Numerous policies have been formulated and implemented to strengthen waste separation management activities in many countries. Waste separation management policies (WSMPs) must be evaluated as the precondition for reducing deviations from policy implementation and improving waste separation performance. Based on text mining technology and the construction of a policy modeling consistency (PMC) index model, we conducted a quantitative evaluation of 22 WSMPs issued by central governmental departments and provinces in the Yangtze River Delta, China from 2013 to 2021 and analyzed their optimization paths. The results suggest that the PMC index of the selected WSMPs has an upward trend. The average PMC index of 22 WSMPs was 6.906, indicating good quality in the policy texts. The PMC index identified seven, nine, five, and one of the policies as being perfect, excellent, good, and acceptable, respectively. The characteristics of WSMPs were further illustrated through PMC surface charts. Based on this, optimization paths for WSMPs with lower PMC indexes are proposed, which indicate that existing WSMPs have great potential for optimization in terms of harsher constraint regulations, context-appropriate incentives, and cultivation of market participants. Finally, this study provides a beneficial reference for similar cities or countries to improve their performance in the management of waste separation and environmental protection.