Dynamic trade elasticities and comparative advantages: Evidence from a PTA

Quantitative trade models predict that trade elasticities vary with comparative advantages, but empirical evidence remains scarce. We provide causal estimates of this relationship by exploiting a quasi-experimental bilateral trade agreement between Colombia and the United States. Using a local projections difference-in-differences approach, extended with state-dependent local projections to account for heterogeneity arising from comparative advantages, we estimate trade elasticities at different horizons. We find that the average short-run trade elasticity is around 1.5, and converges to approximately 4 after seven years. These elasticities increase with comparative advantages: a move from the 25th to 75th percentile in comparative advantages is associated with a 2-percentage-point increase in the long-run elasticity. To measure comparative advantages, we construct a new Revealed Comparative Advantage (RCA) index -- the system bilateral RCA index -- that exploits trade flows across multiple trade partners while preserving the ability to compare Colombia with a specific trade partner regarding their respective comparative advantages. Our findings suggest that comparative advantage shapes the responsiveness of trade flows to policy changes, which has implications for trade theory and quantitative trade models.