An effective priority-setting system has several key attributes which must be manifest both in the system's development and in its implementation. Without them, systems can be developed but will have little impact and not last long. The attributes are important for priority-setting in any organization, but they are particularly important in systems intended to influence decisions in organizations that are as complex and subject to such diverse and changing pressures as EM. Although the subcommittee observed encouraging signs in some attributes—such as stakeholder involvement, work is needed in all of the following key attributes:

• Permanence and consistency. The subcommittee has observed a lack of consistency in EM's efforts to state its goals. Budget documents and other official pronouncements often began (rightly) with a statement of the organization's goals, but the statement commonly differed from those in other documents issued at the same time. Such deviations and inconsistency will substantially inhibit the effectiveness of a new priority-setting system. EM is attempting to modify substantially the operations of a large, complex organization that is known more for its inertia and rigidity than for its agility. To make such a change requires consistency in statements about where the organization should be headed. The subcommittee recommends that the priority-setting system be established through a careful process that involves substantial opportunity for input from the full range of stakeholders. The subcommittee also recommends that once established, the priority-setting system be described and implemented consistently.
• Clarity and transparency. The entire process, its development, exposition, and implementation, should be as clear and transparent to affected stakeholders as possible. Unfortunately, DOE has engendered a substantial legacy of suspicion and mistrust because of decades of operating secretly. It has made much progress already in reducing this mistrust. A clear, transparent priority-setting system will reinforce this trend and should encourage the early exchange of concerns to foster common conclusions.
• Simplicity. The simpler the priority-setting system and its tools are, the more likely it is to be followed and trusted. Although the associated methods should be scientifically defensible, they must also be understandable to affected parties. Because many of the criteria to be rated require subjective judgment and the inventories of toxic materials are incompletely defined, it is not always possible to establish scientific certainty. Hence, the system should not be so complex as to require elaborate analyses and consume an unreasonable amount of time to prepare, especially when the data to support great detail are not available. As more data become available or as the need for greater precision in estimating risks, costs, or benefits arises, then more complex analytical approaches are justified.
• Stakeholder involvement. The legacy of low public trust in and credibility of DOE originated in the need for secrecy in some programs and authorities assigned by Congress for self-regulation. The Secretary has taken steps to involve people affected by DOE's actions in the decision-making process. This has been successful and should be continued formally. Stakeholders do not have authority to determine funding, but they should participate in and understand the basis of funding decisions. The inclusion of state and local stakeholders in this process for FY 1997 is laudable. DOE's experience, and the experience of other organizations in similar circumstances, suggest that regulators and public stakeholders need to participate to the greatest extent feasible in priority-setting that leads to DOE budgetary decisions. (This point is similar to that made in the National Research Council's Building Consensus report (NRC, 1994a) with respect to DOE's use of risk assessment in its environmental remediation program.) However, public and stakeholder interest is strongly aligned with the interests of specific sites, and DOE's current mechanisms for public involvement appear to do a better job of promoting the budgetary stakes of particular sites than of facilitating intersite budgetary tradeoffs. Thus, while allowing for public participation, DOE must also fundamentally alter its budget-allocation process to allow for more-centralized setting of overall goals that promote the national interest. EM should apply the following principles related to stakeholder involvement:

• —For stakeholder involvement in priority-setting to be knowing and intelligent DOE must provide stakeholders with all appropriate planning and budgeting guidance.
• —Stakeholder involvement in priority-setting must occur at the installation level, the field-office level, and the Headquarters level. Therefore, DOE should tailor stakeholder involvement to specific priority-setting requirements, such as different types of stakeholder representatives and different types of stakeholder training.
• —Stakeholder involvement in different EM program components—e.g., Community Leaders Network, the Military Toxics Project (a network of community groups concerned with environmental justice), Transportation External Coordinating Committee, Environmental Management Advisory Board, and State and Tribal Government Working Group (STGWG)—should be integrated into the overall priority-setting effort.
• —Stakeholder involvement in priority-setting must be effective in helping EM to establish and resolve conflicts in field office and installation priorities and Headquarters priorities. Therefore, EM should annually evaluate stakeholder priority-setting efforts. The exercise should not be a one-time effort but should be an iterative process that encourages accountability of DOE field management and contractors.