Economics conventionally assumes that preferences are coherent, i.e. stable, context-independent, and consistent with axioms of rationality. Since these assumptions underpin standard interpretations of cost-benefit analysis (CBA), preference 'anomalies' found in stated preference surveys pose severe problems. This paper proposes an alternative interpretation of CBA as the simulation of the workings of competitive markets in situations in which, in reality, there is market failure. The object of a CBA of a project is to measure the surplus that is created by that project. It is shown that surplus can be defined without assuming that individuals have coherent preferences. Some of the ways in which this approach to CBA can overcome problems caused by preference anomalies are discussed.
anomalies cost-benefit analysis surplus willingness to pay