中国人口年龄结构对地区居民消费的差异影响——基于空间面板数据的地理加权回归方法

Research on Different Influences of Chinese Population Age Structure on Household Consumption Based on Spatial Panel Data Geographically Weighted Regression Model

为研究中国人口年龄结构变化对各地区居民消费率的差异影响及居民消费的空间溢出效应,从局部带宽理论出发构建兼顾空间相关性和空间异质性的空间面板数据地理加权回归模型,并推导出混合效应、固定效应和随机效应设定下的极大似然估计量。实证结果表明,儿童抚养系数的变动对相对落后地区的影响较大,其上升会降低中部地区的居民消费率,并提高西部地区的居民消费率;老年抚养系数的上升会提高中、西部十个省份的居民消费率;儿童和老年抚养系数对东部和东北地区几乎都没有影响。此外,部分省份的居民消费率受周边省份居民消费的溢出影响。

To study the effect of Chinese population age structure on household consumption and the spill- over effect of household consumption, the paper starts from the theory of local bandwidth, constructs the Spatial Panel Data Geographically Weighted Regression （SPDGWR） model, and deduces the maximum like- lihood estimation under the specification of pooled effect, fixed effect, and random effect. The results indi- cate that： the rise of child dependency ratio will reduce the consumption rate of the central region, and in- crease the consumption rate of the western region; the rise of old age dependency ratio will increase the con- sumption rate of the middle and western region; child and old age dependency ratios have no effects on the consumption rate of nearly all the eastern and northeastern regions. Besides, the consumption rate of partial provinces is influenced by the spillover effect of household consumption in their nearby provinces.