Why punishing failure removes reporting, not failure

A hospital introduces penalties for departments reporting medication errors. Reported errors fall by half within a quarter. Independent chart audits find the underlying rate unchanged. The programme succeeded against its metric and failed against its purpose.

What everyone sees

The initial figures look like a successful intervention, and are usually presented as one. The gap only appears when a second measurement channel exists that is not subject to the same incentive, and most organisations do not maintain one, so the apparent success stands unchallenged.

What is actually happening

Campbell’s law holds that the more a quantitative indicator is used for decision-making, the more it will be corrupted. Errors are events; reports are actions taken by people who can choose otherwise, so a penalty applies pressure to the action and not to the event. Edmondson’s research on psychological safety found that teams reporting more errors often had better outcomes, because reporting reflected willingness to disclose rather than incidence. Reason’s work on safety culture argues that reporting depends on the reporter believing disclosure is safe, which a penalty directly removes.

Why it stays hidden

The substitution hides because the numbers move in the intended direction. A falling line on a dashboard reads as improvement, and there is no visible artefact of suppression — an unwritten report leaves nothing behind. The organisation also loses the information it would need to detect the problem, so the intervention degrades the instrument it is being judged by.