Goodhart's Law
Charles Goodhart, 1975, in a paper on monetary policy: “Any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes.”
Marilyn Strathern, 1997, restating it more broadly: “When a measure becomes a target, it ceases to be a good measure.”
The Soviet nail factory story: Central planners in Moscow set quotas by weight. Factories produced fewer, heavier nails — useless for most purposes. Planners switched to quotas by quantity. Factories produced many tiny nails — equally useless. The metric optimized itself into meaninglessness.
The story is likely apocryphal, but the pattern is documented. Soviet sheet steel was heavy and thick. Sheet glass, when planned in tons, was too heavy. Paper, when quotas were set by weight, was too thick.
The mechanism: Measurement selects a proxy for the thing you care about. When you target the proxy, you create incentives to optimize the proxy, not the thing.
Before targeting: the proxy correlates with what matters. After targeting: the proxy gets optimized at the expense of what matters.
Modern examples stack up:
Teaching to the test — Standardized test scores were proxies for learning. Schools started optimizing for scores directly: drilling practice tests, narrowing curricula to tested subjects. Scores rose. Learning didn’t.
Hospital mortality rates — Hospitals that published low mortality rates were considered safer. Some hospitals began refusing high-risk patients. The metric looked better. Care got worse.
Police quotas — Arrest numbers were proxies for crime prevention. Officers started arresting for minor offenses to hit numbers. Arrest counts rose. Public safety didn’t.
Code coverage — Lines covered by tests were proxies for code quality. Developers wrote tests that covered lines without testing behavior. Coverage numbers rose. Bugs didn’t fall.
Measurement remains useful. Targeted measurement corrupts.
Use metrics as signals rather than goals. Watch multiple metrics that are hard to game simultaneously. Change metrics before they calcify into targets.
What you measure, you optimize. What you optimize, you hollow out.
Go Deeper
Books
- The Tyranny of Metrics by Jerry Z. Muller — Comprehensive treatment of metric fixation across institutions.
- Measuring and Managing Performance in Organizations by Robert Austin — How measurement shapes behavior in complex systems.
Essays
- Goodhart’s original 1975 paper is technical (about monetary policy). Strathern’s 1997 formulation is more accessible.
- Campbell’s Law (1979) covers similar ground from social science: “The more any quantitative social indicator is used for social decision-making, the more subject it will be to corruption pressures.”
Related: [[cobra-effect]], [[legibility]], [[systems]]