Contracts & Torts: Private Law

On 27 February 1992 Stella Liebeck, aged seventy-nine, was sitting in the passenger seat of a parked car in Albuquerque, New Mexico. She had bought a cup of coffee from a McDonald's drive-through, and she put it between her knees to take the lid off and add cream. The cup tipped. The coffee was served at somewhere around 180 to 190 degrees Fahrenheit, and it soaked into her cotton sweatpants and stayed against her skin. She suffered third-degree burns over about six per cent of her body, spent eight days in hospital, and underwent skin grafting and later debridement.

She wrote to McDonald's asking for her medical costs and lost income, around twenty thousand dollars. The company offered eight hundred. She sued.

Notice what did not happen. No police officer investigated. No prosecutor decided whether to charge anyone. No one went to prison, and nobody was fined by a regulator. A private individual brought a claim against a private company, in her own name, asking for money. That is private law.

Before going further: this course is general legal education. It is not legal advice, it does not create a lawyer and client relationship, and the law it describes varies by state and by country and changes over time. If you have an actual dispute, consult a lawyer licensed where you are.

Four axes that separate the systems

Students often think the difference between criminal and civil law is severity. It is not. The differences are structural.

	Criminal law	Public law	Private law
Who brings the case	The state, through a prosecutor	Usually a citizen or firm against a government body, or an agency enforcing a statute	One private party against another
What is sought	Punishment: imprisonment, fine, supervision	To strike down, compel, or review official action	Compensation, or an order to perform or stop
Standard of proof	Beyond a reasonable doubt	Varies; often preponderance	Preponderance of the evidence, meaning more likely than not
Who controls it	The prosecutor may drop it whatever the victim wants	The claimant, subject to standing rules	The claimant, who may settle, discontinue, or refuse to sue

The proof standard is the axis that produces the results people find hardest to accept. Beyond a reasonable doubt is a demanding threshold, deliberately set high because the state is proposing to take someone's liberty. Preponderance of the evidence asks only whether the claim is more likely true than not: on the classic image, whether the scales tip at all.

The consequence is that identical conduct can produce opposite outcomes in the two systems without either being wrong. In 1995 a California jury acquitted O. J. Simpson of two murders. In 1997 a civil jury in the same state, on the same underlying events, found him liable for wrongful death and battery and awarded the victims' families damages totalling around thirty-three and a half million dollars. Nothing was overturned. The first jury was asked whether the state had removed all reasonable doubt. The second was asked whether the plaintiffs had shown it was more likely than not.

Key idea: Criminal and civil proceedings ask different questions on different standards, so different answers are not a contradiction.

What private law contains

Private law, sometimes called the law of obligations, is conventionally divided into four bodies. Two of them are this course.

• Contract. Obligations you took on voluntarily. The state enforces them because you chose to be bound, and the central question is always what was agreed.
• Tort. Obligations imposed on you regardless of agreement, mostly amounting to a duty not to injure others in specified ways. You never agreed not to run someone over. The duty is there anyway.
• Property. Rights in things, good against the world rather than against a particular person.
• Unjust enrichment and restitution. The obligation to give back a benefit it would be unjust to keep, which can arise with no agreement and no wrong: a bank wires money to the wrong account, and the recipient must return it.

The contract and tort distinction is the spine of the course, so fix it now with a test. Ask where the obligation came from. If the answer is that the defendant promised, you are in contract. If the answer is that the law imposes this duty on everyone in the defendant's position, you are in tort. A surgeon who operates badly may be liable in both: in tort for falling below the standard of care, and in contract if a specific outcome was guaranteed, which is exactly the situation in a case you will meet in Module 3.

Where the rules come from

An American court deciding a contract or tort case is not reading a code from cover to cover. It is using four kinds of material, and knowing their rank saves a lot of confusion.

1. Binding precedent. Decided cases from courts above the deciding court in the same jurisdiction. This is the common law method: rules are extracted from decisions and refined case by case. Most of contract and tort law is judge-made in this way.
2. Statutes. Where a legislature has intervened, the statute wins. The dominant example in contract is the Uniform Commercial Code, a model statute drafted by the Uniform Law Commission and the American Law Institute, adopted in some form by every state. Its Article 2 governs contracts for the sale of goods, and it displaces the common law rules for that category.
3. Persuasive authority. Decisions from other states, and the Restatements: systematic summaries published by the American Law Institute, including the Restatement (Second) of Contracts of 1981 and the Restatement (Second) of Torts. A Restatement binds nobody. Courts cite it constantly anyway, because it states a rule cleanly and a judge can adopt it by saying so.
4. Scholarship. Treatises and law review articles, which occasionally change a whole field. The rule of strict products liability you meet in Module 5 began as a concurring opinion nobody was required to follow.

Two structural cautions. First, this is a United States and common law course, and Louisiana is the American exception: its private law rests on a civil code in the French and Spanish tradition, and much of what follows does not describe it. Second, most of these rules are state law, so there are fifty versions of them, and a rule described here as settled may be settled differently in the next state.

Back to Albuquerque, and what a jury actually did

The Liebeck jury heard evidence that McDonald's held its coffee at a temperature far above what home brewers and many competitors used, that the company's own quality documents specified that range, and that it had records of several hundred prior burn complaints over about a decade. It found for Liebeck. It also found her twenty per cent responsible for the spill, and reduced her compensatory award accordingly, from two hundred thousand dollars to one hundred and sixty thousand. It then added punitive damages of 2.7 million, a figure the plaintiff's counsel had linked to roughly two days of the company's coffee revenue. The trial judge cut the punitive award to four hundred and eighty thousand. The parties then settled for an undisclosed sum, and the case never produced an appellate opinion.

The version that entered public memory, a woman who spilled coffee on herself and won millions, is wrong in almost every particular, and it was the centrepiece of a lobbying campaign for limits on tort awards that ran through the 1990s. You will meet that argument properly in the last lesson. For now, note the mechanism: the reduction for her own share of the fault, the separation of compensation from punishment, and the judge's power to cut a jury's number.

In short: A private law case is a claim by one party against another for a remedy, decided on the balance of probabilities, and the numbers reported in the press are frequently not the numbers anyone paid.

Remedies, in one paragraph

Private law's main remedy is money, called damages, and its normal aim is not to punish but to put the claimant in a particular position: in contract, the position of performance, and in tort, the position they would have occupied had the wrong not happened. Courts can also order a party to do or stop doing something, through specific performance in contract or an injunction more generally, but those are exceptional in common law systems and require a showing that money is inadequate. Punitive damages, aimed at punishment and deterrence rather than compensation, are available only in narrow circumstances, almost never in contract, and are constrained by constitutional limits on their size.

Common misconceptions

• "Civil cases are the small ones and criminal cases are the serious ones." The difference is who brings the case, what is sought, and the standard of proof. A civil judgment can be far larger in money terms than any criminal fine.
• "If someone is acquitted, a later civil claim is double jeopardy." Double jeopardy protects against repeated prosecution by the state. A private claimant is a different party asking a different question on a lower standard.
• "The Restatements are the law." They are private summaries with no binding force. They matter because courts choose to adopt them.
• "American law is uniform." Contract and tort are mostly state law, so there are fifty bodies of it, plus Louisiana's civil code tradition, which differs at the root.

Where this leaves us

• Private law is a claim by one private party against another, for compensation or an order, decided on the preponderance of the evidence.
• The Simpson outcomes of 1995 and 1997 illustrate that different standards of proof produce different answers on the same facts without contradiction.
• Contract obligations are voluntarily assumed; tort obligations are imposed by law regardless of agreement.
• Courts work from binding precedent, then statutes such as UCC Article 2 for the sale of goods, then persuasive sources including the Restatements, then scholarship.
• Liebeck's award was reduced for her share of the fault, the punitive figure was cut by the judge, and the case settled confidentially.
• This course is general legal education and not legal advice.

Sources

1. Legal Information Institute. (n.d.). Civil law. Cornell Law School. law.cornell.edu
2. Legal Information Institute. (n.d.). Tort. Cornell Law School. law.cornell.edu
3. Legal Information Institute. (n.d.). Contract. Cornell Law School. law.cornell.edu
4. Legal Information Institute. (n.d.). Preponderance of the evidence. Cornell Law School. law.cornell.edu
5. Legal Information Institute. (n.d.). Restatement of the law. Cornell Law School. law.cornell.edu
6. Wikipedia contributors. (n.d.). Liebeck v. McDonald's Restaurants. en.wikipedia.org
7. American Law Institute. (1981). Restatement (Second) of Contracts. American Law Institute Publishers.

In November 1891 the Carbolic Smoke Ball Company took out a newspaper advertisement promising to pay one hundred pounds to anyone who used its device three times a day for two weeks and still caught influenza. To answer the obvious objection, the advertisement added that the company had deposited one thousand pounds with the Alliance Bank in Regent Street to show its sincerity.

Louisa Carlill bought a smoke ball, used it as directed from mid-November until January, and caught influenza. She asked for her hundred pounds. The company refused, and its arguments in Carlill v Carbolic Smoke Ball Co, decided by the English Court of Appeal in 1893, are the arguments every first-year student learns to defeat.

• It was mere advertising puff, not a serious promise. The court pointed at the thousand pounds in the bank. You do not deposit money to demonstrate that you are not serious.
• You cannot make an offer to the whole world. You can, if it is an offer to perform in exchange for an act; the contract then forms with each person who performs, not with everyone who reads it.
• She never communicated her acceptance. In an offer of this kind the offeror plainly does not expect a letter first. Performing the act is the acceptance, and the offer waives notification.

The point: An offer is a communication that gives the other side the power to close the deal by saying yes. If the recipient's yes would still leave something to negotiate, it was not an offer.

Offer, or invitation to negotiate

The general rule is that an advertisement, a catalogue, a price list, and goods on a shelf are not offers. They are invitations to make one. The practical reason is stock: a shop that advertises a coat has not promised a coat to each of the ten thousand people who read the page, and treating the advertisement as an offer would make it liable to all of them.

But the rule bends where the advertisement is specific enough to leave nothing to negotiate. In Lefkowitz v. Great Minneapolis Surplus Store, 251 Minn. 188 (1957), the store advertised three fur coats worth one hundred dollars each, on a first come first served basis, for one dollar apiece. Morris Lefkowitz was first in line. The store refused to sell, saying a house rule limited the offer to women. The Minnesota Supreme Court held that where an advertisement is clear, definite, and explicit, and leaves nothing open for negotiation, it is an offer that the first qualifying person can accept, and the store may not add a condition after the fact.

Compare the two advertisements. Carbolic's set out the exact conduct required and the exact sum payable. Lefkowitz's set out the item, the price, the quantity, and the selection rule. Neither left anything for the parties to work out. That is what distinguishes an offer from an invitation, not the medium it appeared in.

Bilateral and unilateral, and the problem in the middle

A bilateral contract is a promise for a promise: I promise to sell, you promise to buy, and both of us are bound before either performs. A unilateral contract is a promise for an act: I promise a reward if you find my dog, and you are never obliged to look, but if you produce the dog I owe you the money.

The unilateral case creates a famous difficulty. Suppose I offer you a thousand dollars to walk across the Brooklyn Bridge, and revoke when you are three quarters of the way over. If the acceptance is completion, nothing has been accepted and you get nothing for a long walk. The modern answer, set out in section 45 of the Restatement (Second) of Contracts, is that once you begin performance the offeror can no longer revoke; an option contract arises, and the offeror must wait to see whether you finish. Preparing to perform is not enough. The line is between getting ready and starting.

The mirror image rule, and the forms problem

At common law an acceptance must match the offer exactly. Any change makes it a counteroffer, which both rejects the original and makes a new one that the original offeror may accept or refuse. That is the mirror image rule, and its consequence is that the original offer is dead: you cannot reject terms, think again, and then accept them.

This works well for negotiated deals and badly for commerce. Businesses trade on printed forms. A buyer sends a purchase order with its standard terms on the back; the seller returns an acknowledgment with different standard terms on its own back; the goods ship, the buyer pays, and nobody reads either. Under the mirror image rule the acknowledgment is a counteroffer, and the buyer accepts the seller's terms simply by taking delivery. Whoever fires the last form wins, which is why lawyers call it the last shot rule.

For contracts governed by Article 2 of the Uniform Commercial Code, meaning sales of goods, section 2-207 dismantles that. A definite and seasonable expression of acceptance operates as an acceptance even if it states additional or different terms, unless it is expressly made conditional on assent to those terms. Between merchants, the additional terms become part of the contract unless the offer expressly limits acceptance to its own terms, the new terms materially alter the deal, or the other side objects within a reasonable time. And if the writings do not agree but the parties behave as though there is a contract, there is one, on the terms the writings share plus the code's default rules.

Remember: Mirror image for services, land, and employment; section 2-207 for goods. Deciding which body of law applies is the first move, not an afterthought.

Timing: four rules that decide who wins

Most formation disputes are actually disputes about the clock. Four rules do the work.

1. An offer takes effect when received. A letter posted and never delivered offers nothing.
2. A revocation takes effect when received. Saying it out loud in your office does not revoke anything.
3. A rejection or counteroffer takes effect when received, and kills the offer.
4. An acceptance takes effect when dispatched. This is the mailbox rule, which comes from the English case Adams v Lindsell in 1818: the contract forms the moment the acceptance is put into the post, even if it is delayed or lost.

Work an example. On Monday a seller posts an offer, which arrives Tuesday. On Wednesday morning the buyer posts an acceptance. On Wednesday afternoon the seller posts a revocation, which arrives Thursday; the acceptance arrives Friday. Is there a contract? Yes. The acceptance took effect on Wednesday morning when it was posted. The revocation took effect on Thursday when it arrived, by which time the deal was two days old.

Now change one fact. The buyer posts a rejection on Wednesday morning, changes their mind, and telephones an acceptance on Wednesday afternoon; the rejection arrives Thursday. Courts treat this as a race, and the general approach is that the mailbox rule does not protect an acceptance sent after a rejection: whichever the offeror receives first governs, and an offeror who has relied on the rejection is protected.

Instant communications complicate the picture. Where the exchange is effectively face to face, telephone, and generally email and messaging, acceptance is normally effective on receipt rather than on dispatch, because the dispatch rule exists to allocate the risk of postal delay.

Killing an offer

An offer does not last forever. It ends when any of these happens: the stated time expires, or a reasonable time passes if none was stated; the offeror revokes before acceptance; the offeree rejects or counteroffers; or the offeror dies or loses capacity before acceptance.

Two ways to stop the offeror revoking are worth knowing. An option contract is a separate agreement to hold the offer open, and like any contract it must be paid for. Under UCC section 2-205 a merchant's signed written offer that states it will be held open is irrevocable for the stated time, up to three months, with no payment required at all: a rare instance of the code creating a binding promise out of nothing but a signature.

Common misconceptions

• "An advertised price is an offer the shop must honour." Usually it is an invitation to negotiate. It becomes an offer only when it is clear, definite, and explicit, leaving nothing open, as in Lefkowitz.
• "A contract needs a signature." Most contracts need no writing at all. Signatures matter for the categories in the statute of frauds, which is a separate lesson.
• "Silence is acceptance if the offer says so." Generally silence does not accept. Narrow exceptions exist where the parties' prior course of dealing makes silence meaningful, or where the offeree takes the benefit knowing what is expected.
• "An acceptance that adds a term is still an acceptance." Not at common law, where it is a counteroffer that kills the offer. Under UCC 2-207 it can be an acceptance, and the added term is then handled separately.

Putting it together

• Carlill established that an offer can be made to the world, accepted by performance, without prior notification, and that a deposit answers the puff argument.
• Advertisements are usually invitations to negotiate; Lefkowitz shows when one is definite enough to be an offer.
• Under Restatement section 45, beginning performance of a unilateral offer makes it irrevocable; preparing to perform does not.
• The mirror image rule makes any variation a counteroffer that destroys the original offer.
• UCC 2-207 replaces that for goods, so an expression of acceptance with extra terms can still form a contract.
• Offers, revocations, and rejections take effect on receipt; acceptances take effect on dispatch under the mailbox rule, subject to instant communication and the rejection-first situation.

Sources

1. Legal Information Institute. (n.d.). Offer. Cornell Law School. law.cornell.edu
2. Legal Information Institute. (n.d.). Acceptance. Cornell Law School. law.cornell.edu
3. Legal Information Institute. (n.d.). Mailbox rule. Cornell Law School. law.cornell.edu
4. Legal Information Institute. (n.d.). U.C.C. section 2-207: Additional terms in acceptance or confirmation. Cornell Law School. law.cornell.edu
5. Lefkowitz v. Great Minneapolis Surplus Store, Inc., 251 Minn. 188 (1957). CourtListener. courtlistener.com
6. Wikipedia contributors. (n.d.). Carlill v Carbolic Smoke Ball Co. en.wikipedia.org
7. Farnsworth, E. A. (2004). Contracts (4th ed.). Aspen Publishers.

At a family gathering in 1869 an uncle, William E. Story, made his fifteen-year-old nephew an offer. If the boy would refrain from drinking liquor, using tobacco, swearing, and playing cards or billiards for money until he turned twenty-one, the uncle would pay him five thousand dollars.

The nephew did it. On his twenty-first birthday in 1875 he wrote to say so, and the uncle replied that the money was his and that he would hold it at interest until the young man was older. The uncle died in January 1887 without paying. The claim, which had by then passed to a third party, was brought against the executor, Sidway, who argued that the nephew had given up nothing of value and had in fact benefited from the arrangement, so there was no consideration and no contract.

The New York Court of Appeals disagreed in Hamer v. Sidway, 124 N.Y. 538 (1891). The nephew had a legal right to drink, smoke, and gamble, and he had restricted his lawful freedom at the uncle's request. Whether the uncle gained anything, and whether the nephew was better off, are not the court's business.

What matters here: Consideration is not about value. It is about exchange. Something must have been sought by the promisor and given by the promisee in return for the promise.

The test, in three parts

Section 71 of the Restatement (Second) of Contracts states the modern rule as a bargained-for exchange. Break it into three questions and apply them in order.

1. Is there a detriment or a benefit? The promisee must do, or promise to do, something they were not legally obliged to do, or refrain from something they were legally entitled to do. Forbearance counts, as Hamer shows.
2. Was it bargained for? The promisor must have sought it in exchange for the promise, and the promisee must have given it in exchange for the promise. This is what separates a contract from a gift with conditions attached. If I say I will give you my car and you must come to my house to collect it, coming to the house is a condition of the gift, not the price of the car.
3. Is it the price of this promise? Something already done cannot be the price of a promise made afterwards, because it was not given in exchange for it. That is the rule against past consideration.

Two things courts refuse to do. They do not weigh adequacy: if you sell a house for one dollar and mean it, that is a contract, and the traditional phrase is that a peppercorn will do. And they will not enforce a promise that recites a payment that was never made, because a sham recital fails the bargain test rather than the value test.

The pre-existing duty rule, and the hold-up problem

Doing what you are already legally bound to do is not consideration for a new promise. The reason is not formalism, it is leverage.

In Alaska Packers' Association v. Domenico, 117 F. 99 (9th Cir. 1902), a crew signed on in San Francisco to work a salmon season in Alaska for an agreed wage. Once at the remote cannery in Pyramid Harbor, with the season short and no replacement labour available for thousands of miles, they stopped work and demanded roughly double. The company's superintendent, having no alternative, signed. On their return the company paid the original rate. The court held the new promise unenforceable: the men had promised only what they already owed, and the situation was one in which the employer had no real choice.

The English case usually paired with it, Stilk v Myrick (1809), reached the same result when a captain promised to divide two deserters' wages among the remaining crew for sailing the ship home.

The rule has real costs. Circumstances genuinely change, and a contractor who discovers rock where the survey showed soil is not extorting anyone by asking for more. Two escape routes exist. Section 89 of the Restatement (Second) enforces a modification that is fair and equitable in view of circumstances the parties did not anticipate. And for the sale of goods, UCC section 2-209(1) says flatly that an agreement modifying a contract within Article 2 needs no consideration to be binding, leaving good faith to police the hold-up problem instead.

Key idea: The pre-existing duty rule is a blunt instrument aimed at coerced modifications, and both the Restatement and the code have replaced the blunt instrument with a good faith inquiry.

The illusory promise

A promise that leaves the promisor free to do nothing at all is no promise. If I agree to buy as many widgets as I feel like buying, I have not committed to anything, so there is no consideration on my side, and neither of us is bound.

Courts strain to avoid that conclusion, because parties who sign a document usually intend a deal. In Wood v. Lucy, Lady Duff-Gordon, 222 N.Y. 88 (1917), a fashion designer gave Wood the exclusive right to place her endorsements and market her designs, in exchange for half the profits. She then endorsed products herself and kept the money. She argued that Wood had never actually promised to do anything, so the agreement was unenforceable for want of consideration. Judge Cardozo, for the Court of Appeals, held that a promise to use reasonable efforts was implied from the structure of the arrangement: she had given him an exclusive agency, her sole compensation depended on his efforts, and he had undertaken to keep accounts and obtain patents. The promise was there even though nobody had written it down.

The same reasoning saves requirements and output contracts, where one side agrees to buy all it needs or sell all it makes, and the UCC ties those quantities to good faith so they are not illusory.

When a promise binds without consideration

Three routes matter.

Promissory estoppel. Section 90 of the Restatement (Second) makes a promise binding if the promisor should reasonably expect it to induce action or forbearance, it does induce such action, and injustice can be avoided only by enforcement. In Ricketts v. Scothorn, 57 Neb. 51 (1898), a grandfather gave his granddaughter a promissory note for two thousand dollars at six per cent, saying that none of his grandchildren worked and she should not have to either. She quit her job. He paid a year's interest and then died, and the estate refused the principal. There was plainly no bargain: he asked for nothing. The court enforced the note anyway, because he had intended her to give up her employment and she had. Note the remedy question that follows: the section allows a court to limit relief as justice requires, so reliance-based recovery may be smaller than the promise.

Material benefit received. In Webb v. McGowin, 168 So. 196 (Ala. App. 1935), Webb was working on the upper floor of a mill and, to keep a heavy pine block from falling on McGowin below, rode it down himself and was permanently disabled. McGowin promised to pay him fifteen dollars every two weeks for the rest of Webb's life, and did so for about eight years until his own death, after which his estate stopped. On a strict view this is past consideration and unenforceable. The court enforced it, reasoning that a promise made in recognition of a material benefit previously received by the promisor is binding to the extent necessary to prevent injustice. Not every state follows this.

Statutory substitutes. The firm offer under UCC 2-205 from the last lesson, modifications under 2-209, and in some states a signed writing or a surviving statutory seal.

Working a problem

A landlord and a tenant have a lease at two thousand dollars a month with eight months to run. The tenant's business is struggling. The landlord says in an email that he will accept fifteen hundred for the remaining months. The tenant pays fifteen hundred for three months, then the landlord sues for the shortfall.

Step one, consideration. What did the landlord get for giving up five hundred a month? On the face of it, nothing: the tenant already owed the rent and promised nothing new. Under the pre-existing duty rule the modification fails.

Step two, is this Article 2? No. A lease of real property is not a sale of goods, so 2-209 does not rescue it.

Step three, Restatement section 89. Were there unanticipated circumstances making the change fair and equitable? A general downturn the tenant simply misjudged is weaker than a sudden and specific event that neither party foresaw.

Step four, reliance. Did the tenant do something in reliance on the promise that it would be unjust to unwind, such as declining to exercise a break clause or turning down a cheaper unit? If so, promissory estoppel may bar the landlord from claiming the shortfall for the months already paid.

Notice that the answer depends on facts nobody has yet supplied, which is the normal condition of legal analysis. Notice also that this is an illustration of method and not advice on any real tenancy.

Common misconceptions

• "Consideration means the price must be fair." Courts do not weigh adequacy. A peppercorn is consideration; a sham recital of a payment never made is not.
• "A signed promise is enforceable because it is signed." Outside the statutory exceptions, a signature does not supply consideration. It supplies evidence.
• "Promissory estoppel gets you the full promise." Section 90 lets the court limit the remedy as justice requires, and reliance-based recovery is often less than expectation.
• "Doing more work justifies asking for more money." Only if you are not already bound to do it, or the case fits Restatement section 89 or UCC 2-209.

Pulling it together

• Consideration is a bargained-for exchange: a detriment or benefit, sought by the promisor and given in return for the promise.
• Hamer v. Sidway establishes that giving up a legal freedom is consideration even if the promisor gains nothing.
• Courts refuse to weigh adequacy but will reject a sham recital, and past consideration fails the exchange requirement.
• Performing a pre-existing duty is not consideration, as Alaska Packers shows, but Restatement section 89 and UCC 2-209 provide modern escape routes.
• An illusory promise is no consideration, and Wood v. Lucy shows courts implying a reasonable efforts promise to save a deal.
• Promissory estoppel under Restatement section 90 enforces a reliance-inducing promise, with the remedy limited as justice requires.

Sources

1. Legal Information Institute. (n.d.). Consideration. Cornell Law School. law.cornell.edu
2. Legal Information Institute. (n.d.). Promissory estoppel. Cornell Law School. law.cornell.edu
3. Legal Information Institute. (n.d.). U.C.C. section 2-209: Modification, rescission and waiver. Cornell Law School. law.cornell.edu
4. Hamer v. Sidway, 124 N.Y. 538 (1891). CourtListener. courtlistener.com
5. Alaska Packers' Association v. Domenico, 117 F. 99 (9th Cir. 1902). CourtListener. courtlistener.com
6. Wood v. Lucy, Lady Duff-Gordon, 222 N.Y. 88 (1917). CourtListener. courtlistener.com
7. Ricketts v. Scothorn, 57 Neb. 51 (1898). CourtListener. courtlistener.com
8. American Law Institute. (1981). Restatement (Second) of Contracts, sections 71, 89, and 90. American Law Institute Publishers.

On the evening of 20 December 1952, W. O. Lucy stopped at a restaurant in Dinwiddie County, Virginia, run by A. H. Zehmer. The two men had known each other for years, and Lucy had been trying to buy Zehmer's 471-acre Ferguson Farm for about eight years. Both had been drinking. Lucy offered fifty thousand dollars.

Zehmer wrote on the back of a restaurant check that he agreed to sell the Ferguson Farm complete for fifty thousand dollars. He then tore it up, rewrote it so that it read as an agreement by himself and his wife, and persuaded Mrs Zehmer to sign as well. The whole conversation lasted around forty minutes. Lucy put the paper in his pocket, offered five dollars to bind the deal, which Zehmer refused, and left. He then arranged a title examination and had a lawyer prepare to complete.

Zehmer refused to convey. His defence was that the whole thing was a joke, that both men had been drinking heavily, and that he had never intended to sell. The Supreme Court of Appeals of Virginia, in Lucy v. Zehmer, 196 Va. 493 (1954), enforced the sale.

So what?: The law of contract does not ask what you meant. It asks what a reasonable person in the other party's position would have understood you to mean from your words and conduct.

Why the law had to go objective

The older textbooks describe a contract as requiring a meeting of the minds. The phrase survives and the doctrine does not, for the simple reason that nobody can look inside another person's head, and a rule that turned on private intention would let anyone escape any agreement by testifying afterwards about what they had secretly meant.

So the test is external. Section 2 of the Restatement (Second) defines a promise as a manifestation of intention, and section 20 handles the case where two manifestations diverge. What matters is what was said and done, read as a reasonable person in the recipient's position would read it, knowing what that person knew.

Reread the Zehmer facts with that test. Forty minutes of discussion. A written memorandum. A rewritten memorandum to add a party. A signature obtained from a spouse. A refusal to take earnest money, which shows attention to the terms rather than indifference. The next morning Lucy consulted a lawyer. Every visible marker points to a serious transaction. What Zehmer felt about it was invisible and therefore irrelevant.

On the drinking, the court was clear that intoxication is not a get-out unless it is severe enough that the person could not comprehend the nature and consequences of the act. Being lubricated enough to do something you regret is not the standard.

The employee who was told not to worry

Embry v. Hargadine, McKittrick Dry Goods Co., 127 Mo. App. 383 (1907), is the cleanest statement of the principle because the words involved are so slight.

Embry's written employment contract had expired and he had been trying to get a renewal. He went to the company president, McKittrick, and said that unless he was given a contract for another year he would leave at once. By his account McKittrick replied that he should go ahead, that he was all right, that he should get his men out and not let it worry him. Embry went back to work. He was dismissed two months later.

McKittrick testified that he had not intended to make any contract and had merely been brushing off an interruption during a busy week. The court held that the trial judge had put the question the wrong way round. The issue is not what McKittrick intended. It is whether a reasonable person in Embry's position would have taken those words, in that context, as assent to a year's employment. If yes, there was a contract, whatever was in the president's mind.

The point: A party's private mental state becomes relevant only when the other side knows about it, or has reason to know about it. Then the recipient cannot claim to have reasonably understood something else.

When nobody's meaning wins

The objective test does not always produce a contract. Occasionally both parties attach different meanings to the same term and neither has reason to know of the other's meaning, and then there is nothing to enforce.

The example every law student meets is Raffles v Wichelhaus, decided in England in 1864. The parties contracted for a shipment of cotton to arrive from Bombay on a ship called the Peerless. There were two ships of that name sailing from Bombay, one in October and one in December. The buyer meant the October ship; the seller meant the December one. The court held there was no contract at all, because the ambiguity was latent and each party's understanding was equally reasonable.

Section 20 of the Restatement (Second) sets out the three outcomes.

• Neither knows of the other's meaning: no contract, because there is no shared manifestation to enforce.
• One knows, or has reason to know, of the other's meaning and the other does not: the contract is formed on the meaning of the party who did not know. The one with the better information does not get to profit from the confusion.
• Both know of the divergence: no contract, since neither can claim to have been misled.

Notice that the middle outcome is the common one in practice, and that it is not a compromise. It picks a winner, and it picks the less well-informed party.

Offers no reasonable person would take seriously

The objective test cuts both ways. If a reasonable person in the recipient's position would understand a statement as a joke, exaggeration, or advertising bluster, there is no offer, whatever the speaker privately intended.

The best-known modern example is a 1990s television commercial that showed teenagers collecting points for merchandise and ended with a fighter jet landing at a school, captioned as costing seven million points. A viewer assembled the points, sent in the order form with a cheque for the balance, and sued when the company declined to supply a military aircraft. The federal court in New York held that no objective, reasonable person could have understood the commercial as an offer, noting among other things that the jet was shown being flown to school by a teenager and that the aircraft was not available for civilian purchase.

Put that next to Zehmer. Same test, opposite result, because the surrounding circumstances differ. In Zehmer, everything visible pointed to seriousness. In the commercial, everything visible pointed to a joke.

A short procedure

When a party says they did not mean it, work through four questions in order.

1. What was manifested? List the words, the writing, the conduct, and the setting. Facts, not conclusions.
2. How would a reasonable person in the other party's position read that? Include what that person knew about the relationship and any prior dealings.
3. Did the other party know, or have reason to know, of the speaker's actual meaning? If so, the speaker's meaning governs.
4. Is this instead a mutual misunderstanding? If both meanings are reasonable and neither party knew of the other's, section 20 gives no contract.

This procedure will resolve most disputes about whether an agreement exists. What it never asks is what anybody sincerely felt, and that omission is the doctrine.

Common misconceptions

• "There is no contract without a meeting of the minds." The phrase survives in casual speech, but the operative test is the external manifestation of assent, judged by a reasonable person in the recipient's position.
• "I was drunk, so the contract is void." Intoxication excuses only where the person could not comprehend the nature and consequences of the transaction. Lucy v. Zehmer involved considerable drinking and an enforceable sale.
• "I was joking, so nothing happened." Only if a reasonable person would have seen the joke. Zehmer's claimed joke ran to a rewritten memorandum and a spouse's signature.
• "A misunderstanding always destroys the contract." Only where neither party knew of the other's meaning. Where one had reason to know, the contract stands on the other party's meaning.

What you now know

• Assent is judged objectively: what a reasonable person in the recipient's position would understand from words and conduct.
• Lucy v. Zehmer enforced a farm sale written on a restaurant check because every visible marker pointed to a serious deal.
• Embry shows that a few offhand words can form a contract if a reasonable listener would take them as assent.
• A party's undisclosed meaning matters only when the other side knows or has reason to know of it.
• Raffles and Restatement section 20 supply the three outcomes for divergent meanings, including formation on the less-informed party's terms.
• The same objective test defeats offers that no reasonable person would take seriously, as with the fighter jet advertisement.

Sources

1. Lucy v. Zehmer, 196 Va. 493 (1954). CourtListener. courtlistener.com
2. Embry v. Hargadine, McKittrick Dry Goods Co., 127 Mo. App. 383 (1907). CourtListener. courtlistener.com
3. Wikipedia contributors. (n.d.). Lucy v. Zehmer. en.wikipedia.org
4. Wikipedia contributors. (n.d.). Raffles v Wichelhaus. en.wikipedia.org
5. Legal Information Institute. (n.d.). Contract. Cornell Law School. law.cornell.edu
6. Leonard v. Pepsico, Inc., 88 F. Supp. 2d 116 (S.D.N.Y. 1999), affirmed 210 F.3d 88 (2d Cir. 2000).
7. American Law Institute. (1981). Restatement (Second) of Contracts, sections 2, 17, 20, and 201. American Law Institute Publishers.

Between 1957 and 1962 Ora Lee Williams bought furniture and appliances on instalment credit from the Walker-Thomas Furniture Company in Washington, D.C. She was raising seven children on public assistance of two hundred and eighteen dollars a month, and the store knew it, because its own records held her income details.

Every contract she signed contained the same clause. Each payment she made was spread pro rata across every item she had ever bought, so that no single item was ever fully paid off until the balance on all of them reached zero. The effect was that the store retained a security interest in everything, indefinitely.

In April 1962, when the balance from five years of purchases stood at about one hundred and sixty-four dollars, she bought a stereo set priced at five hundred and fourteen dollars and ninety-five cents. She then defaulted, and Walker-Thomas moved to repossess every item she had bought since 1957.

The trial court thought the clause harsh and enforced it anyway, saying relief was a matter for the legislature. In Williams v. Walker-Thomas Furniture Co., 350 F.2d 445 (D.C. Cir. 1965), Judge J. Skelly Wright disagreed, held that the courts had power to refuse enforcement of an unconscionable contract, and sent the case back for findings on whether these contracts were unconscionable.

Why this matters: Every doctrine in this lesson exists because assent that looks real on paper can be produced by incapacity, pressure, lies, or terms nobody could have understood, and enforcing it would make the law an instrument of the stronger party.

Void, voidable, and unenforceable

Three words that students use interchangeably and courts do not.

• Void: no contract ever existed. Nobody can enforce it, and nobody needs to do anything to escape it.
• Voidable: a contract exists, but one identified party has the power to undo it. That party may instead affirm the deal and hold the other side to it. Most of the defences in this lesson produce voidability, which is why it matters who is protected.
• Unenforceable: a valid contract that a court will not enforce, typically for want of a required writing or because a limitation period has expired.

The remedy for a voidable contract is rescission: the deal is unwound and each side returns what it received, which is a restitutionary idea rather than a compensatory one.

Capacity

Three groups lack full contractual capacity, and the protection is asymmetric in each case.

Minors. A contract with a person under eighteen is voidable by the minor, not by the adult. The minor may disaffirm during minority and for a reasonable time after reaching majority, and in most states must return whatever is left of what was received but is not liable for its depreciation. The obvious objection is that this lets teenagers wreck deals, and the obvious answer is that it makes adults careful about contracting with them. There is an exception for necessaries, food, shelter, basic clothing, and medical care, where the minor is liable for their reasonable value so that suppliers are not deterred from providing them.

Mental incapacity. Two tests coexist. The cognitive test asks whether the person could understand the nature and consequences of the transaction. The volitional test, adopted in the Restatement, adds that a person who understands but cannot act reasonably in relation to the transaction may avoid it if the other party had reason to know of the condition. If a court has already appointed a guardian, contracts are typically void rather than voidable.

Intoxication. Available only where the other party had reason to know that the person could not understand or act reasonably, which is a much narrower door than Lucy v. Zehmer defendants tend to hope.

Duress and undue influence

Physical duress, where someone takes your hand and moves the pen, produces a void agreement, because there was never any assent at all. It is rare.

The operative doctrine is duress by improper threat, and it has two elements. First, the threat must be improper: a crime or tort, a criminal prosecution, a bad-faith civil action, or a breach of the duty of good faith under an existing contract. Second, the threat must leave the victim no reasonable alternative. Note what is not enough. A threat to do something you are legally entitled to do, such as refusing to enter a new contract or suing on a genuine claim, is generally proper, even when it is unwelcome.

Economic duress is the commercially important form, and it links directly to the pre-existing duty rule from the last lesson. A supplier who threatens to stop shipping unless the price is raised, at a moment when the buyer cannot source elsewhere in time, is making an improper threat if the threat breaches the existing contract and the buyer has no realistic option. This is what the Alaska Packers facts look like when analysed as a defence rather than as a failure of consideration, and the two doctrines are aimed at the same problem from opposite ends.

Undue influence covers pressure that is not a threat at all: unfair persuasion by a person in a position of dominance or trust, typically a caregiver, a family member, an adviser, or a cleric. Courts look for a cluster of markers, including an unusual time or place for the transaction, insistence that the business be finished at once, emphasis on the consequences of delay, multiple persuaders against one, the absence of independent advice, and a statement that there is no time to consult anyone.

Misrepresentation, and the limits of caveat emptor

A misrepresentation is an assertion not in accord with the facts. To make a contract voidable it must be either fraudulent, meaning the maker knew or believed it was untrue and intended to induce reliance, or material, meaning it would likely induce a reasonable person to agree. It must actually induce assent, and the reliance must be justified.

Three practical points.

• Innocent misrepresentation still counts if material. You do not have to prove the other side lied to rescind. Damages in deceit are a different matter and do require fraud.
• Opinion is not fact, usually. A statement that a car is a great buy is opinion. A statement that it has never been in an accident is fact. The line moves where the speaker has special expertise or a relationship of trust.
• Silence is generally not misrepresentation, with growing exceptions. The old rule of caveat emptor let a seller stay quiet. Modern law imposes disclosure duties where the parties are in a relationship of trust, where a previous statement has become misleading, or where a defect is latent, dangerous, and known to the seller. Many states now require statutory disclosures on residential property sales, which is legislation doing what the common law was slow to do.

Mistake is the cousin of misrepresentation and involves nobody lying. A mutual mistake about a basic assumption on which the contract was made, with a material effect on the exchange, makes the contract voidable by the adversely affected party unless that party bore the risk. In Sherwood v. Walker, 66 Mich. 568 (1887), a cow named Rose 2d of Aberlone was sold for about eighty dollars on the shared assumption that she was barren and fit only for beef. Before delivery she turned out to be with calf and worth perhaps ten times as much. The Michigan Supreme Court allowed the seller to rescind, reasoning that the parties had contracted about a different animal in substance. The dissent is worth reading, because it argued that the buyer had taken a calculated risk on exactly that possibility, which is the modern way of framing the question: who bore the risk of the assumption being wrong.

Unilateral mistake, where only one party is wrong, is much harder. Relief generally requires either that enforcement would be unconscionable, or that the other party knew or had reason to know of the error, which is the familiar case of a contractor's bid containing an obvious arithmetic mistake.

Unconscionability, and its two prongs

Section 2-302 of the Uniform Commercial Code gives a court power to refuse to enforce a contract or a clause it finds unconscionable as a matter of law, and the same doctrine now runs through the common law generally. The code deliberately does not define the term.

Courts have settled on two prongs, and most jurisdictions require some of both.

Prong	What it looks at	Typical evidence
Procedural	The bargaining process	Take it or leave it form, no opportunity to read, fine print, absence of meaningful choice among suppliers, disparity in sophistication, language barriers
Substantive	The terms themselves	Terms that are grossly one-sided, a price far above market, a remedy stripped from one side only, a clause that operates as a trap

Run the Williams facts through it. Procedurally: a standard form, a clause whose operation is not obvious from reading it, a buyer whose limited alternatives the seller knew about. Substantively: a security interest that never released, so that a default on a stereo could take back a bed bought five years earlier. Judge Wright's formulation, an absence of meaningful choice together with terms unreasonably favourable to the other party, has been quoted ever since.

The doctrine attracts serious criticism. It is vague, and vagueness in a commercial rule has costs of its own. Judges asked to strike terms as unconscionable are making distributive judgments with limited information about the market. And a merchant who cannot enforce cross-collateral security may respond by refusing credit to exactly the customers the doctrine was meant to protect, which is an empirical claim that has never been settled either way. Both the doctrine and the criticism are live.

Remember: None of this is legal advice. The doctrines are stated here at the level of general principle, and each state's version differs in detail.

Common misconceptions

• "A contract signed under pressure is automatically void." Most defences make a contract voidable by the protected party, who may prefer to enforce it. Only physical compulsion produces a void agreement.
• "You can only get out of a contract if the other side lied deliberately." An innocent but material misrepresentation supports rescission. Fraud is needed for damages in deceit, not for undoing the deal.
• "If you signed it, you are bound by it." Signing is strong evidence of assent, and it is not conclusive against unconscionability, duress, incapacity, or misrepresentation.
• "Unconscionability means the deal was a bad bargain." A bad bargain is not enough. Courts generally look for both a defective process and grossly one-sided terms.

Summing up

• Void means no contract; voidable means one party may undo it; unenforceable means valid but not enforced. Rescission unwinds a voidable deal.
• Minors' contracts are voidable by the minor, with liability for the reasonable value of necessaries.
• Duress by improper threat requires both an improper threat and the absence of a reasonable alternative; undue influence covers unfair persuasion in relationships of trust.
• A misrepresentation must be fraudulent or material, must induce assent, and must be justifiably relied on; silence is increasingly actionable for latent dangerous defects.
• Sherwood v. Walker frames mutual mistake, and the modern question is which party bore the risk of the assumption failing.
• Unconscionability under UCC 2-302 combines a procedural and a substantive prong, was applied in Williams, and is criticised for vagueness and for possible credit effects.

Sources

1. Legal Information Institute. (n.d.). Unconscionability. Cornell Law School. law.cornell.edu
2. Legal Information Institute. (n.d.). Duress. Cornell Law School. law.cornell.edu
3. Legal Information Institute. (n.d.). Misrepresentation. Cornell Law School. law.cornell.edu
4. Legal Information Institute. (n.d.). Capacity. Cornell Law School. law.cornell.edu
5. Legal Information Institute. (n.d.). U.C.C. section 2-302: Unconscionable contract or clause. Cornell Law School. law.cornell.edu
6. Williams v. Walker-Thomas Furniture Co., 350 F.2d 445 (D.C. Cir. 1965). CourtListener. courtlistener.com
7. Sherwood v. Walker, 66 Mich. 568 (1887). CourtListener. courtlistener.com
8. American Law Institute. (1981). Restatement (Second) of Contracts, sections 151 to 177. American Law Institute Publishers.

In 1677 the English Parliament passed an Act for the Prevention of Frauds and Perjuries, usually cited as 29 Charles II chapter 3 and known ever since as the Statute of Frauds. It required certain kinds of promise to be evidenced by a signed writing, or they could not be sued upon.

The reason was a courtroom that would be unrecognisable today. Parties to a lawsuit were not permitted to testify, on the theory that their interest made them unreliable. Jurors could decide on their own knowledge of the neighbourhood. And a trade existed in professional witnesses who would swear to an oral agreement for a fee. In that setting, a rule that certain valuable promises must be written down was a sensible evidential filter.

England repealed most of it in 1954. Every American state has a version, and it is a live defence in litigation every day.

The core of it: The statute does not make oral contracts invalid. It makes certain oral contracts unenforceable in court, which is a defence somebody has to raise, and which can be defeated in several ways.

The six categories

The classic list is remembered by generations of students with the mnemonic MY LEGS. Each letter is a category of promise that must be in a signed writing to be enforceable.

Letter	Category	Why it is on the list
M	Marriage: promises made in consideration of marriage, such as a prenuptial agreement or a promise of a dowry	Large, long-lived, and easy to fabricate afterwards
Y	Year: a contract that cannot be fully performed within one year of its making	Memory of terms decays over long periods
L	Land: contracts for the sale of land or an interest in land, including most leases over a year	High value and a public record system that depends on documents
E	Executor: a promise by an executor to pay estate debts out of their own pocket	An unusual promise nobody would expect to be made casually
G	Goods: a contract for the sale of goods at a price of five hundred dollars or more, under the standard text of UCC section 2-201	Volume and value; the threshold has not been raised in most states and is now low
S	Suretyship: a promise to answer for the debt or default of another person	The promisor gets nothing, so a casual reassurance should not become a liability

Two notes on the last one. A guarantee is caught by the statute; a promise to pay your own debt is not, however it is phrased. And the main purpose rule takes a guarantee outside the statute when the guarantor's own economic interest is the main reason for the promise, as when a company director guarantees a supplier's invoice in order to keep his own business running.

The one year trap

The Y category catches more students in examinations than anything else in this course, because the test is not what actually happened. It is whether performance could be completed within one year of the moment the contract was made, on any possible course of events consistent with the terms.

Work three examples.

• A two-year employment contract. Within the statute. Two years of work cannot possibly be finished in one.
• Employment for life. Outside the statute, in most states, because the employee might die within the year, which would complete performance according to the terms. This looks morbid and it is the standard answer.
• A promise to build a house, expected to take eighteen months. Outside the statute if nothing in the contract prevents faster completion, because an unusually fast build would perform it inside a year. What matters is the terms, not the expectation.

One more trap: the year runs from the making of the contract, not from the start of performance. An agreement made in January to work for eleven months beginning in March cannot be completed until February of the following year, so it is within the statute.

What the writing has to say

The requirement is modest, and this surprises people expecting a formal document.

At common law the writing must reasonably identify the subject matter, indicate that a contract has been made, state the essential terms with reasonable certainty, and be signed by the party against whom enforcement is sought, called the party to be charged. It does not have to be a single document; several papers can be read together if they refer to each other. It does not have to be signed by both sides, which means an agreement can be enforceable against one party and not the other. A letter, an internal memorandum, or a note written to a third party can serve.

UCC section 2-201 relaxes this further for goods. The writing need only be sufficient to indicate that a contract was made and be signed by the party to be charged; it may omit or misstate terms, and the contract is then enforceable only up to the quantity stated. Quantity is the one term that must be there.

Section 2-201(2) adds a rule that catches out unwary merchants. If both parties are merchants and one sends a written confirmation of an oral deal, sufficient against the sender, the other merchant is bound too unless they object in writing within ten days of receipt. Silence completes the writing requirement against a party who never signed anything.

Bottom line: A signature and a scrap of paper naming the subject matter will usually do. The statute is an evidence rule, not a formality contest.

Electronic records satisfy it. The federal E-SIGN Act of 2000 and the Uniform Electronic Transactions Act, adopted by nearly every state, provide that a record or signature may not be denied legal effect solely because it is electronic. An exchange of emails with a typed name at the bottom has repeatedly been held to satisfy the statute.

Getting out of it

Because the statute can defeat a genuine agreement, courts have built a set of exits. The pattern in all of them is the same: conduct that is itself strong evidence a contract existed removes the need for the writing.

1. Full performance. A contract fully performed on both sides is not undone by the statute.
2. Part performance in land cases. Typically a combination of two or three of: the buyer taking possession, paying some or all of the price, and making substantial improvements. Payment alone is usually not enough, because money can simply be refunded.
3. Specially manufactured goods. Under 2-201(3)(a), where goods are made specially for the buyer, are not suitable for sale to others in the seller's ordinary business, and the seller has substantially begun making them.
4. Judicial admission. Under 2-201(3)(b), if the party being sued admits in a pleading, in testimony, or otherwise in court that a contract was made, the statute cannot be used to deny it. You may not swear that no deal existed and then rely on the absence of a writing.
5. Payment or acceptance. Under 2-201(3)(c), the contract is enforceable for goods that have been paid for and accepted, or received and accepted.
6. Promissory estoppel. Many states will enforce an oral agreement within the statute where one party reasonably relied to their serious detriment, though some refuse this on the ground that it would swallow the statute whole.

Is it worth keeping

The argument against the statute is that it now produces the mischief it was written to prevent. The courtroom conditions of 1677 are gone: parties testify, discovery produces documents and messages, and jurors decide on the evidence. What remains is a rule that lets a party who made a real oral promise walk away from it by pointing at the absence of a signature. That is the reasoning that led England, whose Parliament passed the original, to repeal most of it in 1954.

The argument for keeping it is that it is cheap and it works at the margin. It gives people a reason to write things down before committing to the largest transactions of their lives, it screens out some fabricated claims before trial, and its exceptions handle the cases where genuine reliance has occurred. Both positions are held by serious scholars, and no American state has repealed it outright.

Common misconceptions

• "An oral contract is not worth the paper it is written on." Most oral contracts are fully enforceable. Only the listed categories require writing.
• "A contract within the statute is void." It is unenforceable, and only if the defence is raised. Parties perform such contracts every day without incident.
• "The one year rule asks how long the job actually took." It asks whether the terms permit completion within a year of the making, which is why a lifetime contract falls outside it.
• "The writing must be a formal contract signed by both parties." A note, a letter, or a chain of emails signed by the party being sued can be enough, and under UCC 2-201 the quantity term is the only one that must appear.

The short version

• The Statute of Frauds of 1677 answered a courtroom in which parties could not testify and paid witnesses were a real problem.
• The categories are marriage, contracts not performable within a year, land, executors' personal promises, goods at five hundred dollars or more, and suretyship.
• The one year test looks at what the terms permit, not at what happened, and runs from the making of the contract.
• The writing must identify the subject matter, show a contract was made, and be signed by the party to be charged; UCC 2-201 requires only a quantity term, and electronic records qualify.
• Merchants who fail to object to a written confirmation within ten days are bound by it.
• Full performance, part performance in land, specially manufactured goods, judicial admission, payment and acceptance, and in many states promissory estoppel all defeat the defence.

Sources

1. Legal Information Institute. (n.d.). Statute of frauds. Cornell Law School. law.cornell.edu
2. Legal Information Institute. (n.d.). U.C.C. section 2-201: Formal requirements; statute of frauds. Cornell Law School. law.cornell.edu
3. Wikipedia contributors. (n.d.). Statute of Frauds. en.wikipedia.org
4. Legal Information Institute. (n.d.). Contract. Cornell Law School. law.cornell.edu
5. Farnsworth, E. A. (2004). Contracts (4th ed.), chapter 6. Aspen Publishers.
6. American Law Institute. (1981). Restatement (Second) of Contracts, sections 110 to 150. American Law Institute Publishers.

In May 1957 a Swiss firm, Frigaliment Importing, contracted to buy from a New York seller a quantity of what the documents described as US Fresh Frozen Chicken, Grade A, in two weight ranges: one lot at two and a half to three pounds, another at one and a half to two pounds.

The heavier birds that arrived in Switzerland were stewing chickens, older fowl suitable for soup. The buyer had expected young broilers and fryers. It sued, and Judge Henry Friendly's opinion in Frigaliment Importing Co. v. B.N.S. International Sales Corp., 190 F. Supp. 116 (S.D.N.Y. 1960), opens by announcing that the case requires the court to decide what is chicken.

He then worked the evidence in order. The dictionary supported the broad meaning, covering any bird of that species. Department of Agriculture regulations, incorporated into the contract by its own grading reference, defined chicken to include broilers, fryers, roasters, and stewing chickens alike. The seller was new to the trade and had asked the buyer's agent what was wanted, receiving an answer that used the general word. And, decisively, the contract price for the heavier birds was below the market price for broilers of that weight at the time, so the seller could not have made the deal on the buyer's meaning without losing money on every bird.

The buyer had a real case: in the trade, some witnesses said, chicken alone meant young chicken. But the burden was the buyer's, and Friendly held it had not carried it.

The point: Interpretation is not a search for the parties' feelings. It is a structured argument from the text outwards, and the party asserting the narrower meaning usually has to prove it.

The order of resort

Courts do not consider everything at once. They work through a ranked list, and knowing the ranking is most of the skill.

1. The express terms. What the document says, read as a whole, with each clause given effect if possible.
2. Course of performance. How these parties have actually behaved under this contract, especially where one side accepted the other's performance without objection. This is the strongest extrinsic evidence, because it shows what they thought they had agreed while the deal was running.
3. Course of dealing. How these parties behaved under previous contracts with each other.
4. Usage of trade. What the words mean to people in that trade or place, whether or not these parties knew it, provided the usage is regular enough that they should have known.

Section 1-303 of the Uniform Commercial Code sets that hierarchy out expressly, and the common law follows the same order. Where two conflict, the higher one wins, but courts will first try to read them as consistent.

Beneath these sit the interpretive canons, which are tie-breakers rather than rules.

• Contra proferentem: ambiguity is read against the party who drafted the clause. This does the heavy lifting in insurance and consumer cases.
• Specific over general, and handwritten or typed terms over printed boilerplate, because the more particular expression is likelier to represent this deal.
• Ejusdem generis: a general word at the end of a list takes its colour from the items listed.
• Expressio unius: listing some things suggests the exclusion of others not listed.
• Read the whole: prefer an interpretation that gives every clause work to do over one that makes a clause meaningless.

Plain meaning, and the argument against it

A long tradition holds that if the words are clear on their face, a court should apply them and refuse to hear evidence about what the parties meant. That is the plain meaning rule, and its virtue is predictability: a party who reads a contract can rely on it.

The counter-argument was put by Chief Justice Traynor of California in Pacific Gas and Electric Co. v. G. W. Thomas Drayage and Rigging Co., 69 Cal.2d 33 (1968). A contractor had agreed to indemnify the utility against all loss, damage, expense, and liability resulting from injury to property arising out of the work. During the work, a cover fell and damaged the utility's own steam turbine. The utility said the words plainly covered its property. The contractor said the clause was meant for third-party claims, and offered evidence of the parties' prior dealings and of the practice in the industry.

Traynor held the evidence admissible. His reasoning is that words have no meaning apart from the people using them, so a judge who declares a clause plain is really saying that it is plain to a judge, which is a report about the judge. The test he set is whether the contract language is reasonably susceptible of the meaning contended for. If it is, the extrinsic evidence comes in; the judge is not bound to accept it.

Both approaches survive in American law. New York courts apply a fairly strict plain meaning approach, California a contextual one, and most states sit somewhere between. Critics of the contextual approach point out that if any clause can be reopened with testimony about what was really meant, written contracts stop settling anything and litigation costs rise. That criticism has force, and it explains why the plain meaning rule has not disappeared.

The parol evidence rule

Now a different question. Not what the words mean, but whether a court may hear evidence of terms agreed outside the document at all.

The parol evidence rule provides that where the parties have adopted a writing as the final expression of their agreement, evidence of prior agreements, and of contemporaneous oral agreements, may not be used to contradict or, where the writing is complete, to supplement it.

Three things about it are counterintuitive and worth stating flatly.

• It is not a rule of evidence. It is a rule of substantive contract law about which terms are part of the contract at all.
• Parol does not mean oral. A prior written side agreement is equally excluded.
• It has no application at all to anything agreed after the writing. Later modifications are always admissible, subject to any no-oral-modification clause.

Why this matters: The rule turns on a single prior question, whether the writing was intended as final and, if so, how complete. Everything else follows from that finding.

Integration: the question that decides everything

A writing is integrated if the parties adopted it as final. It is completely integrated if they also adopted it as a complete and exclusive statement of the terms, and partially integrated if it is final as to what it covers but not exhaustive.

Status of the writing	Evidence to contradict	Evidence to add a consistent term
Not integrated	Admissible	Admissible
Partially integrated	Excluded	Admissible
Completely integrated	Excluded	Excluded

How does a court decide? Two schools, and this is where the same split reappears.

The four corners approach, associated with Williston, looks at the document itself. If it appears complete and, in particular, if it contains a merger clause stating that the writing is the entire agreement, it is completely integrated and the inquiry ends.

The contextual approach, associated with Corbin, asks whether the parties actually intended the writing to be exclusive, and lets the court look at the alleged extra term to decide. In Masterson v. Sine, 68 Cal.2d 222 (1968), a couple conveyed a ranch to relatives by a deed that reserved an option to repurchase. When the grantor went bankrupt, the trustee wanted to exercise the option and take the ranch for the creditors. The family argued the option had always been understood as personal to the family, so it could not be exercised by a trustee, and the deed said nothing either way. Traynor admitted the evidence, using the test of whether the alleged term is one that would naturally be made as a separate agreement by parties situated as these were. A deed is a formal instrument that rarely records collateral family understandings, so the silence proved nothing.

UCC section 2-202 codifies a version of this for goods: a confirmatory or final writing may not be contradicted, but it may always be explained or supplemented by course of dealing, usage of trade, and course of performance, and by consistent additional terms unless the writing was intended as complete and exclusive.

What the rule never excludes

These come in regardless of integration, and they are the practical answer to most parol evidence problems.

• Evidence that no contract was ever formed, or that it was void or voidable for fraud, duress, mistake, or illegality. A merger clause does not immunise a lie.
• Evidence of a condition precedent to the writing taking effect at all, for instance that the deal was not to operate until a third party approved.
• Evidence explaining an ambiguity, subject to whichever approach to plain meaning the state follows.
• Evidence of a collateral agreement supported by its own separate consideration.
• Evidence supporting reformation, where the writing fails to record what both sides actually agreed.
• Anything agreed after the writing.

Common misconceptions

• "Parol evidence means spoken evidence." It means evidence outside the writing, including prior written agreements.
• "A merger clause ends the argument." It is strong evidence of complete integration, decisive under a four corners approach and merely persuasive under a contextual one, and it never blocks evidence of fraud.
• "The parol evidence rule keeps out evidence of what happened later." It has nothing to say about subsequent modification.
• "If the words are clear, context is irrelevant." True in plain meaning states; in California and others, extrinsic evidence is admitted whenever the language is reasonably susceptible of the meaning urged.

Looking back

• Frigaliment shows interpretation as a ranked argument from dictionary, regulation, trade usage, and price, with the burden on the party asserting the narrower meaning.
• The order of resort is express terms, then course of performance, then course of dealing, then usage of trade, with canons such as contra proferentem as tie-breakers.
• The plain meaning rule buys predictability; Traynor's contextual test in Pacific Gas admits evidence whenever the language is reasonably susceptible of the meaning urged.
• The parol evidence rule is substantive, covers prior writings as well as oral terms, and does not reach later agreements.
• Integration is the pivotal finding: partially integrated writings may be supplemented, completely integrated ones may not.
• Fraud, duress, mistake, conditions precedent, ambiguity, collateral agreements, and reformation are outside the rule entirely.

Sources

1. Legal Information Institute. (n.d.). Parol evidence rule. Cornell Law School. law.cornell.edu
2. Frigaliment Importing Co. v. B.N.S. International Sales Corp., 190 F. Supp. 116 (S.D.N.Y. 1960). CourtListener. courtlistener.com
3. Pacific Gas and Electric Co. v. G. W. Thomas Drayage and Rigging Co., 69 Cal.2d 33 (1968). CourtListener. courtlistener.com
4. Masterson v. Sine, 68 Cal.2d 222 (1968). CourtListener. courtlistener.com
5. Wikipedia contributors. (n.d.). Parol evidence rule. en.wikipedia.org
6. Corbin, A. L. (1965). The interpretation of words and the parol evidence rule. Cornell Law Quarterly, 50(2), 161-190.
7. American Law Institute. (1981). Restatement (Second) of Contracts, sections 209 to 216. American Law Institute Publishers.

A contractor built a country residence in Westchester County, New York, for about seventy-seven thousand dollars. The specifications required that all wrought iron pipe be of Reading manufacture. When the house was finished the owner's architect discovered that some of the concealed plumbing was pipe of other makes, though of the same grade, quality, and appearance, made in mills that used the same processes.

The owner refused the final payment and demanded that the pipe be replaced. Doing that meant demolishing finished walls and rebuilding them, at a cost out of all proportion to any difference in the house.

In Jacob and Youngs, Inc. v. Kent, 230 N.Y. 239 (1921), Judge Cardozo held that the contractor had substantially performed, that the omission was neither wilful nor material, and that the owner's remedy was the difference in value between what he got and what he was promised, which on those facts was nominal. Judge McLaughlin dissented, on the ground that the owner had specified Reading pipe, had a right to Reading pipe, and should not be told by a court that something else was good enough.

Key idea: Contract law has to choose between two things it wants: holding people to what they promised, and refusing to inflict destruction disproportionate to the loss. Every doctrine in this lesson is a version of that choice.

Promises and conditions

A promise is an undertaking to do something; breaking it is a breach and gives rise to damages. A condition is an event that must occur before a duty to perform arises; if it does not occur, the duty never arises and there is no breach at all, by anyone.

The difference is severe in operation. If a term is a promise and it is broken slightly, the other side must still perform and sue for the small loss. If the same term is a condition and it is not met exactly, the other side's obligation simply never comes into existence.

Because that produces forfeiture, courts interpret doubtful language as a promise rather than a condition. Words such as if, provided that, on condition that, and unless and until create conditions; a bare undertaking usually does not. The Reading pipe specification could have been drafted as an express condition of payment, and if it had been, Cardozo's reasoning would have run into a wall.

Three kinds are worth naming. A condition precedent must occur before the duty arises, such as a mortgage approval clause in a house purchase. Concurrent conditions are mutual and simultaneous, which is the default in a sale: each side's tender is a condition of the other's. A condition subsequent discharges an existing duty when it occurs, which is rarer and usually appears in insurance policies.

Conditions can also be constructive, meaning supplied by the court rather than written by the parties. The most important is the rule that in an exchange where one side's performance takes time, that performance is a constructive condition of the other's payment: a builder must build before an owner must pay, unless the contract says otherwise.

Two escape valves exist. A party who has the benefit of a condition may waive it, expressly or by proceeding as though it were met. And a party who prevents a condition from occurring cannot rely on its non-occurrence, which is an application of the duty of good faith.

Satisfaction clauses

Contracts frequently make payment conditional on one party being satisfied. Courts split these into two categories, and the split is sensible.

• Commercial quality, fitness, or mechanical utility: the standard is objective. Satisfaction means whatever would satisfy a reasonable person, so a buyer cannot reject a functioning machine on a whim.
• Personal taste, aesthetics, or judgement: the standard is subjective, limited only by good faith. A portrait commissioner may reject a portrait they genuinely dislike, but not one they secretly like while pretending otherwise to escape the price.

How bad does a breach have to be

Any failure to perform is a breach. The question that matters is whether it is material, because a material breach lets the injured party suspend its own performance and, if uncured, treat the contract as at an end.

Section 241 of the Restatement (Second) lists the factors, and they are worth learning as a checklist rather than a formula.

1. How much of the expected benefit has the injured party been deprived of?
2. Can that party be adequately compensated in money for what is missing?
3. How much forfeiture will the breaching party suffer if the contract is terminated?
4. How likely is it that the breaching party will cure, given the circumstances?
5. Did the breaching party act in good faith and deal fairly?

Run the Reading pipe through it. The owner got a complete house with functioning plumbing of equivalent quality, so factor one is weak. Factor two is satisfied by a money difference. Factor three is enormous, since terminating or requiring replacement would destroy far more value than was lost. Factor four is irrelevant because the walls are up. Factor five favoured the contractor, since the substitution was found to be inadvertent. That is why the case came out as it did, and it also shows what would have flipped it: a deliberate substitution to save money would have failed on factor five, and courts routinely say so.

In short: Substantial performance is not a licence to do a rough job. It is a doctrine about proportion, and wilfulness removes its protection.

Goods are different: perfect tender

For sales of goods, UCC section 2-601 gives the buyer a much sharper right. If the goods or the tender fail in any respect to conform to the contract, the buyer may reject the whole, accept the whole, or accept any commercial units and reject the rest. That is the perfect tender rule, and it is a deliberate departure from substantial performance.

Three things soften it in practice. Section 2-508 gives the seller a right to cure, either within the time for performance or, where the seller had reasonable grounds to believe the tender would be acceptable, for a further reasonable time. Instalment contracts under 2-612 use a substantial impairment test rather than perfect tender. And a buyer who has accepted goods can revoke acceptance only on the narrower grounds in 2-608. The commercial reason for the harder rule is that goods are usually fungible and resellable, so rejection wastes much less than tearing out a wall.

Repudiation before the time for performance

A party sometimes makes clear in advance that it will not perform. That is anticipatory repudiation, and it requires a clear and unequivocal statement or an act making performance impossible; doubts, grumbling, and requests to renegotiate do not qualify.

When it happens, the injured party may treat the contract as breached and sue at once, or wait a commercially reasonable time to see whether the other side performs. What it may not do is sit indefinitely running up losses, because the duty to mitigate begins immediately. The repudiating party can retract, provided the other side has not already relied on the repudiation or cancelled.

Where a party is merely worried, UCC section 2-609 provides a tool: on reasonable grounds for insecurity, a party may demand adequate assurance of performance in writing and suspend its own performance until it arrives. Failure to provide assurance within a reasonable time, not exceeding thirty days, is itself a repudiation.

When performance is excused

Three related doctrines discharge a duty that has become extraordinary. All three require that the event was not the fault of the party seeking excuse, and that its non-occurrence was a basic assumption of the deal.

• Impossibility. Performance cannot be rendered at all: the specific subject matter is destroyed, the performer of a personal service dies, or performance becomes illegal.
• Impracticability. Performance is technically possible but only at excessive and unreasonable cost, and section 2-615 adopts this standard for goods. The bar is high. A price rise, even a steep one, is normally the kind of risk a fixed-price contract exists to allocate.
• Frustration of purpose. Performance remains possible and pointless. The classic case is the English coronation litigation of 1903, where rooms overlooking the processional route were hired at high prices and the procession was cancelled because the king fell ill. The rooms were still available; the reason for wanting them had vanished.

Courts apply all three narrowly, and the reason is structural. A contract is largely a device for allocating risk, so a doctrine that releases a party whenever things turn out badly would dissolve the instrument it is meant to interpret.

The upshot: Ask first whether the contract allocated this risk. Only if it did not do so, expressly or by implication, does the question of excuse arise.

Common misconceptions

• "Any breach lets you walk away." Only a material breach suspends your own performance and permits termination. A minor breach leaves you performing and suing for the difference.
• "Substantial performance means close enough is fine." It protects the inadvertent shortfall, not the deliberate one, and it applies to services and construction rather than to sales of goods.
• "A condition and a promise are the same thing in practice." A failed condition means no duty ever arose; a broken promise means damages. Courts prefer to read doubtful language as a promise to avoid forfeiture.
• "An unprofitable contract can be escaped as impracticable." Cost increases are usually the very risk the contract allocated. Impracticability requires far more than a bad bargain.

What to carry forward

• A condition determines whether a duty arises; a promise determines whether there is a breach, and courts prefer to find a promise to avoid forfeiture.
• Satisfaction clauses are read objectively for commercial quality and subjectively, limited by good faith, for matters of taste.
• Materiality is assessed on the Restatement section 241 factors, in which proportionality and good faith do most of the work.
• Jacob and Youngs awarded difference in value rather than cost of replacement, and Cardozo stressed that the substitution was not wilful.
• UCC 2-601 imposes perfect tender for goods, softened by the seller's right to cure and by the instalment rules.
• Anticipatory repudiation must be clear and unequivocal, permits immediate suit, and can be met with a demand for adequate assurance under 2-609.
• Impossibility, impracticability, and frustration are narrow, and the prior question is always which party the contract put the risk on.

Sources

1. Legal Information Institute. (n.d.). Breach of contract. Cornell Law School. law.cornell.edu
2. Legal Information Institute. (n.d.). Substantial performance. Cornell Law School. law.cornell.edu
3. Legal Information Institute. (n.d.). Condition precedent. Cornell Law School. law.cornell.edu
4. Legal Information Institute. (n.d.). U.C.C. section 2-615: Excuse by failure of presupposed conditions. Cornell Law School. law.cornell.edu
5. Jacob and Youngs, Inc. v. Kent, 230 N.Y. 239 (1921). CourtListener. courtlistener.com
6. Legal Information Institute. (n.d.). Frustration of purpose. Cornell Law School. law.cornell.edu
7. American Law Institute. (1981). Restatement (Second) of Contracts, sections 224 to 271. American Law Institute Publishers.

George Hawkins had a scar on the palm of his right hand from an electrical burn in childhood. Dr Edward McGee, who had done skin grafting during the First World War, repeatedly solicited the chance to operate, and told the boy's father that he would guarantee to make the hand a hundred per cent perfect. The operation took skin from Hawkins's chest and grafted it onto the palm. Hair grew on the palm.

The New Hampshire Supreme Court, in Hawkins v. McGee, 84 N.H. 114 (1929), treated the guarantee as a contractual promise rather than a matter of medical negligence, and then had to decide what that promise was worth. The trial judge had told the jury to consider the pain of the operation and the damage to the hand. That, the court held, was wrong.

The correct measure was the difference between the value of the perfect hand Hawkins had been promised and the value of the hand he actually ended up with, plus any incidental loss the operation caused. The pain of the operation itself was not recoverable, because Hawkins had agreed to undergo an operation; that pain was part of the price he had consented to pay for the promised result.

What matters here: Contract damages are calculated from the promised position, not from the position before the deal. That single sentence explains most of what follows.

Three interests, one set of facts

In a 1936 article that reorganised the whole field, Lon Fuller and William Perdue identified three distinct interests a court might protect when a contract is broken.

Interest	Aim	Question it answers
Expectation	Put the claimant where performance would have put them	What was the promise worth?
Reliance	Put the claimant back where they were before the contract	What did they spend or give up because of it?
Restitution	Strip the defendant of the benefit received	What did the breaching party get?

Work one set of facts through all three. You contract to buy a food van for forty thousand dollars, pay a ten thousand dollar deposit, spend three thousand on a licence and signage, and turn down another van at forty-five thousand. The seller refuses to deliver, and identical vans now cost fifty thousand.

• Expectation: you were promised a van worth fifty thousand for a price of forty thousand, so the value of the promise is ten thousand, plus your three thousand of wasted expenditure if it is not recoverable elsewhere.
• Reliance: the three thousand spent, and possibly the lost opportunity on the other van, aimed at restoring your pre-contract position.
• Restitution: the ten thousand deposit, which the seller must return whatever else happens, because it has no right to keep it.

These are not cumulative, and a claimant cannot stack expectation on top of reliance for the same loss. They are alternative ways of framing the claim, and reliance is the usual fallback when expectation cannot be proved with enough certainty.

Why expectation is the default

It is worth pausing on this, because the choice is not obvious. Tort law restores the pre-injury position; contract law goes further and delivers a position the claimant never actually occupied. Four reasons are usually given.

1. It is what was bought. The promise itself is the thing exchanged, so protecting anything less means the promisee did not get what they paid for.
2. It makes promises worth relying on. Commerce depends on people committing resources to future performance, which they will not do if a promise is worth only their out-of-pocket costs.
3. It prices breach correctly. A party deciding whether to break a contract faces the full value of the promise, not merely the other side's expenses. Holmes put the point sharply in 1897 when he described the duty to keep a contract at common law as amounting to a prediction that you must pay damages if you do not keep it, which is the root of the later argument about efficient breach: if a promisor can pay the full expectation measure and still be better off, the resources have moved to a higher-valued use and nobody is worse off.
4. It is often easier to measure than reliance, especially where market prices exist.

The efficient breach argument has serious critics, who reply that it treats promises as options to be bought out, ignores the real costs of litigation and of lost trust, and misdescribes what people believe they are doing when they make agreements. That argument runs into the moral question you will meet again in the last lesson of this course.

Four limits on expectation

Foreseeability comes from Hadley v Baxendale, decided in England in 1854. A crankshaft broke at a flour mill in Gloucester, and the millers engaged a carrier to take it to Greenwich as a pattern for a new one. The carrier delayed by several days, and the mill stood idle. The court refused the lost profits. Recoverable losses are those arising naturally from the breach, and those which both parties could reasonably have contemplated at the time of contracting as a probable result. The carrier had not been told that the mill would stop, and mills commonly kept spare shafts. This is the origin of the modern distinction between general damages, which follow naturally, and consequential damages, which require that the special circumstances were communicated.

Certainty requires proof of the amount with reasonable, not mathematical, precision. This is the barrier that defeats most claims by new businesses for lost profits, because there is no trading history from which to project. A claimant with an uncertain expectation often falls back on reliance for exactly this reason.

Mitigation, or the avoidable consequences rule, denies recovery for losses the claimant could have avoided by reasonable effort. A wrongfully dismissed employee must look for comparable work, though the leading American case on the point held that a film actress was not required to accept a different and inferior role in order to reduce the studio's liability. Mitigation is not a duty in the strict sense; nobody can be sued for failing to mitigate. It simply reduces what is recoverable.

Causation requires that the loss actually flow from the breach rather than from an independent cause.

The measure problem: cost of performance or difference in value

When defective or incomplete performance can be put right, the claimant may want the cost of putting it right, and the defendant will argue for the difference in market value. Usually these are close. Occasionally they diverge dramatically.

In Peevyhouse v. Garland Coal and Mining Co., 382 P.2d 109 (Okla. 1962), a farming family leased their land for strip mining and specifically bargained for remedial work at the end of the lease, restoring the surface. The company mined the coal and did not do the restoration. The cost of the remedial work was around twenty-nine thousand dollars. The increase in the market value of the farm if the work were done was about three hundred. The Oklahoma Supreme Court awarded three hundred dollars.

The decision is one of the most criticised in the American contracts canon. The objection is that the family had bargained specifically for the restoration, presumably accepting lower rent in exchange, and that measuring their loss by a market value they had no intention of realising defeats the whole point of the clause. Courts in other states have often reached the opposite result on similar facts, and the case sits in the casebooks alongside Jacob and Youngs as the two poles of the argument. Note that in Jacob and Youngs the shortfall was inadvertent and the demolition would have been pure waste; in Peevyhouse the promise was the whole reason for the clause and the breach was deliberate.

Bottom line: Ask whether the claimant would actually spend the money to complete the work. If they would, cost of performance protects a real expectation. If they would pocket it, difference in value is a fairer measure. Courts do not always ask this, and they should.

Reliance, restitution, and the surgeon's nose

Sullivan v. O'Connor, 363 Mass. 579 (1973), works through all three measures on one set of facts. An entertainer was promised an improved appearance from cosmetic surgery on her nose. Two operations failed to achieve it, a third made things worse, and her appearance ended up disfigured. The court reviewed the choices openly: expectation would give the difference between the promised nose and the one she has, which is speculative and would expose doctors to heavy liability from optimistic reassurances; restitution would give back only the fee. It allowed a reliance recovery that included her out-of-pocket costs, the worsening of her condition, and the pain and suffering of the third operation, which she had not bargained for, while excluding the pain of the first two, which she had.

Restitution deserves one further note. It is available even to a party who has breached. A builder who abandons a job halfway through is liable for damages, but may recover the value of the benefit conferred on the owner, less those damages, so that the owner is not left with a half-built extension for nothing.

Agreed damages, and orders to perform

Parties may fix damages in advance. A liquidated damages clause is enforceable where the anticipated harm was difficult to estimate at the time of contracting and the stated sum is a reasonable forecast of it. If the sum is set to frighten a party into performing rather than to compensate, it is a penalty and is void. Contract law does not permit private punishment.

Specific performance is available only where damages are inadequate. That is presumed for land, since every parcel is treated as unique, and it extends to unique goods, which UCC section 2-716 covers. Courts will not order performance of a personal service contract, both because supervising it is impractical and because compelling labour raises deeper objections; instead they may issue a negative injunction restraining the performer from working for a competitor during the term.

Punitive damages are generally unavailable in contract, however badly the breaching party behaved, unless the conduct is also an independent tort. Attorney's fees are usually borne by each side under the American rule, absent a statute or a contractual fee clause. Emotional distress damages are rarely available, with narrow exceptions where serious distress was a particularly likely result, as in contracts concerning funerals or weddings.

Remember: This is general legal education. Remedy rules vary considerably by state, and any real claim needs advice from a lawyer licensed where you are.

Common misconceptions

• "Damages punish the breaching party." They compensate. Punitive damages are essentially unavailable in contract unless an independent tort is proved.
• "You can recover all losses caused by a breach." Only those arising naturally or within the parties' contemplation at contracting, proved with reasonable certainty, and not reasonably avoidable.
• "A large fixed sum in the contract will be enforced because both sides signed it." Not if it operates as a penalty rather than a genuine pre-estimate of loss.
• "A court will order the other side to perform." Only where damages are inadequate, which normally means land or unique goods, and never for personal services.

Putting it together

• Expectation puts the claimant in the promised position; reliance restores the pre-contract position; restitution strips the defendant of benefits received.
• Hawkins v. McGee measured the loss as the difference between the promised hand and the actual one, excluding the pain the plaintiff had agreed to undergo.
• Expectation is the default because it is what was bought, it makes promises reliable, it prices breach correctly, and it is often easier to measure.
• Hadley v Baxendale confines recovery to losses arising naturally or within the parties' contemplation, which is the origin of the consequential damages rule.
• Certainty, mitigation, and causation cut the claim further, and reliance is the standard fallback when expectation cannot be proved.
• Peevyhouse awarded three hundred dollars where remedial work cost twenty-nine thousand, and is heavily criticised for measuring a bargained-for promise by a market value nobody intended to realise.
• Liquidated damages must be a reasonable forecast rather than a penalty, and specific performance requires that damages be inadequate.

Sources

1. Legal Information Institute. (n.d.). Expectation damages. Cornell Law School. law.cornell.edu
2. Legal Information Institute. (n.d.). Reliance damages. Cornell Law School. law.cornell.edu
3. Legal Information Institute. (n.d.). Consequential damages. Cornell Law School. law.cornell.edu
4. Legal Information Institute. (n.d.). Liquidated damages. Cornell Law School. law.cornell.edu
5. Hawkins v. McGee, 84 N.H. 114 (1929). CourtListener. courtlistener.com
6. Peevyhouse v. Garland Coal and Mining Co., 382 P.2d 109 (Okla. 1962). CourtListener. courtlistener.com
7. Sullivan v. O'Connor, 363 Mass. 579 (1973). CourtListener. courtlistener.com
8. Wikipedia contributors. (n.d.). Hadley v Baxendale. en.wikipedia.org
9. Fuller, L. L., and Perdue, W. R. (1936). The reliance interest in contract damages. Yale Law Journal, 46(1), 52-96.

On the evening of 26 August 1928 May Donoghue went to the Wellmeadow Cafe in Paisley, near Glasgow, with a friend. The friend bought her a bottle of ginger beer. The bottle was dark and opaque. Donoghue drank some, and when the rest was poured out, she said, the decomposed remains of a snail came with it. She became ill.

She had no contract with anyone. Her friend had bought the drink, so any contract claim belonged to the friend, and against the cafe rather than the manufacturer. Under the law as it stood, the manufacturer's obligations ran only to the person it had dealt with, a doctrine called privity. So Donoghue sued the manufacturer, David Stevenson, in tort instead.

In Donoghue v Stevenson, decided by the House of Lords in 1932, a majority held that a manufacturer of products sold in a form showing they will reach the consumer without intermediate examination owes that consumer a duty to take reasonable care. Lord Atkin grounded it in what became known as the neighbour principle: you must take reasonable care to avoid acts and omissions you can reasonably foresee would be likely to injure the persons so closely and directly affected that you ought to have them in contemplation.

Worth knowing: the case was decided on a preliminary point of law, on assumed facts. Whether there was ever a snail in the bottle was never determined, because Stevenson died and the claim settled.

The point: Tort duties reach people you have no relationship with, which is exactly why tort rather than contract governs most accidental injury.

The four elements

A negligence claim requires all four. Fail any one and the claim fails entirely, whatever the sympathy.

Element	What must be shown	Who decides
Duty	The defendant owed this claimant an obligation to take care	The judge, as a question of law
Breach	The defendant fell below the standard of a reasonable person in the circumstances	The jury, or the judge sitting without one
Causation	The breach caused the harm in fact, and the harm was within the scope of the risk	Mixed: factual cause for the jury, proximate cause partly for the judge
Damages	Actual harm, recognised by law	The jury, on evidence

Two consequences of that table are worth stating now. Because duty is a question of law, a judge can dismiss a claim before any jury hears it, which is where much of the real action in tort litigation happens. And because damages are an element rather than a remedy, there is no such thing as negligence without injury: a driver who runs a red light at eighty miles an hour and hits nothing has committed no tort at all.

Duty: the default and its exceptions

The default is broad. Everyone owes a duty of reasonable care to those who might foreseeably be injured by their conduct. If you act, you must act carefully.

The great exception is that there is generally no duty to rescue. A strong swimmer who watches a stranger drown, having done nothing to put them there, commits no tort under the traditional rule. This strikes most students as monstrous, and the standard justifications are worth hearing before deciding. Where would the line fall, given that thousands of people are in danger at any moment? Who among the many bystanders would be liable? How would a court measure the risk a rescuer must accept? And a legal duty to act positively is a heavier imposition on liberty than a duty to refrain from injuring. A minority of states have enacted limited duty-to-rescue statutes, usually requiring no more than calling for help, and most states have Good Samaritan statutes protecting those who do intervene from liability for ordinary negligence.

Duties to act positively do arise in defined situations.

• Special relationships: carrier and passenger, innkeeper and guest, employer and employee, school and pupil, custodian and prisoner, and in some states business and customer.
• Creating the peril: if your conduct, even innocently, put someone in danger, you must take reasonable steps to help.
• Voluntary undertaking: having begun a rescue, you must not abandon it unreasonably, especially if you have deterred others from helping.
• Control over a dangerous person. In Tarasoff v. Regents of the University of California, 17 Cal.3d 425 (1976), a patient told his university psychologist that he intended to kill a young woman, identifiable as Tatiana Tarasoff. Campus police detained and released him; nobody warned her or her family; he killed her two months later. The California Supreme Court held that a therapist who determines, or should determine, that a patient presents a serious danger of violence to another owes a duty to use reasonable care to protect the intended victim. The decision has been adopted, modified, and rejected across the states, and it remains contested precisely because it cuts against therapeutic confidentiality.

Special rules also attach to landowners, historically dividing entrants into invitees, licensees, and trespassers with descending levels of protection. Many states have merged the first two into a single reasonable care standard, and a few have abolished the categories altogether.

Breach, and two shortcuts

Breach means falling below the standard of care, and the next lesson is devoted to defining that standard. Two doctrines let a claimant establish it without the usual evidence.

Negligence per se. Where a defendant violated a statute, the violation establishes breach if the statute was designed to protect the class of persons the claimant belongs to, against the type of harm that occurred. A driver who exceeds a speed limit and hits a pedestrian is in breach without further argument: speed limits protect road users against collision injuries. But a driver who exceeds a speed limit and arrives in time to be struck by falling masonry is not, because the statute was not aimed at that risk. In most states the effect is conclusive on breach; in some it is only evidence of it, and excuses such as emergency or incapacity may be available.

Res ipsa loquitur, the thing speaks for itself. Where an accident of that kind does not ordinarily happen without negligence, and the instrumentality was in the defendant's exclusive control, and the claimant did not contribute to it, the jury may infer breach without direct evidence of what went wrong. The doctrine dates from Byrne v Boadle, decided in England in 1863, in which a barrel of flour rolled out of a first-floor warehouse window in Liverpool and landed on a passer-by. Byrne could not say what the workers had done wrong, because he had been unconscious and had never been inside. The court held that barrels do not fall out of windows unaided.

Key idea: Both doctrines exist because the defendant usually holds the information about what happened, and a rule that always demanded direct proof of the failure would reward the party who kept it hidden.

Causation, briefly, and damages

Causation splits in two, and the next lessons take each in turn. Cause in fact asks whether the harm would have happened but for the breach. Proximate cause asks whether this harm, to this claimant, falls within the scope of the risk that made the conduct careless in the first place. A defendant can plainly be a but-for cause and still escape liability on proximate cause grounds, which is what Module 5 explains.

On damages, the recoverable categories are conventionally split into economic losses, meaning medical costs, lost earnings, and property damage, which are documented and calculable, and non-economic losses, meaning pain, suffering, disfigurement, and loss of enjoyment, which are not. Two limits are worth flagging now. Claims for pure economic loss, financial harm with no accompanying physical injury or property damage, are generally barred, largely because such losses ripple indefinitely: when a bridge closes, every business in the town loses money. And claims for emotional distress alone are restricted, typically requiring the claimant to have been in a zone of physical danger or to have contemporaneously witnessed injury to a close relative.

What the defendant can say back

Two defences are worth previewing, because they change the shape of the whole case.

Comparative negligence reduces the claimant's recovery by their own share of the fault. It replaced the old rule of contributory negligence, under which a claimant even slightly at fault recovered nothing at all. California made the switch judicially in Li v. Yellow Cab Co., 13 Cal.3d 804 (1975), and almost every state has now abandoned pure contributory negligence, usually by statute. Systems divide into pure comparative negligence, where a claimant ninety per cent at fault still recovers ten per cent, and modified systems that bar recovery once the claimant's share reaches fifty or fifty-one per cent.

Assumption of risk bars or reduces recovery where the claimant knowingly and voluntarily accepted a specific risk. Express assumption, by a signed waiver, is enforceable within limits and is regularly struck down where it covers gross negligence or where the service is essential. Implied assumption, once a complete bar, has largely been absorbed into comparative fault in most states.

Why this matters: The move from contributory to comparative negligence changed outcomes more than almost any doctrinal shift in modern tort law, because it converted an all-or-nothing question into an apportionment.

Common misconceptions

• "Carelessness is enough to sue." All four elements are required, and the absence of actual harm defeats the claim entirely, however dangerous the conduct.
• "You have to help someone in danger." Morally, perhaps. Legally, there is generally no duty to rescue a stranger, though creating the peril, a special relationship, or beginning a rescue all change that.
• "Breaking a law automatically means you are liable." Negligence per se requires that the statute protect this class of person against this type of harm, and it establishes breach rather than the whole claim.
• "Res ipsa loquitur means the defendant must prove innocence." It permits an inference of breach; it does not reverse the burden of proof in most states, and the other elements still have to be proved.

Recap

• Donoghue v Stevenson established that a manufacturer owes a duty to the ultimate consumer, decided on assumed facts that were never tried.
• The four elements are duty, breach, causation, and damages, and failing any one defeats the claim.
• Duty is decided by the judge as a matter of law, which is why so many tort cases end before trial.
• There is generally no duty to rescue, subject to special relationships, creation of the peril, voluntary undertaking, and control over a dangerous person as in Tarasoff.
• Negligence per se and res ipsa loquitur are shortcuts to proving breach, each with defined requirements.
• Pure economic loss and standalone emotional distress claims are restricted, and comparative negligence has replaced the all-or-nothing contributory rule almost everywhere.

Sources

1. Legal Information Institute. (n.d.). Negligence. Cornell Law School. law.cornell.edu
2. Legal Information Institute. (n.d.). Duty of care. Cornell Law School. law.cornell.edu
3. Legal Information Institute. (n.d.). Negligence per se. Cornell Law School. law.cornell.edu
4. Legal Information Institute. (n.d.). Res ipsa loquitur. Cornell Law School. law.cornell.edu
5. Wikipedia contributors. (n.d.). Donoghue v Stevenson. en.wikipedia.org
6. Wikipedia contributors. (n.d.). Byrne v Boadle. en.wikipedia.org
7. Tarasoff v. Regents of the University of California, 17 Cal.3d 425 (1976). CourtListener. courtlistener.com
8. Li v. Yellow Cab Co., 13 Cal.3d 804 (1975). CourtListener. courtlistener.com

In January 1944 a tug called the Carroll was working in New York Harbor, shifting barges tied up at the Public Pier at the foot of a Manhattan street. To get at one barge, a deckhand readjusted the mooring lines of the others. The lines were made fast badly, and a barge called the Anna C broke free, drifted down the harbour, struck a tanker, was holed by its propeller, and sank with a cargo of flour belonging to the United States government.

The barge had an employee aboard, called a bargee, whose job was to be present and to raise the alarm. He had been ashore for twenty-one hours without permission or excuse, and later gave an explanation the trial court did not believe. Had he been aboard, the leak might have been reported and the barge saved.

The question for Judge Learned Hand in United States v. Carroll Towing Co., 159 F.2d 169 (2d Cir. 1947), was whether the barge owner was negligent in not having someone aboard. He answered it by writing down an equation.

So what?: Hand's move was to say out loud what courts had been doing implicitly for a century, which is to compare the cost of a precaution against the expected cost of the accident it would prevent.

Why the standard is objective

Before the formula, the standard. In Vaughan v Menlove, decided in England in 1837, the defendant built a hay rick close to the boundary of his land, next to his neighbour's cottages. He was warned repeatedly over five weeks that hay stacked wet can heat and ignite. He put a chimney through it and said he would chance it. It caught fire, and the cottages burned.

His defence was that he had used his own best judgement, and that he could not be liable for failing to exercise a degree of intelligence he did not possess. The court rejected it. The standard is that of a person of ordinary prudence, and it does not vary with the individual's own capacity for judgement.

Three reasons support that, and they are worth separating.

• Administrability. A subjective standard would require a jury to determine each defendant's actual mental capacity, which is close to impossible and easy to fake.
• Notice. People planning their conduct need to know what is expected. A single external standard tells them.
• Fairness to victims. The injured party did not choose the person who hit them, and should not have their compensation depend on that person's intelligence.

The reasonable person is not the average person. Average drivers speed and glance at their phones. The reasonable person is a construct describing the conduct the law expects, not the conduct that is statistically typical.

Where the standard bends

Characteristic	Treatment	Reason
Physical disability	Standard of a reasonable person with that disability	The condition is real, verifiable, and beyond the person's control
Children	Standard of a child of like age, intelligence, and experience	Judgement genuinely develops with age
Children in adult activities	Full adult standard, for driving, flying, operating machinery	Others on the road cannot see who is behind the wheel
Mental illness or low intelligence	Generally no adjustment for adults	Hard to verify, easy to assert, and it would shift the loss onto the innocent victim
Superior skill or professional training	Held to the standard of that profession or skill	Those relying on the professional are entitled to what the credential represents
Sudden emergency	Judged against a reasonable person facing the same emergency	Deliberation is not available in the moment, though the doctrine does not excuse causing the emergency

The mental illness row is the one students most often resist, and the criticism is legitimate: the rule can look like liability without any real fault. Its defenders answer that between two innocents, the loss should fall on the party whose condition caused it, and that verification problems are severe. Some states have begun to carve out exceptions, particularly for patients injuring their own caregivers.

What custom proves

Defendants often argue that they did what everyone in the industry does. Custom is evidence, and it is not conclusive. Learned Hand made the point in a 1932 case about tugs that had no radio receivers and so missed storm warnings, losing two coal barges. The industry did not commonly fit radios. Hand held that a whole calling may have unduly lagged in adopting available devices, and that courts must in the end decide what is required, whatever the industry's habits.

The exception is medical malpractice, where custom largely does set the standard: a doctor is judged against accepted professional practice, established by expert testimony, because courts recognise they lack the competence to design a better standard from scratch. Even there, a minority of states allow a jury to find an accepted practice negligent.

The formula

Hand's formulation compares three quantities.

• B, the burden of taking the precaution: the cost, inconvenience, and lost utility of doing the safer thing.
• P, the probability that the harm will occur if the precaution is not taken.
• L, the gravity of the resulting injury: how bad it is if it happens.

The defendant is negligent if B is less than P multiplied by L. In words: if the precaution costs less than the expected harm it prevents, failing to take it is unreasonable.

Apply it to the barge. B is the cost of keeping a bargee aboard during working hours in a crowded wartime harbour, which is modest, roughly one person's wages and inconvenience. P is meaningful, because barges in a busy harbour do break loose and lines are frequently readjusted by others. L is large: a sunk barge and a lost cargo. B came out well below P times L, so the owner was negligent, and because the tug was also at fault the loss was divided.

Two refinements do most of the real work.

The analysis is marginal. The question is never whether safety in general is worth having. It is whether this specific additional precaution was worth its specific additional cost, given everything already being done. A hospital that already has two backup generators is not negligent for lacking a third, even though the third would reduce risk further.

The values are usually not numbers. Courts rarely have figures for P or L, and juries are not handed spreadsheets. What the formula supplies is a structure for argument: name the precaution, estimate how likely and how serious the harm is without it, and ask whether a reasonable person would have thought it worth doing.

The core of it: Negligence is not carelessness in the abstract. It is the failure to take a specific precaution whose cost was less than the harm it would have been expected to prevent.

The case against the formula

The Hand formula is the foundation of the economic analysis of tort law, in which the aim of liability is to induce parties to take precautions worth taking and no more. Richard Posner and others built a large body of work on it. The criticisms are equally serious, and a student should be able to state them.

1. The numbers are usually unknowable. If P and L cannot be estimated, the formula is a rhetorical frame rather than a calculation, and dressing an intuition in algebra can make it look more rigorous than it is.
2. It requires putting a price on life and limb. Regulators do this openly and defensibly, but a jury asked whether a company was right to value a death at a particular figure tends to react to the pricing itself.
3. It ignores who bears what. B falls on the defendant and L falls on the victim. A rule that permits imposing a large risk on someone else because avoiding it was expensive for you is exactly the kind of transfer corrective justice theorists say tort law exists to reverse.
4. Courts often do not actually do it. Scholars including Richard Wright have argued that judicial reasoning about breach rarely tracks the formula, and that a standard of reasonable care to others explains the decided cases better.

The second criticism has a famous illustration. Ford's Pinto litigation of the 1970s is remembered for an internal cost-benefit memorandum comparing the cost of a fuel tank modification against a valuation of deaths and injuries. Two corrections belong with the story. The document most often quoted was a submission to a federal regulator about rollover fuel leakage across the industry, using a government-derived value of a life, rather than a Pinto-specific decision memorandum, and it was not admitted into evidence at the trial of Grimshaw v. Ford Motor Co., 119 Cal.App.3d 757 (1981). What the Grimshaw jury did hear was evidence about crash testing and about internal knowledge of the fuel system's behaviour, and it returned punitive damages of one hundred and twenty-five million dollars, which the trial judge reduced to three and a half million. The episode is still instructive, but it is often told inaccurately.

Common misconceptions

• "The reasonable person is the average person." The standard is normative rather than statistical. Common conduct can be negligent conduct.
• "Doing what the industry does is a complete defence." Custom is evidence. A whole calling may have lagged, as the tugboat radio case held. Medical practice is the main exception.
• "The Hand formula requires actual numbers." Courts almost never have them. It supplies the structure of the argument about breach, not an arithmetic answer.
• "Ford calculated that burned drivers were cheaper than a fix, and the jury saw the memo." The much-quoted document concerned industry-wide rollover standards for a regulator and was not admitted at the Grimshaw trial.

The takeaway

• Vaughan v Menlove rejected a subjective standard: the test is the person of ordinary prudence, whatever this defendant's own judgement.
• The standard is adjusted for physical disability, childhood, superior skill, and sudden emergency, and generally not for adult mental illness.
• Custom is evidence rather than a defence, except in medical malpractice where accepted practice largely sets the standard.
• Hand's formula in Carroll Towing asks whether the burden of a precaution was less than the probability of harm multiplied by its gravity.
• The analysis is marginal, asking about the next precaution rather than about safety in general, and its inputs are rarely quantified.
• Serious criticisms concern unknowable inputs, pricing life, the distribution of B and L between defendant and victim, and whether courts reason this way at all.

Sources

1. Legal Information Institute. (n.d.). Reasonable person. Cornell Law School. law.cornell.edu
2. United States v. Carroll Towing Co., 159 F.2d 169 (2d Cir. 1947). CourtListener. courtlistener.com
3. Wikipedia contributors. (n.d.). Vaughan v Menlove. en.wikipedia.org
4. Wikipedia contributors. (n.d.). United States v. Carroll Towing Co. en.wikipedia.org
5. Grimshaw v. Ford Motor Co., 119 Cal.App.3d 757 (1981). CourtListener. courtlistener.com
6. Wikipedia contributors. (n.d.). Ford Pinto. en.wikipedia.org
7. Wright, R. W. (2003). Hand, Posner, and the myth of the Hand formula. Theoretical Inquiries in Law, 4(1), 145-273.

On the morning of 24 August 1924 Helen Palsgraf was standing on the platform at the East New York station of the Long Island Rail Road with her two daughters, waiting for a train to Rockaway Beach. A train pulled out. A man carrying a package wrapped in newspaper, about fifteen inches long, ran and jumped for the moving car. He seemed about to fall. A guard on the car reached out and pulled him in; another guard on the platform pushed him from behind. The package was dislodged and fell on the rails.

It contained fireworks. It exploded. At the other end of the platform, some distance away, a set of scales was overturned and struck Mrs Palsgraf.

She sued the railroad, and won at trial and in the intermediate appellate court. In Palsgraf v. Long Island Railroad Co., 248 N.Y. 339 (1928), the New York Court of Appeals reversed by four votes to three. Judge Cardozo held that the guards, in helping a passenger with an ordinary looking parcel, had no reason to foresee any risk to a woman standing yards away, and that a defendant owes a duty only within the range of apprehension. Wrong to her was not established by wrong to somebody else.

Judge Andrews dissented. In his view a person who acts carelessly owes a duty to everyone who is in fact injured, and the real question is one of proximate cause: whether the law, for practical reasons, will trace responsibility this far. He offered a list of considerations rather than a test, including whether there was a natural and continuous sequence, how remote the injury was in time and space, and whether the result was too attenuated for a court to be concerned with.

Why this matters: The two opinions are not really disagreeing about the facts. They are disagreeing about where in the analysis the limit on liability belongs, and both frameworks are still used in American courts.

Two distinct questions

Causation is always two inquiries and students who merge them get lost.

• Cause in fact is an empirical question about the world. Did this conduct make a difference to what happened?
• Proximate cause, better called scope of liability, is a normative question about responsibility. Given that the conduct did make a difference, should the law hold this defendant answerable for this consequence?

A defendant can lose the first and win the second. Someone whose careless driving delays a family by ten minutes, so that they are in the path of a falling tree they would otherwise have missed, is a but-for cause of their injuries and will not be liable for them.

Cause in fact, and the cases that break it

The standard test is but for: would the harm have occurred anyway, without the defendant's breach? If yes, the breach caused nothing. A landlord who fails to install a fire escape is not liable to a tenant who died instantly in the initial explosion and could never have reached it.

The test fails in three recurring situations, and the law has a device for each.

Multiple sufficient causes. Two fires, negligently started by different defendants, converge and burn a house. Each defendant can truthfully say the house would have burned without them, so the but-for test acquits both, which is absurd. Courts substitute the substantial factor test: a defendant whose conduct was a substantial factor in bringing about the harm is a cause, whether or not it was necessary.

Unidentifiable defendant among a small group. In Summers v. Tice, 33 Cal.2d 80 (1948), two hunters negligently fired in the plaintiff's direction at the same moment. One pellet struck his eye. Both had been careless in identical ways, and it was impossible to say whose shot did it. The California Supreme Court shifted the burden: where two or more defendants have been negligent and only one caused the harm, each must exonerate himself, and if neither can, both are liable. This is alternative liability, and it is confined to small groups where all the possible causers are before the court.

The related device of market share liability was developed for the drug diethylstilbestrol, prescribed to pregnant women for decades and later linked to cancers in their daughters. Decades on, no claimant could identify which manufacturer made the pills her mother took. In Sindell v. Abbott Laboratories, 26 Cal.3d 588 (1980), the California Supreme Court apportioned liability among manufacturers by their share of the relevant market. Several states have refused to follow it.

The defendant controls the information. In Ybarra v. Spangard, 25 Cal.2d 486 (1944), a patient went in for an appendectomy and woke with a paralysing injury to his shoulder, an area that had nothing to do with the operation. He was unconscious throughout and could not say who among the surgeons, nurses, and anaesthetists had done it, or how. The court applied res ipsa loquitur against all of those who had control of his body, requiring each to explain. Without that move, the very unconsciousness that made the negligence possible would have made it unprovable.

The point: Each of these devices exists because the ordinary rule, applied literally, would let defendants escape precisely because their carelessness destroyed the evidence.

Proximate cause: two ways to draw the line

The older approach, associated with an English case of 1921, asked about directness: a defendant is liable for all consequences flowing directly from the breach, foreseeable or not.

The modern approach is foreseeability, and it was settled for the common law world by the Privy Council in the Wagon Mound litigation. In 1951 a ship of that name, chartered in Sydney Harbour, negligently discharged furnace oil, which spread across the water to a wharf where welding was in progress. The wharf owners stopped work, took advice that the oil would not ignite on water, and resumed. Molten metal fell, ignited cotton waste floating in the oil, and the wharf burned. In Overseas Tankship (UK) Ltd v Morts Dock and Engineering Co, decided in 1961, the Privy Council held that liability extends only to damage of a kind that was reasonably foreseeable. Fouling of the wharf by oil was foreseeable; destruction by fire, on the evidence as the court took it, was not.

The Restatement (Third) of Torts reframes the same idea as scope of liability: a defendant is liable for harms that result from the risks that made the conduct tortious. Ask what made the act careless, then ask whether this is the sort of harm that reason was worried about. A pharmacist who dispenses pills in an unlabelled container is careless because someone may take the wrong medicine; if the container is thrown and breaks a window, that harm is outside the scope of the risk.

Three rules that decide most disputes

The eggshell skull rule. You take your victim as you find them. If the kind of harm was foreseeable, the defendant is liable for its full extent, even where an unusual susceptibility makes it far worse than anyone could have predicted. A light blow that would bruise most people, but kills a haemophiliac, produces full liability. Note the structure carefully: foreseeability governs the type of harm, and not its magnitude.

Intervening and superseding causes. An intervening act occurs after the breach and contributes to the harm. It is superseding, breaking the chain, only if it was unforeseeable. Foreseeable intervening acts do not help the defendant, and the standard list of foreseeable interventions is long: negligent medical treatment of the injury, the ordinary carelessness of other road users, and rescue attempts. Cardozo made the last point in a 1921 case about a man injured going back along a trestle to look for a cousin thrown from a train, writing that danger invites rescue and that the wrong which imperils life is a wrong to the rescuer too. Deliberate criminal conduct by a third party is more often superseding, unless the very risk created was that such conduct would occur, as with a hotel that leaves rooms unlockable.

Joint and several liability. Where several defendants cause a single indivisible harm, the traditional rule lets the claimant recover the whole sum from any one of them, leaving that defendant to seek contribution from the others. It protects claimants when one defendant is insolvent and shifts that risk onto the solvent co-defendant, which is why many states have modified or abolished it in favour of proportionate several liability. This is one of the most heavily legislated corners of tort law, and it varies enormously by state.

Working the framework

A supermarket leaves a spill unmopped for an hour. A customer slips and breaks a hip. At the hospital, a surgeon negligently operates and the customer loses the leg.

Cause in fact: but for the spill, no fall, no operation, no lost leg. Satisfied.

Scope of the risk: the reason an unmopped spill is careless is that people fall and are injured. A broken hip is squarely within that risk.

Intervening cause: negligent medical treatment of an injury the defendant caused is one of the classic foreseeable interventions, so the supermarket remains liable for the amputation as well, and the surgeon is liable too.

Change one fact: suppose the ambulance is struck by a meteorite. That is not a foreseeable consequence of leaving a floor wet, and it supersedes.

Now notice something about Palsgraf. Under Cardozo's framing the case never reaches this analysis, because there was no duty to Mrs Palsgraf at all. Under Andrews's framing there was a duty, and the case turns on whether a court will trace the sequence from a shove to an explosion to falling scales. Same facts, same result on those facts, different doctrinal machinery, and the choice matters in the harder cases that follow.

Common misconceptions

• "If the defendant caused it, they are liable." Factual causation is necessary and not sufficient. Scope of liability is a separate normative question.
• "Proximate means nearest in time or space." It means within the scope of the risk that made the conduct wrongful. A remote consequence can be within it; an immediate one can be outside it.
• "An unforeseeable extent of injury defeats the claim." The eggshell skull rule says otherwise. Foreseeability applies to the kind of harm, not to how bad it turns out.
• "Any later act by someone else breaks the chain." Only an unforeseeable one. Negligent medical treatment and rescue attempts are treated as foreseeable.

Where this leaves us

• Cause in fact is empirical and normally tested by asking whether the harm would have happened but for the breach.
• Substantial factor, alternative liability from Summers v. Tice, market share liability, and the Ybarra use of res ipsa exist for cases the but-for test cannot handle.
• Proximate cause is a normative limit, framed by Palsgraf as duty by Cardozo and as scope of liability by Andrews.
• The Wagon Mound established foreseeability of the kind of damage as the modern test, and the Restatement (Third) recasts it as harm within the scope of the risk.
• The eggshell skull rule makes the extent of harm irrelevant once the kind is foreseeable.
• Intervening acts supersede only when unforeseeable, and joint and several liability has been modified or abolished in many states.

Sources

1. Legal Information Institute. (n.d.). Proximate cause. Cornell Law School. law.cornell.edu
2. Legal Information Institute. (n.d.). Eggshell skull rule. Cornell Law School. law.cornell.edu
3. Legal Information Institute. (n.d.). Joint and several liability. Cornell Law School. law.cornell.edu
4. Palsgraf v. Long Island Railroad Co., 248 N.Y. 339 (1928). CourtListener. courtlistener.com
5. Summers v. Tice, 33 Cal.2d 80 (1948). CourtListener. courtlistener.com
6. Ybarra v. Spangard, 25 Cal.2d 486 (1944). CourtListener. courtlistener.com
7. Wikipedia contributors. (n.d.). Overseas Tankship (UK) Ltd v Morts Dock and Engineering Co Ltd. en.wikipedia.org
8. Sindell v. Abbott Laboratories, 26 Cal.3d 588 (1980), on market share liability for diethylstilbestrol.

Gladys Escola was a waitress at a restaurant in Merced, California. On a summer day in 1941 she was transferring bottles of Coca-Cola from a case into the restaurant's refrigerator. She had lifted the bottle about eighteen inches when it broke in her hand. The break was clean and the glass severed blood vessels, nerves, and muscles of her thumb and palm, leaving a cut about five inches long.

She could not prove what the bottler had done wrong. She did not know how the bottle had been charged, whether it had been over-pressurised, whether the glass was flawed, or whether it had been damaged in handling. The California Supreme Court let her recover in Escola v. Coca Cola Bottling Co. of Fresno, 24 Cal.2d 453 (1944), by applying res ipsa loquitur.

Justice Roger Traynor agreed with the result and refused the reasoning. In a separate concurrence he argued that a manufacturer should be liable without any finding of negligence when an article it places on the market, knowing it will be used without inspection, proves to have a defect that causes injury. Public policy, he wrote, demanded that responsibility be fixed where it would most effectively reduce the hazards to life and health inherent in defective products that reach the market.

Nineteen years later that concurrence became the law of California, and then of most of the country.

Key idea: Strict liability is not liability for accidents in general. It is a decision that in specific settings the question of fault produces the wrong answers, so the law stops asking it.

Where fault was abandoned first: dangerous activities

In the 1860s John Rylands built a reservoir on his land in Lancashire, employing competent contractors. Beneath it lay disused mine shafts nobody had properly investigated. When the reservoir filled, water broke through into the workings and flooded a neighbouring colliery operated by Thomas Fletcher.

Rylands himself had not been careless; the contractors may have been, but they were independent. In Rylands v Fletcher, decided finally by the House of Lords in 1868, the courts held him liable anyway. The rule, as Blackburn J framed it in the court below, is that a person who for his own purposes brings onto his land and keeps there anything likely to do mischief if it escapes must keep it at his peril, and is answerable for the natural consequences of its escape. Lord Cairns added the limiting idea of a non-natural use of land.

American law absorbed this as strict liability for abnormally dangerous activities. The Restatement (Second) of Torts set out six factors: whether the activity involves a high degree of risk, whether the harm would be great, whether the risk can be eliminated by reasonable care, whether the activity is a matter of common usage, whether it is appropriate to the place where it is carried on, and its value to the community. The Restatement (Third) compresses these to two: a foreseeable and highly significant risk that reasonable care cannot eliminate, and an activity that is not one of common usage.

Blasting is the standard example. So are storing large quantities of explosives, crop dusting with toxic chemicals, and keeping wild animals. Driving is not, despite killing far more people, because it fails the common usage requirement and because reasonable care substantially reduces the risk. Commercial aviation was once treated as abnormally dangerous and no longer is.

The third factor is the one that carries the doctrine. If reasonable care would eliminate the danger, negligence is an adequate tool. Strict liability is for the residual risk that survives all the care anyone can take, and its function is to decide who bears that residue: the person who chose to create it for their own benefit, or the neighbour who did not.

The wall that had to fall: privity

For most of the nineteenth century a manufacturer owed duties only to the person it dealt with. An English case of 1842 held that a coachman injured when a mail coach collapsed could not sue the contractor who had undertaken to keep it in repair, because he was not a party to that contract. The consequence in an industrial economy was that the further a product travelled from its maker, the safer the maker became.

The wall came down in MacPherson v. Buick Motor Co., 217 N.Y. 382 (1916). Buick sold a car to a retail dealer, who sold it to Donald MacPherson. One of the wooden wheels, bought in by Buick from a component supplier, was made of defective wood. It crumbled while the car was in motion and MacPherson was thrown out and injured. Buick argued that its duty ran only to the dealer.

Cardozo rejected it. If the nature of a thing is such that it is reasonably certain to place life and limb in peril when negligently made, then it is a thing of danger, and the manufacturer owes a duty to those who will foreseeably use it, whatever the chain of sale. He also held that Buick could not escape by pointing at its supplier; it was obliged to inspect what it bought in.

The upshot: MacPherson did not create strict liability. It removed privity from negligence, so an injured consumer still had to prove the manufacturer was careless, which in a modern factory is close to impossible from outside.

Greenman, and section 402A

In 1955 William Greenman's wife bought him a combination power tool that could be used as a saw, drill, and wood lathe. While using it as a lathe, a piece of wood flew out of the machine and struck him on the forehead, causing serious injury. He produced evidence that the set screws holding parts of the machine together were inadequate, so that normal vibration caused the piece to move.

In Greenman v. Yuba Power Products, Inc., 59 Cal.2d 57 (1963), Traynor, now Chief Justice, wrote for the court and adopted his own concurrence from Escola. A manufacturer is strictly liable in tort when an article it places on the market, knowing that it is to be used without inspection for defects, proves to have a defect that causes injury. Two years later the American Law Institute published section 402A of the Restatement (Second) of Torts, stating the rule for products in a defective condition unreasonably dangerous to the user, and adding that it applies although the seller has exercised all possible care. Almost every state adopted some version within a decade.

The reasons given for the shift are worth listing, because each is contestable.

• Evidence. The manufacturer holds the information about its own processes; the consumer cannot obtain it.
• Prevention. The manufacturer is best placed to design and inspect out the defect, and liability gives it the incentive.
• Loss spreading. A manufacturer can distribute the cost across the price of every unit and through insurance, converting a catastrophic loss for one family into pennies for many buyers.
• Reliance. Consumers buy sealed goods they cannot inspect, on the strength of the maker's name.

Three kinds of defect

Type	What it means	Test
Manufacturing defect	This unit departs from the maker's own intended design	Compare the item to the specification; genuinely strict
Design defect	Every unit is built as intended, and the design itself is unsafe	Consumer expectations, or risk-utility with a reasonable alternative design
Warning defect	The product is safe if used correctly but the risk was not adequately communicated	Whether a reasonable warning would have reduced the risk, and whether it reached the right person

Only the first is strict in a pure sense. For design defects, states split. The consumer expectations test asks whether the product was more dangerous than an ordinary consumer would expect, which works for a car that catches fire in a low-speed collision and fails for complex machinery about which ordinary consumers have no expectations. The risk-utility test weighs the danger against the usefulness of the design and asks whether a reasonable alternative design was available at acceptable cost. The Restatement (Third) of Torts on Products Liability, published in 1998, adopts the alternative design requirement, and it was and remains controversial: critics say it smuggles negligence back into a doctrine created precisely to escape it, and that requiring a claimant to design a better product is a heavy burden.

Warning cases carry their own rules. The learned intermediary doctrine lets a drug manufacturer discharge its duty by warning the prescribing physician rather than the patient, and it has been narrowed where drugs are advertised directly to consumers. Comment k to section 402A protects unavoidably unsafe products, such as some drugs and vaccines, from design defect liability where they are properly prepared and accompanied by adequate warnings.

What the defendant can still say

Strict liability is not absolute liability, and the defences matter.

• The product was not defective when it left the defendant's hands, which puts alteration and poor maintenance in issue.
• Misuse, if the use was not reasonably foreseeable. Foreseeable misuse, including the standard uses a manufacturer knows people make of a thing, does not help.
• Comparative fault, which most states now apply to reduce recovery in products cases as elsewhere.
• State of the art, meaning the danger was not knowable given the scientific knowledge at the time of manufacture. This is decisive in some states and rejected in others as inconsistent with strict liability.
• Federal preemption, where a federal regulatory scheme displaces state tort claims. This is now a large and technical body of law of its own, particularly for medical devices and pharmaceuticals.

Worth holding on to: The criticism of strict products liability is that it raises prices, deters useful products, and taxes buyers who are never injured to compensate those who are. Vaccines are the clearest case: litigation in the 1980s drove manufacturers out of the market, and Congress responded with the National Childhood Vaccine Injury Act of 1986, creating a no-fault compensation programme funded by an excise tax. That statute is a direct admission that ordinary tort liability was producing an outcome nobody wanted, and it is the strongest single piece of evidence the critics have.

Common misconceptions

• "Strict liability means liable for everything." The claimant still proves a defect, causation, and damages, and the defences remain available.
• "MacPherson created strict products liability." It abolished the privity requirement in negligence. Strict liability arrived with Greenman in 1963 and section 402A in 1965.
• "Driving is abnormally dangerous because so many people die." Total harm is not the test. Common usage and the effectiveness of reasonable care both exclude it.
• "A design defect claim just needs proof that someone was hurt." Most states require either a consumer expectation the product defeated, or a reasonable alternative design that was available and would have avoided the harm.

What to remember

• Rylands v Fletcher imposed liability without fault for the escape of something dangerous brought onto land for the defendant's own purposes.
• Abnormally dangerous activity liability turns above all on whether reasonable care can eliminate the risk, and on common usage.
• Privity shielded manufacturers until MacPherson in 1916 removed it from negligence claims.
• Traynor's Escola concurrence of 1944 became law in Greenman in 1963, and was generalised by section 402A in 1965.
• Manufacturing defects are judged against the maker's own design; design defects by consumer expectations or risk-utility with a reasonable alternative design; warning defects by adequacy and by who received the warning.
• Misuse, alteration, comparative fault, state of the art, and federal preemption remain available, and the vaccine legislation of 1986 shows the system's limits.

Sources

1. Legal Information Institute. (n.d.). Strict liability. Cornell Law School. law.cornell.edu
2. Legal Information Institute. (n.d.). Products liability. Cornell Law School. law.cornell.edu
3. Legal Information Institute. (n.d.). Abnormally dangerous activity. Cornell Law School. law.cornell.edu
4. Escola v. Coca Cola Bottling Co. of Fresno, 24 Cal.2d 453 (1944). CourtListener. courtlistener.com
5. Greenman v. Yuba Power Products, Inc., 59 Cal.2d 57 (1963). CourtListener. courtlistener.com
6. MacPherson v. Buick Motor Co., 217 N.Y. 382 (1916). CourtListener. courtlistener.com
7. Wikipedia contributors. (n.d.). Rylands v Fletcher. en.wikipedia.org
8. American Law Institute. (1998). Restatement (Third) of Torts: Products Liability. American Law Institute Publishers.

On 11 July 1951, in a back yard in Olympia, Washington, a five-year-old boy named Brian Dailey was visiting with his family. Ruth Garratt, an adult, began to sit down in a wood and canvas lawn chair. Brian moved it. She landed on the ground and fractured her hip.

Brian's account, which the trial judge believed, was that he had moved the chair in order to sit in it himself, noticed Mrs Garratt beginning to sit where it had been, and tried to move it back but was too small and too slow. On that finding the trial court dismissed the claim, holding there was no intent to injure.

The Washington Supreme Court sent the case back in Garratt v. Dailey, 46 Wash.2d 197 (1955), on a single point. The question was not whether Brian wanted to hurt her. It was whether he knew, with substantial certainty, that she would attempt to sit where the chair had been. If he knew that, he intended the contact, whatever he hoped would happen. On remand the trial court found that he had known, and entered judgment against him.

Remember: Intent in tort law means purpose or substantial certainty about the consequence, not a desire to cause harm and not a bad motive.

What intent does and does not require

Three clarifications that resolve most confusion.

• Intent attaches to the contact, not to the injury. A person who intends a light shove intends the contact, and is liable for the fractured skull that follows, because the eggshell skull rule applies here too.
• Motive is irrelevant. A doctor who operates without consent, sincerely believing it best for the patient, has committed a battery.
• Children and people with mental illness can be liable. There is no minimum age. The question is only whether this defendant in fact had the required state of mind, which is why Brian Dailey's age mattered to what he knew rather than to whether he could be sued.

Transferred intent completes the picture. If you swing at one person and hit another, the intent transfers to the actual victim. It also transfers between torts within a historic group of five: battery, assault, false imprisonment, trespass to land, and trespass to chattels. Aim to frighten and you connect, and the assault intent supports a battery.

The main intentional torts

Tort	Elements	Damage required?
Battery	Intentional harmful or offensive contact with another's person	No; offensive contact alone suffices
Assault	Intentionally causing reasonable apprehension of imminent harmful or offensive contact	No; the apprehension is the injury
False imprisonment	Intentional confinement within boundaries fixed by the defendant, of which the claimant is aware or by which harmed	No
Trespass to land	Intentional entry onto land in another's possession	No; nominal damages are available
Trespass to chattels	Intentional interference with another's personal property	Yes; some actual harm or dispossession
Conversion	Interference so serious that the defendant should pay the full value	Yes; the remedy is a forced sale
Intentional infliction of emotional distress	Extreme and outrageous conduct, intended or reckless, causing severe distress	Yes; severe distress must be proved

Several of these need no proof of loss, which is the sharpest structural difference from negligence. The reason is that they protect dignitary and possessory interests rather than physical wellbeing, and those interests are invaded the moment the boundary is crossed.

Vosburg v. Putney, 80 Wis. 523 (1891), shows the consequences. A fourteen-year-old lightly kicked an eleven-year-old classmate on the shin, across the aisle, after the class had been called to order. The kick was so slight the boy did not feel it at first. It aggravated an earlier injury, infection set in, and he lost the use of the leg. The kicker was liable for the whole of it. Two features of the reasoning matter: the contact was unlawful because it happened in a classroom after order was called, where no implied licence to horseplay existed, and the extent of the harm was irrelevant once the contact was intentional and unprivileged.

Two further notes. Assault requires apprehension of imminent contact, so words alone are generally not enough, and a threat about next week is not an assault. False imprisonment requires that no reasonable means of escape exist, and it is met by physical barriers, force, threats, or the assertion of legal authority; most states also give shopkeepers a limited privilege to detain a suspected shoplifter for a reasonable time in a reasonable manner on reasonable grounds.

Intentional infliction of emotional distress is the newest and the most restricted. The conduct must be beyond all possible bounds of decency, which excludes insults, indignities, and ordinary cruelty. Where the target is a public figure and the subject is a matter of public concern, the First Amendment adds a further barrier, established when the Supreme Court held that a magazine parody of a well-known minister could not support liability without proof of a false statement of fact made with actual malice.

Consent

Consent defeats these torts entirely, and most of the litigation is about its scope rather than its existence.

• Express or implied. Holding out your arm for an injection is consent; so is stepping onto a rugby pitch, as to the contacts inherent in the game.
• Scope matters. Consent to a tackle is not consent to a punch after the whistle. Consent to surgery on the left knee is not consent to surgery on the right, however sensible the surgeon's judgement in the moment. Emergencies are handled by a separate doctrine of implied consent where the patient cannot communicate.
• Vitiated consent. Consent obtained by fraud about the nature of the act, by duress, or from someone who lacks capacity, is no consent.

Self-defence, others, and property

Self-defence permits reasonable force to prevent imminent harmful or offensive contact. The force must be proportionate, and deadly force is permitted only against a threat of death or serious bodily harm. Jurisdictions differ sharply on whether a person must retreat before using deadly force where retreat is safely possible, with many states having enacted stand your ground legislation. A reasonable mistake about the danger generally does not defeat the privilege.

Defence of others follows the same rules. Where the intervener is mistaken about who is the aggressor, some states protect a reasonable mistake and others put the intervener in the shoes of the person defended, so that a mistake is fatal.

Defence of property is narrower than most people expect, and one case settles it. Edward and Bertha Briney owned an unoccupied farmhouse in Iowa that had been broken into repeatedly. They boarded the windows, and then rigged a twenty-gauge shotgun in an upstairs bedroom, wired to the door, angled at first at the intruder's stomach and then lowered on Mrs Briney's suggestion. On 16 July 1967 Marvin Katko broke in to take old bottles and jars he considered antiques. The gun fired and destroyed much of his right leg.

In Katko v. Briney, 183 N.W.2d 657 (Iowa 1971), the Iowa Supreme Court upheld an award of compensatory and punitive damages. The rule is that the law places a higher value on human safety than on mere rights of property, and no privilege exists to use force calculated to cause death or serious injury to protect unoccupied premises, still less by a device that cannot see who has entered or why. A landowner may use reasonable non-deadly force, and may not delegate to a machine what they could not lawfully do in person.

The point: The prohibition is not about the burglar's deserts. It is that a spring gun cannot tell a thief from a child, a firefighter, or a lost neighbour.

Necessity

Public necessity is a complete privilege: property may be damaged or destroyed to avert a public disaster, such as demolishing a building to stop a fire spreading through a town, and no compensation is owed in tort.

Private necessity is incomplete, and the distinction is elegant. In a 1910 Minnesota case a steamship was unloading at a private dock when a violent storm blew up. Rather than cast off into the lake, the crew kept the vessel moored and replaced the lines as they parted. The ship was repeatedly driven against the dock and damaged it. The court held the shipowner was privileged to stay, so the dock owner could not have cast the ship adrift or sued in trespass, but the shipowner had to pay for the actual damage caused. The privilege protects the right to act; it does not shift the cost of acting onto the person whose property was used.

Common misconceptions

• "Intent means wanting to cause harm." It means purpose or substantial certainty about the contact or confinement. Brian Dailey did not want Mrs Garratt to fall.
• "You cannot sue a young child." There is no minimum age for intentional tort liability. The age affects what the child in fact knew.
• "You may use whatever force is needed to protect your property." Katko holds otherwise: no deadly force to protect unoccupied property, and no delegating to a device what you could not do in person.
• "Necessity means you never pay." Only public necessity is a complete privilege. Private necessity permits the act and still requires compensation for actual damage.

Summing up

• Garratt v. Dailey defines tort intent as purpose or substantial certainty about the consequence, not a wish to injure.
• Transferred intent moves between victims and among battery, assault, false imprisonment, and the two trespass torts.
• Battery, assault, false imprisonment, and trespass to land require no proof of loss, because they protect dignitary and possessory interests.
• Vosburg v. Putney shows full liability for unforeseeable extent following a slight but unprivileged contact.
• Consent is defeated by exceeding its scope, by fraud as to the nature of the act, by duress, and by incapacity.
• Katko v. Briney forbids deadly force, including by mechanical device, to protect unoccupied property.
• Public necessity is a complete privilege; private necessity permits the act but requires payment for actual damage.

Sources

1. Legal Information Institute. (n.d.). Intentional tort. Cornell Law School. law.cornell.edu
2. Legal Information Institute. (n.d.). Battery. Cornell Law School. law.cornell.edu
3. Legal Information Institute. (n.d.). False imprisonment. Cornell Law School. law.cornell.edu
4. Legal Information Institute. (n.d.). Self-defense. Cornell Law School. law.cornell.edu
5. Legal Information Institute. (n.d.). Intentional infliction of emotional distress. Cornell Law School. law.cornell.edu
6. Garratt v. Dailey, 46 Wash.2d 197 (1955). CourtListener. courtlistener.com
7. Vosburg v. Putney, 80 Wis. 523 (1891). CourtListener. courtlistener.com
8. Katko v. Briney, 183 N.W.2d 657 (Iowa 1971). CourtListener. courtlistener.com
9. Hustler Magazine, Inc. v. Falwell, 485 U.S. 46 (1988), on First Amendment limits to emotional distress claims by public figures.

In January 1990 Dr Ira Gore bought a new BMW sports sedan from an authorised dealer in Birmingham, Alabama, for a little over forty thousand dollars. Nine months later he took it to a detailer to be made to look, in his words, snazzier. The detailer told him the car had been repainted.

It had. The car had been damaged by acid rain in transit, and BMW had a nationwide policy: if the cost of repairing damage to a new car came to less than three per cent of the suggested retail price, the car was sold as new with no disclosure. The repaint had cost about six hundred dollars.

An Alabama jury awarded four thousand dollars in compensatory damages for the reduced value of the car, and four million dollars in punitive damages, arrived at by multiplying an estimated diminution in value by the number of cars BMW had sold this way nationwide. The Alabama Supreme Court cut the punitive award to two million.

In BMW of North America, Inc. v. Gore, 517 U.S. 559 (1996), the United States Supreme Court held even that grossly excessive, and for the first time struck down a punitive award as a violation of due process. It set out three guideposts: the reprehensibility of the conduct, the ratio between the punitive award and the actual harm, and how the award compares with civil and criminal penalties for comparable misconduct. Seven years later, in State Farm Mutual Automobile Insurance Co. v. Campbell, 538 U.S. 408 (2003), the Court added that few awards exceeding a single-digit ratio to compensatory damages will satisfy due process.

What matters here: The size of a tort award is not simply a jury's opinion. It is constrained by doctrine, by appellate review, by statute in many states, and ultimately by the Constitution.

What a compensatory award is made of

Compensatory damages divide into two families, and the difference between them structures every argument about reform.

	Economic damages	Non-economic damages
Contents	Medical costs past and future, lost earnings, lost earning capacity, care and equipment, property loss	Pain and suffering, loss of enjoyment of life, disfigurement, emotional harm, loss of consortium
Proof	Bills, pay records, expert projections	Testimony, inference from the injury itself
Predictability	Reasonably consistent between similar cases	Highly variable, which is the reform argument

Future losses are reduced to present value, because a sum paid today can be invested. A claimant who will need thirty thousand dollars a year of care for forty years does not receive 1.2 million dollars; they receive the smaller sum which, invested at an assumed rate, will fund those payments. The discount rate is fought over by economists in the witness box, and small differences in it move awards by large amounts. Life expectancy and work-life expectancy tables supply the periods.

Non-economic damages have no unit of account, and the law has never solved this. Lawyers use the per diem argument, inviting the jury to put a figure on a day of pain and multiply by the remaining life expectancy; some states forbid it as inviting arbitrary arithmetic. The honest description is that a jury is asked to convert something incommensurable into money because the alternative, awarding nothing for it, is worse.

Wrongful death statutes give the family a claim for their own losses when a victim dies. Survival statutes preserve the claim the victim had before death. Both are creatures of statute, because the common law's rule was that a personal action died with the person, and they vary greatly in who may sue and for what.

The collateral source rule

A claimant whose medical bills were paid by their own health insurer may still recover those costs from the defendant. That is the collateral source rule: benefits from a source independent of the defendant are not deducted.

Its defenders argue that the claimant paid the premiums, that a wrongdoer should not profit from the victim's prudence, and that deterrence would be diluted if defendants paid less because their victims were insured. Critics answer that it produces a windfall and inflates awards.

In practice the windfall is often illusory, because health insurers and public programmes assert subrogation rights and reclaim what they paid out of the settlement. Many states have modified or abolished the rule for particular categories, especially medical malpractice.

Insurance, which is who actually pays

Almost nothing in this course happens the way the cases suggest, because in the overwhelming majority of claims the defendant does not pay. An insurer does.

Several consequences follow, and they change the picture.

• Policy limits usually cap the real recovery. A judgment of two million dollars against a driver with a hundred thousand dollar policy and no assets is worth a hundred thousand dollars. The rest is uncollectable, and the defendant is described as judgment-proof.
• The insurer controls the litigation. The duty to defend is broader than the duty to indemnify: an insurer must provide a defence if any allegation in the complaint could conceivably fall within the policy, even where it ultimately owes nothing.
• Settlement is the norm. The vast majority of tort claims settle, and settlement values are shaped by policy limits, litigation cost, and delay far more than by any calculation of the true loss.
• Deterrence is blunted. Insurance spreads the cost of a judgment across all policyholders. It retains some bite through experience rating and premium increases, and none at all where a driver's premium barely moves.

Add to this a fact from the empirical literature that surprises most people: the large majority of people injured by apparent negligence never make a claim at all, whether through ignorance, cost, or an unwillingness to sue. The system is not a machine that compensates everyone injured. It is a machine that compensates a small and unevenly selected subset.

The reform argument, on both sides

Since the 1970s a substantial political movement has argued that tort liability is excessive, unpredictable, and economically damaging. The measures it has secured include caps on non-economic damages, limits on punitive awards, shortened limitation periods and statutes of repose, requirements that a qualified expert certify a malpractice claim before it is filed, and abolition of joint and several liability.

California's medical injury legislation of 1975 is the best-studied example. It capped non-economic damages in medical malpractice at two hundred and fifty thousand dollars, a figure that was not adjusted for inflation for almost fifty years, so that its real value fell to a fraction of the original before legislation in 2022 began raising it on a schedule.

The case for reform: unpredictable awards make insurance hard to price, which raises premiums and pushes practitioners out of high-risk specialties and regions; the threat of suit produces defensive medicine, meaning tests ordered to protect the doctor rather than the patient; and a large share of every dollar spent on the system goes to lawyers and experts rather than to injured people.

The case against: caps bite hardest on the most severely injured, and disproportionately on claimants with low or no earnings, since a homemaker, a retired person, or a child has small economic losses and their entire claim may be non-economic; the empirical evidence that caps reduce total health spending is contested; and the anecdotes that drove the movement, of which the coffee case in the first lesson of this course is the most famous, were frequently misreported.

So what?: The reform debate is not a dispute about facts alone. It turns on who should bear the cost of accidents, which is a question about distribution, and no amount of data settles it by itself.

What is the system for

Three accounts compete, and a student should be able to state each in its strongest form rather than caricaturing two of them.

Corrective justice. The idea traces to Aristotle and has been developed in modern form by scholars including Ernest Weinrib and Jules Coleman. Tort law exists to correct a wrongful transfer between two particular parties: the defendant took something from the claimant that was not theirs to take, and the law's function is to annul that specific gain and loss. Its strongest evidence is the shape of the system itself. A tort case is bilateral. This claimant sues this defendant. The claimant recovers exactly what the defendant's wrong took, no more, and a defendant who injured nobody pays nothing however dangerous their conduct. A regulatory system aimed at reducing accidents would look nothing like this: it would fine dangerous conduct whether or not anyone was hurt, and it would pay victims from a fund without asking who caused the injury.

Deterrence and economic efficiency. Developed by Guido Calabresi and Richard Posner among others, this account says the purpose is to minimise the sum of accident costs and the costs of avoiding accidents. Liability rules should place the loss on whoever could have prevented it most cheaply, so that potential injurers internalise the cost of the risks they create and take precautions up to the point where further precautions cost more than they save. Its strongest evidence is that the doctrine frequently behaves as if this were true. The Hand formula is a cost-benefit test written into the definition of breach. Strict products liability was justified in Traynor's own words by which party could most effectively reduce hazards. Whole doctrines, including the abandonment of privity, are best explained as responses to who could prevent the harm.

Compensation and loss spreading. A third view treats tort law as a mechanism for moving losses from individuals who cannot bear them to institutions that can, through insurance and pricing. Its strongest evidence is what the system has been replaced with where legislatures have looked at it squarely. Workers' compensation removed workplace injuries from tort entirely in the early twentieth century, substituting a scheduled no-fault payment. New Zealand went further: legislation passed in 1972 abolished the right to sue for personal injury altogether and replaced it with a universal no-fault compensation scheme covering everyone injured by accident, however caused. It compensates far more people, far faster, at far lower administrative cost, and at lower individual payments.

Where the accounts break

Each has a problem it cannot easily answer, and holding all three problems in view is the most useful thing this lesson can leave you with.

Corrective justice struggles with insurance. If the point is that this wrongdoer must repair the loss he caused, why is it acceptable, even compulsory, that a stranger pays instead? A defendant who pays a premium and nothing more has not corrected anything personally. Corrective justice theorists answer that the duty is what matters and its discharge may be delegated, which is a real answer and not obviously a complete one.

The deterrence account struggles with the plaintiff. If the aim is optimal precaution, the money should go wherever it best serves that aim, and there is no reason it must go to the injured person rather than into a state fund. The economic account explains why the defendant pays and not why this claimant receives. It also assumes actors who calculate, which fits corporations better than distracted drivers.

The compensation account struggles with what it discards. New Zealand's scheme pays modestly and pays everyone; it also gives an injured person no forum in which anyone is ever held responsible, and no mechanism by which a company that injured hundreds is publicly called to account. Whether that matters is a real disagreement, and people who have been injured are frequently emphatic that it does.

Bottom line: The most defensible position is that tort law is not a single instrument. It is a very old dispute-resolution mechanism onto which each generation has bolted the purposes it needed, and the doctrines you have learned in this course carry the fingerprints of all three.

Two closing words about the boundaries of this course. First, everything here has been general legal education: it describes rules at the level of principle, those rules differ from state to state and country to country, and none of it is legal advice or a substitute for a lawyer licensed where you are. Second, the arguments in this last lesson are genuinely open. If you finish able to state the strongest version of a position you disagree with, you have got the thing this subject is actually for.

Common misconceptions

• "Juries hand out enormous awards and that is the end of it." Awards are reduced by judges, cut on appeal, capped by statute in many states, and constrained by the due process guideposts from Gore and Campbell.
• "Punitive damages are common." They are awarded in a small minority of cases and require conduct well beyond ordinary negligence.
• "Tort law compensates people who are injured." It compensates a small, unevenly selected subset of them, and most injured people never claim.
• "Deterrence and corrective justice are the same thing described differently." They diverge sharply: one cannot explain why the money goes to this claimant, and the other cannot comfortably explain insurance.

Pulling it together

• Compensatory damages split into economic losses, which are documented, and non-economic losses, which are not, and future losses are reduced to present value.
• BMW v. Gore established due process review of punitive awards on three guideposts, and State Farm v. Campbell pointed to single-digit ratios.
• The collateral source rule keeps the defendant from benefiting from the claimant's insurance, though subrogation often reclaims the money anyway.
• Liability insurance determines who really pays, caps real recovery at policy limits, and blunts deterrence while making compensation possible at all.
• Tort reform has secured caps and procedural barriers; its strongest objection is that caps fall hardest on the most severely injured and on claimants with little income.
• Corrective justice explains the bilateral structure, deterrence explains doctrines like the Hand formula and strict products liability, and compensation explains workers' compensation and New Zealand's no-fault scheme.
• This is general legal education and not legal advice.

Sources

1. Legal Information Institute. (n.d.). Punitive damages. Cornell Law School. law.cornell.edu
2. Legal Information Institute. (n.d.). Compensatory damages. Cornell Law School. law.cornell.edu
3. Legal Information Institute. (n.d.). Collateral source rule. Cornell Law School. law.cornell.edu
4. Legal Information Institute. (n.d.). Pain and suffering. Cornell Law School. law.cornell.edu
5. BMW of North America, Inc. v. Gore, 517 U.S. 559 (1996). Legal Information Institute, Cornell Law School. law.cornell.edu
6. State Farm Mutual Automobile Insurance Co. v. Campbell, 538 U.S. 408 (2003). Legal Information Institute, Cornell Law School. law.cornell.edu
7. Wikipedia contributors. (n.d.). Corrective justice. en.wikipedia.org
8. Calabresi, G. (1970). The Costs of Accidents: A Legal and Economic Analysis. Yale University Press.
9. Weinrib, E. J. (1995). The Idea of Private Law. Harvard University Press.