A regression approach for estimating procurement cost
| Date | 01 March 2005 |
| Pages | 187-209 |
| DOI | https://doi.org/10.1108/JOPP-05-02-2005-B003 |
| Published date | 01 March 2005 |
| Author | Gary W. Moore,Edward D. White III |
| Subject Matter | Public policy & environmental management,Politics,Public adminstration & management,Government,Economics,Public Finance/economics,Texation/public revenue |
JOURNAL OF PUBLIC PROCUREMENT, VOLUME 5, ISSUE 2, 187-209 2005
A REGRESSION APPROACH FOR ESTIMATING
PROCUREMENT COST
Gary W. Moore and Edward D. White III*
ABSTRACT. Cost growth in Department of Defense (DoD) weapon systems
continues to be a scrutinized area of concern. One way to minimize unexpected
cost growth is to derive better and more realistic cost estimates. In this vein,
cost estimators have many analytical tools to ply. Previous research has
demonstrated the use of a two-step logistic and multiple regression methodology
to aid in this endeavor. We investigate and expand this methodology to cost
growth in procurement dollar accounts for the Engineering and Manufacturing
Development phase of DoD acquisition. We develop and present two salient
statistical models for cost estimators to at least consider if not use in mitigating
cost growth for existing and future government acquisition programs.
INTRODUCTION
An ongoing problem for over three decades, cost growth during the
acquisition of major weapon systems concerns not only those who work
in the acquisition environment, but also the members of Congress and the
general public. According to reports by the General Accounting Office
(GAO), RAND, and the Institute for Defense Analysis, the average cost
growth in major defense acquisition programs ranges anywhere from 20
– 50 % (Calcutt, 1993, p. i). This fiscal escalation in major acquisition
programs adversely impacts the Defense Department, the defense
industry, and the nation.
--------------------
* Gary W. Moore, Captain, US Air Force, is a cost analyst with the Air Force
Cost Analysis Agency in Washington, DC. Edward D. White III, Ph.D., is an
Associate Professor of Statistics, Department of Mathematics and Statistics, Air
Force Institute of Technology, Wright-Patterson AFB, Ohio. His teaching and
research interests are in design of experiments, linear and nonlinear regression,
and statistical consulting.
Copyright © 2005 by PrAcademics Press
188 MOORE & WHITE III
The Department of Defense (DoD) coined the phrase “realistic
costing” for the current reform being undertaken in the defense
acquisition community. “Under the new costing approach, the Pentagon
will adopt program estimates developed by the Cost Analysis
Improvement Group (CAIG1) in conjunction with a service estimate
(Grossman, 2002, p. 2).” Realistic costing utilizes the CAIG’s cost
estimating expertise to provide higher quality estimates. DoD’s
dedication to realistic costing contributed significantly to the cancellation
of the Navy Area missile defense program, sending a strong message to
the acquisition community. If managers overrun their budget and breach
the Nunn-McCurdy law, their program will be terminated (Grossman,
2002, p. 2).
For managers to understand and to contain cost growth, they must
identify and control the root causes of cost growth. Program managers
often resort to a process known as “buffering” in order to increase the
accuracy of the baseline estimate and to limit the program’s likelihood of
incurring cost overruns. Buffering of an estimate entails assigning a cost
estimate (dollar value) to each of the cost risks, e.g. additional
engineering effort because of a new weapon system, avionics package, or
stealth technology. The plan is to have sufficient funds available if/when
the risk comes to fruition so that the program does not have to request
additional funding. In the past, costs have been assigned to these risks;
however, they have been shown to be sometimes underestimated.
According to McCrillis (2003), who presented the conclusions of a ten-
year study by OSD CAIG (Office of the Secretary of Defense Cost
Analysis Improvement Group), procurement cost growth has occurred
primarily because of optimistic learning curves.
An example of a current acquisition program struggling to keep costs
under control is that of the F/A-22. The production quantity of the Air
Force’s newest air-to-air fighter has fluctuated considerably over the past
decade in the attempt to maintain some modicum of control over
increased cost growth. Because of rising program costs, the F/A-22
program has reduced the number of initial desired aircraft from 750 to
658 in 1991 to a more recently lowered number of 276 in 2002. In 1997
and 2001, the DoD conducted reviews of the F/A-22 Raptor program.
During these reviews the Air Force attributed estimated production cost
growth to increased labor, airframe, and engine costs. These factors
totaled almost 70 percent of the overall cost growth (GAO, 2003).
To continue reading
Request your trial