A Study On The Optimal Size Of China's Cities From The Perspective Of TFP

Expansion of city size have both agglomeration economy side and agglomeration diseconomy side. So in theory there should be an optimum size of city or a suboptimum interval within which marginal revenue of agglomeration offest marginal cost, aggregate net income is maximized.House price is an indicator of agglomeration diseconomy and it is dispersion foce of a city. Especially in China, cost in buying a house accounts for quiet large part of household spending, but the cost is not taken into account in CPI. Therefore, we do not think CPI can well reflect the regional difference of cost of living. In this thesis, we build a weighted cost of living index(COL) by taking account in both house price and CPI, we compared the per capita GDP adjusted by COL, referred to as "real per capita GDP”. And found that, although the nominal per capita GDP of Beijing, Shanghai, Guangzhou, Shenzhen etc ranked top 25, due to the high cost of living, real per capita GDP ranked out of top 25 away.Based on Henderson's city model,we researched the relationship between the TFP of city and the city size, the data used in this article is 2005~2013 China's perfectual cities. We found inverted U-shaped relationship between city size and TFP. We estimated the peak value and 95% confidence leavel interval of optimum city size under current industrial structure, and we found that 15.2% of China city were oversized and 54% of China city were undersized.To make a comparison with existing literature, we add a model studied the relationship between city size and urban labor productivity as others did. Our result and optimal city size was the same with existing literature when using the same model. But the optimal city size was smaller when using TFP as dependent variable.