, it is not (^) difficult to see why A's money damages remedy is likely to be inadequate and on the basis of this insight to develop an economic justification for the unique- ness test. Under a money damages rule, a court must (^) calculate the amount Sotheby's is (^) required to pay A to give A the benefit of his bargain. The amount necessary to fully compensate A is equal (^) to the amount he requires to obtain an appropriate substitute. (^) So in fixing the amount Sotheby's must pay A, the court must first determine what things A would regard as substitutes and then how much of any particular (^) substitute would be required to compensate him for his loss. 4 ' In the hypothetical case, however, (^) it would be very difficult and expensive for a court to acquire the information necessary to make these determinations. Perhaps some information of this sort would

3 The aim of compensation is to put the injured party in (^) the position he would have been in if the invasion of his legally protected interest had not occurred. So stated, the principle of compensation determines the damages for a tortious injury as well as for breach of a contractual obligation. The principle of compensation may be stated in economic terms: "Something compensates X for Y's act if receiving it leaves X on at least as high an indiffer- ence curve as he would have been on, without it, had Y not so acted." R. NozicK, supra note 1, at 57. " Of course, not every (^) imaginable loss is compensable by a payment of money. How can the loss of one's spouse or child be compensated in this way? 40 Specific performance is typically not granted when the seller is no longer in possession of the promised property. Denton v. Stewart, 29 Eng. Rep. 1156 (Ch. 1786) (house had been sold to a (^) third party for valuable consideration). "Even though the impossibility of perform- ing his contract is due to the defendant's own fault, equity will not decree that he shall do what is clearly beyond his power." 11 S. WILLISTON, supra note 14, at § 1422. But when possible, courts will sometimes grant partial specific performance coupled with an abatement in the price. See Skelly Oil Co. v. Ashmore, 365 S.W.2d 582 (Mo. 1963). If the seller has resold the property, a court will occasionally cancel the second sale in order to grant specific per- formance to the first purchaser. In Groves v. Prickett, 420 F.2d 1119 (9th Cir. 1970), the court, in order to enforce plaintiff shareholders' right of first refusal, cancelled a sale of stock to non- shareholders. See also Abdallah v. Abdallah, 359 F.2d 170, 173 (3d Cir. 1966). 1 If there are several substitutes, the court must also identify (^) the least costly one, for the party in breach should not be required to pay more than the smallest amount necessary to fully compensate the disappointed promisee.

[45:

Specific Performance

be produced by the parties. For example, A could introduce evi- dence to establish a past pattern of consumption from which the court might draw^ an^ inference^ as^ to^ what^ would^ be^ a^ satisfactory substitute for^ the^ manuscript.^ Sotheby's^ could^ then^ attempt^ to rebut the evidence and establish some^ alternative^ theory^ of^ prefer- ences and substitutes.^ But^ of^ course^ it^ would be^ time-consuming^ to produce information this way, and any inference a court might draw on the basis of such information would be most uncertain. Moreover, this uncertainty cannot be avoided by simply look- ing to the selling price of other manuscripts or even^ the^ expected resale price of the Hobbes manuscript itself (unless, of course,^ A^ is a professional dealer) .42 It would be risky to infer the value A places on the Hobbes manuscript from the value placed on^ it^ by^ others,^43 and riskier still to infer it from the value others place on the manu- scripts of, for example, Harrington's Oceana or Locke's Second Treatise. If a court attempts to calculate A's^ money^ damages^ on^ the basis of such information, there is a substantial probability that^ the award will miss the mark and be either under- or over- compensatory. Of course, if a court could accurately identify a substitute for the manuscript, it could^ disregard^ the^ fact^ that^ A^ may^ value^ the manuscript in excess of the price that he, or anyone else, has agreed to pay for it. But where it is difficult to identify a satisfactory substitute (as I assume it is here), the goal of compensation requires that an effort be made to determine the value the promisee places on the promisor's performance, as distinct from what the promisee, or anyone else, has offered to pay for^ it.

42 If A is a professional dealer, interested in reselling the manuscript, it is perhaps reasonable to treat his loss as equal to the difference between^ the^ contract^ price^ and^ the^ price at which he could have sold it to someone else at the time of B's breach (or the time^ of performance). Even^ in^ this^ case,^ however,^ money^ damages^ may^ be^ undercompensatory.^ This would be so, for example, if A did not anticipate an immediate resale but planned, instead, to hold the manucript as an investment property. 11 Suppose that A promises to pay $10 for the manuscript, but would only be willing to sell it, if he already owned it, for $15. Suppose, in addition, that B receives an offer to sell the manuscript to C for $18,^ and^ that^ C^ values^ it^ at^ $20.^ To^ fully^ compensate^ A,^ B^ should^ be required to pay him^ $5.^ But^ it^ may be^ very^ difficult^ to^ estimate^ the^ value^ A^ places^ on^ the manuscript. It will be tempting to give A the difference between the contract price and the resale price, $8, but this will^ be^ overcompensatory.^ On^ the^ other^ hand,^ if^ C^ offers^ to^ pay^ $ for the manuscript (still valuing the manuscript at $20), the same measure of damages will be undercompensatory for A. A will receive just $2 in damages,^ will^ be^ unable^ to persuade^ B to sell^ the manuscript^ for^ $15^ or^ less,^ since^ C^ will^ bid the^ price^ up,^ and^ will^ be^ unable^ to convince C-if C has obtained possession of the manuscript-to sell it for a similar amount. In neither case will A get the manuscript. Of course, if the manuscript has already been sold to a third party, a new set of considera- tions (involving the protection of^ good^ faith^ purchasers)^ must^ be^ taken^ into account.^ See^ text and notes at notes 77-96 infra.

19781

The University of Chicago Law Review

Although it is true in a certain sense that all goods compete (^) in the market-that (^) every good has substitutes"4-this is an empty truth. What matters, in measuring money damages, is the volume, refinement, and reliability of the available information about sub- stitutes for the subject matter of the breached contract. When the relevant information is thin and unreliable, there is a substantial risk that an award of money damages will either exceed or fall short of the promisee's actual loss. Of course this risk can always be reduced-but only at great cost when reliable information is diffi- cult to obtain. Conversely, when there is a great deal of consumer behavior generating abundant and highly dependable information about substitutes, the risk of error in measuring the promisee's loss may be reduced at much smaller cost. In asserting that the subject matter of a particular contract is (^) unique and has no established market value, a court is really saying that it cannot obtain, at rea- sonable cost, enough information about substitutes to permit it to calculate an award of money damages without (^) imposing an unac- ceptably high risk of undercompensation on the injured promisee. Conceived in this way, the uniqueness test seems economically sound." The (^) following case will illustrate this point. A contracts with B

for the purchase of 100 ball bearings. B breaches his promise to deliver, and A sues. If there are two or more sellers of ball bearings, and if there is substantial (^) empirical evidence indicating that the cross-elasticity of demand is very high for ball bearings offered by different sellers, a court is warranted in assuming (^) that most pur-

chasers of ball bearings regard those sold by one seller as a satisfac- tory substitute for those sold by any other. For this reason, a court may also justifiably assume that a compensatory damages payment to A which enables him to purchase ball bearings from someone other than B (without incurring costs in excess of those he had originally anticipated), is likely to put A in precisely the position he would have been in had B performed his promise. The better the evidence that most buyers regard one brand of ball bearings as a

4' See Dr. Miles Medical Co. v. John D. Park & Sons Co., 220 U.S. 373, 412 (1911) (Holmes, J., dissenting). " See generally G. STIGLER, supra note 34, at 85-89. In a perfectly competitive market, the goods offered by different sellers are assumed to be homogeneous. Even where two or more goods are not homogeneous, however, there will be many circumstances in which they may usefully be characterized as belonging to the same market or, if the term "market" is reserved for the limiting case of perfect competition, as members of a "product group." See C. FERGU- SON & J. GOULD, MICROECONoMic THEORY 315 (1975). In defining different product groups, one important factor is likely to be the volume and refinement of our information about the cross- elasticity of demand for particular goods.

[45:

Specific Performance

substitute for another, the more confident a court can be that it has

correctly calculated the magnitude of the promisee's loss.

It is^ of^ course^ true^ that^ even^ with^ very^ complete^ information

about substitutes, a court may err in calculating money damages.

It might be that A prefers B's ball bearings to those of any other

seller because he believes they will last longer. Normally this prefer-

ence would be reflected in the market price^ for^ B's^ ball^ bearings;^ if

one brand of ball bearings lasts longer than others, it should com-

mand a premium of some sort. But this may not be the case. It may

be that A's knowledge and^ experience^ with^ B's^ ball^ bearings^ are^ not

shared by other purchasers. If A attaches some special value to^ B's

ball bearings not reflected in their price, he will be undercompen-

sated if his damages are calculated on the assumption that the cost

of a substitute is the cost of obtaining ball bearings from another

seller. To prevent undercompensation, A must prove that he will

suffer a special loss as a result of B's breach, but this may be diffi-

cult and costly to establish.

The conclusion to be drawn from this analysis is a simple one.

Whenever a court calculates money damages, there is some risk that

it will undercompensate the injured party."^ But^ the^ magnitude^ of

this risk is inversely related to the completeness and reliability of

the information on which^ the^ court^ bases^ its^ award.^ At^ one^ extreme,

where there is a well-developed market^ generating^ evidence of^ sub-

stitutability, this risk is minimal. At the other extreme, where there

is no market or at most a few isolated transactions, this risk is

substantial. There is a point between these two extremes at which

the risk becomes unacceptably large (or, what amounts to the same

thing, at which the risk can only be reduced by incurring unaccepta-

ble costs). This is the point separating those contracts that are

specifically enforceable from those^ that^ are^ not-the^ point^ to^ which

the uniqueness test obliquely refers.

There is an additional, perhaps less obvious, reason why money

damages are most likely to be undercompensatory when the subject

matter of a contract is unique. In searching for a particular good,^ a

consumer will^ almost^ always^ incur certain^ costs-the^ costs^ of^ locat-

ing the good, obtaining information about it, and so on.^47 When

these costs are^ incurred^ before^ formation^ of^ a^ contract,^ they^ are

generally not compensable under a money damages rule.^4 "^ Under^ a

46 Calabresi and Melamed make this point in the context of criminal sanctions. See Calabresi & Melamed, supra note 1, at 1125. 47 G. STIGLER, THE ORGANIZATION (^) OF INDUSTRY 171-90 (1968). 43 See Chicago Coliseum Club v. Dempsey, 265 Ill. App. (^) 542, 553 (1932); 5 A. ComN, supra note 14, at §§ 992, 1034.

19781

The University of Chicago Law Review

money damages rule, a disappointed promisee (^) is usually unable to recover pre-contractual search costs, and is limited, instead, to (^) com- pensation for the increased cost of obtaining a substitute after the promisor breaches. Where a good is not unique, information acquired in searching for the (^) good is likely to have -some independent usefulness, which will survive a breach by the promisor. For example, if A wishes to purchase ball bearings, he will first obtain information about the ball bearing market. If he then makes a contract with B, which B breaches, A will of course have to obtain (^) some additional informa- tion before he can-arrange for a substitute purchase.^49 But much of the information acquired prior to his contract with B will still have value for A-it will still be economically useful. On the other hand, where A contracts for the purchase of a unique good, say, a one-of- a-kind stamp, this is less likely to be the case. A may have expended a substantial sum in locating the stamp, and although it is possible that some of the information acquired in the course of his search will be generally useful, it is more likely that much of the information is valuable to A only because it aided the discovery of the stamp in question. Under a money damages rule, A will not be compensated for these search costs if his seller breaches. But of course he will be compensated (that is, he will obtain the desired return) if he can compel specific performance of the contract. (^) That pre-contractual search costs are not compensable under a money damages rule is less worrisome when'the information generated by the search (^) repre- sents a capital stock that can be exploited in subsequent (^) transac- tions. 50 But when (^) it does not, the likelihood is increased that money damages will be undercompensatory.5 1

11 For example, A may need to locate (^) another seller. If he already knows several sellers of ball bearings, A may have to acquire only a small amount of new information before arranging a substitute transaction (looking up the telephone number of one of B's competi- tors, confirming price and delivery terms, and so forth). Before contracting with B, however, A is likely to have obtained a great deal of general information about the properties and relative advantages of different sorts of ball bearings, and this information will continue to be of use after B's breach. 10 If the information represents a capital stock of this sort, its value must be amortized over a number of individual transactions. 51 As part of a money damages award, a court could include compensation for the esti- mated cost of finding a substitute even where the subject matter of the particular contract is unique. But when the promised performance is unique, the cost of locating a substitute will be difficult to estimate, and the likelihood is rather high (^) that any court-determined award will be undercompensatory. Of course a court could simply give the promisee carte blanche to locate a satisfactory substitute on his own and then recover his search costs from the promisor. (^) This is an untenable alternative, however, for it would encourage an overinvest- ment in searching and (^) invite fraud by the promisee. See alsd note 41 supra.

[45:

Specific Performance

The uniqueness test reflects the unwillingness of courts to im-

pose a risk of undercompensation on promisees when that risk is

substantial. What justifies this reluctance? One plausible answer is

that a promisor should not be permitted to benefit from his own

misconduct by placing that risk on someone who is, after all, an

innocent victim of his breach.

This justification is attractive because it appeals to a powerful

moral sentiment. Unfortunately, when stated in its most abstract

form, it proves too much. If the fact of breach is an adequate reason

for protecting the promisee from a risk of undercompensation, it is

unclear why a promisor should ever be permitted to substitute

money damages for the actual performance of his obligation. The

moral justification is not wrong; it is merely unhelpful since it fails

to explain why some contracts are specifically enforceable and oth-

ers are not. Before a court concludes that it would be wrong to

impose a particular risk on the promisee, it should first determine

that the promisee has not agreed to bear the risk, nor been compen-

sated for doing so. The moral justification for specific enforcement

presupposes a solution to this initial problem of risk allocation.

A second, essentially economic, justification for the uniqueness

test consists in showing that the test draws the line between specific

performance and money damages in the way that most contracting

parties would draw it were they free to make their own rules con-

cerning remedies for breach and had they deliberated about the

matter at the time of contracting. If this is true, the uniqueness test

promotes efficiency by reducing the costs of negotiating contracts."

In general, this way of thinking about the rules of contract law

requires consideration of the ex ante interests of parties engaged in

a hypothetical bargaining process, struggling with a problem of ra-

tional choice under conditions of uncertainty. As I shall attempt to

show in the next section, an analysis based upon ex ante considera-

tions does suggest, if only somewhat tentatively, that contracting

parties would be more likely to provide for specific performance

where the subject matter of their contract is unique, and for money

damages where it is not.

II. SPECIFIC PERFORMANCE AND THE Ex Ante Interests of Promisor

and Promisee

When would the parties to a contract freely agree to a judicially

52 See generally R. PosNER, supra note 16, at 69.

1978]

The University of Chicago Law Review

enforceable 3 provision giving the promisee an option5 4^ to specifically enforce the other party's promise? Other things equal, a promisee will always prefer to have such a provision included in the contract for it gives him an additional right which he would not otherwise possess. Other things equal, a promisor will always prefer a contract without such a provision-a contract, in other words, which he may unilaterally breach on the condition that he make a subsequent compensatory payment to the promisee. Consequently, a promisee intent upon writing a specific performance provision-a property rule-into the contract will have to pay to secure the promisor's consent. Similarly, a promisor must make a payment of some sortI^5 in order to exclude a provision for specific enforcement from the contract. If and only if the benefit which the promisee realizes from a specific performance provision exceeds the cost of the provision to the promisor will the provision be included in the final contract. When the subject matter of a contract is unique, the risk is greater that the promisee's money damage remedy will be under- compensatory." Since a right to compel specific performance re- duces this risk, promisees-as a class-should be willing to pay more for a provision giving them a right of this sort when there is no developed market generating information about the value of the subject matter of their contract. However, if a specific performance provision is likely to be more beneficial to a promisee when the subject matter of his contract is unique, it is also likely to be more costly to his promisor under the same circumstances. In the first place, a right in the promisee to compel specific performance increases the probability of costly ne- gotiations for transfer of the promisee's contract rights.^57 This of

13This is a counterfactual assumption.^ See^ text^ and^ notes^ at^ notes^ 62-69^ infra. " The provision should be thought of as an option because a promisee who is entitled to specific performance may always forego this right and pursue a money damages remedy instead. Although the payment will probably (^) assume the form of a straightforward reduction in the contract price, it may be made in other ways as well (for example, by the inclusion of warranty or delivery terms more favorable to the promisee). 0 See text and notes at notes 41-51^ supra. 51 Of course, the^ inclusion^ of^ a^ specific^ performance^ provision^ will^ not^ make^ a voluntary transfer of the promisee's contract rights inevitable. The promisor may always breach without having first negotiated his release. On the whole, however, it is more likely that the promisor will attempt to buy his way out of the contract before breaching under a specific performance rule than under a money damages rule. See text and notes at notes 1-9 supra. One complication not discussed in this article concerns post-breach negotiations. Under either a money damages rule or a specific performance rule, negotiations of this sort are likely to occur if both parties think it in their best interest to avoid litigation. However, the parties may be more likely to conduct post-breach negotiations (rather than litigate or do nothing) under one rule than under the other. This will depend upon the predicted costs of negotiating

[45:

Specific Performance

course always reinforces the promisor's preference for a money dam- ages rule. However, a promisor is likely to regard this reason as especially compelling where the subject matter of his contract is unique, since the lack of information about substitutes will almost certainly make the parties' negotiations longer and more compli- cated and thus more costly. Second, if the promisee is entitled to specifically enforce the promisor's obligation, the promisor who wishes to breach will have to make a release payment to the promisee and buy his way out of the contract. The amount of the release payment demanded (^) by the promisee will be greater than what the promisor would have to pay the promisee under a money damages rule." This is so whether or not the subject matter of the contract is unique. But the difference between what the promisee would accept in exchange for a release and what he may be expected to receive under a court-administered money damages rule is likely to be larger where the subject matter of his contract is unique, because the risk that court-awarded dam- ages will be undercompensatory is greater. For these two reasons, a specific performance provision will be more expensive to the promi-

sor when the subject matter of his contract is unique. Thus far, it would appear that the benefits to the promisee and the costs to the promisor of a specific performance provision are proportional; both are greater when the subject matter of the con- tract is unique. There is, however, an additional consideration influ- encing their ex ante deliberations that provides some basis for thinking that the parties to a contract will be more likely to provide for specific performance when the subject matter of their agreement is unique. The cost of a specific performance provision to the promisor will be determined, in part, by his own estimate of the likelihood that he will want to breach the contract. If he fully intends to perform,

a settlement and the anticipated benefits of litigation in each case. These problems are sufficiently complicated to warrant separate treatment, but I do not believe they affect the basic soundness of my argument. 11 The amount of the release payment will also be greater than the benefits the promisee expects from performance. There will, of course, always be a ceiling on what the promisor will agree to pay the promisee. This ceiling will be determined by two things: the penalty the promisor will incur if he breaches without having purchased a release from the promisee, and the amount he stands to lose if he performs the contract. The promisor will never pay the promisee more than the lesser of these two amounts for a release. The amount which the promisor actually pays the promisee will fall somewhere between the maximum payment he is prepared to make, and the minimum payment the promisee is prepared to accept (assuming, of course, that the former exceeds the latter-if it does not, there can be no negotiated release). This price will be determined primarily by the bargaining skills and relative informational advantages of the parties.

1978]

The University of Chicago Law Review

and thinks breach unlikely, (^) a promisor will be less hostile to a con- tract (^) with a specific performance provision than he would otherwise be. (^) One important factor influencing the promisor's thinking in this regard is the probability that he will receive a better offer for his goods or services in the interim between formation of the contract and performance. The higher the (^) probability, the greater the likeli- hood he will want to breach. The probability of receiving an attrac- tive alternative offer may be especially low (^) where the subject matter of the contract is unique." In this case there is by definition no developed market, transactions are spotty at best, and therefore (^) a promisor will often justifiably think (^) it highly unlikely that he will receive any alternative offer (let alone a better one) for the promised goods or services. Indeed, where the subject (^) matter of his contract is genuinely unique, a promisor may estimate the likelihood of a preferable alternative (^) offer as close to zero, and thus be nearly indif- ferent as to what remedies (^) the promisee will enjoy in the highly unlikely event of breach. Although the promisor (^) thinks breach highly improbable, the promisee may not. Despite (^) the promisor's insistence that he intends to perform, the promisee may be (^) skeptical. As long as he is anxious about the promisor's performance, the promisee will be concerned about the adequacy of his own remedies, (^) and where the subject matter of his contract is unique he will (^) likely have a decided prefer- ence for a contract that gives him the right to specifically enforce the other party's promise. Consequently, in (^) the case of a contract for a unique good (^) or service, the benefits the promisee derives from a specific performance provision are apt (^) to outweigh its costs to the promisor, who, free of doubts about his own reliability, may regard the inclusion of such (^) a provision as a relatively costless way of enticing the promisee to enter the contract on (^) advantageous terms. In the case of a contract for non-unique goods or services, by contrast, the existence of a developed market increases the likeli- hood that the promisor will receive alternative offers before he has performed the contract. The promisor will therefore be anxious to retain the freedom and flexibility (^) enjoyed under a money damages rule. Moreover, the promisor will be especially anxious in this case

' This conclusion does not ineluctably follow from the definition of uniqueness. There may be great demand for a good in short supply. For example, (^) there may be several collectors willing to pay more than the contract price for a manuscript of Hobbes's (^) Leviathan. In the case of currently produceable goods, however, widespread demand for a good should lead to increased production of the good. So (^) in the typical commercial context, the assertion seems plausible.

[45:

Specific Performance

to avoid the additional transaction costs that would be incurred if

he had to negotiate a voluntary transfer of the promisee's contract

rights. Although these costs will tend to be smaller where the sub-

ject matter of the contract is not unique, they can never be less than

some fixed minimum (the cost of contacting the promisee, notifying

him of an intention to breach, obtaining a release statement of some

sort, and so on). Where there is an established market in the goods

or services involved, prices will ordinarily be grouped rather closely

around a single point. The probability is therefore greater that any

alternative offer the promisor does receive will not be sufficiently

high to cover the cost of negotiating a release plus the amount he

must pay the promisee for the release. Thus the likelihood increases

that a promisor who has agreed to a specific performance provision

will find himself in the undesirable position of having to decline an

alternative offer that he would accept under a money damages rule.

In some cases the alternative offer will cover the release payment

but will be refused solely because the transaction costs of negotiat-

ing a transfer of the promisee's contract rights are prohibitively

high. The promisor should therefore be willing to make a small

payment to the promisee, perhaps in the form of a slightly reduced

contract price, in order to exclude a specific performance provision

and thus avoid these potential transaction costs. Because the trans-

action costs avoided by the promisor would not have benefited the

promisee, the latter will be better off with the reduced contract price

if he regards the risk of undercompensation under a money damages

rule as minimal. The promisee will generally regard this risk as

slight where there is a developed market generating information

about suitable substitutes.

In sum, promisors and promisees will typically favor a money

damages rule if the subject matter of their contract is not unique.

When the contract is for unique goods or services, on the other hand,

the benefit to the promisee of a specific performance provision is

likely to be substantial and the promisor may well regard his own

breach as only a remote possibility, so the opposite conclusion seems

more plausible. There is thus some basis for believing the unique-

ness test reflects the typical solution that contracting parties would

arrange for themselves in light of their ex ante interests.

IlI. SPECIFIC PERFORMANCE AND FREEDOM OF CONTRACT

I have argued that ex ante considerations provide some basis

for thinking the uniqueness test is economically rational. 0 The argu-

41 See text and notes at notes 52-59 supra.

1978]

The University of Chicago Law Review [ 5

ment attempted to show that if contracting parties were entirely free (^) to select their own remedies, their choice would generallly corre- spond to the mix of property and liability rules adopted by the law of contracts. But ex ante arguments (^) for the efficiency of a particular legal rule assume that individuals remain free to contract around that rule, and a legal system that denies private parties the right to vary rules in this way will tend to be less efficient (^) than a system that adopts the same rules but (^) permits contractual variation." This raises an important question. (^) If, under the uniqueness standard, a promisee does not have the right to specifically enforce a particular promise, can he create a right of this sort by private agreement?" It is easy to imagine situations in which a promisee might wish to do this. For example, A may wish to enter a specifi- cally enforceable contract for the purchase of ball bearings (^) because he believes that special considerations, which will be difficult to establish in a lawsuit, are likely to make his normal money damages remedy undercompensatory. Similarly, (^) some sellers of ball bear- ings may be willing to agree to (^) a specific performance provision in order to secure a better contract price or a new customer. For ex- ample, a new entrant in an industry, lacking an established reputa- tion, may conclude that agreeing to a specific performance provi- sion is the least costly way to communicate to prospective clients his confidence in his ability to perform. In the well-known case of Stokes v. Moore,"' the plaintiffs, part- ners operating a small loan business, hired and later dismissed de- fendant Stokes. They then sued to enforce a covenant not to com- pete, which had been incorporated in the contract of employment. The contract contained a clause stating that if Stokes breached his promise not to compete, "a restraining order or injunction [might] be issued and entered against [him] in any court of equity jurisdic- tion."6 4^ Although it affirmed the trial judge's decision to issue a temporary injunction, the Supreme Court of Alabama viewed unfa- vorably the parties' attempt to create a right to injunctive relief by

private agreement:

We do not wish to express the view that an agreement for the issuance of an injunction, if and when a stipulated state of facts

', Posner & Rosenfield, Impossibility and Related Doctrines in Contract Law: An Eco- nomic Analysis,, 6 J. LEG. STUD. 83, 89 (1977); R. POSNER, ECONOMIC ANALYSIS OF LAW § 4. (2d ed. 1977). 62 Clearly a promisee of a contract that would be specifically enforceable can agree to give up his right to compel performance. 63 262 Ala. 59, 77 So. 2d 331 (1955). " Id. at 61, 77 So. 2d at 334.

[45:

Specific Performance

arises in the future,^ is^ binding^ on^ the^ court to^ that^ extent.^ Such

an agreement would serve to oust the inherent jurisdiction of

the court to determine whether an^ injunction^ is^ appropriate

when applied for and to require its^ issuance^ even^ though^ to^ do

so would be contrary to the opinion of the court. 5

The court added, however,^ that^ "the^ provision^ for^ an^ injunction^ is

important in its influence upon an exercise of the discretionary

power of^ the^ court^ to^ grant^ a^ temporary^ injunction."^6

The few cases directly addressing this question reach a^ similar

result. 7 A^ private^ agreement^ that^ purports^ to^ give^ one^ party^ the

right to specifically enforce the promise of another will be given

some weight by courts in deciding^ whether^ to^ grant^ injunctive^ relief.

But no court will consider itself foreclosed by the parties' contract

from refusing specific relief. A contractual provision^ accompanied

by a lengthy description of those aspects of the transaction that

make specific performance desirable is^ likely^ to^ carry^ more^ weight

than a provision unadorned by supporting explanation. But in no

event will the contract provision^ prevent^ a^ court^ from^ independently

determining the appropriateness of injunctive relief.^69

Perhaps judicial unwillingness to honor provisions such^ as^ the

one in Stokes reflects a desire to avoid private abuse of a powerful

and intrusive remedy. This is a legitimate concern.^ But^ if^ the pur-

pose in scrutinizing a^ private^ agreement^ of^ the^ Stokes^ variety^ is^ to

prevent abuse by an overreaching promisee, this end could be served

as adequately and more directly by other legal tools-for example,

,s Id. at 64, 77 So. 2d^ at^ 335. I8 Id. 7 See generally McNeil, Power of Contractand Agreed Remedies, 47 CORNELL L.Q. 495, 520-23 (1962). Professor McNeil collects^ the^ principal^ cases^ in^ notes^ 88-92^ of^ his^ article.^ See also Denkin v. Sterner, 10 Pa.^ D.^ &^ C.^ 2d^203 (Ct.^ C.P.^ 1956)^ (refusing to^ enforce^ a^ contrac- tual provision purporting^ to^ give^ the^ seller^ of^ refrigeration^ equipment^ the^ right^ on^ buyer's default price). to^ confess^ judgment^ against^ the^ buyer^ for^ the^ full^ amount^ of^ the^ unpaid^ contract pc (^) See Peters, (^) Remedies for Breach (^) of Contracts (^) Relating to the Sale (^) of Goods (^) Under the Uniform Commercial Code:^ A^ Roadmap^ for^ Article^ Two,^^73 YALE^ L.^ J.^ 199,^252 (1963). " This is more frequently (^) expressed by saying that specific performance is a discretion- ary remedy, and that "the right to specific performance is^ not^ absolute,^ like^ the^ right^ to recover the legal judgment." POMEROY, supra^ note^ 17,^ at^ §^ 35.^ See^ also^ id.^ at^ §^ 46.^ It^ is^ well established that "[n]either party to a contract can^ insist,^ as^ a^ matter^ of^ right,^ upon^ a decree for its^ specific^ performance."^ Snell^ v.^ Mitchell,^^65 Me.^ 48,^^50 (1876).^ Only^ if^ a^ promise^ would be specifically enforced in the absence of a contractual provision^ purporting^ to^ give^ the promisee the power to enjoin its performance,^ will^ a^ court^ compel^ the^ promisor^ to^ do^ what he initially agreed to do^ and^ not^ permit^ him^ to^ substitute^ money^ damages:^ "If^ one who contracts to render personal service agrees that^ in^ case^ of^ breach^ the^ remedies^ of^ specific performance and imprisonment shall be available to the employer, the agreement would not be effective." 5 A. CORBIN, supra note 14, at § 1432.

1978]

The University of Chicago Law Review

by traditional common law doctrines of fraud, duress, and good

faith. 70 If the concern is abuse of the contracting process, courts

should focus on the voluntariness of the parties' agreement.

It may be, however, that courts prohibit the private creation of

injunctive remedies not because specific performance provisions

evidence some procedural unfairness in the parties' dealings, but

rather because they are perceived to be substantively unacceptable

limitations on personal freedom. A provision of the kind involved in

Stokes might be viewed as a modified contract of self-enslavement,

an attempt to transfer an entitlement whose transfer is prohibited

by law (an "inalienable" right or entitlement in the scheme pro-

posed by Calabresi and Melamed). 71 This idea is echoed in some of

the older specific performance cases involving construction and

employment contracts.^

72

Such an argument carries little weight in a case like Stokes,

where the promise to be enforced is a negative one-a promise to

refrain from doing something. More importantly, the argument is

overdrawn. It is true that certain forms of domination (for example,

slavery and peonage) are regarded as inherently bad. Our legal sys-

tem prohibits these forms of domination, whether they are created

by consensual act or by force. On the other hand, there are many

relations of domination recognized and protected by law so long as

they are voluntarily established and maintained. The relation cre-

ated by a contract of employment is an important example of legally

protected domination.

The nature, completeness, and duration of self-imposed limita-

tions on personal freedom determine their legal and moral accepta-

bility. Slavery is objectionable largely because it involves near-total

control. By contrast the domination an employer exercises is partial

and limited-the employer only controls certain aspects of his em-

ployee's life. Nevertheless, employees are not generally required by

judicial order to submit to employer control. The judicial order, it

may be argued, makes a crucial difference: if the employment rela-

tion is created or maintained by the threat of judicial sanctions, it

is almost certain to be plagued by acrimony and ill-will. But al-

though the unpleasantness of a forced employment relation should

certainly be taken into account by an employer contemplating a suit

for specific performance, it should not be a basis for refusing to

impose such a relation upon parties who have agreed to an injunc-

7' See Epstein, Unconscionability:A CriticalReappraisal, 18 J. LAw & ECON. 293 (1976).

7 See Calabresi (^) & Melamed, supra note 1, at 1111-15. 72 See 11 S. WILLISTON, supra note 14, at § 1423.

[45:351

Specific Performance

tive provision in their contract. Moreover, if the party in breach anticipates that the relation will be unbearable, he can buy his release from the contract. Judicial insistence that the specific enforcement of certain con- tracts would create an objectionable form of personal servitude is made yet more puzzling by the numerous cases in which courts have been perfectly willing to negatively enjoin the party in breach from employing his^ time^ or^ talents^ save^ in^ performance^ of^ the^ contract.

7 3

This sort of decree will often have the same effect as a positive injunction to perform. There is another common explanation for the reluctance of courts to enforce private injunctive agreements: the specific enforce- ment of contracts (especially employment and construction con- tracts) entails special administrative costs which normally can be avoided under a money damages rule, and private individuals should not be allowed to shift the special costs associated with this form of relief to the taxpayers who subsidize the legal system. 74 The assumption on which this argument rests, however, may be mis- taken. It is ancient dogma that specific performance necessarily means increased judicial involvement in the enforcement and super- vision of contractual duties. This might be true, but so might the opposite conclusion: if all promises were specifically enforceable, or if private parties were permitted to contract into a specific perform- ance rule at their discretion, a resulting increase in the voluntary transfer of contract rights might lower the number of breaches-and perhaps even of lawsuits-and in this way reduce the actual involve- ment of courts in contractual relationships. In comparing the administrative costs, broadly defined, of a property rule mandating specific performance with a liability rule directing an award of money damages, the following factors must be

73 See, e.g., (^) Philadelphia Ball Club v. LaJoie, 202 Pa. 210, (^51) A. 973 (1902). 11 See, e.g., Marble Co. v. Ripley, 77 U.S. 339, 358-59 (1870); Northern Delaware Indus. Dev. Corp. v. E.W. Bliss Co., 245 A.2d 431 (Del. Ch. 1968); Edelen v. Samuels, 126 Ky. 295, 306-07, 103 S.W. 360, 363 (1907). Northern Delaware, which denied specific performance of a "massive, complex, (^) and unfinished construction contract," has been vigorously (^) criticized. Comment, Specific Performance of Construction Contracts-Archaic Principles Preclude Necessary Reform, 47 NOTRE DAME LAW. 1025, 1029-33 (1972). In recent years, courts have exhibited greater willingness to grant specific performance in situations in which the remedy has traditionally been disfavored on the grounds that it is either too intrusive or too difficult to administer. See In re Grayson-Robinson Stores, Inc. & Iris Constr. Corp., 8 N.Y.2d 133, 168 N.E.2d 377, 202 N.Y.S.2d 303 (1960) (enforcement of an arbitration award ordering specific performance of a contract to construct a portion of a shopping center); In re Staklinski & Pyramid Elec. Co., 6 N.Y.2d 159, 160 N.E.2d 78, 188 N.Y.S.2d 541 (1959) (enforcement of arbitration award ordering employer to reinstate a discharged employee).

1978]

The University of Chicago Law Review

taken into account: (1) the frequency of litigation under both rules; (2) the administrative cost of resolving litigated disputes under both rules; (3) the portion of the administrative costs involved in dispute resolution borne by the parties and the portion borne by society under both rules; (4) the likelihood and cost, under each rule, (^) of pretrial settlement; and (5) special institutional costs, such as the potential loss in court prestige that results from non-compliance with a direct order to perform, and the cost of invoking the court's contempt powers. To say the least, it is unclear how these administrative costs add up under the two rules. Until these questions have been ex- plored, it is unwise to assume that the legal (^) invalidity of private agreements purporting to (^) create a right of specific enforcement is justified on the ground that specific relief is a more costly remedy than money damages and that a greater portion of the costs of specific performance is borne by third parties. A promisee could attempt to achieve (^) indirectly the special pro- tection associated with a property rule by insisting, as a condition of entering the contract, that the other party promise to pay a pen- alty upon breach-a penalty in excess of any damages payment necessary to compensate the promisee. A provision of this sort, if enforceable, would have the same effect as the sanctions that back up both injunctive orders and criminal prohibitions: it would en- courage the prospective taker of an entitlement to purchase the entitlement (^) from its owner in a voluntary market transaction. How- ever, a contractual provision that is deliberately designed to be penal will not be enforced.^7 5 The legal prohibition of penal clauses is an important obstacle facing parties who wish to contract into a

71 The enforcement of penal provisions has been carefully discussed, from an economic point of view, in a recent, and excellent, article. See Goetz & Scott, Liquidated Damages, Penalties, and the Just Compensation (^) Principle:Some Notes on an Enforcement Model and a Theory of Efficient Breach, 77 COLUM. L. Rlv. 554 (1977). Goetz and Scott argue that the "penalty rule" (^) (the rule that a liquidated damages provision will not be enforced (^) if it is penal in nature) is economically unsound since "its uncritical application frequently induces a costly re-examination of the [parties'] initial allocation of risks and may also deny the non- breaching party either adequate compensation for the harm caused by the breach or the opportunity to insure optimally against such harm." Id. at 556. They suggest that "the modern development of unconscionability" offers "a less costly alternative to the sweeping invalidation powers exercised under the penalty rule." Id. at 594. In their attack on the penalty rule, Goetz and Scott emphasize the importance of what they call "non-compensable idiosyncratic value." I have described the same phenomenon in this article as the "risk of undercompensation." See text and notes at notes 41-46 supra. The analysis of penalty clauses Goetz and Scott advance in their article complements, in many ways, my treatment of specific performance, and I find myself in general agreement with both their theoretical conclusions and proposals for change.

[45:351

Specific Performance

property rule. 6 In sum, although the parties to a contract may have legitimate reasons for wishing to contract into a property rule, the courts will

"' Hostility to the use of penal clauses might be based in part upon the conviction that breaking a promise and committing (^) a crime are qualitatively different wrongs and that the legal system (^) ought to prevent private individuals from obliterating the line between them. This conviction is more puzzling than it first appears. Crimes and breaches of contract are both invasions of legally protected interests. In both cases the taker typically knows in advance whose right he is invading and has the time to negotiate a voluntary transfer of that right. Given this important similarity, one might conclude that the right to the performance of a promise and the right to be free from criminal attack should be protected in the same manner. I have argued that (^) a mix of property and liability rules appears to be the best way of protecting contract rights. See text and notes at notes 37-59 supra. The criminal law could adopt a similar mix: crimes involving (^) the appropriation of a unique good (for example, rape or murder) would be punished by imprisonment or fine while crimes involving the taking of a non-unique good (for example purse-snatching) would give rise only to a claim for compen- sation. Of course, the criminal law does not employ this approach-nearly all rights to be free from crime are protected by a property rule. Focusing on the similarity between breaking a promise and committing a crime conceals differences of a more important kind, differences which explain why property rules are used more sparingly in contract law than in criminal law. To explore these differences fully is beyond the scope of this article, but a few suggestive comments can be made. The widespread use of property rules in the criminal law might be explained on the ground that criminal takings not only harm the victim, but inspire fear in the community. Cf. Michelman, Property, Utility and Fairness:Comments on the Ethical Foundationsof "Just Compensa- tion" Law, 80 HARv. L. Rlv. 1165, 1214 (1967) (discusses "demoralization" resulting from injury inflicted by "deliberate social action"). To some extent, however, general confidence among promisees is undermined whenever a contract is broken. See Llewellyn, What Price Contract?-An Essay in Perspective, 40 YALE L.J. 704, 725 n.47 (1931). Of course the costs to the community of a crime may be much greater than the costs of a breach of contract because a crime is usually thought to be a more alarming invasion of rights than breach of a contractual obligation. Many crimes exhibit three features which help explain this perception. First, a criminal may well avoid detection, but a promisor will rarely be able to conceal his breach from the promisee. For this reason, the punishment for a crime must be increased to reflect the risk of non-apprehension, but damages for breach of contract need not be likewise inflated. See generally Becker, Crime and Punishment: An Economic Approach, 76 J. POL. ECON. 169 (1968). (^) Moreover, even if apprehended, a criminal may have disposed of the property ac- quired in the crime and is likely to be insolvent, Calabresi & Melamed, supra note 1, at 1125 n.69; thus compensation from the criminal is often impossible. In the second place, breach of contract rarely entails a risk of physical violence; a crimi- nal taking often does. Most people are likely to view their own life and physical well-being as paradigmatic examples of unique goods. Potential victims of crime are thus justifiably concerned that they may suffer a noncompensable (^) loss. Third, the relation between promisor and promisee is almost always voluntarily estab- lished. In contrast, most crimes can be committed by strangers. A taking (^) which occurs in the context of a voluntarily created relationship often seems less offensive than one which does not. Perhaps this (^) is because we believe an individual has greater control ovr the risks to which he is exposed in the former situation than he does in the latter; perhaps because a taking by a stranger is thought to involve some additional harm to the special set of interests defined by the elusive concept of privacy. These three features distinguish most serious criminal takings from breaches of con- tracts. They also help explain and justify the widespread use of property rules in the criminal

1978]