Who Gets Heeded? An Obligation-Level Audit of Responsiveness in EPA Rulemaking

This paper introduces a novel framework for auditing regulatory responsiveness by shifting the unit of analysis from the entire rulemaking document to the discrete regulatory obligation. While previous research has struggled to measure influence at the aggregate level, this study uses an AI-assisted pipeline to extract specific duties—such as deadlines, reporting requirements, and compliance thresholds—from both proposed and final EPA rules. By matching public comments to these specific obligations, the authors measure whether engagement correlates with changes in the final text, validating their findings through blind human audits.

First, the authors find that engagement is associated with a modest increase in the probability of revision. Obligations addressed by five or more commenters are roughly 19.5 percentage points more likely to be revised than those with less engagement. However, the magnitude of this effect is relatively small, suggesting that public comments are not the primary driver of all regulatory changes.

Second, the study reports a significant null result: the direction of a comment (whether it supports or opposes a provision) does not statistically differentiate the likelihood of revision. This challenges simple preference-aggregation models that assume agencies respond primarily to the volume of pressure from one side or the other.

Third, the analysis reveals an upstream structural inequity. When looking at cross-docket patterns, organizational commenters are more likely to influence editorial refinements than substantive modifications. The authors argue that the true equity asymmetry lies in the differential capacity of various groups to identify, interpret, and contest specific legal obligations, rather than just the ability to submit comments.

This research reframes the debate over administrative equity. By demonstrating that formal access to the rulemaking process does not guarantee substantive influence, the authors highlight that the "playing field" is uneven long before a comment is even written. The ability to navigate complex regulatory text and target specific obligations is a form of capacity that is not distributed equally, suggesting that procedural openness alone is insufficient to ensure democratic accountability in the administrative state.