题目：Spatial Correlation in Migration: Evidence from China
Abstract: The previous research does not focus enough on the theory and empirical evidence of spatial relevance in migration. This paper explores New economic geography(NEG) model and applies Taylor series expansion to provide theoretical explanation for the source of migration. The gravity equation, NEG linear model and spatial model are used with the data of 1985-1990, 1995-2000 and 2005-2010 Chinese Census to test the existence and impact factor of spatial Correlation. The results suggests that there exists the ‘third party’ effect in migration, including the cluster effect by the scale and the competitive effect by the regional wage. Meanwhile, the spatial correlation of disturbance tells that the economic shock transfers through geographical proximity, and the spatial correlation caused by the transfer also affects the migration. The result is robust when changing the spatial weighted matrix. Furthermore, the forward linkage mechanism of NEG theory has more effect on the migration in middle and western regions than in eastern regions. We also find that people in the middle and western regions are more sensitive for the economic factor when making the decision of migration and they burden the higher cost of migration.
Key Words: Migration; Spatial ; Correlation ;Gravity Equation; NEG Linear Model; Spatial Model