Sunday, January 4, 2015

Importer-Specific Elasticities of Demand: Evidence from U.S. Exports



Importer-Specific Elasticities of Demand: Evidence from U.S. Exports


One sentence summary: Elasticity of demand for U.S. exports is higher for lower income and nearby importers.

The corresponding paper by Hakan Yilmazkuday has been published at International Review of Economics and Finance.



Abstract
This paper investigates whether the elasticity of demand systematically changes from one importer country to another in an international trade context. Evidence from U.S. exports supports this view by suggesting that the elasticity of demand in an importer country among the products purchased from the U.S. significantly decreases in GDP per capita and distance to the U.S. of the importer country. In terms of policy implications, using a common elasticity measure would overestimate the gains from reducing trade costs with developed or geographically distant countries and underestimate them with developing or geographically close countries.


Non-technical Summary
In the context of the static applied general equilibrium trade literature, the elasticity of demand is a key parameter that is used by policy-makers to derive quantitative results, because the effects of an international trade policy change are evaluated by the conversion of policy changes into price effects. These price effects (i.e., price changes) are the key in determining the effects of trade policies on the real macroeconomic variables such as output, employment, trade flows, and economic welfare, as well as other important variables of interest. Therefore, there is no question that the measurement of the elasticity of demand is of fundamental importance in determining the response of trade models to policy experiments.

This paper investigates whether the elasticity of demand, which corresponds to the (price) elasticity of demand in the context of CES aggregators under the assumption of a large number of varieties, systematically changes from one importer country to another in the context of international trade. In terms of modeling, a partial equilibrium trade model is introduced where each country has a distinct import demand for different countries' goods (represented by a sub-utility). For instance, the United Kingdom (U.K.) has a certain demand (and a corresponding elasticity of demand) among the goods imported from the United States (U.S.), while Germany has a different demand (and a corresponding elasticity of demand) among the very same U.S. goods. The sub-demand of each importer country is represented by a constant elasticity of substitution (CES) aggregator that is a combination of goods imported from the U.S.. Although the elasticity of substitution is constant for each importer, it is allowed to change across importers, which is the key to this paper.

Using the U.S. export data (at the SITC 4-digit good category) that cover the value and unit prices of exports from the United States to 237 destination countries around the globe between 1996-2013, this paper shows that the elasticity of demand varies significantly across importer countries. In the benchmark case that ignores zero-trade observations, the common elasticity is estimated as about 0.90, while it ranges between 0.75 and 1.32 across importers when importer specific elasticities are considered. Similarly, when zero-trade observations are also included in the analysis, the common elasticity is estimated about 0.86, while it ranges between 0.05 and 1.51 across importers when importer specific elasticities are considered.

The heterogeneity of importer specific elasticities corresponds to important policy implications: Individual responses of importers through importer-specific elasticities, rather than an imposed average response through a common elasticity, should be taken into account, because each importer has its own demand characteristics. Just to give two examples, among other importers, using a common elasticity would overestimate the gains from reducing trade costs (of organized exchange goods) with Finland and underestimate them with Ukraine, both by about twofold. When the reasons behind the heterogeneity of elasticities are further investigated, it is found that importer specific elasticity estimates decrease with the development level and the distance to the U.S. of the importer country. Therefore, a common elasticity measure would overestimate the gains from reducing trade costs with developed or distant countries and underestimate them with developing or nearby countries.