On a Forgotten Challenge in the International Trade Literature

This paper scrutinizes a necessary and sufficient criterion to examine whether currency devaluation improves the trade balance in the models with the exports-to-imports ratio as a dependent variable. We attempt to rectify a common error in the existing literature, which assumes that a positive exchange rate elasticity of trade balance is evidence for the efficacy of currency devaluation. In addition, we re-investigate the Marshall-Lerner condition and demonstrate that the conventional criterion is conditional. However, the McPheters and Stronge (1978) criterion, which claims that the exchange rate elasticity of trade balance needs to be greater than one (M-S criterion), is unconditional. The empirical results derived using the pooled mean group estimation support the idea that currency devaluation, in the long run, improves the trade balance in Asian countries but not in Latin American and OECD countries