2. Does the budget use evidence as a tool, not a weapon?
For evidence to be regularly used by government program managers, they need to trust that isolated negative results of their programs could be used as a tool for program improvement, not just as an excuse for reducing or eliminating program funding. If evidence is used primarily to justify program funding cuts, it is in effect being weaponized, which can lead to ineffective uses of evidence to improve programs and also cause program managers and political champions to become wary of generating evidence at all.
While the budget makes conceptual references to using evidence for program improvements, in practice there are numerous instances where evaluation findings were suggested as the basis to propose terminations of selected programs. The Major Savings Volume of the Trump budget lists programs for termination based on the conclusions of one or more studies, or where limited or incomplete evidence was perceived.
TABLE 1: Select Programs Listed for Termination in the President’s Budget, Citing Evidence
|Program Title||Reference to Evidence in the President’s Budget|
|Education’s Comprehensive Literacy Development Grants||One study found “…no discernible effects on reading achievement.”|
|Education’s Gaining Early Awareness and Readiness for Undergraduate Programs (GEAR UP)||An existing evaluation did not “report on high school graduation or college enrollment outcomes.”|
|Education’s Supporting Effective Instruction (SEI) State Grants||“…there is limited evidence that…[the program] has led to increases in student achievement.”|
|Education’s Technical Assistance Programs||“…little evidence that the Comprehensive Center program increases student achievement in partner districts.”|
|Health and Human Service’s workforce programs||The “programs that lack evidence of significantly improving the Nation's health workforce.”|
|Health and Human Service’s Low Income Home Energy Assistance Program (LIHEAP)||The program is “unable to demonstrate strong performance outcomes.”|
|Labor’s training grants for the Occupational Safety and Health Administration||The agency has “no evidence that the program is effective.”|
Other programs seem to include support for generating evidence, and assume that the results of future studies and research will be used for subsequent budget cuts. For example, the Social Security Disability Insurance program includes a proposal with an up-front investment of $500 million over the next 5 years to be followed immediately by $49 billion in program cuts, presumably determined based on the conclusions of those studies. Such a heavy-handed tactic does not reflect evidence-based decision-making. Instead, the approach assumes that findings of future evaluations will provide conclusive evidence that justify large program cuts. In this instance, the budget prospectively suggests decisions based on findings that do not yet exist.
Other areas do see prioritization of funding toward areas described as increasingly evidence based, including some post-secondary education programs, medication assistance options in Medicaid, and community treatment strategies for mental health, among others. In multiple places the Trump budget documents describe evidence-based decisions were made, though they do not describe the evidence or the decision.
On the one hand, it is positive to see that the administration has identified and described potential uses for evidence. On the other hand, the programs listed above represent a small list of proposed eliminations or program reductions on the back of assertions about evidence.
On the one hand, it is positive to see that the administration has identified and described potential uses for evidence, including identifying that in cases where evidence does not exist programs may not be operating as effectively as possible. On the other hand, the programs listed above represent a small list of proposed eliminations or program reductions on the back of assertions about evidence, and all also happen to be in federal agencies recognized as leaders in supporting approaches to generate evidence about policies. Such an approach must make program managers across the country wonder why they should even bother producing evidence when it can selectively be used to cut program funding.