Improving Comptroller Benchmarks on Program Spending

Abstract

A common tool for overseeing program execution is to compare spending against linear benchmarks to identify programs that may be falling behind or unable to fully use their funding. These benchmarks identify candidates for further investigation and potential budgets reduced and reallocated. Pressures to meet benchmarks can drive bad behaviors, such as premature spending before good prices and intellectual property rights can be negotiated. This paper analyses business theory, program manager observations, and historical trends of DoD obligation and expenditure rates to assess ways to improve these benchmarks. Regressions of historical obligation data find that recent spending has an underlying linear trend, but temporal variables, theory, and execution realities indicate that S-shaped curves are better benchmarks. Also, benchmarks should be adjusted when Congress provides Continuing Resolutions (CRs) in lieu of full appropriations at the start of the fiscal year. Also, as expected by theory, historical expenditure patterns for Research, Development, Test, and Evaluation (RDT&E), Procurement, and Operation and Maintenance (O&M) funds follow S-shaped curves rather than the linear profiles in DoD benchmarks. Recommendations are provided, including adjustments for variable effects on obligations, S-curve profiles for improved benchmarks, and leveraging improved DoD data environments to switch to plan-based benchmarking.