Trends in Nunn-McCurdy Cost Breaches for Major Defense Acquisition Programs
GAO-11-295R: Published: Mar 9, 2011. Publicly Released: Mar 29, 2011.
For nearly 30 years, the statutory provision, known as Nunn-McCurdy, has been an oversight tool for Congress to hold the Department of Defense (DOD) accountable for cost growth on major defense programs. A Nunn-McCurdy breach occurs when a program's unit cost exceeds certain thresholds. When that happens, DOD must notify Congress of the breach. There are a number of statutory provisions that help implement cost growth reporting under Nunn-McCurdy. For the purposes of this report, we refer to these statutory provisions as the Nunn-McCurdy process. In September 2010, Congress requested that we examine trends in Nunn-McCurdy breaches and factors that may be responsible for these trends. In this report, we also discuss changes DOD is making or proposing to make to the Nunn-McCurdy process. To identify trends in Nunn-McCurdy breaches, we collected and analyzed existing data on breaches from DOD's Defense Acquisition Management Information Retrieval system, which contains data on breaches since 1997. DOD officials also provided us with a list of programs that breached the cost growth thresholds since 1997, which we analyzed to remove duplicate entries. In addition, we reviewed analyses by the Office of the Director of Cost Assessment and Program Evaluation to verify our data.
Since 1997, there have been 74 Nunn-McCurdy breaches involving 47 major defense acquisition programs. There were a larger number of breaches in 2001, 2005, 2006, and 2009, which coincides with changes in statute or presidential administration. The statutory changes added a program's original baseline estimate as a new benchmark against which to measure cost growth. During the last two changes in presidential administration, DOD did not submit annual comprehensive Selected Acquisition Reports (SAR), which, along with other factors, may have affected when breaches were reported. The Air Force had a higher proportion of total breaches compared to its proportion of total programs, whereas the Navy had a smaller proportion of breaches compared to its proportion of programs. Aircraft, satellite, and helicopter programs have experienced the largest number of breaches. Thirty-four different prime contractors were listed in the SARs for the programs that breached. Of the 47 programs that breached, 18 programs breached more than one time. Nunn-McCurdy breaches are often the result of multiple, interrelated factors. Our analysis of DOD data and SARs showed that the primary reasons for the unit cost growth that led to Nunn-McCurdy breaches were engineering and design issues, schedule issues, and quantity changes. Cost increases resulting from engineering and design issues may indicate that those programs started without adequate knowledge about their requirements and the resources needed to fulfill them. Many programs also cited revised cost estimates as a factor behind breaches, suggesting estimates were based on inaccurate assumptions. Our previous work shows that without the ability to generate reliable cost estimates, programs are at risk of experiencing cost overruns, missed deadlines, and performance shortfalls. DOD has instituted a process to provide earlier warning of potential breaches and plans to propose changes to try to limit the effect of breaches caused by quantity changes. Specifically, the Joint Staff has implemented a Nunn-McCurdy trip wire process to evaluate the factors that are contributing to cost growth so that programs can take mitigating actions. Our analysis shows nearly 40 percent of Nunn-McCurdy breaches occurred after a production decision had been made--when a program has fewer options for restructuring. DOD also plans to propose a legislative amendment to reduce several statutory requirements added in 2009 for Nunn-McCurdy breaches when it determines the breach was caused primarily by quantity changes that were unrelated to poor performance. Tracking changes in research and development costs, which are not sensitive to quantity changes, would be one way DOD could evaluate program performance in this context.