Statistical Decision Theory in a Macro Simulation Model: Feed Graiin Sector

Abstract

A method for taking uncertainty into account when formulating aggregate agricultural policies is applied to the feed grain program. The impact of alternative feed grain pro­ grams on net farm income, Government payments, and feed grain production in the Southeastern Coastal Plains is shown. A model is developed to explain planted acreages of the major competing crops. The effects of alternative feed grain programs are evaluated using Mqnte Carlo simulation to account for random variation. Confidence intervals are placed on estimates of income and production resulting from selected feed grain programs. Policymakers have at their disposal a wide (. array of policy instruments capable of affecting U.S. agricultural production and farm income. Considerable progress has been made in con­ structing aggregative models which can be used to fOrecast production changes resulting from use of these policy instruments. Results of these models have been severely limited by the nature of their forecasts. For a given combination of expected prices and Government programs, these models provide a single estimate of ex­ pected production response. Although this esti­ mate is an important ingredient in rlecision­ making, other valuable information is ignored. More specifically, risk and uncertainty have not been incorporated into these models. As a result, no estimate is made of the distributions of production or farm income. Such distribu­ tions would show the probability of obtaining a specified level of production or income. Infor­ mation on such probability distributions could aid policymakers in choosing among alternative policies. E'or il'lstance, programs with similar expected levels of net farm income may have different probabilities of producing an unac­ ceptably low level of net farm income. Objectives