ABSTRACT
Richard Day’s presentation outlines an interesting set of issues directed at micromodeling in agriculture. How could we not concur with his recommendations that actual behavioral patterns of farmers should be studied and documented more intensely and carefully, that more efforts are needed on conceptual models of agricultural firm behavior, that the assumptions underlying these models should correspond more closely with reality, and that more emphasis should be given to understanding the behavior of the agricultural sector rather than its estimation and prediction? To be sure, these recommendations have a familiar ring. The real question is balancing the tradeoff between simplicity and accuracy in micromodel specifications for various purposes. The purpose for a particular model effort will largely dictate the tradeoff between complexity and inaccuracy in the modeling of the reality. Armed with this perspective, we are more optimistic than is Day on the profession’s ability to predict and estimate the impact of policies on the evolution of the agricultural sector and on the potential of economics to determine the shape and form of agricultural policies.
