Hierarchical Bayes Estimation Of The Urban Registered Unemployment Rates

In recent years, China’s unemployment situation is very serious, have caused the government and the community wide concern. The problem of unemployment is not only an economic problem, is a political problem that matters to social stability, on the establishment of a harmonious society also posed a great threat to. The urban registered unemployment rate and unemployment relief and social security subsidies are closely related, is also China’s central government and local governments at all levels in the setting of short-term, medium-term and long-term economic and social development plan, the important control index of.The paper using the2000to2009the national and regional urban registered unemployment rate and growth rate of GDP historical data, based on hierarchical Bayesian analysis method to construct Fay-Herriot model, time series model and generalized Fay-Herriot model based on time series. The use of R software using MCMC dynamic simulation technology of the three models respectively under the eastern coastal areas of China, the central region and the western region of the registered urban unemployment rate of data fitting, and using posterior predictive deviation method of the model’s validity and robustness are compared.Conclusion:Fay-Herriot model the posterior predictions from difference with respect to time series model and time series based on generalized Fay-Herriot model the posterior predictions from larger difference, time series model the posterior predictions from the minimum value, illustrate three model time series model for optimal. However, the revised based on the time series of the generalized Fay-Herriot model the posterior predictions from difference compared with time series model has less significant slants big, show our country market to the economic area labor resources and effective configuration, resulting in each economy area registered unemployment rate in cities and towns do not related, that is based on time series generalized Fay-Herriot model in the introduction of regional factors, to simulate the effect of reduced instead.In addition the paper in eastern coastal area as an example use time series model to calculate the area in2010registered urban unemployment rate estimates, the results show that the prediction result is good.