Concept
Failure of Essential Purpose
Limitation of Remedies
Failure of Essential Purpose
A limited remedy that looked fair when the parties signed can become worthless in performance, and § 2-719(2) answers that problem: where circumstances cause an exclusive or limited remedy to fail of its essential purpose, remedy may be had as provided in the Act.1 The provision works on a clause that has already cleared the earlier hurdles. Subsection (1)(a) let the parties substitute repair or refund for the Code's remedies, and subsection (1)(b) made the substitute the sole remedy only if the parties expressly agreed it was exclusive, so a clause meeting both requirements is valid on its face.2 Subsection (2) then asks what no one can know at signing. The word circumstances points to events in performance, such as the defect that appears and the repairs that follow, and when those events leave the buyer without what the clause promised, the limit gives way and the Code's general remedies return.
The valve exists because a limited remedy is a promise about the future, written before anyone knows what will go wrong, and it assumes that repair will work. The buyer accepts the narrower remedy for a lower price, but that trade rests on a premise: the substitute will deliver what it says. When the seller cannot or will not deliver, the premise collapses, and enforcing the clause anyway would let the seller keep the price advantage of the limit while the buyer receives none of the performance that justified it.
Contractual limits take two forms: terms that preclude some element of damages, such as consequential damages, and terms that substitute a different remedy, such as the cost to repair or replace.3 Subsection (2) targets the substitute remedy and leaves the rest of the contract in place. A separate clause excluding consequential damages faces a different test under subsection (3), which enforces the exclusion unless it is unconscionable,4 and a later section asks what becomes of that exclusion when the repair remedy fails. This unit takes up the working test first, then repairs that do not succeed, remedies that cannot operate, remedies that still work despite the buyer's loss, and finally the contested cases in which a defect surfaces only after the remedy's time limit has run.
The Substantial-Value Test
The statute supplies a trigger and a consequence but no definition. The trigger is a failure that circumstances cause, which ties the inquiry to events after signing, and the consequence is that the buyer regains the remedies Article 2 supplies.5 Nothing in the Code says what a remedy's “essential purpose” is, so courts have built the working test themselves.6
The Official Comment gives them the phrase they reach for first. Comment 1 says that when an apparently fair and reasonable clause, because of circumstances, fails in its purpose or deprives either party of “the substantial value of the bargain,” it must give way to the general remedy provisions of the Article.7 That language gives the doctrine its common label, the substantial-value test. The Comment grounds the label in a premise about sales contracts: parties who conclude a contract for sale must accept at least “minimum adequate remedies,” a “fair quantum of remedy” for breach of the obligations the contract outlines.8 Phrases that broad decide little by themselves, and courts admit that they struggle to say what it means for a limited remedy to fail of its essential purpose.9
The method that has emerged starts from the remedy, not from the buyer's loss. A court identifies the purpose the remedy was meant to serve and asks whether applying it in the circumstances that occurred furthers that purpose; only if it does not has the remedy failed.10 For a repair-or-replace clause, the purpose has two sides. The seller wants an opportunity to make the goods conform while limiting its exposure by excluding other damages, and the buyer wants goods that conform to the contract within a reasonable time after a defective part is discovered.11 Behind those purposes sit assumptions the parties rarely state: that the seller will try diligently and in good faith to repair, that repairs will succeed within a reasonable time, and that any loss during the repair period will be small.12 When events defeat those assumptions, the buyer's half of the exchange never arrives, while the seller keeps the limited exposure it bargained for. Some courts frame the same inquiry as a question about what the buyer received in return for accepting the limit: where the goods' capabilities are predictable, a promise of timely and successful repair, and where the goods are experimental or of uncertain capability, only a promise to try.13
Three features keep the test from drifting. First, the measure is the remedy's own purpose. Subsection (2) is triggered when the remedy fails of its essential purpose, and the essential purpose of the Code, of equity, or of contract law generally is beside the point.14 A court therefore does not ask whether the result seems fair in some broad sense or whether the buyer would have done better under the Code's default rules. A remedy that is less generous than the Code's has not failed for that reason, because the Code leaves the parties free to shape their remedies and gives effect to reasonable agreements limiting them; failure appears only when the remedy no longer provides the minimum adequate relief the Code assumes every sales contract carries.15
Second, the question is usually one of fact. Whether a remedy has failed depends on how many defects appeared, how the seller responded, and how long the buyer was left waiting, so the issue ordinarily goes to the jury.16
Third, timing matters. Subsection (2) addresses the application of an agreement to novel circumstances the parties did not contemplate, while a clause that was oppressive at its inception raises a question of unconscionability.17 The line between the two is not perfectly clean. The standard illustration of a worthless remedy, an exclusive “remedy” consisting of a signed letter of apology from the seller's president, is used to show a failure of essential purpose.18 Yet that clause is empty from the moment of signing, and other authority assigns a seller's overreach in what it offers as a remedy to unconscionability, leaving subsection (2) for remedies that were reasonable when made and became worthless as events unfolded.19 A clause hollow on its face may fairly raise both provisions, and the cleanest subsection (2) cases are those in which an apparently fair clause meets events that empty it.
Two decisions show how the test sorts cases. In Martin Rispens & Son v. Hall Farms, Inc., a watermelon-seed buyer whose recovery was limited to the purchase price argued that bacteria in the seed, which produced a fruit disease, were a novel circumstance and that it had bargained for seed that would yield a money crop. The Indiana Supreme Court held that the limitation did not fail of its essential purpose, because the clause was written to hold liability to the price of the seed and enforcing it did exactly that.20 The court added that the possibility that seed will not conform to the warranty should occur to both buyer and seller, so the disease was no novel circumstance.21 The paradigm case for the buyer comes out the other way. In Riley v. Ford Motor Co., a new car developed more than fourteen defects in its first few weeks, the dealer did not satisfactorily repair them, and the exclusive repair-or-replace clause was held to deprive the buyer of the substantial value of the bargain.22 Rispens itself describes Riley this way: the remedy failed because the car could not be repaired to operate free of defects as the express warranty promised.23 Both courts ran the same test and reached opposite results because the clauses had different purposes. The seed clause existed to cap liability at the price, and a capped recovery was the clause working as written, however disappointing the lost crop. The repair clause existed to put the car in working order within a reasonable time, and the buyer received none of that.
Step two carries the weight, because every later step is measured against the purpose identified there. A buyer who is merely unhappy with what the remedy produced has not shown failure at step four, as the seed buyer in Rispens learned; a buyer who shows that the remedy never produced what it promised has.
When the test is met, the remedy fails and the contract remains. The buyer may pursue the remedies Article 2 provides,24 and in a lemon case may also be entitled to revoke acceptance if the defects substantially impair the goods' value.25 Whether the buyer may also recover consequential damages when a separate clause excludes them is a different question. The Code does not answer it, and at least some courts hold that a limitation on consequential damages remains valid even if an exclusive remedy fails.26 A later section takes up the competing approaches.
Unsuccessful or Delayed Repairs
The paradigm failure is the repair-or-replace clause that neither repairs nor replaces in any way that works. Courts apply § 2-719(2) most often to lemons: products with many recurring defects that the seller cannot or will not repair within a reasonable time, leaving the owner indefinitely without full and effective use.29 The fit follows from what the clause promises. A buyer who accepts repair or replacement as the sole remedy has in effect been assured that the seller will cure the defects within a reasonable time and so bring the product into workable condition.30 Three questions therefore organize the cases: whether the seller had a fair chance to cure, how the defects add up, and how long the repairs took.
Bishop Logging Co. v. John Deere Indus. Equipment Co. shows the failure in its plainest form. A logging company bought equipment for swamp logging under a New Equipment Warranty that promised to repair or replace defective parts and excluded consequential damages.31 The machines broke down continuously. One experienced operator had never seen equipment break down more often, a repair mechanic testified that the skidders failed within weeks of entering the field and were never repaired well enough to work over the long term, and the company could never make production.32 Deere answered that the warranty required only repair and replacement, which it had performed promptly on request through more than $110,000 of warranty work.33
The court looked past that bill to the purpose of the clause. For Deere, the repair promise bought an opportunity to make the equipment conform while limiting its risk; for the logging company, it was meant to ensure that the equipment would work in the swamp and that Deere would cure defects within a reasonable time after discovery.34 The court then stated the rule: where a seller has a reasonable chance to correct defects and the equipment still fails to function properly, the buyer is deprived of the benefits of the limited remedy, which therefore fails of its essential purpose, and the buyer may pursue the Code's other remedies if the defect substantially affects the value of the bargain.35 The evidence allowed the jury to find that Deere had not performed its repair obligation properly and within a reasonable time, so the court affirmed.36 What decided the case was what the repairs accomplished. A repair bill records what the seller spent; the remedy's purpose is measured by whether the machines came to work.
The rule in Bishop Logging assumes that the seller had a reasonable chance to cure, and that assumption is the first marker. A repair-or-replace promise by definition implies that the seller must have the opportunity to repair, so the buyer ordinarily must give the seller a reasonable opportunity before arguing that the remedy failed, and a buyer who simply gives up on the seller's efforts has not thereby shown failure.37 What counts as reasonable varies with the goods; some call for a lengthy repair period and several attempts.38 A buyer who defeats the opportunity forfeits the argument. In Belcher v. Versatile Farm Equipment Co., a tractor dealer repaired the tractor, replaced defective parts, and returned it in good condition, but the buyer made unauthorized alterations without the seller's knowledge and refused to return the tractor until it broke down, and the court held that the repair remedy had not failed.39
The seller's chance to cure is real but finite. In McCullough v. Bill Swad Chrysler-Plymouth, Inc., the Ohio Supreme Court held that a new car's warranty failed of its essential purpose where the car was severally and severely flawed and the dealer proved incapable of curing the defects.40 The court adopted an earlier Alabama statement that at some point after the purchase of a new car it should be put in good running condition, because the seller does not have unlimited time to perform its obligation to replace and repair parts, and it concluded that “the hour glass has run” on the dealer's efforts.41 The court also held that the buyer's continued driving of the car did not close off the claim.42 That holding protects the usual consumer, who often must keep using a defective car; a rule treating continued use as acceptance of the repair remedy would force the owner to choose between parking the car and abandoning the claim. The time allowed has no fixed length. In Ex parte Miller, the seller spent 21 months trying to repair a large variety of problems in a new tractor-type vehicle, and the court struck the exclusive repair-or-replace clause, observing that a seller does not have unlimited time to cure.43
The second marker is that defects count cumulatively. In Riley v. Ford Motor Co., a new car showed at least fourteen major and minor defects in its first few weeks, and neither the dealer nor the manufacturer's representatives could satisfactorily repair them.44 The jury returned a verdict far above the cost of repairs, and the Fifth Circuit held that the jury could find that the exclusive remedy deprived the buyer of the substantial value of the bargain, because even after numerous repair attempts the car did not operate as a new car free of defects should.45 The Wisconsin Supreme Court reached the same kind of result in Murray v. Holiday Rambler, Inc., holding that a repair-or-replace remedy failed where the cumulative effect of a substantial number of defects substantially impaired the goods' value to the buyer.46 Counting defects together closes an easy escape: judged one at a time, each problem can be shown fixed once even as new ones keep appearing and the buyer never receives goods that run as warranted. The count also marks the line between the lemon and the ordinary new car. In the usual case, a few repairs and replacements bring a new car into good working condition, and the remedy has done what it was designed to do; Riley fits subsection (2) because the volume of defects and the repeated failures to repair were unusual.47
Delay is the third marker, because the buyer's side of the bargain includes receiving working goods within a reasonable time. Repair or replacement can fail because it takes too long, and in Ehlers v. Chrysler Motor Corp. the South Dakota Supreme Court held that Chrysler's delay in performing the repairs caused the remedy to fail of its essential purpose.48 Sellers may expect a reasonable chance to repair, and buyers may expect that repairs will be done with reasonable dispatch.49 A jury instruction quoted in the dissent in Mercury Marine v. Clear River Construction Co. gathers these routes in one place: a limited warranty fails of its essential purpose if the seller was unwilling to repair or replace, if the goods were unrepairable, or if the seller unreasonably delayed in repairing or replacing.50 The decision shows how much turns on the repair record. The majority found no evidence of repeated attempts to correct defects and repeated failure to do so; the dissent counted five repair attempts, only three of them successful, and would have upheld the jury's finding of failure.51
On one point courts divide: whether the seller's good faith and diligence count. Some courts treat them as relevant, but most hold them irrelevant, which fits the statutory text, because subsection (2) asks whether circumstances caused the remedy to fail and says nothing about the seller's effort or state of mind.52 The majority view also follows from the remedy's purpose. A buyer who needs working goods ends up in the same position whether the seller whose repairs fail was diligent or indifferent, and Bishop Logging reached its result despite more than $110,000 of repair work. The majority view therefore looks to results, not effort.
Step four settles the outcome, and the earlier steps supply its evidence. Whether a repair history adds up to failure is ordinarily a question of fact for the jury, which is why the verdicts in Bishop Logging and Riley were affirmed on the strength of the record.53
When the remedy fails, the buyer may pursue the remedies the Act provides.54 One example is the recovery allowed in Royal Lincoln-Mercury Sales, Inc. v. Wallace, where the jury found that the dealer's repairs had not brought a new car into conformity with the warranty.55 The Mississippi court reasoned that a repair warranty gives the buyer confidence of receiving conforming goods and gives the seller a limit on its liability, and that when the warranty fails as a remedy, its function as a limit on liability fails with it.56 The court allowed recovery of the purchase price less the car's depreciation and placed the burden of proving depreciation on the party relying on it, so the jury's verdict at the full price stood because no evidence of diminished value was offered.57 Whether that reasoning also carries away a separate clause excluding consequential damages is a different question, which the later section on failed remedies and damage exclusions addresses.
Remedies That Cannot Operate
Lemon cases involve a remedy that was tried and did not work.61 A second family of failures needs no history of attempts, because the remedy cannot operate at all: the defect has destroyed the goods, the clause demands an act the seller's breach has made impossible for the buyer, or the seller refuses to perform. In each case the inquiry turns on whether the remedy can be carried out.
The standard illustration pairs an automobile manufacturer's limit to repair and replacement of defective parts with a requirement that the buyer deliver the defective parts to its plant. If defective wiring destroys the entire car, the buyer cannot deliver the wiring to the plant, and even a buyer who could would gain nothing, because repairing or replacing the wiring would not restore the car to working condition.62 The exclusive remedy therefore fails of its essential purpose. The same result follows whenever the seller is willing to repair but the goods cannot be repaired, as when they have been destroyed.63
The reason lies in what a repair-or-replace clause assumes: that goods remain to be made right and that the buyer can bring them to where the repair happens. The buyer accepts the limit for a lower price on the strength of that assumption, and the seller who drafted the clause wrote the delivery step into it. When the seller's own defective product destroys the goods, the step becomes impossible because of the breach, and holding the buyer to the clause turns the seller's default into the seller's defense.
Russo v. Hilltop Lincoln-Mercury, Inc. applied that reasoning to a burned car. The buyers' car caught fire, they returned it to the dealer immediately, and the trial court found it “junk” and “worthless.”64 The dealer argued that its liability was limited to repairing or replacing defective parts and that the buyers had failed to deliver the car to it for that purpose. The Missouri Court of Appeals answered that a promise to repair implies the car is repairable, that this car could not be repaired, and that it would be “ludicrous” to require the buyers to ask the dealer to do the impossible.65 The court added that warranty law should be administered fairly to all parties and that warranties should be meaningful; it held that the buyers had satisfied the warranty's condition and that the dealer was liable for its breach.66 Because the car was worthless after the fire, the court fixed basic damages at its pre-fire value, the $3,918.61 purchase price.67
The same opinion marks what impossibility does not accomplish. The warranty expressly excluded the expense of renting a car, and the trial court had nevertheless allowed the buyers $109.72 for a rental. The buyers defended the award by invoking the exception in § 2-719(2) for a remedy that fails of its essential purpose; the appellate court rejected the argument, saying it did not believe there was such a failure of purpose there, and reversed the rental allowance.68 The court thus excused the buyers from the impossible repair step and still enforced the dealer's separate exclusion of rental expense. Whether a failed repair remedy should carry such an exclusion down with it is contested and belongs to a later section, so Russo is best read here for a narrower lesson: an excused repair step does not automatically erase every other limit in the warranty.
The same conclusion follows when the goods can be repaired but the seller will not repair them. The essential purpose of a repair-or-replace warranty is to repair or replace the goods, so a seller unwilling to do so leaves the buyer with a remedy in name only.69 The limit rested on the seller's promise to cure, and a seller who flatly refuses has repudiated the very obligation that justified confining the buyer to it. The logic that excused the Russo buyers from asking for the impossible applies with similar force to a buyer who would otherwise have to keep requesting a repair the seller has already refused to perform.
A remedy that still works is a different matter. A repair-or-replace remedy does not fail of its essential purpose merely because the buyer believes that going back for repairs will be too much trouble.70 Its purpose is to repair the goods, and when the goods can be repaired and the seller stands ready to do the work, that purpose is being served. The line therefore runs between a remedy that cannot be carried out, or that the seller will not carry out, and a remedy that can be carried out but that the buyer would prefer to avoid. The next section takes up cases in which a limited remedy does exactly what it was designed to do.
Steps two and three examine the same question from the buyer's side and the seller's side, and step four keeps the doctrine from swallowing every ordinary repair clause. A buyer who clears step two or step three may disregard the repair limit and pursue the Code's remedies, while a buyer who reaches only step four has a complaint about convenience that the statute does not reach.
Remedies That Still Work
A limited remedy that works as written is enforced, however much the buyer regrets the bargain. The preceding sections showed remedies that delivered nothing. The cases here run the other way: the remedy was available, or was used, or never had a fair chance, and the buyer's real complaint is the result. A limitation fails when it deprives a party of the substantial value of its bargain, and courts have noted that few situations meet that description; the most familiar is the repair-or-replace clause whose seller, through negligent repair or because the goods are beyond repair, cannot put them in warranted condition.73 The buyer must also come forward with evidence of failure, and the buyer in Figgie lost because the court found no evidence that the limited remedy had failed.74
Two reasons explain why a working remedy is enforced. The first is price. A seller that narrows its exposure to repair, refund, or a capped sum can charge less, and a buyer who accepts the narrower remedy takes the lower price in exchange, so a buyer who later obtains the Code's broad remedies receives something it never paid for. Subsection (2) answers a remedy that stopped working; it gives no relief to a buyer who wishes the bargain had been better. The second reason concerns the steps a buyer must take. A repair promise assumes that the seller will be asked to repair and given the chance, and the Code's policy of encouraging the parties to work out their differences and so minimize losses from defective performance is defeated when a buyer skips that step and demands full relief.75
Mercury Marine v. Clear River Construction Co. shows the first reason at work. The trial court there was prepared to instruct the jury that a repair-or-replace warranty failed of its essential purpose because the warranty did the buyer no good when engines quit in the middle of ocean fishing tournaments, and the Mississippi Supreme Court called that reasoning untenable.76 A repair remedy by definition anticipates some failure of the goods, so the occurrence of a failure, even one that costs the buyer profits, does not by itself defeat the remedy's purpose; the court noted that the trial judge's concern about the warranty's uselessness at sea hinted at unconscionability, which is a separate question.77
The simplest working remedy is the one the buyer never used. In United States Achievement Academy, LLC v. Pitney Bowes, Inc., a business leased document-printing software under a contract that gave a ninety-day warranty of repair or replacement. The software eventually printed the buyer's documents, but only through a workaround that took as many as 270 mouse clicks per print job, and the buyer argued that the remedy failed because the software could never be repaired into a three-click system.78 The court rejected the argument. The buyer had signed a customer acceptance document after the warranty expired, knowing how the workaround operated, and if it was dissatisfied it should have invoked its sole remedy of repair or replacement within the ninety days.79 The buyer could not explain how a remedy it never used had deprived it of the substantial value of the bargain when it had agreed that it was satisfied, so the court granted summary judgment on the claim.80 Failure of essential purpose is measured by what happened when the remedy was tried, and a buyer who let the warranty period run without trying it, and then signed an acceptance, has no failure to point to.
A remedy that ends in a refund can likewise work exactly as written. In Figgie, the seller installed bottle-labeling equipment in April and spent the following months trying to resolve the problems that appeared. In November, when those attempts had still failed, the equipment was returned and the price refunded.81 The buyer argued that by first attempting repair the seller had elected repair as the exclusive remedy, so that the failure of repair made the whole remedy fail. The court of appeals found no support for that theory, because the buyer's remedies were limited to repair, replacement, or return, and the remedies invoked were exactly those that § 2-719(1)(a) contemplates and the agreement provided.82 The court added that a seller is expected to try repair first and to resort to return and refund only if repair does not succeed, particularly in a sale of sophisticated equipment between commercial entities.83
Figgie and Bishop Logging involved the same kind of repair clause and came out differently, and the difference lies in what stood behind the repair promise. A limited remedy of repair or replacement can fail when the seller's repairs are unsuccessful, as they were for the logging equipment.84 The logging company's warranty offered nothing but repair or replacement, so once the repairs failed there was nothing left for the buyer to collect. The Figgie agreement added a third alternative, return of the goods for repayment of the price, and that alternative was carried out. A repair clause fails when repairs never work and nothing else is on offer, and it survives when the contract supplies a fallback and the fallback is honored. Even a refund is not immune. The Figgie court took care to say that a remedy of repair, replacement, or refund may fail if the defects cannot be detected until so late that a refund of the price would be totally inadequate, and it found that risk absent because the buyer knew of the problems from the outset and took part in the repair efforts.85 The late-discovery problem is contested, and a later subsection of this unit gives it separate treatment.
A price cap shows a third way a remedy can work. In Rispens, described above, the Indiana Supreme Court identified the purpose of the seed clause as limiting liability to the purchase price and concluded that enforcing the cap served precisely that purpose.86 The buyer had bargained for seed and not for a full-grown crop, and a buyer who thought recovery of the price inadequate was free to bargain for a more comprehensive remedy.87 The seed cap promised the seller a ceiling on liability, and a recovery held to the price is that ceiling doing its job. Whether a latent defect makes such a limit unconscionable is a separate question with a separate standard, and the Rispens court rejected that argument as well.88
The buyer's own conduct can also defeat a failure argument. The Belcher buyer, whose alterations and refusal to return the tractor appear above, interfered with a repair that had already been done.89 In Web Press Services Corp. v. New London Motors, Inc., the seller had repaired every defect in a vehicle except a defective axle that was not immediately discernible. The buyer never asked the seller to repair the axle and simply sought return of the purchase price, and the Connecticut Supreme Court upheld a trial court's award of half the cost of repairing the axle as adequate and fair relief that best served the Code's policy of minimizing losses.90 The two buyers defeated the cure in different ways: one interfered with a completed repair, the other never gave the seller a chance to look at the axle at all. A seller's chance to cure does not last indefinitely, as Web Press itself recognizes, but it must be given before the remedy can be called a failure.91 Bishop Logging marks the opposite pole, since the seller there had every chance to cure and its repairs still did not produce working machines.
The nature of the goods also shapes what a repair promise means. When the goods are experimental, the remedy is less likely to fail of its essential purpose, because its purpose was never a guarantee that the goods would work as hoped, and the same holds where a buyer orders goods for a special purpose and supplies detailed specifications, since the limited remedy may then amount to a promise that the seller will attempt to make the goods serve.92 This qualifies the earlier point that a seller's diligence does not save a repair remedy that never cured the goods. Buyers of ordinary new goods expect a repaired product to match what they originally expected, and it is unlikely that sellers assume buyers care only whether the seller tried, which is why the lemon cases favor buyers.93 Where the contract's promise was only an attempt, the attempt is the performance.
The decisions also show what a seller can do to keep a limited remedy from failing. The first tool is a backup remedy for the case in which repair does not succeed, such as a refund of the price, replacement with different goods, reasonable liquidated damages, or a combination.94 In Computerized Radiological Services v. Syntex Corp., the seller of an X-ray scanner provided a price refund as the backup, and the court found that the refund prevented failure of essential purpose.95 Once repair has run its course without success, the contract itself names a defined result, so the buyer is never left holding broken goods with nothing to claim. The second tool is candor about the promise. For used goods, experimental goods, or sales in a business setting, a seller that promises only its best efforts, in explicit language the buyer understands, has placed the risk of unsuccessful repair on the buyer, and courts should then find no failure.96 Where a seller limits a consumer to repair or replacement of new goods, there is little it can do to avoid the failure cases other than actually making the goods work.97
Latent Defects and Time Limits
A harder question arises when a remedy works exactly as written but runs out of time before the buyer can learn there is anything to remedy. A latent defect is one that ordinary inspection cannot reveal and that shows itself only after use, processing, or the passage of time, and when the contract gives the buyer a short window in which to claim, such a defect can surface after the window has closed. Courts divide over whether that sequence is a failure of essential purpose, so the answer on an exam depends on which line the governing jurisdiction follows.
The question is hard because of what a time limit does. A time limit allocates risk at signing: the seller answers for defects that appear within the window, the buyer bears whatever appears afterward, and the seller can price the goods and plan its exposure because the window ends. For defects that announce themselves quickly, the allocation is fair, since a buyer who spots the problem inside the window can use the remedy. A latent defect changes the arithmetic. The defect exists from the day of delivery, yet no one can see it until later, so the clause that allocated the risk now operates as a complete bar to a claim the buyer had no chance to make. Courts that apply subsection (2) rely on that result, and courts that refuse answer that the clause did precisely what it was written to do.
The decision most often cited for the first view is U.S. Aprons, Inc. v. R-Five, Inc., a federal district court case from Nebraska. U.S. Aprons ordered red and white fabric from R-Five to make disaster-relief vests headed for the American Red Cross, and the contract required any claim to be brought within 30 days or before the fabric was cut.101 The fabric yellowed only after it was cut and sewn, so the defect could not have appeared before the clause shut off the claim. The court denied the seller's motion for summary judgment because enforcing the clause would deprive U.S. Aprons of any remedy, and it held that the contract language had to yield to the Code's general remedy provisions, which allow a buyer a reasonable time to notify the seller after it discovers, or should have discovered, the defect.102 On that reasoning, the buyer's time to complain runs from discovery of the defect and not from delivery of the goods.
Prairie River Home Care, Inc. v. Procura, LLC carried the same reasoning to software. The contract capped Procura's liability at the fees Prairie River had paid in the twelve months before any claim, and Procura delivered the software on time, but Prairie River did not build it into its computer system until eight to ten months after the parties contracted, and only then did the latent defects appear.103 Procura repeatedly assured Prairie River that it could fix the problems, and each attempted fix produced new ones, so by the time Prairie River abandoned the software more than twelve months had passed since it had paid Procura, and the cap would have left it with no remedy whatsoever.104 The ruling came on a motion to dismiss, so the court decided only that these allegations were enough to show a failure of essential purpose and let the case go forward.105 Software fits the pattern because its defects often emerge only after the buyer has spent heavily to integrate it, and subsection (2) reaches latent defects that cannot be discovered before the buyer has already suffered damage.106
The doctrine is not confined to repair clauses. In Tokyo Ohka Kogyo America, Inc. v. Huntsman Propylene Oxide, LLC, an Oregon federal court observed that failure analysis comes more easily with a repair-or-replace clause, because an unrepaired product shows that the remedy has not worked, but that a refund remedy can fail as well, as when the seller conceals relevant facts or the defect is latent.107 The mechanism is the one at work in the fabric and software cases. A promise to refund the price is of no use to a buyer who cannot learn of the defect, or has been kept from learning of it, until the time to invoke the promise has passed.
The opposing line starts from a different view of what the clause is. In Arkwright-Boston Manufacturers Mutual Insurance Co. v. Westinghouse Electric Corp., the Fifth Circuit, applying Texas law, considered a turbine whose equipment warranty expired on April 13, 1981, twenty-one months after shipment, and whose fourth-row blade failed on October 19, 1981, more than six months after that date.108 Arkwright argued that the time limit was itself a remedy limitation that failed of its essential purpose under § 2-719(2), so that the court should strike it.109 The court declined, drawing on decisions holding that a warranty period operates just as intended when it allocates the risk of loss both before and after expiration, and that a purchaser cannot claim failure of essential purpose merely because a potential claim does not arise until after the warranty period has run.110 One of those decisions put the point in terms of what subsection (2) governs: § 2-719 concerns the circumscription or modification of remedies and has nothing to say about language that defines the scope of a seller's liability.111
That reasoning echoes the distinction drawn earlier between disclaimers and remedy limitations. A time limit on the warranty itself defines how long the seller's promise lasts, much as a disclaimer defines whether the promise exists, so a seller who promised to answer for defects through a fixed date never promised more. A defect that appears after the date lies outside the promise, and nothing has failed. The two lines therefore disagree about characterization before they disagree about outcome: one sees a remedy that ran out and left the buyer empty-handed, and the other sees a promise that ended on schedule.
The decision that gave the failure line its sharpest form is also its most criticized. In Wilson Trading Corp. v. David Ferguson, Ltd., a contract for the sale of yarn barred claims made after weaving, knitting, or processing, or more than ten days after receipt of the shipment, and the New York Court of Appeals reasoned that latent defects in the yarn might not be discoverable in that time, in which case the limited remedy would fail of its essential purpose because the buyer would have no remedy for them.112 The Arkwright court observed that in the twenty years since, no other court had followed that decision, and it repeated the objection made to it.113 Subsection (2) asks whether later circumstances changed how a remedy works, and a time bar that denies a late claim works exactly as its drafter intended, since the denial is its purpose.114 If a ten-day window is objectionable, the critics say, the objection existed at signing, and the proper attack is that the clause fails to leave a minimum adequate remedy under subsection (1), that it is unconscionable, or that it is a manifestly unreasonable time limit under the Code's general rule on such limits.115 A concurring judge in Wilson Trading would have decided the case on that last ground alone.116
The same critique reaches Neville Chemical Co. v. Union Carbide Corp., where the court held that a return-of-price remedy failed of its essential purpose when the defect was latent and the buyer's consequential losses far exceeded the price.117 Avoiding enormous consequential damages is the purpose of every limited remedy, so a refund limit does not fail merely because the losses it excludes are large.118 Both buyers were merchants who could have bargained for better remedies or looked to other suppliers, and a yarn maker has sensible reasons to want defects raised before processing makes the yarn impossible to examine, so rewriting the deal afterward hands the buyer a windfall it never chose to pay for.119 The point connects to the earlier rule that subsection (2) governs the application of an agreement to circumstances the parties did not foresee, while overreaching built into a clause at its inception belongs to unconscionability.
The route matters because the same clause can be attacked through different provisions. In Trinkle v. Schumacher Co., the seller's invoice stated, “Positively no claims allowed after goods are cut,” and the defects in the fabric could not be determined until fabrication work revealed them, which is the situation in U.S. Aprons.120 The Wisconsin trial court treated the clause as a conscionable limit on consequential damages under subsection (3) and denied all consequential damages except the cost of the goods, and the appellate court, viewing the claim as one for breach of warranty, held that limiting the award to the cost of the goods was improper on that theory.121 The facts alone therefore do not choose the doctrine. One court asked whether the remedy failed, and another asked whether the exclusion of consequential damages was unconscionable.
Neither line treats the mere existence of a latent defect as enough. A buyer who discovers the defect within the time the remedy allows, and then fails to use the remedy, has not shown a failure of essential purpose.122 A buyer who notifies the seller in time is in a stronger position when the seller assures that it will cure and then fails to, and a court may find the remedy failed on those facts.123 The dividing question is whether the buyer could have used the remedy within its window, and the jurisdictions part company only when the answer is no.
Step two screens out the buyer who sat on a known problem, and step four is where the jurisdictions part company. When a question does not say which line governs, present both and show how each resolves the facts.