How Much Infrastructure Is Too Much? A New Approach and Evidence from China

This paper extends the Ackerberg–Caves–Frazer approach to a nonparametric aggregate production function to address both the endogeneity and the function misspecification issues in estimating the returns to infrastructure and private capital and thus the optimal allocation between them. Based on Chinese provincial data over 1995–2011, we find that in 1997 most Chinese provinces were under-invested in infrastructure, whereas in 2008 most of the western provinces were over-invested in infrastructure. Such findings suggest that the nationwide large-scale infrastructure investment enacted by the Chinese government after the 1997 and 2008 financial crises may be of different economic efficiency.