Research On Economic Convergence Of Chinese Cities Based On Spatial Spillover Effects

Since Solow and Swan put forward the neoclassical growth model,the theoretical exploration and empirical research on economic convergence has been the focus of scholars.They attempt to find the evidence on how poverty areas catch up with affluent areas,in the dialectical relationship between efficiency and fairness.However,the neoclassical growth model shows limits in explaining why some poverty area remains poor.Correspondingly,the theory of endogenous growth could interpret the phenomenon of path dependence among areas.It does not mean that endogenous growth theory denies the possibility of economic convergence.For example,the expanded model of Schumpeter based on technology transfer and Nelson and Phelps method based on human capital accumulation provide the theoretical evidence of the economic catch-up.However,the traditional neoclassical growth model and endogenous growth model do not take into account the spatial property of the areas.Once realizing that the area is open spatially,ingoring the spatial dependency will lead to some bias.The concept of space in economic theory could be embodied until Krugman introduced the monopoly competition model into the analytical framework.So,will the areas converge?So far,there is no definite answer despite there is plenty of sample inspection.Since China's reform and opening up,the relaxation of production factors promoted the eastern region to gather a large number of superior resources,forming the unbalanced spatial economic distribution.Due to the economic reform from the concept of'takeing count of equity simultaneously' to 'paying more attention to social equity',the economic gap is supposed to be relieved,which is also confirmed in a large number of empirical studies.It is found that most of the research either neglected the interregional spatial correlation among areas,or started the study from the provincial administrative units,which is considered as inappropriate.The cities have become important carriers of economic growth due to rapid urbanization since the year 2002.In addition,the spatial dependence among the cities has been enhanced with the improvement of the transportation infrastructure.Therefore,it is necessary to study the economic convergence from the angle of cities based on the spatial spillover effect.Based on this,the study focuses on cities' economic convergence based on different spatial econometric methods and multi-dimensional indicators,showing extension in three aspects.Firstly,the spatial economic convergence is analyzed through different spatial weight matrices and spatial econometric methods.The construction of spatial weight matrix is the key of spatial econometric analysis.In this study,three kinds of spatial weight matrices are constructed according to the different relationship among cities,and the empirical results are analyzed according to the operation mechanism of different spatial econometric models.Secondly,the spatial convergence of productivity and its decomposition is analyzed.Productivity plays an important role in analyzing economic convergence.Based on the estimation of cities' depreciation rate,the study uses the stochastic frontier model to measure the productivity and its decomposition.Based on this,the study further study the convergence of productivity,and the relationship between productivity convergence and economic convergence.Thirdly,using the data of cities' nighttime lights to correct the convergence results.GDP shows subjectivity and limitation although it is considered as important measure of economic development.The study uses the data of corrected global nighttime light to replace of GDP,in order to study the cities' economic convergence objectively.In terms of specific context,it is found that the convergence trend of per capita GDP and total factor productivity(TFP)is not obvious from 2002 to 2014 in the absence of spatial relevance.The club convergence of different regions is still unobvious,indicating that the economic differences among cities have not been effectively relieved.However,the data of cities' nighttime lights shows obvious convergence,especially in the neighboring cities.It is found that there is not only significant spatial dependency among cities,but also an increasing trend over time,through analyzing the index of Moran and the figure of LISA agglomeration.Therefore,the study further analyses the economic convergence based on spatial spillover effect.Firstly,it is found that there is absolute spatial ? convergence and condition spatial(3 convergence of per capita GDP,despite the convergence trend is weak.The convergence rate based on economic distance weight matrix is largest,showing that the cities with economic connection are easier to converge.Compared with the results of GDP per capita,the convergence trend of urban nighttime light data is obviously enhanced,especially in the neighboring cities.The distinct results are related to the data source.The corrected data of urban nighttime lighting has been widely concerned by scholars it can reflect economic scale from a more objective perspective.The conclusion implies that the real difference among cities may not be as large as economic statistics index.Therefore,there is some boundedness of single index analysis of economic convergence.With the rapid expansion of population and land,the administrative boundaries among cities(especially among neighboring cities)become obscure,bringing in strong urban integration.The influence mechanism of urban expansion on economic convergence is constructed under the framework of spatial Solow model.It is found that the increasing population of the neighboring cities is not conducive to the cities' own economic growth,concerning the fact that there is competition of human resources among cities.Conversely,the expansion of urban land scale is able to to weaken the boundary effect among cities,which is conducive to the economic convergence among cities.TFP determines the trend of the economy to a large extent.Further analysis of the producitivity convergence shows week convergence trend under the condition of different spatial weights,showing similar trend of the per capita GDP.However,there is significant convergence of productivity after introducing control variables,showing that factor aggregation leads to the difference of productivity,in which human capital plays an important role.The result shows that the productivity does not promote the convergence of per capita GDP,but is similar with the result of cities' night light.In terms of the decomposition index of productivity,the unbalanced scale distribution of cities leads to divergence of scale efficiency and technical efficiency,implying that scale efficiency and technical efficiency are likely to become potential factors to reduce the difference of cities'productivity in the future.Plenty of studies have shown that the probability of convergence is significantly smaller than that among regions with similar structural characteristics.The study takes the five state-level urban agglomerations located in different locations,in order to analyse the club convergence.It draws the following conclusions.Firstly,there is no convergence trend of per capita GDP due to the large difference in the economic scale within the Pearl River Delta urban agglomeration,and among the other four urban agglomerations the economic difference of Chengyu urban agglomeration is the smallest.The results of nighttime light data were not fundamentally changed,but only in the strengthening convergence trend.Secondly,there is no convergence of productivity due to the disparity of the productivity in Beijing-Tianjin-Hebei cities and the uneven spatial distribution.Among the other four urban agglomerations,the Yangtze River Delta urban agglomeration has the strongest convergence trend.Thirdly,further analysis shows that the Pearl River Delta urban agglomeration with the highest level of marketization has the largest convergence rate of technological progress,while the Beijing-Tianjin-Hebei urban agglomeration and the Yangtze River Delta urban agglomeration have strong convergence tend of scale efficiency.When it comes to the Chengyu urban agglomeration and urban agglomeration in the Middle Reaches of the Yangtze River,the convergence trend of productivity decomposition is not obvious.In addition,it is also found that there are uneven levels of resource competition or resource mismatch in the five urban agglomerations,in which the scale structure of the urban system needs to be further improved.The study puts forward some policy suggestions to promote the coordinated development of cities based on the studies,such as the accumulation of human capital,breaking the administrative boundary,strengthening the industrial integration,improving the structure of city system and building the city value chain.