Finally, the analyst must consider whether the outcomes of interest are inter-temporal. This is a complication that often arises, as formal decision making frequently requires comparing costs and benefits obtained at different points in time. Economists typically use a discounting function to decrease the importance of costs or benefits occurring further in the future. Discount rates relate future monetary values to the present, corresponding to the empirical reality that actors prefer current consumption to future consumption. Discount rates arise for two reasons. First, there is an macroeconomic basis to discount rates, whereby economic growth and inflation rates mean that the real purchasing power of a unit of wealth decreases over time. Second, there is a moral (or social) element of discount rates when seen from an intergenerational perspective, whereby the discount rate represents the preference of consumption of this generation over consumption of future generations.

---

Uncertainty, expected outcome and expected utility

In cases where there is uncertainty about outcomes, further complications arise. The analyst may consider whether there is probabilistic information on potential outcomes. If outcomes can be represented through a probability density function, then an option can be assigned a value according to the expected outcome. However, uncertainty in outcomes raises a further issue as both economic theory and empirical evidence has shown that people generally have preferences on uncertainty. Therefore valuation can be applied to the uncertainty in outcomes. In other words, a relevant question to consider is how much more would people pay for an outcome that is certain than they would for an uncertain outcome, but with the same expected value? In order to address this, it is necessary to estimate a utility function for an individual respect to the outcome. In general, the utility function is shaped by diminishing marginal utility, which reflects the principle that past a certain threshold increasing quantities of the same good bring little additional utility. Because of this the expected utility of option will differ from the expected outcome, as outcomes which are at the tail end of an expected outcome distribution contribute little to expected utility. This is another way of saying that people are generally risk averse, and in general prefer a certain outcome to an uncertain outcome with an equal expected value.

For a public actor, the utility functions of affected actors must be aggregated into a social welfare function. These considerations apply to situations in which outcomes can be represented probabilistically. When future outcomes cannot be represented probabilistically, valuation methods are not applicable.

While the tasks and methods discussed in this section have been applied extensively, it is important to note that they have also been subjected to substantial criticisms. The valuation tasks and methods described in this subsection are largely based on the neoclassical economics approaches of welfare economics. Criticism of these approaches has focused on the unrealistic assumptions made about actor’s choice processes in order to support valuation methods. Critics point to the empirical fact that regularities in cognitive biases exist in individual decision making, so that framing effects may influence valuation (Kahneman et al. 1982). Others have criticised valuation methods for enabling trade-offs to be made between outcomes should be seen as incommensurable. There are, for example, arguments to be made against the valuation of species extinction or human suffering (Vatn, 2005). In this sense, applying valuation may encompass a strong normative component and the analyst should be aware of these issues when deciding whether to apply valuation methods.