| Long term development of regional economy is characterized by growth of productsã€labor services and evolution of economic structureã€social and political system. Since reform and opening up, Chinese economy keeps rapid growth for years, what accompanyes this and promotes economic growth is economic structure, especially evolution of industrial structure based on factor mobility. Although most regions have established industrial structures which accordes with the principles of comparative advantage and regional divisionã€cooperation, there are some defects such as unreasonable production valueã€employment structureã€undesirable convergence of industrial structureã€low structure benefit. Under the background of transformation of economic structure caused by financial crisis〠regional coordinated development advocated strongly by central governmentã€optimizing industrial structure and enhancing regional competitiveness proposed by local governments, how to optimize industrial structure according to economic objective rules and actual conditions is very important for regionally long termã€stable and sustainable development.The dissertation aims to explore evolution law of regional specializationã€diversity, inter-relationship between region specializationã€diversity and development of regional economy, also give suggestions for policies of adjustment of industrial structure and regional harmonious development.Based on new classical trade theoryã€new trade theoryã€new economic geography〠agglomeration economy theory and theory of industrial structure, combinding with related researches, this dissertation analyzes firstly characters and evolution of regional specializationã€diversity, also relationship between them and per capita incomeã€regional size and geographical location by using data of quantity of employmentã€population size and so on.After that, this dissertaion analyzes influence of city sizeã€population densityã€the density of professional personnelã€economic development level and location on regional specialization by using FGLS method. The results suggest that, following increasing of city size and population density, the level of regional specialization decreases and this negative effect increases from east to west, it reflects the trend that influence of regional division and city crowding on regional specialization increase from east to west. The density of professional personnel improves significantly regional specialization, the trend of decreasing from east to west verifies benefits of externalities in specialization; Following increasing of income per capita, regional specialization showes the trend of "U" shape. We test this trend by using LOWESS method, the result is robust.Subsequently, to understand how specialization and diversity influence knowledge spillover and then promote innovation output also economic growth, using patent data of sity-four key cities from year2006to year2009to measure innovation output, we analyze empirically influences of regional specialization diversity on innovation ability and economic growth by constructing simultaneous equations and using3SLS method.The results from regression of procduction functions show that, the innovation ability of a city will increase0.33percent if GDP increases one percent, flexibility of capital stock to economic growth reaches0.66, it indicates that the growth is still promoted by capital mainly for most Chinese cities at present; After controlling the effect of specialization and diversity in innovation ability, diversity (the regression coefficient is0.13)is beneficial to growth, role of specialization is not significant, it suggests that knowledge Spillovers happen between different industries, concentration of many diversified industries at adjacent regions stimulate growth more than similar industries, the regression results surpport Jacobus externalities.The regression of innovation ability indicates that, economic growth is a important driving force of innovation ability, there is apparently accumulative-cycle causation between them; After controlling the impact of specialization and diversity in growth, regional specialization (the efficient is0.64) promotes growth significantly, but the rule of diversity is not; Elastic coefficient of population density to innovation ability reachs0.179and be significant, which verifies that the rule of city density in knowledge spillover, dense population distribution makes communication becoming more frequently, then knowledge spillover become more easy; R&D expenditures promote innovation too; Furthermore, there is substitutability between regional specialization〠diversity and population density, change in population density will influence locations of some industries especially manufacturing industries, then induces change of industrial structure, it influences innovation effects.Finally, Using specialization and diversity to measure rationalization of industrial structure, in combination with city sizeã€physical capital per capita, this dissertation analyzes the relationship between specializationã€divesity and economic efficiency (measured by per capita income and growth rate of GDP) by using GMM method. This part of analysis includes two samples:two hundred and eighty-six cities and forty-seven resource-based cities. Empirical results of output per capita show that, for the national sample, capital per capita promotes strongly output per capita, the regression coefficient is0.7433and be highly significant; Bigger city size, Higher output per capita, this result surpports the present policy of encouraging developing big city and pushing agglomeration of population towords cities; But synergetic effect between diversity and city size, also specialization and city size are negative, the absolute value of multiply variable between diversity and city size is more big, which suggests that, compared with diversity, regional specialization is more helpful when emphasizing the role of city size on attaining agglomeration effectiveness. Excessively attention on industrial species may causes more loss of agglomeraion effects.When using growth rate of product per capita as the explained variable, we find that, the P value of diversityã€city size and the multiply variable between them all approach ten percent, and the signs of are inverse campared with the situation when using output per capita as the explained variable, which reflects diversity is more beneficial to stable growth fo output per capita.For resource-based cities, the partial effect of output per capita to city size is positive, enlargement of city size is helpful to rapid growth of output per capita; But the coefficients of regional specialization and the multiply variable between specialization are not significant, which suggests that, from year2003to2009, specialization is not the main cause of economic efficiency in resource-based cities because of resources exhaustionã€backwardness of leading industries and so on; The partial effect of of output per capita to diversity, when population size exceeding the critical of two hundred and sixty-five thousand, per capita product will increase following improvement of diversity degree; Furthermore, when exceeding seven hundred and fifteen thousand, GDP growth rate will be more fast following increase of diversity, so do city size.In conclusions, this dissertation presents main findings of our research and their implications for policy applications, also points out the shortcomings and future research orientations. |