Automation, Trade Policy Uncertainty, and Export Composition

We examine whether pre-existing automation shapes the response of export composition to trade policy uncertainty. We are developing a heterogeneous firm model in which technology can reduce baseline operating costs and attenuate the transmission of adverse shocks into costs, thereby moderating firms' export quantity and quality responses under specified conditions. Using data for five manufacturing economies over 2012-2023, we examine the interaction between lagged industrial-robot exposure and a common U.S. trade policy uncertainty index. The baseline results indicate that economies with greater robot exposure exhibit a more favorable response of the high-technology share of manufactured exports. Export value estimates suggest that this pattern reflects the relative resilience of high-technology exports, with more negative differential responses among other manufactured exports. These estimates do not establish an absolute expansion in high-technology exports or within-product quality upgrading. Although the baseline interaction remains statistically significant under alternative small-sample inference procedures, it becomes statistically insignificant when country-specific trends are included or China is excluded. The findings point to a conditional association between embodied automation and export composition under trade policy uncertainty.