用DEA-Malmquist方法测算2000年和2010年中国209个地级市市辖区的城市效率,用非参数方法估计了我国城市效率在不同规模城市间的分布,并实证了城市效率变化的机制。研究发现:2000~2010年我国城市最优规模从350万人增加至717万人,增长104.86%,而城市规模均值仅增长50.47%,城市集聚不足的情况加剧,城市效率整体下降;城市效率变化和城市规模呈U型曲线关系,处于544万左右的城市效率降低最快,改善基础设施有利于提高城市效率;半参数估计表明计量设定和回归结果是稳健的。
This paper uses DEA-Malmquist method to measure the city efficiency of 209 cities in China from 2000 to 2010,estimates the distribution of city efficiency in cities with different scale using nonparametric regression to work out the optimum size of cities,and investigates the mechanism of city efficiency change. It finds that,from 2000 to 2010,the optimal scale of Chinese cities change from 3. 5 million to 7. 17 million,increasing by 104. 86%; but the average scale only increases by 50. 47 percent; the results in the undersize of Chinese cities become more serious and the overall efficiency of Chinese cities decline during these years. It finds that there is a U shape curve between city efficiency change and city initial size; improving urban infrastructure can upgrade city efficiency change. It finds that the U shape curve relationship is robust using semiparametric regression to examine the accurate relationship between city efficiency change and its initial size.