Contracts & Obligations: What Makes a Promise Bind

You tapped "I agree" on a ride-hailing app this morning, or bought a train ticket, or let a landlord keep a deposit in exchange for a key. Nobody shook a hand and nobody signed in ink, yet each of those was a promise that a court somewhere would, in the right circumstances, force someone to keep. Most of the time nobody ever thinks about this. The contract works because everybody knows it could be enforced.

Every society has had to answer the same question: which promises does the community back with force, and what happens when one is broken? Too few enforceable promises and nobody can trade with a stranger. Too many and people are trapped by every careless word. The law of contracts, which many civil-law countries file under the wider heading of obligations, is the long answer. It is also the subject where the world's legal families sound most different while arriving, most of the time, at the same place.

This describes how legal systems work. It isn't legal advice; laws differ by country and change.

The common ground, and where the families part

Strip away the vocabulary and every legal system asks five questions about a promise. Did the parties really agree? Was the agreement the kind the law will enforce, or only a social arrangement? Was the agreement fairly reached, with free and informed consent? How much honesty do the parties owe each other on the way in and while performing? And if it is broken, what does the injured side get? The table shows how three big families answer. Remember that "United States" and "England and Wales" share the common-law tradition (the US inherited it) but have diverged in details, and that Latin America, Spain and much of Europe descend from the Roman-law codes, with France's 1804 Code civil and Germany's BGB (in force 1900) as the two great models.

Question	US and England	Latin America	France and Germany
Where the rules live	Judges' decisions, plus statutes such as the US Uniform Commercial Code (UCC) for sales of goods	National civil and commercial codes (Mexico, Argentina 2015, Brazil 2002, Chile 1855, Colombia)	The codes: France's Code civil (contract law rewritten 2016), Germany's BGB
Why a promise binds	A bargain: something given or promised in return (consideration)	Consent, an object, and in many codes a lawful causa (the reason for the promise)	Consent and lawful content; France dropped the separate cause in 2016
Good faith	Required in performance in the US; England resists a general duty	A general principle in the codes, explicit in Argentina's 2015 code	A central principle (France art. 1104; Germany BGB §242)
Default remedy	Money damages	Choice of enforcement or termination, with damages	Enforcement in kind is available; Germany treats it as the primary right

At world level, cross-border sales between businesses are harmonised by the UN Convention on Contracts for the International Sale of Goods (CISG, concluded in Vienna in 1980 and in force since 1988). Its parties include the US, China, Germany, France, Mexico, Argentina and Chile (not the UK), and it applies to a sale between businesses in two of them unless the contract opts out. It is a useful neutral reference below: its authors were deliberately blending the families.

Offer and acceptance: when does the deal exist?

Most contracts are formed by an offer (a proposal the other side can accept to create a deal) and an acceptance. When both people are in the same room there's no puzzle. The puzzle appears at a distance. If you post an acceptance and the offeror posts a withdrawal the same day, which wins? In the 1818 English case Adams v Lindsell a wool dealer's acceptance was held to bind him when posted, even though it arrived late: the "mailbox rule". The common law (US and England) keeps it to this day: the acceptance takes effect when sent, but the withdrawal of an offer takes effect only when it arrives.

The civil-law tradition and the CISG lean to receipt: a message counts when it reaches the other side. Germany goes a step further: under BGB §145, a person who makes an offer is bound by it unless they expressly excluded the binding effect, so a withdrawal doesn't free them. Under CISG art. 16, an offer can be revoked only if the revocation reaches the other side before they send their acceptance, and not at all if the offer says it stays open for a fixed time.

Play with the clock. Post takes two days each way, and the offer was posted on day 0, so the earliest an acceptance can go out is day 2.

With the starting settings (acceptance posted day 3, withdrawal posted day 2, arriving on day 4), all three rules say there is a contract: the acceptance was sent before the withdrawal arrived. Move the acceptance to day 5 and the paths split: the common law and the CISG say the withdrawal got there first, so there is no deal, while Germany's binding offer means the deal still exists. Now lose the letter: the common law still finds a contract (the English Household Fire v Grant, 1879, held that a lost acceptance still bound the offeror), while a receipt rule does not. Both approaches are defensible; each puts the risk of the post on someone different.

Consideration, causa and form: which promises count?

Not every agreement is a contract. If you promise your aunt a lift to the airport and forget, nobody expects a court. The law needs a filter, and here the families use different filters. The common law demands consideration: each side must give or promise something in return, a "bargained-for exchange". The classic statement is Currie v Misa (1875): a benefit to one side or a detriment to the other. Notably, the law doesn't ask whether the exchange is fair. A car for one pound is still a bargain, although other doctrines (fraud, duress, unconscionability) may be available.

The civil-law filter is different. Chile's 1855 Código Civil (art. 1445), like those that followed it, lists consent, capacity, a lawful object and a lawful causa as conditions of validity. Causa is the reason a person undertook the obligation. France had the same idea in the old art. 1108 until the 2016 reform replaced it with a requirement of lawful and certain content (new art. 1128); the old worries about a contract with no purpose are now handled by other rules, such as one that voids a derisory counter-performance. Germany never used causa for contracts: its filters are consent, then form and morality (§138 voids contracts contrary to good morals).

The bigger difference is the promise to give for nothing. Because there is no exchange, the common law won't enforce a plain gift promise. The civil-law world won't either, but for a different reason: gifts are a special act that needs form. Germany's BGB §518 requires a notary for a promise to give, which cures itself once the gift is performed. French law requires a notarial deed for most gifts. Which promises count? Try the scenarios.

Notice what comes out. For ordinary commercial bargains all families agree. They differ at the edges: gifts, grossly lopsided prices, and promises that someone relied on. The US has a special doctrine for that last case, promissory estoppel (Restatement (Second) of Contracts §90), under which a promise that the promisor should expect to induce reliance, and does, can be enforced to the extent justice requires. It was invented to patch the hole that the consideration rule leaves.

Good faith: how much must you care about the other side?

Does a negotiator who walks away at the last minute owe anything? This is the sharpest transatlantic contrast. The English House of Lords held in Walford v Miles (1992) that a promise to negotiate in good faith is unenforceable, because negotiators are allowed to pursue their own interests. The US imposes a duty of good faith in performing a contract (UCC §1-304, Restatement (Second) of Contracts §205), but generally leaves parties free to leave talks, with the exception that reliance can matter: in Hoffman v Red Owl Stores (Wisconsin, 1965), a would-be franchisee who sold his bakery on the franchisor's assurances recovered his losses.

Continental systems start from the opposite end. Germany's BGB §242 requires performance as Treu und Glauben (good faith) and custom demand; since the 2002 reform, §311(2) makes pre-contract negotiations a relationship with duties of care, following Rudolf von Jhering's 1861 doctrine of culpa in contrahendo (fault in forming a contract). France's art. 1104 makes good faith a public-order rule for negotiation, formation and performance. Predict first, then check.

Mistake and duress: when "yes" isn't really yes

A contract is the meeting of two wills, so what if one will was defective? Two kinds of defect run through every system. Mistake: one or both sides were wrong about something fundamental. The English chestnut is Raffles v Wichelhaus (1864): cotton was sold "to arrive ex Peerless from Bombay", but there were two ships called Peerless, sailing months apart, and each side meant a different one. The court would not enforce a deal in which the parties had never meant the same thing. Duress: one side was forced, by threats or by exploiting a desperate situation.

The families file these under different names. The French code speaks of vices du consentement (defects of consent): erreur (error), dol (deceit) and violence, and the 2016 reform added abuse of a state of dependence. Germany lets a person avoid ("anfechten") a declaration made under a mistake (§119) or induced by deceit or unlawful threats (§123). The common law uses mistake, misrepresentation, duress and undue influence, each with its own tests.

A subtle case shows how the legal families approach "pressure". A builder, halfway through a job, demands more money and the owner, who has no time to find another, agrees. In Alaska Packers' Association v Domenico (US, 1902) sailors who refused to work unless their wage was raised were held to have given no consideration for the raise, since they were already bound. The English Court of Appeal in Williams v Roffey Bros (1989) found that a practical benefit to the promisor can count, as long as there was no duress. The point for a beginner: the question is never just "did they say yes?" but "was the yes free, and was it given for something?"

Remedies: damages or the thing itself?

Finally, the loser of a broken promise needs something. The default in the common law is expectation damages: money to put you where you would have been had the promise been kept. The limit comes from Hadley v Baxendale (1854): a mill-owner whose crankshaft was delayed in transit could not recover profits the carrier hadn't been told about; damages cover losses that were foreseeable at the time of the contract. A court will order specific performance (do what you promised) only if money is inadequate, as for land or unique objects.

Civil law starts at the other end. A creditor has a right to the performance itself: the German BGB (§241) states that the creditor may demand it, and the French reform of 2016 lists enforcement in kind first, with limits for impossibility or manifest disproportion. The CISG compromise (art. 28) tells a court it need not order specific performance if its own law wouldn't. In practice, with ordinary goods, all systems end up with money.

The common law's money-first approach has a famous defender. Oliver Wendell Holmes wrote in 1897 that the duty to keep a contract at common law "means a prediction that you must pay damages if you do not keep it". Law-and-economics scholars turned this into efficient breach: if a third party values the goods more than your buyer does, the best outcome for everyone is to breach and pay the buyer's loss. Critics, such as the legal philosopher Seana Shiffrin, reply that a promise is a moral commitment and that treating it as an option to pay misdescribes what promising is. The simulation below uses the arithmetic; the argument over its morality is still open.

With the starting numbers (price 200, buyer values it at 300, replacements cost 300, third party offers 350), the buyer's expectation damages are 100, the seller's gain from selling to the third party is 150, so the seller pays 100 and keeps 50 more than performing would have given. Under a specific-performance regime the seller would have to negotiate a release with the buyer, who would want at least 100: a deal is still possible because the third party values it at 350 > 300. The two regimes end in the same place, with a different split of the surplus. That is the "Coase" insight: when bargaining is cheap, the remedy changes who gets the money more than who ends up with the machine.

What a graduate just sees

• Every deal has a moment when it starts to exist; ask which rule fixes that moment, and who carries the risk of the post.
• A contract is a filter on promises. The common law filters by bargain (consideration), the codes by consent, content and sometimes causa, and every system filters gifts by form.
• "Good faith" is not one thing: honest performance (US), fair dealing in negotiation (Germany, France), or almost nothing (English negotiators).
• A "yes" must be free and informed; pressure, deceit and shared mistakes undo it.
• Damages aim at the position you would have been in had the promise been kept, not at punishing the breaker.
• Money is the common remedy in every system; the question is whether the court begins with the thing or with the price.
• When a legal rule looks arbitrary, ask who the rule makes bear the risk.

Where it shows up next

• Torts & Civil Liability (35/05): when a duty comes from the law and not from a promise, the same questions (fault, causation, remedy) come back in a different guise.
• Property (35/06): buying a flat is a contract, but the ownership it transfers needs the land registry.
• European Union Law (35/11): consumer rules, such as the 14-day withdrawal right for distance purchases, override ordinary contract doctrine in the name of fairness.
• The real world: the terms you click through, the lease you sign, the freelance invoice you chase, and every international sale of goods under the CISG.

If you want to go deeper

• 📘 Textbook. Ewan McKendrick, Contract Law: Text, Cases, and Materials (Oxford University Press). A standard English course text.
• 📘 Textbook, comparative. Hein Kötz, European Contract Law (Oxford University Press, 2nd ed., 2017, translated by Gill Mertens). Compares common-law and continental answers question by question and covers the 2016 French reform.
• 📘 Gentler. Charles Fried, Contract as Promise: A Theory of Contractual Obligation (Oxford University Press, 2nd ed.). A short, readable argument that contract law is the law of keeping promises, the view Shiffrin and others defend and law-and-economics scholars contest.
• 🌐 Free course. Contract Law: From Trust to Promise to Contract (HarvardX, Charles Fried and Daniel Markovits), listed on Harvard Online and edX. Free to audit when a session is open; check the page for current availability.