Understanding the burden of government oversight on engineering work: Adding empirical data to the debate

This paper presents results from a 6-month time-allocation study of the impact of oversight-related activities on engineering work at a major aerospace contractor. Previous studies have reported a wide range of estimated burdens - from 2% of a total system's cost to factors of 5 times the cost of commercially available alternative products. The wide range of estimates of the burden of oversight can be attributed to measurement challenges and to the phenomenon being measured. Our new data provides an empirically valid estimate of the time spent on these activities, allowing us to reconcile differences between previous measures of oversight. We observe that when the definition of oversight is limited to non-value added external monitoring, the extent of burden is on the order of 6% of total time spent on work performed. However, when the definition of oversight includes both externally driven burden and the government-support infrastructure internal to the contractor, the burden ranges from 1.2 to 1.6 times. In addition, we use this data to test widely held beliefs about the impact of oversight on daily contractor work. Specifically, we found that the particular customers who are generally perceived to drive oversight-related burden actually have a small impact on resultant work time; but they drive more non-value-added requests than others. Additionally, while communications and administrative tasks are perceived as the main content of oversight-driven work, most of the time spent on these tasks was not driven by oversight. Implications of these findings for how the acquisition process can be improved are discussed.