中国地区全要素生产率的Bayesian分析  

Bayesian Analysis of Total Factor Productivity in Chinese Provinces

作　　者：魏学辉[1,2] 白仲林[3] 

机构地区：[1]南开大学数量经济研究所,天津300071 [2]河北工业大学管理学院,天津300130 [3]天津财经大学,天津300222

出　　处：《数理统计与管理》2012年第4期585-594,共10页Journal of Applied Statistics and Management

基　　金：国家自然科学基金项目(批准号:70771072)的阶段性成果

摘　　要：本文利用全国28个省市自治区的相关数据,以索洛增长方程为基础,采用Bayesian SUR模型以及Gibbs-Importance抽样算法,估算了其资本产出弹性,并在此基础上计算了各地区全要素生产率及其增长率。研究结果表明,科技发展战略对全要素生产率的提高具有显著正效应;随着产业结构调整,资本产出弹性和全要素生产率的关系从正相关逐渐变为负相关,并且由此所表明的地区分工协作特征正逐步显现;内陆地区的地缘经济特征制约了其全要素生产率的进一步提高。Based on the input and output data of 28 provinces in China, this paper implements the CibbsImportance sampling algorithm to estimate the output elasticity of capital within the Bayesian SUR model of Solow growth equations. The Total Factor Productivity in the 28 provinces is also calculated with the estimated output elasticity of capital. The empirical analysis shows that the strategies for technological development have positive effects on the improvement of the Total Factor Productivity. Along with the adjustment of the industry structure, the relationship between the output elasticity of capital and the Total Factor Productivity gradually turns negative, and the industrial division and cooperation in regional economies have been showing up. The geographic features of the inland regions restrict the further improvement of their Total Factor Productivity.

关 键 词：全要素生产率 SUR模型 Bayesian分析 GIBBS抽样 

分 类 号：F224.0[经济管理—国民经济] O212[理学—概率论与数理统计]