Statutes of limitation

Why does the law stop prosecuting a serious wrong merely because enough years have gone by?

A fraud was committed. Nothing about it improves with age: the money is still gone, the victim still poorer, the perpetrator no more innocent than on the day. And yet at some fixed anniversary the state's power to prosecute simply switches off, and the only thing the wrongdoer did to earn that is remain undetected long enough.

The standard justification is that evidence goes stale. That answer is true and insufficient, and you can see why in one move: courts handle weak evidence every day without needing a deadline. A jury can be told that a witness is recalling a night twenty years ago and should be believed accordingly. So what does a hard cut-off achieve that "treat old evidence as old" cannot?

Start by asking whose evidence decays. It is tempting to imagine time eroding the case evenly, like weather on a wall. It does not. A prosecutor knows which day matters, holds the exhibits, can take statements while they are fresh, and — decisively — chooses when to bring the charge. A defendant knows none of this until the knock at the door. What would he need? A receipt from an ordinary Tuesday, a colleague who remembers he was at work, a diary nobody kept, because nothing about that day announced that it would one day need proving. Time degrades the evidence of innocuous normality, which is the defence's whole stock, faster than the evidence of a specific event, which is the prosecution's.

Now the sharper step. Could a judge simply examine, case by case, whether this particular defendant has been unfairly disadvantaged by the delay? Consider what that inquiry would require: showing what the dead witness would have said, what the destroyed ledger contained, what the vanished alibi established. The prejudice is by its nature unprovable, because the proof is exactly what time removed — so the question is unanswerable in precisely the cases where the answer is yes. The law therefore substitutes a crude, checkable proxy, years elapsed, for a fact it cannot establish. That is why this is a deadline rather than a factor to be weighed.

Test that reading against the doctrine's own oddities. If the period measured how old the evidence is, the discovery rule would make no sense: in most civil claims and many criminal ones the clock starts not when the act occurred but when the claimant knew, or reasonably should have known, of the claim. Likewise it is suspended while a defendant absconds, or while a victim is a child. Every one of those adjustments measures the claimant's opportunity to act, not the evidence's age. The period is a diligence rule wearing the costume of an epistemic one. A third job runs alongside, clearest in civil cases: repose. If claims never expired, records would be kept forever and no one could ever stop being a provisional defendant.

Finally, look at where the rule stops. Most systems set no period at all for murder, and England and Wales imposes none on indictable offences generally — which proves a functioning system need not have one. So the period is not a necessity of proof but a judgement about where the public interest in punishing this wrong stops outweighing the risk of convicting on evidence the defence can no longer meet. Move either side and the line moves: cheap DNA made some old cases newly provable, and the recognition that delayed reporting of sexual offences is ordinary rather than suspicious has led many jurisdictions to extend or abolish those periods — genuinely contested, since it buys real prosecutions at the price of the protection just described.

A limitation period does not declare the wrong forgiven or expired. It bars the remedy, not the conduct — which is why in many systems it is a defence the accused must raise rather than something the court notices on its own, and why a debt outside its period can still be honoured, just not enforced.

It is worth separating this from two neighbouring ideas. The presumption of innocence tunes the standard of proof inside a trial, on the ground that convicting the innocent is worse than acquitting the guilty; the limitation period governs whether the trial happens at all, on the different ground that the evidence has decayed one-sidedly. And though a deadline is a bright-line rule of the familiar kind, the usual defence of bright lines — that case-by-case judgement is expensive and unpredictable — is not the operative reason here. The operative reason is stronger: the case-by-case inquiry is not merely costly but impossible, because the missing evidence cannot testify to its own absence.

• Why civil limitation periods are usually short and criminal ones long, when the stakes run the other way.
• How laches — an equitable doctrine barring claims brought unreasonably late — reaches the same result without a fixed number.

Every term the collection defines is gathered in the glossary.