Federal agencies should use a three-step performance-management cycle to determine whether programs are solving the problems they were created to address, the Government Accountability Office said in a July overview.
The cycle begins with clear, measurable goals, continues with collecting relevant performance data and ends with managers using those results to improve decisions. GAO says the work must be repeated rather than treated as a one-time compliance exercise.
The stakes reach across programs that collectively spend trillions of dollars each year on health care, public safety, disaster support and other services. Without goals and outcome data, Congress and agency leaders cannot reliably judge whether those dollars produce intended results.
GAO distinguishes inputs, such as money available, and outputs, such as the number of people served, from outcomes that describe whether people's lives or public conditions improved in the intended way.
Programs also frequently overlap across agencies. When related programs use different goals or incomplete data, policymakers have less ability to compare their performance, streamline duplicative work or direct resources toward stronger results.
The report points to 15 programs serving pregnant women, young children and their families. Twelve had established performance-management processes; GAO recommended that one program each at Agriculture, Health and Human Services and Veterans Affairs fully develop the missing pieces.
Performance measures can serve as an early-warning system, but GAO says additional evidence is needed to determine effectiveness. Process evaluations test implementation, outcome evaluations examine alignment with desired results and impact evaluations compare results with what would have happened without the program.
In a 2020 survey, about one-third of federal managers reported access to robust program evaluations. That historical survey does not establish the current share, but it illustrates the evidence gap GAO says agencies must close.
The framework is not a finding that every federal program fails. It is a test for whether leaders have enough evidence to identify problems, improve delivery and explain the return on public spending.