Does College Location Affect the Location Choice of New College Graduates in China?

Based on a representative survey of new college graduates in China, we present a job location choice model that allows for a self-selection effect in college regions by considering region-dependent heterogeneous preference for job locations and unobservable location features. We also treat endogenous housing costs using the method developed by Berry et al. (1995). Our estimation results show that college education significantly increases graduates' likelihood of working in the city where their college is located. However, a model that does not consider the self-selection effect overestimates this impact. The impact of college location on a graduate's migration decision varies considerably across cities. There is significant heterogeneity between students from universities of different tiers and rural versus urban areas. Meanwhile, the graduates show a marked home preference when choosing their job locations, and they tend to avoid working in cities with high housing costs. These findings shed light on debates on place-based higher education policies and spatial distribution of human capital with a high level of education.

Automated summary

Based on a representative survey of new college graduates in China, this study presents a job location choice model that considers the self-selection effect and region-dependent heterogeneous preference for job locations. The results show that college education significantly increases graduates' likelihood of working in the city where their college is located. Furthermore, the impact of college location on a graduate's migration decision varies considerably across cities, and there is significant heterogeneity between students from universities of different tiers and rural versus urban areas.