An Economic Analysis of Alternative Cropping Decisions Under Uncertainty

This project has examined after tax gross margin net present values accruing to Albertawheatf armers under three fertilizer and crop rotation systems; a fixed rotation traditional fertilizer system, a static economic fertilizer decision system within a fixed rotation, and a static economic fertilizer decision system within a dynamic flex-cropping framework. Decision rules appropriate to each system were developed for case farms in three Alberta agro-climatic regions; Medicine Hat, Lethbridge and Olds. The flex-cropping issue is expressed in a dynamic programming framework and incorporates elements not fully explored in previous studies; income taxation, variable input level decisions and stochastically determined moisture conditions and crop prices. Decisions are compared by simulating net present values of after tax gross margins for each system. The traditional system generatedthe lowest net present value, approximately 5 to 17 per cent below the static economic system. Greater improvements, on the order of 14 to 31 per cent above the static economic system, were observed by following dynamic flex-cropping decision rules. Not only did the dynamic flex-cropping decision rules generate superior decision rules regarding mean net present values, the rules were also risk efficient. The probability of low gross margins was minimized in all cases by following the dynamic flex-cropping decision rules. The results of this and related studies indicate that dynamic flex-cropping models are viable for solving crop scheduling problems. The prescriptive power of the model is limited by available data, limitations which reside primarily in the agronomic components. The relationship between spring soil moisture, soil nutrients, and yield must be more clearly defined. This may be accomplished through extensive and long term field trials or through use of emerging biophysical models. Standardization of soil moisture classifications, including method of sampling and depth of measurement, would make field data more adaptable for making fertilizer and recropping decisions. The production functions defining the relationship between spring soil moisture levels and yields are particularly important. These require continued empirical attention. The model developed lends itself readily to extensions such as additional crops,fertilizer inputs, erosion costs and soil degradation issues, financial structure of the farm, and evaluating the influence of government programs. Modern computers with large computational and storage capabilities make the implementation of stochastic dynamic programming methodology a viable farm management tool.