Statistical Analysis of the Dynamic Structure of China's Economic Sectors Based on Bayesian Modeling
This paper aims to develop an alternative production function-based approach for analyzing economic fluctuations at the sectoral level, by applying Bayesian techniques. To estimate total factor productivity (TFP) and elasticities of output with respect to factors of production, we incorporate smoothness priors into statistical models based on sectoral production functions. In addition, we consider that TFP generally varies smoothly; however in some situations there may be abrupt changes. Therefore, to relieve difficulties resulting from abrupt changes in TFP, a new method, termed the random grouping method, is introduced. Compared with the conventional production function approach, a main advantage of our proposed methods is to make detailed analysis of complex movements of TFP possible by flexible modeling. As an illustrative example, we examine TFP trends in primary, secondary, and tertiary sectors in China during 1978- 2004. Estimation results suggest remarkable differences in TFP trends among the three sectors. For the period 2000-2004, the secondary and tertiary sectors have experienced stagnation in TFP growth, although in differing degrees.
Information 13(3(B)), 923-939, 2010-05
International Information Institute