Working Paper: CEPR ID: DP13602
Authors: Peter Egger; Katharina Erhardt
Abstract: Structural quantitative work in international economics typically treats trade policy as log-linearly related to trade costs and as exogenous. This paper proposes a structural approach that allows for a non-parametric relationship and treats tariff and non-tariff trade-policy variables as potentially endogenous in log-linear estimation. We document that the data reject the assumption of log-linearity of trade costs in both the tariff- and the non-tariff-policy domains. Specifically, the partial impact of a change in tariffs is strongest for low policy barriers and medium levels of tariffs but generally decreases in the level of both non-tariff barriers and tariff barriers. To give a relevant illustration, we assess the effects of a unilateral increase of US tariffs on Chinese imports by 10 percentage points and document that the estimated effects on real bilateral trade-flow changes would be largely underestimated by standard approaches.
Keywords: trade policy; gravity models; semiparametric methods; nonparametric methods; generalized propensity scores
JEL Codes: C14; F14; F13
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Increase in tariffs (F69) | Increase in trade costs (F19) |
| Increase in non-tariff barriers (F13) | Increase in trade costs (F19) |
| Increase in non-tariff barriers (medium levels) (F13) | Decrease in trade costs (F19) |
| Increase in tariffs (low policy barriers) (F13) | Stronger marginal effect on trade costs (F12) |
| High non-tariff barriers (F13) | Weaker marginal effects of tariff increases (F69) |