Industrial Agglomeration Mechanism, Under The New Economic Geography Perspective

The dissertation focuses on two aspects. On a theoretical level, the dissertation tries to establish a research framework of the mechanism of industrial agglomeration by making use of monopolistic competition and increasing return, which are the basic assumption of new economic geography. The aim is to generalize a simple, clear and refined theory of the mechanism of industrial agglomeration, which involves dimensions both of time and space and can be used to explain all kinds of industrial agglomeration. On a practical level, by using the above theory of the mechanism of industrial agglomeration, the dissertation tries to explain the basic configuration of regional economic difference since china's reform and opening up especially during the the recent more than ten years- the eastern coastal region's economic growth being fast while the middle and western inland region being much slower (i.e., "the eastern region rising") and the economic growth of the generalized defined middle region being slower than not only the east region but also the west region after 1997(i.e., "the middle region sinking").The definition of "industry" here is extended from manufacturing industry or secondary industry to non-agricultural industries including service industry or tertiary industry. The definition of "industrial agglomeration" has two meanings. One is that industry gathers spatially by transferring. The other is that industry develops inherently in all kinds of locations. In other words, we can conclude that once a region's industries share increase, there occurs industrial agglomeration. The dissertation stresses that "disperse" and "dispersion" are the opposite side of "agglomerate" and "agglomeration". The terminologies express their dynamic and static implications respectively. The opposite two sides are dialectical unity and can not be viewed static and isolated.The concept of "industry" is not limited to manufacturing industry. "Industrial agglomeration" is closely relevant to natural resources, location factors, industrial linkages, thick market for specialized skills, technological spillover, scale economy, increasing return, land rent, land utilization, transport cost, market area, innovation, commuting cost, external economies, market potentiality, division of labour, specialization, trade cost etc. The dissertation gives a literature review on these aspects as different earlier economists are involved. Those earlier economists mainly include the great economist of Neo-classical economics Alfred Marshall, several important German representatives of classical location theory, and many others of different research mainstreams before the appearance of new economic geography. As Dixit-Stiglitz model of monopolistic competition, new trade theory and new economic geography being put forward, the important theoretical basis or basic assumptions of monopolistic competition and increasing return for explaining the phenomena of industrial agglomeration are well founded. Having summarized further, the basic research framework of the mechanism of industrial agglomeration is established.Then, based on the literature review and basic economic theory, the dissertation divides the fundamental factors that affect industrial agglomeration and dispersion into four kinds: scale economy, consumer preference, spatial cost and expectation. The dissertation analyzes each factor independently and indicates that technological and institutional progress driven by labor division can also be thought as separate fundamental factors. The dissertation continues to analyze quite a number of models of new economic geography and illustrates that the mechanism of industrial agglomeration in all the models can be treated as some dynamic circular causality. The circular causality originates from a series of effects: the market access effect, the cost-of-living effect, the cost-of-production effect, sharing effect, matching effect, learning effect or spillover effect. The mechanism of industrial agglomeration in reality is an intertwining of all kinds of dynamic circular causality or agglomeration forces originated by these effects. While, the mechanism of industrial agglomeration is also restrained by dispersion forces originated by the following factors: the immobility of land, spatial cost (including transport cost, trade cost and crowding cost), market crowding effect and some expectations.The increase of trade freeness caused by technological and institutional progress is in favor of agglomeration forces to conquer dispersion forces. This promotes industrial agglomeration and achieves scale economies of production and consumption better. It is a long-time dynamic process. In the short-run, there usually exists a critical condition about trade freeness which determines whether the agglomeration force can conquer dispersion force. The increase of trade freeness near a critical condition will lead to a catastrophic agglomeration. Therefore the occurrence of industrial agglomeration has the characteristic of having stages, which is inclined to widen the regional economic difference due to the market crowding-out effect of interregional trade. The paper also matures the Smith theorem and Yang theorem further by adding spatial dimension.Theoretically, the dissertation also indicates that although the mechanism of industrial agglomeration itself is a positive feedback effect, it generally operates under the external conditions of location or economic geography factors, effects of accidental factors and its historical accumulation, policy changes, and expectations. The mechanism of industrial agglomeration exists despite of the change of time and space. As time goes by so that trade freeness caused by technological and institutional progress increases, the largest spatial level or scope the mechamsm of industrial agglomeration can affect enlarges. As time of the modern information and globalization occurs in the global level recently, the mechanism of industrial agglomeration exists and overlaps in different kinds of spatial levels, dominating industrial agglomerations in different spatial levels. The strength of industrial agglomeration will generally increase as time goes by, and decrease as distance of space extends. The notable phenomena of modern industrial agglomeration are based on the free market economic system.Then, based on the data of different provinces and the relevant literatures, the dissertation gives statistical and descriptive analysis and has an econometric empirical test between the flow of labor, capital, narrowly defined natural resources and economic changes, the agglomeration degree of non-agriculture industry. The dissertation proves that the basic configuration of regional economic difference since china's reform and opening up especially during the recent more than ten years (i.e., "the eastern region rising" and "the middle region sinking") is superficially the result of lopsided distribution of non-agriculture industry in inter-provincial or national level. It is also the result of non-agriculture industry's agglomeration trends in national level. By further analysis, it is found that "the eastern region rising" is due to the interactions of three kinds of factors: location advantages, policy advantages and the circular causality mechanism of industrial agglomeration (also called factors of economic geography, policy and new economic geography). It is a long-time process. As the circular causality mechanism of industrial agglomeration strengthens during "the eastern region rising", the eastern coastal region strengthens its effect on the middle and western inlands more and more. Large amount of production factors in the middle and western inlands flow to eastern region faster and faster. The industries of the eastern region occupy the market in the middle and western regions more and more. The industrial agglomeration internally in the middle and western regions are negatively affected. The middle region is more negatively affected because of its spatial proximity to the eastern coastal region. Accumulated to a certain extent, the economic growth rate of the middle region becomes lower than both the eastern and western regions (i.e., "the middle region sinking"). Correspondingly, the basic configuration of regional economic difference is dominated by the above trend. The strategies of western development and the revitalization of old industrial bases in northeastern do have their promoting effects but can not balance the negative effects caused by the mechanism of interregional industrial agglomeration. The strategy of western development has more positive effects on some most western provinces because far distance stops the large-scale production factors flow out to the eastern region. The market of some most western provinces has not suffered shocking occupying and the production factors together with local market there support the development of local industry better.The dissertation incidentally discusses the broadly concerned phenomenon of "resource curse". It indicates that without considering the location factors comprehensively, the causality will be farfetched and narrow-minded if only it is inferred by economists from the negative correlation between resource richness and economic growth displayed by the phenomenon of "resource curse". It also indicates that the regions with abundant energy resources and mineral resources usually do not have location advantages (e.g., near coast), abundant water resources, suitable climate for densely populated. So resources are often attracted by other regions different from their original regions while not put into production locally to form large-scale industrial agglomeration. The Chinese researchers' identification of the law of "resource curse", which is tightly related to Chinese regional economic difference, can be largely explained by the circular causality mechanism of industrial agglomeration.Lastly, the dissertation provides policy suggestions for promoting harmonious development of regional economy. The core viewpoints include promoting interregional trade, deepening interregional labour division, strengthening interregional exchanges, eliminating obstacles of urbanization, promoting industrial agglomerations in all the regional or spatial levels, encouraging migration and population flow, reducing spatial cost, and guiding expectations. The dissertation also provides more concrete suggestions for promoting the development of the middle and western regions especially "the middle region rising".