Concept
Intangible and Remote Losses
Certainty of damages
Intangible and Remote Losses
Reasonable certainty is tested one element of loss at a time, and some elements give the claimant almost nothing to test it with. Section 352 of the Restatement (Second) of Contracts permits recovery only up to the amount of loss that the evidence permits to be established with reasonable certainty, and comment a adds that the rule “merely excludes those elements of loss that cannot be proved with reasonable certainty.”1 A claim can therefore lose one element and keep the rest.
The standard restates an ordinary burden of proof. The breach victim must prove its loss, so the evidence must persuade the factfinder that the loss more likely occurred than not and must give the factfinder enough basis to calculate a monetary award.2 The clearer the proof that some injury occurred, the harder the factfinder will work to find a figure. With no plausible and relevant evidence of amount, it cannot award damages at all.3
Three kinds of loss strain that standard hardest: injury to a business's good will, income from books and recordings, and losses that would have arrived only through events outside the contract or that exist only as the claimant's private satisfaction. Each asks the factfinder to predict what absent people would have done, whether customers, readers, listeners, or prospective employers, or to put a price on something no market prices. Every added contingency gives the factfinder one more thing to guess at and gives the defendant one more place to attack the estimate.
No category is barred by its name. Section 352 asks the same question of each: whether usable evidence ties a real loss to the breach and lets the factfinder fix an amount rationally. The three subsections that follow apply that question in turn, beginning with good will, which sits closest to the lost-profit claims already covered because its effects, if they appear anywhere, appear in the business's own dealings with customers.
Good Will
Good will is the value a business draws from its reputation and from its customers' readiness to keep buying, and comment a to § 352 names it as an item for which “great precision cannot be expected.” Damages need not be calculable with mathematical accuracy, and they are often at best approximate.4 That relaxation concerns the size of the figure only. The claimant must still prove that customers were lost because of the breach.
Good will is hard to prove because of the way it is lost. A lost sale shows up in the seller's books: a particular buyer refused a particular delivery at a known price. Good will is lost through later decisions by customers who saw defective goods or unreliable service and bought elsewhere or bought less, and they do not always say why. The injury is real, since a business whose dealers have soured on it earns less, but its size depends on what many people outside the lawsuit would have done. The certainty requirement is accordingly likely to be particularly troublesome for a claimant seeking recovery for loss of business reputation.5 A court faces two risks. Refusing all recovery leaves a genuine injury uncompensated, and accepting a bare assertion of damaged reputation invites guesswork. The modern compromise accepts approximate proof and insists on evidence that ties the lost business to the breach.
Courts once met this difficulty with a flat refusal. For many years Pennsylvania denied recovery for loss of good will because “damages of this nature would be entirely too speculative,” as in Harry Rubin & Sons v. Consolidated Pipe Co.6 A rule of that form decides the claim as a matter of law, so the claimant never reaches the stage of offering sales records or customer testimony. In 1990 Pennsylvania joined the overwhelming majority of states in AM/PM Franchise Ass'n v. Atlantic Richfield Co., a gasoline franchisee's claim for loss of good will after its franchisor breached a warranty, and overruled the earlier cases “to the extent they prohibit a plaintiff from alleging a claim for damage to good will as a matter of law.”7
Sales of goods followed the same course. Most courts traditionally refused to allow lost profits from loss of customer good will under the Code, reasoning that damage to business reputation is too speculative. Several courts, including the Southern District of New York applying New York law, now recognize that a buyer may, in appropriate circumstances, show injury to its reputation with sufficient certainty.8 The change moves the fight from the rule of law to the evidence. Good will is no longer excluded by its nature, though proving such damages can still be considerably difficult.9
Illustration 4 to comment a shows what usable proof looks like. A manufacturer sells plastic to a wholesaler, who resells it to dealers. The plastic proves defective, many dealers complain, and some refuse to place further orders. The wholesaler can recover for lost good will if the loss can be estimated with reasonable certainty by such evidence as its business records from before and after the transaction and the testimony of its salespersons and of dealers.10 Each kind of evidence does separate work. The records show the drop in ordering that followed the shipment, which gives the factfinder a baseline and a measured decline to estimate from, much as the past performance of an established business supports a prediction about its future.11 The testimony supplies the link the records cannot. A decline in orders has many possible explanations, and dealers who say they stopped buying because the plastic failed connect the decline to the breach. Neither kind of evidence yields an exact figure, and the comment does not ask for one. It adds that courts have grown more receptive to sophisticated economic data and expert opinion, so a claimant with good records may also call an expert to analyze them.12
A claim that is available in principle can still fail for want of that link. In Hangzhou Silk Import and Export Co. v. P.C.B. International Industries, Inc., an American start-up contracted to buy garments of its own design from a Chinese manufacturer, several shipments proved defective, and the buyer sought damages for harm to its reputation and good will.13 The court accepted that such damages can be recovered but denied them. The buyer had not shown that the cancellation of several of its accounts was caused by the defects: it offered no testimony from disaffected customers and too little other evidence to prove its case.14 The denial came even though the seller had committed fraud by telling one of the buyer's principals that a plant belonging to another company was its own.15
The wholesaler of Illustration 4 and the start-up in Hangzhou Silk share a skeleton: defective goods, a buyer that resells, and customers who stopped ordering. They differ in what each could put before the factfinder. The wholesaler had before-and-after records and dealers who could say why they stopped ordering, while the start-up had cancelled accounts and nothing that tied them to the garments. Decisions of this kind are typical, because the certainty principle cannot be separated from the quantity and quality of the plaintiff's evidence, and many cases decided under the certainty label could as easily be described as a failure to prove damages, causation, or the case as a whole.16
The fraud did not rescue the start-up, and that result fits the principle taught earlier in this chapter. Resolving doubts against the breaching party puts the risk of imprecision about amount on the wrongdoer once a significant loss is established, but it supplies no proof that a loss occurred or that the breach caused it.17 A buyer seeking consequential damages under the Code must establish causation, foreseeability, reasonable certainty as to amount, and that mitigation doctrines do not bar the claim.18 Lost accounts that cannot be tied to the defects stall at causation, before amount is ever reached.
That split result follows from comment a's statement that the requirement merely excludes those elements of loss that cannot be proved with reasonable certainty. The accounts that are proved recover, and the one that is not proved does not pull the others down.19 An exam stem that pairs defective goods with a claim for damaged reputation therefore calls for more than the observation that good will is recoverable today. The answer turns on the evidence in the stem: whether business records and testimony from customers or salespeople link the lost accounts to the breach, and whether the record gives a rational basis for estimating how much business was lost.
Royalties and Artistic Works
Royalties depend on how many strangers decide to buy a work, and public taste cannot be checked until the work meets its audience. A wholesaler's lost-profit claim can be assembled from the prices and costs in the wholesaler's own records, but a lost-royalty claim for a book that never appeared must be assembled from a prediction about readers. The certainty requirement accordingly operates with particular severity on books, movies, and, by analogy, records.21
Comment a supplies the plainest example. A publisher contracts to publish a novel, repudiates the contract, and the author cannot get the book published elsewhere. If the evidence does not permit the author's loss of royalties and of reputation to be estimated with reasonable certainty, the author cannot recover for that loss, although nominal damages remain available.22 The illustration assumes a record with nothing a factfinder could compute from, because a novel that was never published has sold no copies and built no readership to measure. It therefore states the consequence of missing proof and leaves open what happens when proof exists. The comment also notes that a claimant whose profit evidence fails can often recover the amount actually spent in reliance, which is usually easy to prove, and the section on when profits cannot be proved develops that fallback.23
Royalties are not speculative by their nature, and the conditional wording of Illustration 1 reflects that. If the author establishes the fact of damage, perhaps through testimony that people would have bought the book, a court or jury can generally determine damages even though the amount is uncertain.24 That is the two-step structure taught earlier in this chapter applied to one kind of income. The claimant first shows that the breach cost it something, and the factfinder then receives latitude about how much.25 Creative works make the first step harder than usual, because nothing about a book or a record guarantees that anyone would have wanted it. Evidence from the market itself can lift a claim over that step, and one decision shows both what such evidence looks like and where it runs out.
In Contemporary Mission, Inc. v. Famous Music Corp., a group of Roman Catholic priests who wrote musical compositions sued a record company for breaching its contract to make and sell records, on a royalty basis, from the master tape of their rock opera “Virgin.”26 The Second Circuit acknowledged the requirement's particular severity, yet held that the trial court erred in excluding a statistical analysis and expert testimony about how successful the opera's most successful single, “Fear No Evil,” would have been.27 The state of the market at the time of the breach explained why: “the record was real, the price was fixed, the market was buying and the record's success, while modest, was increasing.”28 Even after promotion ended and the record was withdrawn from the marketplace, its own momentum carried it to about 10,000 additional sales and lifted it from roughly number 80 to number 61 on Billboard's “Hot Soul Singles” chart.29 Facts like those make the fact of loss plain, and they give an expert observed sales and a chart trajectory to extrapolate from, so the defendant can test the projection against the same data.
The same plaintiffs, suing on the same breach, also pressed what the opinion calls a “domino theory”: had “Fear No Evil” become a hit, it would have generated opportunities for concert and theatrical tours and similar benefits.30 The court rejected that claim because those additional benefits were “too dependent upon taste or fancy to be considered anything other than speculative and uncertain.”31 Each domino is a separate contingency. The single must first become a hit, which is itself a prediction. Promoters must then book the group, and audiences must then buy tickets, and every step depends on the preferences of people who are not before the court. The chance that all of them fall the right way shrinks with each added step. Nothing had yet happened in the market for tours that an expert could extrapolate from, so the record-sales projection reached forward from observed behavior while the tour claim reached forward from a hypothetical in which behavior changed and then changed again. The doubts-against-the-breacher principle does not rescue the tours, because it shifts only the risk of imprecision about the amount of a loss already established as significant, and it supplies no proof that the tours would have existed.32 The two halves of the decision also show the rule working element by element: one element reaches the factfinder with expert support, and the other is excluded without disturbing it.
Illustration 1 and the tour claim fail for different reasons. The unpublished novel never reached the market, so no sales trend or buying public existed to project from, and that is the ingredient the single supplied. The tours lacked something else: an unsupported chain of further taste-driven events layered on a loss that was itself still hypothetical. A royalty claim therefore fails when the market has said nothing and when it depends on a long chain of what the market might yet have said, and it succeeds to the extent the market has already spoken in numbers.
A stem about lost royalties is asking where the work stands in the market. A work already selling at a fixed price with a rising trend supports statistical and expert proof of its direct losses. A work that never reached the market supports little beyond nominal damages and reliance expenditures unless some other evidence of demand exists. A claim that adds further layers of taste-dependent benefit, such as concert tours, must prove the first layer and then each added one, and it usually fails at the added layers.
Remote and Subjective Losses
Two kinds of loss defeat the certainty requirement even when the breach is plain. A remote loss would have come into being only if a series of events outside the contract each turned out favorably, with other people making the decisions along the way. A subjective loss has its value in the claimant's private satisfaction, which no market prices and no witness can translate into dollars. The harm may be entirely real in the claimant's experience, and the element still fails, because the standard asks what the evidence will support and a sincere sense of injury does not show the injury's size. Because § 352 applies element by element, a claimant who loses one such element can still recover others that the evidence supports.34
The simplest failure is a loss with a name and no number. In ESPN v. Office of the Commissioner of Baseball, a broadcaster breached its contract with major league baseball by preempting scheduled Sunday-night games in favor of NFL games, and the commissioner claimed that the league had lost television exposure, prestige, and opportunities for promotions and sponsorships.35 The commissioner offered no evidence of how much the league lost, and the court dismissed the damages claim as too speculative. A court that sees a clearly harmed claimant will do what it can to make an award, but it needs some foundation for a reasonable estimate and cannot construct a figure out of the air for a speculative or intangible loss.36 The case shows the second step of the fact-then-amount sequence taught earlier in this chapter. The latitude courts give over amount presupposes some relevant evidence to reason from, and exposure and prestige come with no price list.
The league in that scene and the commissioner in ESPN claim the same kind of injury, and only the proof differs. A sponsor's refunded fee is a figure created by a contract and recorded before anyone thought of suing, so the broadcaster can test it against the same documents. Exposure and prestige have no counterpart, because no document says what a viewer's impression of the league was worth. The element-by-element rule applies here as well: a league with sponsor records recovers that element even if its separate claim for general prestige stays too speculative to reach the jury.
ESPN failed for want of any number, and Central America Health Sciences University v. Norouzian failed because the number depended on events that had not happened. A medical student completed his studies, but the school refused to grant a diploma over a dispute about unpaid tuition, and the student established at trial that the school was not justified in withholding it.37 The delay pushed the start of his career as a doctor back by about three years, and the trial court awarded roughly $400,000 for lost income. The appellate court overturned the award as too speculative, because the student could not show that, had the diploma arrived on time, he would have passed the medical boards and been hired by a hospital as a resident, so his anticipated salary was subject to contingencies that made the claim too uncertain.38
The breach was established and the delay was real. What the evidence could not supply was each link between the diploma and the salary. Passing the boards depends on an examination, and being hired depends on a hospital's choice among applicants, so each link rests on a decision by someone other than the parties that nobody had observed. A figure built from several such links is only as probable as the whole series, and the probability that every link holds shrinks as links are added, which is the same logic that defeated the concert-tour claim in the record case. Proof that closed the links, such as a passed board examination or a signed offer from a hospital, would have shortened the chain and given the factfinder something to compute from.
An older case compresses the same reasoning into a few lines. In Newsome v. Western Union Telegraph Co., the plaintiff had told the defendant's operator that he needed whisky in order to get his rafting done, and the court refused to hold the defendant to damages that “from the very nature of the case must be purely speculative and remote.”39 Damages, the court said, are not to be based on mere conjectural probability of future loss or gain, and “[s]omething more than a possible result must appear.”40 The plaintiff's warning about his purpose went to what the defendant knew. It did not cure the speculation about what that purpose would have produced.
A long time horizon multiplies the contingencies. In Kenford Co. v. County of Erie, New York's highest court reviewed an award of $25.6 million in lost profits under a 20-year management contract and held those damages unrecoverable on two separate grounds. The county's liability for the profits was not within the parties' contemplation when they contracted, and the damages were too speculative to meet the requirement of proof with reasonable certainty.41 A twenty-year contract forces the claimant to prove demand, prices, and costs for every year of the term rather than for one.
Evidence can close a link that looks open. One claimant's contract contained a sixty-day termination provision, so profits beyond sixty days depended on the claimant's not ending the relationship anyway. The claimant's president testified, in essence, that he regarded the relationship as a long-term one and that performance problems were the most important reason for his decision to terminate. The court found that testimony, though not absolute, showed the relationship in all probability would have continued long after termination had the defendant performed adequately, and it declined to conclude that the claimant would have exercised its termination right for some other reason.42 A possible contingency does not defeat a claim when testimony shows it probably would not have occurred.
The subjective version of the problem appears in cosmetic-surgery disputes, which mark the line between private and measurable value. If the claimant's expectation was an improved appearance or greater comfort, its value is private. If the damaged feature was a productive instrument, such as a hand an artisan used in daily work, the expectation may relate to future earning capacity, which is a measurable quantity and not merely a matter of comfort and appearance.43 Earnings have a market and a record, while the comfort of looking better has neither, so a claimant who offers no evidence of earnings has nothing for the factfinder to compute from.
These claims often travel with arguments about foreseeability and causation, and the decisions above show the overlap. Kenford rested on two grounds, the parties' contemplation and certainty, and in Newsome the plaintiff's telling the operator why he needed the whisky did not make the defendant answerable for damages that were purely speculative and remote.44 A later section compares certainty with those two limits. For present purposes, certainty is an independent hurdle. A buyer seeking consequential damages under the Code must clear causation, foreseeability, and reasonable certainty as separate elements,45 so a type of loss can be foreseeable when the parties contracted and caused by the breach, and the evidence at trial may still be unable to fix its amount. The doubts-against-the-breacher principle does not repair a failure at this stage. It shifts only the risk of imprecision about the amount of a loss already shown to be significant, and it supplies neither the missing links in a chain of contingencies nor a market value for private satisfaction.46
A claimant who loses an element at this stage is not necessarily left with nothing. The next section takes up the two fallbacks, reliance damages and nominal damages.