本文分析了2000-2010年中国城市规模的变化规律,借鉴集聚经济理论对上述变化规律给出合理解释,并指出以不同产业为主的城市之间的适度规模差异巨大。首先,使用第五次人口普查、第六次人口普查中全国地级市的城镇地区人口数据,文章首次计算并比较了2000年和2010年中国城市的规模位序。结果发现总体上我国的城镇体系发展较为平衡、首位度低,10年间中国城市体系中人口往大城市集中的特征明显,高位次城市发展突出;但是在规模前100位的城市中,人口分布则更为平衡。这表明10年间中国一批大城市发展迅速,人口分布存在着明显的“集中与扩散并存”的趋势。进一步地.基于集聚经济理论,文章构造了城市单位生产成本函数,指出城市的单位生产成本伴随着城市规模的扩大呈现出“U”型曲线特征,城市规模不能无限制的扩大。本文将城市的适度规模界定为城市单位生产成本最小时所对应的规模,利用中国地级市的面板数据,采用异方差稳健性估计的固定效应模型,测度了中国城市的适度规模。测算结果表明,在其他条件不变的情况下,城市的适度规模为420万人。这一结果也同时证实了目前中国的城市规模是普遍偏小的,人口往大城市迅速集中可以降低城市的生产成本,当前城市体系的发展趋势符合经济理性。此外,文章认为,现实中各个城市之间的产业结构差别巨大,这使得上述对城市适度规模的估计只是一个“一般情况”下的估计。一个城市的适度规模部分地取决于这个城市的产业特征。一个城市中追求集聚效益的产业越多,那么这个城市的适度规模就会越大。在界定制造业集聚型城市和生产性服务业集聚型城市的基础上,文章的实证结果进一步表明,与“一般情况”下的城市适度规模相比,制造业集聚型城市的适度规模较小,生产性服务业集聚型城市
This article analyzes the changes of Chinese city scale from 2000 to 2010, and explains it based on agglomeration economy rule. It is pointed out that there is a big difference between the moderate scales of cities with different industries. Firstly, through the analysis of the fifth and sixth national population census data of Chinese municipalities, the paper for the first time calculates and compares the rank-size of Chinese cities in 2000 and 2010. We find that Chinese whole city system was developed in a relatively balanced way, with a low first degree. In the 2000 -2010, population tends to concentrate in the big cities, and those big cities in high rank were better developed. However, in the top 100 cities, population was found distributed more evenly. These results demonstrate that there are a lot of big cities that developed rapidly in 2000 -2010 and the tendencies of concentration and diffusion coexist in population distribution. Secondly, we construct the urban unit production cost function based on agglomeration economy rule, and point out that the city' s unit production cost presents a characteristic of "U"-type curve and city scale cannot be expanded unlimited. Then we define the appropriate scale of a city as the scale of the city with minimum unit production cost. Using the panel data of Chinese municipalities, based on fixed effect model with estimation of the robustness of Heteroscedasticity, the moderate scale of Chinese cities was calculated. The result shows that, other things equal, the moderate scale of Chinese cities is about 4.2 million. It means most of cities in China are undersized, and the current development trend of urban system accords with economic rationality. Since the industrial structure of different cities is different, our estimation of moderate city scale is only for general condition. The moderate scale of a city depends partly on the industry structure of a city. The more a city has industries in pursuit of agglomeration benefits, the bigger the moderate scale is.