Returns on agricultural research and development in the South African sugar industry.

Abstract

The lack of information on returns to R&D is considered a handicap to effective decision-making by policy-makers and managers in agricultural commodity organisations. Results of studies reported in the literature, mostly using economic analysis of aggregated and multi-product data, are usually insufficiently detailed to assist decision-making at institute level. The objective of this study is to find an empirical and practical method of estimating returns for that purpose. Returns on sugarcane production R&D in the South African Sugar Industry are estimated as the factor share of technology in a production function analysis of productivity, as yield per unit area per annum, in which the other significant variables were found to be rainfall, costs of production and area under crop. Eight other variables were excluded from the analysis for lack of significance or collinearity. Under a user pays policy, advisory services are considered self-financing, leaving the estimated returns to be divided between the other two primary functions of an R&D institute, research and extension. It is suggested that the increase in yields obtained by technologists in field trials can represent technology (the output of research) while the increase in the Industry's yield over the same period represents technology plus the transfer of technology (the function of extension) . In percentage terms the ratios of research to extension, for three successive decades to 1986, were found to be 65%:35%, 37%:63% and 17%:83%, indicating decline in the contribution of research and increase in the contribution of extension to the Industry's declining productivity. Research's contribution (17% of the total return on R&D during the last decade) was then apportioned among research programmes in the proportions of the subjective estimates made of their returns, after deducting the return on plant breeding, the only programme whose productivity could be quantified directly from production data. Returns and costs are then compared in terms of percentage net returns [(returns - costs) /costs x 100) and benefit:cost ratios (return/cost). The returns estimated on research, extension and whole Station activities, were similar, in terms of benefit : cost ratios, to those obtained in the few other comparable studies. The advantages of the methods proposed are their empirical simplicity and applicability down to programme (project) level.