Cost–benefit analysis has a long hi

Cost–benefit analysis has a long history. Its originsprobably begin with the work of the French engineerand economist, Jules Dupuit (1844, 1853). Efforts toclaim earlier origins with US Secretary of theTreasury, Albert Gallatin, in 1808 (Hanley andSpash, 1993) do an injustice to the profound intellec-tual effort produced by Dupuit. Dupuit establishedthe foundations of what today would be called‘marginal analysis’, defined the way in which ben-efits and costs should be measured, and firmlyembraced the principle that an investment decision,such as building a road or bridge, should meet acriterion that benefits exceed costs. The exactmeaning of benefits and costs is explored below.While the theoretical foundations of marginal analy-sis were greatly expanded and developed in the late19th century, little happened to apply these princi-ples in practice. In 1936, however, the United StatesFlood Control Act stated that water projects in theUSA should proceed ‘if the benefits to whosoeverthey accrue are in excess of the estimated costs’,embracing the idea that all gainers and losers shouldbe considered. The stimulus was public and politicalconcern over the effects of major hydro-powerconstruction. The formalization of these notionscontinued after the Second World War: in 1946 asubcommittee of the Federal Interagency RiverBasin Committee was set up to consider benefitsand costs, and in 1950 the subcommittee producedthe ‘Green Book’ of guidance on the benefit–costappraisal of water projects. Circular A-47 of theBureau of Budget produced further guidance in1952. In hindsight, all of these efforts were rudimen-tary and lacked proper theoretical foundations. Nonethe less, they were catalytic in effect.At the same time as the practical guidelines werebeing drawn up in the water sector, considerableattention was being devoted to ‘efficiency in gov-ernment’, especially military spending, by bodiessuch as the Rand Corporation. In 1958 three seminalworks appeared: Otto Eckstein’s Water ResourceDevelopment; John Krutilla and Otto Eckstein’sMultipurpose River Development; and RolandMcKean’s Efficiency in Government ThroughSystems Analysis (Eckstein, 1958; Krutilla andEckstein, 1958; McKean, 1958). The feature ofthese works was the synthesis of practical concernswith the theoretical ‘welfare economics’ literaturethat had emerged in the 1930s and 1940s, andnotably the work of Kaldor and Hicks (Kaldor,1939; Hicks, 1939, 1943). The essential step herewas the justification for the benefit–cost principle:justifying projects or policies on the basis that ben-efits exceed costs is wholly consistent with therebeing losers, i.e. those who suffer the costs. TheKaldor–Hicks criterion established that projectswere none the less justified because gainers couldcompensate losers, such that losers would be noworse off and gainers would still have a net benefit.This implies that, provided the compensation takesplace, no one is actually worse off, thus meeting thelong-established ‘Pareto criterion’ for an improve-ment in overall wellbeing.2However, actual com-pensation need not occur: it is necessary only that it