A Note on the Treatment of Foreign Exchange in Project Evaluation

TLDR: In this article, the authors compare the analytical underpinnings of the two methods of treating foreign exchange using a very simple general equilibrium model of the economy and infer what implicit assumptions are being made by each.

Abstract: There is apparently still some unease concerning the analytical treatment of foreign exchange in the OECD or Little-Mirrlees (1968, 1974) manual (eg the recent review of Little and Mirrlees by Bottomley, 1975) What, for example, distinguishes the Little-Mirrlees shadow pricing rules for traded commodities from that of Dasgupta, Sen, and Marglin (1972) in the Unido manual, if anything? In the Unido approach, domestic (non-traded) resources are evaluated by the domestic "willingness to pay", and foreign exchange used or earned via traded goods is evaluated using a shadow price of foreign exchange Little-Mirrlees, on the other hand, dispense with a shadow price of foreign exchange by evaluating traded goods at world prices and non-traded goods at their "foreign exchange equivalent" It has been suggested that when Unido uses a uniform shadow exchange rate for all tradeables and Little-Mirrlees use a single conversion rate for nontradeables they are equivalent, differing only in the choice of numeraire, the numeraire in Little-Mirrlees being foreign exchange and in Unido domestic consumption (eg Dasgupta, 1972; Scott, 1974) The purpose of this note is to clarify the analytical underpinnings of the two methods of treating foreign exchange using a very simple general equilibrium model of the economy In this way, we shall be able to establish the essential similarities and differences between the two approaches and to infer what implicit assumptions are being made by each Both were, of course, written primarily for practitioners of project evaluation, and their comparative usefulness depends to a certain extent upon the practical ease with which they may be applied Our interest is of a purely theoretical nature It is simply to clarify the analytical bases for the two sets of rules Our model is a static general equilibrium neoclassical model with an exportable, an importable, a non-traded good and two factors (labour and capital) The only distortions are taken to be tariffs on the traded goods These could equally well be interpreted as quotas or any other distortion that makes the domestic and world prices different Consumption and production taxes could easily be added to the model at the expense of simplicity Since they add little to the result they are omitted In order to concentrate solely on the foreign exchange issue we have ignored several other problems raised in both manuals and elsewhere-the qualitative distinction between traded and non-traded goods, distortions on the capital market (shortage of savings), the shadow price of labour in a dual economy, and income distribution issues No problems of a dynamic nature are considered