Water Works: How Goodhart’s Law Is Reshaping Stormwater Management

Data-driven performance verification is the backbone of the stormwater treatment industry. Engineers, regulators and municipalities rely on quantitative metrics such as pollutant removal rates and runoff reduction volume to evaluate the success of stormwater control measures (SCMs) and estimate regulatory compliance. We’re entering a promising era in which the standardization of SCM performance evaluation can significantly improve our understanding of how well these practices meet regulatory goals.

But as these metrics increasingly influence funding, research, compliance and design decisions, a troubling dynamic is emerging: we may be optimizing to the measure, not the mission. This phenomenon is best described by Goodhart’s Law: a principle that states, “when a measure becomes a target, it ceases to be a good measure.” As the stormwater industry becomes more sophisticated in tracking the performance of SCMs, we must remain mindful of how metric-driven management can distort outcomes and even paradoxically undermine the very water-quality improvements we aim to achieve.

Goodhart’s Law and Stormwater

Goodhart’s law is not a new concept. Originally introduced by economist Charles Goodhart in his 1975 paper on UK monetary policy (iimag.link/eZGks), the law has since been widely applied across data-driven fields, including systems management, software development, healthcare and finance. The core idea is that when a metric becomes a decision-making target, it can encourage short-term box-checking behaviors—or worse, data manipulation—that erode its original usefulness.

Stormwater regulations often emphasize easily quantifiable metrics intended to represent environmental benefits: runoff reduction volumes, impervious area treated, pollutant removal efficiency and number of green stormwater infrastructure SCMs installed. But when these metrics become the primary or sole target for compliance, a disconnect can arise between meeting numeric goals and delivering meaningful in-field water-quality benefits.

The Limits of Prescriptive SCM Selection

Green stormwater infrastructure (GSI) has been a hallmark of sustainable stormwater management for more than a decade. While GSI offers numerous benefits, its effectiveness depends heavily on factors such as design, application, site suitability and long-term maintenance. An expanding body of research supports the value of GSI, which has led to its widespread adoption in technical standards and permit requirements.

However, some jurisdictions have gone further by prescriptively requiring GSI in new developments—often under the assumption that water-quality and quantity goals will inherently be met. But as the 2025 Infrastructure Report Card by American Society of Civil Engineers notes, “although green infrastructure is critical in addressing localized drainage and water quality, these systems cannot provide sufficient capacity to manage large-scale stormwater events.” This underscores a broader issue: when regulations focus on the type of practice instead of the performance outcome, we risk prioritizing the wrong metric.

Misaligned Incentives and Performance Shortcuts

When stormwater funding is tied to specific metrics such as dollars per pound of pollutant removed or dollars per gallon of runoff reduced, it can incentivize projects in areas where those targets are easiest to hit, rather than where interventions would have the most environmental impact. Given the inherent variability of stormwater systems due to land use, geography and seasonality, accurately measuring these metrics can be difficult.

As a result, many programs rely on presumed performance rather than actual measured data. This may streamline reporting but can undermine long-term effectiveness if important site-specific factors are overlooked. The challenge is especially acute in jurisdictions with stormwater credit-trading systems, where it’s vital to balance the purchase of credits with tangible improvements in the watershed.

The Pitfall of Proxy Pollutants

Total Suspended Solids (TSS) is one of the most common “proxy pollutants” used in stormwater regulations. The common assumption is that removing TSS will also capture other pollutants of concern. While TSS removal contributes significantly to improved water quality, it’s also one of the easiest parameters to address.

By centering regulations and performance evaluations around TSS, we risk neglecting more complex or emerging pollutants (e.g., phosphorus, nitrogen or 6PPD-quinone) for which removal may require different treatment approaches. Focusing on just one pollutant often will not solve the whole problem.

Designing to the Test Protocol

The movement toward the standardized evaluation of SCM performance using well-defined protocols and standards—like those recently developed by ASTM that are rooted in the NJDEP laboratory protocols—represents a significant advancement in generating repeatable and comparable data across practices. These standards help ensure rigorous and consistent testing, particularly for pollutants such as TSS.

However, development of these protocols also highlighted a constraint: laboratory testing often focuses almost exclusively on TSS, in part because it’s easier to simulate in a controlled environment using surrogate materials such as silica. As a result, some SCMs now are being designed primarily to perform well in lab-based TSS tests, even if such optimization doesn’t translate to real-world effectiveness or lifecycle resilience. When TSS removal becomes the only benchmark, other considerations such as capture of other pollutants and long-term maintainability fall by the wayside.

Mitigating the Effects of Goodhart’s Law

Goodhart’s Law manifests in many areas of stormwater management, but one broadly achievable solution to its mitigation stands out: diversification of performance metrics. By balancing quantitative indicators (e.g., numeric targets such as measured pollutant load reduction) and qualitative indicators (e.g., narrative outcome-based targets such as stream health indicators), we can better align how we measure success with the actual environmental benefits we seek.

Adaptive management—where monitoring and data feedback guide future decisions in a feedback loop—can also help us adjust strategies in real time and address unintended consequences before they become systemic issues. Above all, the industry must remain critically aware of the limitations of narrowly focused data-driven metrics. Only then can we ensure our methods of performance evaluation truly support the broader mission: clean water, resilient infrastructure and thriving ecosystems.