Goodhart's Law
Why measurement fails when indicators become targets
In short
When a measure becomes a target, it ceases to be a good measure. Formulated by British economist Charles Goodhart in 1975, the law states that when an indicator is turned into a direct target to control a system, humans or algorithms optimize their behavior to maximize that metric rather than the underlying purpose. Consequently, the metric loses its value as an objective reflection of reality. To prevent distortion, measurements should be used primarily for evaluation and diagnosis rather than direct reward or punishment.
Management by target numbers often produces unintended consequences in everything from healthcare and education to software engineering and artificial intelligence. When the metric itself becomes rewarded, focus shifts from actual quality to manipulating or optimizing the score.
The Mechanism Behind Goodhart's Law
Goodhart's law arose in the context of British monetary policy in the 1970s. When the central bank attempted to control the economy by targeting specific money supply indicators, financial market actors adapted their behavior, causing the statistical correlation between the target and inflation to break down. Anthropologist Marilyn Strathern later popularized the phrase: 'When a measure becomes a target, it ceases to be a good measure.' The issue occurs because complex systems consist of adaptive actors responding to incentives and finding shortcuts around rules.
From Emergency Healthcare to AI
Separating Diagnosis from Control
Sources
Read also
Reflection exercises
Use these exercises to apply the chapter's ideas. You don't need to write anything down — just pause and reflect on each question.
Identify Distorted Metrics
Examine a KPI or incentive structure in your organization or daily life.
- Which metric is being used to evaluate success?
- How could an actor optimize this metric without improving the underlying value?
Separate Diagnosis from Reward
Redesign a measurement process to prevent gaming the system.
- How can the same phenomenon be observed without linking it directly to bonuses or penalties?
- What qualitative measures can complement the quantitative data?
Summary
Goodhart's law shows that incentives change how metrics function. By separating diagnosis from rewards, using balanced indicators, and including qualitative assessments, organizations prevent systems from gaming their own goals.
Read the short version in the archive.