Working Papers
Until 2025, U.S. imports valued below $800 entered free of tariffs and outside official trade statistics. This paper studies how this de minimis exemption shaped the response to the Section 301 tariffs on Chinese goods. Using customs records of low-value shipments obtained through a Freedom of Information Act request and monthly product-level trade data from 2016 to 2024, it compares products that could ship as parcels with products facing the same tariffs that could not. A 25 percentage point tariff increase lowers recorded imports by 50 percent for de minimis products and by 39 percent for other products, a gap that tariff exclusions do not explain. The missing trade does not reappear as recorded imports from third countries. In a multi-sector general equilibrium trade model with input-output linkages, parcels carried 9 percent of China's sales to the United States and cost \$1.6 billion a year in tariff revenue. Together with duty relief on recorded imports, they lowered the effective U.S. tariff on Chinese goods from 16.2 to 12.7 percent.
Numerous researchers have documented wage losses and local labor markets hit by rising import competition. In a wide variety of settings, labor incomes in locations dependent on import-competing industries fall relative to incomes in other locations. This paper attempts to see if there are differences in these effects between demographic groups in the US case in the context of the most studied example, commonly referred to as the ‘China Shock.’ We examine the ‘matched CPS,’ which allows us to observe year-to-year economic transitions of a sample of US workers, to see if the losses in labor income fall more heavily on identifiable demographic groups. We find that income losses are more likely in the face of a trade shock for workers in manufacturing; workers with a bachelor’s degree, and workers with high household income. By contrast, losses are smaller or zero for workers with low family income and for Black workers. We argue that these last findings may be due to liquidity constraints, causing workers to scramble to keep their incomes from falling below the level required to meet essential expenses. These effects seem to be new to the trade literature.
This paper quantifies the unintended consequences of trade policy uncertainty triggered by President Trump’s 2017 renegotiation of NAFTA on Mexico’s foreign direct investment (FDI) inflows. Using panel data on 92 industries across 32 Mexican states from 2009 to 2019, the analysis exploits variation in industry- and state-level exposure to the renegotiation through difference-in-differences, triple-difference, and Poisson Pseudo Maximum Likelihood estimation strategies. The results show that the threat of reverting to Most-Favored Nation (MFN) tariffs led to a significant decline in FDI: a 1% increase in tariff exposure is associated with a $2.5 million reduction in FDI for industries that are the sole exporters of their state. While the U.S.-China trade war increased FDI inflows to Mexico by 24%, the uncertainty surrounding NAFTA reduced them by 3%. These findings highlight how trade agreement renegotiations can discourage investment, suggesting that future agreements include safeguards—such as gradual tariff adjustments or clauses discouraging sudden termination—to mitigate uncertainty-induced shocks.