Two common pitfalls to be avoided in benefit-cost analysis are:
Double-counting of certain costs and benefits by including the same economic impact more than once, in what erroneously seem to be different measures
Including as benefits or costs monetary exchanges which are actually transfer payments, that is, transactions where money moves around without anything of economic value being created or consumed
Although these pitfalls may seem obvious, they sometimes assume subtle forms.
A study includes the cost of constructing a sound wall in the initial project costs, while also counting the disbenefit of noise from the project without a sound wall. This is a mistake of double-counting. Only the disbenefit of noise with the sound wall should be counted, since the rest of the noise impact is eliminated by the sound wall.
A study of a proposed toll decrease on a publicly owned highway counts the benefit of lower vehicle operating costs from reduced toll, even though the costs of toll collection itself remain unchanged. This is a mistake since for government projects tolls and taxes are just transfers between different members of society (as are fares paid to public transit operators). Tolls, fares, and (to a lesser extent) taxes are financial tools used to transfer some or all of a project's cost to its direct beneficiaries and away from society as a whole. While they may be useful for identifying winners and losers, they do not correspond to net impacts on society as a whole.
A study of a proposed downtown transit improvement includes as a benefit the downtown parking fees avoided by former auto users. This may be a mistake since parking fees may also just be transfer payments. On the other hand, if reduced demand for downtown parking actually frees up land or structures for other uses, the value of those released resources would properly count as a benefit. (To the extent that parking fees reflect the value of the resources thus freed up, parking fees could indeed be a reasonable surrogate measure of that benefit.)
To avoid including transfer payments, focus on whether or not there is consumption or savings of real resources with economic value (time, land, materials, clean air, etc.). When resources are consumed or saved, that generally indicates a true cost or benefit. This applies whether or not there is an actual market for those resources. (For example, while there is a market for land, there is no market for cleaner air.) When money is merely moved around among members of the society on whose behalf a project is proposed or to/from their government(s), these movements are usually just transfer payments.
The viewpoint of the analysis helps identify transfer payments. If the purpose of a benefit-cost analysis is to evaluate a private sector investment, its viewpoint would be that of the individual firm, where revenues and expenditures on taxes are properly counted as costs and benefits. In this narrow viewpoint, anything that affects company profitability matters. For government investments, however, accepted practice evolved from the US Flood Control Act of 1936, which states that "benefits to whomsoever they may accrue" should exceed estimated costs (Grant et al, Chapter 7). The correct approach for public investments is therefore "to take a viewpoint at least as broad as those who pay the costs and those who receive the benefits" (Newnan et al, p. 578). This avoids the classic errors in evaluation of local infrastructure proposals where analysts ignore the federally funded portions of costs (regarded as "free money") or environmental consequences "exported" to far outside the local region. In the broad viewpoint properly used to evaluate public investment decisions, we seek to identify net differences among alternatives as they impact the overall society. In such cases, many financial outcomes are actually internal transfers and therefore irrelevant to the decision at hand.
A complicating factor for some public sector investment decisions occurs when the project under consideration might also be a viable candidate for private investment as a for-profit enterprise. In such cases, the loss of tax revenues might properly be treated as a cost component of the project (Fabrycky et al, p. 153). Typically, however, public investment does not directly compete for projects which are appropriate private sector investments, and tax impacts (which include fuel tax expenditures) should be ignored.
The key to avoid double counting is to make sure that, once transfers are set aside, each dollar of impact on society is counted only once, either as a benefit or a cost. In categorizing a particular dollar of impact as a benefit or a cost, it may be helpful to focus on the basic distinction that costs are impacts on those who provide a project, and benefits are impacts on its users and others who might be affected.
Grant, E. L., W. G. Ireson, R. S. Leavenworth. Principles of Engineering Economy. Eighth Edition. John Wiley & Sons. New York. 1990.
Fabrycky, W. J., G. J. Thuesen, D. Verma. Economic Decision Analysis. Third Edition. Prentice Hall. Upper Saddle River, NJ. 1998.
Newnan, D. G., J P. Lavalle, T. G. Eschenbach. Engineering Economic Analysis. Eighth Edition. Engineering Press. Austin TX. 2000.