This paper examines the trade-offs faced by central decision-makers responsible for delivering local public goods. From 2000 to 2010, China closed more than 50% of its primary schools. While smaller and academically weaker schools were closed and students' test scores increased, students and their families paid additional costs due to a longer home-to-school commute post school closure. This paper builds and estimates a structural model of local governments' decision-making to understand these stylized facts and to recover the preferences of local authorities. I pay particular attention to cases when there may be heterogeneous productivity across schools, increasing returns in schooling production, and home-to-school commuting costs. The estimation shows that schools that were closed had lower mean productivity. Some very small schools with relatively high productivity were also closed. I use a moment-inequalities approach to estimating bounds on the weight local governments placed on the cost to families attending a more distant school. The very low upper bound of this estimate implies that distance cost was under-weighted in local governments' decision.
The Effects of Tax Reduction on Agricultural Households’ Consumption and Investment, with Binzhen Wu
From 2000 to 2006, China gradually reduced and finally eliminated its agricultural output tax in rural areas. This tax only depended on the amount of land, and the amount of land was allocated according to the number of family members by rural collectives. Agricultural staple prices were set by the national government during this period. Therefore, this tax exemption directly increased households’ permanent incomes. Exploiting this policy change, we use panel data and a difference-in-difference identification strategy to investigate how the increase in after-tax disposable income affected rural households’ consumption and investment decisions. The results show that the effect of tax reduction on consumption was inversely U-shaped. Moreover, the tax reduction had a positive effect on local non-agricultural production and migration and had no significant negative effect on grain output. These facts can be explained by a model of rural households with liquidity constraints. If non-agricultural activities had higher returns than agricultural activities but required some start-up investment, the tax-reduction relaxed the liquidity constraints. Households could start non-agricultural businesses, even sacrificing consumption in the short run. In conclusion, the short run and the long run impacts of the agricultural tax exemption depended on heterogeneity in liquidity constraints facing farmers.