ABSTRACT

CGE models have now been used to examine the impact of a variety of price reforms (including trade, marketing and shifts in agricultural technology) in a large number of different countries.4 However, the data requirements for such models can be considerable. Unless a recent Social Accounting Matrix (SAM) is already available, the construction of a useable SAM can take substantial time and expertise. Furthermore, the econometric estimation of behavioural parameters (as opposed to merely calibrating such parameters from the original SAM) can be complex and time-consuming and the choice of functional form for the behavioural relationships, although based upon plausible arguments, is essentially arbitrary (Deaton, 1997).