Formation Of Industrial Assemblage In China: Theoretical And Empirical Analysis

Now the phenomenon of industrial agglomeration an important power to develop Chinese economic, and the change of them will influence the development of our country.The paper studies the distribution of agglomeration of Chinese industry and the factors important to its distribution. We attempt to address the following questions: Why are some industries concentrated in a few locations while others scattered across many provinces? What decides industrial agglomeration or dispersion and the important factors to agglomeration of Chinese industry.This paper base on the cost-surplus realized path model of industrial cluster. We use the panel data model using the data which issued from state statistical council, and estimate the local resources, external effect of agglomeration, special transaction cost and other relevant factor who works on the phenomenon of industry agglomeration. We found that the industrial agglomeration in many places are related to the local resources, but the local resources not the key reason. When the condition changed, including the change of special transaction cost and the external effect result from the agglomeration the patens of industrial agglomeration and the kind local industrial will change.In addition to the introduction and conclusion, this paper is divided into four chapters. Basic structure is as follows:The chapter 1 Base on the existing research literature and the existing research, there we present different definition and the types for industrial agglomeration. At the same time, the paper describes the basic situation of China's industrial agglomeration and the economic role of f industrial agglomeration in China..In Chapter 2, the thesis takes review and comparative study on the agglomeration theory and its development in history, we conclude the basic factors of industrial agglomeration and its operating mechanism. The scholars on this issue focus on the practical experience. I will focus on the practical experience of some scholars. Chinese scholars also made a lot research in this area, and I will describe the more representative views as well.The paper then presents a theoretical analysis frame work. Industrial agglomeration is result of same site choice of enterprises themselves. It is the conduct of enterprises, which are driven by the same objective (max the profit) produce agglomeration of enterprises. This paper base on the cost-surplus realized path model of industrial cluster. This paper based on the surplus integrality model of industrial cluster. By analyzing the cost-surplus realized path model of two types of industrial cluster.Suppose there a choice for a enterprises: Part in or not, the result rest on if or not they could max their the profit.π₁=R₁-C₁ (R₁,C₁ is the income and costwhen the enterprises take part in the cluster) andπ₂=R₂-C₂ (R₂,C₂ is the incomeand cost when the enterprises don't take part in the cluster) . Compare above, the result as followed:Suppose there are n enterprises in the cluster, so the cost-surplus:For S_i≥0 is the necessary condition for a company part in the cluster, so:Base on the exchange cost, the cost of a enterprises should include input factor cost C_F inner cost C_O, cost and exchange cost C_M , the cost-surplus divided intothere part: Empirically, it is very hard to disentangle these partial effects and therefore the net effect of technology spillovers remains uncertain.Chapter 3 presents the research methodology and the design of the econometric model. In this paper, I adopt the most popular panel data model recently to conduct a study. Panel data are also known as time series and cross section data or pool data. Panel data are two-dimensional data both in time series and cross section. I use pooled OLS model and Intercept Change model with N individual members. Intercept Change model can be divided into two types of fixed and random effects model. If the model is not set correctly, the results will be far from the economic reality simulation. The result may be different with different models, and the right model can be chosen through the following tests: F test can determine whether it is necessary to establish individual fixed effect model; LM test can determine whether it is necessary to establish random effects model; Hausman test can determine whether Intercept Change model is chosen for a fixed effect model or choose random effects model.So, there are the basic models:Pooled OLS model:Fixed effect model:Random effect: Where, the value add from i th province at the time period t. Y_it for industry agglomeration.In Chapter 4, the paper use Eviews5.0 software to estimate the model with a total of 248 samples where the data are from 2000 to 2007 and in 31 sectors. An econometric analysis further supports scale economics, local protectionism and transaction costs are important factors to agglomeration of Chinese industry.In light of the above evidence, the paper has put forward policy proposal accordingly, hoping that foreign direct investment can be a more effective access to the technology spillover effects.