Cost overruns have been shown to occur in a significant number of transportation investment projects. While there is some literature on the causes of risk of cost overruns in transportation capital investment, little is available on how properly to estimate this risk and how to incorporate risk estimates into project assessment. Thus, the main objectives of this paper are to propose methods for estimating the risk of cost overruns and suggest ways to incorporate it into project decision-making. To these ends we propose distribution-fitting models to elicit the probabilities of cost overruns. These probabilities, in turn, are used as inputs into a decision tree model of project selection. Using a database, made up of observations on a large number of highway projects, the paper shows how empirically this analysis can be carried out. A key conclusion is that the approach proposed here can provide realistic risk estimates, thereby reducing subjective biases in project cost benefit analysis.
Abstract