The Effects of Aggregation on Estimated Import Price Elasticities: The Role of Imported Intermediate Inputs

Abstract

This paper examines how one response of domestic producers to an exogenous change in relative prices may reduce estimates of import price elasticities made for aggregate classifications of imports. This response occurs as follows: A reduction in the relative price of an imported good would decrease demand for the domestically produced substitute. The consequent reduction in production of this substitute would cause reduced imports of intermediate inputs used in its production, even though the relative prices of these inputs remain unchanged. Similarly, an increase in the relative price of an imported good would cause increased imports of intermediate inputs used in production of the domestic substitute. We term these effects on imports of intermediate inputs effects. To the extent that imported goods and intermediate inputs used in production of their domestic substitutes are in the same aggregate classification, output effects will reduce the net changes in imports that accompany relative price changes for this aggregate, and hence will reduce the estimated import price elasticity. If some goods in the aggregate classification are exported, changes in exports due to relative price changes for the aggregate may also cause output effects that would act to reduce the estimated import price elasticity. Clearly, the importance of output effects for estimated elasticities depends upon the level of aggregation used in the estimation and on the input-output structure of the importing country. As a case study, we estimate the magnitude of output effects for import price elasticities estimated for Canada, using the aggregate classifications shown in table 1. We show that output effects provide an important additional reason beside those given by Orcutt (1954) as to why aggregation may reduce estimated import price elasticities.' Our results thus illustrate the magnitude of the bias caused by output effects when elasticities estimated for aggregate classifications of goods are used to predict the effects of a change in relative prices on imports of more disaggregate classifications. Our results also indicate the degree to which a general equilibrium model may understate the effects on trade of exogenous changes in relative prices (such as tariff cuts or depreciation) if estimated import price elasticities are used and the model explicitly accounts for output effects.2 The understatement occurs because estimated elasticities are likely to account for some output effects, and these output effects would be double-counted. Although our analysis is presented in terms of partial elasticities, it also applies to what are called total or semi-total elasticities.3 The next section presents the model used to estimate the influence of output effects on estimated import price elasticities. In section III, the model is estimated. Concluding remarks are presented in section IV.