Existing studies on the environmental Kuznets curve (EKC) neglect the inverse effect of pollution transfer from environmental regulation interactions on pollution reduction from a risk analysis perspective. Based on the regional differentiated attitudes on the environmental regulation reached in risk communication by the risk awareness biases of multiple interest groups, this article clarifies the causality between risk communication and risk transfer based on multistakeholder engagement processes; furthermore, the article incorporates the simultaneous action of the technological innovation effect and pollution risk transfer effect to construct a spatial environmental hyperbolic model with a bidirectional correlation between pollution emissions and economic growth in different regions. To verify our model, we select the pollution from agricultural watersheds in China as a sample to examine the two inverse effects. The results demonstrate that (1) agricultural watershed pollution and economic growth show an inverted U‐shaped relation and a U‐shaped relation in the local region and adjacent regions, respectively; (2) the pollution reduction assessment of the classical EKC model can be largely attributed to pollution risk transfer behavior; and (3) the turning point of the U‐shaped curve appears earlier than that of the inverted U‐shaped curve in the spatial hyperbola model. The findings suggest that stakeholders should consider the risk awareness bias caused by the imbalance of regional economic development and the scenarios that provide a “haven” for pollution risk transfer. Moreover, our study expands the theoretical connotation of the classical EKC hypothesis and is more suitable for pollution reduction scenarios in developing countries.