In Cost-Benefit Analysis the bottom line that determines whether a proposed project is worthwhile is its Net Social Benefit. If the Net Social Benefit is positive the project is socially worthwhile and if it is negative it is not. If the Net Social Benefit is zero then society neither gains or loses from the implementation of the project.

Usually the bottom line in the evaluation of public projects is expressed in terms of the Benefit/Cost ratio. If the Benefit/Cost ratio is greater than unity then the project is worthwhile and if it is less than unity then the project is not worthwhile. The magnitude of the Benefit/Cost ratio is to a degree arbitrary so that it is generally not possible to say that a project with a Benefit/Cost ratio of 2.0 is better than one with B/C rato of 1.5. The arbitrariness however does not extend to changing a value that is less than 1.0 to a value that is greater than 1.0.

The arbitrariness stems from the extent to which the operating costs are netted out from the benefits and thus disappear from the cost figure used in the denominator of the Benefit/Cost ratio.

An algebraic example will illustrate the problem. Suppose a project has an initial cost of C₀ and constant annual operating cost of c₁. Based upon the life of the project and the interest rate there is an annuity factor A that converts the annual operating cost into the present value of the operating cost C₁; i.e., C₁=c₁A. If the constant annual benefits of the project is b then the present value of the benefits would be B=bA. Thus the Net Social Benefit of the project is (B&minus(C₀+C₁)) and the B/C ratio would be B/(C₀+C₁).

But suppose the annual operating costs are netted out (deducted from) the annual benefit. The annual net benefit would be (b−c₁) and its present value would be (b−c₁)A, which is just equal to B−C₁. The cost of the project would justbe its initial cost C₀. The Net Social Benefit would be the same as before ((B−C₁)−C₀)=(B−C₀−C₁) but the B/C ratio would be the higher value of (B−C₁)/C₀.

Since some operating costs could be netted out of benefits and others not netted out the B/C ratio could be anything between the two extremes. Thus in comparing two worthwhile projects the degrees of netting out of operating cost are generally not known it is not appropriate to make any comparison of the projects based their B/C ratios.

If there is not enough funds to fund all of the worthwhile projects then it is not valid to select for funding the projects based upon their B/C ratios, other than to reject those projects with B/C ratios less than or equal to 1.0. To eliminate projects the present values of the benefits and costs must be recalculated using a higher cost of capital (interest rate). If the required funding for the projects deemed worthwhile is still greater than what is available the cost of capital must be raised and the present values recalculated.