Research on public policy formulation and evaluation mechanisms under the PDCA cycle model

This paper proposes a PDCA cycle model for public policy formulation, and proposes a public policy implementation effect prediction model and a public policy implementation effect evaluation model for the situation before and after the implementation of public policy, respectively. In the public policy implementation effect prediction model, simulation is carried out based on Data Farming of Agent simulation, and a genetic algorithm is used as the basis to determine the input parameters of the loop operation and complete the adaptive search process of the optimal solution. The DID model sets dummy variables to calculate the value of DID, and the common trend assumption is applied to link it to the policy effect. The endogenous choice model is used to evaluate public policies, and the regression synthesis method is utilized to reduce subjective interference in the selection of control group samples. The STI public policy of Hebei Province, China, is evaluated from 2019 to 2023, and its STI policy effect improves from -0.3661 in 2019 to 0.4778 in 2023, showing a continuous upward trend. In terms of policy publicity, the science, technology and innovation policy scores higher than the transportation policy by 0.16 and 0.26 in terms of the fit between public services and public needs and policy performance, while the rest of the publicity aspects are all lower than the transportation policy.