空间效应视角下中国城市土地城镇化的驱动因素

Driving factors of urban land urbanization inChina from the perspective of spatial effects

从空间效应视角下分析城市土地城镇化驱动因素可以为优化城市土地资源可持续开发利用提供新参考。基于2003—2017年中国283座城市面板数据,构建经济地理空间权重矩阵,借助空间面板杜宾模型及其偏分分解方法分析中国城市土地城镇化驱动因素。研究表明:(1)研究期内中国城市土地城镇化水平呈持续上升特征,城际差异呈持续扩大特征,并伴随着极化现象。(2)中国城市土地城镇化在地理空间上并非随机分布,而是存在显著正向全局空间自相关性,局域空间格局特征表现为高值集聚、低值塌陷、低值集聚和高值凸起四种类型,数量上以HH和LL为主,以LH和HL为辅,并且存在"低者恒低、高者恒高"的空间格局特征。(3)中国城市土地城镇化存在显著正向空间溢出效应,本地城市土地城镇化水平提升会驱动邻近城市土地城镇化水平提升,这打破了传统计量研究中样本相互独立的基本假设,只有将空间效应纳入城市土地城镇化的计量回归模型之中,才能合理解释城市土地城镇化驱动因素。(4)经济发展、第二产业和土地财政对本地城市土地城镇化具有正向直接效应;经济发展、人口城镇化和对外开放对邻近城市土地城镇化具有负向溢出效应,土地财政对邻近城市土地城镇化具有正向溢出效应。据此,管控城市土地城镇化不可"孤城而为",需要综合考虑邻近城市之间空间溢出效应的存在,发挥城际之间在推进土地城镇化理性发展中的合作联动治理作用。

Analyzing the driving factors of urban land urbanization from the perspective of spatial effects can provide a new reference for optimizing the sustainable development and utilization of urban land resources.Based on the panel data of 283 cities in China from 2003 to 2017,by constructing the economic geography spatial weight matrix,the spatial panel Durbin model and its partial decomposition method were used to analyze the driving factors of urban land urbanization in China.The results indicated that:①During the study period,the level of urban land urbanization in China continued to rise,and inter-city differences continued to expand,accompanied by polarization.②Urban land urbanization in China was not randomly distributed in geographical space,but there was a significant positive global autocorrelation.The characteristics of the local spatial pattern demonstrated the following four types:high-value clustering,low-value collapse,low-value clustering,and high-value convex.The quantity was mainly HH and LL,supplemented by LH and HL,and there was a spatial pattern characteristic of‘the low being constantly low and the high being constantly high’.③There was a significant positive spatial spillover effect of Chinese urban land urbanization.The increase in the level of local urban land urbanization would drive the improvement of the level of urban land urbanization in the neighboring cities,which breaks the basic assumption that the samples are independent of each other in traditional econometric studies.Only by incorporating spatial effects into the econometric regression model of urban land urbanization can we reasonably explain the driving factors of urban land urbanization.④Economic development,secondary industry,and land finance had positive direct effects on local urban land urbanization;economic development,population urbanization,and opening up had negative spillover effects on urban land urbanization in the neighboring cities,and land finance had positive spillover effects on urban land urbanization in the neighboring cities.According to this,the control of urban land urbanization should not be applied to‘solitary cities’.It is necessary to comprehensively consider the existence of spatial spillover effects between neighboring cities and to play the role of inter-city cooperation and linkage governance in promoting rational development of land urbanization.