Abstract
This paper describes a new approach to the theory of agricultural research benefits based on component benefits rather than, as is customary, on surpluses. The surplus concept nevertheless emerges when the distribution of the national benefit is considered. Various published methods (Grossfield‐Heath, Lindner‐Jarrett, Akino‐Hayami, Griliches I and II) are derived as special cases of the theory and numerically compared.Applications are briefly considered; Griliches II is used for the variability of crop yields and some typical results for the distribution of benefits are given.