Distribution Comparisons of EAC Cost Growth
WBS. Some initial EACs report as late as 60% complete, and some final EACs report as early as 40% complete. To minimize the chance of fitting a probability distribution to unrepresentative data, we established initial and ending thresholds for each reported WBS element. For this, we turn to published literature for these thresholds. Christensen and Heise (1992) report that cumulative Cost Performance Index stability has been found for DoD contracts at or after 20% of a program’s completion. Even though recent work by Kim et al. (2019) suggests this percentage might be later, we accept this 20% as the maximum initial threshold percentage for a reported WBS element. For the ending threshold, we turn to Tracy and White (2011). Their analysis suggests that EACs at 92.5% or more complete are not statistically different from the final program cost. Consequently, we require the final EACs to report to at least this threshold of 92.5% completion. The cost growth for each WBS element of a program is intended to capture the ratio of the final cost of the WBS element to its initial projected cost at the beginning or prior to starting the program. This can be considered budgeted growth. To capture this information, EAC data are used. We investigate all the DD 1921s of the 59 DoD ACAT I RDT&E programs from Project Ginger II. For each WBS element, we only include EACs reported as 20% complete or earlier and those final EACs reported at 92.5% complete or later. Cost growth is calculated as the latest EAC divided by the earliest EAC. Equation 1 is used to calculate the cost growth for each WBS element of a program. Cost Growth = EAC Final EAC Initial (1) The ratio of these two becomes unitless. It is important to note that the EACs in Equation 1 were not normalized and instead were calculated with Then-Year dollars. This major limitation to our calculation process is due to an assumption that the Period of Performance (PoP) from initial EAC to final EAC remains constant. If the PoP changes, then the cost growth factors may be incorrect. Ideally, we would have PoPs for all programs in our study. We did not. In fact, a majority of the DD 1921s in our study had no information in this box. Hence, the calculation process assumption was necessary. From a conceptual perspective, a change of PoP likely reflects schedule slippage within a program. Here, we do assume that the programs do not slip.
36
Defense ARJ , Spring 2025, Vol. 32 No. 1
Made with FlippingBook - Online Brochure Maker