Class and source substitutability in the demand for imported wheat. (Volumes I and II)

There is an increasing recognition among wheat market modelers that wheat is a differentiated product. Three recent studies have analyzed the impacts of various policies under assumptions of differentiation of wheat on either a class or a supplier basis. This research extends these studies by allowing importers to differentiate on both levels. Wheats shipped from the five major wheat exporters are identified and grouped into classes based on common physical characteristics. This yields four wheat classes with two to four suppliers within each class. Three structural characteristics of the market are identified including the applicability of aggregating wheat importers into geographic regions, the applicability of assumptions of price responsiveness in demand and the impact that non-commercial sales of wheat have on market shares of classes and suppliers of wheats. This information is used to develop a theoretical model of differentiated demand for imported wheat. The model is empirically implemented by estimating two specifications of the demand functions with differing restrictions on the values of the substitution elasticities. The estimated substitution elasticities reveal that the common practice of aggregating countries into geographic regions is invalid because aggregation often creates either a false sense of competition in import markets or yields an over- or understatement of substitution potential among countries within a region. It is also found that there are three distinct behavioral patterns exhibited by importers with substitution elasticities generally differing for these groups. Bias is introduced in demand function estimation if import data are not adjusted for non-commercial sales of wheat, but data constraints are limiting in determining the magnitude of bias. In many cases, both demand specifications are rejected on the basis of non-significance of estimated functions or due to perverse quantity-price relationships. This result may reflect a lack of reliance on relative prices in class and supplier import determination. Overall, there are few countries where substitution between classes or suppliers is high