Revisiting reliability of estimates at completion for department of defense contracts

Abstract

Purpose Within earned value management, the cost performance index (CPI) and the critical ratio (CR) are used to generate the estimates at completion (EACs). According to the research in the 1990s, estimating the final contract’s cost at completion (CAC) using EACCR is a quicker predictor of the actual final cost versus using EACCPI. This paper aims to investigate whether this trend stills holds for modern department of defense contracts. Design/methodology/approach Accessing the Cost Assessment Data Enterprise (CADE) database, 451 contracts consisting of 863 contract line item numbers (CLINs) were initially retrieved and analyzed in three stages. The first replicated the work conducted in 1990s. The second stage entailed calculating 95 per cent confidence intervals and hypothesis tests regarding percentage accuracy of EACs for a contract’s final CAC. Lastly, regression analysis was conducted to characterize major, moderate and minor influencers on EAC reliability. Findings For modern contracts, EACCR aligns more with EACCPI and no longer demonstrates early accuracy of a contract’s final CAC. Contract percentage completion strongly reduced the per cent error of estimating CAC, while cost-plus-fixed-fee contracts and those with no work breakdown structure greater than Level 2 negatively affected accuracy. Social implications To militate against optimism of early assessment of a contract's true cost. Originality/value This paper provides empirical evidence that EACCR behaves more like EACCPI with respect to modern contracts, suggesting that today’s contracts have relatively high SPI. Therefore, caution is warranted for program managers when estimating the CAC from contract initiation up to and slightly beyond the mid-point of completion.