Product-Level Trade Diversion or Broader Export Expansion? DID Evidence from Mexico and Vietnam under the U.S.–China Tariff Conflict

This dataset supports a study of whether the U.S.–China tariff conflict generated product-level trade diversion toward Mexico and Vietnam. The central hypothesis is that, if U.S. Section 301 tariffs on Chinese goods redirected demand toward third countries, products covered by the tariff lists should have experienced stronger post-policy growth in U.S. imports from Mexico or Vietnam than products not covered by the lists.
The dataset contains quarterly U.S. import data at the HTS6 product level for Mexico and Vietnam from 2015Q1 to 2024Q4. Import values were obtained from the USITC DataWeb. Information on Section 301 tariff coverage and implementation timing was collected from the Office of the United States Trade Representative. Each HTS6 product was classified as either a treatment product or a control product. Treatment products correspond to products covered by the U.S. Section 301 tariff lists on China, while non-listed products form the control group. Product-specific post-treatment indicators were constructed using the effective quarter of the relevant tariff list.
The data include product identifiers, year and quarter, import values, treatment status, post-treatment status, and the interaction term used in the difference-in-differences analysis. Additional variables capture pre-policy product characteristics, including average product size, growth trend, volatility, market share, within-sector share, and observation frequency. These variables are used to control for differences in pre-treatment product characteristics.
The results show clear differences between Mexico and Vietnam. For Mexico, tariff-listed products grew significantly faster than non-listed products after the policy shock, providing evidence consistent with product-level trade diversion. For Vietnam, the estimated policy effect is positive but statistically insignificant. A pooled-sample interaction test further shows that Vietnam’s estimated effect is significantly smaller than Mexico’s.
The descriptive decomposition also shows that Vietnam’s export expansion was driven mainly by non-listed products. Non-listed products accounted for 86.5% of the total increase in average quarterly U.S. imports from Vietnam, while listed products contributed 13.5%. The share of listed products declined from 19.2% before the policy period to 16.1% afterward. These patterns are more consistent with broader export expansion than with trade diversion concentrated in Section 301-listed products.
Users should interpret the DID estimates as relative changes between listed and non-listed products within each country. The results do not directly identify firm relocation, foreign investment, or production-transfer mechanisms. The dataset can be used to replicate the main DID estimates, conduct robustness checks, examine sectoral heterogeneity, and study changes in export structure.