公共支出结构偏向的经济波动效应研究——兼论新常态下的公共支出结构调整

The Economic Fluctuation Effect of Public Expenditure Structure

现有文献对于公共支出是否能平滑经济波动的争论仅考虑了支出总量的作用,而忽略了支出结构的影响,而支出结构偏向是中国的公共支出中的一个显著特征。本文利用1992-2012年的中国省际面板数据,通过构建一个空间动态面板数据模型,实证研究了中国的公共支出结构对省际经济波动的影响。研究结果表明,生产性支出和消费性支出占比越高,经济波动越大,而服务型支出占比的提高则能起到抚平经济波动的作用。此外,本文还发现,地区经济波动存在基于经济距离而非地理距离的空间相关性和时间上的惯性特征。

Existing literatures on the debate of whether public expenditure smooth economic fluctuation or promote economic fluctuation only consider the effect from the total amount of public expenditure, but ignore the effect from the structure of public expenditure. Public expenditure structure bias is a marked feature of China＇ s public expenditure, however, the existing literatures main focus on the effects of biased public expenditure on economic growth, income inequality, economic growth quality, etc, while few literatures pays attention to the effects of biased public expenditure on economic fluctuation. Based on China＇ s provincial panel data from 1995 to 2009, this paper empirically analyzes the effect of China＇ s local government＇ s public expenditure structure on regional economic fluctuation. Taking the spatial dependence and inertia property of regional economic fluctuation into consideration, this paper constructs a dynamic spatial panel data model to make empirical analysis. In order to get robust research conclusions, this paper construct two types index to measure regional economic fluctuation. The first method uses the rolled standard variance of the real GDP in China＇ s provinces to directly measure the provincial economic fluctuation. The second method uses the rolled standard variance of cyclical fluctuation of provincial real GDP, which is separated from long-run trends in the real GDP series, to measure the provincial economic fluctuation. Meanwhile, this paper uses two types spatial distance to construct spatial weight matrix, one type is the traditional geographic distance, and the other is economic distance. This paper uses the SYS-GMM method to estimate the coefficients in dynamic spatial panel data model. The research result reveals that the proportion of productive expenditure and consume expenditure in local government can promote regional economic fluctuation, while the proportion of service expenditure can significantly smooth the regional economic fluctuation. Meanwhile, we find that regional economic fluctuation existing spatial correlation based on the economic distance, while the spatial correlation of regional economic fluctuation based on geographic distance is not significant. In addition, regional economic fluctuation also exist inertial characteristics. The above results are robust even though we change the measure of regional economic fluctuation and spatial weight matrix. In addition to the above crucial research results, this paper also finds that the technological shock which is measured by total factor productivity has no significant effect on regional economic fluctuation, while the fiscal shock can significantly affect the regional economic fluctuation. Meanwhile, the research shows that monetary shock can motivate the economic fluctuation, but a perfect regional finance system can smooth regional economic fluctuation. The research results in this paper reveal that, government can not only employ the quantity tool, but also can adjust the public expenditure structure when using fiscal policy to smooth the economic fluctuation. These findings have great significance to adjust the traditional fiscal policy, because traditional expansionary fiscal policy will extrude private consumption while the adjustment of public expenditure structure has no effect on the private consumption.