States' Experiences Can Inform Federal Efforts
GAO-05-215, Feb 28, 2005
With a number of challenges facing the nation--including a long-term fiscal imbalance--agencies need to maximize their performance and leverage available resources and authorities to achieve maximum value while managing risk. Examining state efforts to increase the focus on performance and their experiences in responding to recent fiscal stress can offer insights into practices that may assist federal decision makers in addressing the challenges ahead. GAO described for five selected states--Arizona, Maryland, Texas, Virginia, and Washington--legislators' use of performance information in budget deliberations, how performance information helped to inform choices during fiscal stress, challenges these states face in implementing and sustaining their efforts, and the potential for state experiences to inform initiatives to improve the use of performance information at the federal level. Among other factors, these states were selected because they have established histories of performance budgeting efforts and represent a variety of approaches to implementing those efforts.
Performance measures and the reporting of program performance are regularly included in the budget processes of the five states GAO visited. Legislators' expectations that this information will be collected and reported are supported through both statutory requirements and executive initiatives. GAO found that the continuing efforts to improve data collection and to relate this information to structures and processes used to make resource decisions were reinforced by the increasing capacity of staff to analyze, synthesize, and incorporate performance information in ways that make this information more accessible and useful to decision makers. State officials described ways in which performance information, including outcome measures and performance evaluations, was used in budget deliberations to identify potential impacts of a proposed policy change, make policy decisions that reduced costs while maintaining effectiveness, and make changes to improve program effectiveness. However, when determining funding levels and defining desired levels of service relative to funding, legislators currently rely most on workload and output measures. In addition to using some traditional tools or approaches to address budgetary shortfalls, such as across-the-board cuts or tax increases, most states GAO visited also developed new initiatives that considered existing performance information among other factors to respond to revenue shortfalls. For example, three states created prioritization initiatives that framed trade-offs according to how programs contributed to achieving statewide goals. Three of the states also established efficiency commissions to identify opportunities for cost savings by improving the structure and function of state government. A lthough the states GAO visited all demonstrated ways in which performance information was used in budget deliberations, officials in every state described challenges in developing and presenting performance information that is both credible and useful. Despite these challenges, these states have demonstrated a commitment to performance budgeting efforts by continuing to refine their approaches in response to those challenges. Success in performance budgeting requires time, agreement on the selection of measures reported, and understanding of different perspectives. The states are working toward this success. GAO convened a panel to discuss the implications of our state findings for the federal government. Panelists were encouraged that performance information has influenced legislative budget decisions in the states, but advised that demand for performance information in Congress may take longer because of the complexity of its processes and committee structures. Most also agreed that the federal government will need to transcend agency borders and take a more crosscutting view of performance to address fiscal challenges, but did not reach consensus on a model or method for doing so.