A Macro-Stochastic Approach to Improved Cost Estimation for Defense Acquisition Programs

Abstract

Inaccurate cost estimates are a recurrent problem for Department of Defense (DoD) acquisition programs, with cost overruns exceeding billions of dollars each year. These estimate errors hinder the ability of the DoD to assess the affordability of future programs and properly allocate resources to existing programs. In this research, the author employs a novel approach called "macro-stochastic" cost estimation for significantly reducing cost estimate errors in Major Defense Acquisition Programs (MDAPs). To achieve this reduction, the author first extracts and catalogs key programmatic data from 936 Selected Acquisition Reports. The author then analyzes historical trends in the data using mixed-model regression with high-level descriptive program parameters. Based on these trends, the model is found to reduce estimate errors by 18.7 percent on average, when applied to a randomly selected, historical cost estimate. However, the model is most beneficial when applied early in program life; when applied to the first cost estimate of each program in the database, the macro-stochastic technique reduces cost estimate error by over one-third. This statistically and economically significant reduction could potentially allow for reallocation of $6.25 billion, annually, if applied consistently to the DoD's portfolio of MDAPs.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Mar 27, 2014
Accession Number
ADA610513

Entities

People

  • Allen J. Deneve

Organizations

  • Air Force Institute of Technology

Tags

Communities of Interest

  • Energy and Power Technologies
  • Ground and Sea Platforms
  • Weapons Technologies

DTIC Thesaurus Topics

  • Acquisition
  • Air Force
  • Cost Analysis
  • Cost Estimates
  • Cost Overruns
  • Cost Reductions
  • Databases
  • Department Of Defense
  • Governments
  • Information Retrieval
  • Information Science
  • Military Acquisition
  • Military Procurement
  • Predictive Modeling
  • Procurement
  • Test And Evaluation
  • United States Government

Readers

  • Defense Acquisition Program Management
  • Public Financial Management and Budgeting
  • Regression Analysis.