Project selection with sets of mutually exclusive alternatives

Abstract

We study the problem to maximise the net economic benefit of an investment plan by selecting from a portfolio of candidate projects within a given budget constraint. As is well known, with independent projects the economic efficiency of the entire investment plan is maximised if projects are selected according to their benefit-cost ratio until the budget is exhausted. Often, however, the planning of a project involves a stage where a set of alternative concepts or designs are considered. A best alternative is chosen, and the plan is composed from the pool of all such best alternatives. This procedure violates the assumptions underlying the benefit-cost ratio criterion. In this paper, we set out the correct criterion to use. A real-life example from Norwegian transport planning is provided to show how the global setting into which the project is going to compete, matters for the selection criterion to be used.