From macro to micro and macro back: Macroeconomic trade elasticities in a developing economy

A long tradition in economic theory has seen the real exchange rate (RER) as a key determinant of trade performance. Several empirical studies, however, have found low estimates of macro trade elasticities and, as a result, questioned this argument. In this paper, we show that to understand and estimate the effect of RER on trade performance, it is crucial to explore the existence of heterogeneous responses of individual products to RER movements. Using trade data from Argentina, we employ the Mean Group method to estimate macro trade elasticities by individual products disaggregated at four digits of the SITC; rev.2. We find a wide range of heterogeneous responses of exports and imports to RER movements at the individual product level. We find that the estimated RER-elasticities in differentiated products and labor-intensive manufacturing goods are substantially larger than those of primary and homogeneous products. Based on the estimated elasticities at the product level, we obtain a "low"aggregate RER-elasticity of exports when we weigh them by the country's trade basket. We show that this result is a consequence of Argentina's economic structure, a country whose exports are heavily specialized in primary and homogeneous products. Our results are important because they help build a bridge between two conflicting views in the RER-economic development literature. Even in countries with "low"aggregate macro RER-elasticities, the real exchange rate may be an important variable for economic performance. An undervalued RER may facilitate economic growth by stimulating investment in activities producing more complex/differentiated goods in countries that specialize in the production of primary and homogeneous goods and, as a result, have "low"trade elasticities. This seems to be the case of several countries in Latin America, including Argentina.