Should the government allow drilling for oil in the Alaskan wilderness? How much arsenic should be permitted in municipal drinking water? What level of pollution-abatement technology should factories be re quired to utilize? How much money, if any, should the state invest in a new vaccine program? How robust should floodwalls in low-lying coastal cities be? These are difficult, vexing questions. Cost-benefit analysis (CBA) is a policy tool that—according to its proponents at least—is suited to answering these questions and the untold others like them. It is intended and frequently employed to evaluate public policy proposals, especially those concerning regulations and investment projects. However, CBA's many critics charge that it is undemocratic and morally bankrupt. Whether these charges are well founded is a pressing question. It thus seems that cost-benefit analysis is itself in need of further analysis. Advocates of CBA maintain that a policy proposal should be en dorsed when it would result in more benefits than costs to the affected individuals, with the gainers being able to compensate the losers such that the losers would be no worse off, and the gainers still better off, than each was originally. This principle, known as the Kaldor-Hicks Criterion,1 leaves open the question of what counts as a cost and what counts as a benefit. CBA's supporters hold that whether a change counts as a benefit or a cost to an affected individual depends on the individual's assessment of the change. More specifically, if the change satisfies a person's pref erence, then it counts as a benefit for that person; and if the change frus trates the individual's preference, it counts as a cost for the individual. In order to apply the Kaldor-Hicks Criterion, and more generally to weigh costs and benefits against one another rationally, it is said that all benefits and costs must be expressed in a common unit, ideally a mone tary one. This means that each individual's preferences must be expressed