我国经济增长与运输业的联动关系研究——基于贝叶斯VAR方法

Research on the Linkage between China’s Economic Growth and Transportation Industry--Based on Bayesian VAR Method

利用我国的GDP、货运量和客运量时间序列数据,采用贝叶斯模型比较方法选取最优VAR模型来分析三者间联动关系,进而捕捉变量结构的时变性和周期性特征。首先比较六种时变参数向量自回归(TVP-VAR)模型,再比较各种机制转换向量自回归(RS-VAR)模型,发现一类特殊的TVP-SV模型的对数边际似然最大。研究表明,近年来运输业与经济增长具有同期同向发展的关系,经济增长对货运量和客运量的影响非常稳定,并且经济增长有利于带动货运量和客运量需求上涨;货运的繁荣发展有利于推动客运量的上涨;改革开放以后,客运量增长是推动经济增长的重要来源,并且经济增长对客运量冲击的脉冲响应走势具有明显的时变特征;我国运输业和经济增长之间的联动关系在很大可能性上不存在机制转换,即不存在阶段性特征。

Based on the time series data of GDP,freight volume and passenger volume in China,the optimal vector autoregressive model is selected by the Bayesian model comparison method to analyze the linkage relationship among the three,so as to capture the time-varying and periodic characteristics of variable structure.Firstly,six kinds of time-varying parameter vector autoregressive(TVP-VAR) models were compared,and then various regime switching vector autoregressive(RS-VAR) models were compared.It was found that a special TVP-SV model had the maximum logarithmic marginal likelihood.The research shows that the transportation industry and the economic growth in recent years have the same development in the same period.The economic growth has a very stable impact on the freight volume and passenger volume.The prosperous development of freight is conducive to the promotion of passenger volume.After the reform and opening up,passenger volume growth is an important source of economic growth,and the impulse response trend of economic growth to passenger volume impact has obvious time-varying characteristics.The linkage between China’s transportation industry and economic growth has no regime switching,that is,no stage characteristics.