中国省份消费水平指数的波动源头研究  

Research on Fluctuation Source of China's Provinces CPI

作　　者：尹静茹 李兴绪[1,2] 胡小朋[1] 刘丹[1] 

机构地区：[1]云南财经大学统计与数学学院 [2]云南省经济社会大数据研究院,昆明650000

出　　处：《统计与决策》2017年第15期11-17,共7页Statistics & Decision

基　　金：国家自然科学基金资助项目(71263055)

摘　　要：文章通过构建基于序贯主成分方法估算的状态空间分层动态因子模型,对我国1999年1月到2015年6月31个省份的CPI即消费水平指标的三层动态因子进行估算。将CPI波动来源分解为相互独立的国家、区域和省份动态因子,进而分析其基本特征,结果表明全国动态因子和区域动态因子基本反映了各省份消费水平的波动趋势,可作为CPI波动的主要源头,然而省份动态因子差异较大,其差异性与各省份经济发展水平相关。并利用方差分解来解释CPI波动源头,其中,全国动态因子是CPI波动的主要驱动力,区域动态因子影响力最弱,省份动态因子的影响程度差异较大,在一定程度上与各区域及各省份的特殊特征有关。Through constructing a hierarchical state-space dynamic factor model （DFM） based on sequential principal components estimation, this paper estimates and calculates three dynamic factors of the CPI of China＇ s 31 provinces from January 1999 to June 2015. The paper decomposes CPI fluctuation source into three independent factors： national, regional and provinces dynamic factors, and then analyzes their basic characteristics. The results show that the national dynamic factors and regional dynamic factors, which basically reflect the fluctuation trend of the consumption level among the provinces, can be taken as the main source of CPI fluctuations. However, provinces dynamic factors are quite different, which is related to the economic development level of the provinces. The paper also uses variance decomposition method to explain CPI fluctuation sources, of which national dynamic factor is the main driving force. Regional dynamic factor is of the weakest influence. The degree of influence of provinces dynamic factors are quite different, which to some extent is related to the particular characteristics of each region and each province as well.

关 键 词：省份CPI波动源头 分层动态因子模型 方差分解 

分 类 号：F064.1[经济管理—政治经济学]