The estimation of stochastic cost-quantity frontiers for discretionary cost and budgeted-expenditure systems

Two alternative approximating functions for representing the unknown composite residual density function of a cost-quantity frontier are examined. The frontier is one in which cost is regarded as preselected. This is appropriate for public sector economics and discretionary cost situations. One function requires the estimation of a single distribution parameter, whereas the other makes no a priori assumption on the shape of the density function.