Brekke [Brekke (1997) Journal of Public Economics 64, 117-123] claims that
using money as the numeraire in cost-benefit analysis "is systematically fa
vourable to those who value money the least, relative to alternative numera
ires." This note attempts to show that the outcome of a properly undertaken
cost-benefit analysis is independent of the choice of numeraire. However,
since the marginal social utility of income (or, say, an environmental good
) is unobservable, one faces a formidable problem in assessing the social p
rofitability of a project which affects more than a single person. (C) 1998
Elsevier Science S.A. All rights reserved.