Technological and Price Uncertainty in a Ricardian Model of International Trade

Abstract

Despite the fact that international trade is one of the riskiest economic activities, there has been remarkably little attempt to incorporate uncertainty into the formal theory. One of the few published papers dealing with this issue is the important contribution by Brainard and Cooper [3]. However, their formal analysis is rather restrictive in that they deal only with the quadratic utility and focus primarily on the case where the country is required to make all its decisions-both production and consumption-before the uncertainties are resolved. The present paper introduces uncertainty into a Ricardian model of international trade. While this technology is somewhat more restrictive than the neoclassical production possibility frontier considered by Brainard and Cooper, it nevertheless is an interesting case to start with in view of the fundamental role the Ricardian case has played in the development of international trade theory. Furthermore, it is of considerable interest to determine to what extent the extreme implications regarding the patterns of specialization associated with the classical Ricardian model remain after the introduction of uncertainty. In other respects the present analysis is more general than the Brainard-Cooper model. In the first place we try as far as possible to derive propositions based on general utility functions, although where appropriate, we use special cases to illustrate particular propositions. In establishing these propositions a crucial role is played by the indirect utility function which enables us to analyse the behaviour most conveniently in terms of the familiar Arrow-Pratt measures of risk aversion. The model is developed for a small country, in which domestic technology and the international trading price (which it takes as given) are both subject to uncertainty. We develop a kind of two-stage analysis in which production decisions are made before the random technological parameters and the actual trading price are known, while trading and thus consumption decisions are made after the uncertainties are resolved. In view of the lags associated with production and the fact that most international trade takes place at previously established prices, this seems to be the most plausible case to consider. Brainard and Cooper did not consider it on the grounds of analytical intractability, but in fact the analysis is no more difficult than the case they examine where all decisions are made under uncertainty.