Economic decision models for agricultural pest control, with special reference to the control of powdery mildew in barley

Abstract

Introduction: The fundamental importance of systematic research for agricultural development has been widely stressed (e.g. by Mellor, 1966; Moseman, 1970). It is therefore surprising that agricultural research has been exposed to so little critical economic appraisal. This is especially true where we might least expect it to be so: in the design of research programmes for the less-developed countries of the Third World. Some of the ways in which agricultural research is deficient have been described recently, in an East African context, by Belshaw and Hall (1972). This effect may be exaggerated by the less well developed capacity of peasant farmers to bear risk: they can not afford to be very far wrong in the decisions they make. However, the phenomenon is a general one, not only in the Third World, but in developed agricultural systems too. Agricultural research must concern itself not only with the provision of technical innovations, such as new types of seed or pesticide, but also with the provision of sufficient information for their economic appraisal, tailored to the conditions on each farm and to the attitude that the individual farmer has towards the risk that the long-run average response will not be realised in any particular year. This dissertation examines in some detail a particular innovation: the control of the disease, powdery mildew, of spring barley in the United Kingdom. Its approach is that of technological economics (Bradbury, 1969), a point of view peculiar to the Departments of Industrial Science and Economics at Stirling University but particularly concerned with problems of resource allocation under conditions of uncertainty. Its theme is the use, and usefulness, of agricultural research and the experimental data it provides. It confirms many of the deficiencies noted by Belshaw and Hall (op. cit.) in the special case of a particular innovation. But it also shows how, in the particular case of agricultural pest control, the micro-economic allocation problems facing the farm firm can be formulated in such a way that they become susceptible to experimental analysis. A summary account of barley production in the United Kingdom is presented in Appendix 1. Data concerning the importance of powdery mildew as a factor limiting yield are discussed in the text. More detailed specifications of the pesticides used against barley mildew are given in Appendix 2. Chapter 1 reviews the literature dealing with the economics of pest control and presents the substance of the farm-level decision models that have been proposed so far. Chapter 2 discusses the choice of variety, both with and without a systemic mildewcide seed dressing. The decision is formulated as a problem in portfolio selection. A practical approach using linear programming is illustrated. Chapter 3 extends the farmer's range of choice to whether or not he should apply the pesticide as a spray after the disease has entered his crop. The time of pesticide application is considered as an explicit decision variable. Chapter 4 considers the aggregate costs and benefits of pest control on the national farm, including the possibility of imposing a legislative ban on the cultivation of winter barley which provides a "green bridge" between successive spring crops. It stresses some of the problems of applying social cost-benefit analysis to any national investment in agricultural pest control. Chapter 5 summarises the deficiencies in the experimental data provided by agricultural research and described in previous chapters. An attempt is made to construct an analytical framework for research into the economics of pest control. It is suggested that mathematical simulation is the appropriate technique for the analysis of many problems of this nature, and that agricultural research might be more properly directed towards the determination of the coefficients in an analytical model of the situation than to empirical small-plot trials of doubtful relevance. It is not claimed that model building is a novel way of looking at problems (like M. Jourdain in Moliere's "Le Bourgeois Gentilhomme", who found to his amazement that he had been talking prose for more than forty years), only that direct experimentation on the physical yield response at selected locations provides an insufficient model of the economic decision problem facing the farmer. Chapter 6 discusses the potential usefulness of properly calibrated economic decision models for agricultural pest control to various groups of people. It also considers possible objections to the use of models, and suggests ways in which the preliminary models developed in the text might be extended to other situations such as the control of soil-borne pests and the analysis of complete spray programmes.