Abstract
The paper investigates the distributional basis for benefit-cost analysis when utility functions are independent. The change in utility for any individual caused by a given project can be written as the product of the marginal utility of income (Λ) and the income equivalent (measured by the appropriate area under the demand curve). Three theoretical positions are possible to deal with the distributional issue. First, one may assume Λ to be the same for all those affected by the project. A second position admits that Λ is dependent on income and tries to determine how different people view their own marginal utilities of income. Finally, it is possible to define the marginal utility of income in terms of social value judgments. This requires a social welfare function which reflects either the collectivity's expressed consensus or the decision maker's own set of judgments. The first approach is most commonly adopted and most convenient for actual analysis. The paper advocates more explicit analysis of distributional effects than is commonly undertaken and disputes the claim that such analysis is outside the professional realm of the economist.

This publication has 17 references indexed in Scilit: