2023-12-01-IMF-全球紧张局势对贸易转移的影响——比我们想象的要高_62页_2mb

Trade Diversion Effects from Global Tensions—Higher Than We Think

Authors: Mengqi Wang and Swarnali Ahmed Hannan
Institution: International Monetary Fund Working Paper WP/23/234
Date: September 2023

Abstract

New dataset combining Mexico's input-output data with cross-country sources (WIOD, UN Comtrade) reveals higher trade diversion effects from global tensions (U.S.-China trade tensions and U.S. sanctions on Russia) compared to cross-country studies.
Uses difference-in-differences, local projections, and empirical methods to show output and downstream tariffs drive trade diversion, while substitution elasticity and GVC integration vary effects.

Key Methodology

1. Data:

   • Industry-level input-output dataset (Mexico) matched with UN Comtrade and WIOD.
   • Insight: National data captures supply linkages more accurately than cross-country sources (WIOD only covers 56 vs. INEGI’s 258 industries).

2. Trade Diversion Measurement:

   • Output tariff, upstream/downstream tariffs (using input-output linkages) to measure exposure.
   • Empirical methods: Difference-in-differences and local projections to isolate trade diversion from other shocks.

Findings

U.S.-China Trade Tensions (2018)

• Aggregate Effect:

  • Net tariff increase by the U.S. on China (+5.28 pp) led to a 6.4% boost in Mexico’s exports to the U.S.
  • Output tariffs had the strongest positive impact; downstream tariffs also contributed.

• Industry-Level Variation:

  • Trade diversion depended on:
    • U.S. tariff changes on Chinese goods (larger increases boosted exports).
    • Decline in U.S. imports from China.
    • Product substitutability (more substitutable goods saw larger gains).
    • GVC integration (weaker correlation).

• Comparison with Cross-Country Data:

  • INEGI-based estimates (6.4% increase) vs. WIOD (1.4%)—national data shows magnitudes are higher due to granularity.

U.S. Sanctions on Russia (2014)

• Positive Trade Diversion:
  • Proxy analysis using Russia as a controlled variable.
  • Mexico’s exports to the U.S. increased ~10% in real terms four months post-sanctions.

Determinants of Trade Diversion

• Import Decline from China: Negative correlation with trade diversion (U.S. diversion fills China’s export gap).
• Tariff Exposure: Output and downstream tariffs had statistically significant positive effects.
• Substitutability: Easily substitutable goods drove larger trade diversion.
• GVC Integration: Weaker correlation, but higher integration indirectly beneficial (via input availability).

Conclusion

• Trade diversion effects on Mexico are higher than estimated by cross-country studies, primarily due to supply linkages and substitutability between countries.
• Policy Implications:
  • Cannot overlook short-term gains from trade diversion.
  • Geoeconomic fragmentation risks inefficiencies and reduced global welfare.
  • Sanctions and tariffs fragment trade flows but should be assessed carefully due to negative spillovers.