An investigation into the use of mathematical models for resource allocation in agricultural research

Abstract

The management of state financed agricultural research have the responsibility of selecting a portfolio of projects which will provide the greatest benefit to society for the resources invested. The purpose of this investigation was to examine the possibility of using mathematical models to aid in the making of resource allocation decisions within agricultural research. Both the data and criteria on which a quantitative project selection and resource allocation procedure could be based were found to be inadequate. Consequently, a formal Resource Allocation System for Agricultural Research (RASAR) was developed as a framework within which mathematical models could be developed and used. RASAR was conceptualized as in iterative system with the purpose of selecting a portfolio of research projects such that the research outputs would provide society with the potential power to change the agricultural system in ways that are expected to bring about the greatest improvement in social welfare. The ultimate goal of agricultural research was tentatively identified as having nine dimensions in three broad categories: Consumption category — (1) quantity, (2) quality, (3) availability; Security category — (H) human safety, (5) economic defence, (6) food sources security, (7) conservation; Equity category — (8) distribution, (9) individual rights. Subsystems within RASAR for generating socio-economic data relating to these dimensions were specified and tested with four case study research projects. A mathematical programming model which could provide management with a tool for assimilating the complexity of criteria and data into a form which is readily usable for decision making was developed and evaluated. A number of conclusions emerged from the research: (a) mathematical models can be effectively used to assist agricultural research management with the complex problem of resource allocation, providing an adequate system for specifying selection criteria and for generating data is utilized; (b) RASAR offers considerable scope for development into an effective system for the more effective and rational allocation of research resources; (c) the multiplicity of objectives or reasons for research and the lack of adequate socio-economic data which tend to make resource allocation decisions difficult can be adequately brought together in a well defined resource allocation system and used to improve the decision making process; (d) projects which appear to have beneficial outputs and be justifiable in terms of their immediate objectives may, in fact, have obscure but substantial social costs that are not apparent without a rigorous socio-economic assessment.