Academic journal article Contemporary Economic Policy

Sensitivity Analysis in Benefit-Cost Analysis: A Key to Increased Use and Acceptance

Academic journal article Contemporary Economic Policy

Sensitivity Analysis in Benefit-Cost Analysis: A Key to Increased Use and Acceptance

Article excerpt


Nearly everyone generally agrees that decision making is a comparison of benefits and costs. Beyond the abstract, there is much controversy, especially about giving benefit-cost analysis (BCA) a formal role. For example, the formal role of BCA was expanded by a 1995 federal regulatory reform bill. It passed the House, but not the Senate. Controversy is especially intense for formal decision making that involves large environmental impacts.

The environment could benefit greatly from expanded use of BCA, but the major environmental groups are adamantly opposed to expanded use of BCA. They believe that most BCAs would either be biased against the environment or be ignored when they were pro-environment. Fear of BCA has driven environmental organizations to lobby for legal language that mandates certain actions regardless of expense. The Clean Air Act, the Water Pollution Control Act, and the Endangered Species Act have such language.

The sources of controversy include the identification of effects, deciding which are costs and which are benefits, and their future size and present dollar value. Randall (1981), Portney and Harrington (1995), and Zerbe (forthcoming) describe the controversies in more detail. Hines (1973) contains a nice story that illustrates the controversies. BCA practices must be formally revised specifically to address these issues. The BCA process, and its published findings, must be transparent enough to discourage deliberate bias, and it must alert readers to inadvertent bias, and to the significance of data gaps.

The aim of this article is to show how more extensive use of sensitivity analysis (SA) would address such concerns. SA is an open, public investigation of the sensitivity of net present value (NPV) and benefit-cost ratio (BCR) calculations to changes in assumptions, numbers, and computation methods. The desired outcome is a more informative, less controversial, more frequently used decision making tool. More frequent use probably will depend on achieving the first two goals. Being more informative means making BCA less subject to abuse and misuse, such as use of BCA to justify political decisions (see, for example, Hines, 1973; De Alessi, 1996), rather than more thoroughly to help examine choices. Controversy and more frequent use depend, in part, on the difference between informative and decisive. Zerbe (forthcoming) explains in detail how BCAs inform decision making and why BCA alone does not determine the appropriate choice. Because of especially troublesome issues, and the organized resistance to BCA, the potential gains are especially high for policy decisions with environmental impacts.


A comparison of political costs and benefits is the natural driving force of government behavior. Such informal comparisons often create strong incentives to manipulate formal, economic BCA findings to justify the decision implied by the political calculus. BCA authors can bias their findings in the direction that would please their sponsors by carefully selecting data, assumptions, or computation methods, and many probably have (see Hines, 1973; De Alessi, 1996, for good examples). As Schmid (1989) points out, "everyone in government is part of the political process, the hired analyst as well as the elected politician." The GAO (1984) concludes that because of political pressures and agency self-interest, BCA "will not be consistently prepared in an unbiased, systematic manner" even if the GAO's recommended changes in organizational structures and BCA authors' incentives were made.

Despite the major conceptual and computational innovations in BCA that economists have made (see Zerbe and Dively, 1994; Hanley and Spash, 1993, for a summary), BCA findings could be misleading, even without deliberate bias. Environmental impacts are still especially vulnerable to being shortchanged by traditional BCA's shortcomings. For example, since environmental impacts can be longlasting, the controversial discounting issue (Baumol, 1968; Ramsey, 1969; Usher, 1969; Arrow and Lind, 1970; Hirshleifer and Shapiro, 1969; Arrow and Kurz, 1970) is especially important. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.