How Can We Improve Evaluation Methods for Public Infrastructure?

Given the smaller total budget for public expenditure and the fact that the cost of public funds to Ireland has increased, it is more important than ever to ensure that public investment is prioritised properly in order to derive maximum benefit. Inorder to prioritise we must evaluate. A variety of evaluation methods can be utilised, but perhaps the most widely used is cost benefit analysis. The usefulness of costbenefit analysis crucially depends on a number of parameters and inputs. This paper considers the international literature on two issues, namely, the impact of risk in theform of inaccurate cost or benefit estimates and the setting of the appropriate discount rate, both of which can impact significantly on the usefulness of costbenefit analysis. The evidence on the expected costs and benefits of projects highlights that projects often do not go according to plan and that these estimates are subject to systematic optimism bias, which, while not universal, appears to be widespread. In relation to the appropriate choice of a discount rate, a riskless rate should in general not be used unless all risks have been properly assessed and costed within the analysis. This paper concludes that the discount rates that are currently used in Ireland appear to be low. Furthermore, the paper highlights that if the conventional exponential discounting is used then costs and benefits that occur in the distant future are essentially ignored. A declining discount rate accounts better for costs/benefits that occur in the distant future and is consistent with the observed pattern of time preference of individuals. This paper recommends a hybrid approach be adopted where costs and benefits are discounted using exponential discounting up to a point at which the discounting is switched to declining discounting.