Purpose – The purpose of this paper is to model rural-to-urban migration in China with endogenous policy change and then simulate the economic performance with policy discussion. Design/methodology/approach – Three important problems of urbanization in China are analyzed in a unified framework: within-city inequality, under-urbanization, and urban economic growth. Our model emphasizes the costs of potential conflicts caused by the de facto income disparity between the urban natives and the rural-to-urban migrants, which is the result of providing urban public services to natives exclusively. Findings –We show that, in the early stage of urban economic growth when the losses of potential conflicts are relatively small, the exclusive provision of urban public services may be beneficial to native urban residents. However, the losses under such unequal public service provision policy increase in the process of urban growth, and beyond a certain stage of development, opening public service access equally to the migrants will be a better choice, even if only the natives’ utility is considered. Such an endogenous policy change not only decreases with-in city inequality and conflict, but also advances urbanization and urban economic growth. Originality/value – In this paper, we regard development not only as a process of urbanization and industrialization, but also that facilitating the integration of different social groups, including native urban residents and migrants. It helps to better understand some key aspects of Chinese economy such as under-urbanization and increasing urban-rural inequality during urbanization by examining the rural-to-urban migration with endogenous policy change. Our model also helps us better understand the currently heated debate on Lewis turning point in China. The endogenous policy change in the model implies that China has not run out of surplus labor since rural migrants are still somewhat part of the underprivileged class in urban society in China, which some empirical studies have also pointed out.
