Where Emissions Travel: A Source–Receptor Approach to Evaluating Environmental Policy

Pollution from regulated sources can cross administrative boundaries; consequently, jurisdiction-based treatment assignment may fail to accurately capture policy exposure at a given receptor. To address this, we construct a predetermined treatment intensity by combining regulatory assignment with pre-policy source characteristics and atmospheric source–receptor linkages, and use it in a continuous treatment difference-in-differences design. A formal framework and simulations clarify when cross-boundary transport attenuates jurisdiction-based comparisons and when the source–receptor design recovers policy effects. Applied to China’s Two Control Zones policy, both designs indicate post-policy reductions in local SO2 emissions. For SO2 dry deposition at receptors, the source–receptor estimate is negative and statistically significant, whereas the jurisdiction-based estimate is also negative but not statistically distinguishable from zero. The same exposure measure reveals declines in crop value per hectare in more exposed villages, which are less apparent in jurisdiction-based comparisons. Crop-specific response patterns are consistent with a sulfur-nutrient channel.