Interpreting Dual Variables for Optimization with Nonmonetary Objectives

Abstract

An approach to the interpretation of dual variables for mathematical programs with nonmonetary objectives is described. The approach is general, and in particular may be applied to problems of goal programming, expected utility maximization, intertemporal utility maximization and risk minimization subject to minimum income constraints. The technique is designed to transform shadow prices, expressed in marginal increases in the objective per unit of resource, into easily interpreted units, such as dollars per unit of resource.