XGBoost-SHAP-IPA: An explainable algorithm for influencing factors on resident willingness to pay for recycled water

Amid intensifying global water scarcity, enhancing public utilization and willingness to pay (WTP) for recycled water has become imperative. Traditional statistical methods exhibit fundamental limitations in capturing behavioral complexity, modeling nonlinear relationships, and explaining key drivers, consequently failing to support targeted policy interventions. To address these gaps, this study develops an integrated XGBoost-SHAP-IPA framework — synthesizing Extreme Gradient Boosting (XGBoost), SHapley Additive exPlanations (SHAP), and Importance-Performance Analysis (IPA) — to systematically identify and interpret determinants of recycled water WTP in China. Utilizing nationally representative survey data, we executed sequential analytical phases: XGBoost modeling for WTP nonlinearities, SHAP value decomposition for factor contribution quantification, and IPA for policy priority establishment. Results demonstrate that public WTP exhibits strong dependence on six critical dimensions. Enhancing societal recognition and WTP necessitates a comprehensive strategy integrating three pillars: strengthening public trust through transparent project governance; optimizing dynamic pricing models and targeted subsidy mechanisms; implementing evidence-based civic education programs.