Corps of Engineers:
Observations on Planning and Project Management Processes for the Civil Works Program
GAO-06-529T: Published: Mar 15, 2006. Publicly Released: Mar 15, 2006.
Through the Civil Works Program, the Corps of Engineers (Corps) constructs, operates, and maintains thousands of civil works projects across the United States. The Corps uses a two-phase study process to help inform congressional decision makers about civil works projects and determine if they warrant federal investment. As part of the process for deciding to proceed with a project, the Corps analyzes and documents that the costs of constructing a project are outweighed by the benefits. To conduct activities within its civil works portfolio, the Corps received over $5 billion annually for fiscal years 2005 and 2006. During the last 4 years, GAO has issued five reports relating to the Corps' Civil Works Program. Four of these reports focused on the planning studies for specific Corps' projects or actions, which included a review of the cost and benefit analyses used to support the project decisions. The fifth report focused on the Corps management of its civil works appropriation accounts. For this statement, GAO was asked to summarize the key themes from these five studies. GAO made recommendations in the five reports cited in this testimony. The Corps generally agreed with and has taken or is taking corrective action to respond to these recommendations. GAO is not making new recommendations in this testimony.
GAO's recent reviews of four Corps civil works projects and actions found that the planning studies conducted by the Corps to support these activities were fraught with errors, mistakes, and miscalculations, and used invalid assumptions and outdated data. Generally, GAO found that the Corps' studies understated costs and overstated benefits, and therefore did not provide a reasonable basis for decision-making. For example: (1) for the Delaware Deepening Project, GAO found credible support for only about $13.3 million a year in project benefits compared with the $40.1 million a year claimed in the Corps' analysis; (2) for the Oregon Inlet Jetty Project, GAO's analysis determined that if the Corps had incorporated more current data into its analysis, benefits would have been reduced by about 90 percent; and (3) similarly, for the Sacramento Flood Control Project, GAO determined that the Corps overstated the number of properties protected by about 20 percent and used an inappropriate methodology to calculate the value of these protected properties. In addition, the Corps' three-tiered internal review process did not detect the problems GAO uncovered during its reviews of these analyses, raising concerns about the adequacy of the Corps' internal reviews. The agency agreed with GAO's findings in each of the four reviews. For three projects the Corps has completed a reanalysis to correct errors or is in the process of doing so; it decided not to proceed with the fourth project. GAO's review of how the Corps manages its appropriations for the civil works program found that instead of an effective and fiscally prudent financial planning, management, and priority-setting system, the Corps relies on reprogramming funds as needed. While this just-in-time reprogramming approach can provide funds rapidly to projects that have unexpected needs, it has also resulted in many unnecessary and uncoordinated movements of funds, sometimes for reasons that were inconsistent with the Corps' own guidance. Because reprogramming has become the normal way of doing business at the Corps, it has increased the Corps' administrative burden for processing and tracking such a large number of fund movements. For example, in fiscal years 2003 through 2004 the Corps moved over $2.1 billion through over 7,000 reprogramming actions. In response to GAO's findings, the Congress directed the Corps to revise its procedures for managing its civil works appropriations, starting in fiscal year 2006, to reduce the number of reprogramming actions and institute more rational financial discipline for the program.