Campbell's Law

Campbell's Law states that the more a metric is used for high-stakes decisions, the more likely it is to be manipulated, distorting the very process it's meant to monitor. In the school scenario, teachers start 'teaching to the test'—focusing narrowly on tested material—instead of fostering deep, genuine understanding. The test score, intended as a measure of learning, becomes the goal itself and ceases to be a reliable indicator. For example, a support center that rewards employees for the lowest 'average call time' may find them rushing customers off the phone, hurting satisfaction.

Ask this: "What important qualities does this metric fail to capture?"

Campbell's Law happens when we substitute a simple, measurable proxy for a complex, hard-to-measure goal. Because people respond to incentives, they naturally focus their efforts on improving the proxy metric, especially when rewards like money or status are involved. This focus often comes at the expense of the true, underlying objective.

The core of the problem is goal displacement. The proxy (e.g., test scores) displaces the real goal (e.g., education). This phenomenon is visible across many different domains.

In software development, if a manager measures productivity solely by 'lines of code written,' developers might write long, inefficient code to hit their target, even though it hurts the product's quality and maintainability.

In healthcare, a hospital system judged by low patient readmission rates might avoid admitting high-risk patients or transfer them elsewhere to keep the numbers down, rather than actually improving their long-term health outcomes.

This doesn't mean all metrics are bad. Campbell's Law is most powerful when the metric is a poor representation of the goal and the stakes are high. Using a balanced set of metrics that are harder to game can help mitigate the risk.

Coined by social psychologist Donald T. Campbell, this principle emerged from his work evaluating social programs in the 1970s. It serves as a critical warning against naive quantitative assessments in complex social systems and is closely related to Goodhart's Law ('When a measure becomes a target, it ceases to be a good measure'). The law highlights the inevitable tension between measurement for accountability and the potential for that measurement to corrupt the system it evaluates.

• Campbell (1976) first articulated the principle, stating, "The more any quantitative social indicator is used for social decision-making, the more subject it will be to corruption pressures and the more apt it will be to distort and corrupt the social processes it is intended to monitor." [1]
• Müller & de Rijcke (2017) demonstrated how research evaluation metrics, like citation counts, push scientists toward 'strategic publication' behaviors, prioritizing journal impact over potentially more innovative or risky research. [2]
• Van Thiel & Leeuw (2002) found in public administration that performance indicators often lead to behaviors like 'cream-skimming' (serving only the easiest cases to succeed on the metric) and 'tunnel vision' (ignoring all unmeasured aspects of a job). [3]

Campbell's Law is a powerful heuristic, not an ironclad rule. Its effects can be mitigated. Using a dashboard of multiple, counter-balancing metrics can create a more holistic and harder-to-game picture of performance. Furthermore, the law primarily applies in high-stakes contexts where metrics are tied to rewards or punishments. When metrics are used for low-stakes informational feedback and learning, they are far less likely to become corrupted. Some simple, direct metrics (like company revenue) are also less susceptible if they are tightly aligned with the organization's ultimate goal.