In this paper the empirical implications of altruism for cost-benefit
analysis of projects involving health changes are investigated. It is
shown that a willingness-to-pay question allowing the respondent to st
ate her total willingness to pay (irrespective of what reasons she may
have for paying), subject to everybody else paying so as to stay at t
heir initial levels of utility, produces, as a special case, the proje
ct evaluation rules derived by Jones-Lee (1991, 1992) and others. The
implications of alternative formulations of the valuation question in
a contingent valuation study are also explored.