Social Formation of Geographic Proximity Effects: Understanding Unequal Access to Higher Education in China

Recent literature has demonstrated that college proximity (such as the number of colleges within commuting distance) enhances college attendance in the United States. But not all differences in proximity affect college attendance. We argue that the causal role of proximity is crucially shaped by the educational institution and family logic of schooling the child generation. We thus focus on the questions of what proximity matters and why. The question of what requires a distinction of different types of schools and measures of proximity. The question of why looks at signs of the underlying mechanism of a true proximity effect.

We have created a new database by combining (1) school address data from official censuses of organizations in China, (2) provincial educational and labor market statistics from official published sources, and (3) individual-level data from the 2012 Chinese Labor Dynamic Survey (CLDS), which is the first wave of a nationally representative panel survey with over 16,000 respondents. CLDS provides detailed address on the place of residence for respondents when they were in secondary school. The combined database allows us to directly address the research questions. Organizational censuses provide school addresses and founding dates that allow us to estimate the national spatial distribution of secondary schools and colleges over time. Provincial data facilitate the control for confounding variables.

Preliminary results provide striking evidence for the critical role of proximity to secondary schools (the number of schools within commuting distance), viz. the gateway to college. Gateway proximity, not college proximity, enhances college attendance. This finding contrasts starkly with the focus on college proximity by the recent U.S. literature on the geography of access to higher education. To further identify the possible mechanisms underlying the gateway proximity effect, we will also conduct a detailed examination of alternative measures of proximity.