Management bias and returns to scale in a Cobb-Douglas production function for agriculture

The purpose of this paper is to reconsider the specification and estimation of agricultural production functions. By assuming a criterion of expected profit maximization and incorporating a management input into the function using dummy variables, least squares estimates of the parameters of the production function are unbiased and consistent. Functions are estimated for four years of annual Farm Management Survey data for the North-West of England. Results presented reveal both reduced marginal productivity estimates and returns-to-scale compared with conventional cross-section estimates.