When a Court Sets the Chosen Law Aside

Parties draft a governing law clause to remove uncertainty, and it does remove a great deal of it. What it cannot do is displace the interest a state has in applying its own protective law to people within its borders, and the four grounds on which courts set these clauses aside are all versions of that single idea.

No connection to the chosen state

The threshold requirement. Courts generally require that the chosen state have a substantial relationship to the parties or the transaction, or that some other reasonable basis exists for the choice.

Easily met in ordinary commerce. A party's headquarters or incorporation, the place of performance, the location of the goods or the land, or the place of contracting each supply the relationship.

Not met by a bare preference. Selecting a state because its law is favorable, with no other connection to anything, is the case that fails, and the requirement exists to prevent exactly that.

A recognized body of commercial law can be a reasonable basis. Parties in unrelated states selecting a well-developed commercial jurisdiction for its predictability have frequently satisfied the test, and several states have legislated to permit it for large transactions.

This ground is rarely the operative one. Most clauses that fail do so on policy grounds rather than for want of a connection, because most drafters name a state with an obvious link.

Fundamental policy of a more interested state

Two elements, both required. The other state must have a materially greater interest in the issue, and applying the chosen law must contravene a fundamental policy of that state.

A different rule is not a fundamental policy. States differ constantly and choice-of-law clauses exist to select among differences. Something more is needed.

Non-waivable statutory rights are the clearest case. Where a legislature has made rights unwaivable, allowing them to be waived by a governing law clause would defeat the statute directly.

Protective legislation supplies most examples. Employment, consumer, insurance and franchise statutes exist to protect parties who did not negotiate the terms, and courts read them as embodying policies that survive a clause — which is why the wage and non-compete outcomes described in how far a non-compete travels turn out as they do.

The interest question usually answers itself. For an employee the state where they live and work has the greater interest; for a consumer, their own state. For two businesses in different states, the answer is genuinely contested and clauses survive far more often.

Express statutory prohibition

Some statutes void the clause directly. Rather than leaving the matter to policy analysis, several states have enacted provisions declaring that a choice-of-law clause selecting another state's law is voidable for defined contracts.

Employment covenants are the common target. Statutes voiding out-of-state choice of law and forum for employees who live and work in the state have spread quickly.

Consumer and construction contracts too. Several states void clauses in consumer agreements and in construction contracts for work performed in the state.

These are the strongest form of the objection. A court applying an express prohibition is not weighing interests; it is applying a rule, which makes the outcome predictable and the clause worthless for that category.

They are frequently voidable rather than void. The protected party may elect to rely on the clause where it happens to suit them, which is a drafting detail with real consequences.

The clause itself fails

Fraud or overreaching in obtaining it. A clause procured by misrepresentation about its effect can be set aside independently of the rest of the agreement.

Unconscionability. Where the clause was imposed in circumstances of gross inequality and operates oppressively — a consumer form contract selecting a distant state's law and forum together — courts have declined to enforce it.

The agreement is invalid. Where the contract containing the clause is unenforceable in its entirety, the clause generally falls with it, though arbitration provisions are treated differently under their own doctrines.

Scope. The commonest practical outcome is not that the clause is overridden but that it never covered the claim. A clause governing the agreement does not automatically govern a tort or statutory claim, as discussed in what a choice-of-law clause achieves.

Which makes drafting scope the highest-value edit. Parties spend their negotiating energy on which state is named and rarely on what the clause covers, and the second question decides more disputes than the first — a pattern that repeats in the contract analysis set out in the test courts use for contracts.

What happens after the clause falls

The court applies its ordinary conflicts analysis. Setting a clause aside does not mean the forum's law automatically applies. The court falls back to the test it would have used had no clause existed, which for a contract means weighing the significant contacts and for a tort means locating the relationship.

The result is frequently the same state anyway. Where the chosen state was chosen because it had the strongest connection, the fallback analysis often points at it too, and the argument about the clause turns out to have been about nothing.

Partial displacement is the usual outcome. Courts commonly apply the chosen law to the contract generally while applying local protective law to the specific issue the statute addresses. One agreement, two bodies of law, each doing the job it was designed for.

The remedy may change even where the rule does not. A statute that survives the clause frequently brings its own remedies — statutory damages, fee-shifting, penalties — which can dwarf the contractual measure of loss and are the practical reason the point is worth litigating.

And the forum question becomes live again. Where a statute voids the choice-of-law clause it frequently voids the forum clause alongside it, so a party that expected to litigate at home finds itself defending in the other side's state under the other side's law.

The practical conclusion for drafters is a modest one. Governing law clauses work well between commercial parties of comparable sophistication, in states with genuine connections to the transaction, on issues no legislature has removed from private ordering. They work poorly, and frequently not at all, where the counterparty is a person the law has decided to protect. Drafting on that basis — a clear clause for the commercial terms, and an assumption that the counterparty's own state governs anything protective — produces agreements that survive contact with a court, which is a better outcome than a clause that reads impressively and is set aside in the first paragraph of a judgment, after both sides have paid to argue about it.