Revisiting the role of trade and automation in US labor market polarization

The increase in the share of high- and low-wage employment at the expense of middle-wage employment has been a striking feature of the US economy. We exploit differences across local labor markets in the exposure to Global Value Chains (GVCs), Chinese import competition and automation to study the drivers of this labor market polarization. Using value added trade data, we are able to correctly assign trade-related shocks to local labor markets, based on the source of value added. Across the 722 commuting zones that approximate US local labor markets, we find that employment polarization is mainly driven by their exposure to automation. GVCs lead to an increase in the employment share of relatively high-wage occupations (which we call ‘skill upgrading’), while import competition from China leads to an increase in the employment share of relatively low-wage occupations (which we call ‘skill downgrading’). Trade as a combination of the two thus may contribute to employment polarization but to a lesser extent than automation.