The Research On City Size And Urban Productivity Across Chinese Cities

Urban productivity determines the future of a nation’s economy and urbanization is considered the inevitable result of economic development. By the end of 2011, the number of China’s urban dwellers has first time exceeded fifty percent of the nation’s population. However, China’s urbanization is still far behind the process of industrialization. The number of officially designated cities grew rapidly during the past years, but many cities are too small to benefit from agglomeration economies. In the next 5-6 years, additional 100 million residents will settle permanently in cities. Therefore, it is of great theoretical and practical significance to find the optimal sizes and the sources of urban productivity of Chinese cities with different types and structures. It is also important to reveal spatial distributional patterns of capital, industries, and heterogeneous firms across cities.Built on the previous studies in new economic geography(NEG), this paper constructs a simple NEG model that links producer services and manufacturing sectors, analyzes the synergy effects of industrial structure and city size on productivity of Chinese cities, and estimates the optimal size of cities for a given industrial structure and the marginal benefit of structural upgrading under a constraint of urban scale. The panel data regression estimates of prefectural or higher level cities show that the effects of industrial structure on productivities depend on size of the cities; a city must surpass a threshold size in order to benefit from the home market effects which stem from the linkages of upstream-downstream industries. As the urban population size increases, urban productivity experiences an inverted-U shaped progress and the marginal benefit of urban growth increases as the industrial structure shifts from manufacturing to producer services. Most prefectural or higher level cities in China are smaller than their optimal sizes. Therefore, small and medium sized cities should develop manufacturing and promote agglomeration, while developed large cities transition from manufacturing-oriented economies to service economies. The estimation also shows that the structure-size effects on urban productivity are negligible for county-level cities, since most county-level cities are still in early stages of industrialization and are too small to produce the scale economies and agglomeration economies, which are common to prefecture or higher-level cities.City size and agglomeration density are two related yet different variables, which obviously have different empirical measurements and policy implications. This paper constructs an urban agglomeration model which includes both size and density. It is demonstrated that city size and productivity have an inverted-U shaped relationship, while the relationship between density and productivity takes a U shaped curve first and an inverted-U shaped curve after city density exceeds a threshold value. There is a positive interaction effect between size and density. The panel data estimation of prefectural or higher level cities shows that population densities of all cities are larger than the expected threshold, indicating that actually productivity has an inverted-U shaped relationship with density and the optimal density grows as city size increases. Since cities hardly achieve optimal size and optimal density at the same time, the distributions of population and industries within each city should be adjusted according to city type.The size of capital accumulation varies widely across cities with different sizes and hierarchies. Part of this dissertation theoretically analyzes the effects of city size on capital accumulation based on a NEG style footloose capital model and empirically estimates a panel model of all county or higher level cities in China. The econometric results show that small cities export capital to large ones because of the agglomeration effects. Therefore, large cities may be more efficient on capital accumulation, and the growth of city size promotes capital deepening. Ceteris paribus, an one percent increase in urban population will generate a 0.22% increase in capital per capita, and the administrative function of provincial or higher level cities strengthens the influence of city size. The estimation of quantile regression model shows that the effect of city size reaches the maximum at the median quantiles of the conditional distribution of capital per capita. Policies that promote agglomeration of population and industries in less developed areas will help reverse the backwash effect of capital accumulation in large cities.Urban productivity results from firm productivities in the city. However, agglomeration of firms generates agglomeration externalities, which is also widely considered a primary cause of higher productivities of larger cities. This paper proposes a modified heterogeneous firm model to examine the effects of firm selection/sorting on productivity distribution within and across Chinese cities, and to examine the source of difference in urban productivities. The theoretical analysis shows that highly productive firms are able to set lower prices, take larger shares of markets, and thus tend to locate in larger cities, whereas less efficient firms tend to locate in smaller cities; the larger the city, the higher is the threshold of productivity required for a firm, and urban productivity results from the combined effects of firm selection and agglomeration. This paper designs a two-stage econometric method and uses China’s firm level data to estimate the selection effects. The results show that the threshold level of firm productivity increased as city size grew and the sorting effect was at least one-fifth to two-fifth of agglomeration economies. The strength of selection effects relative to agglomeration economies varies across industries. The paper also demonstrates that reducing transaction cost helps close the threshold gap between large cities and small ones and that market integration will raise productivity of small cities and close intercity gaps.Besides the market size, the linkages between heterogeneous firms and urban intermediate input sectors also influence the expected profits of firms. This paper modifies a “new” new economic geography model to take into account intermediate products described by the new economic geography. City size exerts an endogenous selection effect on the threshold productivity of manufacturing firms through downstream market demands and upstream intermediate supplies, and the linkage between manufacturing sectors and the intermediate sectors supported by a city creates a home market effect. A micro dataset of 118988 manufacturing firms in 653 county or higher level cities is used to estimate a quantile regression model and the results show that the selection effect of city size at low quantiles is at least 80% of agglomeration externalities. The positive effects of intermediate products may neutralize the selection effect on local firms in large cities. The estimation further proves that inefficient state-owned enterprises are more difficult to be eliminated by the city threshold productivity.The findings in this research suggest that as city size grows, the urban business environment becomes more inclusive and competitive. Differentiated firms and industries in large cities are supported by upstream input sectors and downstream markets and at the same time the firms must face the challenges of more entrants. The policy makers should consider the combined effects and choose a suitable development path according to each city’s size, functional orientation, and economic capacity.