Goodhart's Law

Goodhart’s Law is a fundamental principle in the literature of social sciences and economics stating that when a measure becomes a target, it ceases to be a good measure. Developed by Charles Goodhart in 1975, this law emerged particularly in the context of economic policy and monetary policy. The core proposition of the law is: “When a measure becomes a target, it ceases to be a good measure.” By explaining the behavioral and systemic changes that arise when measures are turned into targets, the law has found applications in both economics and the social sciences.

The importance of Goodhart’s Law lies in its ability to objectively analyze the consequences of the interaction between measures and targets. Measures carry informational value only when they remain independent; however, when they become targets, the behavior of individuals and institutions shifts toward manipulating or optimizing those measures. This causes the measure to no longer accurately reflect the underlying reality. This phenomenon, observed in many fields such as economics, education, corporate governance, and digital platforms, has elevated Goodhart’s Law to one of the foundational concepts of modern social science.

Goodhart’s Law is grounded in monetary policy experiments conducted at the Bank of England. In the 1970s, monetary aggregates were adopted as policy targets in the United Kingdom, and it was observed that economic actors could manipulate these indicators. Charles Goodhart argued that certain monetary aggregates had lost their reliability and were undermining the predictability of economic policy.

Goodhart’s 1975 work focused on the British economic structure but aimed to frame the law as a universal economic and social phenomenon. According to Goodhart, measures do not merely reflect current conditions; they also shape the behavior of individuals and institutions. Therefore, setting measures as targets weakens their capacity to accurately represent true performance.

The Bank of England’s experiments revealed the impact of targeting monetary aggregates on economic stability. Banks and credit institutions developed strategies to improve the targeted indicators, thereby reducing the reliability of those measures. For instance, artificially inflating credit volumes to meet monetary supply targets caused economic indicators to diverge from the actual economic reality. This constitutes a concrete application of Goodhart’s Law in economic contexts, moving it beyond an academic theory into the realm of practical policy-making.

Goodhart’s Law demonstrates its influence across many fields beyond economics. Turning measures into targets alters the behavior of individuals and institutions, thereby diminishing the reliability of measurement.

Using exam scores as a measure of educational success directs teachers and students toward exam-focused study. This reduces the quality of learning and prevents the measure from accurately reflecting students’ true knowledge and skills. As teachers restructure curricula around exam requirements, students shift toward short-term memorization. This approach is a classic example of the negative consequences that arise when measures become targets.

When Key Performance Indicators (KPIs) are used as performance measures in corporate governance, employees may focus on short-term achievements. This can negatively affect long-term quality, innovation, and organizational culture. For example, a sales employee prioritizing sales targets may neglect customer satisfaction or resort to unethical methods to achieve results. This illustrates another application of Goodhart’s Law in a corporate context.

Today, digital platforms and algorithms are widely used in data-driven decision-making processes. Goodhart’s Law gains particular relevance here: when social media platforms direct content distribution through algorithms to optimize engagement metrics, user behavior changes and information pollution increases. E-commerce sites that prioritize click-through and conversion rates may harm user experience and long-term customer loyalty.

Goodhart’s Law is based on the interaction between measures and targets. Measures provide reliable information only when they remain independent; when they become targets, they alter the behavior of individuals and institutions. This reduces their informational value and creates a mismatch between expected outcomes and actual performance.

Turning measures into targets leads to both intentional and unintentional manipulation. Intentional manipulation occurs when institutions alter reports or prioritize short-term results. Unintentional manipulation arises from individuals exhibiting target-driven behavior. This interaction can rapidly erode the informational value of measures and lead to systemic distortions.

Goodhart’s Law parallels the Lucas Critique. Economic models produce invalid predictions if they fail to account for how actor behavior changes in response to policy shifts. Therefore, understanding this law is critical for policy design and economic modeling.

Goodhart’s Law is a powerful principle but has limitations:

Risk of Overgeneralization: Not every measure loses its reliability when it becomes a target. Some indicators may retain their reliability even when targeted.

Target-Measure Interaction: In some cases, the interaction between targets and measures can be predictable and manageable.

Goodhart’s Law illustrates how turning measures into targets alters system behavior in the context of digitalization and data-driven decision-making. Targeting engagement rates in social media algorithms leads platforms to prioritize attention-grabbing content over content quality. Over-optimizing specific performance metrics in artificial intelligence models reduces generalizability and increases the risk of bias. These dynamics reveal the impact of targeting measures on measurement outcomes.

Goodhart’s Law is a critical principle for understanding the behavioral and systemic effects of turning measures into targets. Targeting measures in economics, education, corporate governance, and digital platforms can lead to unexpected and often negative outcomes. Evaluating the reliability of measures and their impact on behavior is a structural component of decision-making processes. The law remains a fundamental theoretical concept in both academic literature and practical applications, providing a foundational conceptual framework for the design of measurement systems.