TERMS OF TRADE VULNERABILITY AND NATIONAL WELFARE IN A SMALL OPEN ECONOMY

Abstract

trade was presented by Brainard and Cooper (1967). They described lucidly how the presence of risk might imply an optimal re-allocation of resources so as to diversify that risk. The subsequent literature on trade and uncertainty divides into two camps: the first models trade choices as being made prior to the realisation of the random variables; while in the second, production decisions are made prior while consumption and trade decisions are made after the realisation. In their illuminating survey, Helpman and Razin (1978, Ch. 4) dub these two as Ex-Ante Trading Decision Models and Ex-Post Commodity Trading Decision Models. They show that the former is appropriate for one country if its trading partners follows the latter, and that if both are feasible, the latter will always be preferred. The model in this paper falls into the latter camp and involves a small open economy with all sectors operating in a perfectly competitive environment, and for simplicity, with only one sector facing an exogenous stochastic price process. We address three issues directly in this paper; they have been touched on tangentially elsewhere but we believe that they deserve direct treatment. The first and second issues concern the effects on expected national welfare of changes in the first two moments of the distribution of the terms of trade. The third issue examines government intervention, in the form of an ad-valorem production tax or subsidy, with an objective of achieving an optimal allocation of resources. In the absence of intervention, it is important to understand the welfare effects of an improvement in the mean terms of trade. In an undistorted world under certainty, an improvement in the terms of trade always improves welfare. As one might expect, the outcome is ambiguous when an uncertainty distortion is introduced. The reasons are somewhat complicated because the mean terms of trade positively influences the variance of national income and covaries with it. The relative degree of risk aversion between producers and consumers matters because the risk-averse producer effect supports the certainty result. In the same manner, we study how an increase in price riskiness will affect expected national welfare. At first sight, one might be tempted to guess that greater riskiness must reduce national welfare. There are two reasons why this view is mistaken; firstly, in general we do not know whether consumers dislike or like price risk; and secondly, although national income positively covaries