<p>Agricultural green transformation requires policies that address pollution control and carbon mitigation within the same production process. This study investigates whether China’s Agricultural Social Services (ASS) and Policy-based Agricultural Insurance (PAI) create additional environmental gains when they operate as a coordinated policy package. Based on a balanced panel of 31 Chinese provinces from 2007 to 2023, we construct a coordinated environmental performance index that integrates carbon-reduction performance and non-point source pollution-control performance. A double/debiased machine learning framework is then used to estimate the effect of ASS-PAI coordination while allowing for high-dimensional covariates and nonlinear confounding structures. The evidence shows a steady improvement in coordinated environmental performance over time, accompanied by a more compact spatial distribution and a southeastward movement of the performance center. Provinces implementing both ASS and PAI achieve higher environmental performance than those exposed to only one policy. This result remains stable after changing model specifications, cross-fitting folds, and machine-learning algorithms. Mechanism analysis indicates that coordinated policies work partly by expanding mechanized agricultural operations and strengthening farmer cooperatives. The effect is stronger where agricultural service systems are more developed, insurance markets are less mature, grain-crop dominance is lower, and irrigation conditions are better. These findings show that production-service support and risk-protection instruments can become mutually reinforcing when they are institutionally matched. The study provides evidence for designing agricultural policy packages that help developing economies control diffuse pollution and lower production-related carbon emissions within the same production system.</p>