Cost Estimation Issues

The U.S. Department of Transportation’s subcommittee on Transportation,
Treasury, and Independent Agencies presented a testimony regarding the cost drivers on highway projects to the United States House of Representatives in 2003. Kenneth Mead, the Inspector General, stated “if the efficiency with which the $500 billion invested by the Federal Government and States over the last 6 years had been improved by only 1 percent, an additional $5 billion would be made available - enough to fund 4 of the 17 active major highway projects.” During the investigation, Mead discovered that cost increases occurred because costs such as inflation, preliminary engineering, and construction management were not included in the estimate.

The General Accounting Office (GAO) performed several audits on how the Federal Highway Administration (FHWA) manages cost inflation on large bridge and highway projects. The GAO (2002) reported a majority of the cost escalation for a project occurs between the initial environmental review estimate and the detailed estimate. The FHWA often approves large projects in segments. As a result, a large public investment has already been made before proceeding segments have been approved. Furthermore, the process that a project must go through to be approved contributes to cost escalation. Initial estimates are completed for an environmental review of the best project alternative. These estimates are not focused on an accurate estimate of total cost, but rather focused on reviewing different design alternatives. Another problem is the FHWA has no statutory requirement to focus on project cost containment (GAO 2002).
Although cost estimating programs have improved due to advances in technology, estimating inaccuracy has remained the same for 70 years. A recent study analyzed the reasons for cost underestimation in public works projects (Flyvbjerg, Holm, and Buhl 2002). The study indicated that projects are 86 percent more likely to be underestimated than overestimated. Furthermore, the actual cost of road projects are on average 20 percent higher than their estimated costs. Flyvbjerg et al. (2002) stated that some promoters and forecasters intentionally deceive the public by claiming a project
costs less than it actually will in order to gain support and approval for a project. Many of the estimates that decision makers evaluate do not include important details such as environmental and safety concerns that could result in large cost increases.

Schexnayder et al. (2003) listed numerous reasons for cost estimating issues in their synthesis on project cost estimating. Some of the reasons cited were project scope changes, unforeseen engineering complexities, changes in economic and market conditions, changes in regulatory requirements, and local government pressures. Flyvbjerg et al. (2002) also provided several issues that result in cost underestimation. One issue resulting in underestimation is technical mistakes such as lack of experience or insufficient data. Another explanation for underestimation is economic incentives such as self-interest and public interest. Finally, project promoters and forecasters could underestimate projects for shear political gain and power.

When a project is reviewed for potential inclusion in a State Highway Agencies’s (SHA) project program, the earlier estimate is not always the most accurate reflection of the project’s cost. As a result, underestimated projects that are initiated prevent other projects that are economically feasible from being constructed (Flyvbjerg et al. 2002). Inadequate estimating invariably leads to misallocation of scarce resources. If estimates are consistently high, compared to bid costs and ultimately final costs, fewer projects will be authorized than could have been performed with the resources available, resulting in loss of benefits. If estimates are consistently low, more projects can be authorized than can be fully funded, resulting in project slowdowns, scope changes, performance shortfalls, and generally higher costs and lower benefits. If estimates are consistently neither high nor low, but still inaccurate, the estimated benefit/cost ratios will not be correct and the most beneficial projects may not be authorized, while less beneficial projects are authorized. All of these conditions result in misallocation of funds and a loss in benefits to the public (Flyvbjerg et al. 2002).
Other reasons for cost increases are inaccuracy of the scope and schedule, political pressure to complete a project within a certain budget, and decreasing or extending the project’s schedule. Chang (2002) found that the owner was responsible for one-third of cost and time increases; where as the consultant contributed the least tocost and time increases.