for helpful comments on an earlier draft. 1 The availability of trade data was surely a contributing factor to the advanced nature of the early empirical work on trade models. It is sufficient to note that by 1957 there already existed, inter alia: (i) at least 42 books and articles containing estimates of income and price elasticities for imports and exports [see Cheng's (1959) survey]; (ii) several superb methodological criticisms [Orcutt (1950), Harberger (1953)] that explained why estimated price elasticities could differ from the true elasticities; and (iii) at least one complete simultaneous model of import demand and supply [Morgan and Corlett (1951)], actually estimated by limited-information maximum likelihood methods, that displays the authors' awareness of many of the methodological issues that still entertain current research.