Thus far we have mainly been concerned with the definition and measurement of money measures at the individual level. In a few cases, average WTP measures, such as the WTP per capita, have been used. There is of course nothing wrong with such averages per se, but the central question is what, if any, useful information the decision-maker can extract from the average WTP for a proposed measure. After all, behind an average, there are typically both those who gain and those who lose. It is sometimes claimed that a positive aggregate CV or EV, which is equivalent to a positive average WTP, means that those who gain from a project can, at least hypothetically, compensate those who lose from it. If true, this claim would mean that a distributional analysis of a project's consequences is more or less superfluous. However, more than 20 years ago, Boadway (1974) proved this claim to be false. Thus, the aggregation issue is still of central importance in project evaluations.

In this chapter we will take a look at the aggregation problem and examine the conditions under which the sum of WTPs for a measure indicates whether a proposed project is socially profitable. The chapter should also be useful for those working with other output measures such as qalys and hyes; there are always some basic ethical assumptions behind an aggregation procedure, and the investigator should clarify these assumptions to the decision-maker and other users of the investigation.