【摘要】：Rapidly developing and prosperous urban area is contrastive with economic disparity. Spatial inequalities in income, health, education are observed all over the world. Inequalities change with rapid economic growth. For example, Fujita and Hu (2001) report that the absolute gap between coastal and interior regions in China increased by 10 times in the early years of its economic reform. An important question is whether economic growth always accelerates spatial inequality. Since Williamson (1965), many empirical studies find systematic evidence about the uneven regional and urban development. In particular, an inverted U-pattern relationship between inequality and development is observed in the literature for countries, regions and cities: regional disparity increases in early development stages, then stabilize, and then diminish in mature periods of growth. Nevertheless, such a fact is not verified theoretically with economic mechanisms yet. Spatial inequality is usually criticized from the viewpoint of equity, which may result in various social problems and cause social instability. However, inequality is also beneficial because it increases economic efficiency. Regional specialization based on comparative advantage improves the productivity. A theoretical welfare analysis is necessary to conclude which one is larger. This paper examines spatial inequalities in industrial location and income, answering the following questions. What are the causes of spatial inequality? Does economic growth accelerate spatial inequality? Is the economic inequality beneficial or harmful to local residents? From the perspective of theory, spatial inequality is fundamentally determined by the location of firms and labor. Therefore, we use a New Economic Geography (NEG) approach inthis study to incorporate the production technology of increasing returns to scale and monopolistic competition in the industry. As shown in Krugman and Venables (1995), NEG approach is useful in explaining why firms agglomerate in large cities, resulting to spatial inequality in firm location. International and interregional trade plays an important role which explicitly depends on the transportation (trade) costs. However, existing models assumes a homogeneous agricultural good, which is traded costlessly across regions. The nominal wages of labor in all regions become equal, so the authors fail to disclose the spatial inequality in income. For this reason, we reconstruct an NEG model by deleting its agricultural sector. Specifically, we revise the footloose-capital model of Martin and Rogers (1995). We choose this framework because the assumption of two production factors, mobile capital and immobile labor, fits Chinese economic situation well. Due to the Hukou system, labor mobility is subject to tight restrictions. However, capital mobility is predominant among regions and cities. The simple model allows us to observe the spatial inequalities in both firm location and income simultaneously. Spatial disparity turns out to be pervasive and persistent. The inverted U-pattern results from a balance of market size effect and production cost effect. These results are further generalized to a space of multiple regions/cities. Welfare analysis shows the possibility of a conflict of interest among regions/cities. Our results have many policy implications with respect to urban transformation. Firstly, lump sum subsidies to residents in rural area might exacerbate the industrial agglomeration in cities, and fail to reduce regional inequality. Secondly, since spatial inequalities will diminish in the future, it is important for regions and cities to develop their own resource correctly to attract manufacturing firms to locate there during the urban transformation process.