Concept
Anticipated or Actual Harm
Liquidated Damages or Penalty
Anticipated or Actual Harm
Restatement (Second) of Contracts § 356(1) and UCC § 2-718(1) measure a stipulated sum against the harm a breach causes, and both state that harm in the disjunctive: the "anticipated or actual loss" in the Restatement, the "anticipated or actual harm" in the UCC.1 "Anticipated" points to the forecast the parties could make when they signed. "Actual" points to the loss the breach in fact produced. Because "or" joins the two, a court may test the sum at contracting, at breach, or at both. The text does not say which view controls when they disagree, and courts fill that gap in different ways.
The views can disagree because every forecast is made under uncertainty. A sensible estimate can sit far above the loss the breach later causes, and a careless one can land close to it. Each view also guards a different value. A rule that looked only to actual loss would make the injured party prove the loss to validate the clause, which is the expense the clause was written to avoid. A rule that looked only to the forecast would enforce a large payment for a breach that cost nothing, which strains the compensation principle behind the penalty rule.
Three problems follow, and the subsections below take them in order. The first is how far a clause can rest on what the parties reasonably anticipated at signing. The second is whether the loss that actually followed can save a clause the forecast would not support, and whether it can also sink one the forecast would support. The third is the hardest case: a reasonable forecast followed by a breach that caused no loss at all. Any answer should begin by identifying which vantage point the governing jurisdiction adopts, because one clause can pass under one view and fail under the other.
Forecast at Formation
Under the forecast-at-formation view, a court asks what the parties could reasonably expect a breach to cost when they signed, and a sum that was a reasonable forecast on that date stays valid whether the later loss runs higher or lower. Lawyers call this the ex ante view, a judgment made before the events it predicts. The ex post view looks back from the breach and belongs to the next subsection. The word "anticipated" in both standards supports the ex ante reading.2
The forecast carries this weight because of what a forecast is. When the parties contract, they know the transaction and can weigh its risks, but the harm a breach will cause lies in the future, and the loss later proved will almost certainly land above or below the stipulated figure.3 A rule demanding a match would doom every clause, since the uncertainty that led the parties to write it would also guarantee that it never matched. The inquiry therefore rests on the factors the parties objectively contemplated at formation.4 A rule keyed to later events would also defeat the clause's purpose. The parties fix the amount in advance to avoid proving damages, and a clause that survived only when the proved damages came close to it would send them to trial on the very question they meant to remove.
Maryland states the formation view plainly. Its older cases treat a sum named in the contract in place of uncertain damages, incapable of exact ascertainment, as liquidated damages unless the sum is grossly excessive and out of all proportion to the damages that might reasonably have been expected. They judge that question from the contract, its subject matter, and the circumstances known to the parties at signing.5 The rule cuts both ways. If damages later prove larger than the stipulated sum, the promisee cannot collect more, because the parties agreed that the named sum would stand in place of damages.6 Absent a statute providing otherwise, Maryland courts look to the stipulated loss at formation and not to the losses that followed the breach.7
Barrie School v. Patch shows the rule at work. Parents enrolled their daughter in a private school for the 2004-2005 year under an agreement that required a $1,000 non-refundable deposit plus $13,490 in tuition, allowed cancellation by certified letter received before May 31, 2004, and made the parents liable for the full year's tuition as liquidated damages if they withdrew after that date.8 They withdrew late. The Maryland Court of Appeals upheld the clause as a reasonable forecast of just compensation, neither grossly excessive nor out of all proportion to the harm that might have been expected when the agreement was made.9
The school's budgeting explained why its loss could not be pinned down in advance. Its chief financial officer testified that the school builds its budget in November and December of the preceding year by totaling the expenses of instructing the students, then setting a tuition level and enrollment number that meet its revenue goal, so that isolating the effect of one student is very difficult.10 The circuit court agreed that assigning an exact amount to one child's departure would be next to impossible.11 Because the budget spreads expenses across the whole student body before the year opens, one family's departure does not obviously reduce what the school must spend, and a year's tuition was a sensible estimate of what a late withdrawal would cost. The same facts thus did double duty: they showed that the loss was hard to measure, and they gave the parties a principled number to use anyway.
The parents answered that the school had admitted another student in their daughter's place and so lost little or nothing. The court rejected the argument. Because the forecast was reasonable when the contract was made, it did not matter that actual damages came in below the prediction.12 And because the clause was valid, the school had no duty to mitigate.13 An argument that asks what the school really lost skips to a different test, since the formation view asks only whether the figure was a sensible forecast at signing.
The forecast approach reaches beyond consumer contracts. After an acquisition agreement is signed, one party may refuse to close, and corporate codes let the agreement preserve the board's power to abandon the merger even after a shareholder vote.14 Shareholders may also vote the deal down, and the delay of a vote may let new bidders appear.15 A termination fee prices those risks. In Brazen v. Bell Atlantic Corp. the Delaware Supreme Court treated a $550 million termination fee in a merger agreement as liquidated damages and upheld it as a reasonable estimate rather than a penalty.16 The fee was around 2 percent of the corporation's market capitalization, Delaware Chancery decisions had approved fees of 2 to 3 percent, and the parties had analyzed the costs they incurred in setting it, counting not only the out-of-pocket expense of negotiating the deal but also the opportunity cost of dealing only with the other merger party.17 The court's contract test asked whether the damages were uncertain and whether the amount agreed upon was reasonable and not punitive.18 Each feature gave the forecast an anchor outside the parties' say-so: a market benchmark, an itemized cost analysis, and a recognized category of harm, lost opportunities, that is real but hard to prove at trial.
A defending party shows a reasonable basis for the amount by pointing to market conditions, industry custom, prior dealings, or other risk factors, and a clause with such a basis is typically upheld under principles of freedom of contract.19 Barrie's budget and Brazen's benchmark and cost analysis are examples. A formula with no visible tie to loss falls short. In Wasserman's Inc. v. Township of Middletown a lease measured the lessee's recovery for the township's termination by a percentage of the lessee's gross receipts. The New Jersey Supreme Court observed that gross receipts do not account for the lessee's ordinary expenses or for the expenses the breach causes, so it could not tell whether the figure reflected damages likely to flow from a breach or was an arbitrary number.20 It remanded for a determination of reasonableness and placed the burden of production and persuasion on the township, because stipulated damages clauses are presumptively reasonable.21 The lesson holds under the formation view: a forecast needs a visible tie to the loss. Tuition in Barrie was the revenue the school's budget was built to collect, while gross receipts say nothing about what a lessee loses when a lease ends.
Courts divide on whether the forecast is the only look. In Kelly v. Marx, the Massachusetts court allowed the sellers of a house to keep the buyer's five percent deposit as liquidated damages and rejected a second look at the actual loss. It reasoned that a single look matches the parties' expectations and resolves disputes without waiting for proof of damages, while a second look undermines the certainty the parties bargained for and invites the aggrieved party to offer evidence of harm. The court held to that view even though the sellers suffered no loss from the breach.22 Maryland's rule in Barrie is the same kind of single-look approach, but it rests on the common law and yields to statute; the court noted that UCC § 2-718(1) speaks of anticipated or actual harm for sales of goods.23 The New Jersey court in Wasserman's described the modern trend as assessing reasonableness either at formation or at breach.24 Whichever vantage point a court adopts, it first asks whether the clause was a genuine attempt to settle damages in advance, and it treats the clause as a penalty if it instead punishes or deters breach by imposing liability beyond reasonably anticipated loss.25 The next subsection takes up what actual loss can add.
Actual Loss After Breach
The word actual in both standards gives a court a second vantage point beyond the forecast the parties made at signing.27 Judging the sum by what the breach in fact cost is the ex post view, the counterpart of the formation view. Its simplest effect is a rescue: a sum that was not a careful forecast can still stand when it proves reasonably close to the actual loss, because an enforceable clause need satisfy only one of the two tests.28 A sum that tracks the real loss compensates the injured party instead of punishing the breaching one, and punishment is what makes a clause a penalty.29
Because reasonableness can be assessed at two moments, a clause can land in four positions. The sum may be reasonable at both moments, unreasonable at both, unreasonable at signing but reasonable in hindsight, or reasonable at signing but unreasonable in hindsight.30 The first two need little analysis. Reasonableness at both moments is strong evidence of a fair bargain, and unreasonableness at both is strong evidence of the unfairness that unconscionability reaches. The standard does real work only when the views come apart. The third position is the rescue just described. The fourth asks whether hindsight can condemn a clause that foresight approved, and courts answer that question differently.
The rescue follows from the disjunctive wording. A sum that wildly exceeds the harm anticipated at signing is valid if it is reasonable in light of the harm that actually followed, so a clause that looked excessive when written may stand if the breach proves serious enough.31 The UCC speaks of harm, not of the damages a court would otherwise award, so the comparison can include losses that the ordinary rules of foreseeability and certainty would exclude.32 What counts as actual harm is itself unsettled. One defensible reading includes all harm that mitigation could not have avoided, whether or not damages law would have compensated it.33
The rescue has limits. Some courts apply an intent test: a clause that looks excessive at signing shows that the parties meant a penalty, even when the clause captured the real loss well. That approach can condemn clauses the disjunctive reading would save, and courts could instead impute unusual foresight to the parties.34
Authorities divide on whether actual loss can only save a clause or can also sink one. The disjunctive wording supports the first answer, because a sum that need satisfy only one of the two views should stand once it passes either. The Ninth Circuit read the text that way in California & Hawaiian Sugar Co. v. Sun Ship, Inc., enforcing a clause that yielded more than $4 million as a reasonable forecast although the actual harm was apparently under $400,000, and stressing that the parties were sophisticated and the extent of harm was difficult to measure. The Seventh Circuit in XCO International, Inc. v. Pacific Scientific Co. likewise treated a clause as enforceable if reasonable in light of actual harm even when it was not a reasonable forecast.35 On that reading, actual loss is an extra route to enforcement and never a route to defeat.
Wassenaar v. Panos reflects the opposite answer. The Wisconsin court explained that damages grossly disproportionate to the actual harm usually lead courts to conclude that the parties' original expectations were unreasonable, so facts available at trial can undo a clause that looked reasonable at signing.36 The New Jersey court relied on that reasoning and added that hindsight is frequently better than foresight, so a court assessing the genuineness of a pre-estimate cannot help being influenced by what it knows of later events.37 A court that checks actual loss is protecting the oldest concern of the penalty rule. The doctrine grew from equity's practice of relieving defaulting obligors from forfeiture of the penalty due under conditioned penal bonds, and that practice led to the modern distinction between a penalty and liquidated damages.38 An award far above the real harm is a forfeiture however sensible the forecast looked.
Two statements about the second moment show why the answer turns on which reading a court adopts. One holds that adding a second moment clearly works toward less frequent enforcement, since a clause that once faced a single test must now survive another look.39 The New Jersey court observed the opposite, that judging a clause at either the making or the breach encourages more frequent enforcement.40 Each statement is accurate for the rule it assumes. If the two moments are alternatives, a second chance to pass makes enforcement more frequent. If actual loss is an added hurdle, the second moment makes enforcement less frequent. Courts that let actual loss only save a clause favor certainty, because the parties know that a sensible forecast will hold. Courts that let it also sink a clause favor compensation, and they accept some uncertainty in return.
Barrie puts the sink question on one set of facts. The school had enrolled more students than its budget projections called for, so mitigation had in effect already occurred, although no evidence of mitigation was presented.41 The dissent would have weighed all the surrounding circumstances, those existing at contracting and at breach, including the damages actually sustained, and it objected that enforcing the clause manifestly overcompensated the school.42 The majority declined to take the school's slight loss into account, because the clause was a reasonable forecast and a valid clause fixes the damages.43 The facts were the same for both opinions. What differed was whether actual loss could condemn a clause that was reasonable when made. UCC Article 2A closes the question for leases of goods, because it speaks only of "anticipated harm" and so gives a lease no ex post route.44 The extreme version of the sink question, a reasonable forecast followed by a breach that caused no loss at all, is the subject of the next subsection.
When No Loss Occurs
Restatement (Second) of Contracts § 356(1) permits liquidated damages only at an amount reasonable in light of the anticipated or actual loss and the difficulties of proof of loss, and UCC § 2-718(1) states the same test for sales of goods.46 The hardest test of that wording is a breach that caused no loss under a clause that was a sensible forecast at signing. The clause passes the formation view, because the parties priced a real risk in good faith. It fails the compensation principle, because the promisee would collect for a harm that never happened. The text does not say which consideration wins, so courts divide. On an exam, identifying that split matters more than any single outcome.
The Restatement answers through the interaction of its two factors. The anticipated or actual loss and the difficulty of proving it combine to decide whether a sum is unreasonably large: great difficulty of proof allows considerable latitude, slight difficulty allows little, and where it is clear that no loss at all has occurred, a provision fixing a substantial sum is unenforceable.47 The reason follows from what a liquidated damages clause is for. Parties liquidate so that they need not prove a loss that is hard to measure, and the clause earns deference to the extent it spares that proof. When the loss is zero and the records show it, nothing remains to be spared, and the only job the clause can still perform is to deliver a payment that compensates nothing, which is the punishment the penalty rule forbids. The Restatement's racetrack illustration shows the result. A builder completes a grandstand nine days late under a clause of $1,000 per day, but the owner could not have opened the track sooner because permission to operate arrived a month late. The delay cost the owner nothing, and the clause is unenforceable because that absence of loss is easy to prove.48 Had the owner's loss been hard to prove, the clause would have had room to operate, so difficulty of proof is the variable that decides the case.
That route works only because the Restatement measures difficulty of proof after the breach. When the parties signed, they could not know whether the permit would arrive on time, so any loss from delay was uncertain. By trial that uncertainty is gone, and the owner's lost revenue is a matter of records. The Restatement therefore examines difficulty of proof at the time of trial and not at the time of contracting, which departs from the earlier prevailing view that uncertainty of damages is tested at contracting.49 The proper moment remains disputed, which is one more reason the no-loss case produces conflicting decisions.
Garden Ridge, L.P. v. Advance International, Inc. reached the same result in a sale of goods by a different route. The buyer's evidence showed that it earned the same profit on the nonconforming goods as it would have earned on conforming ones, so the buyer had lost nothing, and the Texas court of appeals refused to enforce the clause.50 To get there, the majority described the UCC test as asking whether the sum was unreasonable in light of the anticipated or actual harm, which in effect changed the statute's or into and and made actual harm a second hurdle.51 A concurring judge objected that this reading contradicted the plain text, under which the sum may be reasonable based on either anticipated harm or actual harm.52 The objection has force, and the clause could have failed on firmer ground. Section 2-718(1) also requires reasonableness in light of the difficulties of proof of loss, and a buyer with no loss presents no difficulty at all, so the sum fails the combined test as the Restatement analyzes it, with no need to rewrite the statute.53
Courts that judge reasonableness only at formation reach the opposite answer on the same fact. In Barrie, parents who withdrew a child after the deadline argued that the school suffered no harm because it filled their daughter's place. The Maryland Court of Appeals rejected the defense, reasoning that it would negate the benefit of a stipulated sum and depart from the principle that reasonableness is judged from the time of contract formation.54 Kelly v. Marx let sellers keep a deposit though they suffered no loss.55 Under the traditional approach of assessing enforceability solely at contracting, a total absence of loss may be irrelevant, as in Frick Co. v. Rubel Corp., and Wallace Real Estate Investment, Inc. v. Groves treats actual damage as no prerequisite to upholding a clause while allowing unconscionably disproportionate actual damage to bear on the reasonableness of the pre-estimate. Against them, a number of courts, among them the First Circuit in Colonial at Lynnfield, Inc. v. Sloan and a Missouri appellate court in Valentine's, Inc. v. Ngo, hold that the complete absence of actual loss makes a clause invalid as a penalty.56
The two lines of authority share the no-loss fact and differ in the rule each treats as governing. Garden Ridge worked from statutory text that names actual harm and let the compensation principle control when loss is nil, so a breach could defeat a clause that looked sound at signing. Barrie fixed reasonableness at formation and treated the clause as a risk allocation made before anyone knew how the breach would unfold. That allocation binds the promisee too, since the school could not have claimed more than the stipulated tuition had its loss run higher, and the family cannot claim less because the loss ran lower. The competing risk is that this certainty makes the promisee's windfall enforceable, and the Restatement approach exists to prevent exactly that. One commentary summarizes the cases this way: a substantial clause that was a reasonable pre-estimate would seem enforceable even if no damage ensues, some cases so hold, and others refuse enforcement in the extreme case because the absence of loss is readily proved.57
Land sales produce some of the cleanest no-loss cases. Specific performance is usually the best remedy for breach of a land sale contract because it gives the non-breaching party exactly what she bargained for.58 Now suppose that a buyer agrees to pay $200,000 for a house and posts a $10,000 deposit that the seller may keep as liquidated damages. The buyer defaults, and the seller resells for $250,000, so the seller ends the episode with a profit and no loss. Even if $10,000 was a reasonable estimate at signing, a court that examines actual loss may treat the forfeiture as a penalty, while a court that tests reasonableness only at formation lets the seller keep the deposit, as in Kelly. A finding of zero loss also does not always end the inquiry. The facts should still be examined to see whether the breaching party would be unjustly enriched, as where a seller collects a premium price for prompt delivery and then delivers late without harming the buyer.59