Decision-making in agriculture : a farm-level modelling approach

Abstract

In the past decade South Africa experienced major political and economic changes. In addition to these major changes, South Africa is a highly diverse country and a country of extremes in many respects. Within this dynamic and diverse environment the agricultural sector has to survive and grow financially. In order to survive and grow, good decision-making within the agricultural sector in terms of policies and business strategies is extremely important and necessary. However, within the dynamic and extreme environment it is very difficult for decision-makers to make correct decisions since the likely impact of changes in markets and policies is difficult to quantify. The general objective of this dissertation is to identify and construct a type of farm-level model that will have the ability to quantify the likely impact of change in markets and policies on the financial viability of a representative farm. The specific objective is to construct a model of a representative grain and livestock farm in the Reitz district, Free State province, South Africa. The approach to farm-level modelling that is followed is a positivistic approach since questions of “what is the likely impact” is asked, and not “what ought to be” questions. Apart from behavioural equations, this farm-level simulation model also consists of accounting identities. The model is of a deterministic type since explanatory and descriptive types of questions need to be answered. The development of this farm-level model contributes to research in the field of farm-level modelling in South Africa due to the fact that it has the ability to simulate the impact of changes in markets and policies on a representative farm’s financial position. This is done by linking the farm-level model to a sector-level model developed by Meyer (2002) as well as outputs from several other institutions in terms of macro-economic variables and social variables. There are, however, several issues that became clear in this study. Firstly, positivistic simulation models have the disadvantage that validation and verification are difficult and time consuming due to lack of accurate and detailed data. Secondly, due to the positivistic nature of the model, the assumption is made that very little adjustment in terms of the farm structure takes place during the simulation process. One possible solution to this problem of not being able to simulate adaptation to changing conditions is to develop a model following a normative approach. The third problem with specifically the deterministic type of model is the fact that the model and simulation process assumes no risk. Lastly, in following the positivistic approach, the modeller needs theoretical as well as practical knowledge and understanding of the system modelled and simulated, in order to simulate reality as closely as possible.