The Identification And Measurement Of Agglomeration Economies

At present, China is faced with a great pressure in economic transition and structural adjustment. Agglomeration economies, whose cultivation has become a strong driving force of urban sustainable development and the development of urban and rural, regional coordination, is a kind of effective resource allocation effect in the economic transition and the structural adjustment process. At theoretical level, agglomeration is a core proposition in the study of urban economics, regional economics and new economic geography. But because of its invisibility, agglomeration economies have been regarded as a "black box". In addition, in the degree of economic integration and the liquidity of production factors, there are a big gap and great differences in economic geography, compared with Europe and the United States and other developed countries. Thus the agglomeration economic theory and its relevant experience originated in Europe and the United States and other developed countries, need to be further analyzed seriously and studied on its application and development in the domestic academic circle in China.Based on Marshall's classic interpretation in three aspects of agglomeration economies, and with the aim of confirming the existence of agglomeration economies and investigating the internal mechanism of agglomeration economies, this thesis carries on the identification and measurement in agglomeration economies study from multiple dimensions. On research perspectives, on the one hand, given the importance of space factor in identifying and measuring agglomeration economies, space factor is divided into two types---independent space and link space. From the viewpoint of link space, this thesis examines the geographical proximity effect and the proximity effect of organizational culture; on the other hand, the main body of agglomeration economies are divided into three levels-macro or urban level, meso or enterprise level and micro or labor force level. This thesis emphasizes the close links between behavioral agents at the three levels and strives to explore a variety of effects on agglomeration economies caused by behavioral agents'heterogeneity at each level.On technical route, considering the Circular and Cumulative Causation plays a key role in the process of production factors'agglomeration, based on the theory of matrix from deep excavation of urban spatial effect, heterogeneity and endogeneity, this thesis employs spatial analysis, multi-level analysis and micro econometric tools, to try to solve the problems, such as spatial effect, heterogeneity and endogeneity, which are contrary to the basic assumptions of classical regression. Among them, this thesis holds identification and effective treatment of endogeneity problem of agglomeration economies as the pinnacle of empirical analysis. This thesis has five chapters besides introduction part and conclusion part.In chapter one from agglomeration economies's three aspects---its connotation, its dimension and its function mechanism, this thesis defines the concept of agglomeration economies, and makes a detailed explanation on agglomeration index and econometric estimation method. Furthermore it elaborates the basis of agglomeration economies theory from three angles---the polarization theory and location selection, externality theory, and spatial economic theory.In chapter two, treating China's urban and rural economic sectors, industry, urban and geographical space as the measurement object, it analyzes the related indicators about the measurement of the agglomeration economies, adopting original geographic concentration index, location quotient and Gini coefficient, location Gini coefficient, EG index, urbanization economy, local economy and the density of employment, market potential and Moran's I index respectively.In chapter three, according to the classic model of urban economics, the new model of economic geography and unstructured method analysis framework, combined with the micro individual data and urban panel data, using instrumental variable method, the fixed effects and system GMM method, it deals effectively with endogenous problems in the process of identification and estimation of the agglomeration economies. The empirical results show that the coefficient on wage elasticity of employment density is about0.10, and the coefficient on the wage elasticity of market potential is between0.23and0.34. In other words, the results show that cities with a higher degree in employment density or market potential will get a higher capacity to pay labor force. Under the condition of controlling employment density or market potential, labor wages in larger urban land area are higher than that in smaller urban land area. It reveals the deep reason why big cities will be much more strongly attractive to labor.In chapter four, treating the identification and measurement of agglomeration economic and technological externalities as the theme, and introducing the geographic proximity and the technology proximity into spatial econometric model at the same time, it systematically analyzes the different roles that the geographic proximity and the technology proximity play respectively in the process of spatial knowledge spillover, resulting in an effective combination that the spillover effect makes in the two dimensions, the spatial dimension and the industry dimension. Based on the first and the second Economic Census data in all regions all over the nation, the Bayesian Markov Chain Monte Carlo (MCMC) methods are used to estimate the spatial Durbin model and the spatial externalities are divided into direct effects and indirect effects, so as to avoid explanations for the model parameter error. The empirical results not only confirm the presence of spatial externalities, but also find that localization economies based on technological proximity promote innovation output effectively, and spatial spillover effects based on geographical proximity are relatively weak. There are obvious conflicts between the central government and the local government on the orientation of promoting the innovation output policy.In chapter five, centered on the vicinity of organizational culture in agglomeration economies, according to the perspective of regional heterogeneity and the typical data structure that industry nests in city, using the analysis method of the Hierarchical Linear Models (HLM) and indirect estimation strategy of agglomeration economies,this thesis attempts to reverify the existence form of agglomeration economies to capture unobservable heterogeneity of agglomeration economies from the industry dimension and reveal the internal close contacts among the dimensions in agglomeration economies. The empirical results not only verify that the localization economy (MAR externality) exists in manufacturing, but also confirm that at industry-level, agglomeration economies are both influenced by industry's own characteristics, and deeply affected by the social environment of industry nested urban economy.