The promised machine has acquired a second claimant. Before delivery, another buyer offers more for it, and the seller begins to reckon what it would cost to disappoint the first. The calculation is easy to condemn and harder to dismiss. If the machine can do more useful work elsewhere, something may be gained by changing its destination. Yet the first buyer has already arranged his affairs around a claim that was meant to be settled. A new opportunity for the seller is not, by itself, a new permission to dispose of that claim. The question of breach begins where these two facts meet: performance can cease to be the best use of resources, while the promisee can still have a right that the better use does not extinguish.
The remedy gives the promise a practical shape, but it does not exhaust its meaning. In his 1897 account of the common law, Oliver Wendell Holmes asked what someone indifferent to morality would predict the courts would do. A promise seen from that vantage could appear as a choice between performance and a compensatory payment. Holmes himself separated that inquiry from a judgment about how one ought to live, and expressly distinguished the cases in which equity would compel obedience to an order. His method directs attention to consequences that solemn language can obscure. It does not establish that every profitable breach is morally innocent, or that all the law requires in every case is a check. The price of disobedience and the justification for it remain different questions.
Expectation damages attempt to protect the value the agreement gave the disappointed party. The reference point is the position performance would have produced, with its costs as well as its receipts. A buyer who would have obtained something worth one hundred and thirty by paying one hundred has an expected gain of thirty; if the price has not been paid and nothing is delivered, thirty restores that gain in the simplest monetary account. Giving him the full one hundred and thirty while also excusing the price would count the same hundred twice. This modest subtraction matters. The economic argument depends on compensating the actual loss from nonperformance, not on attaching an arbitrary fine to the broken word or confusing the promised thing’s value with the net benefit of the bargain.
A constructed transaction shows what full compensation can accomplish. Let the first buyer value an existing machine at one hundred and thirty and agree to pay one hundred on delivery. A second buyer values it at one hundred and eighty and offers one hundred and sixty. Ignore transaction costs, risk and every effect outside these three parties; assume that money measures each party’s gain and that damages can be paid at once. Performance gives the seller one hundred and the first buyer thirty. Breach followed by a payment of thirty gives the seller one hundred and thirty, leaves the first buyer with the same thirty, and gives the second buyer twenty. Their combined gain rises from one hundred and thirty to one hundred and eighty. The machine reaches a higher-valued use, and the stipulated payment protects the first buyer’s stipulated interest. These assumptions are the substance of the result.
A profitable breach can instead be a way of transferring an unrecognized loss. Suppose the first buyer’s true value is two hundred, because the machine is unusually well suited to a use that the damages assessment misses. Keep the award at thirty and the second buyer’s value at one hundred and eighty. The seller still gains thirty by diverting the machine. The first buyer, who would have gained one hundred from performance, receives thirty and loses seventy; the newcomer gains twenty. Combined value falls by twenty. Nothing about the seller’s incentive signals that reversal, because his account includes the award rather than the whole loss. Costs of delay or collection can create a similar separation. Calling an award compensatory does not make its amount equal to every interest that performance would have served.
The legal measure itself needs a jurisdiction and a scope. California’s Civil Code section 3300 states a general measure of compensation for detriment proximately caused by contractual breach or likely to result in the ordinary course, subject to other express provisions of the Code. That is a legal standard for adjudicating a claim, not a promise that every disappointed expectation will be converted into money. The circumstances, the proof and the governing rules matter. Nor does a higher price elsewhere demonstrate the value of the first buyer’s loss: price and value were deliberately different in the machine example. A theory that needs exact compensation must face the difficulty of establishing it, rather than take the name of the remedy as evidence that the difficulty has disappeared.
Specific performance can preserve the promisee’s choice without preventing a valuable reallocation. Return to the first version of the example, in which the machine is worth one hundred and thirty to its buyer. If he can obtain it and then sell it to the newcomer for one hundred and sixty, the machine reaches the same final use. The original seller receives one hundred, the original buyer receives sixty after paying the price, and the newcomer receives twenty. Total value is still one hundred and eighty; the additional thirty now belongs to the first buyer. An enforceable claim to delivery has changed who can bargain over the opportunity. When bargaining and resale are feasible, compelled delivery need not imprison the machine in its first use. When they are costly or unavailable, the comparison changes again.
A promise to produce something can present a different problem from a promise to convey what already exists. If a contractor discovers that making the promised object will consume resources worth more than the object’s value, insisting on production can destroy a gain that a release would preserve. The parties may negotiate that release, but negotiation uses time and may fail. A damages rule that accurately charges the contractor for the buyer’s loss can make avoiding the waste privately attractive without requiring a new bargain. The alternatives matter: if another producer can fulfill the same obligation at an ordinary cost, the original contractor’s difficulty need not make performance wasteful. The comparison must include obtaining that substitute performance. Nor does the production-cost argument carry unchanged into a case where the object already exists and the dispute concerns who gets the profit from selling it.
The remedy also shapes the agreement made before either party knows what will happen. An assurance of delivery can support an investment the buyer would otherwise refuse to make; an open-ended exposure to extraordinary production costs can make a seller demand a higher price or decline the work. The parties may prefer different allocations of that risk, and a payment that seems small after a loss may have been part of a cheaper bargain beforehand. Even so, a hypothetical bargain is not a record of actual consent. The analyst must distinguish an arrangement both parties could have preferred from the arrangement they understood themselves to have made. Otherwise the imagined agreement becomes a convenient way for the party who benefits from breach to supply the missing permission himself.
The moral question survives a successful calculation because a promise can give another person a claim over one’s conduct. Charles Fried grounds that claim in the capacity of free people to undertake obligations and enlist one another’s reliable cooperation. In revisiting his argument, he also recognizes that the morality of promising cannot dictate every detail of the law’s remedial machinery. That leaves room for agreed options, releases and sensible adjustments without making obligation disappear. A seller who obtains the buyer’s informed release has recognized the buyer as someone entitled to answer. A seller who merely predicts an affordable award may have avoided that encounter. Whether the difference is justified depends on the promise, the changed circumstances and the loss actually borne, not solely on the seller’s gain.
A payment can replace performance in an account. Whether it answers the promise depends on what the account has counted.
The broken promise is worth neither whatever the seller happens to earn by breaking it nor necessarily whatever a court can most readily measure. Its worth begins with the undertaking and the position it gave the other party reason to expect. Damages, performance and renegotiation are different ways of protecting or revising that position, each with costs and possibilities the others lack. The useful economic insight is that a change of performance can create value; the indispensable qualification is that the gain must be distinguished from a loss shifted onto someone else. The next plate turns from an obligation broken to an obligation deliberately left open: the employment agreement, whose promised performance includes a later decision about what the work shall be. The argument remains open.