| Official statistics is one of the most important and basic information resources. The quality of statistical data influences the science of Macroeconomic policy-making. However, the quality of official statistical data is a focus of concern at home and abroad, either in theory or in practice. A long period of time,the implementation of statistical systems at the existing system is showing a lot of drawbacks. The most salient point is beyond the control of the quality of statistical data. In the process of statistical data generation, there are many factors that lead to data distortion. One of the impacts is human factors and the most serious and difficult is the prevention and control of human factors. Because of external intervention, such as the involvement of non-technical factors, the data quality problem of Chinese official statistics is more complex than the Western country's,which challenge China's statistical system reform.This paper bases on the literature review both at home and abroad. Firstly, law enforcement and the empirical analysis reflected problems. Big difference between our provinces, municipalities and autonomous regions reported GDP and total population statistics and the central government reported GDP and total population shows that low quality of statistical data of our government. From the perspective of empirical data quality found the non-technical causes of the problem, which may be more statistical system and incorrect view of the performance and so on.And then the article explores the basic relationship and the interests and conflict of stakeholders from the statistics practice, statistical laws and regulations in the process of data generation. Based on the experience, use game theory to build game model of statistical data, which roughly simulate the main game process of statistical data.In this paper, construct statistical data game models, don't arbitrary assumptions or ignore the relationship between game players of statistics, randomly set function. The paper find out the statistics on the actual game and the interests of their respective concerns through the analysis of the economic interests, which made the statistical model of the game's decision-making shift from discrete variables to continuous variables; game's ration from the entirely rationality to limited rationality; the order of the game actions shifted from static to dynamic, and make models more practical.Based on analysis of the statistical data game model, the paper inspects how the relevant statistics influence the data quality in the data generation process, and seized the interest adjustment and make in-depth discussions in the balance, counter-measures of management. The main conclusions are: the system of government, the Census and Statistics management system caused the basis and conditions of the occurrence of unreasonable Game. Inside and outside of statistics data production process, game player or institutions had motives and opportunity to interfere data generation process. In the data generation process, there is a variety the contradictions and conflicts of individual rationality and collective rationality about the level of data quality. Official intervention is the main source of problem, which make behavior of statistical data producers complex, and which the trust of users and the data generation quality destructive. Finally, this article summarized the statistical data generation balanced view of the game and put forward corresponding data quality management measures and the direction of further study. |
