Research On The Effectiveness Of The Simulation Of Public Policy

Public policy making is decisive for the national economic growth and the improvement of civilians’ quality of life, and it is critical for the governmental administration, national economy, and social stability. In recent years, there has been a growing interest on public policy research in the field of administrative management. In developed countries, a large number of data and findings showed that scientific public policy-making methods could greatly reduce the error rate of decision-making, and improve the results of public policy implementation. Today, scientific methods of public policy-making also draw the attention of the academia of public policy in China. Using simulation modeling techniques, formulation and simulation of public policy outcomes became possible. If we can ensure the whole process of public policy simulation to be correct and reliable, the public policy simulation can be closer to the actual demand and further enhance the public policies. Therefore, verification and validation work has the important theory significance and practical value for the field of public policy.In this paper, a lot of research work had been carried out based on the VV&A test for simulation of public policy as the core research content such as the study on the theoretical details, on the architecture of VV&A test for simulation of public policy, on V&V testing process and method, on confirming work process and method, and on the verification validity of the design of the corresponding VV&A method and the process for actual public policy simulation.With respect to the fundamental theories, four dimensions of the research work had been carried out. Firstly, we clarified the definition of the three basic concepts of public policy, public policy simulation, and simulation credibility of public policy. Secondly, the complexity of public policy was studied. Thirdly, simulation theory and process based on agent modeling was introduced. Finally, the concept, principle, and method of VV&A test had been summarized.In the research on the architecture of VV&A test for simulation of public policy, the research work conducted as the following. Firstly, the simulation of public policy process based on agent modeling was decomposed into steps including demand analysis, agent modeling, environment modeling, relation modeling, and language modeling. Secondly, simulation work was carried out and concept model was delineated in view of the actual problem from the refined oil pricing policy in China. Finally, the system structure of four levels of the VV&A test was explored based on the simulation process of public policy.In the research on the V&V test of process and method of public policy simulation, the research work carried out is the following. First of all, the VV&A for the public policy was divided into the V&V process and the verification process, and the parallel V&V process was designed for public policy simulation. Secondly, the V&V process was divided into five steps including checking simulation requirements, conceptual model validation, mathematical model checking, simulation model validation, simulation model validation. Finally, according to the V&V process of public policy simulation, function test method based on the RBF neural network and structure test method based on hierarchical weight were designed for implementing the V&V.In the study of validation processes and methods of public policy simulation, we conducted the research work from two aspects. Firstly, we identified the confirmation of participation, the implementation content and the implementation process. Secondly, according to the characteristics of subjectivity of the confirmation, the hierarchical subdivision index weight fuzzy evaluation method and the questionnaire reliability and validity analysis method were proposed to make subjectivity change to objectivity for more scientific conclusions.In the research on validate the effectiveness of the VV&A process in this paper, three aspects of research work were carried out. Firstly, mathematical model and simulation model were established based on the concept model of the refined oil pricing policy in China. Secondly, the V&V test work were carried out by using function test method based on RBF neural network and structure test method based on hierarchical weight. Finally, validation processes were adopted using hierarchical subdivision index weight fuzzy evaluation method and questionnaire reliability and validity analysis method.Public policy intervention is also studied to resolve high risk economic problems, and a general model for resoving this kind of problem is proposed. In this model, manufacturer loss function is introduced and conditional value at risk is used in measuring financial risk. Empirical results show that this model can select optimal public policy for high risk economic problems.