Do border effects alter regional development: evidence from a quasi-natural experiment in China
利用1997年四川分设为四川和重庆的准自然实验，通过双重差分法估计新省界对附近县域人口的影响，发现边界效应显著抑制了人口增长。
Abstract Market access/potential is main explanations for spatial variation in economic activity. Past research has used quasi-natural experiments such as the imposition and removal of the Iron Curtain to assess how changes in market access influence economic outcomes. Rather, we focus on key quantity effects of market access by tracking population changes induced by the creation of a subnational border. We exploit a quasi-natural experiment in China and use a difference-in-difference identification strategy to estimate the effects of introducing a new border when Sichuan province was split into Chongqing and Sichuan in 1997. We find that the new border had negative population effects on Sichuan counties located near the new border. Further investigation finds that such border effects are unique to the new border region and are not related to other factors such as being more rural. We also provide additional evidence to exclude alternative explanations including differences in industry composition or access to transportation.