中国城市超线性规模效应研究

On the Super Linear Scaling Effects of the Population Size in China Cities

社会网络理论是解释城市发展动力机制的重要理论。该理论认为城市规模对城市经济发展具有超线性规模效应,即城市收入水平增长远大于城市规模增长。但它的直接推论——城市规模越大越好,和中国城市的现实情况和未来发展前景似乎略有冲突。本文的主要工作,便是利用中国287个地级以上城市1993-2011年的数据,检验该理论对中国城市发展的解释力。研究发现:(1)若简单考察城市人口规模与城市经济之间的关联,则中国存在超线性规模效应。市辖区人口增加1个百分点,市辖区实际GDP增加1.72个百分点;(2)一旦控制教育、医疗、税收、基础设施、公共品等因素后,超线性规模效应便不复存在:城市经济发展的重要促进因素是投资、教育和基础设施等,人口过快增长反而成为城市经济发展的不利因素。市辖区人口增长率增加1个百分点,市辖区人均实际GDP至少减少0.04个百分点。

The social network theory is an important theory on the mechanism of city development. The theory indicates that the city population size has a super linear scaling effect on the city output, i.e., the increment of the city output is much larger than the city population size increment. The main work in this paper is to test the theory using the data in 287 cities from 1993 to 2011 in China. The empirical results show that： if we only consider the relation between the city size and the city output, the super linear scaling effect exists. When the population size increases by 1 percent, the output increases by 1.72 percent when the education, health and infrastructure variables are controlled for. The results coincide with the classical growth theory further, and the population size may be an obstacle to the city development. The population growth rate increases by 1 percent, the real GDP per capita will decrease by 0.04 percent significantly.