The AI Journal

In April 1902, two parties to a contract in London were waiting on the same event. Paul Krell had rented a room on Pall Mall, at a price that reflected one thing only: the view it gave onto the coronation procession of Edward VII. The Herne Bay Steamboat Company had chartered its vessel to a hirer who wanted two things at once, a cruise to view the King’s naval review at Spithead and a cruise round the assembled fleet regardless. When the King fell ill and the ceremonies were postponed, both contracts had, on the face of it, been struck by the same accident: an external event had removed the reason either party wanted what they had bargained for. The Court of Appeal decided the two cases within months of each other, and let one bargain fail while holding the other to its terms. Krell did not have to pay for the room. The steamboat hirer did have to pay for the boat. Nothing about either party’s ability to perform had changed in either case; the room was still there, the boat still sailed. What had changed, and changed unevenly, was the point of paying.

That is frustration of purpose, and English and American law both set it beside an older, differently shaped doctrine: impossibility, and its more forgiving modern relation, impracticability. Forty years before the coronation cases, a music hall hired out for a run of concerts burned to the ground before the first performance, and the court in Taylor v Caldwell excused the owner from liability, because the thing actually promised, a building standing and usable on the appointed nights, had ceased to exist. Nobody argued that the concerts had lost their point. They argued, successfully, that a particular performance had become impossible because its physical subject was gone. The music hall could not be rebuilt in time; the coronation room could still be occupied; only one of the two excuses asks whether the promised act can still happen at all.

That is the distinction that gets lost. Casebooks and practitioner guides tend to present frustration and impossibility as a single family of discharge, differing mainly in how strict a court will be, with impracticability as a slightly looser American statutory cousin under section 2-615 of the Uniform Commercial Code. The American Restatement (Second) of Contracts keeps them in adjoining sections, 261 for impracticability and 265 for frustration of purpose, as though they were two grades of the same excuse rather than two different questions put to two different facts. Impossibility and impracticability ask whether the promisor can still render the performance that was promised, at anything short of a change in degree the parties did not bargain to bear. Frustration asks something else: whether the exchange, fully performable, still delivers what both sides understood it to be for. A seller who can no longer obtain the goods is excused, if at all, under the first question. A buyer who can obtain goods he no longer has any use for is asking the second question, and the goods being perfectly available is not an answer to it.

The two questions get run together because they arrived on the law’s desk together and have shared a remedy ever since. Herne Bay and Krell were decided by the same court in the same season, reasoning from the same premise, that a contract rests on assumptions neither side stated, and both excuses discharge the loser from the same obligation to pay or perform further, generally with the same restitutionary mopping-up for whatever was transferred in the meantime. Richard Posner and Andrew Rosenfield’s 1977 analysis treats the family as one problem in risk allocation: when an unbargained-for contingency strikes, the law’s job is to decide, after the fact, which party would have borne the risk had they thought to price it into the deal. Read that way, impossibility and frustration look like variations on a theme, not separate doctrines, and a court under time pressure can be forgiven for pleading them, and deciding them, as interchangeable.

They are not interchangeable, because the thing being risk-allocated differs. Impossibility and impracticability allocate the risk that the performance itself turns out to be harder, costlier or physically unavailable than assumed; the Uniform Commercial Code’s official commentary to section 2-615 is explicit that an increase in cost is not enough unless some unforeseen contingency has altered the essential nature of the performance, because ordinary market risk is a risk the parties are taken to have already priced. Frustration allocates a different risk entirely: that the shared assumption behind why the exchange mattered will hold. A tenant whose restaurant lease remains fully performable, premises available, rent payable, keys in hand, is not making an impossibility argument merely by having lost his customers; he is making, if he is making anything, a frustration argument about the assumption that people would be allowed to gather there, and that argument succeeds or fails on whether that assumption was truly shared and truly basic to the deal, not on whether performing the lease has become harder. Treating the two as a single sliding scale invites a court to ask the impossibility question, “is performance still feasible”, of a case that only ever raised the frustration question, “does performance still mean anything”, and a merely feasible but pointless performance will survive a test built for a different complaint.

The strongest objection to insisting on this is that the doctrines were never cleanly apart to begin with, and still are not. Herne Bay’s own judgment discusses the hirer’s purpose and the ship’s availability in the same breath, because the two facts were genuinely entangled in that case; a long line of English commentary treats frustration of purpose as a subcategory of frustration generally, which itself contains the English analogue of impossibility, so that the species-genus relationship is not a sloppy collapse but a considered taxonomy, and remedies for both have in practice converged toward the same restitutionary rules regardless of which label a court reaches for. If the outcomes are the same and the facts are frequently mixed, demanding a sharp line can look like insisting on a distinction that does no further work.

The outcomes converge only once a court has already decided which test to apply, and that prior decision is exactly where the conflation does its damage. A court asked whether a nightclub’s rent obligation survived a closure order is not choosing between two paths to the same destination; it is choosing whether to ask if the club could still open its doors, which it could, or whether the point of opening them had gone, which is a different and separate inquiry with a different evidentiary target, a shared basic assumption rather than a physical or economic obstacle. Collapsing that choice into a single sliding-scale “frustration family” does not merely blur vocabulary. It lets a party resisting discharge argue the easier of the two questions, performance remained possible, against an opponent who was never required to prove impossibility at all, and it lets a court that has found performance feasible stop there, never reaching the question the facts actually raised. The coronation cases survive as a pair precisely because they show a court holding two different tests apart under pressure to blend them; losing that discipline a century later is not tidying up old law. It is deciding cases by the wrong standard.