Armington Elasticities: Estimates for Uruguayan Manufacturing Sectors

Abstract

The high degree of openness of the Uruguayan economy and the fact that the country is characterised by being constantly subject to shocks, strongly suggest the use of simulation tools that allow for predicting its future evolution under different scenarios. Currently, Computable General Equilibrium Models (CGEMs) applied to trade are one of the most potentially powerful available instruments for doing such analyses. Although there already exists some research performing this methodology for Uruguay, accumulation in the area is still scarce. As it is well known, operational CGEMs critically rely on the availability of many parameters, especially related to elasticities, among which Armington elasticities of substitution are crucial when the models are applied to trade. In the case of Uruguay, these parameters have been imposed in order to have coherent simulations and/or using available estimates for other economies. Hence, in order to obtain more robust results, it is of upmost interest to specify and estimate Armington models that would provide estimates of the elasticities using Uruguayan data. The work here summarized intends to start filling in this gap. We performed the task at a 4 digit level of disaggregation (ISIC, Rev.2) for 32 manufacturing sectors along 1989 to 2001, using monthly and quarterly data. We specify our models following Armington’s original proposal. In doing so, we took special care in methodological issues that might hinder the robustness of results. Finally, it is worth mentioning that, in many cases, the value obtained as the upper bound of the estimated intervals is very similar to the magnitude needed for CGEMs simulations to be sensible, as reported in the literature and, particularly, when compared with those stemming from GTAP.