A Critique of Exchange Rate Treatment in Agricultural Trade Models: Comment

Abstract

In the May 1979 issue of the Journal, Chambers and Just (CJ) critically reviewed theoretical and empirical work on the role of exchange rates in agricultural prices and trade. CJ contend that the assumption of zero cross-price elasticities between commodities leads to an inaccurate portrayal of import demand and export supply functions and therefore an inaccurate portrayal of the effects of exchange rate changes on trade.1 This author agrees with that contention. Incorporation of possible cross-price effects will clarify the effect of all variables in the system. However, this author feels that the remedy CJ have proposed for incorporating these crossprice effects in order to analyze the effect of exchange rate fluctuations is less than desirable. The proposition will be put forth that the inclusion of prices for commodities which are close substitutes (complements) to the good studied and an index of all domestic prices in the country studied is more sound in and practice than the cure proposed by CJ. CJ argue that separate exchange rate variable (involving a weighted price index of other traded goods if feasible) (p. 256), should be included in the equation for an excess demand function. This separate exchange rate variable would serve as a price index for all other traded goods. The structure under which this suggestion is derived is unclear. The exchange rate, in and of itself, is only relevant to excess demand as a deflator. It is a way of transforming prices into a currency which is relevant to producers and purchasers of goods in the country. Japanese producers and purchasers are interested in the yen price, not the dollar price. Therefore the idea of including the exchange rate as a separate regressor for an excess demand function is structurally unsound. If the exchange rate is multiplied by a weighted price index of other traded goods, other problems develop. The first problem is that the power parity theory may not hold. Exchange rates may not reflect the purchasing power of the currencies involved. Therefore the exchange rate multiplied by a weighted price index of other traded goods does not accurately reflect the domestic currency price of goods in the country. Second, there is the problem of price transmission for traded goods in a world where trade barriers are common. A change in the export price of a commodity may not be completely reflected in the internal price of the commodity in an importing country. Finally, the exchange rate change also can have indirect effects on all goods in the economy. The exchange rate directly affects traded goods, but because of cross-price effects the price of nontraded goods also could change. These price movements for nontraded goods could then influence the excess demand function for the country. These cross-price effects between traded and nontraded goods are important, but certainly outside the scope of an empirical agricultural trade model.