Goodhart’s Law states: “When a measure becomes a target, it ceases to be a good measure.”
Originally formulated by economist Charles Goodhart in 1975 in the context of British monetary policy, the law highlights how a performance indicator loses its predictive or control value once it becomes the very goal that people optimize for. In other words, the moment a metric is tied to rewards, penalties, or success criteria, human behavior shifts to game that metric—undermining the original intent.
Origin and Purpose
Goodhart observed that statistical indicators such as money supply aggregates stopped correlating with economic outcomes once used for policy targets. Social scientist Marilyn Strathern and later organizational researchers generalized this principle to many other fields, including education, healthcare, and agile product development.
Core Elements
- Central idea: Every metric is initially a proxy for the desired outcome. Once turned into a goal, the proxy itself becomes the focus of optimization.
- Mechanisms: Research distinguishes several failure patterns—regressional (optimization of measurement noise), causal (interventions break causal links), adversarial (deliberate gaming), and extremal (optimization outside the data range).
- Relation to other concepts: Goodhart’s Law is related to, but distinct from, Campbell’s Law (corruption of social indicators) and the Lucas Critique (behavioral changes under policy pressure).
Practical Relevance
Goodhart’s Law is critical wherever organizations reduce complex goals to a few metrics. Typical patterns include:
- Agile development: Over-emphasizing “velocity” encourages teams to split stories artificially and compromise quality.
- Customer service: Focusing only on “average handling time” can lower call duration while damaging customer satisfaction.
- Healthcare: Strict emergency-room wait targets may lead to administrative “clock-stopping” rather than improved patient flow.
- Education: “Teaching to the test” optimizes exam scores at the expense of deeper learning.
The law applies to companies of all sizes—from start-ups working with simple OKRs to global enterprises running complex KPI systems.
Implementation in Practice
Organizations can design metrics that drive genuine value without falling into the Goodhart trap by:
- Multi-dimensional measurement: Use a balanced set of indicators, combining quantitative and qualitative signals, so no single number dominates behavior.
- Iterative review: Treat every KPI as a hypothesis, subject to regular inspection and adaptation through retrospectives and audits.
- Careful incentive design: Align bonuses and rewards to long-term customer value, quality, and learning, not just easy-to-count outputs.
- Transparent governance: Clearly communicate the purpose and limitations of each metric. Make side effects explicit to create shared accountability.
- Technical safeguards: In AI and algorithmic decision-making, deliberately test for gaming or proxy optimization (e.g., red-team stress tests).
These practices shift the role of metrics from rigid control instruments to tools for shared learning and continuous improvement.
Real-World Examples
- Healthcare: Hospitals under strict four-hour ER wait limits met the target on paper by moving patients to non-counted holding areas, without improving care.
- Financial services: Call centers measured only on call duration saw customer repeat calls rise and Net Promoter Score fall until qualitative metrics were added.
- Software development: Teams rewarded for the number of tickets closed optimized for ticket count instead of customer value, until product-centric metrics were introduced.
Limitations, Weaknesses, and Criticism
- Not an argument against measurement: The law warns against misuse, not against metrics themselves.
- Overgeneralization: Well-designed composite indicators and holistic measurement systems can remain robust even when targets are set.
- Different failure modes: Regressional, causal, adversarial, and extremal forms require different countermeasures.
- Single-metric risk: Reducing complex objectives to one figure encourages “surrogation,” where people confuse the proxy with the goal.
vCultural effects: Over-reliance on numbers can crowd out intrinsic motivation and distort organizational culture.
- Governance cost: Continuous metric reviews and multiple layers of measurement require sustained effort and discipline.
- Safety-critical exceptions: In domains like aviation or nuclear safety, strict numeric targets remain necessary but must be paired with qualitative oversight.
CALADE Perspective
CALADE sees Goodhart’s Law as an invitation to smarter metric design, not a call to abandon measurement. In transformation programs and agile portfolios we:
- Use multi-dimensional control systems instead of single KPIs to capture the richness of value creation.
- Treat metrics as hypotheses that must be validated or replaced as conditions change.
- Coach leadership teams to keep strategy and culture aligned with business outcomes rather than letting a few numbers dictate behavior.
Related Terms
- Campbell’s Law
- Lucas Critique
- Balanced Scorecard
- Metric Anti-Patterns
- Surrogation
← back to list