Tradeoff between air pollution and economic benefits in migration dynamics: Evidence from China

This paper investigates how environmental amenity and economic benefits are balanced in migration dynamics from the individual perspective. We construct a structural model based on the dataset that combines PM2.5 concentration data from NASA and migration information from the China Migration Dynamic Survey (CMDS). Thermal inversion is used as a source of exogenous shock in air pollution to address the omitted variable bias. The results indicate that if the ambient concentration of PM2.5 at destinations increases 1 μg/m3, the probability of migration at the individual level will decrease by 0.90 %. A battery of robustness checks is implemented, including different approaches to measuring migration decision, using imputed wage instead of disposable income, using different IV and proxy variable. Heterogeneity analysis reveals that the deterrence effect of air pollution is more pronounced for the following groups: migrants who are male, move from a rural residence, or to a coastal destination, belong to working-age population, engage in a tertiary industry, have a postgraduate-level education, and support two or more elders.