Working Paper: CEPR ID: DP13122
Authors: Andreas M. Fischer; Philipp Herkenhoff; Philip Saur
Abstract: In a seminal paper, Autor et al. (2013) estimate the effect of Chineseexports on U.S. labor markets. To establish causality, they instrumentChinese exports to the United States with Chinese exports toother advanced economies, assuming that demand shocks to advancedeconomies are uncorrelated. Our paper documents robust empiricalpatterns that are inconsistent with this identifying assumption. Basedon a parsimonious structural model, we identify the part of sectoralChinese export growth that is driven by China-specific supply shocks.An identification strategy based on our approach essentially preservesthe estimates from the reduced form regression in Autor et al. (2013).However, in a general equilibrium model from Caliendo et al. (2019),our identification of the China shock implies more pronounced andmore dispersed manufacturing employment losses and welfare gains.Finally, our identification realigns the sectoral employment losses withstandard Heckscher-Ohlin theory.
Keywords: international trade; employment; instrumental variable
JEL Codes: F10
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Chinese supply shocks (F69) | growth in Chinese exports to the U.S. (F10) |
| Chinese supply shocks (F69) | employment losses in manufacturing sectors (O14) |
| Chinese supply shocks (F69) | higher welfare gains (D69) |
| Chinese supply shocks (F69) | differential impacts across commuting zones in the U.S. labor market (J69) |
| Chinese supply shocks (F69) | employment losses in labor-intensive sectors (F66) |