Abstract
The typical United States (US) weapons system has grown in cost, capability, and the time required for development and production.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Defense Science Board (1978). The consequence is that for many large programs, when future year defence acquisition expenditures are charted against current year expenditures, the pattern begins to resemble the “bow wave” ahead of a ship.
- 2.
For Peck and Scherer’s analysis, “Implicit in the notion of cost is the aspect of quantity. The cost per unit of a weapon system determines the quantity of systems which can be obtained with any specified amount of resources” (Peck and Scherer 1962; Note 1, 251). This is in line with the prevailing industry and military practice of normalizing cost for any quantity changes in comparisons (e.g., quantity-adjusted cost). This has the tendency of casting quantity change as a predictor of cost change, but not the reverse, in which quantity change is understood as the dependent variable. See Scherer (1964).
- 3.
Drezner et al. (1993) is one fairly common example of research practices in reviewing the key factors affecting cost growth, including budget trends, system performance, management complexity, and schedule-related issues. Their reliance on univariate and bivariate views of selected acquisition report (SARs) data unfortunately allows important explanatory factors to be confounded.
- 4.
See Tyson et al. (1994) “Examining the relationship between cost and schedule growth in development, we concluded that it was not appropriate to consider the major independent variable [Development Schedule Growth] to be exogenous…Therefore we estimated the tactical missile development relationship as a simultaneous system of equations.”
- 5.
RAND released their version of SARs data (Jarvaise et al. 1996); the electronic version is available on the RAND website (www.rand.org) as an electronic data appendix to the Jarvaise et al. (1996) monograph.
- 6.
For example, schedule data—perhaps a single missing planned date for MS B—impacts several variables calculated using that data as an end point for one interval, the start point for the next interval, and various ratios of program phase length and schedule slip ratios that could be calculated from that variable. MDAPs that began reporting by the 1969 SAR had only fragmentary or inconsistent data covering their early years before the SARs began. See Hough (1992) and Jarvaise et al. (1996). For the present analysis, an MDAP sample of 244 programs is reduced by 50 % or more in multivariate estimation procedures as a result of the missing endpoints on schedule data.
- 7.
The discussion is simplified to focus on the essential concepts required rather than burden the reader with DoD program terminology and acronyms. The reader is referred for additional technical detail of acquisition cycles, program structure and weapon program reporting metrics to the studies by Hough (1992) and Jarvaise et al. (1996).
- 8.
Because there have been many different purposes in calculating cost growth, it is possible to find cost analyses that focus on a specific time period, i.e., Development cost growth, or Production cost growth. This analysis calculates cost growth from the Design Estimate to the last available Current Estimate, or to project completion, whichever happens first, and is a more comprehensive measure of overall program cost growth.
- 9.
All SAR cost estimates throughout the life of a defence contract must be adjusted (or normalized) for effective cost (total and per unit) given the changing number of units to be purchased. How to best achieve this has been a topic of lengthy debate and refinements among defence analysts. The best exposition is in Hough (1992, 30–41), as well as in the earlier report by Dews et al. (1979, Appendix A).
- 10.
A measure of kurtosis = 3 and skewness = 0 would describe a normally distributed variable.
- 11.
Only two variables (B1950SYR and the PEO dummy variable) show even mild collinearity, but the condition index is 10.86. All other variables produce condition indexes under a value of 2.97. The Belsley-Kuh-Welsch guidelines are that a condition index value greater than 15 indicates a possible problem and an index greater than 30, a serious problem with multicollinearity (Belsley et al. 1980). Serial correlation in a traditional sense is not present, although it is possible that there may be a cohort effect for groups of programs going through the same program stage in the same years, but with limited sample available, the sample partitioning would leave few degrees of freedom for hypothesis testing.
- 12.
Other variables defined by McNicol (2006) and McCrillis (2003) also related to time periods in which DoD cost estimation requirements were tightened, or in which DoD budgets were plentiful, but coefficient estimates for these additional variables were statistically insignificant in estimation. In contrast, in the cost growth equation the highly significant PEO variable coefficient has a p value of 0.066, in contrast to p-values of 0.001 and 0.000 in the schedule change and quantity change equations, respectively.
- 13.
Including the variable PEO in each equation is a compromise because the right hand side variables across the set of equations in the SUR model are more similar, which reduces the gain from using SUR methods. Ideally, each equation in a SUR system should have dissimilar explanatory variables, but correlated dependent variables (Greene 2000).
References
Belsley D, Kuh E, Welsch R (1980) Regression diagnostics: identifying influential data and sources of collinearity. John Wiley, New York, NY
Chin W (2004) British weapons acquisition policy and the futility of reform. Ashgate, Burlington, VT
Defense Science Board (1978) Report of the acquisition cycle task force. Office of the Under Secretary of Defense for Research and Engineering, Washington D.C
Dews E, Smith G, Barbour A, Harris E, Hesse M (1979) Acquisition policy effectiveness: department of expense experience in the 1970s. RAND Corporation, Santa Monica, CA, R-2516
Drezner JA, Smith G (1990) An analysis of weapon system acquisition schedules. RAND Corporation, Santa Monica, CA, R-3937
Drezner JA, Jarvaise JM, Hess R, Hough WPG, Norton D (1993) An analysis of weapons system cost growth. RAND Corporation, Santa Monica, CA, MR-291
Greene WH (2000) Econometric analysis, 4th edn. Prentice Hall, Englewood Cliffs
Hough PG (1992) Pitfalls in calculating cost growth from selected acquisition reports. RAND Corporation, Santa Monica, CA, N-3136
Jarvaise JM, Drezner JA, Norton D (1996) The defense system cost performance database. RAND Corporation, Santa Monica, CA, MR-625
Kmenta J (1986) Elements of econometrics, 2nd edn. Macmillan, New York
McCrillis J (2003) Cost growth of major defense programs. Annual Department of Defense Cost Analysis Symposium (ADoDCAS). Williamsburg, VA, 30 Jan
McNicol DL (2006) Cost growth in major weapon procurement programs. Institute for Defense Analysis, Alexandria, VA, IDA P-3832
Peck M, Scherer FM (1962) The weapons acquisition process: an economic analysis. Harvard, Boston
Scherer FM (1964) The weapons acquisition process: economic incentives. Harvard, Boston
Tyson K, Nelson JR, Om N, Palmer PR (1989) Acquiring major systems cost and schedule trends. Institute for Defense Analysis, Alexandria, VA, IDA P-2201
Tyson K, Harmon B, Utech D (1994) Understanding cost and schedule growth in acquisition programs. Institute for Defense Analysis, Alexandria, VA, IDA P-2967
U.S. Congressional Budget Office (1987) Effects of weapons procurements stretch-outs on costs and schedules. Washington, D.C
U.S. General Accounting Office (1992) Defense weapons systems acquisition. Washington, DC. HR 93-7
Zellner A (1962) An efficient method of estimating seemingly unrelated regressions and tests for aggregation bias. J A Stat Assoc 57(298):348–368
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2014 Springer Science+Business Media New York
About this chapter
Cite this chapter
Williamson, R. (2014). A Systems Estimation Approach to Cost, Schedule, and Quantity Outcomes. In: Alleman, J., Ní-Shúilleabháin, Á., Rappoport, P. (eds) Demand for Communications Services – Insights and Perspectives. The Economics of Information, Communication, and Entertainment. Springer, Boston, MA. https://doi.org/10.1007/978-1-4614-7993-2_14
Download citation
DOI: https://doi.org/10.1007/978-1-4614-7993-2_14
Published:
Publisher Name: Springer, Boston, MA
Print ISBN: 978-1-4614-7992-5
Online ISBN: 978-1-4614-7993-2
eBook Packages: Business and EconomicsEconomics and Finance (R0)