我国居民消费季节调整和节日效应测算  

China's Consumption Seasonal Adjustment and Holiday Effect Estimation

作　　者：桂文林[1,2] 韩兆洲[1,3] 

机构地区：[1]暨南大学 [2]中国社会科学院数量经济与技术经济所 [3]暨南大学教育学院

出　　处：《统计研究》2015年第2期60-68,共9页Statistical Research

基　　金：国家哲学社会科学基金项目"新型工业化指标体系及测度方法研究"(10BJY050);教育部人文社会科学基金青年项目"中国居民消费价格指数数据质量优化与通货膨胀治理"(13YJC910004);中国博士后科学基金项目"中国季节调整模型设计与环比增长率测算技术研究"(2012M520509)资助

摘　　要：Bayes季节调整方法因有坚实的理论基础,调整效果优于其他方法等,日益受到广泛的重视与应用。本文将Bayes季节调整模型引入国内,同时在模型中补充贸易日和闰年的影响。用R软件Timsac包中的Bayesian程序实现对社会消费品零售额的季节调整和环比增长率测算,测算结果表明,长期我国社会消费品零售总额具有稳定的指数增长趋势和U型季节特征,得到的月环比增长率反应灵敏。通过季节指数抛物线拟合,得到"五一"和"十一"节日经济效应和比例。总体上"五一"的节日效应显著;"十一"有正面效应,但影响不显著。With solid theoretical foundation and better adjustment results, the Bayes seasonal adjustment method is increasingly widely paid attention to and applied. This paper introduces the Bayes seasonal adjustment model and adds effects of the trade day and the leap year to the model. Seasonal adjustment and chain growth rates of total retail sales of consumer goods are estimated by the Bayesian procedure of R Timsae software. It is shown that in the long term China＇s total retail sales have a stable trend of exponential growth and seasonal characteristics of U-shaped, and estimated growth rates on a month-to-month basis are sensitive. The “May Day”and “National Day” holiday effect and the proportions are obtained by parabola fitting of seasonal index. “May Day” holiday effect is significant, and “National Day” holiday effect is positive hut not significant.

关 键 词：贝叶斯模型 季节调整 居民消费 节日效应 

分 类 号：F222.3[经济管理—国民经济]