Early Warning Model for Acquisition Program Cost and Schedule Growth
Earned Value Management/Schedule Track
This paper uses the non-linear Rayleigh function to model the behavior underlying the problem-solving and engineering activities contained in research and development contracts and reflected in the cumulative cost accrual during the execution of those contracts. This paper examines 115 contract datasets. Using non-linear least squares with monotonically transformed and restricted parameters and the techniques of numerical computation, we examine the ability of the Rayleigh function to explain the variation of the actual cost data over time in each of these contracts. We then document the results and show that on average the Rayleigh model explains over 93% of the variation in actual cost data. Further, in each of the datasets we estimate model parameters and from these derive usable independent estimates of the final cost at completion of the contract and the final contract duration. Then we employ the Rayleigh model progressively to predict final contract cost and duration by doing regression analyses of over 2500 information sets in these 115 contract datasets. These predictions are then compared to standard cost and schedule estimation techniques. We have validated that our implementation of the Rayleigh model outperforms all other common estimation techniques and provides very usable results very early in the life of the program. Armed with these results, we further develop a technique to use the Rayleigh model to assess plan realism before actual cost data have even been collected. We also compute confidence regions, which generate measures of overall contract cost and schedule risk at each prediction cycle. Finally we develop computational tools, which link economic and business insights to the trends of Rayleigh-generated estimates through the prediction cycles of a contract. This work has important economic implications for acquisition decision-making, contract execution, program management, and analyses of alternative offers and contract plans when programs are begun.
Dr. Dan Davis
Dan Davis is a research analyst and economist working with Gary Christle and Wayne Abba at the Center for Naval Analyses (CNA) on cost and acquisition issues and management studies. Dan has worked on studies of acquisition issues for the U.S. Air Force, the U.S. Navy, the U.S. Army, and the Department of Health and Human Services. A recent study for the Navy included the results of extensive research on ways of improving the analysis of existing earned value data to catch problems with projects early and to assist senior decision-makers in their oversight and management of major programs. Dan has also presented other work on the economics of subcontracting and procurement at the Western Economics Association International conference in Seattle in July 2007.
Prior to joining CNA, Dan retired after twenty-two years of service in the U.S. Army. Upon retirement, he earned his Ph.D. in economics from the University of North Carolina at Chapel Hill. His dissertation used game theory, auction theory, mechanism design, and signaling theory to explore the bidding behavior of contractors and their use of subcontractors in government procurement. His work also explored the optimal mechanism design and the optimal reservation price for the government, when it acted as a surplus maximizer and when it acted as a social welfare maximizer. Dan is a graduate of the U.S. Military Academy at West Point, he is a Rhodes Scholar, and he has a master’s degree from the University of Oxford at Oxford, England.