Elasticity of substitution in the U.S. market with endogenous transport costs: a sectoral approach.

• Author: Molina, Danielken
• Position: Report
• Pages: 95(18)

1. Introduction

   Within the last ten years, the world has probably experienced the most important change on bilateral trade relations. In 2002, the second most important market for trade was created; the European Union began the circulation of a common currency within all the members. As we know, the creation of the Eurozone entitled new economic rules common to all country members, which applied to several topics, being trade one of them. At the same time, the world entered into a phase of globalization that implied, among many other changes, the general reduction of barriers of trade, paying special attention to tariff barriers of trade. As a whole, WTO members have engaged in a general reduction of trade barriers.

   As an example of this trend, by 1999-2001, Mexico and Canada were already trading with the U.S. under the free trade agreements of NAFTA and CUFTA. Latin America has not been absent to these processes. Brazil, Argentina, Bolivia, Uruguay and Chile are all engaged in the region's most important trade agreement: MERCOSUR, and at the moment of writing this paper, Chile was negotiating bilateral trade agreements with Colombia and other Andean countries.

   It could be said that the most interesting negotiations for Latin American countries have been taking place within the last three years; as the U.S. negotiated bilateral free trade agreements with several countries of the region, (1) and as expected, the main objective of these treaties has been to reduce bilateral tariffs and bilateral non-tariffs barriers of trade.

   It may be taken as some sort of redundancy, but it must be remarked that the effects of trade costs are a very important determinant of trade patterns between countries. Future reductions of trade costs will have important effects on the type and amount of goods exported from foreign countries, and in the present context this would imply an important source of trade diversion or trade creation for those countries that could potentially decrease the trade costs of exporting products to foreign locations. As an example of this pattern, several Latin American countries have made important airport modernization decisions in order to increase the efficiency of their cargo handling. A perfect example of this strategy is Colombia, which in 2006 approved and started upgrading of El Dorado, the most important cargo airport of South America. Panama also, in 2004, initiated an investment plan to upgrade the inter-ocean Channel to satisfy the current cargo requirements of maritime ships.

   Taking all these into account, the present paper focuses on the estimation of the elasticity of substitution of imported products to the U.S. market for the period comprehended between 1990 and 2003. Taking advantage of the quality of the data available for imports to the U.S., we use information available for 4512 different products imported to the U.S. from all over the world, at the six digit level of disaggregation of the Harmonized System Code. Our empirical exercise builds up on the theoretical model developed by Anderson and Wincoop (op. cit) and we estimate the average elasticity of substitution for our sample of products imported by the U.S. market. Given the features of the data, we also were able to classify each product within eight economic sectors. (2) We use this structure to estimate elasticities of substitution by sector. Finally and as an additional extension, we endogenize transport costs, and compare these estimates against the standard estimates that were obtained under the assumption on exogenous transport costs. As previous trade literature has already explained, trade costs are determined by all the costs contained on the process of delivering a final good to a consumer. In particular, the cost structure is determined by a whole set of factors that include transportation costs, policy barriers of trade, information costs, contract enforcement, currency costs, other legal and regulatory costs and local market distribution costs. Rough estimates of trade costs for industrialized countries have found that their importance for trade is equivalent to an ad-valorem tax rate of around 170%, which can be decomposed in the following aggregate components: "21 percent is explained by transportation costs, 44 percent is related to border trade barriers and the other 55 percent is related to retail and wholesale distributional costs". (3) So trade costs do matter!

   The structure of this paper goes as follows: Section 2 describes the model and the endogenous trade costs specification to be used to determine the estimation equations of transport costs and imports demand, which will be used to estimate the elasticities of substitution. Section 3 describes the data. Section 4 reports the results and finally Section 5 presents some conclusive remarks and directions for future research.

2. The Model

   Following Anderson and Wincoop (op. cit), we develop a general equilibrium model that is used to estimate the elasticity of substitution of the products imported by the U.S., within the period comprehended between 1990-2003. Our model builds on two facts. As the authors explain, the impact of trade costs over trade is important, and is equivalent to the effect of an ad-valorem tax of 170%. Therefore, it would be interesting to take a step forward and try to endogenize it. The gain is given by the conceptual idea that imports' levels and transport costs are both endogenous since obviously imports are affected by transporting costs; i.e. insurances.

   For estimation purposes, the direct consequence of this problem is that the estimates of the elasticity of substitution among goods are biased. As expected, we could potentially reduce this problem by modeling the transport cost structure of goods. As we know, modeling transport costs is not a new concept, Hummels (1999) and Micco (2006), among others, have already modeled transport costs. But the implications of the estimates of the elasticity of substitution might give a new conception of the effects of transport costs on trade, not only for their direct effect on imports but also for the effect of the assumption of the level of substitution between imported-competing goods.

   Second, and from a policy perspective view, low transport costs for exporters which are near to the U.S. market would give some advantage to nearby producers in comparison to producers located in remote locations that are directly associated by distance. Thereby, countries that are located far away form the American market could potentially reduce the bias implied by transport costs by implementing policies aimed to decrease their transport methods; i.e. increasing infrastructure level or implementing new and more efficient methods of transportation. As a general policy example, if Chinese exporters could reduce their transporting costs to the U.S. market, this would potentially reduce the level of exports of other competing countries to the American market.

   In the following section we proceed to derive the model. First we develop Anderson and Wincoop's (op. cit) general equilibrium model, and then we endogenize the transport cost structure. This would...