ESTIMATING VARIABILITY IN THE IMPLEMENTATION COST GROWTH OF MILITARY BASE REALIGNMENTS AND CLOSURES USING HISTORIC DATA

Abstract

The Department of Defense (DoD) periodically conducts a Base Realignment and Closure (BRAC) round to improve the stationing of its force structure, eliminate excess infrastructure, and attain cost savings. The most recent BRAC round in 2005 far exceeded its estimated cost to implement; in a 2012 report, the Government Accountability Office reported that the 2005 BRAC implementation cost grew from the original estimate by 67%. The DoD requires an improved cost estimate and understanding of inherent uncertainty. Using data from 58 observations of BRAC 2005 recommendations, this thesis examines trends in cost growth. The thesis does not find any statistically significant differences in cost increases among subsets of data analyzed by type of DoD recommending agency, presence of Commission amendments, BRAC action complexity, or size of estimate. Variation in implementation cost growth is mildly narrower for BRAC actions that were amended by the Commission and for actions that were more complex. The analysis detects a bias in estimating large BRAC actions, which indicates a systematic hesitancy or inability to fully estimate the most expensive BRAC actions. The distribution of BRAC 2005 actions’ cost increases is used to inform an improved, three-point estimate for future BRAC rounds. Under conditions comparable to BRAC 2005, this thesis shows that the true mean of future BRAC actions’ cost increases may be expected to be 93% with a 95% confidence interval of [57%, 129%].